48 THE VOYAGE OF H.M.S. CHALLENGER. 



much as 71 per cent, of silica (the tridymite trachyte of New Zealand, for example), 

 but this large proportion is due in great part to the infiltration of siliceous matter 

 subsequent to the consolidation of the rock. To infiltration of this kind we have 

 recourse in order to explain the anomaly in the present case. We have already said 

 that the trachytes of Ascension contain little veins of quartz of secondary origin, and 

 the ground-mass is sometimes penetrated by silica. Darwin remarked the frequency 

 with which siliceous veins occur in the whole region, and infiltration of silica of 

 secondary origin accounts for the divergence in the analysis before us. The small 

 proportion of ferrous oxide, magnesia, and lime clearly shows that pyroxene is a 

 very subordinate constituent of the rock. We see besides, as analyses of trachytes 

 often show, that soda predominates over potash in a marked degree. Perhaps we have 

 here a monoclinic felspar which would approach those described by Forstner (2 - l mol. 

 Na 2 AP Si e O' 6 , with 1 mol. K 2 A1 G Si 6 O 16 ). Vom Rath showed that in the sanidines of 

 Laacher-See soda may lie present in larger amount than potash ; perhaps small plagio- 

 clases are hidden in the ground-mass, which may itself contain a glass more or less rich 

 in soda. 



We have considered the pyroxenic trachytes, and now turn to the specimens which 

 show a transition to obsidian. 



The ever-increasing predominance of base over crystalline elements is shown very 

 well in a specimen from Red Hill (?). The external appearance is still quite that of 

 ordinary trachyte ; to the naked eye it shows a rather pronounced schistoid appearance. 

 The colour is grey, darker than the ordinary trachytes of the island ; it is still slightly 

 rugose, and has not assumed a vitreous texture. Crystals of sanidine from 3 to 4 

 millimetres long determine a porphyritic structure in the rock. When a thin section is 

 examined, the large share which the vitreous mass has in its constitution becomes 

 apparent. The schistose appearance is also found in the preparation, — it is produced 

 by lines of vesicles, which, like those of pumice, are due to the liberation of gases 

 during cooling. Well-developed felspar and augite microliths are ranged in the same 

 direction as the vesicles. It is without doubt to its fluidal structure that the lamina- 

 tion of this rock must be attributed. Large sections of sanidine and augite, but mostly 

 the former, detach themselves from the vitreous ground-mass, which is light brown 

 in colour with slightly darker bands. The sanidine is sometimes found crystallised as a 

 Carlsbad twin. Plagioclase is occasionally detected. Besides the minerals already 

 mentioned, the ground-mass is filled with little bundles of crystals extremely minute and 

 only appearing under the highest powers. Relying on microscopic analyses only, these 

 obsidians could not be separated from the augitic trachytes. In fact, one sees that the 

 latter rock is related through all its transitions with the former, and that the constituent 

 minerals are the same in both ; only the vitreous element tends gradually to take the 

 place of the minerals, which grow smaller as the trachyte approaches the vitreous 



