All this done, I am afraid that there can be little doubt that the more general climatic investigations will be long and vexatious. Even in South America, with extremely favourable conditions, the result is far from being definite. Glancing over the table published by M. Becquerel in his book on climates, from the observations of Humboldt, Hall, Boussingault, and others, it becomes evident, I think, that nothing can be founded upon the comparisons therein instituted; that all reasoning, in the present state of our information, is premature and unreliable. Strong statements have certainly been made; and particular cases lend themselves to the formation of hasty judgments. "From the Bay of Cupica to the Gulf of Guayaquil," says M. Boussingault, "the country is covered with immense forest and traversed by numerous rivers; it rains there almost ceaselessly; and the mean temperature of this moist district scarcely reaches 78.8 deg. F.... At Payta commence the sandy deserts of Priura and Sechura; to the constant humidity of Choco succeeds almost at once an extreme of dryness; and the mean temperature of the coast increases at the same time by 1.8 deg. F."[57] Even in this selected favourable instance it might be argued that the part performed in the change by the presence or absence of forest was comparatively small; there seems to have been, at the same time, an entire change of soil; and, in our present ignorance, it would be difficult to say by how much this of itself is able to affect the climate. Moreover, it is possible that the humidity of the one district is due to other causes besides the presence of wood, or even that the presence of wood is itself only an effect of some more general difference or combination of differences. Be that as it may, however, we have only to look a little longer at the table before referred to, to see how little weight can be laid on such special instances. Let us take five stations, all in this very district of Choco. Hacquita is eight hundred and twenty feet above Novita, and their mean temperatures are the same. Alto de Mombu, again, is five hundred feet higher than Hacquita, and the mean temperature has here fallen nearly two degrees. Go up another five hundred feet to Tambo de la Orquita, and again we find no fall in the mean temperature. Go up some five hundred further to Chami, and there is a fall in the mean temperature of nearly six degrees. Such numbers are evidently quite untrustworthy; and hence we may judge how much confidence can be placed in any generalisation from these South American mean temperatures.