o hasten the change which had been inaugurated by arousing the
natural appetite. Suggestion was introduced merely because the
experiment was limited in time. In no case was it allowed to override
the dictates of appetite.

Careful records of the amount of food taken and the constituents in (1)
protein, (2) fats and (3) starches and sugars, were kept for each man
for each day. In order to avoid weighing the food at the table and the
annoyance which such a procedure involves, the food was all weighed in
the kitchen and served in definite portions of known food value. From
the records thus supplied, it was easy, by means of a "mechanical diet
indicator" devised for the purpose, to find the proportions of food
elements. The first result of the experiment was a reduction in the
amount of protein consumed.

During the first four weeks, the men consumed an average of from 2,760
to 3,030 calories per day, of which 120 to 240 were in the flesh foods,
such as meats, poultry, fish and shell-fish, and that 2.4 to 2.7
calories of protein were ingested for each pound of body-weight.
Translating Professor Chittenden's figures for the physiological
requirement of ingested protein, we find it to be from 1.3 to 1.7
calories per pound of body-weight. Thus the men were at this time
consuming nearly double the Chittenden allowance. During the last four
weeks of the experiment all these magnitudes were lower. The per capita
calories ranged from 2,220 to 2,620, of which only 40 were in flesh
foods, and the protein had fallen to 1.4 to 1.9 calories per pound of
body-weight, which corresponds closely to the Chittenden standard.

Gymnasium tests were made at the beginning, middle and end of the
experiment. These tests were of two kinds--tests of strength and tests
of endurance.

During the first period there was a slight increase in strength (from an
average "total" strength of 1,076 to 1,118), and during the second
period a slight fall to 995, which is about 12 per cent. from the
mid-year's 1,118, and about 8 per cent. from the original 1,076. Thus
the strength of the men remained nearly stationary throughout the
experiment.

It is fortunate that the strength of the men remained so nearly
stationary; for it demonstrates the more clearly that the increase in
endurance which will be shown below was an increase in endurance
_per se_, and not in any degree due to an increase in strength. Strength
and endurance are entirely distinct and should