28 . THE NATUEE OF THEIU FOOD. 



The natural food of plants consists of carbon in the state of 

 carbonic acid, of nitrogen, certain earths and salts,- and water. 

 The latter, if distilled, has little power, by itself, of sustaining 

 vegetable life : but, as in nature it is universally mixed with 

 various other substances, it conveys to the roots the nutritious 

 matters that are required ; and it furnishes, by its decomposi- 

 tion, a considerable supply of the oxygen consumed in the 

 formation of carbonic acid, as well as much of the hydrogen that 

 is assimilated by plants. It has been proved, experimentally, 

 that plants cannot long exist upon pure water ; but, if they are 

 so circumstanced as to be able to obtain and decompose 

 carbonic acid, they wUl grow in the absence of other matters. 

 It is only, however, when the peculiar principles, whether 

 earthy or saline, on which they naturally feed, are presented to 

 them, that they become perfectly healthy : and especially when 

 they have the means of obtaining nitrogen, which appears, from 

 its great abundance in the youngest parts, to be indispensable 

 to plants upon the first formation of their tissue. 



The researclies of chemists have shown that all rain-water contains 

 ammonia, a compoiind of hydrogen and nitrogen, and thus the source 

 of the nitrogen absorbed by plants was explained. But it has also 

 been shown, especially by M. Barral, that other substances, upon which 

 plants feed, are contained in rain-water to a much greater amount than 

 was suspected. This observer was led, during sis months of 1851, to 

 examine minutely the water collected in the rain-gauges of the 

 Observatory at Paris. His mode of investigation is declared by 

 Messrs. Dumas, Bonssinganlt, Gasparin, Regnault, and Arago, names 

 foremost in French Science, to be free from all objection, and to bear 

 the most severe counter trials to which they could expose it. M. Barral 

 states, that although the quantities of the following substances varied 

 in different months, yet the monthly average, from Jidy to December 

 inclusive, was as follows : — 



