1For a list of the plants collected here, see a paper in the Journal of the LinnÆan Society, vol. xxii.

2Much cinchona bark, coming from the interior, was formerly shipped at Tumaco; but between horrible roads and the reckless waste of the forests through mismanagement, but little is now conveyed by this way.

3For a list of the species collected, see the Journal of LinnÆan Society, vol. xxii.

4The abrupt change in the vegetation on this part of the American coast has been noticed by Humboldt, Weddell, and other scientific travellers. In a note to the French edition of Grisebach (“Vegetation du Globe,” traduit par P. de Tchihatcheff, ii. p. 615), M. AndrÉ expresses the opinion that this, as well as some other cases of abrupt change in the vegetation observed by him in Colombia, are to be explained by the nature of the soil, which in the arid tracts is sandy or stony, and fails to retain moisture. Admitting that in certain cases this may afford a partial explanation of the facts, it is scarcely conceivable that the limit of the zone wherein little or no rain falls should exactly coincide with a change in the constitution of the soil, and I should be more disposed to admit a reversed order of causation, the porous and mobile superficial crust remaining in those tracts where, owing to deficient rainfall, there is no formation of vegetable mould, and no accumulation of the finer sediment forming a retentive clay.

5The only detailed account of the operations that I have seen is in a work entitled, “Histoire de la Guerre du Pacifique,” by Don Diego Barros Arana. Paris: 1881. It appears to be fairly accurate as to facts, but coloured by very decided Chilian sympathies.

6The heights given in the text are those of the railway stations.

7Of 138 genera of HelianthoÏdeÆ 107 are exclusively confined to the American continent, 18 more are common to America and distant regions of the earth, one only is limited to tropical Asia, and two to tropical Africa, the remainder being scattered among remote islands—the Sandwich group, the Galapagos, Madagascar, and St. Helena.

8See noteto page 184.

9In Nature for September 14, 1882.

10The only accurate information that I have found respecting the climate of Lima is contained in a paper by Rouand y Paz Soldan, “Resumen de las Observaciones Meteorologicas hechas en Lima durante 1869,” quoted in the French translation of Grisebach’s “Vegetation du Globe.” Reduced to English measures, they give the following results:—

There is reason to think that the temperature for July, 1869, given above was exceptionally low, and although the months during which fogs prevail are abnormally cool for a place within 13° of the equator, I believe that the thermometer rarely falls below 60° Fahr.

11See Appendix A, On the Fall of Temperature in ascending to Heights above the Sea-level.

12It is a curious illustration of the utterly untrustworthy character of statements made by unscientific travellers to read the following passage in a book published by a recent traveller in South America, who visited Chicla in November, the beginning of summer. He declares that the fringe of green vegetation “dwindles and withers at a height of nine or ten thousand feet;... while on the upper grounds, where sometimes rain is plentiful, the air is too keen and cold for even the most dwarfish and stunted vegetation to thrive.”

13“Versuch einer Entwicklungsgeschichte der Pflanzenwelt.”

14The heights are certainly incorrect. The base of the hill of Amancaes is nearly seven hundred feet above sea-level, and Mr. Nation states that the two localities mentioned by Mr. Cruikshank are at about the same elevation.

15Two small Chilian wooden ships, the Esmeralda, of 850 tons, mounting eight guns, commanded by Arturo Prat, and the Covadonga, of 412 tons, with two guns, commanded by Condell, were engaged in the blockade of Iquique, when, on the 21st of May, 1879, they were attacked by the Peruvian ironclad Independencia, of 2004 tons, mounting 18 (chiefly heavy Armstrong) guns, commanded by J.G. Moore, and the monitor Huascar, of 1130 tons, mounting two 300-pounder Armstrong turret guns, besides two deck guns, under Miguel Grau, the most skilful and enterprising of the Peruvian commanders. The Chilian captains resolved on a desperate defence. After maintaining for two hours the fight against the Huascar, Arturo Prat resolved on the attempt to board his adversary. Bringing his ship alongside, he sprang on the deck of the Huascar; but the ships were separated at once, and two men only fell along with him, while the Esmeralda went to the bottom with her crew of 180 men, of whom several were picked up by the boats of the Huascar. The Independencia, following the little Covadonga, ran on the rocks in the shallows south of Iquique, and became a total wreck; while the Covadonga, though shattered by her enemy’s guns, was able to reach Autofogasta. The heroism of the Chilian commanders saved their country, and at the critical moment changed the fortune of the war.

16In the preface to his “Florula Atacamensis,” Dr. Philippi, who has explored this region more thoroughly than any other traveller, states that on the range of coast hills between the Pan de Azucar (lat. 26° 8' south) and Miguel Diaz (lat. 24° 36') the fogs, called in Peru garua, or garruga, deposit during a great part of the year some moisture which occasionally takes the form of fine rain, such as is familiarly known to occur on the hills near Lima. He remarks as singular the fact that the same phenomenon is not observed on the coast north or south of those limits. From more recent observations, it would appear that this is not strictly true as regards the higher coast hills near Coquimbo, but it seems to hold as regards the tract of coast to the northward, between the neighbourhood of Taltal and that of Iquique, a distance of about four degrees of latitude. It may be that the coast hills are lower here than further south, and that as the desert region inland rises very gradually, and has a higher temperature inland than near the coast, the formation of fog is prevented. Whatever be the cause, the absence of fog would go far to account for the utter sterility of this region.

17The four species of Encelia described in De Candolle’s “Prodromus” appear to me to be but slightly modified forms of a single species. Since the publication of that work, several other and quite distinct species have been ranked under the same generic name.

18While botanizing in the Tajo de Ronda, the singular cleft which cuts through the rocky hill on which the town is built, I was once for some time in positive danger. The boys, having espied me, assembled on the bridge that crosses the cleft, some three hundred feet above my head, and commenced a regular fire of stones, that drove me to take shelter under an overhanging rock until, being tired of the sport, they turned their attentions elsewhere.

19One of the difficulties felt by all students of geographical distribution arises from the imperfect or careless indications given both in books and in herbaria, and this is more felt in regard to South America than as to any other part of the world. A very large proportion of the earlier collections bear simply the label “Brazil,” forgetting that the area is as great as that of Europe. In other cases local names of places, not to be found on maps or in gazetteers, embarrass the student and weary his patience. It is mainly from Darwin that naturalists have learned that geographical distribution is the chief key to the past history of the earth.

20The last season of excessive rainfall was that of 1877. I have seen no complete returns, but it appears that the rain of that year commenced in Central Chili in February, a very rare phenomenon; that more than six inches of rain fell in April, of which, at Santiago, four inches fell in twenty-four hours. More heavy rain fell in May, and finally in July a succession of storms flooded large districts, destroying property and life, the fall for the month being more than fourteen inches at Valparaiso. Much interesting information respecting the climate of Chili will be found in a work by Don B. VicuÑa Mackenna, “Ensayo Historico sobre el Clima de Chile” (Valparaiso: 1877), from which I have borrowed the above-mentioned particulars.

21I believe that in the column for rainfall at Punta Arenas, snow has not been taken into account.

22The recent untimely death of this valuable official is deplored by all classes in Chili.

23This is doubtless the summit described by Darwin under the name Campana de Quillota. He gives the height as 6400 feet above sea-level. The figures in the text are taken from the Chilian survey.

24The mapping of the Andean chain is a task of immense difficulty, and although the Chilian survey is the best that has yet been executed, it leaves much to be desired. Even in the small district which I was able to visit, I found several grave errors in Petermann’s map, reduced from the Chilian survey, which is, nevertheless, the best that has been published in Europe. One of the most serious is the omission of the Uspallata Pass, the most frequented of those leading from Central Chili to the Argentine territory, which is neither named nor correctly indicated by the tints adopted to mark the zones of elevation.

25“Origin of Species,” 3rd edit., p. 410.

26Molina, one of the most pernicious blunderers who have brought confusion into natural history, grouped together under the generic name Peumus several Chilian plants having no natural connection with each other. Misled by his erroneous description, botanists have applied the name peumus to a fragrant shrub, common about Valparaiso and elsewhere, which is known in the country by the name boldu.

27The Baths of Cauquenes are said to be 2523 feet above the sea; the Morro, by aneroid observation, is about 2000 feet higher.

28As happens with many other plants described by early botanists, there has been much confusion in regard to the species named by LinnÆus Lobelia Tupa. The plant was first made known to Europeans by the excellent traveller, Father FeuillÉe, whose “Journal des Observations Physiques MathÉmatiques et Botaniques faites sur les cÔtes de l’AmÉrique meridionale, etc.,” published in 1714, is a book which may still be consulted with advantage. His descriptions of plants are usually careful and accurate, but the accompanying plates all ill-executed and often misleading. LinnÆus, followed by Willdenow, refers to FeuillÉe’s work, but gives a very brief descriptive phrase which suits equally well FeuillÉe’s plant and several others subsequently discovered. Aiton, in the “Hortus Kewensis,” gives the name Lobelia Tupa to a plant which is plentiful about Valparaiso, where I found it still in flower, the seeds of which were received at Kew about a century ago from Menzies. This is now generally known by the not very appropriate name Tupa salicifolia of Don, but was first published by Sims in the Botanical Magazine, No. 1325, as Lobelia gigantea, which name it should now bear. The plant which I found near Cauquenes appears to be the Tupa Berterii of Decaudolle, a rare species, apparently not known to the authors of the “Flora Chilena.” No doubt could have arisen as to the plant intended by LinnÆus as Lobelia Tupa if writers had referred to FeuillÉe’s full and accurate description. His account of the poisonous effects of the plant was probably derived from the Indians, and may be exaggerated. The whole plant, he says, is most poisonous, the mere smell causing vomiting, and any one touching his eyes after handling the leaves is seized with blindness. I may remark that the latter statement, which appears highly improbable, receives some confirmation from the observations of Mr. Nation, mentioned above in page 77. The plant which I saw in Peru, but failed to collect, is much smaller than most of the Chilian species, and has purple flowers, but is nearly allied in structure. It is probably the Tupa secunda of Don. I gather from a passage in one of Mr. Philippi’s writings that the word tupa in Araucanian signifies poison. We are yet, I believe, ignorant of the chemical nature of the poisonous principle contained in the plants of this group.

29The measurements of the height of the peak of Aconcagua vary considerably in amount, but I believe that the most reliable is that adopted by Petermann—6834 metres, or 22,422 English feet.

30The inconvenience of using a periphrasis for the name of so important a country may warrant my adoption of the obvious name Argentaria in place of Argentine territory, or Argentine Confederation, and I shall adhere to the shorter designation in the following pages.

31It is quite possible that the bird which I took for the black albatross was the giant petrel, common, according to Darwin, in these waters, and closely resembling an albatross.

32See an interesting paper in the Journal of Botany for July, 1884.

33The estimates given by Pissis do not rest on accurate observations, and seem to me exaggerated. I should be inclined to reckon the difference of height of the snow-line between the extreme stations as nearer to two thousand than to three thousand feet.

34I am not aware that the concurrent conclusions as to the height of this mountain have been verified by accurate observations, but the height commonly given appears to be a close approximation to the truth.

35“Flora Antarctica,” vol. ii. p. 289.

36See Appendix B.

37It is unfortunate that the Spaniards who had the naming of so large a part of the American continent should have shown so little inventive faculty. When they did not adopt a native name for a river, they rarely got beyond Red River, Black River, or Big River, and wherever we turn we encounter a Rio Colorado, a Rio Negro, or a Rio Grande.

38The constant inconvenience of employing such cumbrous expressions as Argentine Confederation or Argentine territory for a state of such vast extent and such yearly increasing importance must be felt by every one who has occasion to speak or write about this region of America. I trust that I shall be forgiven if in this book, as well as elsewhere, I have taken the liberty of applying a single name, which has nothing about it so strange as that it should not long since have come into use.

39The ParanÀ, with its great tributary the Paraguay, drains an area of more than 1,100,000 square miles; the basin of the Uruguay is reckoned at 153,000 square miles.

40The term provinces, commonly applied to the federated States, is misleading, and should be laid aside.

41Much information respecting this country is to be found in a volume entitled, “The Argentine Republic,” published in 1876 for the Centenary Exhibition at Philadelphia. It contains a series of papers prepared by Mr. Richard Napp, assisted by several German men of science.

42Dr. Hann (“Klimatologie,” p. 657, et seq.) has discussed the causes of the prevalent high barometric pressure on both coasts of temperate South America, and has shown that in winter the area of maximum pressure moves northward towards the Tropic of Capricorn.

43The species common here is allied to T. stricta, but is not, I think, identical.

44The best general account of the geology of Brazil that I have seen is contained in a short paper by Orville A. Derby, entitled, “Physical Geography and Geology of Brazil.” It was published in the Rio News, in December, 1884, and, through the kindness of Mr. Geikie, i have seen a reprint in the library of the School of Mines.

45Proceedings of the Royal Geographical Society for 1879, p. 564.

46See his valuable work, “Climats, GÉologie, Faune et GÉographie Botanique du Bresil.”

47“Klimatologie,” p. 382.

48Darwin’s estimate of the height was one thousand feet, while Professor Moseley gives double that amount. I incline to think that the lower figure is nearer to the truth.

49I borrow this statement from the excellent “Lehrbuch der Klimatologie,” by Dr. Julius Hann. Stuttgart, 1883.

50See Reports of the British Association for the Advancement of Science for 1882, pp. 451–453.

51It is remarkable that there is no reference to the investigations of M. de St. Robert, and the formula deduced from them, in the article on the “Barometrical Measurement of Heights,” in the new edition of the EncyclopÆdia Britannica.

52Published by the War Department, United States Army, Professional Papers of the Signal Service, No. xv.

53Air nearly saturated with vapour is lighter than air relatively dry; and hence it may happen that, when a current of moist air meets one relatively dry, it will flow over the latter if they are nearly at the same temperature, but if the drier current be much warmer, it may flow beneath it.

54On this subject see Handbuch der Klimatologie, by Julius Hann, pp. 141, et seq. See also Tables I. and II. in a report on thermometric observations in the Alps, by J. Ball, in Reports of the British Association for the Advancement of Science for 1862, pp. 366–368.

55See “Die Barometrischen HÖhenmessungen und ihre Bedeutung fÜr die Physik der AtmosphÄre,” Leipzig, 1870, by R. RÜhlmann.

56I use the term “eccentricity” in the popular sense, to express the distance of the focus from the centre of the ellipse.

57Viewed in the light of Mr. Langley’s recent researches on solar radiation, all these numerical determinations are probably far from the truth; but the errors do not much affect the present argument.

58The observations at Stanley Harbour, which are those adopted by Dr. Hann (Klimatologie, p. 697), show temperatures notably lower than those recorded for a place in the islands lying farther south, which are given in the Zeitschrift der Œsterreichischen Gesellschaft fÜr Meteorologie, vol. v. p. 369. The mean of the two is probably nearly correct.

59These figures are derived from the tables given in the Anales de la Oficina Meteorologica Argentina, by B. Gould, vol. iii. The figures show a considerable amount of annual variation. The monthly means of the six months from February to July, 1879, exceed those of the same period in 1878 by more than 2° Fahr.