BOEIC ACID AND BORAX. 21 



for even if the same kind or character of clothing be worn the varia- 

 tion in weight is very great because _of changes in the hygroscopic 

 condition of the atmosphere. Thus a given amount of clothing would 

 show very different weights on a dry and on a wet day. 



Certain variations in weight from the normal have already been 

 explained, and in the general discussion of the influence of weights it 

 is always advisable to take the average weight of a period of days 

 rather than the separate weight for any one day. In the interpreta- 

 tion of the value of the body weight it should not be forgotten that a 

 loss in weight must not be interpreted to mean always defective nutri- 

 tion, nor a gain in weight be attributed always to conditions favorable 

 to health. The accumulation of an excessive amount of fat is not an 

 evidence of excellent digestion or normal increase. It may be due to 

 a perversion, to some extent, of the processes of assimilation. On the 

 other hand, a loss of weight is not always to be interpreted as indi- 

 cating an unfavorable condition of nutrition, because in persons who 

 indulge in overfeeding or who have accumulated excessive fat for 

 other reasons a diminution of weight may be distinctly favorable to 

 better digestion and health. Nevertheless, in a state of normal equi- 

 librium, when the food supply remains constant any marked variations 

 in weight can not be regarded as wholly normal. 



METHODS OF ANALYSIS. 



The methods of analysis employed were those, in so far as they 

 applied, of the Association of Official Agricultural Chemists, and, in 

 other cases, those commonly used by physiological chemists. 



ANALYSIS OF FOODS- 



So far as possible foods were composited or received in large quan- 

 tities and in such condition that a fair sample could be taken of a lot 

 which would last the table for considerable time. For instance, in 

 the case of soups, vegetables, and fruit products (canned fruit, jellies, 

 and jams) arrangements were made with manufacturers to receive the 

 product of a single kettle canned in such a manner that the contents 

 of the separate cans would be as nearly uniform as possible. In this 

 way one or two cans of each lot were taken as a sample, and a single 

 analysis was made to answer as long as that shipment lasted. A great 

 amount of analytical work which would have attended the examina- 

 tion of each food at each meal was thus avoided. 



Potatoes were cooked without seasoning, and a composite sample of 

 all the potatoes served during the subperiod was subjected to analysis. 

 The bread was purchased of a neighboring baker. Throughout the 

 entire experiment only one variety of bread, of very constant compo- 

 sition, was employed. In all cases the water content of each food for 



