obtained to the nature of the deficiencies. His early results in this
direction confirmed the results of other investigators, animals lived no
longer on these diets than when allowed to fast. What was missing? Up to
1911 the main result of these experiments had been to call attention to
the peculiar deficiencies of cereals and especially in mineral salts, but
without unlocking the mystery.

These collateral investigations show how in all parts of this country and
on the other side of the ocean events were marching toward the same goal.
The year 1911 then is a significant epoch, for from this time the various
independent efforts began to link up and the next few years carried us far
toward the goal.

In 1912 McCollum was working with a mixture consisting of 18 per cent.
purified protein in the form of milk curd or casein, 20 per cent. lactose
or milk sugar, 5 per cent. of a fat and a salt mixture made up to imitate
the salt content of milk. The remainder of that mixture was starch. With
this mixture McCollum found that growth could be produced if the fat were
butter fat but not if it were olive oil, lard, or vegetable oils of
various sorts. Carrying out the lead here suggested he tried egg yolk
fats. They proved as effective as butter fat.

[Illustration: FIG. 1. COMPOSITE CHART OF MCCOLLUM AND DAVIS PUBLICATIONS

I (from _Journ. Biol. Chem._, 1913, xv, 167). This chart shows the
effect in period III of the addition of an ether extract of egg, 1 gram
being given every other day. The diets for periods I-IV were as follows:

Periods . . . . . . . . . . . . . . . I II III IV
Salt mixture . . . . . . . . . . . . 6 6 6 6