126 STORRS AGRICULTURAL EXPERIMENT STATION. 
products may be considered as representing the cost of digestion 
in terms of food ingredients. From this point of view the total 
feces, including both these products and the undigested residue, 
may be considered as representing the portion of the food un- 
available for the building of tissue and the yielding of energy. 
The difference between the total food and total feces may 
therefore be designated as the availability of the food. In 
this discussion, therefore, the term availability is so used, and 
the coefficients obtained as explained in detail are designated 
throughout as coefficients of availability. 
Availability of nutrients of the fish or poultry alone.—Since 
the object of these experiments was to ascertain the availability 
of the fish and poultry alone, the coefficients for this portion 
of the diet must be determined. For this purpose, the remain- 
ing articles in the diet were those the coefficients of availability 
of which were already known from previous experiments. It 
was assumed that in the series here reported, their availability 
would be unchanged. By the use of these coefhicients, there- 
fore, it was possible to estimate the amounts of each nutrient 
of the feces derived from the foods other than fish or poultry. 
The remainder of the feces could be ascribed to the fish or 
poultry alone. From these amounts and corresponding weights 
of each nutrient in the fish or poultry eaten, the coefficients 
of availability of the fish or poultry could be determined as 
explained above for the total food. 7 
In making these computations it was assumed that 88 per 
cent. of tHe protein of bread, and 97 per cent. of the protein 
of the milk and butter, and in all cases, 95 per cent. of the 
fat was available; or in other words, 12 per cent. of the protein 
of bread and 3 per cent. of the protein of milk and butter, 
together with 5 per cent. of the fat, was not available and 
would be found in the feces. This method may be illustrated 
by the use of figures from one of the experiments. 
Thus, in experiment No. 558 (Table 14), with salmon, 80.8 
grams of the total protein was supplied by the bread of which 
i2 per cent. or 9.7 grams was assumed to be not available and 
so found in the feces. ‘The milk and butter supplied 63.8 
grams of protein, of which 3 per cent. or I.g grams would be 
found in the feces. The total protein in the feces from the 

