68 PROCEEDrN"GS OE THE GEOLOGICAL SOCIETY. 



addition of an enstatitic mineral *. For this Dr. Wadsworth. proposes 

 the name saxonite t. Another marked variety adds angite or horn- 

 blende; to this we may aj^ply the term IJierzolite t, though hitherto 

 it has heen limited to one containing an augitic mineral, because 

 I think that here the substitution of hornblende has no important 

 significance. This group, then, consists of rocks containing little or 

 no alkaline constituent, little hme or alumina, the last named often 

 occuiTing in very small quantities, a considerable proportion of iron- 

 oxide, and magnesia and silica in nearly equal proportions, the former 

 generally shghtly exceeding the latter, and each not far from 40 

 per cent. Hemicrystalline and vitreous representatives of these 

 rocks are extremely rare. I have never met with one of either in 

 my own experience, though I should expect them to occur. Thus 

 no names have heen proposed for them. They are largely repre- 

 sented in past time by the true serpentines §, their metamorphic 

 representatives, into which they pass by insensible gradations ; 

 among these, I have suspected, though I cannot prove it, the 

 presence of glassy representatives, which, however, are rare and 

 local. 



The Picrite group may be regarded as a transitional one, formed 

 by the introduction of a small and variable amount of felspar, such 

 as auorthite or labradorite ; the amount of olivine is diminished : 

 enstatite and a pyroxenic mineral, with biotite, become important 

 constituents. The chemical composition, as might be expected, is 

 variable, but the percentage of silica is slightly greater than in the 

 last group, though generally not more than 45. Magnesia, though 

 still one of the two dominant constituents, is present in an amount 

 distinctly less than the silica, commonly from about 17 to 27 per 

 cent. Alumina is always present, sometimes in considerable 

 quantities, with alkaline constituents ; but chrome and nickel, which 

 are generally detected in the normal peridotites, are now often 

 absent. These rocks, so far as we at present know, are generally 

 ancient, and often more or less altered |{ ; but I think the name 



* I will use this term for breyity to include either enstatite, bronzite, or 

 hypersthene. minerals very closely related, if not Tarieties of one species. 

 " t Litliological Studies, pp. 84, 193. 



t Dr. WadsTvorth proposes {loc. cit. p. 193) to limit the term Iherzolife to 

 the variety with diallage, and call that with augite bvxhnerite, applying to a 

 variety wilh diallage and without enstatite the name ev.lysife. I doubt, how- 

 ever, the possibility of separating the first and second. Eidysite also has 

 hitherto been apphed to a garnet -bearing peridotic rock. 



§ I am, of course, well aware that this statement has been disputed, but 

 have more than once given my reasons for it, so need not repeat them. On 

 the principles of reasoning which I have endeavoured to establish above, I am 

 unable to understand how the derivation of a true serpentine, i. e. that of which 

 the Lizard serpentine is a type, from a peridotite, or the igneous nature of the 

 latter rock can be doubted. 



II The serpentinous rock of Duporth (Cornwall), for which the name Dupor- 

 thite has been needlessly proposed, appears to be a member of the picrite 

 group. Many of the picrites certainly hang on very closely to the dolerite 

 o-roup (described hereafter), and can be seen to graduate into representatives 

 of it. 



