practically the same amount of feed, and digestion tests showed that there
was no difference in the digestibility of the different rations. Exercise
was provided by allowing them the run of a yard free of all vegetation. It
was a year or more before any distinct change appeared in the different
groups. At that time the cornfed animals were in fine condition. On the
contrary, the wheat-fed group were rough coated, gaunt in appearance and
small of girth. The oat-fed group were better off than the wheat-fed but
not in so good shape as the corn-fed. In reproduction the corn-fed animals
carried their young well. They were carried for the full term and the
young after birth were well formed and vigorous. The wheat-fed mothers
gave birth to young from three to five weeks before the end of the normal
term. The young were either born dead or died within a few hours after
birth. All were much under weight. The oat-fed mothers produced their
young about two weeks before the normal period. Of four young, so born,
one was born dead, two so weak that they died within a day or two and the
fourth was only saved by special measures. The young of the oat-fed
mothers were of nearly the same size, however, as those of the corn-fed
mothers. After the first reproduction period, the mothers were kept on
this diet another year and the following year repeated the same process
with identical results. During the first milk-producing period the average
production per day was 24.03 pounds per day for the corn-fed, 19.38 pounds
for the oat-fed, and 8.04 pounds for the wheat-fed. During the second
period it was 28.0, 30.1, and 16.1 pounds per day respectively during the
first thirty days.

Every chemical means was now employed to determine the causes of these
differences and without success. McCollum then decided to attempt to solve
the problem by selecting small animals (the rat was used) and experiment
with mixtures consisting of purified proteins from different sources,
combined with fats, carbohydrates and mineral salts until a clue was