XIII. ] ORIGIN OF CRYSTALLINE ROCKS. 289 







enclosing granitic vein-stone.* Similar shells of galenite and 

 of zircon, having the external forms of these species, are also 

 found filled with calcite. In many of these cases the process 

 seems to have been first the formation of a hollow mould or 

 skeleton-crystal (a phenomenon sometimes observed in salts 

 crystallizing from solutions), the cavity being subsequently 

 filled with other matters. (Ante, page 212.) Such a process 

 is conceivable in free crystals formed in veins, as, for example, 

 galenite, zircon, tourmaline, beryl, and some examples of gar- 

 net, but is not so intelligible in the case of those garnets im- 

 bedded in mica-schist, studied by Delesse, which enclosed 

 within their crystalline shells irregular masses of white quartz, 

 with some little admixture of garnet. Delesse conceives these 

 and similar cases to be produced by a process analogous to that 

 seen in the crystallizations of calcite in the Fontainebleau sand- 

 stone ; where the quartz grains, mechanically enclosed in well- 

 defined rhombohedral crystals, equal, according to him, sixty- 

 five per cent of the mass. Very similar to these are the crys- 

 tals with the form of orthoclase, which sometimes consist in 

 large part of a granular mixture of quartz, mica, and ortho- 

 clase, with a little cassiterite, and in other cases contain two 

 thirds their weight of the latter mineral, with an admixture of 

 orthoclase and quartz. Crystals with the form of scapolite, 

 but made up, in a great part, of mica, seem to be like cases of 

 envelopment, in which a small proportion of one substance in 

 the act of crystallization compels into its own crystalline form 

 a large portion of some foreign material, which may even so 

 mask the crystallizing element that this becomes overlooked, as 

 of secondary importance. The substance which, under the 

 name of houghite, has been described as an altered spinel, is 

 found by analysis to be an admixture of vb'llknerite with a 

 variable proportion of spinel, which, in some specimens, does 

 not exceed eight per cent, but to which, nevertheless, these 

 crystalloids (to use the term suggested by Naumann) appear to 

 owe their more or less complete octohedral form.t 



* Report Geol. Survey of Canada, 1866, p. 189. 

 t Ibid., pp. 189, 213 ; American Journal of Science (3), I. 188. 

 13 s 



