The most probable theory seems to be that put forward by Noellner. The origin of the nitric acid is, according to him, to be ascribed to the decay of great masses of sea-weed, which, by means of hurricanes such as are still prevalent in these districts, were driven into the lagoons. The chief difficulty in the way of accepting this theory is the enormous quantity of sea-weed required to produce the millions of tons of nitric acid these deposits contain. It must be remembered, however, as bearing upon this point, that the occurrence of gigantic masses of sea-weed in the Pacific Ocean[210] is by no means uncommon even at the present time. If, to understand the formation of coal, we must suppose the Carboniferous period to be one during which exceptionally luxuriant growth of vegetation took place, we may be permitted to suppose a similar luxuriant growth of sea-weed during the formation of the nitrate deposits. Very strong confirmation of the truth of this theory is further afforded by the presence in large quantities, in the raw nitrate of soda, of iodine, a substance characteristic of sea-weed; while pieces of sea-weed still undecomposed are met with here and there. On the whole, therefore, this theory, while not free from difficulties, seems to be the most worthy of acceptance as regards the origin of the nitrate deposits.[211] _Appearance of Nitrate-fields._ Having thus discussed the origin of the nitrate-fields, we may now give a more detailed description of their appearance. The chief deposits at present being worked are those lying in the Pampa de Tamarugal, in the province of Tarapaca. They stretch to a distance of thirty or forty miles inland, from Pisagua southwards to somewhat beyond the town of Iquique. This huge desert, as has been already indicated, seems to be entirely destitute of all vegetation and animal life. Even in the immediately adjoining country the only kind of vegetation that seems to grow is a species of _acacia_. The few streams that are found in this neighbourhood are entirely fed by the melting snow from the Cordilleras. Darwin describes the appearance presented by these pampas as resembling "a country after snow, before the last dirty patches are thawed." The _caliche_, or raw nitrate of soda, is not equally distributed over the pampas. The most abundant deposits are situated on the slopes of the hills which probably formed the shores of the old lagoons. An expert can tell from the external appearance of the ground where the richest deposits are likely to be found. The _caliche_ itself is not found on the surface of the plain, but is covered up by two layers. The uppermost, known technically as _chuca_, is of a friable nature, and consists of sand and gypsum; while the lower, the _costra_, is a rocky conglomerate of clay, gravel, and fragments of felspar. The _caliche_ varies in thickness from a few inches to 10 or 12 feet, and rests on a soft stratum of earth called _cova_.