OLDER PERIDOTITES OF SCOTLAND. 4013 



serpentine, and magnetite. But while the hardness of the micaceous 

 mineral is very variable, depending on the extent of alteration it has 

 undergone, I found it, in fresh specimens, to be no less than 3 of Mohs' 

 scale. The specific gravity was determined, by taking the density 

 of a solution of the boro-tungstate of cadmium, in which particles of 

 the mineral remain just suspended, to be 2*8 ; examined with the 

 blowpipe it is found to be extremely infusible, only the very thinnest 

 splinters showing traces of whitening and fusion on the edges after 

 prolonged action with a powerful flame. In the blowpipe-beads it 

 gave a faint iron-reaction. It is scarcely acted upon at all by 

 either boiling hydrochloric or sulphuric acid. 



The general appearance and many of the properties of this 

 mineral so strongly resembled those of the altered enstatites 

 (diaclasite or bastite, &c.) that I was for some time inclined to 

 refer it to one of those species. This conclusion appeared to be 

 confirmed by a study of slices under the microscope ; for while its 

 extinction appeared to be that of a rhombic mineral, it was seen to 

 contain great numbers of tabular inclusions, like those found in, 

 and often regarded as characteristic of, enstatite, bronzite, and 

 hyperstheue. A thin flake of substance which was examined by 

 Mr. Fletcher in a polariscope with convergent light seemed to indi- 

 cate that the mineral is a uniaxial one, however, and I accordingly 

 determined to attempt to isolate the mineral and submit it to analysis. 



The isolation of the mineral proved, in this case, to be a task of 

 considerable difficulty. Prom the powdered rock the magnetite and 

 the minerals which adhered to it were removed by an electro-magnet. 

 The irregular grains of serpentine, it was found, could be separated 

 from the flaky particles of the micaceous mineral and of hornblende 

 by repeated washings in water. But the latter two minerals 

 were found to be only very imperfectly separated, owing to their 

 similar density, by the boro-tungstate solution ; and enough material 

 for a partial analysis could in the end be only obtained by laborious 

 picking under the microscope. I am greatly indebted to my 

 assistants in the geological laboratory of the i^ormal School of 

 Science, Mr. G. Cole and Mr. W. Atkinson, for the patience and 

 care with which they carried out this process of separation and the 

 subsequent chemical operations. These operations were necessarily 

 performed upon extremely small quantities of material, so that the 

 results obtained from the several different determinations did not 

 show a very close agreement. They were, however, sufficient to 

 prove that the mineral had a composition which may be approximately 

 represented as follows : — 



Silica 38-0 



Alumina 13-0 



Iron oxide calculated as ferric oxide .... 4*5 



Lime 5*0 



Magnesia 24*0 



Water 6-5 



Alkalies ? by difference 3*0 



