232 Royal Society : — 



distinguished from acetate of rosaniline, which they in other respects 

 much resemble. The crystals are insoluble in anhydrous ether, diffi- 

 cultly soluble in water, but dissolve readily in alcohol, The solution 

 has a magnificent blue colour, with a coppery iridescence on its sur- 

 face. Addition of acids destroys this colour. Ammonia and the fixed 

 caustic alkalies leave the colour apparently untouched ; but it is now 

 produced by a finely divided deep-blue precipitate suspended in the 

 liquid, which may be filtered off, the filtrate separated from it being 

 colourless. 



The green crystals were found to be the iodide of a peculiar basic 

 compound. The iodine is rather firmly held in this compound ; but it 

 may be precipitated from the alcoholic solution by oxide of silver, and 

 exchanged for bromine or chlorine by treatment of this solution with 

 bromide or chloride of silver, when the bromide or chloride corre- 

 sponding to the iodide are produced. The analysis of the crystals 

 gave results indicating unequivocally the formula 



G30 H 39 N 2 1, 

 which received a close confirmation by the examination of a fine 

 platinum-salt crystallizing in rhombic tablets, which is obtained by 

 precipitating the chloride corresponding to the iodide, strongly acidu- 

 lated with hydrochloric acid, by dichloride of platinum. Neverthe- 

 less slight discrepancies between the theoretical values of the formula 

 and the results obtained led me to assume the existence in the crystals 

 of a compound containing less carbon and hydrogen, indeed of a 

 homologous iodide, 



C 28 H 3S N 2 I. 



This hypothesis, not countenanced at first by the remarkable con- 

 stancy which the composition of the iodide presented even after three 

 or four crystallizations, was fully confirmed when the chloride was 

 submitted to a systematic partial precipitation by dichloride of 

 platinum. After several repetitions of the process, the partially 

 precipitated platinum-salt being decomposed by sulphuretted hydro- 

 gen and the chlorides again partially precipitated, two platinum-salts 

 were obtained, one of which, the less soluble one, proved to be the 

 pure platinum-salt corresponding to the iodide with 30 equivalents 

 of carbon, whilst the other one was sufficiently pure to show that it 

 belonged in reality to the homologous iodide with 2 equivalents of 

 carbon less. 



The amount of the iodide 



C 28 H 35 N 2 1, 

 which contaminated (if the term may be applied to so beautiful a 

 substance) the iodide 



C 3 o H 39 N 2 1, 

 is, however so small that its presence did not materially influence 

 the analytical results obtained in the further examination of the 

 compound. 



The explanation of the formation of the iodide presents no diffi- 

 culty ; this substance obviously derives from lepidine, 



C 10 H 9 N, 



