6b PEOCEEDIIS'GS OF THE GEOLOGICAL SOCIETY. 



I. II. III. 



SiO, .... 64-70 69-65 SiO^ .... 53-55 



Alfi^ .... 3-50 1-82 Al.Og 6-01 



CaO .... 12-00 13-31 ^Eep^ .... 5-74 



Ka,0 .... 19-80 15-22 CaO .... 29^22 



■ K,0 .... 5-48 



100-00 100-00 • 



100-00 



In flint-glass the percentage of silica is often higher, that of alumina 

 lower, and those of lime and soda more nearly equal than in (I.). 

 In fact, we might take as an approximate composition of a typical 

 flint-glass about 70 per cent, of silica and about 27 per cent, of lime 

 and soda, in nearly equal proportions, the residue consisting of 

 alumina, iron, manganese, &c. The presence of alumina, as is well 

 known, renders the glass less fusible. Hence, in order to obtain 

 devitrified specimens, Mr. Claudet was in the habit of sprinkling a 

 little clay on the residual glass in a pot. (This, I think, acted 

 mechanically as well as chemically, as probably it was not wholly 

 melted.) Compound silicates also, he informs me, were more fusible 

 than simple. In the comparative study of artificial glasses and of 

 igneous rocks there are, however, two obvious difiiculties — one, that 

 the former, as indicated above, contain a much larger percentage 

 of lime and of an alkali than we find in obsidians, pitchstones, or 

 rhyolites, and that in regard to them we are dealing with dry fusion 

 and dry solidification (Mr. Siemens informs me that it is an important 

 matter to allow the " metal" to boil for a considerable time, i. e. to 

 eliminate from it vapours and gases). As I had been informed that 

 Messrs. Siemens were making bottle-glass from " granulite " or granite, 

 I had great hopes of obtaining from them an artificially produced 

 obsidian, and thus being enabled to study the devitrification of a 

 substance chemically identical with a natural rock. This, I regret 

 to say, is not quite the case. Their staple material is a granite or 

 granulite, of which analyses are given below*, but a certain quantity 

 of calcium fluoride and sodium chloride or sulphate are added. I 

 was not informed of the exact amount ; Mr. Siemens says it is 

 done rather roughl}-, but "so as to make a good alkaline glass." 

 Hence, even in this glass, we have more lime and alkali, as well as 

 less magnesia, than we should find in such a rock as an andesite, 



*I. II. 



Light-coloured Dark-coloured 



variety. yariety. 



SiO„ .. 69-30 60-90 



A1,0, 17 70 19-10 



FeA 1-10 3-10 



CaO 1-60 2-90 



MgO 0-13 0-80 



K.,0 3-45 4-10 



Na,0 5-20 5-53 



Diff. 1-25 3-27 



Difference contains HF, MnO, &e. 



