known, alkaloids in general give pyrrol on dry distillation; morphine, 

 for instance, on being heated with 10 parts of zinc dust gives off 

 pyrrol, ammonia, trimethylamine, pyridine, phenanthrene, etc.* Dur- 

 ing the past winter we made several attempts to prove the presence of 

 pyridine among the products of dry distillation of the active principle 

 as above isolated, as its detection would prove that our principle was to 

 be classed among the alkaloids. 



In attempting to show the presence of pyridine we heated the active 

 principle with zinc dust in a current of dry hydrogen and passed the 

 distillation products through a test tube and two small flasks, the tube 

 being placed in a cooling mixture, the first flask being partly filled 

 with dilute sodium hydrate, and the second with dilute hydrochloric 

 acid. 



At the close of the experiment the test tube gave off alkaline vapors 

 that smelled strongly of trimethylamine and had in addition a peculiar 

 sweetish odor. The contents of the test tube diluted with water gave 

 a precipitate with bromine water and with copper sulphate, but with 

 the latter reagent the fine blue color characteristic of pyridine did 

 not appear. The presence of amines seems to render difficult the de- 

 tection of small quantities of pyridine. The contents of the third flask 

 on exposure to the air took on a rose-red color and after a time threw 

 out a precipitate of the same tint. Both flasks gave off the odor of 

 benzaldehyde. 



The following experiment was also made with a little of the sul- 

 phate, which no longer gave the rose-red reaction with iodine but 

 which still contained a little of the coniine-like substance. For two 

 hours it was kept at 150° C. in a sealed tube with 25 per cent hydro- 

 chloric acid. On shaking out with ether a considerable crystalline 

 residue was obtained which, without much purification, melted 

 sharply at 120° C. and had all the properties of benzoic acid. 



In the above decomposition experiments the quantity of material 

 at command was small, perhaps not more than one-fifth of a gramme 

 at a time, and not entirely pure. We do not, therefore, lay much stress 

 on our failure to detect pyridine. 



The fact that the nitrogen of our compound is given off in the form 

 of amines and pyrrol gives strong gTound to believe that our substance 

 is to be classed with the pyridine bases, using this term in a broad 

 way. Its basic character, its ability to take up acid radicles, as illus- 



* We are well aware that certain salts of glutaminic, pyromuci'c and its re- 

 lated acids also yield pyrrol on dry distillation. These compounds, however, like 

 the proteids and their allies, appear to us to be excluded. 



52 



