ioo SCIENCE PROGRESS 



but the other nitrogen atom does not appear to be in a true 



nitro group, since on reduction ammonia is lost and a saturated 



keto-alcohol, C27H44O2, is formed. This behaviour recalls the 



observations of Wallach and others on a and jB phellandrene, 



anethol, etc., which they showed could be converted, through 



nitro-derivatives, into saturated ketones or aldehydes. In the 



phellandrene compound, for example, Wallach assumes the 



grouping 



CH 



l>0 



C-NO 



and however this may be, the effect of reduction is certainly 

 to transform the group CH : CH into the group CO . CH 2 , and 

 we may perhaps assume the same change with the cholesterine 

 derivative. 



The keto-alcohol cholestanonol, therefore, probably contains 

 CO . CH 2 instead of the CH : CH of the parent substance. It is 

 a white crystalline body melting at 142-143 . The keto-group is 

 easily characterised by the formation of a crystalline nitrophenyl- 

 hydrazone and the hydroxyl by the corresponding acetyl and 

 benzoyl derivatives. This hydroxyl is easily attacked by most 

 oxidising agents, and a diketone, cholestandione, C27H42O2, is pro- 

 duced. This very important body, which, as mentioned above, 

 is also prepared by the reduction of oxycholestenone, melts at 

 169 and gives a well-defined dioxime. Bromine acts as a sub- 

 stituent, giving a dibromcholestandione, C2 7 H 40 O 2 Br 2 , which in 

 melting-point and other properties proves identical with the 

 dibrom-addition product of oxycholestenone. The close con- 

 nection between the two compounds is thus emphasised, and 

 a valuable connecting link given us from which conclusions may 

 be drawn as to formulae. Since the two bromine atoms are 

 added at the double link of oxycholestenone they should be in 

 the 1 : 2 position to each other. Again, if the bromination of 

 cholestandione follows the usual course with such ketones, each 

 should be in the a /3 position to a carbonyl group, giving us the 

 grouping CO . CH 2 . CH 2 . CO, and making cholestandione a 

 7-diketone. This view is to some extent confirmed by its 

 behaviour towards hydrazine and its oxidation. Treated ener- 

 getically with chromic acid it gives cholestanone dicarboxylic 

 acid, which still contains one keto-group. If cholestandione 

 were a /3-diketone the acid would be a /3-ketonic acid ; but 

 it is stable on heating to 200° with potash, which is all against 



