256 



Mr. J. Jolj. 



[Nov. 18, 



water. I came on no exceptions. There is, indeed, a more con- 

 spicuous kind of variation shown in the specific heats of minerals not 

 always determinable by external appearances, but this cannot, I think, 

 be classed with the foregoing. It is, however, difficult at present to 

 draw the line between them. But so far it seems highly probable that 

 in slight differences of thermal capacity may be recognised a mole- 

 cular restraint or discord or whatever molecular condition results in 

 opacity, roughness, incompleteness, formlessness. 



That other kind of variation to which I allude is of a different 

 order of magnitude, and, I venture to suggest, points to causes of 

 different order. It suggests the possibility that more than one 

 molecular arrangement may obtain in bodies chemically and crystal I o- 

 graphically, to all appearance, identical. For example, in hexagonal 

 prisms of beryl two very different degrees of molecular freedom would 

 appear to be possible. Certainly two very different thermal capacities 

 obtain, unexplained by chemical differences or crystallographic form, 

 so far as I can ascertain. The variation, too, seems in no way con- 

 nected with the density of the substance, for this may be invariable 

 from one specimen to another, while the difference of specific heat is 

 remarkable. In many cases this differance of structure, if such it be, 

 is. revealed in the behaviour of the substance to light, in the opacity 

 or transparency of the crystal. So marked is this that specific heats 

 may be at once assigned with certainty to different crystals of known 

 minerals on their opacity or transparency. In some cases again the 

 substances present no differences whatever in appearance. The case 

 of beryl stands thus : — 



Sp. h. Sp. gr. 



r Transparent, blue, six small crystals 0-20587 2*676 



*\ „ blue crystal 0-20705 2-666 



I „ greenish crystal 0-20636 2-660 



r Sub-translucent, green crystal 0'21257 2'706 



„ „ „ 0-21236 2-706 



L Opaque, dull-greenish crystal 0'21306 2 - 644 



Between the x and (3 specimens above there is no difference 

 apparent in crystallographic development save in the case of a couple 

 of the small x crystals of the first experiment, which show pyramidal 

 as well as the basal and prismatic faces common to all the specimens. 

 Examination of the twenty analyses collected by Dana (' Mineralogy ') 

 suggests no explanation on Chemical grounds. The percentages 

 throughout vary in no adequate degree, ascribing to the molecules the 

 thermal values recorded previously in arriving at a result by calcula- 

 tion agreeing with the (3 specific heat. I do not think, too, that 

 results independent, as they constantly seem, of density can with 

 probability be ascribed to differences in the amounts of the constituent 



