336 Proceedings of Royal Society of Edinburgh. [sess. 
The method employed in making the present studies was that 
adopted by Atwater and his colleagues. A detailed account of 
this is given by Bryant in “ Some Results of Dietary Studies 
in the United States ” (reprint from the Yearbook of the Depart- 
ment of Agriculture for 1898). 
An inventory is made of all the food in the house. Everything 
is weighed accurately on a tested balance. Each article of food 
as it is purchased is weighed and added to the inventory. At the 
end of the study, everything is again weighed. The amount 
actually used is then calculated. But all the food is not actually 
consumed. The “ refuse ” (that is, the inedible part) is allowed for 
in the chemical analyses, and so in the determination of calorific 
value. A certain amount of the food is also inevitably wasted. 
All the “ waste ” (that is, nutritive material which has not been 
actually consumed) is also sorted out, and it is carefully weighed 
and deducted from the food used. In this way the actual con- 
sumption during the period of study is arrived at. The cost of 
the food is also taken into account, and the expenditure on this 
item per man per diem is calculated. In giving the cost of the 
“waste,” an estimate was obviously all that could be given. A 
note is also made when anyone is absent from a meal, and the 
presence of guests is also taken into consideration, in order to 
make the results as accurate as possible. 
The analyses used were chiefly taken from Atwater’s “ Chemical 
Composition of American Food-Stuffs ” (. Bulletin 28 — revised 
edition — of the U.S. Department of Agriculture). Konig’s Ghemie 
der menschlichen Nahrungs und Genussmittel , Wynter Blyth’s 
Foods , their Composition and Analysis , Noble and Firth’s Text- 
book of Hygiene , Mitchell’s Flesh Foods , and Hutchison’s Food 
and Dietetics were also consulted. Several analyses made by 
Dr Craufurd Dunlop in the Laboratory of the Royal College of 
Physicians, Edinburgh, were also used. In two cases, galantine 
and haggis, the quantities used were so small that an analysis 
was not made, but an estimate was made from their composition. 
The present studies were conducted for one week — during the 
month of February in one case, and March in the others. 
In Studies A, B, C, and D the students in residence were men. 
All the servants were women, and they were allowed for by 
