50 



As it seemed likely that the study of any one of these 

 reactions would explain all the rest, the author began with 

 the analysis of g and h, of which he had a larger supply 

 than of the others. 



g is lemon yellow, transparent, soluble in water, insoluble 

 in alcohol, crystallizing in square tables much resembling 

 those of ferrocyanide of potassium. By exposure to a heat 

 of 212°, it loses 13.5 per cent, water of crystallization, and 

 becomes opaque. More strongly heated it is decomposed, 

 giving off hydrocyanate of methylene, = c 2 h 3 cy or me cy. 

 The analysis corresponds with the formula 4 k cy, 3 ve cy, 

 M cy, 8 Aq. 



h is white, very soluble in water and alcohol, crystallizing 

 in square shining tables. It closely resembles sulphome- 

 thylate of potash, but differs from it in being anhydrous, in 

 containing cyanogen, and in yielding hydrocyanate of me- 

 thylene when decomposed by heat. Its analysis agrees with 

 the formula 6 s o 3 , 3 k o, m o, m cy. 



If 3 equivalents of ferrocyanide of potassium be supposed 

 to act on 3 of sulphomethylate of potash there is the follow- 

 ing equation : 



3 equiv. Ferrocyanide 3 equiv. Sulphomethylate 



6 k cy, 3 Fe cy + 6 s o, 3 k o, 3 m o = 



1 equiv. g 1 equiv. h 



= 4 k cy, 3 Te cy, u cy + 6 s o 3 , 3 K o, m o, M cy 



+ 2ko. i. e. 2 equiv. potash. In conformity with this ex- 

 planation, the liquid in which g crystallizes is alkaline. 



If this explanation be admitted, it will of course apply, 

 mutatis mutandis, to the salts a b, c d, e f. The author, 

 however, is not yet satisfied that the salts which he analyzed 

 may not have been mixtures, perhaps in definite proportions. 

 No doubt can be entertained that new salts have been 

 formed, but the close resemblance between their properties 

 and those of the salts which yield them, renders the task of 

 purifying and analyzing them one of great difficulty. 

 (To be continued.) 



