A Study of " Protagon. 



99 



suits which have lately made protagon appear to be a mechanical 

 mixture. It is very noticeable, however, that Wilson and Cramer, 

 in spite of their insistence upon this conclusion, have failed to give 

 their opinion a thorough test. They have not ascertained whether 

 fractional recrystallization of protagon by their own non-decom- 

 position method does not yield results similar to those they ascribe 

 to hydrolytic decomposition in the experiments in this laboratory. 

 They declared that treatment with boiling absolute alcohol, for 

 from one to two minutes at a time by their new method, does not 

 bring about such hydrolytic decomposition. Then why did they 

 not show that fractional recrystallization, by their non-decompo- 

 sition method, yields fractions of identical composition ? 



We have endeavored to settle the doubt in this matter on the 

 ground selected by Wilson and Cramer. We have prepared pro- 

 tagon by carefully following every detail of their very simple method, 

 as given by them. We cannot believe that defects in our technique 

 have accounted for any of our results. Nevertheless, we obtained 

 protagon which steadily decreased in phosphorus content each time 

 it was recrystallized by the Wilson and Cramer method (i .3 per 

 cent, to 0.4 per cent, in four recrystallizations) ; which never com- 

 pletely redissolved in the moderate excess of boiling absolute 

 alcohol we used for the purpose ; which, on recrystallization, each 

 time left in the mother liquor material having a much higher phos- 

 phorus content than the corresponding crystallized portion ; and 

 which, on fractional division by recrystallization at 20°C. and o°C, 

 yielded two crystalline products, and material in the final mother 

 liquor, that were strikingly unlike, physically and chemically. In 

 short, simple fractional recrystallization of protagon made by the Wil- 

 son-Cramer method, in which treatment, according to Wilson and 

 Cramer, decomposition does not occur, has given results which con- 

 firm conclusively the data and deductions published by Gies and 

 his collaborators, that Wilson and Cramer have sought unsuccess- 

 fully to explain away. 



Rosenheim and Tebb 1 also recently examined Wilson and 

 Cramer's results, but have not been able to confirm Wilson and 

 Cramer's conclusions. 



1 Rosenheim and Tebb : Proceedings of the Physiological Society, Journal of 

 Physiology, 1908, xxxvii. 



