466 Dr. Anderson on the Products of the 



Analyses were made of the different portions of oil boiling 

 about the temperature of 310°, with the following results : — 



I. 3-840 grs. of the base, boiling between 310° and 815°, gave 

 11*007 carbonic acid and 3*060 water. 



II. 4*012 grs. of the base, boiling between 315° and 320°, gave 

 11*516 carbonic acid and 3*160 water. 



III. 4*319 grs. of the base, boiling between 316°and 320°, gave 

 12*430 carbonic acid and 3*576 water. 



IV. 4*430 grs. of the base, boiling between 320° and 324°, gave 

 12*812 cai*bonic acid and 3*405 water. 



100*00 100*00 10000 100*00 



These results agree very closely with the formula C^"^ H^ N, 

 as is shown by the following comparison of the mean experimental 

 and calculated numbers. 



Mean. Calculation. 

 ^ 



Carbon . . . 78*45 78*50 C^^ 84 



Hydi'ogen . . 8*81 8*41 H^ 9 



Nitrogen . . 12*54 13*09 N 14 



10000 100*00 107 



Notwithstanding the close correspondence of these results, 

 however, further experiment showed that some of the fractions, 

 especially those of lower boiling-points, contained appreciable 

 quantities of picoline, the presence of which was established by 

 the analysis of the platinum salts. When, for instance, a portion 

 of any of these fractions was saturated with dilute hydrochloric 

 acid and bichloride of platinum added, fine prismatic crystals 

 were slowly deposited, which, as the result of numerous experi- 

 ments, were found to contain about 32*8 per cent, of platinum, 

 which is exactly the quantity present in the picoline salt, of which 

 the theoretical per-centage is 32*92. On evaporation of the 

 mother -liquor, crystals were deposited which gave quantities 

 of platinum vaiying from 32*5 to 32*0 per cent., and which were 

 obviously mixtures of the picoline and lutidine salts. When the 

 last mother-liquor, however, was evaporated to a small bulk, and 

 alcohol and aether added, another salt altogether distinct from 

 that of picoline, and crystallizing in flattened tables, was de- 

 posited, which analysis proved to have the constitution of the 

 lutidine salt. 



This platinum salt crystallizes from its solutions in square 



