Let us present the reader with some quantitative studies. Turn to page 189 and examine the percentage of amino-acids isolated from various proteins.

What does such a list show? Some of the proteins do not contain the amino-acid glycocoll; others do not contain tryptophane; still others are deficient in lysine and cystine. (Strictly speaking, it is incorrect to speak of a protein as "containing" such and such amino-acids. These amino-acids become evident only after the protein has been decomposed.) To what extent are these amino-acids needed by the organism, and what happens if the only source of protein in the diet is deficient in one or more of these amino-acids?

The answer has already been supplied for gelatin. Unless the missing amino-acid tryptophane (and tyrosine) is added, death occurs.

Now suppose we arrange a diet in such a way as to vary from time to time the protein it may contain. Let us assume that we have fixed upon the type and amount each of fat, carbohydrate, mineral salts; and that we allow plenty of water and plenty of air; and that we adhere to such a diet throughout the series of trials. From time to time, however, we shall replace one protein, say casein, with another, say zein. If the substitution of zein for casein causes ill-effects, we may reasonably assume that this is due to the absence of one or more amino-acids from the diet. Suppose we add one or more of such amino-acids to this otherwise deficient diet and the animals that we experiment with begin to thrive again; are we not justified in concluding that our assumption was correct?

This, in fact, is the procedure that Professor Hopkins of Cambridge, Dr. Osborne of the Connecticut Experimental Station and Professor Mendel of Yale have adopted. The last two investigators in particular, who have done all their work jointly, have enriched this phase of the science with much of value. Let us follow Osborne and Mendel in their path.