o66 Messrs. J. A, Gardner and F. W. Fox. Cyrigin and 



015 to 0-16 per cent. If we take 550 c.c. as the amount of bile secreted per 

 day and the cholesterol-content as 0'16, each of the subjects of our experi- 

 ment should have passed into the intestine about 0"9 grm. per day, more or 

 less, in bile solution. It is evident, therefore, that the subjects have reabsorbed 

 a considerable amount of the cholesterol along with the bile salts in the 

 intestine. 



It has been shown in earlier papers of this series that in the case of 

 herbivora (1912, 2), and also carnivora such as dog and cat (1913), that 

 cholesterol given with the food appears in the blood, the cholesterol having 

 been absorbed in the intestine with the bile salts. This is more difficult to 

 demonstrate in the human subject, but Widal, Weill and Laudat (1912) 

 have shown that heavy fat meals produce hypercholesterinemia, though this 

 state is transitory. The adult human subject is marked off from other 

 animals by the fact that the cholesterol passed into the intestine undergoes 

 reduction at some stage to coprosterol and y8-cholestanol, probably by 

 bacterial action, though a small quantity of cholesterol escapes this process. 

 It would seem a probable assumption that this reduction limits the 

 reabsorption of the cholesterol. This has not been definitely proved, 

 though feeding experiments are in progress with herbivorous animals to 

 gain evidence on this point. 



We think, however, that the considerations detailed above fully explain 

 the very variable negative balances recorded in Table I. 



It also follows that since cholesterol is an integral constituent of all cells 

 of the body, and there is an excess of output over intake, there must he some 

 organ in th-e hod.y capable of syntliesising cholesterol. This question we are at 

 present investigating. 



It will be seen from Table II that the intake of unsaponifiable matter not 

 precipitated by digitonin is very much larger than the output, and that, as an 

 average of the twenty-six experiments, there is a percentage utilisation of 

 82-8-3 per cent. The ratio of the average amount of fpecal sterols precipi- 

 table by digitonin to unsaponifiable matter not so precipitated was 1 : 0"2o to 

 1 : 1'18, but in the majority of cases the individual ratios are not far from 

 the mean value. 



As already mentioned, the faecal unsaponifiable matter not precipitated was 

 volatile in superheated steam without appreciable decomposition ; thus 9 grm. 

 only left in the distilling flask 0'14 grm. of carbonaceous matter. 



On fractionation of the oil from the whole of the experiments under a 

 pressure of 1 mm., the ratio of the lower boiling portions which did not give 

 the Burchardt-Liebermann reaction to the high boiling " sterol " portion which 

 did was about 1 : 4. 



