4 I'ls/ier — T/ie Ejf'ect of Diet on Endurance. 



the order of tlie proportion of proteid. The inen were then asked, 

 when and only wlien the a])petite was entirely xc ill in g, to choose the 

 better and puicr foods and the low proteid foods in preference to 

 those high in proteid. In tliis way the men gradually sliifted their 

 diet upward in the two lists, and thereby pursued a little faster the 

 same direction in which they had already been found to be uncon- 

 sciously moving under the influence of thorough mastication and 

 implicit obedience to appetite. 



It would too greatly lengthen this report if any attempt were made 

 to repeat in detail all the specific advice given to the experimenters 

 under Rule '-i. What has been said covers in a general waj"- all the 

 points except the advice (subject always to the consent of appetite) 

 to eat light and quickly digested suppers in order to go to bed on an 

 empty stomach. 



Careful record of the amounts of food eaten and the constituents 

 of proteids, fats and carbohydrates was kejjt for each man each day, 

 certain days being omitted if for any reason the record was incom- 

 plete, as when, for instance, the men were out of town or took their 

 meals away from the club.' To avoid weighing at the table, the 

 food was all weighed in the kitchen and served in "standard portions" 

 of 100 calories each, or simj^le fractions or multiples thereof, and the 

 men merely recorded the number of portions eaten. The proportions 

 of proteids, fats and carbohydrates were found by means of the 

 writer's " Mechanical Diet Indicator. " '^ Atwater and Bryant's 

 tables were used as a basis for calculation. For the first few weeks 

 the figures were probably subject to some errors, and in all cases 

 more or less guessing had to be practiced with reference to the 

 amount of lean and fat of meats ; but the influence of any errors on 

 the results must necessarily be small, because meat supplied, at 

 the highest, o\)\y a small fraction of the total calories. It is believed 

 that the results are in general correct to two significant figures. 



For the first two weeks of the first half of the experiment, the men 

 ate in their ordinary way. During the following eight weeks they 

 masticated moi'e thoroughly and followed the leadings of taste more 

 carefully. Most persons, while nominally following taste, are largely 



^ The number of days each week on which the record of diet was kept was sel- 

 dom under six. 



-For a description of this instrument, seethe writer's "A New Metliod of 

 Indicating Food Values," American Journal of Fhyaioloyy, April, 1906. For a 

 description of its practical uses see "A Graphic Method in Practical Dietetics," 

 Jour, of the Amer. Med. Assoc, Apr. 20, 1907. 



