SOURCES OF KNOWLEDGE 209 



digested from the food and the theoretical fat equivalent 

 of the decomposed protein as measured by the urine 

 nitrogen were charged up against the milk-fat, and a large 

 quantity of the latter could be accounted for only as 

 having had its source in carbohydrates. 



Another method of investigating fat formation has 

 been used with dogs. It is well known that when an 

 animal is deprived of food the expenditure of energy 

 by the body is maintained at the expense of body sub- 

 stance. Both muscular tissues and fatty substance are 

 broken down and used in this way, the latter being 

 regarded as furnishing the most natural and available 

 supply of fuel. It was found in the case of dogs that 

 after a certain number of days of starvation there oc- 

 curred a sudden and large increase in the waste of nitro- 

 gen compounds as shown by the urine excretion, the 

 explanation for this being that the body fat had become 

 exhausted and a demand was at once made upon the 

 protein tissues for the necessary supply of energy. As 

 soon as this rise of nitrogen waste appeared, then the 

 dog was allowed to eat, and whatever fat was found in 

 the body at the end of the feeding-period was regarded as 

 having been formed from the food taken after the star- 

 vation period. If, for instance, the ration was wholly 

 protein and fat was found to have become deposited in 

 the body, this was regarded as proof of the formation of 

 fat from protein. Such experiments as these have not 

 always been conclusive, although they are regarded by 

 some scientists as having furnished proof that protein 

 may be a source of fat. 



295. The respiration apparatus. After all, the investi- 

 gations of the kinds described fail to furnish data so 

 accurate and so complete as are necessary for entirely 



N 



