GROWTH AFTER INANITION FOR VARIOUS PERIODS. 19 



than thirteen or fourteen weeks, and keep .them alive; so as a 

 rule they were permitted to increase slightly in body weight after 

 that time. Aron ('n) and Jackson ('15) similarly found it 

 increasingly difficult to hold animals at constant body weight as 

 the experiment progressed. 



Individual weight records were kept, the individuals rats being 

 identified by staining the integument with an aqueous solution 

 of picric acid. The weight of the rats was always recorded 

 immediately before feeding. The test rats were weighed daily; 

 whereas the controls were weighed at gradually increasing 

 intervals (about once a week after reaching 200 grams body 

 weight). 



Previous to reaching sexual maturity the control males and 

 females were usually separated in order to prevent pregnancy. 

 In the case of the stunted rats it was found unnecessary to separ- 

 ate the sexes while being underfed, for in no case did a pregnancy 

 result, although of litter No. S8 the control female and also those 

 rats refed after one and three weeks of maintenance each bore 

 one litter. 



All the rats were fed on whole wheat (Graham) bread soaked in 

 whole milk. The control rats were given an abundant amount of 

 food, whereas the test rats received during the underfeeding just 

 that amount of food required to hold them at nearly constant 

 body weight. It was observed that the test rats did not eat their 

 entire maintenance ration immediately until after about two 

 weeks of underfeeding. Aron ( ' 1 1 ) noted that at first it was neces- 

 sary to feed a dog its maintenance ration in two portions, but 

 after a short time it was able to eat its entire allowance at a single 

 feeding. Water in abundance was given to all the animals. 



It was found, as was observed by Jackson ('15), that the rats 

 are for a time held at constant body weight upon a gradually 

 diminishing ration. It was further observed, however, that 

 after about fifty days of maintenance there is no further decrease 

 in the necessary amount of food. In one litter, the average 

 amount was found to remain practically constant from the 

 6oth to the i2Oth day of the experiment. 



After being held at maintenance, the test rats were generously 

 fed for different periods as indicated in Table I. At the end of 



