Io6 PEOCEEDINGS OF THE GEOLOGICAL SOCIETY. 



rocks crystals of considerable size of such minerals as pyrite, 

 magnetite &c., are readily formed. So are quartz and calcite, also 

 hornblende, chlorite, epidote, various zeolites, garnet (generally 

 microscopic), tourmaline in special circumstances, &c. In like way, 

 pure limestone readily assumes a crystalline structure. Hence we find 

 fairly crystalline limestones, and especially dolomites of Secondary 

 or even later age. It is, however, Noteworthy that any one of these 

 presents marked differences from a rock of Archaean age which has 

 a like chemical composition. The most crystalline dolomites from 

 the English Permians or from the Alpine Secondary strata, the lime- 

 stones from the latter and from Socotra (probably Tertiary), differ 

 greatly from any which are certainly of Archaean age, and, if I mis- 

 take not, there are even differences between the older and younger of 

 the last. Such minerals as mica, some varieties of pyroxene, quartz 

 — not obviously of fragmental origin — garnet, not to name others, 

 are frequent in crystalline limestones of Archaean age, but are rare 

 in those certainly of later date, in which the non -calcareous con- 

 stituents still appear unchanged. In reference to this point I may 

 remark that it will surprise me if the comparatively late date 

 assigned by some geologists to the marbles of Carrara proves to be 

 correct. 



The facility with which crystalline quartz is deposited, in the 

 manner often described, causes a sandstone to be readily converted 

 into a quartzite ; but I may remark that for this change it seems to 

 be essential that the rock should be very clean, that is, the quartz- 

 grains should be as nearly as possible free from all earthy or ferru- 

 ginous admixture. Clay, indeed, appears to be generally a very 

 refractory material, except sometimes under great pressure, when 

 silica is separated, and minute micaceous minerals are formed from 

 the residue. As I have said, clay remains seemingly unaltered 

 in some crystalline limestones and quartzites, and, if abundant, 

 certainly impedes crystallization. Prom what I have seen, I am led 

 to think that mixtures of different mineral substances and different- 

 sized particles do not so readily change as those which are homo- 

 geneous. Thus I have noticed that, in the same series, gritty slates 

 seem a little less altered than those of finer and more uniform 

 materials. 



We find a good example of this variation in the Palaeozoic group 

 near Loch Maree. The purest quartzites are the most highly altered. 

 The quartzites, as a whole, are more metamorphosed than the dirty 

 limestone above or the heterogeneous Torridon Sandstone below. In 

 the case of the last named we might have expected that quartz 

 would have been attracted to quartz, felspar to felspar, mica to mica, 

 hornblende to hornblende. These accretions are familiar in the case 

 of quartz, they have been claimed for felspar and hornblende, and 

 they are in accordance with the principle of elective affinity. Thus 

 we should expect that the conversion of a suitable arkose into a 

 gneissic rock would be comparatively easy. It is quite possible that 

 this may be the origin of some gneisses ; but in the case of the 

 Torridon Sandstone, we do not see much approach towards it. 



