530 M. E. Breunlin on the Constitution of 



I made analyses of five kinds of blue ultramarine (I. to V.), 

 and of two kinds of green (VI. VII.), with the following 

 results : — 



III. IV. V. VI. VII. 



Silica . . . 37-4 40-9 35-5 363 366 38-4 38-8 



Alumina . . 30-0 242 28-4 259 25-0 27-4 28-3 



Oxide of ii-on . 1-3 0-5 06 3-1 0-9 0-6 0-9 



Soda . . . 14-9 16-3 19-2 21-0 17-2 16-9 13-9 



Sodium. . . 2-8 32 19 2-1 3-2 53 5-5 



Q , , fa. 2-0 2-2 1-3 1-4 22 37 38 



bUlphUr ^^^ ^.j g.g ^.g g.g g.y 3.5 5.7 



Lime ... Oo 08 0-6 11 I'O 08 0-9 

 Sulphuric acid 2-3 1-3 31 27 2-0 Oo 06 

 Clay ... 2-8 1-5 20 23 2-8 1-7 10 



101-1 99-4 100-5 101-7 99-6 98-7 99*4 



la these analyses, sulphur a is the quantity of sulphur liberated 

 as sulphuretted hydi-ogen when ultramarine is treated with acid, 

 and sulphur b the quantity of sulphur which falls as milk of 

 sulphur. 



The analyses were executed in the following manner : — 



A quantity of ultramarine, di-ied at 100° C, was drenched in 

 a beaker with hydrochloric acid, and completely decomposed. 

 It was filtered ; silica, sulphur, and uudecomposed clay remained 

 on the filter, and were washed out, and the silica separated from 

 the clay by solution in caustic potash, from which it was sepa- 

 rated by acid, and determined. The filtrate from the clay, sul- 

 phm- and silica was evaporated to dryness, the residue heated 

 above 100° C, then moistened with HCl, and digested with 

 water. The silica which had been dissolved in hydrochloric acid 

 remained behind, was filtered, and estimated along with the 

 other. 



In the filtrate, alumina and oxide of ii-ou were precipitated by 

 ammonia, and separated from each other by potash ; the lime 

 was precipitated by oxalate of ammonia, and the soda determined 

 as sulphate of soda. 



Another portion of ultramarine was fused with soda and nitre 

 in a platinum crucible ; the mass swelled up from the carbonic 

 acid hberated, and became bluish-green, green, and then colom-- 

 less. 



From the solution in acidulated water, sihca was separated in 

 the usual manner, and the sulphui-ic acid determined as sulphate 

 of baryta. This analysis gave the whole of the sulphur, along 

 with the sulphuric acid contained already as such. 



Another portion was treated with hydrochloric acid, the silica 

 separated, and in the filtrate sulphuric acid precipitated by 



