36C ISIDOK GEEEXWALD 



in cities, which always lead in the consumption of wheat as compared 

 with rye, barley or oats. But Sherman (b) (1018) has collected figures 

 showing that in Russia in the period from 1804 to 1809, there were 1.82 

 bushels of wheat and 4.76 bushels of rye consumed per person per annum. 

 During the following five years, these figures were 2.40 and 4.78, respec- 

 tively, and from 1011 to 1013 were 2.80 and 4.47. The magnitude of 

 these changes in a coiintry with, relatively, so small an urban population 

 indicates that the use of wheat was increasing in the country as well as 

 in the cities. 



Dietary Studies 



Manner of Conducting Studies and of Calculating Results. The 

 amount and composition of the food consumed per person may be deter- 

 mined in various ways. As in the calculations of Sherman and of 

 Holmes, the total amount of food raised in and imported into a given 

 area, less that exported and used otherwise than as human food, may be 

 divided by the number of people. The method is, at best, only an 

 approximation but it serves very well to indicate the relative importance 

 of the different food materials. Next, studies may be made of groups 

 such as families, eating clubs, public institutions, military and naval 

 organizations, etc., in which the total amount of food is weighed and, 

 with or without deduction for waste, is divided by the total number of 

 people participating. Finally, the food consumed by an individual may 

 be weighed. 



The composition of the food may be calculated in different ways. 

 For such gross calculations as those relating to the food consump- 

 tion of an entire city or country, it is obvious that only the average 

 of a considerable number of analyses can be used. In the other cases, 

 the same procedure may be followed but it is also possible, and prefer- 

 able, to secure sufficient of most, of the materials to last through all, 

 or a considerable part, of the experiment and to analyze representative 

 samples of these. Still greater accuracy may be obtained by taking to 

 the laboratory and analyzing a composite sample of the food consumed, 

 weighed as served, and mixed in exactly the same proportion as consumed. 



Assuming the trustworthiness of the subjects, many factors influence 

 the accuracy and significance of the results. Studies made under labora- 

 tory conditions with accurate weighing and analysis of the food are the 

 most accurate but are obviously expensive and difficult to make in large 

 number for a long period. Studies made in the home can be carried out 

 in larger number, can be continued for a longer period and come nearer 

 to "normal" conditions but the accuracy of the weighings and the ap- 

 plicability of the analytical data employed are more questionable. Daily 

 and seasonal variations in food consumption must also be considered. The 



