BASALT-GLASS OF THE WESTEKN ISLES OE SCOTLAND. 453 



digestion of a specimen in concentrated hydrochloric acid for thirty- 

 hours, found a residue of 39*62 per cent, and, with Rosenbusch, 

 classes these Hawaiian lavas as hyalomelanes. Krukenberg, how- 

 ever, regards it as doubtful whether they should be placed with 

 tachylyte or with hyalomelane. 



Boricky * found the tephrite-glass of Klein Priesen to be decom- 

 posed partially and with difficulty by hydrochloric acid, some floccu- 

 lent silica being separated. Many of the German basalt-glasses, on 

 the other hand, appear to be entirely decomposed by boiling hydro- 

 chloric acid, that of the Sasebiihl, according to Mohlf, after two 

 days' digestion, and some even more readily than this. 



Classified according to the test of solubility in acids, the rock 

 of Screpidale might perhaps be placed with the tachylytes, and 

 the Beal, Some, and Gribun rocks with the hyalomelanes ; but 

 our own results, like those of Eosenbusch and Penck, tend to the 

 conclusion that it is advisable to abandon altogether a distinction 

 founded on such an uncertain character. 



6. Chemical Composition. 



If, as there seems reason to believe, basalt-glass is merely a 

 rapidly cooled portion of a basalt-lava, we may expect the two 

 rocks to have the same chemical composition. Delesse's analysis of 

 the dyke of Lamlash and of its vitreous selvage appears to indicate 

 that this is really the case. The analyses are as follows X '• — 



Basalt of Basalt-glass 



centre of side of 



of dyke. dyke. 



Silica 55-20 56-05 



Alumina 16-98 17-13 



Oxide of iron 11-00 10-30 



„ manganese traces. traces. 



Lime 6-80 6-66 



Magnesia : 0-52 1-52 



Soda \ 5-65 (by \ 3-29 



Potash J difference). J 0-98 



Water and volatile matter .... 3-85 3-50 



100-00 99-43 



The differences in these two results, judging by their character, 

 are such as may be fairly supposed to fall within the limits of error 

 in analysis. 



It has frequently been pointed out, however, that the average 

 composition of the basalt-glasses differs from that of the basalts, 



* Petrograpkische Studien an den Basaltgesteinen Bohmens, p. 182. 



t Die Gesteine der Sababurg, (Cassel, 1871) p. 41. 



$ Annales des Mines, vol. xiii. p. 369 (1851). In the analysis of the basalt- 

 glass as given by Delesse, the total (99"53) does not agree with the figures of 

 the column, which amount, as above, to 9943. A misprint may have occurred 

 in the column itself. 



Q.J.G.S. No. 155. 2 k 



