77 
originally contained : water, therefore, cannot be considered as the 
sole food of plants. 
If water does not constitute, alone, the food of plants, it is, how- 
ever, absolutely necessary to their existence ; serving as the vehicle 
by which those substances, from which they derive a larger portion 
of nutriment, are conveyed through their system of vessels. Water 
is also, most probably, so acted on in the vegetable system, that it 
is resolved into its two simple principles, hydrogen and oxygen : 
both of which are well known to exist in almost every vegetable 
substance ; and are presumed to be absolutely necessary to the 
performance of various important functions, in the vegetable 
economy. 
In all vegetables is discovered, by analysis, a considerable por- 
tion of carbon ; particularly in those, in which the woody fibre is 
most abundant : thus, in trees, it constitutes, upon an average, 
nearly a fifth of their whole substance. Analysis has also mani- 
fested, that the water which drains from a dung-hill, and all fertile 
soils, contain a very large proportion of carbonaceous matter. That 
carbon, therefore, is a food of plants, appears to be established 
almost beyond a doubt. 
How indispensible to vegetation are the various combinations of 
earths, must be evident to every one. They form a proper medium 
by which the due quantity of water is regularly diffused, and ad- 
ministered, to the various plants which the ground bears. They 
also form a mass sufficiently yielding to allow the ready extensions 
of the roots of plants, in every direction; at the same time pos- 
sessing sufficient tenacity and firmness, to secure to them stability 
in their situation. 
But chemical analysis has demonstrated, that all plants contain 
more or less of earth in their composition. Saussure, jun. planted 
several trees in a granitic, and in a calcareous soil : and after ana- 
lysing the soil, analysed the trees ; when he found, that the trees 
