152 MB. G. BAKROW ON THE OCCURRENCE OF [May 1 898, 



for some hours — in fact till it ceased to tinge the acid solution 

 j'ellow.^ It was again elutriated under the tap, and the residue, after 

 drying, passed through a borotungstate solution of sp. gr. 2*9. 

 This practically eliminated nearly all the quartz and white mica, 

 except such as adhered to the chloritoid. The residue was now 

 passed under a powerful electro-magnet, which picked up the 

 chloritoid and dropped nearly all the white-spot material. (This it 

 will not do unless the material ha? been first strongly acted upon 

 by acid, and a large part of the iron removed.) Some of the white 

 material was still taken up by the magnet, but examination showed 

 that its surfaces were greatly etched and corroded. About 2 grammes 

 of the now nearly pure chloritoid was gently ground in an agate 

 mortar, washed, and dried. The powder was placed in a concen- 

 trated solution of cadmium borotungstate, and now the etched 

 surfaces of white material, owing to their air-retaining nature, 

 floated much longer than the chloritoid, which was drawn off almost 

 pure. An analysis was made of this material and is given below. 

 It is so nearly identical with that given by Barrois, to show the 

 composition of the large crystals fi-om the schist in the lie de Groix, 

 that the two are printed side by side : — 



SiO.^ ... 





Kincardine. 



26-00 



Ile de Groix, 

 24-90 



ALO,.. 



4005 



40-36 



2 3 



FeO ... 





19-50 





Fe,0^ 



505 



I 26-17 



MgO... 



Loss on 



ignition .... 

 Total 



2-88 

 6-00 



2-54 



6-23 





99-48 .... 







100-20 



It will thus be seen that, both optically and chemically, the 

 mineral here described is substantially identical with that analysed 

 by Dr. Barrels from the lie de Groix. Put there is a great dif- 

 ference between the rocks containing the mineral, for the chloritoid 

 in the foreign specimen occurs in a decomposed sillimanite-gneiss, 

 and, as our experience from Kincardine would lead us to expect, the 

 crystals are large in proportion to the very high metamorphism 

 that the nature of the rock suggests. In a specimen kindly supplied 

 to me by Dr. Barrois one crystal weighed 2 grammes. Curiously 

 enough, films of the rock-matrix traverse these crystals in exactly 

 the same manner as do minute threads in the exceptional crystals 

 obtained on the Kincardineshire coast ; and further, both show 

 polysynthetic twinning. The part played by this mineral seems fairly 

 evident. Owing to the exceptional composition of the containing 

 rock, it entirely replaces the characteristic brown mica so abundant 

 in the Highland schists. Indeed, in a single slide from the margin 

 of the rock, we may find alternate films, one of which contains 



^ After this prolonged heating the flakes were examined carefully, to make sure 

 that they had not been acted on by the acid. The four analyses that have been 

 made of the mineral in diffei-ent stages of purity show this quite clearly, for in 

 two cases acjd was not employed ; yet these two correspond exactly with the 

 others, allowing for more impui'ity. 



