1913.] Contributions to the Biochemistry of (rrowth. 303 



accurately by these methods. Quantities of glycogen below this limit are 

 therefore indicated in the following tables by the expression " < 02 per cent." 



Since the glycogen store of the liver is known to be dependent upon and 

 influenced by external conditions, an attempt was made to equalise these as 

 far as possible. All the animals were of about the same age and weight ; 

 they were kept on a constant diet of bread and milk in the special metabolism 

 cages devised by Professor Schiifer.* In some series the supply of food 

 was limited to a definite quantity, which was sufficient to cover the require- 

 ments of the animal with regard to the growth of both the host and the 

 tumour; in other series the supply was ad libitum. 



The residts of these first observations, which have been grouped together 

 in Table I, show that the glycogen- content of the liver of both normal and 

 tumour-bearing animals varies within very wide limits. 



The following data were also known, but since they were found to have 

 no relation to the variations observed in this series, they have been omitted 

 from the table : — sex of host, weight of host. The fact that the observations 

 extended over several years also excludes the possibility of seasonal changes 

 on the part of the host and of cyclical changes on the part of the tumour-cells 

 as being responsible for these wide fluctuations. These must have been 

 caused, therefore, by factors which had not been controlled by the conditions 

 under which the observations had been carried out. 



A closer analysis of the results appeared to give a clue as to the nature of 

 these factors. It will be seen that the variations in the glycogen-content of 

 the liver are, as a rule, not so marked in animals killed on the same day. 

 Now when several estimations were carried out on the same day, the animals 

 were killed at practically the same time. Since all the animals were always 

 fed at the same time, it seemed possible that the influence of the time which 

 elapsed between the last meal of the animal and the moment when the liver 

 was subjected to analysis was operative within narrower limits — in such 

 small animals as rats at any rate — than the data given in the literature for 

 larger animals would lead one to suppose. 



Some estimations carried out with normal rats weighing about 100 grm. 

 showed that the glycogen-content of the liver was always relatively high 

 three to five hours after a meal, and that 15 hours after a meal it fell so low 

 (below 0'2 per cent.) that with the small amount of liver available it could 

 not be determined. 



In order to be able to compare the conditions in tumour-bearing animals 

 with those in normal animals it was therefore necessary to compare animals 

 in the same state of digestion and assimilation. This object was attained in 

 * For a detailed description, see 'Quart. Journ. Exper. Physiol.,' 1912, vol. 5, p. 204. 



z 2 



