of Ferrous Oxide in Silicates. 115 



are possible ; first, that oxidation should take place; second, 

 that the mineral should not be perfectly decomposed. To set- 

 tle the first point two soluble silicates, and a sample of dunyte 

 from North Carolina, and the other a rolling mill cinder, were 

 treated by our process, and afterwards by decomposition with 

 sulphuric acid in a flask, in an atmosphere of carbon dioxide, 

 with results as follows : 



l. 2. 



Dunite _ 2-72 2'72 



Cinder 59-60 59-88 



Showing closely accordant results. To settle the question as to 

 complete decomposition duplicate determinations were made on 

 pyroxene from Vesuvius, and on the crocidolite from Cumber- 

 land, E. I. 



1. 2. 



Pyroxene 4*59 459 



Crocidolite 21-19 21-25 



That the decomposition is complete and the results obtained 

 are to be relied on, is proved not only by the agreement shown 

 above, but also by the absence of all gritty particles in the 

 beaker after solution, and by the fact that complete solution 

 takes place on heating, or even after a few hours standing in 

 the cold, after titration. 



The whole operation requires about one hour from the time 

 the mineral is put into the crucible to the final determination 

 of the iron. The time required for pulverization cannot be 

 lessened, however, for unless this part of the work is carefully 

 done, decomposition will not be complete. In several experi- 

 ments on the cinder named above the loss was found to be 

 from three to six per cent of the ferrous oxide known to be con- 

 tained in it, and it was at first feared that the method would 

 prove a failure on certain silicates ; but finer pulverization com- 

 pletely overcame the difficulty and there was afterwards no 

 trouble with any of the silicates on which we experimented. 



Hamilton College. March 29th, 1887. 



Note. — Since writing the above the attention of the authors 

 has been called to an article by C. E. Avery* entitled " Decom- 

 position of refractory Silicates by Fluorides." In this article the 

 use of a metallic fluoride is proposed for the purpose mentioned, 

 but no suggestion is made as to ferrous oxide, nor are any details 

 of the process given. In Crooke's " Select Methods of Chemical 

 Analysis "f the application of this method to the determination of 

 ferrous oxide is suggested, but here only meagre details are given 

 and in both cases objection is made to the use of ammonium 

 fluoride, which we have found most convenient and satisfactory. 



* Chemical News, vol. xix, p. 270, 1869. \ Edition of 1886, p. 133. 



