Effect upon Appetite of Air of Occupied Rooms. 143 



In Series XI there was a slight excess of food consumed on the 

 no-ventilation days. This series was however rendered practically 

 valueless by the fact that after it was under way we discovered 

 that religious dietary laws, observed but loosely at first, but 

 gradually with more strictness, had prevented the use of certain 

 articles of the diet. This influence was quite impossible to measure 

 or to balance. The results however are included for completeness. 



The other four series show a consistent excess of food con- 

 sumption on the ventilation days. In view of the fact that these 

 series represent averages of 71, 80, 196, and 160 meals respectively 

 we believe them to be significant. In Series III and VII where 

 the food left by each subject was separately weighed every one 

 of the 8 different subjects ate more on the ventilation days. 



The reliability of these figures based on the number of observa- 

 tions and their departure from the mean has been determined 

 mathematically by Mr. W. A. McCall, psychologist on the staff 

 of the Commission. 



The chances that the differences in calories consumed between 

 ventilation and no-ventilation days may be zero or may favor the 

 other condition, are expressed in Table II. 



table 11. 



Series. Chance of Different Result. Interpretation. 



Ill 0.38 chance in 100 Very reliable. 



VII 1.0 " " " Highly reliable. 



X 9.0 ' Probable. 



XI 18.0 



XII 12.0 " " " 



This effect on appetite of the absence of ventilation, though 

 slight, is apparently a persistent one. At the beginning of an 

 experiment there is but little difference noted on the two types of 

 days. As the novelty of the environment and the meal wears off 

 however, the effect of breathing the same air over and over again 

 seems to exert an increasing influence. This action is illustrated 

 in Series XII, where the average excess of food consumed on 

 ventilation days was 6.8 per cent. During the first two weeks 

 this excess was but 2.8 per cent. In the last two weeks it amounted 

 to 10.8 per cent. 



A separate computation of the bread and butter consumption at 



