328 Proceedings of Royal Society of Edinburgh. [sess. 
Atwater developed this method of investigation. The numerous 
reports submitted by him to the United States Department of 
Agriculture give his results along three lines of study : 
(1) With a respiration calorimeter ; 
(2) Chemical analyses of food-stuff's ; and 
(3) Dietaries of representative classes of the community. 
He suggests as a dietary standard: proteids, 125 grammes; 
fats, 125 grammes ; and carbohydrates, 450 grammes per man 
per diem. The noticeable difference from the Voit standard is the 
increase in fats and the diminution in the amount of carbo- 
hydrate. Atwater specially emphasises the fact that no allowance 
is made for differences in digestibility, and that personal idiosyn- 
crasy is not considered. 
Professor Chittenden of Yale has recently carried out an 
elaborate investigation into the amount of food required in 
health. These researches are published in book form, Physio- 
logical Economy in Nutrition. Chittenden’s interest in the 
subject was aroused by the contention of Mr Horace Fletcher 
that a high standard of health could be maintained on a low 
proteid intake. Chittenden’s criticisms of the former work on 
this subject are : (1) that the chemical analyses are not accurate, 
and (2) that the amount of food ordinarily taken is by no means 
necessarily the amount required. 
To eliminate possible fallacies in the work, observations were 
made over several months with men of different classes and 
nationalities engaged in different work. The three classes of 
men studied were : (1) professional men, (2) soldiers, and 
(3) college athletes. 
The quantity of food taken was gradually reduced : not only 
proteids, but fats and carbohydrates were diminished. The 
body-weight fell while the change of food was being made, but 
it soon became stationary. The nitrogen equilibrium was then 
tested, and a slight plus balance was found. Systematic tests of 
strength showed an increased vigour, and the subjects of the 
experiment professed to be in improved health. The amount 
of proteid taken in these cases was only about oue-third the 
ordinarily accepted standard. 
Chittenden concludes his report by pointing out that, if health 
