362 Messrs. J. A. Gardner and F. W. Fox. Origin and 



unsaponifiable matter taken. The sterol-digitonide, precipitated from faeces, 

 consisted mainly of coprosterol digitonide with smaller portions of 

 /3-cholestanol digitonide, cholesterol digitonide, and perhaps also phyto- 

 sterol digitonide. 



The unsaponifiable matter not precipitated by digitonin was got by 

 difference, but this was always checked by direct weighing of the oil 

 washed away from the digitonide by ether. Fat in the foods was deter- 

 mined by extraction with ether in the usual way, and the unsaponifiable 

 matter and sterols estimated as described in the case of faeces. In the case 

 of bread and meat, however, the ether extraction was preceded by repeated 

 extraction with boiling alcohol, as it is well known that in such substances 

 the extraction of fat by ether alone is imperfect, even though the exhaustion 

 may be prolonged. The alcohol extracts were then evaporated, taken up in 

 ether, and the ether solution was added to the other ether extract. 



Full details of the nitrogen balance in the subjects of diets A, B, C and D 

 are given in the reports and papers mentioned. 



Results. 



The results of our experiments on the intake and output of unsaponifiable 

 matter are summarised in tlie following Tables. Table I contains the daily 

 intake and output of sterols precipitated by digitonin. The figures represent 

 the daily average over a period of six days. The intake consists of the 

 total cholesterol, in free and ester form, of the food consumed, together with 

 traces of phytosterol. The sterol excreted consisted, as stated above, mainly 

 of coprosterol, with smaller quantities of /3-cholestanol and cholesterol, and 

 perhaps phytosterol. The subjects are indicated by their initials. Eight of 

 the subjects partook of diets A, B and C, except Mr. C, who was omitted 

 from the experiment on diet B owing to an attack of diarrhoea, which came 

 on during the experimental period. 



Only two of these subjects, E and B, were available for the experiments 

 with diet D. Another subject, P, joined in this experiment. 



In Table II, the daily intake and output of the portion of the unsaponifiable 

 matter of the fat which is not precipitated by digitonin is given. These 

 figures are only approximate, and are no doubt slightly too high owing to the 

 presence of traces of resinous matter produced by the action of the alkali on 

 the alcohol during hydrolysis of the fat (Gardner and Fox, 1921). 



Discussion of Besults. 



It will be noticed from the figures in Table I that in every case there is 

 an excess of output over intake, except subject B on diet who shows a 



