-------------+-------+-------+-------+-------+-------+-------+-------

At the time when these analyses were made, a method for the quantitative
estimation of tryptophane had not been devised, although one is now
available. The addition of the percentages of tryptophane and of other
amino-acids for which methods of determination are not yet known, would
bring the total, in each case, more nearly up to the full 100 per cent.
These data will serve to show how widely the different plant proteins vary
in the proportions of the different amino-acids which they contain. Animal
proteins have been found to be still more variable in composition.

In the use of the proteins as food for animals, it appears that the
different amino-acids are in some way connected with the different
physiological functions which the proteins have to perform in the animal
body: thus, _tryptophane_ is absolutely essential to the maintenance of
life, but does not promote growth; _lysine_, on the other hand, definitely
promotes growth, so that animals which have been maintained without any
increase in weight for many months immediately begin to grow when furnished
with a diet in which lysine is a constituent; while _arginine_ seems to be
definitely associated with the reproductive function; and _cystine_, with
the growth of hair, feathers, etc. It is not known whether there is any
similar relation of amino-acids to the functions of different proteins in
plant metabolism.

The separation of the individual amino-acids from the mixture which results
from the hydrolysis of any given protein is a long and tedious process and,
at best, yields only moderately satisfactory results. For that reason, it
has recently been almost entirely abandoned in favor of the separation
devised by Van Slyke, which divides the total nitrogenous matter in the
mixture resulting from the hydrolysis of a protein into the following
groups; ammonia N, humin (or melanin) N, cystine N, arginine N, histidine
N, lysine N, amino N of the filtrate, and non-amino N of the filtrate.
These groups can be conveniently and fairly accurately separated out of the
hydrolysis mixture, by means of various precipitating agents, and the
quantity of N in the several precipitates determined by the usual Kjeldahl
method. The actual process for these separations need not be discussed
here, as it is given in detail in all standard text-books dealing with the
methods of biochemical analysis. The distribution