WATER EXCHANGES OF DOG 



31 



procedure is imposed, no muscular exercise is permitted or other 

 unusual influence of which the experimenter is aware; (5) Random 

 days but similar hours are chosen for recoveries from diverse 

 water loads; and (6) Various factors are arranged, according to 

 experience in physiological experimentation and to acquaintance 

 with the species, so that measurements are as comparable as 

 feasible. 



o 



2 - 



0^ 



0.2 - 



f, Qj 



p 

 q: 







-4 



-3 



-2 



Water Load 



Fig. 12. Mean rates of water output (% of Bo/At) at diverse negative water loads 

 (% of Bo) in 2 dogs. Each measurement represents a 24-hour period, 5 being average for 

 each point of total loss, and 10 for each point of urinary loss. Additional data of 

 Adolph ( '39a). The results are approximately confirmed by fewer data of Straub (1899) 

 and Spiegler ( '01). 



In a few words, the water exchanges are such during recovery 

 from privation that gain greatly exceeds loss, and such that gain 

 is augmented and loss is diminished as compared with the control 

 exchanges. Both factors are in directions (of signs) that tend to 

 restore water content to its control value. Of gains, drinking alone 

 is proportional to the negative water load. Detailed features of 

 the drinking process emphasize its physiological nature. Of losses, 

 those by the urinary path are reduced. Each load is linked to an 



