Vol. 54,] CHLORITOID IN KINCARDINESHIRE. 153 



biotite to the exclusion of chloritoid, while the adjacent film contains 

 chloritoid to the exclusion of biotite. To a certain extent the silli- 

 manite-gneiss from the He de Groix shows the same characteristic. 

 The rock is greatly decomposed, but there does seem to have been 

 a little brown mica in it; still, the total amount must be quite 

 insignificant when compared with the great quantity of chloritoid 

 present in the large crystals. In ordinary silliraanite-gneisses 

 biotite is one of the chief constituents. It will be interesting to 

 learn to what extent biotite is similarly replaced in other areas. 



Variation in Composition of the Chloritoid. 



Examining the list of analyses of this mineral given by Dana, 

 one is struck by the fact that no two of them have the same compo- 

 sition, and consequently either the mineral must vary much in 

 composition, or else the extreme difficulty in eliminating impurities 

 has led to a belief in a variation that does not exist. Is not the 

 latter the more reasonable supposition? Four analyses^ of the 

 Kincardineshire mineral have been made from specimens varying in 

 purity. The first gave nearly 1*5 per cent, of alkalies, and the last 

 merely a trace. On completing the first analysis it was seen that 

 a large quantity of white mica must be adhering to the jagged 

 edges of the chloritoid, but the microscope fails to make this at all 

 clear. Obviously, it was quite improbable that an appreciable amount 

 of alkali would be present. At intervals during the last 3 years 

 efforts have been made to rid the mineral of impurity. This I have 

 practically succeeded in doing, with the result that the chemical 

 composition proves in reality to be substantially identical with that 

 of the mineral from the He de Groix. As the crystals in this rock 

 are so large that a practically pure sample can be obtained in an 

 hour, we may fairly regard it as the standard or type. 



Had not the above efibrts to obtain pure material been made, I 

 should simply have added one more to the list of chloritoids having 

 exactly the same optical properties, but with a very variable com- 

 position. Now, if we turn to the paper by A. Eenard & C. de La Vallee 

 Poussin,^ we see that the fine crystals from Serpent are actually 

 figured. This figure shows that the mineral is permeated with 

 quartz, so disseminated through it that the quartz must be very 

 difficult to eliminate. If our suggestion that the chief variations 

 in composition are really due to impurity be correct, then the 

 Serpont specimen should show an abnormally high silica per- 

 centage (41-65). This is exactly what it does, and Dana lays 

 special stress on this point, using this high silica-percentage to 

 justify a different name for the mineral from that of chloritoid having 

 the composition given by Barrels and myself. The identity of all 

 three minerals, so far as optical properties go, is rendered clear by 

 the careful descriptions given in the paper referred to above. By 

 sedulously noting the composition of the matrix in which these 

 minerals are embedded, and especially the material in contact with 



^ Mr. W. M. Hutchings kindly made the second for me. 



2 'Note 8ur rOttrelite,' Annal. Soc. Geol. Belgique, vol. vi (1879) p. 51. 



