342 Canadian Record of Science. 
digestive function of such plants is reduced, do they 
become incapable of fixing carbon and forming the 
ordinary carbohydrate products such as starch and sugar. 
Some of the most notable of parasites are to be found in 
the celebrated banyans of India, which often begin their 
growth in the tops of lofty trees, upon which they feed 
until killed. 
We again find a very large class of plants feeding upon 
the products of organic decay. These contain no chloro- 
phyll, have no proper roots and no leaves, or at most mere 
rudiments of such organs. Like the parasites, they cannot 
appropriate carbon, except in the form of organic com- 
pounds; their existence thus implies their dependence upon 
previous life. They do not liberate oxygen, but eliminate 
carbon dioxide as one of their characteristic products. Such 
plants are designated by botanists saprophytes, and are rep- 
resented by the mold of stale bread and cheese, by the 
common mushroom and puff-ball, and also by the Indian 
pipe, one of our common wild flowers. 
We thus find that any extended consideration of the sub- 
ject with which we are now dealing, must recognise the 
special characteristics of plants in their relation to the 
appropriation of food, but as more detailed statement 
would lead us too far from our main purpose, we shall for 
the remainder of our discussion, confine ourselves to those 
plants in which the digestive function is fully developed, 
and with which we are more largely concerned as the pro- 
ducers of our food. 
The special functions of the various elements appropri- 
ated by the plant, are not at all well understood, but the re- 
sults of investigations so far made, indicate their value in a 
general way and show in what direction other inquiries 
should be made. For the purpose of determining how far 
each element present is essential to growth, we resort to 
special methods of culture, either in water or pure quartz 
sand, under such conditions that the number of elements 
and the exact quantity of each may be known and controlled. 
From such a series of investigations we learn that potash 
