1818.] Crystalline Form and Chemical Composition. 265 



themselves, we cannot determine, with precision, the amount of 

 the extraneous matter. 



Chemists, on the contrary, do not admit these accidental 

 compounds on so large a scale. 



It appeared to the author that the problem might receive some 

 illustration from a series of experiments on such substances as 

 might be compounded and decomposed at pleasure ; as thess 

 would enable him to determine the proportion of accidental 

 mixture capable of entering into the composition of a body, 

 without affecting its crystalline form. 



Mixtures of Sulphate of Copper and Sulphate of Iron. 



It had been long known that a mixture of equal parts of sul- 

 phate of iron and sulphate of copper would afford rhomboidal 

 crystals similar to those given by sulphate of iron, the diagonals 

 of their faces being to each other as a/ 7 to y 10 ; but this 

 remarkable fact did not determine all the proportions of these 

 two salts capable of producing this rhomboidal crystal. The 

 author's first researches were, therefore, directed to this point. 



He mixed sulphates of iron and of copper in various propor- 

 tions, and produced crystals from the solutions. Equal parts gave, 

 as before, the rhomboidal form ; but by increasing the proportion 

 of sulphate of copper, he at length obtained crystals similar in 

 form to those of the sulphate of copper ; and by varying the 

 experiments he ascertained the proportions that would give either 

 form. 



He at first concluded that the crystals which were produced 

 contained the same proportions of the two salts as the mixture ; 

 but by analyzing the crystals * he found the proportions vary 

 according to the degree of concentration of the solution. The 

 sulphate of iron, being the most soluble salt of the two, remained 

 in the greatest proportion in the solution. 



Frequently after obtaining crystals of the form of sulphate of 

 iron from some given proportions of the two salts, part of the 

 crystals were re-dissolved, and again crystallized ; and by repeat- 

 ing this process, it was found that the proportion of sulphate of 

 iron in the crystals became less and less, and at last was so 

 small that the crystals assumed the form of those of sulphate of 

 copper. Thus crystals which contain 20 parts of sulphate of 

 iron and 80 of sulphate of copper, retained the form of sulphate 

 of iron : then being successively redissolved, and again crystal- 

 lized, were found to retain the same form, although the sulphate 

 of iron was reduced first to 15, then to about 12, and lastly to 

 only 9 per cent., which appears to be the limit of proportions 

 affording the rhomboidal crystal ; for on again redissolving and 



• These crystals were analyzed by dissolving in pure water, and precipitating 

 the iron by pure ammonia ; and the proportion of *ulpb;Ue of Iron was computed 

 according to the analyiii of Kir wan, which give* 28 parts of sulphur. c acid to I if© 

 parti of oiide. 



5 



