CLIMATE OF NORTHERN PERU.

47

Before returning to the Islay, I had the advantage of a short conversation with the very intelligent gentleman who acts as British consular agent at Payta, and whose ability would perhaps be seen to advantage in a more conspicuous post. The information received from him fully confirmed the impressions formed during my short excursion. The appearance of the gullies that furrow the seaward face of the plateau sufficiently showed that, however infrequent they may be, heavy rains must sometimes visit this part of the coast. I now learned that, in point of fact, abundant rain lasting for several days recurs at intervals of three or four years, the last having been seen in the year 1879. As happens everywhere else in the arid coast zone, extending nearly two thousand miles from Payta to Coquimbo in North Chili, abundant rainfall is speedily followed by an outburst of herbaceous vegetation covering the surfaces that have so long been bare. During the long dry intervals slight showers occur occasionally a few times in each year. These are quite insufficient to cause any general appearance of fresh vegetation, but suffice, it would seem, to maintain the vitality of the few species that hold their ground persistently. The ordinary supply of water in Payta, obtained from a stream descending from the Andes seventeen miles distant, is carried by donkeys that are despatched every morning for the purpose. There was something quite strange in the appearance of a few bundles of fresh grass which we saw in the plaza. They had come that morning by the same conveyance for the support of the very few

domestic animals that it is possible to keep in such a place.

The problems suggested by the singular climatal conditions of this region of South America have not, I think, been as fully discussed as they deserve to be, and I here venture on some remarks as a contribution to the subject.

The existence of the so-called rainless zone on the west coast of South America is usually accounted for by two agencies whose union is necessary to produce the result. The great range of the Andes, it is said, acts as a condenser on the moisture that is constantly carried from the Atlantic coasts by the general westward drift of the atmosphere in low latitudes. The copious rainfall thus produced on the eastern slopes of the great range leaves the air of the highlands of Peru and Bolivia relatively dry and cool, so that any portion that may descend to the coast on the western declivity tends to prevent rather than to cause fresh aqueous precipitations. Meanwhile the branch of the Antarctic Ocean current known as the Humboldt current, which sets northward along the sea-board from Western Patagonia, is accompanied by an aërial current, or prevailing breeze, which keeps the same direction. The cold air flowing towards the equator, being gradually warmed, has its capacity for holding vapour in suspension constantly increased, and is thus enabled to absorb a large portion of the vapour contained in the currents that occasionally flow inland from the Pacific, so that the production of rain is a rare event, recurring only at long intervals. Admitting the plausibility of this explanation, a first diffi

CAUSES OF THE ARID COAST CLIMATE. 49

culty presents itself. If the Andes act as a barrier against the vapour-laden atmosphere of eastern tropical America throughout Peru, Bolivia, and Northern Chili why, it may be asked, do they fail to perform the same function in Ecuador and Colombia? Whence the absolute contrast in point of climate that exists between these regions? Why is the littoral zone between the Gulf of Guayaquil and that of Panama, a distance of some eight hundred miles, not merely less dry than that of Peru, but actually more moist that most parts of the coast of Brazil or Guiana ?

Some answer may, I think, be given to these questions. In the first place, comparing the orography of Peru and Bolivia with that of Ecuador, some important differences must be noted. In Eastern Peru, as is at once shown by the direction of the principal rivers, we find no less than four parallel mountain ranges, increasing in mean elevation as we travel from east to west. The westernmost range, to which in Peru the name Cordillera is exclusively applied, does not everywhere include the highest peaks, but has the highest mean elevation. The second range, exclusively called Andes in Peru, rivals the first in height and importance. I know of no collective names by which to distinguish the third range, dividing the valley of the Huallaga from that of the Ucayali, nor the fourth range, forming the eastern boundary of the latter stream. In South Peru and Bolivia the mountain ranges are less regularly disposed, but cover a still wider area; and throughout the whole region it is obvious that the warm and moist currents drifting slowly westward have to traverse a zone of lofty

E

mountains varying from four to six hundred miles in width, and can carry no moisture available to produce rain on the western seaboard. In Ecuador the two principal ranges-the Cordillera and the Andesare much nearer together than they usually are in Peru, and no parallel ranges flank them on the east. The numerous tributaries of the Maranon flow in a tolerably direct course east or south-east, many of them rising within a hundred and fifty miles of the Pacific coast. It follows that the atmospheric currents meeting less preliminary obstruction reach the eastern slopes of the main range still very heavily charged with vapour. In crossing the barrier a large portion of the burthen must be deposited; but it is probable that a large amount is nevertheless carried to the western side of the range.

It may be said that this explanation, whatever it may be worth, cannot apply to the territory of Colombia, where the Andes are broken up into at least three lofty ranges, and the mountains cover as wide a space as they do in Peru. My impression is that the abundant supply of moisture on the west coast of Colombia arises from a different source. The effects of the Isthmus of Panama as a barrier against atmospheric currents must be absolutely insignificant, and I have no doubt that those which flow eastward along the coast of the Caribbean Sea are in part diverted south-east and south along the west coast of Colombia,

There can, however, be little doubt that in determining the climate of the west coast the influence of the Humboldt current, and of the cool southerly

INFLUENCE OF THE HUMBOLDT CURRENT. 51

breezes that accompany it, is far greater than that of the disposition of the mountain ranges. A glance at the map shows that about the fifth and sixth degrees of south latitude the direction of the coast undergoes a considerable change. On the voyage from Panama, we had hitherto steered somewhat west of south; henceforward our course lay between south-south-east and south-east. All the currents of the ocean and atmosphere, whose existence arises from the unequal distribution of heat on the earth's surface, vary somewhat in their course throughout the year with the changes of season, and this doubtless holds good on the American coast. I believe, however, that both the sea and air currents from the south are normally deflected away from the coast at the promontory of Ajulla (sometimes written " Ahuja "), a short distance south of Payta. A further portion is again deflected westward at Cape Parinas, north of which headland they seem not to be ordinarily met. I infer, however, from the testimony of seamen, that at some seasons they are felt near the coast as far north as the equator, and even beyond it. This inference was confirmed by observing the parched appearance of the seaward slope of Cabo Sta. Elena, north of the Gulf of Guayaquil, which apparently does not fully share in the frequent rains that elsewhere visit the coast of Ecuador.

Whatever force there may be in the above suggestions, I confess that they do not seem to me adequate to account for the extraordinary difference of climate between places so near as Payta and Tumbez-not quite a hundred miles apart-and I trust that further