314 



different radicals. The production of the platinum bases, however, 

 shows that they do still contain replaceable hydrogen, so that either 

 the formation of a fixed base by the addition of one equivalent of a 

 radical does not prove that they are nitryl bases, or the received 

 opinion regarding the constitution of the platinum bases must 

 undergo some modification. 



It is clear that at present we cannot attempt any explanation of 

 these apparently anomalous results ; but I am now engaged examin- 

 ing the decompositions of the platinum salts of amide and nitryl 

 bases containing known radicals, which will probably lead to their 

 correct explanation. 



4. On the Volatile Bases produced by Destructive Distilla- 

 tion of Cinchonine. By C. Greville Williams, Assistant 

 to Professor Anderson, Glasgow University. 



In this paper the author shows that Cinchonine by distillation 

 with potash, undergoes a very complex decomposition, and that in- 

 stead of yielding one base, as has hitherto been supposed, gives at 

 least seven. 



The mode of research at first adopted was to convert the basic 

 liquid into platinum salt, and separate the bases by fractional crys- 

 tallization in the manner described in his paper " On the Presence 

 of Pyridine in the Naphtha produced by destructive distillation of 

 the Bituminous shale of Dorsetshire."* 



The experiments made in this manner, indicated that several sub- 

 stances were present, but it was evident that to decide the question, a 

 very large amount of material would be required; the author there- 

 fore subjected 100 ounces of cinchonine to distillation with potash, 

 and thus obtained sufficient of the basic oil to enable him to effect 

 twelve complete fractionations, involving at least 240 distillations. 



Runge's Pyrrol was present in the crude bases, and was removed 

 by protracted boiling of the acid solution. 



The bases were procured free from water by digestion with potash. 

 The following fractions were then analysed. 



Fraction boiling at 310^ F. The basic liquid on analysis gave 

 numbers exactly agreeing with the formula, 

 Ci* H9 N, 

 * Philosophical Magazine, Sept. 18o4. 



