148 STORRS AGRICULTURAL EXPERIMENT STATION. 
All weights and measurements were made and recorded by 
the writer, who, however, checked his work by noting the 
weights unused at each weighing. Since the amounts weighed 
at a time were comparatively small, the chance for serious 
error was not so great as might appear. A further check, 
though a rough one, was afforded by securing as accurately as 
possible the original weight of each package at the time of 
purchase and then noting the weight of the empty package. 
As in most cases there was no waste, the difference was obvi- 
ously the weight of food consumed. 
The amounts consumed of liquids, including milk, chocolate, 
beef tea, birch beer and root beer were estimated by measure- 
ment, and the weights then computed. A large amount of 
water also was drunk, of which no account was taken. 
Composition of food materials. —Theoretically, in studies of this 
kind, each food material used should be sampled and analyzed. 
This was obviously impracticable under the circumstances, and 
recourse was had to analyses of similar food materials compiled 
elsewhere.* It is recognized that the composition of such 
materials varies widely. This is especially true in the case of 
cooked foods, because of differences in the recipes used in their 
preparation. Consequently, the results of these studies must 
be considered less reliable than those based on actual sampling 
and analysis. Nevertheless, they serve the purpose for which 
they were made, and give a fair indication of the food con- 
sumption of the subjects. 
The following table gives the percentage composition and 
energy of each material as employed in the computations. 
The table also includes the amount paid per pound, the mate- 
rials taken from home at the start being estimated at the pre- 
vailing White Mountain prices. . 
A study of this table reveals anew the fact that the cost of 
food is a very unreliable indication of its nutritive value. 
Thus, the malted milk tablets, a proprietary preparation for 
which extraordinary claims are made and selling at the rate of 
$3.50 per pound, actually contain somewhat less of both pro- 
tein and energy than the deviled ham which was procured in 

* U.S. Dept. Agr., Office of Experiment Stations Buls. 28, 152; Division of Chemis- 
try Bul. 13, part9; Conn. (New Haven) Stations Rpt. 1903, 
Lab. Inland Rev. Dept., Canada, Bul. 63. 
