264 BACTEllIAL POISONS. 



lecithin derived thus from these weeds. It is soluble in 

 alcohol, and but slightly in ether. It reddens turmeric, 

 decomposes carbonates and with acids forms salts which 

 precipitate lead acetate and silver nitrate, but not copper 

 acetate, even on warming. 



The hydrochloride is crystalline, and is partially disso- 

 ciated by excess of water. The platinum salt is soluble, 

 and crystallizes in very small cross-shaped prismatic 

 needles. The gold salt is amorphous and is readily altered 

 on heating. 



The properties of this compound show that it is of a 

 pyridine nature, and inasmuch as it does not give a pre- 

 cipitate with copper acetate, it would appear that the carb- 

 oxyl is not directly united to the pyridine nucleus. This 

 does not necessarily follow now that we know that some 

 amido-acids exists which do not give a reaction with copper 

 acetate (see page 231). Its pyridine nature is further- 

 more shown on distillation with lime. An oily alkaline 

 base is thus obtained which forms an iodomethylate, and 

 this with potassium hydrate yields quite an intense red 

 color, resembling lees (De Coninck's reaction). On 

 oxidation with permanganate of potassium it yields a mono- 

 basic acid. According to Gautier and Motjegues the 

 compound is probably identical with De Jungh's gadnine, 

 and they ascribed to it the following constitution, which, 

 it should be said, lacks full confirmation : 



H 

 C 

 ^ \ 

 IIC COH 



I II 

 H^C C— C3H,.CO,H. 



\ / 

 N 

 H 



Compare with tyrosin, CgHjjNOg (page 197). 



