VIEWS OF NATURE:
OR THOUGHTS ABOUT
The Amazing Wonders of Creation;
WITH
Scientific Illustrations.

1. p. 1—“The Lake of Tacarigua.”

On advancing through the interior of South America, from the coast of Caracas or of Venezuela towards the Brazilian frontier (from the 10th degree of north latitude to the equator), the traveller first passes a lofty chain of mountains (the littoral chain of Caracas) inclining from west to east; next vast treeless Steppes or plains (Los Llanos), which extend from the foot of the littoral chain to the left bank of the Orinoco; and, lastly, the mountain range which gives rise to the cataracts of Atures and Maypure. This mountain chain, which I have named the Sierra Parime, passes in an easterly direction between the sources of the Rio Branco and Rio Esquibo, in the direction of Dutch and French Guiana. This region, which is the seat of the marvellous myths of the Dorado, and is composed of a mountain mass, divided into numerous gridiron-like ridges, is bounded on the south by the woody plain through which the Rio Negro and the Amazon have formed themselves a channel. Those who would seek further instruction regarding these geographical relations, may compare the large chart of La Cruz Olmedilla (1775), which has served as the basis of nearly all the more modern maps of South America, with that of Columbia, which I drew up in accordance with my own astronomical determinations of place, and published in the year 1825.

The littoral chain of Venezuela is, geographically considered, a portion of the Peruvian Andes. These are divided at the great mountain node of the sources of the Magdalena (lat. 1° 55′ to 2° 20′) into three chains, running to the south of Popayan, the easternmost of which extends into the snowy mountains of Merida. These mountains gradually decline towards the Paramo de las Rosas into the hilly district of Quibor and Tocuyo, which connects the littoral chain of Venezuela with the Cordilleras of Cundinamarca.

This littoral chain extends murally and uninterruptedly from Portocabello to the promontory of Paria. Its mean elevation is scarcely 750 toises, or 4796 English feet; but some few summits, like the Silla de Caracas (also called the Cerro de Avila), 23which is adorned with the purple-flowering Befaria (the red-blossomed American Alpine rose), rise 1350 toises, or 8633 English feet above the level of the sea. The coast of the Terra Firma everywhere bears traces of devastation, giving evidence of the action of the great current which runs from east to west, and which, after the disintegration of the Caribbean Islands, formed the present Sea of the Antilles. The tongues of land of Araya and Chuparipari, and more especially the coasts of Cumana and New Barcelona, present to the geologist a remarkable aspect. The rocky islands of Boracha, Caracas, and Chimanas rise like beacon-towers from the sea, affording evidence of the fearful irruption of the waters against the shattered mountain chain. The Sea of the Antilles may once have been an inland sea, like the Mediterranean, which has suddenly been connected with the ocean. The islands of Cuba, Hayti, and Jamaica still exhibit the remains of the mountains of micaceous schist which formed the northern boundary of this lake. It is a remarkable fact that the highest peaks are situated at the very point where these islands approach one another the closest. It may be conjectured that the principal nucleus of the chain was situated between Cape Tiburon and Morant Point. The height of the copper mountains (montañas de cobre) near Saint Iago de Cuba has not yet been measured, but this range is probably higher than the Blue Mountains of Jamaica (1138 toises, or 7277 English feet), whose elevation somewhat exceeds that of the Pass of St. Gothard. I have already expressed my conjectures more fully regarding the valley-like form of the Atlantic Ocean, and the ancient connection of the continents, in a treatise written at Cumana, entitled Fragment d’un Tableau géologique de l’Amérique méridionale, which appeared in the Journal de Physique, Messidor, an IX. It is remarkable that Columbus himself makes mention, in his official report, of the connection between the course of the equinoctial current and the form of the coast-line of the Greater Antilles.^[H]

The northern and more cultivated portion of the province of Caracas is a mountainous region. The marginal chain is divided, like that of the Swiss Alps, into many ranges, enclosing longitudinal valleys. The most remarkable among these is 24the charming valley of Aragua, which produces an abundance of indigo, sugar, and cotton, and, what is perhaps the most singular of all, even European wheat. The southern margin of this valley is bounded by the beautiful Lake of Valencia, the ancient Indian name of which was Tacarigua. The contrast presented by its opposite shores gives it a striking resemblance to the Lake of Geneva. The barren mountains of Guigue and Guiripa have indeed less grandeur and solemnity of character than the Savoy Alps; but, on the other hand, the opposite shore, which is covered with bananas, mimosæ, and triplaris, far surpasses in picturesque beauty the vineyards of the Pays de Vaud. The lake is 10 leagues, (of which 20 form a degree of the Equator), i.e., about 30 geographical miles, in length, and is thickly studded with small islands, which continually increase in size, owing to the evaporation being greater than the influx of fresh water. Within the last few years several sandbanks have even become true islands, and have acquired the significant name of Las Aparecidas, or the “Newly Appeared.” On the island of Cura the remarkable species of solanum is cultivated, which has edible fruit, and has been described by Willdenow (in his Hortus Berolinensis, 1816, Tab. xxvii.). The elevation of the Lake of Tacarigua above the level of the sea is almost 1400 French feet (according to my measurement, exactly 230 toises, i.e., 1471 English feet) less than the mean height of the valley of Caracas. This lake has several species of fish peculiar to itself,^[I] and ranks among the most beautiful and attractive natural scenes that I am acquainted with in any part of the earth. When bathing, Bonpland and myself were often terrified by the appearance of the Bava, a species of crocodile-lizard (Dragonne?), hitherto undescribed, from three to four feet in length, of repulsive aspect, but harmless to man. We found in the Lake of Valencia a Typha, perfectly identical with the European bulrush, the Typha angustifolia—a singular and highly important fact in reference to the geography of plants.

In the valleys of Aragua, skirting the lake, both varieties of the sugar-cane are cultivated, viz., the common Caña criolla, and the species newly introduced from the South Sea, the Caña de Otaheiti. The latter variety is of a far lighter and 25more beautiful green, and a field of it may be distinguished from the common sugar-cane at a great distance. Cook and George Forster were the first to describe it; but it would appear, from Forster’s treatise on the edible plants of the South Sea Islands, that they were but little acquainted with the true value of this important product. Bougainville brought it to the Isle of France, whence it passed to Cayenne and (subsequently to the year 1792) to Martinique, Saint Domingo or Haiti, and many of the Lesser Antilles. The enterprising but unfortunate Captain Bligh transported it, together with the bread-fruit tree, to Jamaica. From Trinidad, an island contiguous to the continent, the new sugar-cane of the South Sea passed to the neighbouring coasts of Caracas. Here it has become of greater importance than the bread-fruit tree, which will probably never supersede so valuable and nutritious a plant as the banana. The Tahitian sugar-cane is more succulent than the common species, which is generally supposed to be a native of Eastern Asia. It likewise yields one-third more sugar on the same area than the Caña criolla, which is thinner in its stalk, and more crowded with joints. As, moreover, the West Indian Islands are beginning to suffer great scarcity of fuel (on the island of Cuba the sugar-pans are heated with orange-wood), the new plant acquires additional value from the fact of its yielding a thicker and more ligneous cane (bagaso). If the introduction of this new product had not been nearly simultaneous with the outbreak of the sanguinary Negro war in St. Domingo, the prices of sugar in Europe would have risen even higher than they did, owing to the interruption occasioned to agriculture and trade. The important question which here arises, whether the sugar-cane of Otaheiti, when removed from its indigenous soil, will not gradually degenerate and merge into the common sugar-cane, has been decided in the negative, from the experience hitherto obtained on this subject. In the island of Cuba a caballeria, that is to say, an area of 34,969 square toises (nearly 33 English acres), produces 870 cwt. of sugar, if it be planted with the Tahitian sugar-cane. It is remarkable enough that this important product of the South Sea Islands should be cultivated precisely in that portion of the Spanish colonies which is most remote from the South Sea. The voyage 26from the Peruvian shore to Otaheiti may be made in twenty-five days, and yet, at the period of my travels in Peru and Chili, the Tahitian sugar-cane was not yet known in those provinces. The natives of Easter Island, who suffer great distress from want of fresh water, drink the juice of the sugar-cane, and, what is very remarkable in a physiological point of view, likewise sea-water. On the Society, Friendly, and Sandwich Islands, the light green and thick stemmed sugar-cane is everywhere cultivated.

In addition to the Caña de Otaheiti and the Caña criolla, a reddish African sugar-cane is cultivated in the West Indies, which is known as the Caña de Guinea. It is less succulent than the common Asiatic variety, but its juice is esteemed especially well adapted for the preparation of rum.

In the province of Caracas the light green of the Tahitian sugar-cane forms a beautiful contrast with the dark shade of the cacao plantations. Few tropical trees have so thick a foliage as the Theobroma Cacao. This noble tree thrives best in hot and humid valleys. Extreme fertility of soil and insalubrity of atmosphere are as inseparably connected in South America as in Southern Asia. Nay, it has even been observed that in proportion as the cultivation of the land increases, and the woods are removed, the soil and the climate become less humid, and the cacao plantations thrive less luxuriantly. But while they diminish in numbers in the province of Caracas, they spread rapidly in the eastern provinces of New Barcelona and Cumana, more especially in the humid woody region lying between Cariaco and the Golfo Triste.

2. p. 1—“The natives term this phenomenon ‘banks.’”

The Llanos of Caracas are covered with a widely-extended formation of ancient conglomerate. On passing from the valleys of Aragua over the most southern range of the coast chain of Guigue and Villa de Cura, descending towards Parapara, the traveller meets successively with strata of gneiss and micaceous schist, a probably Silurian transition rock of argillaceous schist and black limestone; serpentine and greenstone in detached spheroidal masses; and lastly, on the margin of the great plain, small elevations of augitic amygdaloid and porphyritic schist. These hills between Parapara and Ortiz appear to me to be produced by volcanic eruptions on the 27old sea-shore of the Llanos. Further to the north, rise the far-famed cavernous and grotesquely-shaped elevations known as the Morros de San Juan, which form a species of devil’s dyke, the grain of which is crystalline, like upheaved dolomite. They are, therefore, to be regarded rather as portions of the shore than as islands in the ancient gulf. I consider the Llanos to have been a gulf, for when their inconsiderable elevation above the present sea level, the adaptation of their form to the rotation current, running from east to west, and the lowness of the eastern shore between the mouth of the Orinoco and the Essequibo are taken into account, it can scarcely be doubted that the sea once overflowed the whole of this basin between the coast chain and the Sierra de la Parime, extending westward to the mountains of Merida and Pamplona (in the same manner as it probably passed through the plains of Lombardy to the Cottian and Pennine Alps). Moreover, the inclination or line of strike of these Llanos is directed from west to east. Their elevation at Calabozo, a distance of 100 geographical (400 English) miles from the sea, scarcely amounts to 30 toises, or 192 English feet; consequently 15 toises (96 English feet) less than the elevation of Pavia, and 45 toises (288 English feet) less than that of Milan in the plain of Lombardy between the Swiss Lepontine Alps and the Ligurian Apennines. This conformation of the land reminds us of Claudian’s expression, “curvata tumore parvo planities.” The surface of the Llanos is so perfectly horizontal that in many parts over an area of some 480 English square miles, not a single point appears elevated one foot above the surrounding level. When it is further borne in mind that there is a total absence of all shrubs, and that in some parts, as in the Mesa de Pavones, there is not even a solitary palm-tree to be seen, it may easily be supposed that this sea-like and dreary plain presents a most singular aspect. Far as the eye can range, it scarcely rests on any object elevated many inches above the general level. If the boundary of the horizon did not continually present an undefined flickering and undulating outline, owing to the condition of the lower strata of air, and the refraction of light, solar elevations might be determined by the sextant above the margin of the plain as above the horizon of the sea. This perfect flatness of the ancient sea-bottom renders the banks even more striking. They are 28composed of broken floetz-strata, which rise abruptly about two or three feet above the surrounding level, and extend uniformly over a length of from 10 to 12 geographical (i.e., 40 to 48 English) miles. It is here that the small rivers of the Steppe take their origin.

On our return from the Rio Negro, we frequently met with traces of landslips in passing over the Llanos of Barcelona. We here found in the place of elevated banks, isolated strata of gypsum lying from 3 to 4 toises, or 19 to 25 English feet, below the contiguous rock. Further westward, near the confluence of the River Caura and the Orinoco, a large tract of thickly grown forest land to the east of the Mission of San Pedro de Alcantara, fell in after an earthquake in the year 1790. A lake was immediately formed in the plain, which measured upwards of 300 toises (1919 feet) in diameter. The lofty trees, as the Desmanthus, Hymenæa, and Malpighia, retained their verdure and foliage for a long time after their submersion.

3. p. 2—“A shoreless ocean seems spread before us.”

The distant aspect of the Steppe is the more striking when the traveller emerges from dense forests, where his eye has been familiarised to a limited prospect and luxuriant natural scenery. I shall ever retain an indelible impression of the effect produced on my mind by the Llanos, when, on our return from the Upper Orinoco, they first broke on our view from a distant mountain, opposite the mouth of the Rio Apure, near the Hato del Capuchino. The last rays of the setting sun illumined the Steppe, which seemed to swell before us like some vast hemisphere, while the rising stars were refracted by the lower stratum of the atmosphere. When the plain has been excessively heated by the vertical rays of the sun, the evolution of the radiating heat, the ascent of currents of air, and the contact of atmospheric strata of unequal density, continue throughout the night.

4. p. 2—“The naked stony crust.”

The deserts of Africa and Asia acquire a peculiar character from the frequent occurrence of immense tracts of land, covered by one flat uniform surface of naked rock. In the Scha-mo, which separates Mongolia and the mountain chain of Ulangom and Malakha-Oola from the north-west 29part of China, such rocky banks are termed Tsy. In the woody plains of the Orinoco they are found to be surrounded with the most luxuriant vegetation.^[J] In the midst of these flat, tabular masses of granite and syenite, several thousands of feet in diameter, presenting merely a few scattered lichens, we find in the forests, or on their margins, little islands of light soil, covered with low and ever-flowering plants, having the appearance of small gardens. The monks settled on the Upper Orinoco, singularly enough regard the whole of these horizontal naked stony plains, when extending over a considerable area, as conducive to fevers and other diseases. Many of the villages belonging to the mission have been transferred to other spots on account of the general prevalence of this opinion. Do these stony flats (laxas) act chemically on the atmosphere or influence it only by means of a greater radiation of heat?

5. p. 2—“Compared with the Llanos and Pampas of South America, or even with the Prairies on the Missouri.”

Our physical and geognostic knowledge of the western mountain region of North America has recently been enriched by the acquisition of many accurate data yielded by the admirable labours of the enterprising traveller Major Long, and his companion Edwin James, but more especially by the comprehensive investigations of Captain Frémont. The knowledge thus established clearly corroborates the accuracy of the different facts which in my work on New Spain I could merely advance as hypothetical conjectures regarding the northern plains and mountains of America. In natural history, as well as in historical research, facts remain isolated until by long-continued investigation they are brought into connection with each other.

The eastern shore of the United States of North America inclines from south-west to north-east, as does the Brazilian coast south of the equator from the Rio de la Plata to Olinda. On both these regions there rise, at a short distance from the coast line, two ranges of mountains more nearly parallel to each other than to the western Andes, (the Cordilleras of Chili and Peru), or to the North Mexican chain of the Rocky Mountains. The South American or Brazilian mountain 30system, forms an isolated group, the highest points of which, Itacolumi and Itambe, do not rise above an elevation of 900 toises, or 5755 English feet. The eastern portion of the ridge most contiguous to the sea is the only part that follows a regular inclination from S.S.W. to N.N.E., increasing in breadth and diminishing in general elevation as it approaches further westward. The chain of the Parecis hills approximates to the rivers Itenes and Guaporé, in the same manner as the mountains of Aguapehi and San Fernando (south of Villabella) approach the lofty Andes of Cochabamba and Santa Cruz de la Sierra.

There is no direct connection between the two mountain systems of the Atlantic and South-sea coasts (the Brazilian and the Peruvian Cordilleras); Western Brazil being separated from Eastern or Upper Peru by the low lands of the province of Chiquitos, which is a longitudinal valley that inclines from north to south, and communicates both with the plains of the Amazon and of the Rio de la Plata. In these regions, as in Poland and Russia, a ridge of land, sometimes imperceptible (termed in Slavonic Uwaly), forms the line of separation between different rivers; as for instance, between the Pilcomayo and Madeira, between the Aguapehi and Guaporé, and between the Paraguay and the Rio Topayos. The ridge (seuil) extends from Chayanta and Pomabamba (19°–20° lat.,) in a south-easterly direction, and after intersecting the depressed tracts of the province of Chiquitos, (which has become almost unknown to geographers since the expulsion of the Jesuits,) forms to the north-east, where some scattered mountains are again to be met with, the divortia aquarum at the sources of the Baures and near Villabella (15°–17° lat.)

This water-line of separation which is so important to the general intercourse and growing civilization of different nations corresponds in the northern hemisphere of South America with a second line of demarcation (2°–3° lat.) which separates the district of the Orinoco from that of the Rio Negro and the Amazon. These elevations or risings in the midst of the plains (terræ tumores, according to Frontinus) may almost be regarded as undeveloped mountain-systems, designed to connect two apparently isolated groups, the Sierra Parime and the Brazilian highlands, to the Andes chain of 31Timana and Cochabamba. These relations, to which very little attention has hitherto been directed, form the basis of my division of South America into three depressions or basins, viz., those of the Orinoco in its lower course, of the Amazon, and of the Rio de la Plata. Of these three basins, the exterior ones, as I have already observed, are Steppes or Prairies; but the central one between the Sierra Parime and the Brazilian chain of mountains must be regarded as a wooded plain or Hylæa.

In endeavouring by a few equally brief touches to give a sketch of the natural features of North America, we must first glance at the chain of the Andes, which, narrow at its origin, soon increases in height and breadth as it follows an inclination from south-east to north-west, passing through Panama, Veragua, Guatimala, and New Spain. This range of mountains, formerly the seat of an ancient civilization, presents a like barrier to the general current of the sea between the tropics, and to a more rapid intercommunication between Europe, Western Africa, and Eastern Asia. From the 17th degree of latitude at the celebrated Isthmus of Tehuantepec, the chain deflects from the shores of the Pacific, and inclining from south to north becomes an inland Cordillera. In Northern Mexico, the Crane Mountains (Sierra de las Grullas) constitute a portion of the Rocky Mountains. On their western declivity rise the Columbia and the Rio Colorado of California; on the eastern side the Rio Roxo of Natchitoches, the Canadian river, the Arkansas, and the shallow river Platte, which latter has recently been converted by some ignorant geographers, into a Rio de la Plata, or a river yielding silver. Between the sources of these rivers rise in the parallels of 37° 20′ and 40° 13′ lat., three huge peaks composed of granite, containing little mica, but a large proportion of hornblende. These have been respectively named Spanish Peak, James or Pike’s Peak, and Big Horn or Long’s Peak.^[K] Their elevation exceeds that of the highest summits of the North Mexican Andes, which indeed nowhere attain the height of the line of perpetual snow from the parallels of 18° and 19° lat., or from the group of Orizaba, (2717 toises, or 17,374 English feet), and of Popocatepetl (2771 toises, or 17,720 English feet) to Santa Fé and Taos in New Mexico. 32James’ Peak (38° 48′ lat.) is said to have an elevation of 11,497 English feet. Of this only 8537 feet have been determined by trigonometrical measurement, the remainder being deduced in the absence of barometrical observations, from uncertain calculations of the declivity or fall of rivers. As it is scarcely ever possible, even at the level of the sea, to conduct a purely trigonometrical measurement, determinations of impracticable heights are always in part barometrical. Measurements of the fall of rivers, of their rapidity and of the length of their course, are so deceptive, that the plain at the foot of the Rocky Mountains, more especially near those summits mentioned in the text, was, before the important expedition of Captain Frémont, estimated sometimes at 8000 and sometimes at 3000 feet above the level of the sea.^[L] From a similar deficiency of barometrical measurements, the true height of the Himalaya remained for a long time uncertain; now, however, science has made such advances in India, that when Captain Gerard had ascended on the Tarhigang, near the Sutledge, north of Shipke, to the height of 19,411 feet, he still had, after having broken three barometers, four equally correct ones remaining.^[M]

Frémont, in the expedition which he made between the years 1842 and 1844, at the command of the United States Government, discovered and measured barometrically the highest peak of the whole chain of the Rocky Mountains to the north-north-west of Spanish, James’, Long’s, and Laramie’s Peaks. This snow-covered summit, which belongs to the group of the Wind River Mountains, bears the name of Frémont’s Peak on the great chart published under the direction of Colonel Abert, chief of the topographical department at Washington. This point is situated in the parallel of 43° 10′ north lat., and 110° 7′ west long., and therefore nearly 5° 30′ north of Spanish Peak. The elevation of Frémont’s Peak, which according to direct measurement is 13,568 feet, must therefore exceed by 2072 feet that given by Long to James’ Peak, which would appear from its position to be identical with Pike’s Peak, as given in the map above referred to. The Wind River Mountains constitute the dividing ridge (divortia aquarum) between the two seas. “From the summit,” 33says Captain Frémont in his official report,^[N] “we saw on the one side numerous lakes and streams, the sources of the Rio Colorado, which carries its waters through the Californian Gulf to the South Sea; on the other, the deep valley of the Wind River, where lie the sources of the Yellowstone River, one of the main branches of the Missouri which unites with the Mississippi at St. Louis. Far to the north-west we could just discover the snowy heads of the Trois Tetons, which give rise to the true sources of the Missouri not far from the primitive stream of the Oregon or Columbia river, which is known under the name of Snake River, or Lewis Fork.”

To the surprise of the adventurous travellers, the summit of Frémont’s Peak was found to be visited by bees. It is probable that these insects, like the butterflies which I found at far higher elevations in the chain of the Andes, and also within the limits of perpetual snow, had been involuntarily drawn thither by ascending currents of air. I have even seen large winged lepidoptera, which had been carried far out to sea by land-winds, drop on the ship deck at a considerable distance from land in the South Sea.

Frémont’s map and geographical researches embrace the immense tract of land extending from the confluence of Kanzas River with the Missouri, to the cataracts of the Columbia and the Missions of Santa Barbara and Pueblo de los Angeles in New California, presenting a space amounting to 28 degrees of longitude (about 1360 miles) between the 34th and 45th parallels of north latitude. Four hundred points have been hypsometrically determined by barometrical measurements, and for the most part, astronomically: so that it has been rendered possible to delineate the profile above the sea’s level of a tract of land measuring 3,600 miles with all its inflections, extending from the north of Kanzas River to Fort Vancouver and to the coasts of the South Sea (almost 720 miles more than the distance from Madrid to Tobolsk). As I believe I was the first who attempted to represent, in geognostic profile, the configuration of entire countries, as the Spanish Peninsula, the highland of Mexico, and the Cordilleras of South America (for the half-perspective 34projections of the Siberian traveller, the Abbé Chappe,^[O] were based on mere and for the most part on very inaccurate estimates of the falls of rivers); it has afforded me special satisfaction to find the graphical method of representing the earth’s configuration in a vertical direction, that is, the elevation of solid over fluid parts, achieved on so vast a scale. In the mean latitudes of 37° to 43° the Rocky Mountains present, besides the great snow-crowned summits, whose height may be compared to that of the Peak of Teneriffe, elevated plateaux of an extent scarcely to be met with in any other part of the world, and whose breadth from east to west is almost twice that of the Mexican highlands. From the range of the mountains, which begin a little westward of Fort Laramie, to the further side of the Wahsatch Mountains, the elevation of the soil is uninterruptedly maintained from five to upwards of seven thousand feet above the sea’s level; nay, this elevated portion occupies the whole space between the true Rocky Mountains and the Californian snowy coast range from 34° to 45° north latitude. This district, which is a kind of broad longitudinal valley, like that of the lake of Titicaca, has been named The Great Basin by Joseph Walker and Captain Frémont, travellers well acquainted with these western regions. It is a terra incognita of at least 8000 geographical (or 128,000 English) square miles, arid, almost uninhabited, and full of salt lakes, the largest of which is 3940 Parisian (or 4200 English) feet above the level of the sea, and is connected with the narrow Lake Utah,^[P] into which the “Rock River” (Timpan Ogo in the Utah language) pours its copious stream. Father Escalante, in his wanderings from Santa Fé del Nuevo Mexico to Monterey in New California, discovered Frémont’s “Great Salt Lake” in 1776, and confounding together the river and the lake, called it Laguna de Timpanogo. Under this name I inserted it in my map of Mexico, which gave rise to much uncritical discussion regarding the assumed non-existence of a large inland salt lake,^[Q]—a 35question previously mooted by the learned American traveller Tanner. Gallatin expressly says in his memoir on the aboriginal races^[R]—“General Ashley and Mr. J. S. Smith have found the Lake Timpanogo in the same latitude and longitude nearly as had been assigned to it in Humboldt’s Atlas of Mexico.”

I have purposely dwelt at length on these considerations regarding the remarkable elevation of the soil in the region of the Rocky Mountains, since by its extension and height it undoubtedly exercises a great, although hitherto unappreciated influence on the climate of the northern half of the new continent, both in its southern and eastern portions. On this vast and uniformly elevated plateau Frémont found the water covered with ice every night in the month of August. Nor is the configuration of the land less important when considered in reference to the social condition and progress of the great North American United States. Although the mountain range which divides the waters attains a height nearly equal to that of the passes of Mount Simplon (6170 Parisian or 6576 English feet), Mount Gothard (6440 Parisian or 6863 English feet), and the great St. Bernard (7476 Parisian or 7957 English feet), the ascent is so prolonged and gradual that no impediments oppose a general intercourse by means of vehicles and carriages of every kind between the Missouri and Oregon territories, between the Atlantic States, and the new settlements on the Oregon (or Columbia) river, or between the coast-lands lying opposite to Europe on the one side of the continent, and to China on the other. The distance from Boston to the old settlement of Astoria on the Pacific at the mouth of the Oregon when measured in a direct line, and taking into account the difference of longitude, is 550 geographical, i.e., 2200 English miles, or one-sixth less than the distance between Lisbon and Katherinenburg in the Ural district. On account of this gentle ascent of the elevated plains leading from the Missouri to California and the Oregon territory (all the resting-places measured between the Fort and River Lamarie on the northern branch of the Platte river to Fort Hall on the Lewis Fork of the Columbia, being situated at an elevation of from five to upwards of seven thousand feet, and that in Old Park even at the height of 9760 Parisian or 10,402 English feet!), considerable 36difficulty has been experienced in determining the culminating point, or that of the divortia aquarum. It is south of the Wind River Mountains, about midway between the Mississippi and the coast line of the Southern Ocean, and is situated at an elevation of 7490 feet, or only 480 feet lower than the pass of the Great Bernard. The emigrants call this culminating point the South Pass.^[S] It is situated in a pleasant region, embellished by a profusion of artemisiæ, especially A. tridentata (Nuttall), and varieties of asters and cactuses, which cover the micaceous slate and gneiss rocks. Astronomical determinations place its latitude in the parallel of 42° 24′, and its longitude in that of 109° 24′ W. Adolf Erman has already drawn attention to the fact, that the line of strike of the great east-Asiatic Aldanian mountain-chain, which separates the basin of the Lena from the rivers flowing towards the Great Southern Ocean, if extended in the form of a great circle on the surface of the globe, passes through many of the summits of the Rocky Mountains between 40° and 55° north lat. “An American and an Asiatic mountain-chain,” he remarks, “appear therefore to be only portions of one and the same fissure erupted by the shortest channels.”^[T]

The western high mountain coast chain of the Californian maritime Alps, the Sierra Nevada de California, is wholly distinct from the Rocky Mountains, which sink towards the Mackenzie River (that remains covered with ice for a great portion of the year), and from the high table land on which rise individual snow-covered peaks. However injudicious the choice of the appellation of Rocky Mountains may be, when applied to the most northerly prolongation of the Mexican central chain, I do not deem it expedient to substitute for it the denomination of the Oregon Chain, as has frequently been attempted. These mountains do indeed give rise to the sources of three main branches constituting the great Oregon or Columbia river (viz., Lewis’, Clarke’s, and North Fork); but this mighty stream also intersects the chain of the ever snow-crowned maritime Alps of California. The name of Oregon Territory is also employed, politically and officially, to designate the lesser territory of land west of the 37coast chain, where Fort Vancouver and the Walahmutti settlements are situated; and it would therefore seem better to abstain from applying the name of Oregon either to the central or to the coast chain. This denomination, moreover, led the celebrated geographer Malte-Brun into a misconception of the most remarkable kind. He read in an old Spanish chart the following passage:—“And it is still unknown (y aun se ignora) where the source of this river” (now called the Columbia) “is situated,” and he believed that the word ignora signified the name of the Oregon.^[U]

The rocks which give rise to the cataracts of the Columbia at the point where the river breaks through the chain, mark the prolongation of the Sierra Nevada of California from the 44th to the 47th degree of latitude.^[V] In this northern prolongation of the chain lie the three colossal elevations of Mount Jefferson, Mount Hood, and Mount St. Helen’s, which rise 14,540 Parisian (or 15,500 English) feet above the sea-level. The height of this coast chain or range far exceeds therefore that of the Rocky Mountains. “During an eight months’ journey along these maritime Alps,” says Captain Frémont,^[W] “we were constantly within sight of snow-covered summits; and while we were able to cross the Rocky Mountains through the South Pass at an elevation of 7027 feet, we found that the passes in the maritime range, which is divided into several parallel chains, were more than 2000 feet higher”—and therefore only 1170 (English) feet below the summit of Mount Etna. It is also a very remarkable fact, and one which reminds us of the relations of the eastern and western Cordilleras of Chili, that volcanoes still active are only found in the Californian chain which lies in the closest proximity to the sea. The conical mountains of Regnier and of St. Helen’s are almost invariably observed to emit smoke; and on the 23rd of November, 1843, the latter of these volcanoes erupted a mass of ashes which covered the shores of the Columbia for a distance of forty miles, like a fall of snow. To the volcanic Californian chain belong also in the far north of Russian America, Mount Elias (according to La Pérouse 1980 toises, or 12,660 feet, and according to Malaspina 2792 toises, or 17,850 feet in height), and Mount Fair 38Weather (Cerro de Buen Tiempo, 2304 toises, or 14,733 feet high). Both these conical mountains are regarded as still active volcanoes. Frémont’s expedition, which has proved alike useful in reference to botany and geognosy, likewise collected volcanic products in the Rocky Mountains (as scoriaceous basalt, trachyte, and true obsidian), and discovered an old extinct crater somewhat to the east of Fort Hall (43° 2′ north lat., and 112° 28′ west long.), but no traces of any still active volcanoes emitting lava and ashes, were to be met with. We must not confound with these the hitherto unexplained phenomenon termed smoking hills, côtes brûlées, and terrains ardens, in the language of the English settlers and the natives who speak French. “Rows of low conical hills,” says the accurate observer M. Nicollet, “are almost periodically, and sometimes for two or three years continually, covered with dense black smoke, unaccompanied by any visible flames. This phenomenon is more particularly noticed in the territory of the Upper Missouri, and still nearer to the eastern declivity of the Rocky Mountains, where there is a river named by the natives Mankizitah-watpa, or the river of smoking earth. Scorified pseudo-volcanic products, a kind of porcelain jasper, are found in the vicinity of the smoking hills.”

Since the expedition of Lewis and Clarke an opinion has generally prevailed that the Missouri deposits a true pumice on its banks; but here white masses of a delicate cellular texture have been mistaken for that substance. Professor Ducatel was of opinion that the phenomenon which is chiefly observed in the chalk formation, was owing to “the decomposition of water by sulphur pyrites and to a reaction on the brown coal floetzes.”^[X]

If before we close these general remarks regarding the configuration of North America we once more cast a glance at those regions which separate the two diverging coast chains from the central chain, we shall find in strong contrast, on the West, between that central chain and the Californian Alps of the Pacific, an arid and uninhabited elevated plateau nearly six thousand feet above the sea; and in the East, between the Rocky Mountains and the Alleghanies. (whose highest points, Mount Washington and Mount Marcy, rise, 39according to Lyell, to the respective heights, of 6652 and 5400 feet,) we see the richly watered, fruitful, and thickly-inhabited basin of the Mississippi, at an elevation of from four to six hundred feet, or more than twice that of the plains of Lombardy. The hypsometrical character of this eastern valley, or in other words, its relation to the sea’s level, has only very recently been explained by the admirable labours of the talented French astronomer Nicollet, unhappily lost to science by a premature death. His great chart of the Upper Mississippi, executed between the years 1836 and 1840, was based on two hundred and forty astronomical determinations of latitude, and one hundred and seventy barometrical determinations of elevation. The plain which encloses the valley of the Mississippi is identical with that of northern Canada, and forms part of one and the same depressed basin, extending from the Gulf of Mexico to the Arctic Sea.^[Y] Wherever the low land falls in undulations, and slight elevations which still retain their un-English appellation of côteaux des prairies, côteaux des bois, occur in connected rows between the parallels of 47° and 48° north lat., these rows and gentle undulations of the ground separate the waters between Hudson’s Bay and the Gulf of Mexico. Such a line of separation between the waters is formed, north of Lake Superior or Kichi Gummi, by the Missabay Heights, and further west by the elevations known as Hauteurs des Terres, in which are situated the true sources of the Mississippi, one of the largest rivers in the world, and which were not discovered till the year 1832. The highest of these chains of hills hardly attains an elevation of from 1500 to 1600 feet. From its mouth (the old French Balize) to St. Louis, somewhat to the south of its confluence with the Missouri, the Mississippi has a fall of only 380 feet, notwithstanding that the itinerary distance between these two points exceeds 1280 miles. The surface of Lake Superior lies at an elevation of 618 feet, and as its depth in the neighbourhood of the island of Magdalena is fully 790 feet, its bottom must be 172 feet below the surface of the ocean.^[Z]

Beltrami, who in 1825 separated himself from Major Long’s 40expedition, boasted that he had found the sources of the Mississippi in Lake Cass. The river passes, in its upper course, through four lakes, the second of which is the one referred to, while the outermost one, Lake Istaca (47° 13′ north lat., and 95° west long.), was first recognised as the true source of the Mississippi, in 1832, in the expedition of Schoolcraft and Lieutenant Allen. This stream, which subsequently becomes so mighty, is only 17 feet in width, and 15 inches deep, when it issues from the singular horse-shoe-shaped Lake Istaca. The local relations of this river were first fully established on a basis of astronomical observations of position by the scientific expedition of Nicollet, in the year 1836. The height of the sources, that is to say, of the last access of water received by Lake Istaca from the ridge of separation, called Hauteur de Terre, is 1680 feet above the level of the sea. Near this point, and at the southern declivity of the same separating ridge, lies Elbow Lake, the source of the small Red River of the north, which empties itself, after many windings, into Hudson’s Bay. The Carpathian Mountains exhibit similar relations in reference to the origin of the rivers which empty themselves into the Baltic and the Black Sea. M. Nicollet gave the names of celebrated astronomers, opponents as well as friends, with whom he had become acquainted in Europe, to the twenty small lakes which combine together to form narrow groups in the southern and western regions of Lake Istaca. His atlas is thus converted into a geographical album, reminding one of the botanical album of the Flora Peruviana of Ruiz and Pavon, in which the names of new families of plants were made to accord with the Court Calendar, and the various alterations made in the Oficiales de la Secretaria.

The east of the Mississippi is still occupied by dense forests; the west by prairies only, on which the buffalo (Bos Americanus) and the musk ox (Bos moschatus) pasture. These two species of animals, the largest of the new world, furnish the nomadic tribes of the Apaches-Llaneros and Apaches-Lipanos with the means of nourishment. The Assiniboins occasionally slay from seven to eight hundred bisons in the course of a few days in the artificial enclosures constructed for the purpose of driving together the wild 41herds, and known as bison parks.^[AA] The American bison, called by the Mexicans Cibolo, is killed chiefly on account of the tongue, which is regarded as a special delicacy. This animal is not a mere variety of the aurochs of the old world; although, like other species of animals, as for instance the elk (Cervus alces) and the reindeer (Cervus tarandus), no less than the stunted inhabitants of the polar regions, it may be regarded as common to the northern portions of all continents, and as affording a proof of their former long existing connection. The Mexicans apply to the European ox the Aztec term quaquahue, or horned animal, from quaquahuitl, a horn. The huge ox-horns which have been found in ancient Mexican buildings near Cuernavaca, south-west of the capital of Mexico, appear to me to belong to the bison. The Canadian bison can be used for agricultural labour, and will breed with the European cattle, although it is uncertain whether the hybrid thus engendered is capable of propagating its species. Albert Gallatin, who, before his appearance in Europe as a distinguished diplomatist, had acquired by personal observation a considerable amount of information regarding the uncultivated parts of the United States, assures us that the fruitfulness of the mixed breed of the American buffalo and European cattle is an undoubted fact: “the mixed breed,” he writes, “was quite common fifty years ago in some of the north-western counties of Virginia, and the cows, the issue of that mixture, propagated like all others.” “I do not remember,” he further adds, “that full-grown buffaloes were tamed; but dogs would at that time occasionally bring in the young bison-calves, which were reared and bred with European cows. At Monongahela all the cattle for a long time were of this mixed breed. It was said, however, that the cows yielded but little milk.” The favourite food of the buffalo is the Tripsacum dactyloides (known as buffalo-grass in North Carolina) and a hitherto undescribed species of clover allied to the Trifolium repens, and designated by Barton as Trifolium bisonicum.

I have elsewhere^[AB] drawn attention to the fact, that according to a passage of the trustworthy Gomara^[AC], there 42lived, as late as the sixteenth century, an Indian tribe in the north-west of Mexico, in 40° north lat., whose greatest wealth consisted in hordes of tamed buffaloes (bueyes con una giba). Yet, notwithstanding the possibility of taming the buffalo, and the abundance of milk it yields, and notwithstanding the herds of Lamas in the Peruvian Cordilleras, no pastoral tribes were met with on the discovery of America. Nor does history afford any evidence of the existence, at any period, of this intermediate stage of national development. It is also a remarkable fact that the North American bison or buffalo has exerted an influence on geographical discoveries in pathless mountain districts. These animals advance in herds of many thousands in search of a milder climate, during winter, in the countries south of the Arkansas river. Their size and cumbrous forms render it difficult for them to cross high mountains on these migratory courses, and a well-trodden buffalo-path is therefore followed wherever it is met with, as it invariably indicates the most convenient passage across the mountains. Thus buffalo-paths have indicated the best tracks for passing over the Cumberland Mountains in the south-western parts of Virginia and Kentucky, and over the Rocky Mountains, between the sources of the Yellowstone and Plate rivers, and between the southern branch of the Columbia and the Californian Rio Colorado. European settlements have gradually driven the buffalo from the eastern portions of the United States. Formerly these migratory animals passed the banks of the Mississippi and the Ohio, advancing far beyond Pittsburgh.^[AD]

From the granitic rocks of Diego Ramirez and the deeply-intersected district of Terra del Fuego (which in the east contains silurian schist, and in the west, the same schist metamorphosed into granite by the action of subterranean fire,)^[AE] to the North Polar Sea, the Cordilleras extend over a distance of more than 8000 miles. Although not the loftiest, they are the longest mountain chain in the world, being upheaved from one fissure, which runs in the direction of a meridian from pole to pole, and exceeding in linear 43extent the distance which, in the old continent, separates the Pillars of Hercules from the Icy Cape of the Tschuktches, in the north-east of Asia. Where the Andes are divided into several parallel chains, those lying nearest the sea are found to be the seat of the most active volcanoes; and it has moreover been repeatedly observed that when the phenomenon of an eruption of subterranean fire ceases in one mountain chain, it breaks forth in some other parallel range. The cones of eruption usually follow the direction of the axis of the chain; but in the Mexican table-land, the active volcanoes are situated on a transverse fissure, running from sea to sea, in a direction from east and west.^[AF] Wherever the upheaval of mountain masses in the folding of the ancient crust of the earth has opened a communication with the fused interior, volcanic activity continued to be exhibited on the murally upheaved mass by means of the ramification of fissures. That which we call a mountain chain has not been raised to its present elevation, or manifested as it now appears, at one definite period; for we find that rocks, varying considerably in age, have been superimposed on one another, and have penetrated towards the surface through early formed channels. The diversity observable in rocks is owing to the outpouring and upheaval of rocks of eruption, as well as to the complicated and slow process of metamorphism going on in fissures filled with vapour, and conducive to the conduction of heat.

The following have for a long time, viz., from 1830 to 1848, been regarded as the highest or culminating points of the Cordilleras of the new continent:—

The Nevado de Sorata, also called Ancohuma or Tusubaya (15° 52′ south lat.), somewhat to the south of the village of Sorata or Esquibel, in the eastern chain of Bolivia: elevation, 25,222 feet.

The Nevado de Illimani, west of the mission of Yrupana (16° 38′ south lat.), also in the eastern chain of Bolivia: elevation, 24,000 feet.

The Chimborazo (1° 27′ south lat.), in the province of Quito: elevation, 21,422 feet.

The Sorata and Illimani were first measured by the distinguished geologist, Pentland, in the years 1827 and 1838; 44and since the publication of his large map of the basin of the Laguna de Titicaca, in June, 1848, we learn that the above elevations given for the Sorata and Illimani are 3960 feet and 2851 feet too high. His map gives only 21,286 feet for the Sorata, and 21,149 feet for the Illimani. A more exact calculation of the trigonometrical operations of 1838 led Mr. Pentland to these new results. He ascribes an elevation of from 21,700 to 22,350 feet to four summits of the western Cordilleras; and, according to his data, the Peak of Sahama would thus be 926 feet higher than the Chimborazo, but 850 feet lower than the Peak of Aconcagua.

6. p. 2—“The desert near the basaltic mountains of Harudsch.”

Near the Egyptian Natron Lakes, which in Strabo’s time had not yet been divided into the six reservoirs by which they are now characterized, there rises abruptly to the north a chain of hills, running from east to west past Fezzan, where it at length appears to form one connected range with the Atlas chain. It divides in north-eastern, as Mount Atlas does in north-western Africa the Lybia, described by Herodotus as inhabited and situated near the sea, from the land of the Berbirs, or Biledulgerid, famed for the abundance of its wild animals. On the borders of Middle Egypt the whole region, south of the 30th degree of latitude, is an ocean of sand, studded here and there with islands or oases abounding in springs and rich in vegetation. Owing to the discoveries of recent travellers, a vast addition has been made to the number of the Oases formerly known, and which the ancients limited to three, compared by Strabo to spots upon a panther’s skin. The third Oasis of the ancients, now called Siwah, was the nomos of Ammon, a hierarchical seat and a resting-place for the caravans, which inclosed within its precincts the temple of the horned Ammon and the spring of the Sun, whose waters were supposed to become cool at certain periods. The ruins of Ummibida (Omm-Beydah) incontestably belong to the fortified caravanserai at the Temple of Ammon, and therefore constitute one of the most ancient monuments which have come down to us from the dawn of human civilization.^[AG]

45The word Oasis is Egyptian, and is synonymous with Auasis and Hyasis.^[AH] Abulfeda calls the Oases el-Wah. In the latter time of the Cæsars, malefactors were sent to the Oases, being banished to these islands in the sandy ocean, as the Spaniards and English transported their malefactors to the Falkland islands and New Holland. The ocean affords almost a better chance of escape than the desert surrounding the Oases; which, moreover, diminish in fruitfulness in proportion to the greater quantity of sand incorporated in the soil.

The small mountain range of Harudsch (Harudje^[AI]) consists of grotesquely-shaped basaltic hills. It is the Mons Ater of Pliny, and its western extremity, known as the Soudah mountain, has been recently explored by my unfortunate friend, the enterprising traveller Ritchie. These basaltic eruptions in the tertiary limestone, and rows of hills rising abruptly from fissures, appear to be analogous to the basaltic eruptions in the Vicentine territory.

Nature repeats the same phenomena in the most distant regions of the earth. Hornemann found an immense quantity of petrified fishes’ heads in the limestone formations of the White Harudsch (Harudje el-Abiad), belonging probably to the old chalk. Ritchie and Lyon remarked that the basalt of the Soudah mountain was in many places intimately mingled with carbonate of lime, as is the case in Monte Berico; a phenomenon that is probably connected with eruptions through limestone strata. Lyon’s chart even indicates dolomite in the neighbourhood. Modern mineralogists have found syenite and greenstone, but not basalt, in Egypt. Is it possible that the true basalt, from which many of the ancient vases found in various parts of the country were made, can have been derived from a mountain lying so far to the west? Can the obsidius lapis have come from there, or are we to seek basalt and obsidian on the coast of the Red Sea? The strip of the volcanic eruptions of Harudsch, on the borders of the African desert, moreover reminds the geologist of augitic vesicular amygdaloid, phonolite, and greenstone porphyry, which are only found on the northern and western limits of the steppes of Venezuela 46and of the plains of the Arkansas, and therefore, as it were, on the ancient coast chains.^[AJ]

7. p. 3—“When suddenly deserted by the tropical east wind, and the sea is covered with weeds.”

It is a remarkable phenomenon, although one generally known to mariners, that in the neighbourhood of the African coast, (between the Canaries and the Cape de Verde islands, and more especially between Cape Bojador and the mouth of the Senegal,) a westerly wind often prevails instead of the usual east or trade wind of the tropics. The cause of this phenomenon is to be ascribed to the far-extending desert of Zahara, and arises from the rarefaction, and consequent vertical ascent of the air over the heated sandy surface. To fill up the vacuum thus occasioned, the cool sea-air rushes in, producing a westerly breeze, adverse to vessels sailing to America; and the mariner, long before he perceives any continent, is made sensible of the effects of its heat-radiating sands. As is well known, a similar cause produces that alternation of sea and land breezes, which prevails at certain hours of the day and night on all sea-coasts.

The accumulation of sea-weed in the neighbourhood of the western coasts of Africa has been often referred to by ancient writers. The local position of this accumulation is a problem which is intimately connected with the conjectures regarding the extent of Phœnician navigation. The Periplus, which has been ascribed to Scylax of Caryanda, and which, according to the investigations of Niebuhr and Letronne, was very probably compiled in the time of Philip of Macedon, contains a description of a kind of fucus sea, Mar de Sargasso, beyond Cerne; but the locality indicated appears to me very different from that assigned to it in the work “De Mirabilibus Auscultationibus,” which for a long time, but incorrectly, bore the great name of Aristotle.^[AK] “Driven by the east wind,” says the pseudo-Aristotle, “Phœnician 47mariners came in a four days’ voyage from Gades to a place where the sea was found covered with rushes and sea-weed (θρύον καὶ φῦκος). The sea-weed is uncovered at ebb, and overflowed at flood tide.” Does he not here refer to a shoal lying between the 34th and 36th degrees of latitude? Has a shoal disappeared there in consequence of volcanic revolution? Vobonne refers to rocks north of Madeira.^[AL] In Scylax it is stated that “the sea beyond Cerne ceases to be navigable in consequence of its great shallowness, its muddiness, and its sea-grass. The sea-grass lies a span thick, and it is pointed at its upper extremity, so that it pricks.” The sea-weed which is found between Cerne (the Phœnician station for merchant vessels, Gaulea; or, according to Gosselin, the small estuary of Fedallah, on the north-west coast of Mauritania,) and Cape Verde, at the present time by no means forms a great meadow or connected group, “mare herbidum,” such as exists on the other side of the Azores. Moreover, in the poetic description of the coast given by Festus Avienus,^[AM] in which, as Avienus himself very distinctly acknowledges, he availed himself of the journals of Phœnician ships, the impediments presented by the sea-weed are described with great minuteness; but Avienus places the site of this obstacle much further north, towards Ierne, the Holy Isle.

When we consider that the sea-weed (fucus), the mud or slime (πηλὸς), the shallowness of the sea, and the perpetual calms, are always regarded by the ancients as characteristic of the Western Ocean beyond the Pillars of Hercules, we feel inclined, especially on account of the reference to the calms, to ascribe this to Punic cunning, to the tendency of a great trading people to hinder others, by terrific descriptions, from competing with them in maritime trading westwards. But even 48in the genuine writings of the Stagyrite,^[AN] the same opinion is retained regarding the absence of wind, and Aristotle attempts to explain a false notion, or, as it seems to me, more correctly speaking, a fabulous mariner’s story, by an hypothesis regarding the depth of the sea. The stormy sea between Gades and the Islands of the Blest (Cadiz and the Canaries) can in truth in no way be compared with the sea, which lies between the tropics, ruffled only by the gentle trade-winds (vents alisés), and which has been very characteristically named by the Spaniards^[AO] El Golfo de las Damas.

From very careful personal researches and from comparison of the logs of many English and French vessels, I am led to believe that the old and very indefinite expression Mar de Sargasso, refers to two fucus banks, the larger of which is of an elongated form, and is the easternmost one, lying between the parallels of 19° and 34°, in a meridian 7° westward of the Island of Corvo, one of the Azores; while the smaller and westernmost bank is of a roundish form, and is found between Bermuda and the Bahama Islands (lat. 25°–31°, long. 66°–74°). The principal diameter of the small bank, which is traversed by ships sailing from Baxo de Plata (Caye d’Argent,) northward of St. Domingo to the Bermudas, appears to me to have a N. 60° E. direction. A transverse band of fucus natans, extending in an east-westerly direction between the latitudes of 25° and 30°, connects the greater with the smaller bank. I have had the pleasure of seeing these views adopted by my lamented friend Major Rennell, and confirmed, in his great work on Currents, by many new observations.^[AP] The two groups of sea-weed, together with the transverse band uniting them, constitute the Sargasso Sea of the older writers, and collectively occupy an area equal to six or seven times that of Germany.

The vegetation of the ocean thus offers the most remarkable example of social plants of a single species. On the main land the Savannahs or grass plains of America, the heaths (ericeta), and the forests of Northern Europe and Asia, 49in which are associated coniferous trees, birches, and willows, produce a less striking uniformity than do these thalassophytes. Our heaths present in the north not only the predominating Calluna vulgaris, but also Erica tetralix, E. ciliaris, and E. cinerea; and in the south, Erica arborea, E. scoparia, and E. Mediterranea. The uniformity of the view presented by the Fucus natans is incomparably greater than that of any other assemblage of social plants. Oviedo calls the fucus banks “meadows,” praderias de yerva. If we consider that Pedro Velasco, a native of the Spanish harbour of Palos, by following the flight of certain birds from Fayal, discovered the Island of Flores as early as 1452, it seems almost impossible, considering the proximity of the great fucus bank of Corvo and Flores, that no part of these oceanic meadows should have been seen before the time of Columbus by Portuguese ships driven westward by storms.

We learn, however, from the astonishment of the companions of the admiral, when they were continuously surrounded by sea-grass from the 16th of September to the 8th of October, 1492, that the magnitude of the phenomenon was at that period unknown to mariners. In the extracts from the ship’s journal given by Las Casas, Columbus certainly does not mention the apprehensions which the accumulation of sea-weed excited, or the grumbling of his companions. He merely speaks of the complaints and murmurs regarding the danger of the very weak but constant east winds. It was only his son, Fernando Colon, who in the history of his father’s life, endeavoured to give a somewhat dramatic delineation of the anxieties of the sailors.

According to my researches, Columbus made his way through the great fucus bank in the year 1492, in latitude 28½°, and in 1493, in latitude 37°, and both times in the longitude of 38°–41°. This can be established with tolerable certainty from the estimation of the velocity recorded by Columbus, and “the distance daily sailed over;” not indeed by dropping the log, but by the information afforded by the running out of half-hour sand-glasses (ampolletas). The first certain and distinct account of the log, (catena della poppa,) which I have found, is in the year 1521, in Pigafetta’s Journal of Magellan’s Circumnavigation of the World.^[AQ] The determination 50of the ship’s place during the days in which Columbus was crossing the great bank is the more important, because it shews us that for three centuries and a half the total accumulation of these socially-living thalassophytes, (whether consequent on the local character of the sea’s bottom or on the direction of the recurrent Gulf stream,) has remained at the same point. Such evidences of the persistence of great natural phenomena doubly arrest the attention of the natural philosopher, when they occur in the ever-moving oceanic element. Although the limits of the fucus banks oscillate considerably, in accordance with the strength and direction of long predominating winds, yet we may still, in the middle of the nineteenth century, take the meridian of 41° west of Paris (or 8° 38′ west of Greenwich) as the principal axis of the great bank. Columbus, with his vivid imaginative force, associated the idea of the position of this bank with the great physical line of demarcation, which according to him, “separated the globe into two parts, and was intimately connected with the changes of magnetic deviation and of climatic relations.” Columbus when he was uncertain regarding the longitude, attempted to determine his place (February, 1493,) by the appearance of the first floating masses of tangled weed (de la primera yerva) on the eastern border of the great Corvo bank. The physical line of demarcation was, by the powerful influence of the Admiral, converted on the 4th of May, 1493, into a political one, in the celebrated line of demarcation between the Spanish and Portuguese rights of possession^[AR].

8. p. 3—“The Nomadic Tribes of Tibbos and Tuaryks.”

These two nations, which inhabit the desert between Bornou, Fezzan, and Lower Egypt, were first made more accurately known to us by the travels of Hornemann and Lyon. The Tibbos or Tibbous occupy the eastern, and the Tuaryks (Tueregs) the western portion of the great sandy ocean. The former, from their habits of constant moving, were named by the other tribes “birds.” The Tuaryks are subdivided into two tribes—the Aghadez and the Tagazi. These are often caravan leaders and merchants. They speak 51the same language as the Berbers, and undoubtedly belong to the primitive Lybian races. They present the remarkable physiological phenomenon that, according to the character of the climate, the different tribes vary in complexion from a white to a yellow, or even almost black hue; but they never have woolly hair or negro features.^[AS]

9. p. 3—“The ship of the desert.”

In the poetry of the East, the camel is designated as the land-ship, or the ship of the desert (Sefynet-el-badyet^[AT]).

The camel is, however, not only the carrier in the desert, and the medium for maintaining communication between different countries, but is also, as Carl Bitter has shown in his admirable treatise on the sphere of distribution of this animal, “the main requirement of a nomadic mode of life in the patriarchal stage of national development, in the torrid regions of our planet, where rain is either wholly or in a great degree absent. No animal’s life is so closely associated by natural bonds with a certain primitive stage of the development of the life of man, as that of the camel among the Bedouin tribes, nor has any other been established in like manner by a continuous historical evidence of several thousand years.”^[AU] “The camel was entirely unknown to the cultivated people of Carthage through all the centuries of their flourishing existence, until the destruction of the city. It was first brought into use for armies by the Marusians, in Western Lybia, in the times of the Cæsars; perhaps in consequence of its employment in commercial undertakings by the Ptolemies, in the valley of the Nile. The Guanches, inhabiting the Canary Islands, who were probably related to the Berber race, were not acquainted with the camel before the fifteenth century, when it was introduced by Norman conquerors and settlers. In the probably very limited communication of the Guanches with the coast of Africa, the smallness of their boats must necessarily have impeded the transport of large animals. The true Berber race, which was diffused throughout the interior of Northern Africa, and to which the Tibbos and Tuaryks, as already observed, belong, 52is probably indebted to the use of the camel throughout the Lybian desert and its oases, not only for the advantages of internal communication, but also for its escape from complete annihilation and for the maintenance of its national existence to the present day. The use of the camel continued, on the other hand, to be unknown to the negro races, and it was only in company with the conquering expeditions and proselyting missions of the Bedouins through the whole of Northern Africa, that the useful animal of the Nedschd, of the Nabatheans, and of all the districts occupied by Aramean races, spread here, as elsewhere, to the westward. The Goths brought camels as early as the fourth century to the Lower Istros (the Danube), and the Ghaznevides transported them in much larger numbers to India as far as the banks of the Ganges.” We must distinguish two epochs in the distribution of the camel throughout the northern part of the African continent; the first under the Ptolemies, which operated through Cyrene on the whole of the north-west of Africa, and the second under the Mahommedan epoch of the conquering Arabs.

It has long been a matter of discussion, whether those domestic animals which were the earliest companions of mankind, as oxen, sheep, dogs, and camels, are still to be met with in a state of original wildness. The Hiongnu, in Eastern Asia, are among the nations who earliest trained wild camels as domestic animals. The compiler of the great Chinese work, Si-yu-wen-kien-lo^[AV], states that in the middle of the eighteenth century, wild camels, as well as wild horses and wild asses, still roamed over Eastern Turkestan. Hadji Chalfa, in his Turkish Geography, written in the seventeenth century, speaks of the very frequent hunting of the wild camel in the high plains of Kashgar, Turfan, and Khotan. Schott finds in the writings of a Chinese author, Ma-dschi, that wild camels exist in the countries north of China and west of the basin of the Hoang-ho, in Ho-si or Tangut. Cuvier^[AW] alone doubts the present existence of wild camels in the interior of Asia. He believes that they have merely “become wild;” since Calmucks, and others professing kindred Buddhist 53doctrines, set camels and other animals at liberty, in order “to acquire to themselves merit for the other world.” The Ailanitic Gulf of the Nabatheans was the home of the wild Arabian camel, according to Greek witnesses of the times of Artemidorus and Agatharchides of Cnidus.^[AX] The discovery of fossil camel-bones of the ancient world in the Sewalik hills (which are projecting spurs of the Himalaya range), by Captain Cautley and Dr. Falconer, in 1834, is especially worthy of notice. These remains were found with antediluvian bones of mastodons, true elephants, giraffes, and a gigantic land tortoise (Colossochelys), twelve feet in length and six feet in height.^[AY] This camel of the ancient world has been named Camelus sivalensis, although it does not show any great difference from the still living Egyptian and Bactrian camels with one and two humps. Forty camels have very recently been introduced into Java, from Teneriffe^[AZ]. The first experiment has been made in Samarang. In like manner, reindeer were only introduced into Iceland from Norway in the course of the last century. They were not found there when the island was first colonised, notwithstanding its proximity to East Greenland, and the existence of floating masses of ice.^[BA]

10. p. 3—“Between the Altai and the Kuen-lün.”

The great highland, or, as it is commonly called, the mountain plateau of Asia, which comprises the lesser Bucharia, Songaria, Thibet, Tangut, and the Mogul country of the Chalcas and Olotes, is situated between the 36th and 48th degrees of north latitude and the meridians of 81° and 118° E. long. It is an erroneous idea to represent this part of the interior of Asia as a single, undivided mountainous swelling, continuous like the plateaux of Quito and Mexico, and situated from seven to upwards of nine thousand feet above the level of the sea. I have already shown in my “Researches respecting the Mountains of Northern India,^[BB]” that there is not in this sense any continuous mountain plateau in the interior of Asia.

54My views concerning the geographical distribution of plants, and the mean degree of temperature requisite for certain kinds of cultivation, had early led me to entertain considerable doubts regarding the continuity of a great Tartarian plateau between the Himalaya and the chain of the Altai. This plateau continued to be characterized, as it had been described by Hippocrates, as “the high and naked plains of Scythia, which, without being crowned with mountains, rise and extend to beneath the constellation of the Bear.”^[BC] Klaproth has the undeniable merit of having been the first to make us acquainted with the true position and prolongation of two great and entirely distinct chains of mountains,—the Kuen-lün and the Thian-schan, in a part of Asia which better deserves to be termed “central,” than Kashmeer, Baltistan, and the Sacred Lakes of Thibet (the Manasa and the Ravanahrada). The importance of the Celestial Mountains (the Thian-schan) had indeed been already surmised by Pallas, without his being conscious of their volcanic character; but this highly-gifted investigator of nature, led astray by the hypotheses of the dogmatic and fantastic geology prevalent in his time, and firmly believing in “chains of mountains radiating from a centre,” saw in the Bogdo Oola (the Mons Augustus, or culminating point of the Thian-schan,) such “a central node, whence all the other Asiatic mountain chains diverge in rays, and which dominates over all the rest of the continent!”

The erroneous idea of a single boundless and elevated plain, occupying the whole of Central Asia, the “Plateau de la Tartarie,” originated in France, in the latter half of the eighteenth century. It was the result of historical combinations, and of a not sufficiently attentive study of the writings of the celebrated Venetian traveller, as well as of the naïve relations of those diplomatic monks who, in the thirteenth and fourteenth centuries (thanks to the unity and extent of the Mogul empire at that time), were able to traverse almost the whole of the interior of the continent, from the ports of Syria and of the Caspian Sea to the east coast of China, washed by the great ocean. If a more exact acquaintance with the language and ancient literature of India were of an older date among us than half a century, the hypothesis of this central plateau, occupying the wide space between the Himalaya and 55the south of Siberia, would no doubt have sought support from some ancient and venerable authority. The poem of the Mahabharata appears, in the geographical fragment Bhischmakanda, to describe “Meru” not so much as a mountain as an enormous swelling of the land, which supplies with water the sources of the Ganges, those of the Bhadrasoma (Irtysch), and those of the forked Oxus. These physico-geographical views were intermingled in Europe with ideas of other kinds, and with mythical reveries on the origin of mankind. The lofty regions from which the waters were supposed to have first retreated (for geologists in general were long averse to the theories of elevation) must also have received the first germs of civilization. Hebraic systems of geology, based on ideas of a deluge, and supported by local traditions, favoured these assumptions. The intimate connexion between time and space, between the beginning of social order and the plastic condition of the surface of the earth, lent a peculiar importance and an almost moral interest to the Plateau of Tartary, which was supposed to be characterized by uninterrupted continuity. Acquisitions of positive knowledge,—the late matured fruit of scientific travels and direct measurements,—with a fundamental study of the languages and literature of Asia, and more especially of China, have gradually demonstrated the inaccuracy and exaggeration of those wild hypotheses. The mountain plains (ὀροπέδια) of Central Asia are no longer regarded as the cradle of human civilization, and the primitive seat of all arts and sciences. The ancient nation of Bailly’s Atlantis, which d’Alembert has happily described as “having taught us everything but its own name and existence,” has vanished. The inhabitants of the Oceanic Atlantis were already treated, in the time of Posidonius, as having a merely apocryphal existence.^[BD]

A plateau of considerable but very unequal elevation runs with little interruption, in a S.S.W.-N.N.E. direction, from Eastern Thibet towards the mountain node of Kentei, south of Lake Baikal, and is known by the names of Gobi, Scha-mo, (sand desert,) Scha-ho. (sand river,) and Han-hai. This swelling of the ground, which is probably more ancient than the elevation of the mountain-chains by which it is intersected, is situated, as we have already remarked, between 81° and 118° 56east longitude from Greenwich. Measured at right angles to its longitudinal axis, its breadth in the south, between Ladak, Gertop, and H’lassa (the seat of the great Lama), is 720 miles; between Hami in the Celestial Mountains, and the great curve of the Hoang-ho, near the In-schan chain, it is scarcely 480; but in the north, between the Khanggai, where the great city of Karakhorum once stood, and the chain of Khin-gan-Petscha, which runs in a meridian line (in the part of Gobi traversed in going from Kiachta to Pekin by way of Urga), it is 760 miles. The whole extent of this elevated ground, which must be carefully distinguished from the more eastern and higher mountain-range, may be approximately estimated, including its deflections, at about three times the area of France. The map of the mountain-ranges and volcanoes of Central Asia, which I constructed in 1839, but did not publish until 1843, shows in the clearest manner the hypsometric relations between the mountain-ranges and the Gobi plateau. It was founded on the critical employment of all the astronomical determinations accessible to me, and on many of the very rich and copious orographic descriptions in which Chinese literature abounds, and which were examined at my request by Klaproth and Stanislaus Julien. My map marks in prominent characters the mean direction and the height of the mountain-chains, together with the chief features of the interior of the continent of Asia from 30 to 60 degrees of latitude, between the meridians of Pekin and Cherson. It differs essentially from any map hitherto published.

The Chinese enjoyed a triple advantage, by means of which they were enabled to enrich their earliest literature with so considerable an amount of orographic knowledge regarding Upper Asia, and more especially those regions situated between the In-schan, the alpine lake of Khuku-noor, and the shores of the Ili and Tarim, lying north and south of the Celestial Mountains, and which were so little known to Western Europe. These three advantages were, besides the peaceful conquests of the Buddhist pilgrims, the warlike expeditions towards the west (as early as the dynasties of Han and Thang, one hundred and twenty-two years before our era, and again in the ninth century, when conquerors advanced as far as Ferghana and the shores of the Caspian Sea); the religious interest attached to certain high mountain summits, 57on account of the periodical performance of sacrifices, in accordance with pre-existing enactments; and lastly, the early and generally known use of the compass for determining the direction of mountains and rivers. This use, and the knowledge of the south-pointing of the magnetic needle, twelve centuries before the Christian era, gave a great superiority to the orographic and hydrographic descriptions of the Chinese over those of Greek and Roman authors, who treated less frequently of subjects of this nature. The acute observer Strabo was alike ignorant of the direction of the Pyrenees and of that of the Alps and Apennines.^[BE]

To the lowlands belong almost the whole of Northern Asia to the north-west of the volcanic Celestial Mountains (Thian-schan); the steppes to the north of the Altai and the Sayanic chain; and the countries which extend from the mountains of Bolor, or Bulyt-tagh (Cloud Mountains in the Uigurian dialect), which run in a north and south direction, and from the upper Oxus, whose sources were discovered in the Pamershian Lake, Sir-i-kol (Lake Victoria), by the Buddhist pilgrims Hiuen-thsang and Song-yun in 518 and 629, by Marco Polo in 1277, and by Lieutenant Wood in 1838, towards the Caspian Sea; and from Lake Tenghiz or Balkasch, through the Kirghis Steppe, towards the Aral and the southern extremity of the Ural Mountains. In the vicinity of mountainous plains, whose elevation varies from 6000 to more than 10,000 feet above the sea’s level, we may assuredly be allowed to apply the term lowlands to districts which are only elevated from 200 to 1200 feet. The first of these heights correspond with that of the city of Mannheim, and the second with that of Geneva and Tübingen. If we extend the application of the word plateau, which has so frequently been misused by modern geographers, to elevations of the soil which scarcely present any sensible difference in the character of the vegetation and climate, physical geography, owing to the indefiniteness of the merely relatively important terms of high and low land, will be unable to distinguish the connexion between elevation above the sea’s level and climate, between the decrease of the temperature and the increase in elevation. When I was in Chinese Dzungarei, 58between the boundaries of Siberia and Lake Saysan (Dsaisang), at an equal distance from the Icy Sea and the mouth of the Ganges, I might assuredly consider myself to be in Central Asia. The barometer, however, soon showed me that the elevation of the plains watered by the Upper Irtysch between Ustkamenogorsk and the Chinese Dzungarian post of Chonimailachu (the sheep-bleating) was scarcely as much as from 850 to 1170 feet. Pansner’s earlier barometric determinations of height, which were first made known after my expedition, have been confirmed by my own observations. Both afford a refutation of the hypotheses of Chappe D’Auteroche (based on calculations of the fall of rivers) regarding the elevated position of the shores of the Irtysch, in Southern Siberia. Even further eastward, the Lake of Baikal is only 1420 feet above the level of the sea.

In order to associate the idea of the relation between lowlands and highlands, and of the successive gradations in the elevation of the soil, with actual data based on accurate measurements, I subjoin a table, in which the heights of the elevated plains of Europe, Africa, and America are given in an ascending scale. With these numbers we may then further compare all that has as yet been made known regarding the mean height of the Asiatic plains, or true lowlands.

No portion of the so-called Desert of Gobi, which consists in part of fine pasture lands, has been so thoroughly investigated in relation to its differences of elevations as the zone which extends over an area of nearly 600 miles, between 59the sources of the Selenga and the Chinese wall. A very accurate barometrical levelling was executed, under the auspices of the Academy of St. Petersburgh, by two distinguished savans—the astronomer George Fuss, and the botanist Bunge. They accompanied a mission of Greek monks to Pekin, in the year 1832, in order to establish there one of those magnetic stations whose construction I had recommended. The mean height of this portion of the Desert of Gobi amounts hardly to 4263 feet, and not to 8000 or 8500 feet, as had been too hastily concluded from the measurements of contiguous mountain summits by the Jesuits Gerbillon and Verbiest. The surface of the Desert of Gobi is not more than 2558 feet above the level of the sea between Erghi, Durma, and Scharaburguna; and scarcely more than 320 feet higher than the plateau of Madrid. Erghi is situated midway, in 45° 31′ north lat., and 111° 26′ east long., in a depression of the land extending in a direction from south-west to north-east over a breadth of more than 240 miles. An ancient Mongolian saga designates this spot as the former site of a large inland sea. Reeds and saline plants, generally of the same species as those found on the low shores of the Caspian Sea, are here met with; while there are in this central part of the desert several small saline lakes, the salt of which is carried to China. According to a singular opinion prevalent among the Mongols, the ocean will at some period return, and again establish its dominion in Gobi. Such geological reveries remind us of the Chinese traditions of the bitter lake, in the interior of Siberia, of which I have elsewhere spoken.^[BF]

The basin of Kashmir, which has been so enthusiastically praised by Bernier, and too moderately estimated by Victor Jacquemont, has also given occasion to great hypsometric exaggerations. Jacquemont found by an accurate barometric measurement that the height of the Wulur Lake, in the valley of Kashmir, near the capital Sirinagur, was 5346 feet. Uncertain determinations by the boiling point of water gave Baron Carl von Hügel 5819 feet, and Lieutenant Cunningham only 5052 feet.^[BG] The mountainous 60districts of Kashmir, which has excited so great an interest in Germany, and whose climatic advantages have lost somewhat of their reputation since Carl von Hügel’s account of the four months of winter snow in the streets of Sirinagur,^[BH] does not lie on the high crests of the Himalaya, as has commonly been supposed, but constitutes a true cauldron-like valley on their southern declivity. On the south-west, where the rampart-like Pir Panjal separates it from the Indian Punjaub, the snow-crowned summits are covered, according to Vigne, by basaltic and amygdaloid formations. The latter are very characteristically termed by the natives schischak deyu, or devil’s pock-marks.^[BI] The charms of the vegetation have also been very differently described, according as travellers passed into Kashmir from the south, and left behind them the luxuriant and varied vegetation of India; or from the northern regions of Turkestan, Samarkand, and Ferghana.

Moreover, it is only very recently that we have obtained a clearer view regarding the elevation of Thibet, the level of the plateau having long been uncritically confounded with the mountain tops rising from it. Thibet occupies the space between the two great chains of the Himalaya and the Kuen-lün, and forms the elevated ground of the valley between them. The land is divided from east to west, both by the inhabitants and by Chinese geographers, into three parts. We distinguish Upper Thibet, with its capital, H’lassa (probably 9592 feet high); Middle Thibet, with the town of Leh or Ladak (9995 feet); and Little Thibet, or Baltistan, called the Thibet of Apricots (Sari-Butan), in which lie Iskardo (6300 feet), Gilgit, and south of Iskardo, but on the left bank of the Indus, the plateau Deotsuh, whose elevation was determined by Vigne (11,977 feet). On carefully examining all the notices we have hitherto possessed regarding the three Thibets, and which will have been abundantly augmented during the present year by the brilliant boundary surveying expedition under the auspices of the Governor-general, Lord Dalhousie, we soon become convinced that the region between the Himalaya and the Kuen-lün is no unbroken table-land, but that it is intersected by mountain groups, which undoubtedly belong to perfectly distinct systems of elevation. 61Actual plains are very few in number: the most considerable are those between Gertop, Daba, Schang-thung (the Shepherd’s Plain), the native country of the shawl-goat, and Schipke (10,449 feet); those round Ladak, which attain an elevation of 13,429 feet, and must not be confounded with the depressed land in which the town lies; and finally, the plateau of the Sacred Lakes, Manasa and Ravanahrada (probably 14,965 feet), which was visited by Father Antonio de Andrada as early as the year 1625. Other parts are entirely filled with compressed mountain masses, “rising,” as a recent traveller observes, “like the waves of a vast ocean.” Along the rivers, the Indus, the Sutledge, and the Yaru-dzangbotschu, which was formerly regarded as identical with the Buramputer (or correctly the Brahmaputra), points have been measured which are only between 6714 and 8952 feet above the sea; and the same is the case with the Thibetian villages Pangi, Kunawur, Kelu, and Murung.^[BJ] From many carefully collected determinations of heights, I think that we are justified in assuming that the plateau of Thibet between 73° and 85° east long, does not attain a mean elevation of 11,510 feet: this is hardly the elevation of the fruitful plain of Caxamarca in Peru, and is 1349 and 2155 feet less than the plateau of Titicaca, and of the street pavement of the Upper Town of Potosi (13,665 feet).

That beyond the Thibetian highlands and the Gobi, whose outline has been already defined, Asia presents considerable depressions, and indeed true lowlands, between the parallels of 37° and 48°, where once an immeasurable continuous plateau was fabulously supposed to exist, is proved by the cultivation of plants which cannot flourish without a certain degree of temperature. An attentive study of the travels of Marco Polo, in which mention is made of the cultivation of the vine, and of the production of cotton in northern latitudes, had long ago directed the attention of the acute Klaproth to this point. In a Chinese work, bearing the title Information respecting the recently conquered Barbarians (Sinkiang-wai-tan-ki-lio), it is stated that “the country of Aksu, somewhat to the south of the Celestial Mountains, near the rivers which form the great Tarim-gol, produces grapes, pomegranates, and numberless other fruits of singular excellence; 62also cotton (Gossypium religiosum), which, covers the fields like yellow clouds. In summer the heat is extremely great, and in winter there is here, as at Turfan, neither intense cold nor heavy snow.” The neighbourhood of Khotan, Kaschgar, and Yarkand still, as in the time of Marco Polo,^[BK] pays its tribute in home-grown cotton. In the oasis of Hami (Khamil), above 200 miles east of Aksu, orange trees, pomegranates, and the finer vines are found to flourish.

The products of cultivation which are here noticed lead to the belief that over extensive districts the elevation of the soil is very slight. At so great a distance from the sea side, and in the easterly situation which so much increases the degree of winter cold, a plateau, as high as Madrid or Munich, might indeed have a very hot summer, but would hardly have, in 43° and 44° latitude, an extremely mild and almost snowless winter. I have seen a high summer heat favour the cultivation of the vine, as at the Caspian Sea, 83 feet below the level of the Black Sea (at Astrakhan, latitude 46° 21′); but the winter cold is there from –4° to –13°. Moreover, the vine is sunk to a greater depth in the ground after the month of November. We can understand that cultivated plants, which, as it were, live only in the summer, as the vine, the cotton plant, rice, and melons, may be cultivated with success between the latitudes of 40° and 44°, on plateaux at an elevation of more than 3000^[BL] feet, and may be favoured by the action of radiant heat; but how could the pomegranate trees of Aksu, and the orange trees of Hami, whose fruit Father Grosier extolled as excellent, endure a long and severe winter (the necessary consequence of a great elevation^[BM])? Carl Zimmerman^[BN] has shown it to be extremely probable that the Tarim depression, or the desert between the mountain chain of Thian-schan and Kuen-lün, where the steppe river Tarim-gol discharges itself into the Lake of Lop, formerly described as an alpine lake, is hardly 1280 feet above the level of the sea, or only twice the elevation of Prague. Sir Alexander Burnes also ascribes to Bokhara only 63an elevation of 1188 feet. It is most earnestly to be desired that all doubt regarding the elevation of the plateaux of Central Asia, south of 45° north latitude, should finally be removed by direct barometrical measurements, or by determinations of the boiling point of water, conducted with greater care than is usual in these cases. All our calculations of the difference between the limits of perpetual snow and the maximum elevation of vine cultivation in different climates, rest at present on too complex and uncertain elements.

In order as briefly as possible to rectify that which has been advanced in the former edition of the present work, regarding the great mountain systems which intersect the interior of Asia, I subjoin the following general review:—We begin with the four parallel chains, which run, with tolerable regularity, from east to west, and are connected together by means of a few detached transverse lines. Differences of direction indicate, as in the Alps of Western Europe, a difference in the epoch of elevation. After the four parallel chains (the Altai, the Thian-schan, the Kuen-lün, and the Himalaya) we must consider as following the direction of meridian, the Ural, the Bolor, the Khingan, and the Chinese chains, which, with the great inflection of the Thibetian and Assam-Birmese Dzangbo-tschu incline from north to south. The Ural divides a depressed portion of Europe from a similarly low portion of Asia. The latter was called by Herodotus,^[BO] and even earlier by Pherecydes of Syros, Scythian or Siberian Europe, and comprised all the countries to the north of the Caspian and of the Iaxartes, which flows from east to west, and may therefore be regarded as a continuation of our Europe, “as it now exists, extending lengthwise across the continent of Asia.”

1. The great mountain system of the Altai (the “gold mountains” of Menander of Byzantium, an historical writer of the seventh century; the Altaï-alin of the Moguls, and the Kin-schan of the Chinese) forms the southern boundary of the great Siberian lowlands, and running between 50° and 52½° north latitude, extends from the rich silver mines of the Snake Mountains, and the confluence of the Uba and the Irtysch, to the meridian of Lake Baikal. The divisions and names of the “Great” and the “Little Altai,” taken from an obscure passage of Abulghasi, should be wholly avoided.^[BP] 64The mountain system of the Altai comprehends—(a) the Altai proper, or Kolywanski Altai, which is entirely under the Russian sceptre: it lies to the west of the intersecting fissures of the Telezki Lake, which follow the direction of the meridian; and in ante-historic times probably constituted the eastern shore of the great arm of the sea, by which, in the direction of the still existing lakes, Aksakal-Barbi and Sary-Kupa,^[BQ] the Aralo-Caspian basin was connected with the Icy sea;—(b) East of the Telezki chains, which follow the direction of the meridian, the Sayani, Tangnu, and Ulangom, or Malakha ranges, all tolerably parallel with each other, and following an east and west direction. The Tangnu, which merges in the basin of the Selenga, has, from very remote times, constituted the national boundary between the Turkish race, to the south, and the Kirghis (Hakas, identical with Σάκαι), to the north.^[BR] It is the original seat of the Samoieds or Soyotes. who wandered as far as the Icy Sea, and were long regarded in Europe as a race inhabiting exclusively the coasts of the Polar Sea. The highest snow-covered summits of the Kolywan Altai are the Bielucha and the Katunia Pillars. The latter attain only a height of about 11,000 feet, or about the height of Etna. The Daurian highland, to which the mountain node of Kentei belongs, and on whose eastern margin lies the Jablonoi Chrebet, divides the depressions of the Baikal and the Amur.

2. The mountain system of the Thian-schan, or the chain of the Celestial Mountains, the Tengri-tagh of the Turks (Tukiu), and of the kindred race of the Hiongnu, is eight times as long, in an east and west direction, as the Pyrenees. Beyond, that is to say, to the west of its intersection with the meridian chain of the Bolor and Kosuyrt, the Thian-schan bears the names of Asferah and Aktagh, is rich in metals, and is intersected with open fissures, which emit hot vapours luminous at night, and which are used for obtaining sal-ammoniac.^[BS] East of the transverse Bolor and Kosyurt chain, there follow successively in the Thian-schan, the Kashgar Pass (Kaschgar-dawan), the Glacier Pass of Djeparle, which leads to Kutch and Aksu in the Tarim basin; the volcano of Pe-schan, which 65erupted fire and streams of lava at least as late as the middle of the seventh century; the great snow-covered massive elevation of Bogdo-Oola; the Solfatara of Urumtsi, which furnishes sulphur and sal-ammoniac (nao-scha), and lies in a coal district; the volcano of Turfan (or volcano of Ho-tscheu or Bischbalik), almost midway between the meridians of Turfan (Kune Turpan), and of Pidjan, and which is still in a state of activity. The volcanic eruptions of the Thian-schan chain reach, according to Chinese historians, as far back as the year 89, A.D., when the Hiongnu were pursued by the Chinese from the sources of the Irtysch as far as Kutch and Kharaschar^[BT]. The Chinese General, Teu-hian, crossed the Thian-schan, and saw “the Fire Mountains, which sent out masses of molten rock that flow to the distance of many Li.”

The great distance of the volcanoes of the interior of Asia from the sea coast is a remarkable and isolated phenomenon. Abel Rémusat, in a letter to Cordier^[BU], first directed the attention of geologists to this fact. This distance, for instance, in the case of the volcano of Pe-schan, from the north or the Icy Sea at the mouth of the Obi, is 1528 miles; and from the south or the mouths of the Indus and the Ganges, 1512 miles; so central is the position of fire-emitting volcanoes in the Asiatic continent. To the west its distance from the Caspian at the Gulf of Karuboghaz, is 1360 miles, and from the east shores of the Lake of Aral, 1020 miles. The active volcanoes of the New World had hitherto offered the most remarkable examples of great distance from the sea coast, but in the case of the volcano of Popocatepetl, in Mexico, this distance is only one hundred and thirty-two miles, and only ninety-two, one hundred and four, and one hundred and fifty-six, respectively in the South American volcanoes Sangai, Tolima, and de la Fragua. All extinct volcanoes, and all trachytic mountains, which have no permanent connexion with the interior of the earth, have been excluded from these statements^[BV]. East of the volcano of Turfat, and of the fruitful Oasis of Hami, the chain of the Thian-schan merges into the great elevated tract of Gobi, which runs in a S.W. and N.E. direction. This interruption 66of the mountain chain continues for more than 9½ degrees of longitude; it is caused by the transversal intersection of the Gobi, but beyond the latter, the more southern chain of In-schan (Silver Mountains), proceeding from west to east, to the shores of the Pacific near Pekin (north of the Pe-tscheli), forms a continuation of the Thian-schan. As we may regard the In-schan as an eastern prolongation of the fissure from which the Thian-schan is upheaved, so we may also be inclined to consider the Caucasus as a western prolongation of the same range, beyond the Great Aralo-Caspian basin or of the lowlands of Turan. The mean parallel or axis of elevation of the Thian-schan oscillates between 40° 40′ and 43° north latitude; that of the Caucasus (inclining, according to the map of the Russian Staff, from E.S.E. to W.N.W.) between 41° and 44°.^[BW] Of the four parallel chains that traverse Asia, the Thian-schan is the only one of which no summit has as yet been measured.

3. The mountain system of the Kuen-lün (Kurkun or Kulkun), including the Hindoo-Coosh, with its western prolongation in the Persian Elburz and Demavend, and the American chain of the Andes, constitute the longest lines of elevation on our planet. At the point where the meridian chain of the Bolor intersects the Kuen-lün at right angles, the latter receives the name of Onion Mountains (Tchsung-ling), a term also applied to a portion of the Bolor at the inner eastern angle of intersection. Bounding Thibet in the north, the Kuen-lün runs in a regular direction from east to west, in the parallel of 36° north latitude; until the chain is broken in the meridian of H’lassa, by the vast mountain node which surrounds the Sea of Stars, Sing so-hai (so celebrated in the mythical geography of the Chinese), and the Alpine lake of Khuku-noor. The chains of Nan-schan and Kilian-schan, lying somewhat further north, and extending to the Chinese wall near Liang-tsheu, may almost be regarded as the eastern prolongation of the Kuen-lün. To the west of the intersection of the Bolor and the Kuen-lün (Tchsung-ling), the regular direction of the axes of elevation (inclining from east to west in the Kuen-lün and Hindoo-Coosh, and from south-east 67to north-west in the Himalaya) proves, as I have elsewhere attempted to show, that the Hindoo-Coosh is a prolongation of the Kuen-lün and not of the Himalaya.^[BX] From the Taurus in Lycia to the Kafiristan, the chain follows the parallel of Rhodes (the diaphragm of Dicæarchus) over a distance of 45 degrees of longitude. The grand geological views of Eratosthenes,^[BY] which were further developed by Marinus of Tyre, and by Ptolemy, and according to which “the prolongation of the Taurus in Lycia was continued, in the same direction, through all Asia as far as India,” appear in part to be based on representations derived by the Persians and Indians from the Punjaub.

“The Brahmins maintain,” says Cosmas Indicopleustes, in his Christian Topography^[BZ], “that a line drawn from Tzinitza (Thinæ) across Persia and Romania, would exactly pass over the centre of the inhabited earth.” It is remarkable, as Eratosthenes observes, that this greatest axis of elevation in the old world passes directly through the basin (the depression) of the Mediterranean, in the parallels of 35½° and 36° north latitude, to the Pillars of Hercules.^[CA] The most eastern portion of Hindoo-Coosh is the Paropanisus of the ancients, the Indian Caucasus of the companions of the great Macedonian. The name of Hindoo-Coosh, which is so frequently used by geographers, does not in reality apply to more than one single mountain pass, where the climate is so severe, as we learn from the travels of the Arabian writer, Ibn Batuta, that many Indian slaves frequently perish from the cold.^[CB] The Kuen-lün still exhibits active fire-emitting eruptions at the distance of several hundred miles from the sea-coast. Flames, visible at a great distance, burst from the cavern of the mountain of Schinkhieu, as I learn from a translation of the Yuen-thong-ki, made by my friend Stanislaus Julien.^[CC] The loftiest summit in the Hindoo-Coosh, north-west of Jellalabad, is 20,232 feet above the level of the sea; to the west, towards Herat, the 68chain sinks to 2558 feet, rising again north of Teheran, in the volcano of Demavend, to the height of 14,675 feet.

4. The mountain system of the Himalaya has a normal direction from east to west, running more than 15 degrees of longitude (from 81° to 97°), or from the colossal mountain Dhawalagiri (28,072 feet) to the intersection of the Dzangbo-tscheu (the Irawaddy of Dalrymple and Klaproth), whose existence was long regarded as problematical, and to the meridian chains, which cover the whole of Western China, and form the great mountain group, from which spring the sources of the Kiang, in the provinces of Sse-tschuan, Hu-kuang, and Kuang-si. Next to the Dhawalagiri, the Kinchinjinga, and not the more eastern peak of Schamalari, as has hitherto been supposed, is the highest point of this portion of the Himalaya, which inclines from east to west. The Kinchinjinga, in the meridian of Sikhim, between Butan and Nepal, between the Schamalari (23,980 feet) and the Dhawalagiri, is 28,174 feet in height.

It is only within the present year that it has been trigonometrically measured with exactness, and as I learn from India through the same channel, “that a new measurement of the Dhawalagiri still leaves it the first place among all the snow-crowned summits of the Himalaya,” this mountain must necessarily have a greater elevation than the 28,072 feet hitherto ascribed to it.^[CD] The point of deflection in the direction of the chain is, near the Dhawalagiri, in 81° 22′, east longitude. From thence the Himalaya no longer follows a due west direction, but runs from S.E. to N.W., as a vast connecting system of veins between Mozufer-abad and Gilgit, merging into a part of the Hindoo-Coosh chain in the south of Kafiristan. Such a turn and alteration in the line of the axis of elevation of the Himalaya (from E.-W. to S.E.-N.W.) certainly indicates, as in the western region of our European Alpine mountains, a different age or period of elevation. The course of the Upper Indus, from the sacred lakes of Manasa and Itavana-hrada, (at an elevation of 14,965 feet,) in the vicinity of which this great river takes its origin, to Iskardo, and to the plateau of Deotsuh (at an elevation of 12,994 feet), measured by Vigne, follows in the Thibetian 69highlands the same north-westerly direction as the Himalaya.

Here are situated the Djawahir, whose height was long since accurately determined at 26,902 feet, and the Alpine valley of Caschmere (never visited by winds or storms), where, at an elevation of only 5346 feet, lies the lake of Wulur, which freezes every winter, and whose surface is never broken by a single ripple.

After considering the four great mountain systems of Asia, which, in their normal geognostic character, are true parallel chains, we must turn to the long series of alternating elevations following a direction from north to south, and which extend from Cape Comorin, opposite to the island of Ceylon, to the Icy Sea, alternating between the parallels of 66° and 77° east longitude, from S.S.E. to N.N.W. To this system of meridian chains, whose alternations remind us of faults in veins, belong the Ghauts, the Soliman chain, the Paralasa, the Bolor, and the Ural range. This interruption of the profile of the elevation is so constituted, that each new chain begins in a degree of latitude beyond that to which the preceding one had attained, all alternating successively in an opposite direction. The importance which the Greeks (probably not earlier than the second century of our era) attached to these chains running from north to south, induced Agathodæmon and Ptolemy (Tab. vii. et viii.) to regard the Bolor under the name of Imaus as an axis of elevation, which extended as far as 62° north latitude into the basin of the lower Irtysch and Obi.^[CE]

As the vertical height of mountain summits above the sea’s level (however unimportant the phenomenon of the more or less extensive folding of the crust of a planetary sphere may be in the eyes of geognosists) will always continue, like all that is difficult of attainment, to be an object of general curiosity, the present would appear to furnish a fitting place for the introduction of an historical notice relative to the gradual advance of hypsometric knowledge. When I returned to Europe in 1804, after an absence of four years, not one of the high snow-crowned summits of Asia (in the Himalaya, the Hindoo-Coosh, or the Caucasus) had been yet measured with any degree of accuracy. I was unable, therefore, to 70compare my determinations of the heights of perpetual snow in the Cordilleras of Quito or the mountains of Mexico, with any results obtained in India. The important travels of Turner, Davis, and Saunders to the highlands of Thibet, were indeed accomplished in the year 1783; but the intelligent Colebrooke justly observed that the height of the Schamalari (28° 5′ north latitude, 89° 30′ east longitude, somewhat north of Tassisudan), as given by Turner, rested on a foundation quite as slight as the assumed measurements of the heights seen from Patna and Kafiristan by Colonel Crawford and Lieutenant Macartney.^[CF] The admirable labours of Webb, Hodgson, Herbert, and the brothers Gerard, have indeed thrown considerable light on the question concerning the heights of the colossal summits of the Himalaya; but yet, in 1808, the hypsometric knowledge of the East Indian mountain chains was still so uncertain, that Webb wrote to Colebrooke, “The height of the Himalaya still remains undetermined. It is true that I have ascertained that the summits visible from the elevated plains of Rohilkand are 21,000 feet higher than that plateau, but we are ignorant of their absolute height above the sea.”

In the year 1820 it first began to be currently reported in Europe that there were not only much higher summits in the Himalaya than in the Cordilleras, but that Webb had seen in the pass of Niti, and Moorcroft in the Thibetian plateau of Daba, and the sacred lakes, fine corn-fields and fertile pasturelands at elevations far exceeding the height of Mont Blanc. This announcement was received in England with great incredulity, and opposed by doubts regarding the influence of the refraction of light. I have shown the unsoundness of such doubts in two printed treatises on the mountains of India, in the Annales de Chimie et de Physique. The Tyrolese Jesuit, Father Tiefenthaler, who in 1766 penetrated as far as the provinces of Kemaun and Nepal, had already divined the importance of the Dhawalagiri. We read on his map: “Montes Albi, qui Indis Dolaghir, nive obsiti.” Captain Webb always employs the same name. Until the measurements of the Djawahir (30° 22′ north latitude, and 79° 58′ 71east longitude, 26,902 feet in elevation), and of the Dhawalagiri (28° 40′ north latitude, and 83° 21′ east longitude, 28,072 feet in elevation), were made known in Europe, the Chimborazo, which, according to my trigonometrical measurement, was 21,422 feet in height,^[CG] was still everywhere regarded as the loftiest summit on the earth. The Himalaya appeared, therefore, at that time, to be 4323 feet or 6620 feet higher than the Cordilleras, according as the comparison was made with the Djawahir or the Dhawalagiri. Pentland’s South American travels, in the years 1827 and 1838, directed attention to two snow-crowned summits of Upper Peru, east of the lake of Titicaca, which were conjectured to be respectively 3824 and 2578 feet higher than the Chimborazo.^[CH] It has been already observed,^[CI] that the most recent computations in the measurements of the Sorata and Illimani have shown the error of this hypsometric assertion. The Dhawalagiri, therefore, on whose declivity in the river-valley of Ghandaki, the Salagrana Ammonites, so celebrated in the Brahminical ritual as symbols of the testaceous incarnation of Vishnu, are collected, still indicates a difference of elevation between both continents of more than 6600 feet.

The question has been asked, whether there may not be still greater heights in the rear of the southernmost chain, which has been as yet measured with more or less exactitude. Colonel George Lloyd, who in 1840 edited the important observations of Captain Alexander Gerard and his brother, entertains the opinion, that in that part of the Himalaya, which he somewhat indefinitely names the “Tartaric Chain” (and consequently in Northern Thibet, in the direction of the Kuen-lün, perhaps in the Kailasa of the sacred lakes or beyond Leh) there are mountain-summits which attain an elevation of from 29,000 to 30,000 feet, one or two thousand feet higher, therefore, than the Dhawalagiri.^[CJ] No definite opinion can be formed on the subject until we are in the possession 72of actual measurements, since the indication which led the natives of Quito, long before the arrival of Bouguer and La Condamine, to regard the summit of the Chimborazo as the culminating point—or the highest point within the region of perpetual snow—is rendered very deceptive in the temperate zone of Thibet, where the radiation of the table-land is so effective, and where the lower limit of perpetual snow does not constitute a regular line of equal level as in the tropics. The greatest elevation above the level of the sea that has been reached by man on the sides of the Himalaya is 19,488 feet. This elevation was gained by Captain Gerard, with seven barometers, as we have already observed, on the mountain of Tarhigang, somewhat to the north-west of Schipke.^[CK] This happens to be almost the same height as that to which I myself ascended up on the Chimborazo (on the 23rd of June, 1802), and which was reached thirty years later (16th of December, 1831) by my friend Boussingault. The unattained summit of the Tarhigang is, moreover, 1255 feet higher than the Chimborazo.

The passes across the Himalaya from Hindostan to Chinese Tartary, or rather to Western Thibet, especially between the rivers Buspa and Schipke, or Langzing Khampa, are from 15,347 to 18,544 feet in height. In the chain of the Andes I found that the pass of Assuay, between Quito and Cuenca, at the Ladera de Cadlud, was also fully 15,566 feet above the level of the sea. A great part of the Alpine plains of the interior of Asia would lie buried throughout the whole year in snow and ice, if the limits of perpetual snow were not singularly elevated, probably to about 16,626 feet, by the force of the heat radiated from the Thibetian plain, the constant serenity of the sky, the rarity of the formation of snow in the dry atmosphere, and by the powerful solar heat peculiar to the eastern continental climate, which characterizes the northern declivity of the Himalaya. Fields of barley (of Hordeum hexastichon) have been seen in Kunawur at an elevation of 14,700 feet and another variety of barley, called Ooa, and allied to Hordeum cœleste, even 73much higher. Wheat thrives admirably well in the Thibetian highlands, up to an elevation of 12,000 feet. On the northern declivity of the Himalaya, Captain Gerard found that the upper limits of the birch woods ascend to 14,069 feet; and small brushwood used by the natives for fuel in their huts is even found within the parallels of 30° 45′ and 31° north latitude, at an elevation of 16,946 feet, and therefore nearly 1280 feet higher than the lower snow-limit in the equatorial regions. It follows from the data hitherto collected that on the northern declivity of the Himalaya the mean of the lower snow-line is at least 16,626 feet, whilst on the southern declivity it falls to 12,980 feet. But for this remarkable distribution of heat in the upper strata of the atmosphere, the mountain plain of Western Thibet would be rendered uninhabitable for the millions of men who now occupy it.^[CL]

In a letter which I have lately received from India from Dr. Joseph Hooker, who is engaged in meteorological and geological observations, as well as in the study of the geography of plants, he says, “Mr. Hodgson, whom we here consider more thoroughly conversant than any other geographer with the hypsometric relations of the snow ranges, recognises the correctness of the opinions you have advanced in the third part of your Asie centrale, regarding the cause of the unequal height of the limit of perpetual snow on the northern and the southern declivity of the Himalaya range. In the trans-Sutledge region (in 36° north latitude) we often observed the snow limit as high as 20,000 feet, whilst in the passes south of Brahmaputra, between Assam and Birmah (in 27° north latitude), where the most southern snow-capped mountains of Asia are situated, the snow limit sinks to 15,000 feet.” I believe we ought to distinguish between the extreme and the mean elevations, but in both we find the formerly disputed difference between the Thibetian and the Indian declivities manifested in the clearest manner.

The local differences vary still more, as may be seen from the series of extremes given in Asie centrale, t. iii., p. 295. Alexander Gerard saw the snow-limit ascend to 20,463 feet on the Thibetian declivity of the Himalaya; and Jacquemont found it as low as 11,500 feet on the south-Indian declivity, north of Cursali on the Jumnautri.

[The recent investigations of Lieutenant Strachey show that M. Humboldt has been led astray, when treating of the Himalaya, by the very authorities on whom he placed the most reliance. The results of his inquiries on this point are given in the first volume of the Cosmos (Bohn’s Ed.), pp. 9 and 338. As the subject is one of considerable interest we give a brief sketch of Lieutenant Strachey’s^[CM] recent labours, confining ourselves to his own views, and omitting (for want of space) his somewhat lengthy exposition of the errors committed by the authorities quoted by Humboldt. The following are his personal observations regarding the southern limit of the belt of perpetual snow.

“In this part of the Himalaya it is not, on an average of years, till the beginning of December, that the snow line appears decidedly to descend for the winter. After the end of September, indeed, when the rains are quite over, light falls of snow are not of very uncommon occurrence on the higher mountains, even down to 12,000 feet; but their effects usually disappear very quickly, often in a few hours. The latter part of October, the whole of November, and the beginning of December, are here generally characterised by the beautiful serenity of the sky; and it is at this season, on the southern edge of the belt, that the line of perpetual snow is seen to attain its greatest elevation.

“The following are the results of trigonometrical measurements 75of the elevation of the inferior edge of snow on spurs of the Treslú and Nandádevi groups of peaks, made, before the winter snow had begun, in November, 1848.

“The points 1, 2 and 3 are in ridges that run in a south-westerly direction. The dip of the strata being to the north-east, the faces exposed to view from the south are for the most part very abrupt, and snow never accumulates on them to any great extent. This in some measure will account for the height to which the snow is seen to have receded on the eastern exposures, that is, upwards of 17,000 feet. On the western exposures the ground is less steep, and the snow is seen to have been observed at a considerable less elevation; but it was in very small quantities, and had probably fallen lately, so that I am inclined to think that its height, viz., about 15,000 feet, rather indicates the elevation below which the light autumnal falls of snow were incapable of lying, than that of the inferior edge of the perpetual snow. It is further to be understood, that below this level of 15,000 feet the mountains were absolutely without snow, excepting those small isolated patches that are seen in ravines, or at the head of glaciers, which, of course, do not affect such calculations as these. On the whole, therefore, I consider that the height of the snow-line on the more prominent points of the southern edge of the belt may be fairly reckoned at 16,000 feet at the very least.

“The point No. 4 was selected as being in a much more retired position than the others. It is situate not far from the head of the Pindur river. It was quite free from snow at 15,300 feet, and I shall therefore consider 15,000 feet as the 76elevation of the snow-line in the re-entering angles of the chain.

“I conclude, then, that 15,500 feet, the mean of the heights at the most and least prominent points, should be assigned as the mean elevation of the snow-line at the southern limit of the belt of perpetual snow in Kumaon; and I conceive that whatever error there may be in this estimate will be found to lie on the side of diminution rather than of exaggeration.

“This result appears to accord well with what has been observed in the Bissehir range. The account given by Dr. Gerard of his visit to the Shátúl Pass on this range, which he undertook expressly for the purpose of determining the height of the snow-line, contains the only definite information as to the limit of the perpetual snow at the southern edge of the belt that is to be found in the whole of the published writings of the Gerards; and the following is a short abstract of his observations. Dr. Gerard reached the summit of the Shátúl Pass, the elevation of which is 15,500 feet, on the 9th of August, 1822, and remained there till the 15th of the same month. He found the southern slope of the range generally free from snow, and he states that it is sometimes left without any whatever. On the top of the pass itself there was no snow; but on the northern slope of the mountain it lay as far down as about 14,000 feet. On his arrival rain was falling, and out of the four days of his stay on this pass it either rained or snowed for the greater part of three. The fresh snow that fell during this time did not lie below 16,000 feet, and some of the more precipitous rocks remained clear even up to 17,000 feet.

“The conclusion to which Dr. Gerard comes from these facts is, that the snow-line on the southern face of the Bissehir range is at 15,000 feet above the sea. But I should myself be more inclined, from his account, to consider that 15,500 feet was nearer the truth; and in this view I am confirmed by verbal accounts of the state of the passes on this range, which I have obtained from persons of my acquaintance, who have crossed them somewhat later in the year. The difference, however, is after all trifling.

“Such is the direct evidence that can be offered on the height of the snow-line at the southern limit of the belt of perpetual snow: some additional light, may, however, be 77thrown on the subject generally by my shortly explaining the state in which I have found the higher parts of the mountains at the different seasons during which I have visited them.

“In the beginning of May, on the mountains to the east of the Rámganga river, near Námik, I found the ground on the summit of the ridge, called Champwá, not only perfectly free from snow at an elevation of 12,000 feet, but covered with flowers, in some places golden with calsha and ranunculus polypetalus, in others purple with primulus. The snow had in fact already receded to upwards of 12,500 feet, behind which even a few little gentians proclaimed the advent of spring.

“Towards the end of the same month, at the end of the Pindur, near the glacier from which that river rises, an open spot on which I could pitch my tent could not be found above 12,000 feet. But here the accumulation of snow, which was considerable in all ravines even below 11,000 feet, is manifestly the result of avalanches and drift. The surface of the glacier, clear ice as well as moraines, was quite free from snow up to nearly 13,000 feet; but the effect of the more retired position of the place in retarding the melting of the snow, was manifest from the less advanced state of the vegetation. During my stay at Pinduri the weather was very bad, and several inches of snow fell; but, excepting where it had fallen on the old snow, it all melted off again in a few hours, even without the assistance of the sun’s direct rays. On the glacier, at 13,000 feet, it had all disappeared twelve hours after it fell.

“On revisiting Pinduri about the middle of October, the change that had taken place was very striking. Now not a sign of snow was to be seen on any part of the road up to the very head of the glacier; a luxuriant vegetation had sprung up, but had already almost entirely perished, and its remains covered the ground as far as I went. From this elevation, about 13,000 feet, evident signs of vegetation could be seen to extend far up the less precipitous mountains. The place is not one at which the height of the perpetual snow can be easily estimated, for on all sides are glaciers, and the vast accumulations of snow from which they are supplied, and these cannot always be readily distinguished from snow in situ; but as far as I could judge, those places which might be 78considered as offering a fair criterion were free from snow up to 15,000, or even 16,000 feet.

“Towards the end of August I crossed the Barjikang Pass, between Rálam and Juhár, the elevation of which is about 15,300 feet. There was here no vestige of snow on the ascent to the pass from the south-east, and only a very small patch remained on the north-western face. The view of the continuation of the ridge in a southerly direction was cut off by a prominent point, but no snow lay on that side within 500 feet of the pass, while to the north I estimated that there was no snow in considerable quantity within 1500 feet or more, that is, nearly up to 17,000 feet. The vegetation on the very summit of the pass was far from scanty, though it had already begun to break up into tufts, and had lost that character of continuity which it had maintained to within a height of 500 or 600 feet. Species of Potentilla, Sedum, Saxifraga, Corydalis, Aconitum, Delphinium, Thalictrum, Ranunculus Saussurea, Gentiana, Pedicularis, Primula, Rheum, and Polygonum, all evidently flourishing in a congenial climate, showed that the limits of vegetation and region of perpetual snow were still far distant.

“In addition to these facts, it may not be out of place to mention that there are two mountains visible from Almorah, Rigoli-gúdri, in Garhwal between the Kailganga and Nandákni, and Chipula, in Kumaon, between the Gori and Dauli (of Darma), both upwards of 13,000 feet in elevation, from the summits of which the snow disappears long before the end of the summer months, and which do not usually again become covered for the winter till late in December.”

These remarks are followed by an exposition of the errors into which Webb, Colebrooke, Hodgson, A. Gerard, and Jacquemont, have fallen. The heights assigned by these travellers “must all be rejected; nor can it be considered at all surprising that any amount of mistake, as to the height of the snow-line, should be made, so long as travellers cannot distinguish snow from glacier ice, or look for the boundary of perpetual snow at the beginning of the spring.”

With regard to the northern limit of the belt of perpetual snow, Lieutenant Strachey’s observations were made in September, 1848, on his way from Milam into Hundes, viâ Unta-dhúra, Kyungar-ghát, and Balch-dhúra, at the beginning of 79the month; and on his road back again, viâ Lakhur-ghát, at the end of the month.

“Of the three passes that we crossed on our way from Milam, all of them being about 17,000 feet in elevation, the first is Wata-dhára, and we saw no snow on any part of the way up to its top, which was reached in a very disagreeable drizzle of rain and snow. The final ascent to the pass from the south is about 1000 feet. The path leads up the side of a ravine, down which a small stream trickles, the ground having a generally even and rounded surface. Neither on any part of this nor on the summit of the pass itself, which is tolerably level, were there any remains of snow whatever. On the ridge to the right and left there were patches of snow a few hundred feet above; and on the northern face of the pass an accumulation remained that extended about 200 feet down, apparently the effect of the drift through the gap in which the pass lies. Below this again the ground was everywhere quite free from snow. On the ascent to Wata-dhára, at perhaps 17,000 feet, a few blades of grass were seen, but on the whole it may be said to have been utterly devoid of vegetation. On the north side of the pass, 300 or 400 feet below the summit, a cruciferous plant was the first met with.

“The Kyungar pass, which is four or six miles north of Wata-dhára, was found equally free from snow on its southern face and summit, which latter is particularly open and level. The mountains on either side were also free from snow to some height; but on the north a large bed lay a little way down the slope, and extended to about 500 feet from the top. On this pass a boragineous plant in flower was found above 17,000 feet; a species of Urtica was also got about the same altitude, and we afterwards saw it again nearly as high up on the Lakhur pass.

“In our ascent to the Balch pass no snow was observed on any of the southern spires of the range, and only one or two very small patches could be seen from the summit on the north side. The average height of the top of this range can hardly be more than 500 feet greater than that of the pass; and as a whole it certainly does not enter the region of perpetual snow. As viewed from the plains of Handes, it cannot be said to appear snowy, a few only of the peaks being tipped.

80“We returned to Milam viâ Chirchun. The whole of the ascent to The Lakhur pass was perfectly free from snow to the very top, i.e. 18,300 feet, and many of the neighbouring mountains were bare still higher. The next ridge on this route is Jainti-dhára, which is passed at an elevation of 18,500 feet, but still without crossing the least portion of snow. The line of perpetual snow is however evidently near; for though the Jainti ridge was quite free, and some of the peaks near us were clear probably to upwards of 19,000 feet, yet in more sheltered situations unbroken snow could be seen considerably below us; and on the whole I think that 18,500 feet must be near the average height of the snow-line at this place.”

A brief recapitulation of the principal results of Lieutenant Strachey’s inquiries shows us that “the snow-line or the southern edge of the belt of perpetual snow in this portion of the Himalaya is at an elevation of 15,000 feet, while on the northern edge it reaches 18,500 feet; and that on the mountains to the north of the Sutlej, or still further, it recedes even beyond 19,000 feet. The greater elevation which the snow-line attains on the northern edge of the belt of perpetual snow is a phenomenon not confined to the Thibetan declivity alone, but extending far into the interior of the chain; and it appears to be caused by the quantity of snow that falls on the northern portion of the mountains being much less than that which falls farther to the south along the line where the peaks, covered with perpetual snow, first rise above the less elevated ranges of the Himalaya.”

The letters of Dr. Joseph Hooker published during the present year (1849) in the Athenæum (pp. 431 and 1039) may also be consulted with advantage.

11. p. 5—“A tawny tribe of Herdsmen.”

The Hiongnu (Hioung-nou), whom Deguignes and with him many other historians long believed to be identical with the Huns, inhabited the vast Tartarian tract of land which is bordered on the east by Uo-leang-ho, the present territory of the Mant-schu, on the south by the Chinese wall, on the west by the U-siün, and on the north by the land of the Eleuthes But the Hiongnu belong to the Turkish, and the Huns to the Finnish or Uralian race. The northern Huns, a rude people of herdsmen, unacquainted with agriculture, were of a blackish 81brown complexion. The southern Huns, or Hajatehah called by the Byzantines Euthalites or Nephthalites, and inhabiting the eastern shore of the Caspian Sea, had fairer skins. These pursued agriculture, and dwelt in towns. They are frequently termed White Huns, and d’Herbelot even regards them as Indo-Scythians. In Deguignes^[CN] an account will be found of the Punu, the leader or Tanju of the Huns, and of the great drought and famine which led to the migration of a portion of the nation northwards about the year 46 A.D. All the details, given in his celebrated work regarding the Hiongnu, have been recently submitted by Klaproth to a rigid and learned scrutiny. From the result of his investigations it would appear, that the Hiongnu belong to the widely diffused Turkish races of the Altai and Tangnu mountain districts. The name of Hiongnu was a general name for the Ti, Thu-kiu or Turks, in the north and north-west of China, even in the third century before the Christian era. The southern Hiongnu submitted themselves to the Chinese, and in conjunction with the latter destroyed the empire of the northern Hiongnu, who were in consequence compelled to flee to the west, and thus appear to have given the first impulse to the migration of nations in Central Asia. The Huns, who were long confounded with the Hiongnu (as the Uigures were with the Ugures and Hungarians) belonged, according to Klaproth,^[CO] to the Finnish race of the Uralian mountains, which race has been variously intermixed with Germans, Turks, and Samoiedes.

The Huns (Οὖννοι) are first mentioned by Dionysius Periegetes, a writer who was able to obtain more accurate information than others regarding the interior of Asia, because, as a learned man and a native of Charax on the Arabian Gulf, he was sent back to the East by Augustus, to accompany thither his adopted son, Caius Agrippa. Ptolemy, a century later, writes the word Χοῦνοι with a strong aspiration, which, as St. Martin observes, is again met with in the geographical name of Chunigard.

12. 82p. 6—“No hewn stone.”

Representations of the sun and figures of animals have certainly been found graven in rocks on the banks of the Orinoco, near Caicara, where the woody region borders on the plain, but in the Llanos themselves not a trace of these rough memorials of earlier inhabitants has ever been discovered. It is to be regretted that no accurate account has reached us of a monument which was sent to Count Maurepas, in France, and which, according to Kalm, was discovered in the prairies of Canada, 900 French leagues (about 2700 English miles) west of Montreal, by M. de Verandrier, while engaged on an expedition to the coast of the Pacific Ocean.^[CP] This traveller met in the plains with huge masses of stone erected by the hand of man, on one of which there was an inscription believed to be in the Tartar language^[CQ]. How can so important a monument have remained uninvestigated? Can it actually have borne an alphabetical inscription, or are we not rather to believe that it must have been an historical picture, like the so-called Phœnician inscription, which has been discovered on the bank of the Taunton river, and whose authenticity has been questioned by Court de Gebelin? I indeed regard it as highly probable that these plains were once traversed by civilised nations, and it seems to me that this fact is proved by the existence of pyramidal grave-works or burrows and bulwarks of extraordinary length, between the Rocky Mountains and the Alleghanys, on which Squier and Davis have now thrown new light in their account of the ancient monuments of the Mississippi valley.^[CR] M. de Verandrier was despatched, about the year 1746, on this expedition by the Chevalier de Beauharnois, Governor-General of Canada; and several Jesuits in Quebec assured Kalm that they had actually had this so-called inscription in their hands, and that it was graven on a small tablet which was found inlaid in a hewn pillar. I have in vain requested several of my friends in France to make inquiries regarding this monument, in the event of its being in the Collection of Count Maurepas. I have also found equally uncertain accounts 83of the alphabetical writing of the American aboriginal races, in a work of Pedro de Cieça de Leon,^[CS] in Garcia,^[CT] and in Columbus’s^[CU] journal of his first voyage. M. de Verandrier maintained also that traces of the ploughshare were observed for days together in travelling over the grassy plains of Western Canada; a circumstance that other travellers, prior to him, likewise profess to have noticed. But the utter ignorance of the primitive nations of North America regarding this implement of agriculture, the want of beasts of draught, and the vast extent of surface over which these tracks extend through the prairie, tend rather to make me adopt the opinion that this singular appearance of furrows is owing to some movement of water over the earth’s surface.

13. p. 6—“It spreads like an arm of the sea.”

The great steppe, which extends from the mouth of the Orinoco to the snowy mountains of Merida, from east to west, deflects towards the south in the parallel of 8° north latitude, and occupies the whole space between the eastern declivity of the elevated mountains of New Granada and the Orinoco, which here flows in a northerly direction. That portion of the Llanos, which is watered by the Meta, Vichada, Zama, and Guaviare, connects as it were the valley of the Amazon with that of the Lower Orinoco. The word Paramo, which I have frequently employed in this work, signifies in the Spanish colonies all alpine regions which are situated from 11,000 to 14,000 feet above the level of the sea, and whose climate is rude, ungenial, and misty. In the higher Paramos hail and snow fall daily for many hours continuously, and yield a beneficial supply of humidity to the alpine plants, not from the absolute quantity of vapour in the higher strata of the air, but by the frequency of the aqueous deposits occasioned by the rapidly changing currents of air, and the variations of the electric tension. The trees found in these regions are low, and spread out in an umbrella-like form, have gnarled branches, which are constantly covered with fresh and evergreen foliage. They are 84mostly large-flowering laurel and myrtle-leaved alpine shrubs Escallonia tubar, Escallonia myrtilloides, Chuquiraga insignis, Araliæ, Weinmanniæ, Frezieræ, Gualtheriæ, and Andromeda reticulata, may be regarded as the representatives of the physiognomy of this vegetation.^[CV] To the south of the town of Santa Fé de Bogota lies the celebrated Paramo de la Suma Paz, an isolated mountain group, in which, according to Indian legends, great treasures are concealed; and hence issues a small stream or brook, which pours its foaming waters through a remarkable natural bridge in the rocky ravine of Icononzo.

In my Latin treatise, De Distributione geographica Plantarum secundum cœli temperiem et altitudinem montium, 1817, p. 104, I have thus endeavoured to characterise these Alpine regions: “Altitudine 1700–1900 hexapod: asperriæ solitudines, quæ a colonis hispanis uno nomine Paramos appellantur, tempestatum vicissitudinibus mire obnoxiæ, ad quas solutæ et emollitæ defluunt nives; ventorum flatibus ac nimborum grandinisque jactu tumultuosa regio, quæ æque per diem et per noctes riget, solis nubila et tristi luce fere nunquam calefacta. Habitantur in hac ipsa altitudine sat magnæ civitates, ut Micuipampa Peruvianorum, ubi thermometrum centes. meridie inter 5° et 8°, noctu –0°.4 consistere vidi; Huancavelica, propter cinnabaris venas celebrata, ubi altitudine 1835 hexap. fere totum per annum temperies mensis Martii Parisiis.”

14.

p. 6—“The Cordilleras of Cochabamba and the Brazilian mountains approximate to one another by means of separate transverse chains.”

The immense space between the eastern coasts of South America and the eastern declivity of the chain of the Andes is contracted by two mountain masses, which partially separate from one another the three valleys or plains of the Lower Orinoco, the Amazon, and the Rio de la Plata. The more northern mountain mass, called the group of the Parime, is opposite to the Andes of Cundinamarca, which, after extending far towards the east, assume the form of one elevated mountain, between the parallels of 66° and 68° west longitude. It is connected by the narrow mountain ridge of Pacaraima with the granitic hills of French Guiana, as I have clearly indicated in the map of Columbia which I drew up from my own 85astronomical observations. The Caribs, in their long expeditions from the missions of Carony to the plains of Rio Branco, and even to the Brazilian frontier, are obliged to traverse the crests of Pacaraima and Quimiropaca. The second group of mountains, which separates the valley of the Amazon from that of La Plata, is the Brazilian, which approximates to the promontory of Santa Cruz de la Sierra, in the province of Chiquitos, west of the Parecis hills. As neither the group of the Parime, which gives rise to the cataracts of the Orinoco, nor the Brazilian group, is directly connected with the chain of the Andes, the plains of Venezuela and those of Patagonia are directly connected with one another.^[CW]

15. p. 6—“Herds of wild dogs.”

In the Pampas of Buenos Ayres the traveller meets with European dogs, which have become wild. They live gregariously in holes and excavations, in which they conceal their young. When the horde becomes too numerous, several families go forth, and form new settlements elsewhere. The European dog barks as loudly after it has become wild, as does the indigenous American hairy species. Garcilaso asserts that, prior to the arrival of the Spaniards, the Peruvians had a race of dogs called Perros gozques; and he calls the indigenous dog Allco. In order to distinguish this animal from the European variety, it is called in the Quichua language Runa-allco, Indian dog, or dog of the natives. The hairy Runa-allco appears to be a mere variety of the shepherd’s dog. It is, however, smaller, has long yellow-ochry coloured hair, is marked with white and brown spots, and has erect and pointed ears. It barks continually, but seldom bites the natives, however it may attack the whites. When the Inca Pachacutec, in his religious wars, conquered the Indians of Xauxa and Huanca (the present valley of Huancaya and Jauja), and compelled them by force to submit to the worship of the sun, he found that dogs were made the objects of their adoration, and that the priests used the skulls of these animals as wind instruments. It would also appear that the flesh of this canine divinity was eaten by the believers.^[CX] The veneration of dogs in the valley of the 86Huancaya is probably the reason why the skulls, and even whole mummies, of these animals are sometimes found in the Huacas, or Peruvian graves of the most ancient period. Von Tschudi, the author of an admirable treatise on the Fauna Peruana, has examined these skulls, and believes them to belong to a peculiar species, which he calls Canis ingæ, and which is different from the European dog. The Huancas are still, in derision, called “dog-eaters” by the inhabitants of other provinces. Among the natives of the Rocky Mountains of North America, cooked dog’s flesh is placed before the stranger guest, as a feast of honour. Captain Frémont was present at such a dog-feast in the neighbourhood of Fort Laramie, which is one of the stations of the Hudson’s Bay Company for trading in skins and peltries with the Sioux Indians.^[CY]

The Peruvian dogs were made to play a singular part during eclipses of the moon, being beaten as long as the darkness continued. The Mexican Techichi, a variety of the common dog, which was called in Anahuac Chichi, was the only completely dumb dog. The literal signification of the word Techichi is “stone-dog,” from the Aztec, tetl, a stone. This dog was eaten according to the ancient Chinese custom, and the Spaniards found this food so indispensable before the introduction of horned cattle, that the race was gradually almost entirely extirpated.^[CZ] Buffon confounds the Techichi with the Koupara of Guiana,^[DA] which is, however, identical with the Procyon or Ursus cancrivorus, the Raton crabier, or the crab-eating Aguara-guaza of the coasts of Patagonia.^[DB] Linnæus, on the other hand, confounds the dumb dog with the Mexican Itzcuintepotzotli, a canine species which has not hitherto been perfectly described, and which is said to be characterised by a short tail, a very small head, and a large hump on the back. The name signifies a hump-backed dog, and is derived from the Aztec itzcuintli, another word for dog, and tepotzotli, humped or a humpback. I was much struck in America, especially in Quito and Peru, with the great number of black hairless dogs. They are termed Chiens turcs by Buffon, and are the Canis ægyptius of Linnæus. This species is common amongst the Indians, who, however, 87generally despise them, and treat them ill. All European dogs multiply rapidly in South America; and if no species are to be met with equal to those of Europe, it is partly owing to want of care, and partly to the circumstance that the finest varieties (as the elegant greyhound and the Danish tiger breed) have never been introduced.

Von Tschudi makes the singular remark, that on the Cordilleras, at elevations of more than 12,000 feet, delicate breeds of dogs and the European domestic cat are exposed to a particular kind of mortal disease. “Innumerable attempts have been made to keep cats as domestic animals in the town of Cerro de Pasco (lying at an elevation of 14,100 feet above the sea’s level); but such endeavours have invariably been frustrated, as both cats and dogs have died in convulsions at the end of a few days. The cats, after being attacked by convulsive fits, attempt to climb the walls, but soon fall to the ground exhausted and motionless. I frequently observed instances in Yauli of this chorea-like disease; and it seems to arise from insufficient atmospheric pressure.” In the Spanish colonies, the hairless dog, which is called Perro chinesco, or chino, is supposed to be of Chinese origin, and to have been brought from Canton, or from Manila. According to Klaproth, the race has been very common in the Chinese Empire from the earliest ages of its culture. Among the animals indigenous to Mexico, there was a very large, totally hairless, and dog-like wolf, named Xoloitzcuintli, from the Mexican xolo or xolotl, a servant or slave.^[DC]

The result of Tschudi’s observations regarding the American indigenous races of dogs are as follows:—There are two varieties almost specifically different—1. The Canis caraibicus of Lesson, totally hairless, with the exception of a small tuft of white hair on the forehead and at the tip of the tail; of a slate-gray colour, and without voice. This variety was found by Columbus in the Antilles, by Cortes in Mexico, and by Pizarro in Peru (where it suffers from the cold of the Cordilleras); and it is still very frequently met with in the warmer districts of Peru, under the name of Perros chinos. 2. The Canis ingæ, which belongs to the barking species, and has a pointed nose and pointed ears; it is now used for watching sheep and cattle; it exhibits many variations of colour, induced 88by being crossed with European breeds. The Canis ingæ follows man up the heights of the Cordilleras. In the old Peruvian graves, the skeleton of this dog is sometimes found resting at the feet of the human mummy, presenting an emblem of fidelity frequently employed by the mediæval sculptors.^[DD] European dogs, that had become wild, were found in the island of St. Domingo, and in Cuba, in the early periods of the Spanish conquest.^[DE] In the savannahs between the Meta, Arauca, and Apure, dumb dogs (perros mudos) were used as food as late as the sixteenth century. The natives called them Majos or Auries, says Alonzo de Herrera, who undertook an expedition to the Orinoco, in 1535. The highly intelligent traveller Gisecke found this variety of non-barking dogs in Greenland. The dogs of the Esquimaux live entirely in the open air, scraping for themselves at night holes in the snow, and howling like wolves, in concert with one of the troop, who sits in the middle, and takes the lead in the chorus. The Mexican dogs were castrated, in order that their flesh might become more fat and delicate. On the borders of the province of Durango, and further north, near the Slave Lake, the natives load the larger dogs with their buffalo-skin tents, (at all events they did so formerly,) when, on the change of seasons, they seek a different place of abode. These various details may all be regarded as characteristic of the mode of life led by the nations of Eastern Asia.^[DF]

16. p. 7—“Like the greater part of the Desert of Sahara, the Llanos lie within the Torrid Zone.”

Significant denominations, particularly such as refer to the form of the earth’s surface, and which arose at a period when there was only very uncertain information respecting different regions and their hypsometric relations, have led to various and long-continued geographical errors. The ancient Ptolemaic denomination of the “Greater and Lesser Atlas”^[DG] has exercised the injurious influence here indicated. There is no doubt that the snow-covered western summits of the Atlas of Morocco may 89be regarded as the Great Atlas of Ptolemy; but where is the limit of the Little Atlas? Are we still to maintain the division into two Atlas chains (which the conservative tendency of geographers has retained for 1700 years) in the territory of Algiers, and even between Tunis and Tlemse? Are we to seek a Greater and a Lesser Atlas between the coast and the parallel chains of the interior? All travellers familiar with geognostic views, who have visited Algeria since it has been in the possession of the French, contest the meaning conveyed by the generally adopted nomenclature. Among the parallel chains, that of Jurjura is generally supposed to be the highest of those which have been measured; but the well-informed Fournel (who was long Ingénieur en chef des Mines de l’Algérie) affirms that the mountain range of Aurès, near Batnah, which even at the end of March was found covered with snow, has a greater elevation. Fournel contests the existence of a Little and a Great Atlas, as I do that of a Little and a Great Altai^[DH]. There is but one Atlas, formerly called Dyris by the Mauritanians, “a name that must be applied to the foldings (rides, suites de crêtes), which form the division between the waters flowing to the Mediterranean and towards the lowland of the Sahara.” The lofty Atlas chain of Morocco inclines from north-east to south-west, and not, like the Eastern Mauritanian portion of the Atlas, from east to west. It rises into summits which, according to Renou, attain an elevation of 11,400 feet, exceeding, therefore, the height of Etna^[DI]. A singularly formed highland, of an almost square shape (Sahab el-Marga), is situated in 33° north lat., and is bounded to the south by high elevations. From thence the Atlas declines in height in a westerly direction towards the sea, about a degree south of Mogador. This south-western portion bears the name of Idrar-N-Deren.

The northern boundaries of the extended low region of the Sahara in Mauritania, as well as its southern limits towards the fertile Sudan, have hitherto been but imperfectly investigated. If we take the parallels of 16½° and 32½° north lat. as the outer limits, we obtain for the Desert, including its oases, an area of more than 1,896,000 square miles; 90or between nine and ten times the extent of Germany, and almost three times that of the Mediterranean, exclusive of the Black Sea. The best and most recent intelligence, for which we are indebted to the French observers, Colonel Daumas, and MM. Fournel, Renou, and Carette, shows us that the Desert of Sahara is composed of several detached basins, and that the number and the population of the fertile Oases is very much greater than had been imagined from the awfully desert character of the country between Insalah and Timbuctoo, and the road from Mourzouk, in Fezzan, to Bilma, Tirtuma, and Lake Tschad. It is now generally affirmed that the sand covers only the smaller portion of the lowlands. A similar opinion had been previously advanced by my Siberian travelling companion, the acute observer Ehrenberg, from what he had himself seen^[DJ]. Of larger wild animals, only gazelles, wild asses, and ostriches are to be met with.

“That lions exist in the desert,” says M. Carette, “is a myth popularised by the dreams of artists and poets, and has no foundation but in their imagination. This animal does not quit the mountains where it finds shelter, food, and drink. When the traveller questions the natives concerning these wild beasts, which Europeans suppose to be their companions in the desert, they reply, with imperturbable sang froid, ‘Have you, then, lions in your country which can drink air and eat leaves? With us lions require running water and living flesh; and therefore they only appear where there are wooded hills and water. We fear only the viper (lefa), and, in humid spots, the innumerable swarms of mosquitoes which abound there.^[DK]’”

While Dr. Oudney, in his long journey from Tripoli to Lake Tschad, estimated the elevation of the Southern Sahara at 1637 feet, and German geographers even ventured to add an additional thousand feet, Fournel, the engineer, has, by careful barometric measurements, based on corresponding observations, made it tolerably probable that a part of the northern desert is below the sea’s level. The portion of the desert which is now called “Le Zahara d’Algérie,” advances to the chains of hills of Metlili and el-Gaous, where lies the most northern of all the Oases, el-Kantara, fruitful in dates. This low basin, which reaches the parallel of 34° lat., receives 91the radiant heat of a stratum of chalk, inclined at an angle of 65° towards the south, and which is full of the shells of Inoceramus^[DL]. “Arrived at Biscara (Biskra),” says Fournel, “an indefinite horizon, like that of the sea, lay spread before us.” Between Biscara and Sidi Ocba the land is only 243 feet above the sea’s level. The inclination increases considerably towards the south. In another work^[DM], where I have brought together all the points that refer to the depression of some portions of continents below the level of the sea, I have already noticed that, according to Le Père, the bitter lakes (lacs amers) on the isthmus of Suez, when they have but little water, and, according to General Andréossy, the Natron lakes of Fayoum, are also lower than the level of the Mediterranean.

Among other manuscript notices of M. Fournel, I possess a geognostic vertical profile, with all the inflexions and inclinations of the strata, representing the surface the whole way from the coast near Philippeville to a spot near the Oasis of Biscara in the Desert of Sahara. The direction of the line on which the barometric measurements were taken is south 20° west; but the points of elevation determined are projected, as in my Mexican profiles, on a different plane, one from N. to S. Ascending uninterruptedly from Constantine, whose elevation is 2123 feet, the highest point is found between Batnah and Tizur, at only 3581 feet. In the part of the desert which lies between Biscara and Tuggurt, Fournel has succeeded in digging a series of artesian wells^[DN]. We learn from the old accounts of Shaw, that the inhabitants of the country were acquainted with a subterranean supply of water, and related fabulous tales of a “sea under the earth (bahr tôhl el-erd).” Fresh waters, which flow between clay and marl strata of the old chalk and other sedimentary formations, under the action of hydrostatic pressure, form gushing fountains when the strata are pierced^[DO]. The phenomenon of fresh water being often found near beds of rock salt, need not surprise the geognosist, acquainted with mining operations, since Europe offers many analogous phenomena.

92The riches of the desert in rock-salt, and its employment for purposes of building, have been known since the time of Herodotus. The salt zone of the Sahara (zone salifère du désert) is the most southern of the three zones which pass through Northern Africa from south-west to north-east, and is believed to be connected with the beds of rock-salt in Sicily and Palestine described by Friedrich Hoffman, and by Robinson^[DP].

The trade in salt with Sudan, and the possibility of cultivating the date-tree in the many Oasis-like depressions, caused probably by earth-slips in the beds of tertiary chalk or Keuper-gypsum, have equally contributed to animate the desert, at various parts, by human intercourse. The high temperature of the air, which renders the day’s march so oppressive across the Sahara, makes the coolness of the night (of which Denham and Sir Alexander Burnes frequently complained in the African and Asiatic deserts) so much the more remarkable. Melloni^[DQ] ascribes this coolness (which is probably produced by the radiation of heat from the ground), not to the great purity of the heavens (irraggiamento calorifico per la grande serenità di cielo nell’ immensa e deserta pianura dell’ Africa centrale), but to the extreme calm, and the absence of all movement in the air throughout the whole night^[DR].

The river Quad-Dra (Wadi Dra), which is almost dry the greater part of the year, and which, according to Renou^[DS], is one-sixth longer than the Rhine, flows into the Sahara in 32° north latitude, from the southern declivity of the Atlas of Morocco. It runs at first from north to south, until in 29° north lat., and 5° 8′ west long., it deflects at right angles to the west, and traversing the great fresh-water lake of Debaid, flows into the sea at Cape Nun, in lat. 28° 46′, and long. 11° 8′. This region, which was first rendered celebrated by the Portuguese discoveries of the fifteenth century, and whose geography has subsequently been shrouded in the deepest obscurity, is now known on the coast as the country of the Scheik of Beirouk 93(whose dominions are independent of the Emperor of Morocco). It was explored, in the months of July and August, 1840, by the French Count, Captain de Bouet-Villaumez, under the orders of his government. From manuscript and official reports it would appear that the mouth of the Quad-Dra is at present so much blocked up by sand as to have an open channel of only about 190 feet. The Saguiel-el-Hamra,—still very little known,—which comes from the south, and is supposed to have a course of at least 600 miles, flows into the same mouth at a point somewhat farther eastward. The length of these deep, but generally dry, river-beds is astonishing. They are ancient furrows, similar to those which I observed in the Peruvian desert at the foot of the Cordilleras, between the latter and the shores of the Pacific. In Bouet’s manuscript narrative^[DT], the mountains which rise to the north of Cape Nun are estimated at the great height of 9,186 feet.

It is generally supposed that Cape Nun was discovered in 1433 by the Knight Gilianez, despatched under the order of the celebrated Infante, Henry, Duke of Viseo, and founder of the Academy of Sagres, which was presided over by the pilot and cosmographer, Mestre Jacomè, of Majorca; but the Portulano Mediceo,—the work of a Genoese navigator of the year 1351,—already contains the name of “Cavo di Non.” The doubling of this Cape was as much dreaded as has been since then the passage round Cape Horn; although it is only 23′ north of the parallel of Teneriffe, and might be reached by a few days’ sail from Cadiz. The Portuguese adage, “Quem passa o Cabo de Num, ou tornarà ou não,” could not intimidate the Infante, whose heraldic French motto of “Talent de bien faire,” well expressed his noble, enterprising, and vigorous character. The name of this Cape, which has long been supposed to originate in a play of words on the negative particle, does not appear to me to be of Portuguese origin. Ptolemy placed on the north-west coast of Africa a river Nuius, in the Latin version Nunii ostia. Edrisi refers to a town, Nul, or Wadi Nun, somewhat further south, and about three days’ journey in the interior, named by Leo Africanus Belad de Non. Several European navigators had penetrated far to the south of Cape Nun before the Portuguese squadron under Gilianez. The Catalan, Don Jayme Ferrer, in 1346, as we learn from the Atlas Catalan, published at 94Paris by Buchon, had advanced as far as the Gold River (Rio do Ouro), in 23° 56′ north lat.; while the Normans, at the close of the fourteenth century, reached Sierra Leone in 8° 30′ north latitude. The merit of having been the first to cross the equator in the Western Ocean incontrovertibly belongs, like so many other great achievements, to the Portuguese.

17. p. 7.—“As a grassy plain, resembling many of the Steppes of Central Asia.”

The Llanos of Caracas, of the Rio Apure and the Meta, which are the abode of numerous herds of cattle, are, in the strictest sense of the word, grassy plains. The two families of the Cyperaceæ and the Gramineæ, which are the principal representatives of the vegetation, yield numerous forms of Paspalum (Paspalum leptostachyum, P. lenticulare), of Kyllingia (Kyllingia monocephala (Rottb.), K. odorata), of Panicum (Panicum granuliferum, P. micranthum), of Antephora, Aristida, Vilfa, and Anthisteria (Anthisteria reflexa, A. foliosa). It is only here and there that any herbaceous dicotyledon, as the low-growing species of Mimosa intermedia and M. dormiens, which are so grateful to the wild horses and cattle, are found interspersed among the Gramineæ. The natives very characteristically apply to this group the name of “Dormideras,” or sleepy plants, because the delicate and feathery leaves close on being touched. For many square miles not a tree is to be seen; but where a few solitary trees are found, they are, in humid districts, the Mauritia Palm, and, in arid spots, a Proteacea described by Bonpland and myself, the Rhopala complicata (Chaparro bobo), which Willdenow regarded as an Embothrium; also the useful Palma de Covija or de Sombrero; and our Corypha inermis, an umbrella palm allied to Chamærops, and used by the natives for the covering of their huts. How much more varied and rich is the aspect of the Asiatic plains! In a great portion of the Kirghis and Kalmuck Steppes which I have traversed (extending over a space of 40 degrees of longitude), from the Don, the Caspian Sea and the Orenburg-Ural river Jaik, to the Obi and the Upper Irtysch, near the Lake Dsaisang, the extreme range of view is never bounded by a horizon in which the vault of heaven appears to rest on an unbroken sea-like plain, as is so frequently the case in the Llanos, Pampas, and Prairies of America. I have, indeed, 95never observed anything approaching to this phenomenon, excepting, perhaps, where I have looked only towards one quarter of the heavens, for the Asiatic plains are frequently intersected by chains of hills, or clothed with coniferous woods. The Asiatic vegetation, too, in the most fruitful pasture lands, is by no means limited to the family of the Cyperaceæ, but is enriched by a great variety of herbaceous plants and shrubs. In the season of spring, small snowy white and red flowering Rosaceæ and Amygdaleæ (Spiræa, Cratægus, Prunus spinosa, Amygdalus nana), present a pleasing appearance. I have elsewhere spoken of the tall and luxuriant Synanthereæ (Saussurea amara, S. salsa, Artemisiæ, and Centaureæ), and of leguminous plants, (species of the Astragalus, Cytisus and Caragana). Crown Imperials (Fritillaria ruthenica and F. meleagroides), Cypripediæ and tulips gladden the eye with their varied and bright hues.

A contrast is presented to this charming vegetation of the Asiatic plains by the dreary Salt Steppes, especially by that portion of the Barabinski Steppe which lies at the base of the Altai Mountains, between Barnaul and the Serpent Mountain, and by the country to the east of the Caspian. Here the social Chenopodiæ, species of Salsola, Atriplex, Salicorniæ, and Halimocnemis crassifolia^[DU], cover the clayey soil with patches of verdure. Among the five hundred phanerogamic species which Claus and Göbel collected on the Steppes, Synanthereæ, Chenopodiæ, and Cruciferæ were more numerous than the grasses; the latter constituting only ¹⁄₁₁th of the whole, and the two former ⅐th and ⅑th. In Germany, owing to the alternation of hills and plains, the Glumaceæ (comprising the Gramineæ, Cyperaceæ, and Juncaceæ) constitute ⅐th, the Synanthereæ (Compositæ) ⅛th, and the Cruciferæ ¹⁄₁₈th of all the German Phanerogamic species. In the most northern part of the flat land of Siberia, the extreme limit of tree and shrub vegetation (Coniferæ and Amentaceæ) is, according to Admiral Wrangell’s fine map, 67° 15′ north lat., in the districts contiguous to Behring’s Straits, while more to the west, towards the banks of the Lena, it is 71°, which is the parallel of the North Cape of Lapland. The plains bordering on the Polar Sea are the domain of Cryptogamic plants. They are called Tundra (Tuntur in Finnish), and are vast swampy districts, covered 96partly with a thick mantle of Sphagnum palustre and other Liverworts, and partly with a dry snowy-white carpet of Cenomyce rangiferina (Reindeer-moss), Stereocaulon paschale, and other lichens. “These Tundra,” says Admiral Wrangell, in his perilous expedition to the Islands of New Siberia, so rich in fossil wood, “accompanied me to the extremest Arctic coast. Their soil is composed of earth that has been frozen for thousands of years. In the dreary uniformity of the landscape, and surrounded by reindeer, the eye of the traveller rests with pleasure on the smallest patch of green turf that shows itself on a moist spot.”

18. p. 7.—“A diversity of causes diminishes the dryness and heat of the New Continent.”

I have endeavoured to compress the various causes of the humidity and lesser heat of America into one general category. It will of course be understood, that I can only have reference here to the general hygroscopic condition of the atmosphere, and the temperature of the whole continent; for in considering individual regions, as for instance, the island of Margarita, or the coasts of Cumana and Coro, it will be found that these exhibit an equal degree of dryness and heat with any portion of Africa.

The maximum of heat, at certain hours of a summer’s day, considered with reference to a long series of years, has been found to be almost the same in all regions of the earth, whether on the Neva, the Senegal, the Ganges, or the Orinoco, namely, between 93° and 104° Fahr., and on the whole not higher; provided that the observation be made in the shade, far from solid radiating bodies, and not in an atmosphere filled with heated dust or granules of sand, and not with spirit-thermometers, which absorb light. The fine grains of sand (forming centres of radiant heat) which float in the air, were probably the cause of the fearful heat (122° to 133° Fahr. in the shade) in the Oasis of Mourzouk to which my unhappy friend Ritchie, who perished there, and Captain Lyon, were exposed for weeks. The most remarkable instance of a high temperature, in an air probably free from dust, is mentioned by an observer who well knew how to arrange and correct all his instruments with the greatest accuracy. Rüppel found the temperature 110°.6 Fahr. at Ambukol, in Abyssinia, with a cloudy sky, a strong south-west wind, and an approaching thunder-storm. 97The mean annual temperature of the tropics, or the actual climate of the region of palms, is on the main land between 78°.2 and 85°.5 Fahr., without any sensible difference between the observations made in Senegal, Pondichery, and Surinam^[DV].

The great coolness, one might almost say coldness, which prevails during a great portion of the year in the tropics, on the coast of Peru, and which causes the mercury to fall to 59° Fahr., is, as I hope to show in another place, not to be attributed to the effect of neighbouring mountains covered with snow, but rather to the mist (garua) which obscures the sun’s disk, and to a current of cold sea-water commencing in the antarctic regions, and which coming from the south-west, strikes the coast of Chili near Valdivia and Concepcion, and is thence propelled with violence, in a northerly direction, to Cape Pariña. On the coast of Lima, the temperature of the Pacific is 60°.2 Fahr., whilst it is 79°.2 Fahr. under the same parallel of latitude when outside the current. It is singular, that so remarkable a fact should have remained unnoticed, until my residence on the coast of the Pacific, in October, 1802.

The variations of temperature, of many parts of the earth, depend principally on the character of the bottom of the aërial ocean, or in other words, on the nature of the solid or fluid (continental or oceanic) base on which the atmosphere rests. Seas, traversed in various directions by currents of warm and cold water (oceanic rivers), exert a different action from articulated or inarticulated continental masses or islands, which may be regarded as the shoals in the aërial ocean, and which, notwithstanding their small dimensions, exercise, even to great distances, a remarkable degree of influence on the climate of the sea. In continental masses, we must distinguish between barren sandy deserts, savannahs, (grassy plains,) and forest districts. In Upper Egypt and in South America, Nouet and myself found, at noon, the temperature of the ground, which was composed of granitic sand, 154° and 141° Fahr. Numerous careful observations instituted at Paris, have given, according to Arago, 122° and 126°.5 Fahr.^[DW] The Savannahs, which, between the Missouri and the Mississippi, are called Prairies, and which appear in the south at 98the Llanos of Venezuela and the Pampas of Buenos Ayres, are covered with small monocotyledons, belonging to the family of the Cyperaceæ, and with grasses, whose dry pointed stalks, and whose delicate, lanceolate leaves radiate towards the unclouded sky, and possess an extraordinary power of emission. Wells and Daniell^[DX] have even seen in our latitude, where the atmosphere has a much less considerable degree of transparency, the thermometer fall to 14°.5, or 18° Fahr. on being placed on the grass. Melloni has most ably shown^[DY] that in a calm, which is a necessary condition of a powerful radiation, and of the formation of dew, the cooling of the stratum of grass is promoted by the falling to the ground of the cooler particles of air, as being the heavier.

In the vicinity of the equator, under the cloudy sky of the Upper Orinoco, the Rio Negro and the Amazon, the plains are covered with dense primeval forests; but to the north and south of this woody district, there extend, from the zone of palms and of tall dicotyledonous trees in the northern hemisphere, the Llanos of the Lower Orinoco, the Meta, and Guaviare; and in the south, the Pampas of the Rio de la Plata and of Patagonia. The area thus covered by grassy plains, or Savannahs, in South America, is at least nine times greater than that of France.

The forest region acts in a threefold manner, by the coolness induced by its shade, by evaporation, and by the cooling process of radiation. Forests uniformly composed in our temperate zone of “social” plants, belonging to the families of the Coniferæ or Amentaceæ (the oak, beech, and birch), and under the tropics composed of plants not living socially, protect the ground from direct insolation, evaporate the fluids they have themselves produced, and cool the contiguous strata of air by the radiation of heat from their leafy appendicular organs. The leaves are by no means all parallel to one another, and present different inclinations towards the horizon; and according to the laws established by Leslie and Fourier, the influence of this inclination on the quantity of heat emitted by radiation is such, that the radiating power of a given measured surface a, having a given oblique direction, is equal to the radiating power of a leaf of the size of a projected on a horizontal 99plane. In the initial condition of radiation of all the leaves which form the summit of a tree, and which partially cover each other, those which are directly presented towards the unclouded sky, will be first cooled.

This production of cold (or the exhaustion of heat by emission) will be the more considerable in proportion to the thinness of the leaves. A second stratum of leaves has its upper surface turned to the under surface of the former, and will give out more heat by radiation towards that stratum than it can receive from it. The result of this unequal exchange will then be a diminution of temperature for the second stratum also. A similar action will extend from stratum to stratum, till all the leaves of the tree, by their greater or less radiation, as modified by their difference of position, have passed into a condition of stable equilibrium, of which the law may be deduced by mathematical analysis. In this manner, in the serene and long nights of the equinoctial zone, the forest air, which is contained in the interstices between the strata of leaves, becomes cooled by the process of radiation; for a tree, a horizontal section of whose summit would hardly measure 2000 square feet, would, in consequence of the great number of its appendicular organs (the leaves), produce as great a diminution in the temperature of the air as a space of bare land or turf many thousand times greater than 2000 square feet.^[DZ] I have thus sought to develope somewhat fully the complicated relations which the action of great forest regions exerts on the atmosphere, because they have so often been touched upon in connection with the important question of the climate of ancient Germany and Gaul.

As in the old continent, European civilization has had its principal seat on the western coast, it could not fail to be early remarked that under equal degrees of latitude the opposite eastern littoral region of the United States of North America was several degrees colder, in mean annual temperature, than Europe, which is, as it were, a western peninsula of Asia, and bears much the same relation to it as Brittany does to the rest of France. The fact, however, escaped notice that these differences decrease from the higher to the lower latitudes, and that they are hardly perceptible below 30°. For the west coast of the New Continent exact observations 100of the temperature are still almost entirely wanting; but the mildness of the winter in New California shows that in reference to their mean annual temperature, the west coasts of America and Europe under the same parallels, scarcely present any differences. The annexed table gives the mean annual temperatures, which correspond to the same geographical latitudes, on the eastern coast of the New Continent and the western coast of Europe:—

In the preceding table the number placed before the fraction represents the mean annual temperature, the numerator of the fraction, the mean winter temperature, and the denominator the mean summer temperature. Besides the more marked difference between the mean annual temperatures, there is also a very striking contrast between the opposite coasts in respect to the distribution of heat over the different seasons of the year; and it is indeed this distribution which exerts the greatest influence on our bodily feelings and on the process of vegetation. Dove^[EA] makes the general remark, that 102the summer temperature of America is lower under equal degrees of latitude than that of Europe. The climate of St. Petersburgh (lat. 59° 56′), or to speak more correctly, the mean annual temperature of that city, is found on the eastern coast of America, in lat. 47° 30′, or 12° 30′ more to the south; and in like manner we find the climate of Königsberg (lat. 54° 43′) at Halifax in lat. 44° 39′. Toulouse (lat. 43° 36′) corresponds in its thermic relations to Washington.

It is very hazardous to attempt to obtain any general results respecting the distribution of heat in the United States of North America, since there are three regions to be distinguished—1, the region of the Atlantic States, east of the Alleghanys; 2, the Western States, in the wide basin between the Alleghanys and the Rocky Mountains, watered by the Mississippi, the Ohio, the Arkansas, and the Missouri; and 3, the elevated plains between the Rocky Mountains and the Coast Range of New California, through which the Oregon or Columbia river wends its course. Since the commendable establishment by John Calhoun of uninterrupted observations of temperature, made on a uniform plan, at thirty-five military stations, and reduced to diurnal, mensal, and annual means, we have attained more correct climatic views than were generally held in the time of Jefferson, Barton, and Volney. These meteorological stations extend from the point of Florida and Thompson’s Island (West Key), lat. 24° 33′, to the Council Bluffs on the Missouri; and if we reckon Fort Vancouver (lat. 45° 37′), among them, they include a space extending over forty degrees of longitude.

It cannot be affirmed that on the whole the second region has a higher mean annual temperature than the first, or Atlantic. The further advance towards the north of certain plants on the western side of the Alleghanys, depends partly on the nature of those plants and partly on the different distribution through the seasons of the year of the same annual amount of heat. The broad valley of the Mississippi enjoys, at its northern extremity, the warming influence of the Canadian lakes, and at the south, that of the Mexican Gulf-Stream. These five lakes (Lakes Superior, Michigan, Huron, Erie, and Ontario,) cover an area of 92,000 square miles. The climate is so much milder and more uniform in the vicinity of the lakes, that at Niagara, for instance (in 43° 15′ 103north lat.), the mean annual winter temperature is only half a degree below the freezing-point, whilst, at a distance from the lakes, in 44° 53′ north lat. at Fort Snelling, near the confluence of the river St. Peter with the Mississippi, the mean winter temperature is 15°.8 Fahr.^[EB] At this distance from the Canadian lakes, whose surface is from five to upwards of six hundred feet above the sea’s level, whilst the bottom of Lakes Michigan and Huron is five hundred feet below it, recent observations have shown that the climate of the country possesses the actual continental character of hotter summers and colder winters. “It is proved,” says Forry, “by our thermometrical data, that the climate west of the Alleghany chain is more excessive than that on the Atlantic side.” At Fort Gibson, on the Arkansas river, which falls into the Mississippi, in lat. 35° 47′, where the mean annual temperature hardly equals that of Gibraltar, the thermometer was observed, in August, 1834, to rise to 117° Fahr. when in the shade, and without any reflected heat from the ground.

The statements so frequently advanced, although unsupported by measurements, that since the first European settlements in New England, Pennsylvania, and Virginia, the destruction of many forests on both sides of the Alleghanys, has rendered the climate more equable,—making the winters milder and the summers cooler,—are now generally discredited. No series of thermometric observations worthy of confidence extend further back in the United States than seventy-eight years. We find from the Philadelphia observations that from 1771 to 1824, the mean annual heat has hardly risen 2°.7 Fahr.;—an increase that may fairly be ascribed to the extension of the town, its greater population, and to the numerous steam-engines. This annual increase of temperature may also be owing to accident, for in the same period I find that there was an increase of the mean winter temperature of 2° Fahr.; but with this exception the seasons had all become somewhat warmer. Thirty-three years’ observations at Salem in Massachusetts show scarcely any difference, the mean of each one oscillating within 1° of Fahrenheit, about the mean of the whole number; and the winters of Salem, instead of having been rendered more mild, as conjectured, from the eradication 104of the forests, have become colder by 4° Fahr. during the last thirty-three years.^[EC]

As the east coast of the United States may be compared, in equal latitudes, with the Siberian and Chinese eastern coasts of Europe, in respect to mean annual temperature, so the western coasts of Europe and America have also very justly been compared together. I will here only adduce a few instances from the western region of the Pacific, for two of which, viz., Sitka. (New Archangel,) in Russian America, and Fort George. (having the same latitudes respectively as Gottenburg and Geneva,) we are indebted to Admiral Lütke’s voyage of circumnavigation. Iluluk and Danzig are situated in about the same parallel of latitude, and although the mean temperature of Iluluk, owing to its insular climate and the cold sea current contiguous to it, is lower than that of Danzig, the winters of the former are milder than those of the Baltic city.

Snow is hardly ever seen on the banks of the Oregon or Columbia river, and ice on the river lasts only a few days. The lowest temperature which Mr. Ball ever observed there (in 1838) was 18°.4 Fahr.^[ED] A cursory glance at the summer and 105winter temperatures given above, suffices to show that a true insular climate prevails on and near the western coasts; whilst the winter cold is less considerable than in the western part of the old continent, the summers are much cooler. This contrast is made most apparent when we compare the mouth of the Oregon with Forts Snelling and Howard, and the Council Bluffs in the interior of the Mississippi and Missouri basin, (44°–46° north lat.,) where, to speak with Buffon, we find an excessive or true continental climate,—a winter cold, which on some days is –32° or even –37° Fahr., followed by a mean summer’s heat, which rises to 69° and 71°.4 Fahr.

19. p. 8.—“As if America had emerged later from the chaotic covering of waters.”

The acute natural inquirer Benjamin Smith Barton, expresses himself thus accurately:^[EE]—“I cannot but deem it a puerile supposition, unsupported by the evidence of nature, that a great part of America has probably later emerged from the bosom of the ocean than the other continents.” I have already elsewhere treated of this subject in a memoir on the primitive nations of America:^[EF]—“The remark has been too frequently made by authors of general and well-attested merit that America was in every sense of the word a new continent. The luxuriance of vegetation, the vast mass of waters in the rivers, and the continued activity of great volcanoes, confirm the fact (say these writers,) that the still agitated and humid earth is in a condition approximating more closely to the chaotic primordial state of our planet than the old continent. Such ideas appeared to me, long before my travels in those regions, no less unphilosophical than at variance with generally acknowledged physical laws. These imaginary representations of an earlier age and a want of repose, and of the increase of dryness and inertia with the increased age of our globe, could only have been framed by those who seek to discover striking contrasts between the two hemispheres, and who do not endeavour to consider the construction of our terrestrial planet from one grand and general point of view. Are we to regard the southern as more recent than the northern part of Italy, simply because the former is almost constantly disturbed by earthquakes and volcanic eruptions? How 106inconsiderable, moreover, are the phenomena presented by our volcanoes and earthquakes, when compared with the convulsions of nature which the geognosist must conjecture to have occurred in the chaotic condition of our globe, when mountain masses were upheaved, solidified, or cleft asunder? Different causes must also occasion a diversity of effects in the forces of nature in parts of the earth remote from one another. The volcanoes in the new continent,” (of which I still count about twenty-eight,) “may probably have continued longer active, because the high mountain ridges on which they are erupted in rows upon long fissures are nearer to the sea, and because this vicinity appears to modify the energy of the subterranean fire, in a manner which, with few exceptions, has not yet been explained. Besides, both earthquakes and fire-erupting mountains act periodically. At present” (this I wrote forty-two years ago,) “physical disquietude and political repose prevail in the new continent, whilst in the old continent the calm repose of nature is contrasted with the dissensions of different nations. The time may however come, when this strange contrast between physical and moral forces may change its theatre of action from one quarter of the world to another. Volcanoes enjoy centuries of repose between their manifestations of activity; and the idea that in the older countries nature must be characterized by a certain repose and quietude, has no other foundation than in the mere caprice of the imagination. There exists no reason for assuming that one side of our planet is older or more recent than the other. Islands, as the Azores and many flat islands of the Pacific, which have been upheaved by volcanoes, or been gradually formed by coral animals, are indeed more recent than many plutonic formations of the European central chain. Small tracts of land, as Bohemia and Kashmeer, and many of the valleys in the moon, inclosed by a ring of mountains, may continue for a long time under the form of a sea, owing to partial inundations, and after the flowing off of these inland waters, the bottom, on which plants would gradually manifest themselves, might indeed be figuratively regarded as of more recent origin. Islands have been connected together into continental masses by upheaval, whilst other parts of the previously existing land have disappeared in consequence of the subsidence of the oscillating ground; but general submersions can, from hydrostatic laws, only be imagined as embracing simultaneously 107all parts of the earth. The sea cannot permanently submerge the vast lowlands of the Orinoco and the Amazon, without at the same time destroying our Baltic lands. Moreover the succession and identity of the floetz strata, and of the organic remains of plants and animals belonging to the primitive world, inclosed in those strata, show that several great depositions have occurred almost simultaneously over the whole earth.”^[EG]

20. p. 8.—“The Southern Hemisphere is cooler and more humid than the Northern.”

Chili, Buenos Ayres, the southern part of Brazil, and Peru, enjoy the cool summers and mild winters of a true insular climate, owing to the narrowness and contraction of the continent towards the south. This advantage of the Southern Hemisphere is manifested as far as 48° or 50° south lat., but beyond that point, and nearer the Antarctic Pole, South America is an inhospitable waste. The different degrees of latitude at which the southern extremities of Australia, including Van Diemen’s Island, of Africa, and America, terminate, give to each of these continents its peculiar character. The Straits of Magellan lie between the parallels of 53° and 54° south lat.; and notwithstanding this, the thermometer falls to 41° Fahr. in the months of December and January, when the sun is eighteen hours above the horizon. Snow falls almost daily in the lowlands, and the maximum of atmospheric heat observed by Churruca in 1788, during the month of December, and consequently in the summer of that region, did not exceed 52°.2 Fahr. The Cabo Pilar, whose turret-like rock is only 1394 feet in height, and which forms the southern extremity of the chain of the Andes, is situated in nearly the same latitude as Berlin.^[EH]

Whilst in the Northern Hemisphere all continents fall, in their prolongation towards the Pole, within a mean limit, which corresponds tolerably accurately with 70°, the southern extremities of America. (in Tierra del Fuego, which is so 108deeply indented by intersecting arms of the sea,) of Australia, and of Africa, are respectively 34°, 46° 30′ and 56° distant from the South Pole. The temperature of the unequal extents of ocean which separate these southern extremities from the icy Pole contributes essentially towards the modification of the climate. The areas of the dry land of the two hemispheres separated by the equator are as 3 to 1. But this deficiency of continental masses in the Southern Hemisphere is greater in the temperate than in the torrid zone, the ratio being in the former at 13 to 1, and in the latter as 5 to 4. This great inequality in the distribution of dry land exerts a perceptible influence on the strength of the ascending atmospheric current, which turns towards the South Pole, and on the temperature of the Southern Hemisphere generally. Some of the noblest forms of tropical vegetation, as for instance tree-ferns, advance south of the equator to the parallels of from 46° to 53°, whilst to the north of the equator they do not occur beyond the tropic of Cancer.^[EI] Tree-ferns thrive admirably well at Hobart Town in Van Diemen’s Land (42° 53′ lat.), with a mean annual temperature of 52°.2 Fahr., and therefore on an isothermal line less by 3°.6 Fahr. than that of Toulon. Rome, which is almost one degree of latitude further from the equator than Hobart Town, has an annual temperature of 59°.7 Fahr.; a winter temperature of 46°.6 Fahr., and a summer temperature of 86° Fahr.; whilst in Hobart Town these three means are respectively 52°, 42°.1, and 63° Fahr. In Dusky Bay, New Zealand, tree-ferns thrive in 46° 8′ lat., and in the Auckland and Campbell Islands in 53° lat.^[EJ]

In the Archipelago of Tierra del Fuego, having a mean winter temperature of 33° Fahr., and a mean summer temperature of only 50° Fahr., in the same latitude as Dublin, Captain King found “vegetation thriving most luxuriantly in large woody-stemmed trees of Fuchsia and Veronica;” whilst this vigorous vegetation, which, especially on the western coast of America (in 38° and 40° south lat.), has been so picturesquely described by Charles Darwin, suddenly disappears south of Cape Horn, on the rocks of the Southern Orkney and Shetland Islands, and of the Sandwich Archipelago. These islands, but scantily covered with grass, moss, and 109lichens, Terres de Désolation, as they have been called by French navigators, lie far to the north of the Antarctic Polar Circle; whilst in the Northern Hemisphere, in 70° lat., on the extremest verge of Scandinavia, fir-trees reach a height of more than 60 feet.^[EK] If we compare Tierra del Fuego, and more particularly Port Famine, in the Straits of Magellan, 53° 38′ lat., with Berlin, which is situated one degree nearer the equator, we shall find for Berlin, 47°.3 38°.9
62°.3; and for Port Famine, 42°.6 34°.7
50°.0 Fahr. I subjoin the few certain data of temperature which we at present possess of the temperate zones of the Southern Hemisphere, and which may be compared with the temperatures of northern regions in which the distribution of summer heat and winter cold is so unequal. I make use of the convenient mode of notation already explained in which the number standing before the fraction indicates the mean annual temperature, the numerator the winter, and the denominator the summer temperature.

21. 110p. 9.—“One connected sea of sand.”

As we may regard the social Erica as furnishing one continuous vegetable covering spread over the earth’s surface, from the mouth of the Scheldt to the Elbe, and from the extremity of Jutland to the Harz mountains, so may we likewise trace the sea of sand continuously through Africa and Asia, from Cape Blanco to the further side of the Indus, over an extent of 5,600 miles. The sandy region mentioned by Herodotus, which the Arabs call the Desert of Sahara, and which is interrupted by oases, traverses the whole of Africa like a dried arm of the sea. The valley of the Nile is the eastern boundary of the Lybian desert. Beyond the Isthmus of Suez and the porphyritic, syenitic, and greenstone rocks of Sinai begins the Desert mountain plateau of Nedschd, which occupies the whole interior of the Arabian Peninsula, and is bounded to the west and south by the fruitful and more highly favoured coast-lands of Hedschaz and Hadhramaut. The Euphrates forms the eastern boundary of the Arabian and Syrian desert. The whole of Persia, from the Caspian Sea to the Indian Ocean, is intersected by immense tracts of sand (bejaban), among which we may reckon the soda and potash Deserts of Kerman, Seistan, Beludschistan, and Mekran. The last of these barren wastes is separated by the Indus from the Desert of Moultan.

22. p. 9.—“The western portion of Mount Atlas.”

The question of the position of the Atlas of the ancients has often been agitated in our own day. In making this inquiry, ancient Phœnician traditions are confounded with the statements of the Greeks and Romans regarding Mount Atlas at a less remote period. The elder Professor Ideler, who combined a profound knowledge of languages with that of astronomy and mathematics, was the first to throw light on this obscure subject; and I trust I may be pardoned if I insert the communications with which I have been favoured by this enlightened observer.

“The Phœnicians ventured at a very early period in the world’s history to penetrate beyond the Straits of Gibraltar. They founded Gades and Tartessus on the Spanish, and Lixus, together with many other cities on the Mauritanian coasts of the Atlantic Ocean. They sailed northward along these shores 111to the Cassiterides, from whence they obtained tin, and to the Prussian coasts where they procured amber found there; whilst southward they penetrated as far as Madeira and the Cape de Verd Islands. Amongst other regions they visited the Archipelago of the Canary Isles, where their attention was arrested by the Peak of Teneriffe, whose great height appears to be even more considerable than it actually is from the circumstance of the mountain projecting directly from the sea. Through their colonies established in Greece, especially under Cadmus in Bœotia, the Greeks were made acquainted with the existence of this mountain which soared high above the region of clouds, and with the ‘Fortunate Islands’ on which this mountain was situated, and which were adorned with fruits of all kinds, and particularly with the golden orange. By the transmission of this tradition through the songs of the bards, Homer became acquainted with these remote regions, and he speaks of an Atlas to whom all the depths of ocean are known, and who bears upon his shoulders the great columns which separate from one another the heavens and the earth,^[EL] and of the Elysian Plains, described as a wondrously beautiful land in the west.”^[EM] Hesiod expresses himself in a similar manner regarding Atlas, whom he represents as the neighbour of the Hesperides.^[EN] The Elysian Plains, which he places at the western limits of the earth, he terms the ‘Islands of the Blessed.’^[EO] Later poets have still further embellished these myths of Atlas, the Hesperides, their golden apples, and the Islands of the Blessed, which are destined to be the abode of good men after death, and have connected them with the expeditions of the Tyrian God of Commerce, Melicertes, the Hercules of the Greeks.

“The Greeks did not enter into rivalship with the Phœnicians and Carthaginians in the art of navigation until a comparatively late period. They indeed visited the shores of the Atlantic, but they never appear to have advanced very far. It is doubtful whether they had penetrated as far as the Canary Isles and the Peak of Teneriffe; but be this as it may, they were aware that Mount Atlas, which their poets had described 112as a very high mountain situated on the western limits of the earth, must be sought on the western coast of Africa. This too was the locality assigned to it by their later geographers Strabo, Ptolemy, and others. As however no mountain of any great elevation was to be met with in the north-west of Africa, much perplexity was entertained regarding the actual position of Mount Atlas, which was sought sometimes on the coast, sometimes in the interior of the country, and sometimes in the vicinity of the Mediterranean, or further southward. In the first century of the Christian era, when the armies of Rome had penetrated to the interior of Mauritania and Numidia, it was usual to give the name of Atlas to the mountain chain which traverses Africa from west to east in a parallel direction with the Mediterranean. Pliny and Solinus were both, however, fully aware that the description of Atlas given by the Greek and Roman poets did not apply to this mountain range, and they therefore deemed it expedient to transfer the site of Mount Atlas, which they described in picturesque terms, in accordance with poetic legends, to the terra incognita of Central Africa. The Atlas of Homer and Hesiod can, therefore, be none other than the Peak of Teneriffe, while the Atlas of Greek and Roman geographers must be sought in the north of Africa.”

I will only venture to add the following remarks to the learned explanations of Professor Ideler. According to Pliny and Solinus, Atlas rises from the midst of a sandy plain (e medio arenarum), and its declivity affords pasture to elephants, which have undoubtedly never been known in Teneriffe. That which we now term Atlas is a long mountain ridge. How could the Romans have recognised one isolated conical elevation in this mountain range of Herodotus? May the cause not be ascribed to the optical illusion by which every mountain chain, when seen laterally from an oblique point of view, appears to be of a narrow and conical form? I have often, when at sea, mistaken long mountain ranges for isolated mountains. According to Höst, Mount Atlas is covered with perpetual snow near Morocco. Its elevation must therefore be upwards of 11,500 feet at that particular spot. It seems to me very remarkable that the barbarians, the ancient Mauritanians, if we are to believe the testimony of Pliny, called Mount Atlas Dyris. This mountain chain 113is still called by the Arabs Daran, a word that is almost identical in its consonants with Dyris. Hornius,^[EP] on the other hand, thinks that he recognises the term Dyris in the word Ayadyrma, the name applied by the Guanches to the Peak of Teneriffe.^[EQ]

As our present geological knowledge of the mountainous parts of North Africa, which, however, must be admitted to be very limited, does not make us acquainted with any traces of volcanic eruptions within historic times, it seems the more remarkable that so many indications should be found in the writings of the Ancients of a belief in the existence of such phenomena in the Western Atlas and the contiguous west coast of the continent. The streams of fire so often mentioned in Hanno’s Ship’s Journal might indeed have been tracks of burning grass, or beacon fires lighted by the wild inhabitants of the coasts as a signal to warn each other of threatening danger on the first appearance of hostile vessels. The high summit of the “Chariot of the Gods,” of which Hanno speaks (the θεῶν ὄχημα), may also have had some faint reference to the Peak of Teneriffe; but farther on he describes a singular configuration of the land. He finds in the gulf, near the Western Horn, a large island, in which there is a salt lake, which again contains a smaller island. South of the Bay of the Gorilla Apes the same conformation is repeated. Does he refer to coral structures, lagoon islands (Atolls), and to volcanic crater lakes, in the middle of which a conical mountain has been upheaved? The Triton Lake was not in the neighbourhood of the lesser Syrtis, but on the western shores of the Atlantic.^[ER] The lake disappeared in an earthquake, which was attended with great fire-eruptions. Diodorus^[ES] says expressly πυρὸς εκφυτήματα μεγάλα. But the most wonderful configuration is ascribed to the hollow Atlas, in a passage hitherto but little noticed in one of the philosophical 114Dialexes of Maximus Tyrius, a Platonic philosopher who lived in Rome under Commodus. His Atlas is situated “on the continent where the Western Lybians inhabit a projecting peninsula.” The mountain has a deep semi-circular abyss on the side nearest the sea; and its declivities are so steep that they cannot be descended. The abyss is filled with trees, and “one looks down upon their summits and the fruits they bear as if one were looking into a well.”^[ET] The description is so minute and graphic that it no doubt sprung from the recollection of some actual view.

23. p. 9.—“The Mountains of the Moon, Djebel-al-Komr.”

The Mountains of the Moon described by Ptolemy,^[EU] σελήνης ὄρος, form on our older maps a vast uninterrupted mountain chain, traversing the whole of Africa from east to west. The existence of these mountains seems certain; but their extent, their distance from the equator, and their mean direction, still remain problematical. I have indicated in another work^[EV] the manner in which a more intimate acquaintance with Indian idioms and the ancient Persian or Zend teaches us that a part of the geographical nomenclature of Ptolemy constitutes an historical memorial of the commercial relations that existed between the West and the remotest regions of Southern Asia and Eastern Africa. The same direction of ideas is apparent in relation to a subject that has very recently become a matter of investigation. It is asked, whether the great geographer and astronomer of Pelusium merely meant in the denomination of Mountains of the Moon (as in that of “Island of Barley,” (Jabadiu, Java) to give the Greek translation of the native name of those mountains; whether, as is most probable, El-Istachri, Edrisi, Ibn-al-Vardi, and other early Arabian geographers, simply transferred the Ptolemaic nomenclature into their own language; or whether similarity in the sound of the word and the manner in which it was written misled them? In the notes to the translation of Abd-Allatif’s celebrated description of Egypt, my great teacher, Silvestre de Sacy,^[EW] expressly says, “The 115name of the mountains regarded by Leo Africanus as furnishing the sources of the Nile, has generally been rendered ‘Mountains of the Moon,’ and I have adhered to the same practice. I do not know whether the Arabs originally borrowed this denomination from Ptolemy. It may indeed be inferred that at the present day they understand the word قمر in the sense of moon, pronouncing it kamar; I do not think however, that such was the practice of the older Arabs, who pronounced it komr, as has been proved by Makrizi. Aboulfeda positively rejects the opinion of those who would adopt the pronunciation kamar, and derive the word from the name of the moon. As, according to the author of Kamous, the word komr, considered as the plural of اقمر, signifies an object of a greenish or dirty white colour, it would appear that some authors have supposed that this mountain derived its name from its colour.”

The learned Reinaud, in his recent excellent translation of Abulfeda (t. ii., p. i., pp. 81, 82), regards it as probable that the Ptolemaic interpretation of the name of Mountains of the Moon (ὄρη σεληναῖα) was that originally adopted by the Arabs. He observes that in the Moschtarek of Yakut, and in Ibn-Said, the mountain is written al-Komr, and that Yakut writes in a similar manner the name of the Island of Zendj (Zanguebar). The Abyssinian traveller Beke, in his learned and critical treatise on the Nile and its tributaries,^[EX] endeavours to prove that Ptolemy, in his σελήνης ὄρος, merely followed the native name, for the knowledge of which he was indebted to the extensive commercial intercourse which then existed. He says, “Ptolemy knew that the Nile rises in the mountainous district of Moezi, and in the languages which are spoken over a great part of Southern Africa (as, for instance, in Congo, Monjou, and Mozambique), the word moezi signifies the moon. A large tract of country situated in the south-west was called Mono-Muezi, or Mani-Moezi, i.e., the land of the King of Moezi (or Moon-land); for in the same family of languages in which moezi or muezi signifies the moon, mono or mani signifies a king. Alvarez^[EY] speaks 116of the ‘regno di Manicongo,’ or territory of the king of Congo.” Beke’s opponent, Ayrton, seeks the sources of the White Nile (Bahr el-Abiad), not as do Arnaud, Werne, and Beke, near the equator, or south of it (in 31° 22′ E. long. from Greenwich), but far to the north-east, as does Antoine d’Abbadie, in the Godjeb and Gibbe of Eneara (Iniara), therefore in the high mountains of Habesch, in 7° 20′ north lat., and 35° 22′ east long. from Greenwich. He is of opinion that the Arabs, from a similarity of sound, may have interpreted the native name Gamaro, which was applied to the Abyssinian mountains lying south-west of Gaka, and in which the Godjeb (or White Nile) takes its rise, to signify a mountain of the moon (Djebel al-Kamar); so that Ptolemy himself, who was familiar with the intercourse existing between Abyssinia and the Indian Ocean, may have adopted the Semitic interpretation, as given by the descendants of the early Arab immigrants.^[EZ]

The lively interest which has recently been felt in England for the discovery of the most southern sources of the Nile induced the Abyssinian traveller above referred to, (Charles Beke) at a recent meeting of the “British Association for the advancement of Science,” held at Swansea, more fully to develope his ideas respecting the connection between the Mountains of the Moon and those of Habesch. “The Abyssinian elevated plain,” he says, “generally above 8000 feet high, extends towards the south to nearly 9° or 10° north latitude. The eastern declivity of the highlands has, to the inhabitants of the coast, the appearance of a mountain chain. The plateau, which diminishes considerably in height towards its southern extremity, passes into the Mountains of the Moon, which run not east and west, but parallel to the coast, or from N.N.E. to S.S.W., extending from 10° north to 5° south latitude. The sources of the White Nile are situated in the Mono-Moezi country, probably in 2° 30′ south latitude, not far from where the river Sabaki, on the eastern side of the Mountains of the Moon, falls into the Indian Ocean, near Melindeh, north of Mombaza. Last autumn (1847), the two Abyssinian missionaries Rebmann and Dr. Krapf 117were still on the coast of Mombaza. They have established in the vicinity, among the Wakamba tribe, a missionary station, called Rabbay Empie, which seems likely to be very useful for geographical discoveries. Families of the Wakamba tribe have advanced westward five or six hundred miles into the interior of the country, as far as the upper course of the river Lusidji, the great lake Nyassi or Zambeze (5° south lat.?), and the vicinal sources of the Nile. The expedition to these sources, which Friedrich Bialloblotzky, of Hanover, is preparing to undertake” (by the advice of Beke), “is to start from Mombaza. The Nile coming from the west referred to by the ancients is probably the Bahr-el-Ghazal, or Keilah, which falls into the Nile in 9° north lat., above the mouth of the Godjeb or Sobat.”

Russegger’s scientific expedition—undertaken in 1837 and 1838, in consequence of Mehemet Ali’s eager desire to participate in the gold washings of Fazokl on the Blue (Green) Nile, Bahr el-Azrek—has rendered the existence of a Mountain of the Moon very doubtful. The Blue Nile, the Astapus of Ptolemy, rising from Lake Coloe (now called Lake Tzana), winds through the colossal Abyssinian range of mountains; while to the south-west there appears a far extended tract of low land. The three exploring expeditions which the Egyptian Government sent from Chartum to the confluence of the Blue and the White Nile (the first under the command of Selim Bimbaschi, in November, 1839; the next, which was attended by the French engineers Arnaud, Sabatier, and Thibaut, in the autumn of 1840; and the third, in the month of August, 1841), first removed some of the obscurity which had hitherto shrouded our knowledge of the high mountains, which between the parallels of 6°–4°, and probably still further southward, extend first from west to east, and subsequently from north-west to south-east, towards the left bank of the Bahr-el-Abiad. The second of Mehemet Ali’s expeditions first saw the mountain chain, according to Werne’s account, in 11° 20′ north lat., where Gebel Abul and Gebel Kutak rise to the height of 3623 feet. The high land continued to approach the river more to the south from 4° 45′ north lat. to the parallel of the Island of Tchenker in 4° 4′, near the point at which terminated the expedition commanded by Selim and Feizulla Effendi. The shallow river breaks its way through 118the rocks, and separate mountains again rise in the land of Bari to the height of more than 3200 feet. These are probably a part of the Mountains of the Moon, as they are given in our most recent maps, although they are not covered with perpetual snow, as asserted by Ptolemy.^[FA] The line of perpetual snow would assuredly not be found in these parallels of latitude below an elevation of nearly 15,500 feet above the sea’s level. It is not improbable that Ptolemy extended the knowledge he may have possessed of the high mountains of Habesch, near Upper Egypt and the Red Sea, to the country of the sources of the White Nile. In Godjam, Kaffa, Miecha, and Sami, the Abyssinian mountains rise from 10,000 to nearly 15,000 feet, as we learn from exact measurements; (not according to those of Bruce, who gives to Chartum an elevation of 5041 feet, instead of the true height, 1524 feet!) Rüppell, who ranks amongst the most accurate observers of the present day, found Abba Jarat (in 13° 10′ north lat.) only 70 feet below the elevation of Mont Blanc,^[FB] The same observer states that a plain, elevated 13,940 feet above the Red Sea, was barely covered with a thin layer of freshly fallen snow.^[FC] The celebrated inscription of Adulis, which, according to Niebuhr, is of somewhat later date than the age of Juba and Augustus, speaks of “Abyssinian snow that reaches to the knee,” and affords, I believe, the most ancient record in antiquity of snow within the tropics,^[FD] as the Paropanisus is 12° lat. north of that limit.

Zimmermann’s map of the district of the Upper Nile shows the dividing line where the basin of the great river terminates in the south-east, and which separates it from the domain of the rivers belonging to the Indian Ocean, viz.; from the Doara which empties itself north of Magadoxo; from the Teb on the amber coast of Ogda; from the Goschop whose abundant waters are derived from the confluence of the Gibu and the Zebi, and which must be distinguished from the Godjeb, rendered celebrated since 1839 by Antoine d’Abbadie, Beke, and the Missionary Krapf. In a letter to Carl Ritter I hailed with the most lively joy the appearance of the 119combined results of the recent travels of Beke, Krapf, Isenberg, Russegger, Rüppel, Abbadie, and Werne, as ably and comprehensively brought together in 1843 by Zimmermann. “If a prolonged span of life,” I wrote to him, “bring with it many inconveniences to the individual himself, and some to those about him, it yields a compensation in the mental enjoyment, afforded by comparing the earlier state of our knowledge with its more recent condition, and of seeing the growth and development of many branches of science that had long continued torpid, or whose actual fruits hypercriticism may even have attempted to set aside. This genial enjoyment has from time to time fallen to our lot in our geographical studies, and more especially in reference to those portions of which we could hitherto only speak with a certain timid hesitation. The internal configuration and articulation of a continent depends in its leading characters on several plastic relations which are usually among the latest to be elucidated. A new and excellent work of our friend, Carl Zimmermann, on the district of the Upper Nile and of the eastern portions of Central Africa, has made me more vividly sensible of these considerations. This new map indicates, in the clearest manner, by means of a special mode of shading, all that still remains unknown, and all that by the courage and perseverance of travellers of all nations (among which our own countrymen happily play an important part), has already been disclosed to us. We may regard it as alike important and useful that the actual condition of our knowledge, should, at different periods, be graphically represented by men well acquainted with the existing and often widely scattered materials of knowledge, and who not merely delineate and compile, but who know how to compare, select, and, where it is practicable, test the routes of travellers by astronomical determinations of place. Those who have contributed as much to the general stock of knowledge as you have done, have indeed an especial right to expect much, since their combinations have greatly increased the number of connecting points; yet I scarcely think that when, in the year 1822, you executed your great work on Africa, you could have anticipated so many additions as we have received.” It must be admitted that, in some cases, we have only acquired a knowledge of rivers, their direction, their branches, and 120their numerous synonymes according to various languages and dialects; but the courses of rivers indicate the configuration of the surface of the earth, and exert a threefold influence; they promote vegetation, facilitate general intercourse, and are pregnant with the future destiny of man.

The northern course of the White Nile, and the south-eastern course of the great Goschop, show that both rivers are separated by an elevation of the surface of the earth; although we are as yet but imperfectly acquainted with the manner in which such an elevation is connected with the highlands of Habesch, or how it may be prolonged in a southerly direction beyond the equator. Probably, and this is also the opinion of my friend Carl Ritter, the Lupata Mountains, which, according to the excellent Wilhelm Peters, extend to 26° south lat., are connected by means of the Mountains of the Moon with this northern swelling of the earth’s surface (the Abyssinian Highlands). Lupata, according to the last-named African traveller, signifies, in the language of Tette, closed, when used as an adjective. This mountain-range which is only intersected by some few rivers would thus be the closed or barred. “The Lupata chain of the Portuguese writers,” says Peters, “is situated about 90 leagues from the mouth of the Zambeze, and has an elevation of little more than 2000 feet. This mural chain has a direction due north and south, although it frequently deflects to the east or the west. It is sometimes interrupted by plains. Along the coast of Zanzibar the traders in the interior appear to be acquainted with this long, but not very high range, which extends between 6° and 26° south lat. to the Factory of Lourenzo-Marques on the Rio de Espirito Santo (in the Delagoa Bay of the English). The further the Lupata chain extends to the south, the nearer it approaches the coast, until at Lourenzo-Marques it is only 15 leagues distant from it.”

24. p. 10.—“The consequence of the great rotatory movement of the waters.”

The waters of the northern part of the Atlantic between Europe, Northern Africa, and the New Continent, are agitated by a continually recurring gyratory movement. Under the tropics the general current to which the term rotation-stream might appropriately be given in consideration of the cause 121from which it arises, moves, as is well known, like the trade wind from east to west. It accelerates the navigation of vessels sailing from the Canary Isles to South America; while it is nearly impossible to pursue a straight course against the current from Carthagena de Indias to Cumana. This bend to the west, attributed to the trade winds, is accelerated in the Caribbean Sea by a much stronger movement, which originates in a very remote cause, discovered as early as 1560 by Sir Humphrey Gilbert,^[FE] and confirmed in 1832 by Rennell. The Mozambique current, flowing from north to south between Madagascar and the eastern coast of Africa, sets on the Lagullas Bank, and bends to the north of it round the southern point of Africa. After advancing with much violence along the western coast of Africa beyond the equator to the island of St. Thomas, it gives a north-westerly direction to a portion of the waters of the South Atlantic, causing them to strike Cape St. Augustin, and follow the shores of Guiana beyond the mouth of the Orinoco, the Boca del Drago, and the coast of Paria.^[FF] The New Continent from the Isthmus of Panama to the northern part of Mexico forms a dam or barrier against the movements of the sea. Owing to this obstruction the current is necessarily deflected in a northerly direction at Veragua, and made to follow the sinuosities of the coast-line from Costa Rica, Mosquitos, Campeche, and Tabasco. The waters which enter the Mexican Gulf between Cape Catoche of Yucatan, and Cape San Antonio de Cuba, force their way back into the open ocean north of the Straits of Bahama, after they have been agitated by a great rotatory movement between Vera Cruz, Tamiagna, the mouth of the Rio Bravo del Norte, and the Mississippi. Here they form a warm, rapid current, known to mariners as the Gulf Stream, which deflects in a diagonal direction further and further from the shores of North America. Ships bound for this coast from Europe, and uncertain of their geographical longitude, are enabled by this oblique direction of the current to regulate their course as soon as they reach the Gulf Stream by observations of latitude only. The bearings of this current were first accurately determined by Franklin, Williams, and Pownall.

122From the parallel of 41° north lat. this stream of warm water follows an easterly direction, gradually diminishing in rapidity as it increases in breadth. It almost touches the southern edge of the Great Newfoundland Bank, where I found the greatest amount of difference between the temperature of the waters of the Gulf Stream and those exposed to the cooling action of the banks. Before the warm current reaches the Western Azores it separates into two branches, one of which turns at certain seasons of the year towards Ireland and Norway, while the other flows in the direction of the Canary Isles and the western coast of Northern Africa.

The course of this Atlantic current, which I have described more fully in the first volume of my travels in the regions of the tropics, affords an explanation of the manner in which, notwithstanding the action of the trade winds, stems of the South American and West Indian dicotyledons have been found on the coasts of the Canary Islands. I made many observations on the temperature of the Gulf Stream in the vicinity of the Newfoundland Bank. This current bears the warmer water of lower latitudes with great rapidity into more northern regions. The temperature of the stream is therefore from about 4°½ to 7° Fahr. higher than that of the contiguous and unmoved water which constitutes the shore as it were of the warm oceanic current.

The flying-fish of the equinoctial zone (Exocetus volitans), is borne by its predilection for the warmth of the water of the Gulf Stream far to the north of the temperate zone. Floating sea-weed (Fucus natans), chiefly taken up by the stream in the Mexican Gulf, makes it easy for the navigator to recognize when he has entered the Gulf Stream, whilst the position of the branches of the sea-weed indicate the direction of the current. The mainmast of the English ship of war, the Tilbury, which was destroyed by fire in the seven years’ war on the coasts of Saint Domingo, was carried by the Gulf Stream to the northern coasts of Scotland: and casks filled with palm-oil, the remains of the cargo of an English ship wrecked on a rock off Cape Lopez in Africa, were in like manner carried to Scotland, after having twice traversed the Atlantic Ocean, once from east to west between 2° and 12° north lat., following the course of the equinoctial current, and once from west to east between 45° and 55° north lat. by help of the Gulf Stream. Rennell, in 123the work already referred to, p. 347, relates the voyage of a bottle inclosing a written paper which had been thrown from the English ship Newcastle in 38° 52′ north lat., and 63° 58′ west long., on the 20th of January, 1819, and which was first seen on the 2nd of June, 1820, at the Rosses in the north-west of Ireland, near the Island of Arran. Shortly before my arrival at Teneriffe a stem of South American cedar-wood (Cedrela odorata), thickly covered with lichens, was cast ashore near the harbour of Santa Cruz.

The effects of the Gulf Stream in stranding on the Azorean Islands of Fayal, Flores, and Corvo, bamboos, artificially cut pieces of wood, trunks of an unknown species of pine from Mexico or the West Indies, and corpses of men of a peculiar race, having very broad faces, have mainly contributed to the discovery of America, as they confirmed Columbus in his belief of the existence of Asiatic countries and islands situated in the west. The great discoverer even heard from a settler on the Cap de la Verga in the Azores “that persons in sailing westward had met with covered barks, which were managed by men of foreign appearance, and appeared to be constructed in such a manner that they could not sink, almadias con casa movediza que nunca se hunden.” There are well authenticated proofs, however much the facts may have been called in question, that natives of America (probably Esquimaux from Greenland or Labrador), were carried by currents or streams from the north-west to our own continent. James Wallace^[FG] relates that in the year 1682 a Greenlander in his canoe was seen on the southern extremity of the Island of Eda by many persons, who could not, however, succeed in reaching him. In 1684 a Greenland fisherman appeared near the Island of Westram. In the church at Burra there was suspended an Esquimaux boat, which had been driven on shore by currents and storms. The inhabitants of the Orkneys call the Greenlanders who have appeared amongst them Finnmen.

In Cardinal Bembo’s History of Venice I find it stated, that in the year 1508 a small boat, manned by seven persons of a foreign aspect, was captured near the English coast by a French ship. The description given of them applies perfectly to the form of the Esquimaux (homines erant septem mediocri statura, colore subobscuro, lato et patente vultu, cicatriceque una 124violacea signato). No one understood their language. Their clothing was made of fish skins sewn together. On their heads they wore coronam e culmo pictam, septem quasi auriculis intextam. They ate raw flesh, and drank blood as we would wine. Six of these men perished during the voyage, and the seventh, a youth, was presented to the King of France, who was then at Orleans.^[FH]

The appearance of men called Indians on the coasts of Germany under the Othos and Frederic Barbarossa in the tenth and twelfth centuries, and as Cornelius Nepos (in his Fragments),^[FI] Pomponius Mela,^[FJ] and Pliny^[FK] relate, when Quintus Metellus Celer was Proconsul in Gaul, may be explained by similar effects of oceanic currents and by the long continuance of north-westerly winds. A king of the Boii, or, as others say, of the Suevi, gave these stranded dark-coloured men to Metellus Celer. Gomara^[FL] regards these Indian subjects of the King of the Boii as natives of Labrador. He writes, Si ya no fuesen de Tierra del Labrador, y los tuviesen los Romanos por Indianos engañados en el color. It may be inferred that the appearance of Esquimaux on the northern shores of Europe was more frequent in earlier times, for we learn from the investigations of Bask and Finn Magnusen, that this race had spread in the eleventh and twelfth century in considerable numbers, under the name of Skrälingers, from Labrador as far south as the Good Vinland, i.e. the shore of Massachussets and Connecticut.^[FM]

As the winter cold of the most northern part of Scandinavia is ameliorated by the action of the Gulf Stream, which carries American tropical fruits (as cocoa-nuts, seeds of Mimosa scandens and Anacardium occidentale) beyond 62° north lat.; so also Iceland enjoys from time to time the genial influence of the diffusion of the warm waters of the Gulf Stream far to the northward. The sea coasts of Iceland, like those of the Faroe Isles, receive a large number of trunks of 125trees, driven thither from America; and this drift-wood, which formerly came in greater abundance, was used for the purposes of building, and cut into boards and laths. The fruits of tropical plants collected on the Icelandic shores, especially between Raufarhaven and Vapnafiord, show that the movement of the water is from a southerly direction.^[FN]

25. p. 10.—“Lecideæ and other Lichens.”

In northern regions, the absence of plants is compensated for by the covering of Bœomyces roseus, Cenomyce rangiferinus, Lecidea muscorum, Lecidea icmadophila, and other cryptogamia which are spread over the earth, and which may be said to prepare the way for the growth of grasses and other herbaceous plants. In the tropical world, where mosses and lichens are only observed to abound in shady places, some few oily plants supply the place of the lowly lichen.

26. p. 11.—“The Care of Animals yielding milk.—Ruins of the Aztek fortress.”

The two oxen already named, Bos americanus and Bos moschatus, are peculiar to the northern part of the American continent. But the natives—

drank the fresh blood, and not the milk, of these animals. Some few exceptions have indeed been met with, but only among tribes who at the same time cultivated maize. I have already observed that Gomara speaks of a people in the north-west of Mexico who possessed herds of tame bisons, and derived their clothing, food, and drink from these animals. This drink was probably the blood,^[FO] for, as I have frequently remarked, a dislike of milk, or at least the absence of its use, appears before the arrival of Europeans to have been common to all the natives of the New Continent, as well as to the inhabitants of China and Cochin China, notwithstanding their great vicinity to true pastoral tribes. The herds of tame lamas which were found in the highlands of Quito, Peru, and Chili, belonged to a settled and agricultural 126population. Pedro de Cieça de Leon^[FP] seems to imply, although assuredly as a very rare exception to the general mode of life, that lamas were employed on the Peruvian mountain plain of Callao for drawing the plough.^[FQ] Ploughing was, however, generally conducted in Peru by men only.^[FR] Barton has made it appear probable that the American buffalo had from an early period been reared among some West Canada tribes on account of its flesh and hide.^[FS] In Peru and Quito the lama is nowhere found in its original wild condition. According to the statements made to me by the natives, the lamas on the western declivity of the Chimborazo became wild at the time when Lican, the ancient residence of the rulers of Quito, was laid in ashes. In Central Peru, in the Ceja de la Montaña, cattle have in like manner become completely wild; a small but daring race that often attacks the Indians. The natives call them “Vacas del Monte” or “Vacas Cimarronas.”^[FT] Cuvier’s assertion that the lama had descended from the guanaco, still in a wild state, which had unfortunately been extensively propagated by the admirable observer, Meyen,^[FU] has now been completely refuted by Tschudi.

The Lama, the Paco or Alpaca, and the Guanaco are three originally distinct species of animals.^[FV] The Guanaco (Huanacu in the Quichua language) is the largest of the three, and the Alpaca, measured from the ground to the crown of the head, the smallest. The Lama is next to the Guanaco in height. Herds of Lamas, when as numerous as I have seen them on the elevated plateaux between Quito and Riobamba, are a great ornament to the landscape. The Moromoro of Chili appears to be a mere variety of the lama. The different species of camel-like sheep found still wild at elevations of from 13,000 to upwards of 16,000 feet above the level of the sea, are the Vicuña, the Guanaco, and the Alpaca; of these the two latter species are also found tame, although this is but rarely the 127case with the Guanaco. The alpaca does not bear a warm climate as well as the lama. Since the introduction of the more useful horse, mule, and ass (the latter of which exhibits great animation and beauty in tropical regions), the lama and alpaca have been less generally reared and employed as beasts of burden in the mining districts. But their wool, which varies so much in fineness, is still an important branch of industry among the inhabitants of the mountains. In Chili the wild and the tame guanaco are distinguished by special names, the former being called “Luan” and the latter “Chilihueque.” The wide dissemination of the wild Guanacos from the Peruvian Cordilleras to Tierra del Fuego, sometimes in herds of 500 heads of cattle, has been facilitated by the circumstance that these animals can swim with great facility from island to island, and are not therefore impeded in their passage across the Patagonian channels or fiords.^[FW]

South of the river Gyla, which together with the Rio Colorado pours itself into the Californian Gulf (Mar de Cortes), lie in the midst of the dreary steppe the mysterious ruins of the Aztek Palace, called by the Spaniards “las Casas Grandes.” When, about the year 1160, the Azteks first appeared in Anahuac, having migrated from the unknown land of Aztlan, they remained for a time on the borders of the Gyla river. The Franciscan monks, Garces and Font, who saw the “Casas Grandes” in 1778, are the last travellers who have visited these remains. According to their statement, the ruins extended over an area exceeding sixteen square miles. The whole plain was covered with the broken fragments of ingeniously painted earthenware vessels. The principal palace, if the word can be applied to a house formed of unburnt clay, is 447 feet in length and 277 feet in breadth.^[FX]

The Tayé of California, a delineation of which is given by the Padre Venegas, appears to differ but inconsiderably from the Ovis musimon of the Old Continent. The same animal has also been seen in the Stony Mountains near the source of the River of Peace, and differs entirely from 128the small white and black spotted goat-like animal found on the Missouri and Arkansas. The synonyme of Antilope furcifer, A. tememazama. (Smith,) and Ovis montana is still very uncertain.

27. p. 11.—“The culture of farinaceous grasses.”

The original habitat of the farinaceous grasses, like that of the domestic animals which have followed man since his earliest migrations, is shrouded in obscurity. Jacob Grimm has ingeniously derived the German name for corn, Getraide, from the old German “gitragidi,” “getregede.” “It is as it were the tame fruit (fruges, frumentum) that has fallen into the hands of man, as we speak of tame animals in opposition to those that are wild.”^[FY]

“It is a most striking fact that on one half of our planet there should be nations who are wholly unacquainted with the use of milk and of the meal yielded by narrow-eared grasses, (Hordeaceæ and Avenacecæ) whilst in the other hemisphere nations may be found in almost every region who cultivate cereals and rear milch cattle. The culture of different cereals is common to both hemispheres; but while in the New Continent we meet with only one species, maize, which is cultivated from 52° north to 46° south lat., we find that in the Old World the fruits of Ceres, (wheat, barley, spelt, and oats,) have been everywhere cultivated from the earliest ages recorded in history. The belief that wheat grew wild in the Leontine plains as well as in other parts of Sicily was common to several ancient nations, and is mentioned as early as Diodorus Siculus.”^[FZ] Cereals were also found in the alpine meadow of Enna. Diodorus says expressly, “The inhabitants of the Atlantis were unacquainted with the fruits of Ceres, owing to their having separated from the rest of mankind before those fruits were made known to mortals.” Sprengel has collected several interesting facts from which he is led to conjecture that the greater number of our European cereals originally grew wild in Northern Persia and India. He supposes for instance that summer wheat was indigenous in the land of the Musicani, a province of Northern India;^[GA] barley, antiquissimum 129frumentum, as Pliny terms it, and which was also the only cereal known to the Guansches of the Canaries, originated, according to Moses of Chorene,^[GB] on the banks of the Araxes or Kur in Georgia, and according to Marco Polo in Balascham, in Northern India;^[GC] and Spelt originated in Hamadan.

My intelligent friend and teacher, Link, has however shown in a comprehensive and critical treatise,^[GD] that these passages are open to much doubt. In a former essay of my own,^[GE] I expressed doubts regarding the existence of wild cereals in Asia, and considered them to have become wild. Reinhold Forster, who before his voyage with Captain Cook made an expedition for purposes of natural history into the south of Russia by order of the Empress Catherine, reported that the two-lined summer barley (Hordeum distichon) grew wild near the confluence of the Samara and the Volga. At the end of September in the year 1829, Ehrenberg and myself also herborised on the Samara, during our journey from Orenburg and Uralsk to Saratow and the Caspian Sea. The quantity of wheat and rye plants growing wild on uncultivated ground in this district was certainly very remarkable; but the plants did not appear to us to differ from the ordinary kinds. Ehrenberg received from M. Carelin a species of rye, Secale fragile, that had been gathered on the Kirghis Steppe, and which Marshal Bieberstein for some time conjectured to be the mother plant of our cultivated rye, Secale cereale. Michaux’s herbarium does not show (according to Achill Richard’s testimony), that Spelt (Triticum spelta) grows wild at Hamadan in Persia, as Olivier and Michaux have been supposed to maintain. More confidence is due to the recent accounts obtained through the unwearied zeal of the intelligent traveller, Professor Carl Koch. He found a large quantity of rye (Secale cereale var. β, pectinata) in the Pontic Mountains, at heights of more than 5000 or 6000 feet above the level of the sea, on spots where this species of grain had not within the memory of the inhabitants been previously cultivated. “Its appearance here is the more important,” he remarks, “because with us this grain never propagates 130itself spontaneously.” Koch collected in the Schirwan part of the Caucasus a kind of grain which he calls Hordeum spontaneum, and regards as the originally wild Hordeum zeocriton. (Linn.)^[GF]

A negro slave of the great Cortes was the first who cultivated wheat in New Spain, from three seeds which he found amongst some rice brought from Spain for the use of the troops. In the Franciscan convent at Quito I saw, preserved as a relic, the earthen vessel which had contained the first wheat sowed in Quito by the Franciscan monk, Fray Jodoco Rixi de Gante, a native of Ghent in Flanders. The first crop was raised in front of the convent, on the “Plazuela de S. Francisco,” after the wood which then extended from the foot of the Volcano of Pichincha had been cleared. The monks, whom I frequently visited during my stay at Quito, begged me to explain the inscription on the cup, which according to their conjecture contained some hidden allusion to wheat. On examining the vessel, I read in old German the words “Let him who drinks from me, ne’er forget his God.” This old German drinking cup excited in me feelings of veneration! Would that everywhere in the New Continent the names of those were preserved who, instead of devastating the soil by bloody conquests, confided to it the first fruits of Ceres! There are “fewer examples of a general affinity of names in terms relating to the different species of corn and objects of agriculture than to the rearing of cattle. Herdsmen when they migrated to other regions had still much in common, while the subsequent cultivators of the soil had to invent special words. But the fact that in comparison with the Sanscrit, Romans and Greeks seem to stand on the same footing with Germans and Slavonians, speaks in favour of the very early contemporaneous emigration of the two latter. Yet the Indian java (frumentum hordeum), when compared with the Lithuanian jawai, and the Finnish jywa, affords a striking exception.”^[GG]

28. p. 11.—“Preferring to keep within a cooler climate.”

Throughout the whole of Mexico and Peru we find the trace of human civilisation confined to the elevated table-lands. 131We saw the ruins of palaces and baths on the sides of the Andes, at an elevation of from 10,230 to 11,510 feet. None but northern tribes migrating from the north towards the equator could have remained from preference in such a climate.

29. p. 12.—“The history of the peopling of Japan.”

I believe I have succeeded in showing, in my work on the monuments of the American primitive races,^[GH] by an examination of the Mexican and Thibetian-Japanese calendars, by a correct determination of the position of the Scansile Pyramids, and by the ancient myths which record four revolutions of the world and the dispersion of mankind after a great deluge, that the western nations of the New Continent maintained relations of intercourse with those of Eastern Asia, long before the arrival of the Spaniards. These observations have derived additional weight, since the appearance of my work, from the facts recently published in England, France, and the United States, regarding the remarkable pieces of sculpture carved in the Indian style, which have been discovered in the ruins of Guatimala and Yucatan.^[GI] The ancient architectural remains found in the peninsula of Yucatan testify more than those of Palenque, to an astonishing degree of civilization. They are situated between Valladolid, Merida, and Campeche, chiefly in the western portion of the country. But the monuments on the island of Cozumel, (properly Cuzamil,) east of Yucatan, were the first which were seen by the Spaniards in the expedition of Juan de Grijalva in 1518, and in that of Cortes in 1519. Their discovery tended to diffuse throughout Europe an exalted idea of the advanced condition of ancient 132Mexican civilization. The most important ruins of the peninsula of Yucatan (unfortunately not yet thoroughly measured and drawn by architects) are those of the “Casa del Gobernador” of Uxmal, the Teocallis and vaulted constructions at Kabah, the ruins of Labnan with its domed pillars, those of Zayi which exhibit columns of an order of architecture nearly approaching the Doric, and those of Chiche with large ornamented pilasters. An old manuscript written in the Maya language by a Christian Indian, which is still in the hands of the “Gefe politico” of Peto, Don Juan Rio Perez, gives the different epochs (Katunes of 52 years) at which the Toltecs settled in different parts of the peninsula. Perez would infer from these data that the architectural remains of Chiche go back as far as the fourth century of our era, whilst those of Uxmal belong to the middle of the tenth century; but the accuracy of these historical deductions is open to great doubt.^[GJ]

I regard the existence of a former intercourse between the people of Western America and these of Eastern Asia as more than probable, although it is impossible at the present time to say by what route and with which of the tribes of Asia this intercourse was established. A small number of individuals of the cultivated hierarchical castes may perhaps have sufficed to effect great changes in the social condition of Western America. The fabulous accounts formerly current regarding Chinese expeditions to the New Continent refer merely to expeditions to Fusang or Japan. It is, however, possible that Japanese and Sian-Pi may have been driven by storms from the Corea to the American coasts. We know as matters of history that Bonzes and other adventurers navigated the Eastern Chinese seas in search of a remedial agent capable of making man immortal. Thus under Tschin-chi-huang-ti three hundred young couples were dispatched to Japan in the year 209 before our era, who, instead of returning to China, settled on the Island of Nipon.^[GK] May not accident have led to similar expeditions to 133the Fox Islands, to Alaschka, or New California? As the western coasts of the American continent incline from north-west to south-east, and the eastern coasts of Asia from north-east to south-west, the distance between the two continents in the milder zone, which is most conducive to mental development (45° lat.), would appear too considerable to admit of an accidental settlement having been made in this latitude. We must therefore assume that the first landing took place in the ungenial climate of 55° and 65°, and that cultivation, like the general advance of population in America, progressed by gradual stations from north to south.^[GL] It was even believed in the beginning of the sixteenth century that the fragments of ships from Catayo, i.e. from Japan or China, had been found on the coasts of the Northern Dorado, called also Quivira and Cibora.^[GM]

We know as yet too little of the languages of America entirely to renounce the hope that, amid their many varieties, some idiom may be discovered, that has been spoken with certain modifications in the interior of South America and Central Asia, or that might at least indicate an ancient affinity. Such a discovery would undoubtedly be one of the most brilliant to which the history of the human race can hope to attain! But analogies of language are only deserving of confidence where mere resemblances of sound in the roots are not alone the object of research, but attention is also directed to the organic structure, the grammatical forms, and those elements of language which manifest themselves as the product of the intellectual power of man.

30. p. 12—“Many other forms of animal life.”

The Steppes of Caracas abound in flocks of the so-called Cervus mexicanus. This stag when young is spotted, and resembles the roe. We have frequently met with perfectly white varieties, which is a very striking fact when the high temperature of this zone is taken into consideration. The Cervus mexicanus is not found on the declivities of the Andes in the equatorial region, at an elevation exceeding from 4476 to 5115 feet, but another white deer, which I could scarcely distinguish by any one specific characteristic from the 134European species, ascends to an elevation of nearly 13,000 feet. The Cavia capybara is known in the province of Caracas by the name of Chiguire. This unfortunate animal is pursued in the water by the crocodile, and on land by the tiger or jaguar. It runs so badly that we were often able to catch it with our hands. The extremities are smoked and eaten as hams, but have a most unpleasant taste, owing to the flavour and smell of musk by which they are impregnated; and on the Orinoco we gladly ate monkey-hams in preference. These beautifully striped animals—the Viverra mapurito, Viverra zorilla, and Viverra vittata—exhale a fetid odour.

31. p. 12—“The Guaranes and the fan-palm Mauritia.”

The small coast tribe of the Guaranes (called in British Guiana, the Warraws, or Guaranos, and by the Caribs U-ara-u) inhabit not only the swampy delta and the river net-work of the Orinoco (more particularly the banks of the Manamo grande and the Caño Macareo), but also extend, with very slight differences in their mode of living, along the sea-shore, between the mouths of the Essequibo and the Boca de Navios of the Orinoco.^[GN] According to the testimony of Schomburgk, the admirable observer referred to in the note, there are still about 1700 Warraus or Guaranos living in the vicinity of Cumaca, and along the banks of the Barime river, which empties itself into the gulf of the Boca de Navios. The social habits of the tribes settled in the delta of the Orinoco were known to the great historian Cardinal Bembo, the cotemporary of Christopher Columbus, Amerigo Vespucci, and Alonzo de Hojeda. He says^[GO] quibusdam in locis propter paludes incolæ domus in arboribus œdificant. It is hardly probable that instead of the Guaranos at the mouth of the Orinoco, Bembo should here allude to the natives of the country near the mouth of the gulf of Maracaibo, where Alonzo de Hojeda, in August, 1499, (when accompanied by Vespucci and Juan de la Cosa) found a population having their dwellings fondata sopra l’acqua come Venezia (“built like Venice on the water”).^[GP] Vespucci, in the account of his 135travels, in which we meet with the first traces of the etymology of the name of the province of Venezuela (Little Venice) as used for the province of Caracas, speaks only of houses built on a foundation of piles, and makes no mention of habitations in trees.

Sir Walter Raleigh bears a subsequent and incontrovertible evidence to the same fact, for he says expressly in his description of Guiana, that on his second voyage in 1595, when in the mouth of the Orinoco, “he saw the fire of the Tivitites and Qua-rawetes” (so he calls the Guaranes), “high up in the trees.”^[GQ] There is a drawing of the fire in the Latin edition of this work,^[GR] and Raleigh was the first who brought to England the fruit of the Mauritia palm, which he very justly compared, on account of its scales, to fir-cones. Father José Gumilla, who twice visited the Guaranes as a missionary, says, indeed, that this tribe have their dwelling in the Palmares (palm groves) of the morasses; but while he speaks more definitely of pendent habitations supported by high pillars, makes no mention of platforms attached to still growing trees.^[GS] Hillhouse and Sir Robert Schomburgk^[GT] are of opinion that Bembo, through the relations of others, and Raleigh, by his own observation, were deceived into this belief in consequence of the high tops of the palm trees being lighted up in such a manner by the fires below them, that those sailing by thought the habitations of the Guaranes were attached to the trees themselves. “We do not deny,” says Schomburgk, “that in order to escape the attacks of the mosquitos, the Indian sometimes suspends his hammock from the tops of trees, but on such occasions no fires are made under the hammock.”^[GU]

According to Martius, the beautiful Palm, Moriche, Mauritia 136flexuosa, Quieteva, or Ita Palm,^[GV] belongs, together with Calamus, to the family of the Lepidocaryæ or Corypheæ. Linnæus has described it very imperfectly, as he erroneously considered it to be devoid of leaves. The trunk is 26 feet high, but it probably does not attain this height in less than 120 or even 150 years. The Mauritia extends high up the declivity of the Duida, north of the Esmeralda mission, where I found it in great beauty. It forms, in moist places, fine groups of a fresh and shining verdure, reminding us of that of our alders. The trees preserve the moisture of the ground by their shade, and hence the Indians believe that the Mauritia draws water around its roots by some mysterious attraction. In conformity with an analogous theory they advise, that serpents should not be killed, because the destruction of these animals is followed by the drying up of the lagoons. Thus do the rude children of nature confound cause and effect! Gumilla calls the Mauritia flexuosa of the Guaranes the tree of life (“arbol de la vida”). It is found on the mountains of Ronaima, east of the sources of the Orinoco, as high as 4263 feet. On the unfrequented banks of the Rio Atabapo, in the interior of Guiana, we discovered a new species of Mauritia having a prickly stem; our Mauritia aculeata.^[GW]

32. p. 13.—“An American Stylite.”

The founder of the sect of Stylites, the fanatical Pillar-saint, Simeon Sisanites of Syria, the son of a Syrian herdsman, is said to have passed thirty-seven years in holy contemplation, elevated on five columns, each higher than the preceding. He died in the year 461. The last of the pillars which he occupied was 40 ells in height. For seven hundred years there continued to be followers of this mode of life, who were called Sancti Columnares, or Pillar-saints. Even in Germany, in the see of Treves, attempts were made to found similar aërial cloisters; but the dangerous practice met with the constant opposition of the bishops.^[GX]

33. 137p. 14.—“Towns on the banks of the Steppe-rivers.”

Families who live by raising cattle and do not take part in agricultural pursuits have congregated together in the middle of the Steppe, in small towns, which, in the cultivated parts of Europe, would scarcely be regarded as villages. Among these are Calabozo, which, according to my astronomical observations, is situated in 8° 56′ 14″ north lat., and 67° 43′ west long.; Villa del Pao (8° 38′ 1″ north lat., and 66° 57′ west long.); Saint Sebastian, and others.

34. p. 14.—“Funnel-shaped clouds.”

The singular phenomenon of these sand-spouts, of which we see something analogous on the cross roads of Europe, is especially characteristic of the Peruvian sandy desert between Amotape and Coquimbo. Such dense clouds of sand may endanger the safety of the traveller who does not cautiously avoid them. It is remarkable that these partial and opposing currents of air should arise only when there is a general calm. The aërial ocean resembles the sea in this respect; for here, too, we find that the small currents (filets de courant) in which the water may frequently be heard to flow with a splashing sound, occur only in a dead calm (calme plat).

35. p. 14.—“Increases the stifling oppression.”

I have observed in the Llanos de Apure, at the cattle farm of Guadalupe, that the thermometer rose from 92°.7 to 97°.2 Fahr. whenever the hot wind began to blow from the desert, which was covered either with sand or short withered grass. In the middle of the sand-cloud the thermometer stood for several minutes together at 111° Fahr. The dry sand in the village of San Fernando de Apure had a temperature of 126° Fahr.

36. p. 15.—“The phantom of a moving undulating surface.”

The well known phenomenon of the mirage is called in Sanscrit “the thirst of the gazelle.”^[GY] All objects appear to float in the air, while their forms are reflected in the lower stratum of the atmosphere. At such times the whole desert 138resembles a vast lake, whose surface undulates like waves. Palm trees, cattle, and camels sometimes appear inverted in the horizon. In the French expedition to Egypt, this optical illusion often nearly drove the faint and parched soldiers to distraction. This phenomenon has been observed in all quarters of the world. The ancients were also acquainted with the remarkable refraction of the rays of light in the Lybian Desert. We find mention made in Diodorus Siculus of strange illusive appearances, an African Fata Morgana, together with still more extravagant explanations of the conglomeration of the particles of air.^[GZ]

37. p. 15.—“The Melocactus.”

The Cactus melocactus is frequently from 10 to 12 inches in diameter, and has generally 14 ribs. The natural group of the Cactaceæ, the whole family of the Nopaleæ of Jussieu, belongs exclusively to the New Continent. The Cactus assumes a variety of shapes, being ribbed and melon-like (Melocacti); articulated (Opuntiæ); upright-like columns (Cerei); of a serpentine or creeping form (Rhipsalides); or provided with leaves (Pereskiæ). Many extend high up the slopes of the mountains. Near the foot of the Chimborazo, in the sandy table-land around Riobamba, I found a new species of Pitahaya (Cactus sepium), even at an elevation of 10,660 feet.^[HA]

38. p. 16.—“The scene suddenly changes in the Steppe.”

I have endeavoured to describe the approach of the rainy season, and the signs by which it is announced. The deep blue of the heavens in the tropics is occasioned by the imperfect solution of vapour. The cyanometer indicates a lighter shade of blue as soon as the vapours begin to fall. The dark spot in the constellation of the Southern Cross becomes indistinct in proportion as the transparency of the atmosphere decreases, and this change announces the approach of rain. The bright radiance of the Magellanic clouds (Nubecula major and Nubecula minor) then gradually fades away. The fixed stars which had before been shining with a calm, steady, planetlike 139light, are now seen to scintillate in the zenith.^[HB] All these phenomena are the result of the increased quantity of aqueous vapour floating in the atmosphere.

39. p. 16.—“The humid clay soil is seen to rise slowly in a broad flake.”

Drought produces the same phenomena in animals and plants as the abstraction of heat. During the dry season many tropical plants lose their leaves. The crocodile and other amphibious animals conceal themselves in the mud and lie apparently dead, like animals in cold regions who are thrown into a state of hybernation.^[HC]

40. p. 17.—“A vast inland sea.”

Nowhere are these inundations on a larger scale than in the net-work of streams formed by the Apure, the Arachuna, the Payara, the Arauca, and the Cabuliare. Large vessels sail across the country over the Steppe for 40 or 50 miles.

41. p. 17.—“To the mountainous plain of Antisana.”

The great mountain plateau which surrounds the volcano of Antisana is 13,473 feet above the level of the sea. The pressure of the atmosphere is so inconsiderable at this height, that blood will flow from the nostrils and mouth of the wild bull when hunted with dogs.

42. p. 17.—“The marshy waters of Bera and Rastro.”

I have elsewhere more circumstantially described the capture of the gymnotus.^[HD] Mons. Gay Lussac and myself were perfectly successful in the experiments we conducted without a chain on a living gymnotus, which was still very vigorous when it reached Paris. The discharge of electricity is entirely dependent on the will of the animal. We did not observe any electric sparks, but other physicists have done so on numerous occasions.

43. p. 18.—“Awakened by the contact of moist and dissimilar particles.”

In all organic bodies dissimilar substances come into 140contact with each other, and solids are associated with fluids. Wherever there is organization and life, there must be electric tension, or, in other words, a voltaic pile must be brought into play, as the experiments of Nobili and Matteucci, and more especially the late most admirable labours of Emil Dubois, teach us. The last-named physicist has succeeded in “manifesting the presence of the electric muscular current in living and wholly uninjured animal bodies:” he shows that “the human body, through the medium of a copper wire, can at will cause the magnetic needle at a distance to deflect first in one direction and then in another.”^[HE] I have myself witnessed these movements produced at will, and have thus unexpectedly seen much light thrown on phenomena, to which I had laboriously and ardently devoted so many years of my earlier life.

44. p. 19.—“The myth of Osiris and Typhon.”

Respecting the struggle of two human races, the Arabian shepherd tribes of Lower Egypt and the cultivated agricultural races of Upper Egypt; on the subject of the fair-haired Prince Baby or Typhon, who founded Pelusium; and on the dark-complexioned Dionysos or Osiris; I would refer to Zoëga’s older and almost universally discarded views as set forth at p. 577 of his masterly work “De origine et usu obeliscorum.”

45. p. 19—“The boundaries of European semi-civilization.”

In the Capitania General de Caracas, as well as in all the eastern part of America, the civilization formerly introduced by Europeans is limited to the narrow strip of land which skirts the shore. In Mexico, New Granada, and Quito on the other hand, European civilization has penetrated far into the interior of the country and advanced up to the ridges of the Cordilleras. There existed already in the fifteenth century an earlier stage of civilization among the inhabitants of the last-named region. Wherever the Spaniards perceived this culture they pursued its track, regardless whether the seat of it was at a distance from the sea, or in its vicinity. The ancient cities were enlarged and their former significant 141Indian names mutilated, or exchanged for those of Christian saints.

46. p. 19—“Huge masses of leaden-coloured granite.”

In the Orinoco, and more especially at the cataracts of Maypures and Atures (not in the Black River or Rio Negro), all blocks of granite, even pieces of white quartz, wherever they come in contact with the water, acquire a grayish black coating, which does not penetrate beyond 0·01 of a line into the interior of the rock. The traveller might almost suppose that he was looking at basalt, or fossils coloured with graphite. Indeed, the crust does actually appear to contain manganese and carbon. I say “appears” to do so, because the phenomenon has not yet been thoroughly investigated. Something perfectly analogous to this was observed by Rozier in the syenitic rocks of the Nile (near Syene and Philæ); by the unfortunate Captain Tuckey on the rocky banks of the Zaire; and by Sir Robert Schomburgk at Berbice.^[HF] On the Orinoco these leaden-coloured rocks are supposed when wet to give forth noxious exhalations, and their vicinity is believed to be conducive to the generation of fevers.^[HG] It is also remarkable that the South American rivers generally, which have black waters (aguas negras), or waters of a coffee brown or wine yellow tint, do not darken the granite rocks; that is to say, they do not act upon the stone in such a manner as to form from its constituent parts a black or leaden-coloured crust.

47. p. 20—“The rain-foreboding howl of the bearded ape.”

Some hours before the commencement of rain, the melancholy cries of various apes, as Simia seniculus, Simia beelzebub, &c., fall on the ear like a storm raging in the distance. The intensity of the noise produced by such small animals can only be explained by the circumstance that one tree often contains a herd of seventy or eighty apes. I have elsewhere spoken of the laryngeal sac, and the ossification of the larynx of these animals.^[HH]

48. 142p. 20—“Its uncouth body often covered with birds.”

The crocodiles lie so motionless, that I have often seen flamingoes (Phœnicopterus) resting on their heads, while the other parts of the body were covered, like the trunk of a tree, with aquatic birds.

49. p. 20—“Down its dilating throat.”

The saliva with which the boa covers its prey tends to promote rapid decomposition. The muscular flesh is rendered gelatinously soft under its action, so that the animal is able to force entire limbs of its slain victim through its swelling throat. The Creoles call the giant boa Tragavenado (stag-swallower), and fabulously relate that the antlers of a stag which could not be swallowed by the snake have been seen fixed in its throat. I have frequently observed the boa constrictor swimming in the Orinoco, and in the smaller forest streams, the Tuamini, the Temi, and the Atabapo. It holds its head above water like a dog. Its skin is beautifully speckled. It has been asserted, that the animal attains a length of 48 feet, but the longest skins which have as yet been carefully measured in Europe do not exceed from 21 to 23 feet. The South American boa (a Python) differs from the East Indian.^[HI]

50. p. 20—“Living on gums and earth.”

It is currently reported throughout the coasts of Cumana, New Barcelona, and Caracas (which the Franciscan monks of Guiana are in the habit of visiting on their return from the missions,) that there are men living on the banks of the Orinoco who eat earth. On the 6th of June, 1800, on our return from the Rio Negro, when we descended the Orinoco in thirty-six days, we spent the day at the mission inhabited by these people (the Otomacs). Their little village, which is called La Concepcion de Uruana, is very picturesquely built against a granite rock. It is situated in 7° 8′ 3″ north lat.; and according to my chronometrical determination, in 67° 18′ west longitude. The earth which the Otomacs eat, is an unctuous, almost tasteless clay, true potter’s earth, of a yellowish grey 143colour, in consequence of a slight admixture of oxide of iron. They select it with great care, and seek it in certain banks on the shores of the Orinoco and Meta. They distinguish the flavour of one kind of earth from that of another; all kinds of clay not being alike acceptable to their palate. They knead this earth into balls measuring from four to six inches in diameter, and bake them before a slow fire, until the outer surface assumes a reddish colour. Before they are eaten, the balls are again moistened. These Indians are mostly wild, uncivilized men, who abhor all tillage. There is a proverb current among the most distant of the tribes living on the Orinoco, when they wish to speak of anything very unclean, “so dirty that the Otomacs eat it.”

As long as the waters of the Orinoco and the Meta are low, these people live on fish and turtles. They kill the former with arrows, shooting the fish as they rise to the surface of the water with a skill and dexterity that has frequently excited my admiration. At the periodical swelling of the rivers, the fishing is stopped, for it is as difficult to fish in deep river water as in the deep sea. It is during these intervals, which last from two to three months, that the Otomacs are observed to devour an enormous quantity of earth. We found in their huts considerable stores of these clay balls piled up in pyramidal heaps. An Indian will consume from three-quarters of a pound to a pound and a quarter of this food daily, as we were assured by the intelligent monk, Fray Ramon Bueno, a native of Madrid, who had lived among these Indians for a period of twelve years. According to the testimony of the Otomacs themselves, this earth constitutes their main support in the rainy season. In addition, they however eat, when they can procure them, lizards, several species of small fish, and the roots of a fern. But they are so partial to clay, that even in the dry season, when there is an abundance of fish, they still partake of some of their earth-balls, by way of a bonne bouche after their regular meals.

These people are of a dark, copper-brown colour, have unpleasant Tartar-like features, and are stout, but not protuberant. The Franciscan who had lived amongst them as a missionary, assured us that he had observed no difference in the condition and well-being of the Otomacs during the periods in which they lived on earth. The simple facts are therefore 144as follows:—The Indians undoubtedly consume large quantities of clay without injuring their health; they regard this earth as a nutritious article of food, that is to say, they feel that it will satisfy their hunger for a long time. This property they ascribe exclusively to the clay, and not to the other articles of food which they contrive to procure from time to time in addition to it. If an Otomac be asked what are his winter provisions—the term winter in the torrid parts of South America implying the rainy season—he will point to the heaps of clay in his hut. These simple facts do not, however, by any means decide the questions: whether clay can actually be a nutritious substance; whether earths can be assimilated in the human body; whether they only serve as ballast; or merely distend the walls of the stomach, and thus appease the cravings of hunger? These are questions which I cannot venture to decide.^[HJ] It is singular, that Father Gumilla, who is generally so credulous and uncritical, should have denied the fact of earth being eaten by and for itself.^[HK] He maintains that the clay-balls are largely mixed with maize-flour, and crocodile’s fat. But the missionary Fray Ramon Bueno, and our friend and fellow-traveller, the lay-brother Fray Juan Gonzales, who perished at sea off the coast of Africa (at the time we lost a portion of our collections), both assured us, that the Otomacs never mix their clay cakes with crocodile’s fat, and we heard nothing in Uruana of the admixture of flour.

The earth which we brought with us, and which was chemically investigated by M. Vauquelin, is quite pure and unmixed. May not Gumilla, by confounding heterogeneous facts, have intended to allude to a preparation of bread from the long pod of a species of Inga? as this fruit is certainly buried in the earth, in order to hasten its decomposition. It appears to me especially remarkable, that the Otomacs should not lose their health by eating so much earth. Has this tribe been habituated for generations to this stimulus?

In all tropical countries men exhibit a wonderful and almost irresistible desire to devour earth, not the so-called alkaline or calcareous earth, for the purpose of neutralizing acidity, but unctuous, strong-smelling clay. It is often found 145necessary to shut children up in order to prevent their running into the open air to devour earth after recent rain. The Indian women who are engaged on the river Magdalena, in the small village of Banco, in turning earthenware pots, continually fill their mouths with large lumps of clay, as I have frequently observed, much to my surprise.^[HL] Wolves eat earth, especially clay, during winter. It would be very important, in a physiological point of view, to examine the excrements of animals and men that eat earth. Individuals of all other tribes, excepting the Otomacs, lose their health if they yield to this singular propensity for eating clay. In the mission of San Borja we found the child of an Indian woman, which, according to the statement of its mother, would hardly eat anything but earth. It was, however, much emaciated, and looked like a mere skeleton.

Why is it that in the temperate and cold zones this morbid eagerness for eating earth is so much less frequently manifested, and is indeed limited almost entirely to children and pregnant women, whilst it would appear to be indigenous to the tropical lands of every quarter of the earth? In Guinea the negroes eat a yellowish earth, which they call caouac; and when they are carried as slaves to the West Indies they even endeavour there to procure for themselves some similar species of food, maintaining that the eating of earth is perfectly harmless in their African home. The caouac of the American islands, however, deranges the health of the slaves who partake of it; for which reason the eating of earth was long since forbidden in the West Indies, notwithstanding which a species of red or yellowish tuff (un tuf rouge jaunâtre) was secretly sold in the public market of Martinique in the year 1751.

“The negroes of Guinea say that in their own country they habitually eat a certain earth, the flavour of which is most agreeable to them, and which does not occasion them any inconvenience. Those who have addicted themselves to the excessive use of caouac are so partial to it, that no punishment can prevent them from devouring this earth.”^[HM] In the island of Java, between Sourabaya and Samarang, Labillardière saw 146small square reddish cakes publicly sold in the villages. The natives called them tana ampo (tanah signifies earth in Malay and Javanese); and on examining them more closely, he found that they were cakes made of a reddish clay, and intended for eating.^[HN] The edible clay of Samarang has recently (1847) beep sent, by Mohnike, to Berlin in the shape of rolled tubes like cinnamon, and has been examined by Ehrenberg. It is a fresh-water formation deposited in tertiary limestone, and composed of microscopic polygastrica (Gallionella, Navicula) and of Phytolitharia.^[HO] The natives of New Caledonia, to appease their hunger, eat lumps as large as the fist of friable steatite, in which Vauquelin detected an appreciable quantity of copper.^[HP] In Popayan and many parts of Peru calcareous earth is sold in the streets as an article of food for the Indians. This is eaten together with the Coca (the leaves of the Erythroxylon peruvianum). We thus find that the practice of eating earth is common throughout the whole of the torrid zone among the indolent races who inhabit the most beautiful and fruitful regions of the earth. But accounts have also come from the north, through Berzelius and Retzius, from which we learn, that in the most remote parts of Sweden hundreds of cartloads of earth containing infusoria are annually consumed by the country people as bread-meal, more from fancy (like the smoking of tobacco) than from necessity. In some parts of Finland a similar kind of earth is mixed with the bread. It consists of empty shells of animalcules, so small and soft, that they break between the teeth without any perceptible noise, filling the stomach without yielding any actual nourishment. Chronicles and archives often make mention during times of war of the employment as food of infusorial earth, which is spoken of under the indefinite and general term of “mountain meal.” Such, for instance, was the case in the Thirty Years’ War, at Camin in Pomerania, Muskau in the Lausitz, and Kleiken in the Dessau territory; and subsequently in 1719 and 1733, at the fortress of Wittenberg.^[HQ]

51. 147p. 20.—“Images graven in rocks.”

In the interior of South America, between the parallels of 2° and 4° north lat., lies a wooded plain inclosed by four rivers, the Orinoco, the Atabapo, the Rio Negro, and the Cassiquiare. Here we find granitic and syenitic rocks, which, like those of Caicara and Uruana, are covered with colossal symbolical figures of crocodiles, tigers, utensils of domestic use, signs of the sun and moon, &c. This remote portion of the earth is at present wholly uninhabited throughout an extent of more than 8000 square miles. The neighbouring tribes, who occupy the lowest place in the scale of humanity, are naked wandering savages, who could not possibly have carved hieroglyphics in stone. A whole range of these rocks covered with symbolical signs may be traced from Rupunuri, Essequibo, and the mountains of Pacaraima, to the banks of the Orinoco and of the Yupura, extending over more than eight degrees of longitude.

These carvings may belong to very different periods of time, for Sir Robert Schomburgk even found on the Rio Negro representations of a Spanish galliot,^[HR] which must necessarily have been of a date subsequent to the beginning of the sixteenth century, and that in a wilderness where the inhabitants were probably as rude then as they now are. But it must not be forgotten, as I have already elsewhere observed, that nations of very different descent, but in similarly uncivilized conditions, possessed of the same disposition to simplify and generalize outlines, and urged by identical inherent mental tendencies, may be led to produce similar signs and symbols.^[HS]

At the meeting of the Society of Antiquaries in London a memoir was read on the 17th of November, 1836, by Sir Robert Schomburgk, “On the religious traditions of the Macusi Indians, who inhabit the Upper Mahu, and a portion of the Pacaraima mountains,” and who have therefore not changed their habitation for a century (since the journey of the intrepid Hortsmann). “The Macusis,” says Sir Robert Schomburgk, “believe that the only being who survived a 148general deluge, repeopled the earth by converting stones into human beings.” This myth, which is the fruit of the lively imagination of these tribes, and which reminds us of that of Deucalion and Pyrrha, shows itself in a somewhat modified form among the Tamanacs of the Orinoco. When these people are asked how the human race survived this great flood, the age of waters of the Mexicans, they unhesitatingly reply, “that one man and one woman were saved by taking refuge on the summit of the lofty mountain of Tamanacu, on the banks of the Asiveru, and that they then threw over their heads the fruits of the Mauritia palm, from the kernels of which sprang men and women, who again peopled the earth.” Some miles from Encaramada there rises in the midst of the savannah the rock of Tepu-Mereme; i.e., the “painted rock,” which exhibits numerous figures of animals and symbolical signs, having much resemblance to those which we observed at some distance above Encaramada, near Caycara. (7° 5′ to 7° 40′ north lat., and 66° 28′ to 67° 23′ west long.) Similarly carved rocks are found between the Cassiquiare and the Atabapo (2° 5′ to 3° 20′ lat.); and what is most striking, also 560 miles further eastward in the solitudes of the Parime. The last-named fact is proved beyond a doubt, by the journal of Nicolas Hortsmann of Hildesheim, of which I have seen a copy in the handwriting of the celebrated d’Anville. That simple and modest traveller wrote down every day on the spot whatever had struck him as worthy of notice; and his narrative deserves perhaps the more confidence from the fact that the great disappointment he experienced in having failed in the object of his researches, which was the discovery of the Lake of Dorado, with its lumps of gold and a diamond mine (which proved to be merely rock crystal of a very pure kind), led him to look with a certain degree of contempt on all that fell in his way. On the bank of the Rupunuri, at the point where the river, winding between the Macarana mountains, forms several small cascades; and before reaching the country immediately surrounding the Lake of Amucu, he found, on the 16th of April, 1749, “rocks covered with figures,” or, as he says in Portuguese, “de varias letras” (with various letters or characters). We were shown, at the rock of Culimacari, on the banks of the Cassiquiare, signs said to be characters drawn by line and rule: 149but they were merely ill-formed figures of the heavenly bodies, crocodiles, boa-constrictors, and utensils used in the preparation of manioc-meal. I found among these painted rocks (piedras pintadas) neither a symmetrical arrangement nor any trace of characters drawn with a regard to regularity in space and size. The word “letras” in the journal of the German Surgeon (Hortsmann) must not, therefore, I am disposed to think, be taken in the strictest sense.

Schomburgk did not succeed in finding the rocks observed by Hortsmann, but he has described others which he saw on the bank of the Essequibo, near the cascade of Waraputa. “This cascade,” he says, “is celebrated not only for its height, but also for the great number of figures hewn in the rock, which bear a great resemblance to those that I have seen on the island of St. John, (one of the Virgin Islands,) and which I consider to be without doubt the work of the Caribs, by whom this part of the Antilles was peopled in former times. I made the most strenuous efforts to hew away a portion of the rock carved with inscriptions, which I was desirous of taking with me; but the stone was too hard, and my strength had been wasted by fever. Neither threats nor promises could prevail on the Indians to aim a single stroke of the hammer against these rocks—the venerable monuments of the culture and superior skill of their forefathers. They regard them as the work of the Great Spirit; and all the different tribes we met were acquainted with them, although living at a great distance. Terror was painted on the faces of my Indian companions who seemed to expect every moment that the fire of heaven would fall on my head. I now saw clearly that all my efforts were fruitless, and I was therefore obliged to content myself with bringing away a complete drawing of these monuments.”

The last resolution was undoubtedly the best, and the editor of the English journal, to my great satisfaction, subjoins in a note the remark, “that it is to be wished that others may succeed no better than Schomburgk, and that no traveller belonging to a civilized nation will in future attempt the destruction of these monuments of the unprotected Indians.”

The symbolical signs which Sir Robert Schomburgk found in the fluvial valley of the Essequibo, near the rapids of Waraputa,^[HT] resemble, indeed, according to his observation, 150the genuine Carib carvings of one of the smaller Virgin Islands (St. John); but notwithstanding the wide extent of the Carib invasions, and the ancient power of that fine race, I cannot believe that this vast belt of carved rocks which intersects a great portion of South America from west to east, is actually to be ascribed to the Caribs. These remains seem rather to be traces of an ancient civilization, which may have belonged to an epoch when the tribes, whom we now distinguish by various names and races, were still unknown. The veneration which is everywhere shown by the Indians for these rude carvings of their predecessors, proves that the present races have no idea of the execution of similar works. Nay, more than this, between Encaramada and Caycara, on the banks of the Orinoco, many of these hieroglyphic figures are found sculptured on the sides of rocks at a height which can now only be reached by means of extremely high scaffolding. When asked who can have carved these figures, the natives answer with a smile, as if it were a fact of which none but a white man could be ignorant, that “in the days of the great waters their fathers sailed in canoes at this height.” Here we find a geological dream serving as a solution of the problem presented by a long extinct civilization.

I would here be permitted to subjoin a remark, which I borrow from a letter addressed to me by Sir Robert Schomburgk, the distinguished traveller already mentioned. “The hieroglyphic figures are much more widely extended than you probably have conjectured. During my expedition, the object of which was the exploration of the river Corentyn, I not only observed several gigantic figures on the rock of Timeri (4° 30′ north lat. and 57° 30′ west long.), but I also discovered similar ones in the vicinity of the great cataracts of the river Corentyn (in 4° 21′ 30″ north lat. and 57° 55′ 30″ west long.) These figures have been executed more carefully than any others which I met with in Guiana. They are about 12 feet in height and appear to represent human figures. The head-gear is extremely remarkable; it surrounds the entire head, spreads far out, and is not unlike the glory represented round the heads of Saints. I left drawings of these images in the colony, which I hope some day to be able to lay collectively before the public. I have seen less complete figures on the Cuyuwini, a river which, flowing from the 151north-west, empties itself into the Essequibo in 2° 16′ north lat.; and I subsequently found similar figures on the Essequibo itself in 1° 40′ north lat. These figures, therefore, as appears from actual observations, extend from 7° 10′ to 1° 40′ north lat., and from 57° 30′ to 66° 30′ west long. The zone (or belt) of the sculptured rocks (as far as it has yet been investigated) thus extends over an area of 192,000 square miles, and includes within its circuit the basins of the Corentyn, Essequibo, and Orinoco—a circumstance that enables us to judge of the former population of this portion of the continent.”

Remarkable relics of a former culture, consisting of granitic vessels ornamented with beautiful representations of labyrinths, and the earthenware forms resembling the Roman masks, have been discovered among the wild Indians on the Mosquito coast.^[HU] I had them engraved in the picturesque Atlas appended to the historical portion of my travels. Antiquarians are astonished at the resemblance of these algreco vessels to those which embellish the Palace of Mitla (near Oaxaca, in New Spain). The large-nosed race, who are so frequently sculptured in relief on the Palenque of Guatimala and in Aztec pictures, I have never observed in Peruvian carvings. Klaproth recollects having noticed that the Chalkas, a horde of Northern Mongolia, had similar large noses. It is universally known, that many races of the North American, Canadian, and copper-coloured Indians, have fine aquiline noses, which constitute an essential physiognomical mark of distinction between them and the present inhabitants of New Granada, Quito, and Peru. Are the large-eyed, fair-skinned natives of the north-west coast of America, of whom Marchand speaks as living in 54° and 58° north lat., descended from the Usuns, an Alano-Gothic race of Central Asia?

52. p. 20.—“Deal certain death with a poisoned thumb-nail.”

The Otomacs frequently poison their thumb-nails with curare. The mere impress of the nail proves fatal, should the curare become mixed with the blood. We have in our possession the creeping plant, from the juice of which the curare is prepared, in the Esmeralda Mission, on the Upper Orinoco, 152but, unfortunately, we did not find the plant when in blossom. From its physiognomy, it seems to be allied to Strychnos.^[HV]

Since I wrote the above notice of the Curare, or Urari, as the plant and poison were called by Raleigh, the brothers Robert and Richard Schomburgk have rendered important service to science by making us accurately acquainted with the nature and mode of preparing this substance, which I was the first to bring to Europe in any considerable quantity. Richard Schomburgk found this creeping plant in flower in Guiana, on the banks of the Pomeroon and Sururu, in the territory of the Caribs, who are, however, ignorant of the mode of preparing the poison. His instructive work^[HW] gives the chemical analysis of the juice of the Strychnos toxifera, which, notwithstanding its name and organic structure, contains, according to Boussingault, no trace of strychnine. Virchow’s and Münter’s interesting physiological experiments show that the curare or urari poison does not appear to destroy by resorption from without, but chiefly when it is absorbed by the animal substance after the separation of the continuity of the latter; that curare does not belong to tetanic poisons; and that it especially produces paralysis, i.e., a cessation of voluntary muscular movement, while the function of the involuntary muscles (as the heart and intestines) continues unimpaired.^[HX]

53. p. 154—“Across that pacific arm of the sea.”

The Atlantic Ocean, between the parallels of 23° south lat. and 70° north lat., has the form of a furrowed longitudinal valley, in which the advancing and receding angles are opposite to each other. I first developed this idea in my work entitled Essai d’un Tableau Géologique de l’Amérique méridionale, which was published in the Journal de Physique, t. liii. p. 61.^[IC] From the Canary Isles, especially from 21° north lat., and 23° west long., to the north-east coast of South America, the surface of the ocean is so calm, and the waves so gentle, that an open boat might navigate it in safety.

54. p. 155—“Fresh springs among the Islands of the Antilles.”

On the southern coast of the island of Cuba, south-west of the harbour of Batabano, in the Gulf of Xagua, at a distance of eight to twelve miles from the shore, springs of fresh water gush from the bed of the ocean, probably from the action of hydrostatic pressure. The jet is propelled with such force that boats use extreme caution in approaching this spot, which is well known for its counter current producing a heavy swell. Trading vessels sailing along the coast, which do not purpose putting into port, sometimes visit these springs, in order to provide themselves, in the midst of the ocean, with a supply of fresh water. The freshness of the water increases with the depth from which it is drawn. River cows (Trichecus manati), which do not generally inhabit salt water, are frequently killed here. This singular phenomenon (the fresh springs), of which no mention had hitherto been made, was most accurately investigated by my friend, Don Francisco Lemaur, who made a trigonometrical survey of the Bahia de Xagua. I did not myself visit Xagua, but remained in the insular group situated further to the south (the so-called Jardines del Rey), to make astronomical determinations of their latitude and longitude.

55. p. 155—“Ancient rocky barrier.”

Columbus, whose unwearied spirit of observation was directed 175on every side, proposes in his letters to the Spanish monarchs, a geognostic hypothesis regarding the configuration of the larger Antilles. Being fully impressed with the idea of the strength of the Equinoctial current, which has often a westerly direction, he ascribes to it the disintegration of the group of the smaller Antilles, and the singularly lengthened configuration of the southern coasts of Porto Rico, Haiti, Cuba, and Jamaica, all of which follow almost exactly the direction of parallels of latitude. On his third voyage (from the end of May, 1498, to the end of November, 1500), when, from the Boca del Drago to the Island of Margarita, and afterwards from that island to Haiti, he felt the whole force of the equinoctial current, “that movement of the waters which accords with the movement of the heavens—movimiento de los cielos,” he says expressly that the violence of the current has torn the Island of Trinidad from the mainland. He refers the sovereigns to a chart which he sends them—a “pintura de la tierra,” drawn by himself, to which frequent reference is made in the celebrated lawsuit against Don Diego Colon respecting the rights of the first Admiral. “Es la carta de marear y figura que hizo el Almirante señalando los rumbos y vientos por los quales vino á Paria, que dicen parte del Asia.”^[ID]

56. p. 156—“Across the snow-crowned Paropanisus.”

In Diodorus’ description of the Paropanisus,^[IE] we seem to recognise a delineation of the Peruvian chain of the Andes. The army passed through inhabited districts in which snow daily fell!

57. p. 156—“Herrera in his Decades.”

Historia general de las Indias Occidentales, Dec. i. lib. iii. cap. 12 (ed. 1601, p. 106); Juan Batista Muñoz, Historia del Nuevo Mundo, lib. vi. c. 31, p. 301; Humboldt, Examen Crit., t. iii. p. 111.

58. p. 158—“The Sources of the Orinoco have never been visited by any European.”

Thus I wrote respecting these sources in the year 1807, in 176the first edition of the Ansichten der Natur, and I repeat with equal truth the same statement after an interval of forty-one years. The travels of the brothers Robert and Richard Schomburgk, so important in reference to all departments of natural science and geography, have established other and more interesting facts; but the problem of the situation of the sources of the Orinoco has been only partially solved by Sir Robert Schomburgk. M. Bonpland and myself advanced from the west as far as Esmeralda, or the confluence of the Orinoco with the Guapo; and I was enabled, by the aid of well-attested information, to describe the upper course of the Orinoco to above the mouth of the Gehette, and to the small waterfall (Raudal) de los Guaharibos. From the east Sir Robert Schomburgk, proceeding from the mountains of the Majonkong Indians, the inhabited portion of which he estimated by the boiling point of water to be 3517 feet in height, succeeded in reaching the Orinoco by the Padamo River, which the Majonkongs and Guinaus (Guaynas?) call Paramu.^[IF] I had placed this confluence of the Padamo with the Orinoco in my Atlas, in 3° 12′ N. lat., and 65° 46′ W. long. but Schomburgk found it by direct observation in 2° 53′ lat. and 65° 48′ W. long. The main object of this traveller’s journey was not ‘natural history,’ but the solution of the prize question proposed by the Royal Geographical Society of London, in November, 1834,—on the connection of the coast of British Guiana with the easternmost point which I had reached on the Upper Orinoco. After undergoing many sufferings, this object was thoroughly achieved. Robert Schomburgk reached Esmeralda, with his instruments, on the 22nd of February, 1839. His determinations of the latitude and longitude of the place agreed more closely with mine than I had anticipated. Let us here allow the observer to speak for himself:—“Words are inadequate to describe the feelings which overwhelmed me when I sprang on shore. My object was attained; my observations, begun on the coast of Guiana, were brought into connection with those of Humboldt at Esmeralda, and I freely admit that at a time when my physical powers had almost entirely deserted me, and when I was surrounded by dangers and difficulties of no ordinary kind, the recognition which I hoped 177for from him, was the sole inducement which inspired me with a fixed determination to press forward towards the goal which I had now reached. The emaciated figures of my Indian companions and my faithful guides proclaimed more fully than any words could do, what difficulties we had had to surmount, and had surmounted.” After citing expressions so gratifying, I must be permitted to subjoin the opinions I expressed regarding this great undertaking promoted by the Royal Geographical Society of London, in my Preface to the German edition of Robert Schomburgk’s Account of his Travels, published in 1841. “Immediately after my return from Mexico, I indicated the direction and the routes by which the unknown portion of the South American Continent between the sources of the Orinoco, the mountain chain of Pacaraima, and the sea-shore near Essequibo, might be explored. These wishes, so strongly expressed in the personal narrative of my journey, have at length, after the lapse of nearly half a century, been for the most part fulfilled. I rejoice that I have been spared to see so important an enlargement of our geographical knowledge; I rejoice too in seeing a courageous and well-conducted enterprise, requiring the most devoted perseverance, executed by a young man, to whom I feel bound no less by the ties of similarity of pursuits than those of country. These circumstances were alone able to overcome the aversion and disinclination which I entertain, perhaps unjustly, for introductory prefaces by a different hand than that of the author himself. But I could not resist the impulse of expressing thus publicly my sincere esteem for the accomplished traveller who, led on by the meritorious idea of penetrating from east to west, from the Valley of the Essequibo to Esmeralda, has succeeded, after five years of efforts and of sufferings (the extent of which I well appreciate from my own experience), in attaining the object of his ambition. Courage for the sudden accomplishment of a hazardous undertaking is easier to find, and implies less inward strength, than the resolution to endure with resignation long-continued physical sufferings, excited by absorbing mental interest; and still to press forward, undismayed by the certainty of having to retrace his steps under equally great privations and with enfeebled powers. Serenity of mind, which is almost the first requisite for an 178enterprise in inhospitable regions, a passionate love for any department of scientific labour (be it natural history, astronomy, hypsometrics, or magnetism), a pure feeling for the enjoyment which nature is capable of imparting, are elements which, when they combine together in one individual, ensure valuable results from a great and important journey.”

I will preface my consideration of the question of the sources of the Orinoco with my own conjectures in relation to the subject. The perilous route travelled in 1739 by the surgeon Nicolas Hortsmann, of Hildesheim; in 1775 by the Spaniard Don Antonio Santos, and his friend Nicolas Rodriguez; in 1793 by the Lieutenant-Colonel of the 1st Regiment of the Line of Para, Don Francisco José Rodriguez Barata; and (according to manuscript maps, for which I am indebted to the former Portuguese Ambassador in Paris, Chevalier de Brito) by several English and Dutch settlers, who in 1811 travelled from Surinam to Para by the portage of the Rupunuri and by the Rio Branco;—divides the terra incognita of the Parime into two unequal parts, and serves to mark the position of a very important point in the geography of those regions—viz., the sources of the Orinoco, which it is no longer possible to remove to an indefinite distance towards the east, without intersecting the bed of the Rio Branco, which flows from north to south through the fluvial district of the Upper Orinoco; while this portion of the great river itself pursues for the most part a direction from east to west. The Brazilians, since the beginning of the present century, have from political motives manifested a vivid interest in the extensive plains east of the Rio Branco.^[IG] Owing to the position of Santa Rosa on the Uraricapara, whose course appears to have been pretty accurately determined by Portuguese engineers, the sources of the Orinoco cannot be situated east of the meridian of 63° 8′ west long. This is the eastern limit beyond which they cannot be placed, and taking into consideration the state of the river at the Raudal de los Guaharibos (above Caño Chiguire, in the country of the strikingly fair-skinned Guaycas Indians, and 52′ east of the great Cerro 179Duida), it appears to me probable that the Orinoco in its upper part does not extend, at the utmost, beyond the meridian of 64° 8′ west long. This point is, according to my combinations, 4° 12′ west of the little lake of Amucu, which was reached by Sir Robert Schomburgk.

I will now detail the conjectures of that traveller, after having first given my own earlier ones. According to him the course of the Upper Orinoco, to the east of Esmeralda, is directed from south-east to north-west; my estimations of latitude for the mouths of the Padamo and the Gehette appear to be respectively 19′ and 36′ too small. Schomburgk conjectures that the sources of the Orinoco are situated in lat. 2° 30′, and the fine “Map of Guayana, to illustrate the route of R. H. Schomburgk,” which accompanies the splendid English work entitled Views in the Interior of Guiana, places its geographical sources in 64° 56′ west long., i.e., 1° 6′ west of Esmeralda, and only 48′ of longitude nearer to the Atlantic than I had determined the position of this point. Astronomical combinations led Schomburgk to place the mountain of Maravaca, which is about ten thousand feet high, in 3° 41′ lat. and 65° 48′ west long. The Orinoco was scarcely three hundred yards wide near the mouth of the Padamo or Paramu, and more to the west, where it expands to a width of from four to six hundred yards, it was so shallow, and so full of sandbanks, that the expedition was obliged to dig channels, as the river bed was only fifteen inches deep. Fresh-water dolphins were still to be seen in great numbers everywhere—a phenomenon which the zoologists of the eighteenth century would not have expected to find in the Orinoco and the Ganges.

59. p. 158—“The most luxurious product of a tropical climate.”

The Bertholletia excelsa (Juvia), of the family of Myrtaceæ (and placed in Richard Schomburgk’s proposed division of Lecythideæ), was first described in Plantes Equinoxiales, t. i. 1808, p. 122, tab. 36. This colossal and magnificent tree offers, in the perfect development of its cocoa-like, round, close-grained, woody fruit, inclosing the three-cornered and also woody seed-vessels, the most remarkable example of luxuriant organic development. The Bertholletia grows in 180the forests of the Upper Orinoco, between the Padamo and the Ocamu, in the vicinity of the mountain of Mapaya, as well as between the rivers Amaguaca and Gehette.^[IH]

60. p. 158—“Grass stalks, whose joints measure upwards of eighteen feet from knot to knot.”

Robert Schomburgk, when visiting the small mountainous country of the Majonkongs, on his route to Esmeralda, was fortunate enough to determine the species of Arundinaria, which furnishes the material for these blowing-tubes. He says of this plant: “It grows in large tufts, like the bambusa; the first joint rises, in the old cane, without a knot, to a height of from 16 to 17 feet before it begins to bear leaves. The entire height of the Arundinaria, growing at the foot of the great mountain-cluster of Maravaca, is from 30 to 40 feet, with a thickness of scarcely half an inch in diameter. The top is always inclined; and this species of grass is peculiar to the sandstone mountains between the Ventuari, the Paramu (Padamo), and the Mavaca. The Indian name is Curata, and, therefore, from the excellence of these celebrated long blowing-tubes, the Majonkongs and Guinaus of these districts have acquired the name of the Curata nation.”^[II]

61. p. 159—“Fabulous origin of the Orinoco from a lake.”

The lakes of these regions (some of which are wholly imaginary, while the real size of others has been much exaggerated by theoretical geographers) may be divided into two groups. The first of these groups comprise those situate between Esmeralda (the most easterly mission on the Upper Orinoco), and the Rio Branco; to the second, belong the lakes presumed to exist in the district between the Rio Branco and French, Dutch, and British Guiana. This general view, of which travellers should never lose sight, proves that the question of whether there is another Lake Parime eastward of the Rio Branco, besides the Lake Amucu, seen by Hortsmann, Santos, Colonel Barata, and Schomburgk, has nothing whatever to do with the problem of the sources of the Orinoco. As the name of my distinguished friend the former Director of the Hydrographic Office at Madrid, Don Felipe Bauza, is 181of great weight in questions of geography, the impartiality which ought to influence every scientific investigation makes it incumbent on me to mention that this learned man was inclined to the view that there must be lakes west of the Rio Branco, at no great distance from the sources of the Orinoco. He wrote to me from London shortly before his death, “I wish you were here that I might converse with you respecting the geography of the Upper Orinoco, which has occupied you so much. I have been fortunate enough to rescue from entire destruction the papers of the General of Marine, Don José Solano, father of the Solano who perished in so melancholy a manner at Cadiz. These documents relate to the settlement of the boundary line between the Spaniards and Portuguese, with which Solano had been charged since 1754, in conjunction with the Escadre Chef Yturriaga and Don Vicente Doz. In all these plans and sketches I find a Laguna Parime sometimes as a source of the Orinoco, and sometimes as wholly detached from it. Are we then to assume that there is another lake further eastward to the north-east of Esmeralda?”

Löffling, the celebrated pupil of Linnæus, accompanied the last-named expedition to Cumana in the capacity of botanist. He died on the 22nd of February, 1756, at the mission of Santa Eulalia de Murucuri (somewhat to the south of the confluence of the Orinoco and Caroni), after traversing the missions on the Piritu and Caroni. The documents of which Bauza speaks are the same as those on which the great map of De la Cruz Olmedilla is based. They have served as the foundation of all the maps of South America, which appeared in England, France, and Germany, before the end of the last century; and have also served for the two maps executed in 1756 by Father Caulin, the historiographer of Solano’s expedition, and by M. de Surville, Keeper of the Archives in the Secretary of State’s Office at Madrid, who was but an unskilful compiler. The contradictions abounding in these maps show the little reliance that can be placed on the results of this expedition. Nay more, Father Caulin, above referred to, acutely details the circumstances which gave rise to this fable of the lake of Parime; and the map of Surville, which accompanies his work, not only restores this lake, under the name of the White Lake, and the Mar Dorado, but indicates another smaller one, from which flow partly by means 182of collateral branches, the Orinoco, Siapa, and Ocamo. I was able to convince myself on the spot of the following facts well known in the missions; that Don José Solano did not do more than cross the cataracts of Atures and Maypures; that he did not reach the confluence of the Guaviare and the Orinoco in 4° 3′ north lat., and 68° 9′ west long.; and that the astronomical instruments of the boundary expedition were neither carried to the isthmus of the Pimichin and the Rio Negro, nor to the Cassiquiare; and even on the Upper Orinoco, not beyond the mouth of the Atabapo. This vast extent of territory was not made the scene of any accurate observations before my journey, and has subsequently to Solano’s expedition been traversed only by some few soldiers who had been sent on exploring expeditions; while Don Apolinario de Fuente, whose journal I obtained from the archives of the province of Quixos, has gathered without discrimination everything from the fallacious narratives of the Indians that could flatter the credulity of the Governor Centurion. No member of the expedition had seen a lake, and Don Apolinario was unable to advance beyond the Cerro Yumarique and Gehette.

Although a line of separation, formed by the basin of the Rio Branco, is now established throughout the whole extent of the country, to which we are desirous of directing the inquiring zeal of travellers, it must yet be admitted, that our geographical knowledge of the district west of this valley between 62° and 66° long., has made no advance whatever for at least a century. The repeated attempts made by the Government of Spanish Guiana since the expeditions of Iturria and Solano, to reach and to pass over the Pacaraima Mountains, have been attended by very unimportant results. When the Spaniards, in proceeding to the missions of the Catalonian capuchins of Barceloneta, at the confluence of the Caroni and the Rio Paragua, ascended the last-named river southward to its junction with the Paraguamusi, they founded at this point the mission of Guirion, which, at first, bore the pompous appellation of Ciudad de Guirion. I place it in about 4° 30′ north latitude. From thence the Governor Centurion, in consequence of the exaggerated accounts given by two Indian chiefs, Paranacare and Arimuicapi, respecting the powerful tribe of the Ipurucotos, was excited to search for ‘El Dorado,’ and in carrying what were then called spiritual conquests still further, founded, beyond the 183Pacaraima Mountains, the two villages of Santa Rosa and San Bautista de Caudacacla. The former was situate on the upper eastern bank of the Uraricapara, a tributary of the Uraricuera, which I find in the journal of Rodriguez under the name of the Rio Curaricara; the latter, at from 24 to 28 miles further east-south-east. The astronomo-geographer of the Portuguese Boundary Commission, Captain Don Antonio Pires de Sylva Pontes Leme, and the Captain of Engineers, Don Ricardo Franco d’Almeida de Serra, who between 1787 and 1804, surveyed with the greatest care the whole course of the Rio Branco and its upper tributaries, call the most western part of the Uraricapara, “The Valley of Inundation.” They place the Spanish mission of Santa Rosa in 3° 46′ north lat., and mark the route that leads from thence northward across the mountain chain to the Caño Anocapra, a branch of the Paraguamusi, which forms a connecting passage between the basin of the Rio Branco and that of the Caroni. Two maps of these Portuguese officers, embracing all the details of the trigonometrical survey of the bends of the Rio Branco, the Uraricuera, the Tacutu, and the Mahu, were most kindly communicated to Colonel Lapie and myself by the Count of Linhares. These valuable unpublished documents, of which I have availed myself, are still in the hands of the learned geographer, who long since began to have them engraved at his own expense. The Portuguese sometimes call the whole of the Rio Branco by the name of Rio Parime, and sometimes limit this appellation to one branch only, the Uraricuera, somewhat below the Caño Mayari and above the old mission of San Antonio. As the words Paragua and Parime alike imply water, great water, lake, and sea, we cannot wonder at finding them so often repeated among tribes living at great distances from each other; as, for instance, by the Omaguas on the Upper Marañon, by the Western Guaranis, and by the Caribs. In all parts of the world, as I have already remarked, large rivers are called by those who live on their banks “the River,” without any specific denomination. Paragua, the name of a branch of the Caroni, is also the term applied by the natives to the Upper Orinoco. The name Orinucu is Tamanakish; and Diego de Ordaz first heard it used in the year 1531, when he ascended to the mouth of the Meta. Besides the Valley of Inundation above mentioned 184we find other large pieces of water between the Rio Xumuru and the Parime. One of these bays is a branch of the Tacutu, and the other of the Uraricuera. Even at the base of the Pacaraima Mountains the rivers are subject to great periodical overflowings; and the Lake Amucu, of which we shall subsequently speak more fully, exhibits exactly the same character at the commencement of the plains. The Spanish missions, Santa Rosa and San Bautista de Caudacacla, or Cayacaya, founded in the years 1770 and 1773, by the Governor Don Manuel Centurion, were destroy ed before the close of the last century; and since that time, no new attempt has been made to advance from the basin of the Caroni to the southern declivity of the Pacaraima Mountains.

The territory east of the valley of the Rio Branco has of late years been made the subject of several successful explorations. Mr. Hillhouse navigated the Massaruni as far as the Bay of Caranang, whence, as he says, a path would lead the traveller, in two days, to the source of the Massaruni; and, in three days, to the tributaries of the Rio Branco. With respect to the windings of the great river Massaruni, described by Mr. Hillhouse, he himself observes, in a letter addressed to me from Demerara, 1st January, 1831, that “the Massaruni, reckoning from its sources, flows first to the west, then for one degree of latitude to the north; afterwards nearly 200 miles eastward; and, finally, to the north and north-north-east till it merges in the Essequibo.” As Mr. Hillhouse was unable to reach the southern declivity of the Pacaraima chain, he was not acquainted with the Amucu Lake; and he says himself, in his printed report, that “from the accounts given him by the Accaouais, who are continually traversing the country between the shore and the Amazon River, he is convinced there is no lake in this district.” This assertion occasioned me some surprise, as it was directly opposed to the views I had previously formed regarding the Lake Amucu, from which flows the Caño Pirara, according to the accounts given by the travellers Hortsmann, Santos, and Rodriguez (and which had inspired me with the more confidence, because they entirely coincide with the recent Portuguese manuscript charts). Finally, after five years of expectation, Schomburgk’s journey has removed all farther doubt.

“It is difficult to believe,” says Mr. Hillhouse, in his interesting 185memoir on the Massaruni, “that the tradition of a large inland sea is wholly unfounded. According to my views, the following circumstance may have given rise to the belief in the existence of the fabulous lake of the Parime. At some distance from the rocky fall of Teboco the waters of the Massaruni present to the eye as little motion as the calm surface of a lake. If at a more or less remote period the horizontal granitic strata of Teboco had been totally compact and without fissures, the waters must have been at least 50 feet above their present level, and there would have been formed an immense lake 10 or 12 miles in width, and 1500 or 2000 miles in length.”^[IJ] The extent of this supposed inundation is not the only reason which prevents me from acceding to this explanation; for I have seen plains (Llanos), where, during the rainy season, the overflowing of the tributaries of the Orinoco annually covered a surface of 6400 square miles. The labyrinth of ramifications between the Apure, Arauca, Capanaparo, and Sinaruco (see maps 17 and 18 of my Physical Atlas), is then wholly lost sight of; the configuration of the river beds can no longer be traced, and the whole appears like one vast lake. But the locality of the fabulous Dorado, and of the Lake Parime, belongs historically to quite a different part of Guiana, namely, that lying south of the Pacaraima mountains. This myth of the White Sea and of the Dorado of the Parime, has arisen, as I endeavoured thirty years ago to show in another work, from the appearance of the micaceous rocks of the Ucucuamo, the name Rio Parime (Rio Branco), the inundations of the tributaries; and especially from the existence of the lake Amucu, which is in the neighbourhood of the Rio Rupunuwini (Rupunuri), and is connected by means of the Pirara with the Rio Parime.

I have had much pleasure in finding that the travels of Sir Robert Schomburgk have fully corroborated these early views. The section of his map which gives the course of the Essequibo and of the Rupunuri is quite new, and of great importance in a geographical point of view. It places the Pacaraima chain between 3° 52′and 4° north lat., while I had given its mean direction from 4° to 4° 10′. The chain reaches the confluence of the Essequibo and Rupunuri in 3° 57′ north lat., and 58° 1′ west longitude; I had placed it half a degree too 186far to the north. Schomburgk calls the last-named river Rupununi, according to the pronunciation of the Macusis; and gives as the synonymes Rupunuri, Rupunuwini and Opununy, which have arisen from the difficulty the Carib tribes of these districts find in pronouncing the letter “r.” The position of the lake Amucu and its relations to the Mahu (Maou) and Tacutu (Tacoto) correspond perfectly with my map of Colombia drawn in 1825. We agree equally well regarding the latitude of the lake of Amucu, for while he places it in 3° 33′, I considered it to be in 3° 35′; the Caño Pirara (Pirarara) which connects the Amucu with the Rio Branco, flows from it towards the north, and not to the west as I had marked it. The Sibarana of my map, the sources of which Hortsmann placed to the north of the Cerro Ucucuamo near a fine mine of rock crystal, is the Siparuni of Schomburgk’s map. His Waa-Ekuru is the Tavaricaru of the Portuguese geographer Pontes Leme, and is the branch of the Rupunuri which lies the nearest to the lake of Amucu.

The following remarks from the report of Sir Robert Schomburgk throw some light on the subject in question. “The lake of Amucu,” says this traveller, “is without doubt the nucleus of the Lake of Parime and of the supposed White Sea. In December and January, when we visited it, it was scarcely a mile in length, and was half covered with reeds.” The same observation occurs on D’Anville’s map of 1748. “The Pirara flows from the lake to the W.N.W. of the Indian village of Pirara and falls into the Maou or Mahu. The last-named river rises, according to the information given me, north of the ridge of the Pacaraima mountains, which in their eastern portion do not attain a greater elevation than about 1600 feet. The sources of the river are on a plateau, from whence it is precipitated in a beautiful waterfall, known as the Corona. We were on the point of visiting this fall, when on the third day of our excursion to the mountains, the indisposition of one of my companions compelled me to return to the station at the lake Amucu. The Mahu has black coffee-coloured water, and its current is more impetuous than that of the Rupunuri. In the mountains through which it pursues its course it is about 60 yards in breadth. Its environs are here extremely picturesque. This valley as well as the bank of the Buroburo, which flows into the Siparuni, are inhabited by the Macusis. 187In April the whole Savannahs are overflowed, and then present the peculiar phenomenon of the waters belonging to two different river basins commingling together. It is probable that the vast extent of this temporary inundation may have given rise to the fable of the lake of Parime. During the rainy season a water communication is formed in the interior of the country between the Essequibo, the Rio Branco, and the Gran Para. Some groups of trees, rising like Oases on the sand-hills of the Savannahs, present, at the time of the inundation, the appearance of islands scattered over a lake; and these are without doubt the Ipomucena islands of Don Antonio Santos.”

In D’Anville’s manuscripts, which his heirs kindly allowed me to examine, I find that Hortsmann of Hildesheim, who described these districts with great care, saw a second Alpine lake, which he places two day’s journey above the confluence of the Mahu with the Rio Parime (Tacutu?). It is a black water lake, situated on the summit of a mountain. He explicitly distinguishes it from the lake of Amucu, which he describes as “covered with rushes.” The descriptions given by Hortsmann and Santos coincide with the Portuguese manuscript maps of the Marine Bureau at Rio Janeiro, in not indicating the existence of an uninterrupted connection between the Rupunuri and the lake of Amucu. In D’Anville’s maps of South America, the rivers are better drawn in the first edition published in 1748, than in the more extensively circulated one of 1760. Schomburgk’s travels fully confirm the independence of the basin of the Rupunuri and Essequibo; but he draws attention to the fact that, during the rainy season, the Rio Waa-Ekuru, a tributary of the Rupunuri, is in connection with the Caño Pirara. Such is the condition of these river-channels, which are still but little developed, and almost entirely without separating ridges.

The Rupunuri and the village of Anai, 3° 56′ north latitude, 58° 34′ west longitude, are at present recognised as the political boundaries between the British and Brazilian domains in these desert regions. Sir Robert Schomburgk was compelled by severe illness to make a protracted stay at Anai. He bases his chronometrical determinations of the position of the lake of Amucu on the mean of many lunar distances, east and west, which he measured during his sojourn at Anai. His 188determinations of longitude for these points of the Parime are in general one degree more east than those in my map of Colombia. While I am far from calling in question the result of these lunar observations taken at Anai, I may be allowed to observe that the calculation of these distances is of importance, when it is desired to carry the comparison from the lake of Amucu to Esmeralda, which I found in 66° 19′ west longitude.

Thus then we see the great Mar de la Parima, (which it was so difficult to remove from our maps, that even after my return from America it was still supposed to be 160 miles in length,) reduced by recent investigations to the lake of Amucu, measuring only two or three miles in circumference. The illusions entertained for nearly two hundred years, and which in the last Spanish expedition, in 1775, for the discovery of El Dorado, cost several hundred lives, have finally terminated by enriching geography with some few results. In the year 1512 thousands of soldiers perished in the expedition, undertaken by Ponce de Leon, to discover the “Fountain of Youth,” on one of the Bahama Islands, called Binimi, which is hardly to be found on any of our maps. This expedition led to the conquest of Florida, and to the knowledge of the great oceanic current, or gulf-stream, which flows through the Straights of Bahama. The thirst after gold, and the desire of rejuvenescence—the Dorado and the Fountain of Youth—stimulated, to an almost equal extent, the passions of mankind.

62. p. 161—“The Piriguao, one of the noblest forms of the Palm.”

Compare Humboldt, Bonpland, and Kunth, Nova Genera Plantarum, and Plant. æquinoct., t. i. p. 315.

63. p. 171—“The grave of an extinct race.”

During my stay in the forests of the Orinoco, researches were being made, by royal command, in reference to these bone-caves. The missionary of the cataracts had been falsely accused of having discovered in these caves treasures which the Jesuits had concealed there prior to their flight.

64. p. 172—“When his language perished with him.”

The parrot of the Atures has been made the subject of a 189charming poem by my friend Professor Ernst Curtius, the tutor of the promising young Prince Friedrick Wilhelm of Prussia. The author will forgive me for closing the present section of the “Views of Nature” with this poem, which was not designed for publication, and was communicated to me by letter.

65. p. 191.—“Characteristic denominations in Arabic and Persian.”

More than twenty words might be cited by which the Arabs distinguish between a Steppe (tanufah), according as it may be a Desert without water, entirely bare, or covered with siliceous sand, and interspersed with spots of pasture land (Sahara, Kafr, Mikfar, Tih, Mehme). Sahl is a depressed plain; Dakkah a desolate elevated plateau. In Persian Beyaban is an arid sandy waste (as the Mongolian Gobi and the Chinese Han-hai and Scha-mo); Yaila is a Steppe covered with grass rather than with low-growing plants (like the Mongolian Küdah, the Turkish Tala or Tschol, and the Chinese Huang). Deschti-reft is a naked elevated plateau.^[IK]

66. p. 191.—“The old Castilian dialects.”

Pico, picacho, mogote, cucurucho, espigon, loma tendida, mesa, panecillo, farallon, tablon, peña, peñon, peñasco, peñoleria, roca partida, laxa, cerro, sierra, serrania, cordillera, monte, montaña, montañuela, cadena de montes, los altos, malpais, reventazon, bufa, &c.

67. p. 194.—“Where the map had indicated Montes de Cacao.”

On the range of hills from which the lofty Andes de Cuchao have originated, see my Relation historique, t. iii. p. 238.

68. p. 197.—“Hermesia.”

The genus Hermesia, the Sauso, has been described by Bonpland, and is delineated in our Plantes équinoxiales, t. i. p. 162, tab. xlvi.

69. p. 199.—“The fresh-water dolphin.”

These are not sea dolphins, which, like some species of Pleuronectes (flat fish which invariably have both eyes on one side of the body), ascend the rivers to a great distance, as, 203for instance, the Limande (Pleuronectes Limanda), which is found as far inland as Orleans. Some forms of sea fish, as the dolphin and skate (Raia), are met with in the great rivers of both continents. The fresh-water dolphin of the Apure and the Orinoco differs specifically from the Delphinus gangeticus as well as from all sea dolphins.^[IL]

70. p. 199.—“The striped nocturnal monkey.”

This is the Douroucouli or Cusi-cusi of the Cassiquiare which I have elsewhere described as the Simia trivirgata,^[IM] from a drawing made by myself of the living animal. We have since seen the nocturnal monkey living in the menagerie of the Jardin des Plantes at Paris.^[IN] Spix also met with this remarkable little animal on the Amazon River and called it Nyctipithecus vociferans.