50 



characters are not those of any known variety of chlorite, nor 

 yet of chloritoid, but they agree more closely with those of 

 the former mineral. 



The mode of occurrence of andalusite is noteworthy ; it is 

 in ovoid to irregular porphyroblasts up to 3 mm. in diameter 

 and exhibits well the sieve structure of Becke and Gruben- 

 mann, being crowded with inclusions chiefly of quartz and 

 iron ore. These are arranged along definite lines, generally 

 curved, in the host, but the directions apparently do not bear 

 any definite or constant relationship to the cleavage planes or 

 to the schistosity of the rock, to which, indeed, they are in- 

 differently parallel and perpendicular. The appearance in places 

 reminds one of helicitic relict texture, but it is plain that 

 the two things have nothing in common. The phenomenon 

 may possibly be attributed to some deorientating force pos- 

 sessed by the growing andalusite and exerted against the 

 minerals which remained undigested in the process of forma- 

 tion of the andalusite itself, a process which occurred subse- 

 quently to the schistose recrystallization of the rock and as a 

 result of the igneous intrusion. 



The andalusite in its growth has made room for itself 

 partly by mechanical and partly by chemical means, the 

 schistose laminae being sometimes pushed out, and at other 

 times cut across and corroded. Both effects are well seen in 

 the microphotograph (pi. ii., fig. 2). There is no reason to 

 belive that rotation of the andalusite under pressure has 

 occurred. Indeed, the evidence is all against this. 



It has been noted that, whereas tiny inclusions of iron 

 ore ( ? ilmenite) are quite a feature of the andalusite, they 

 are by no means so abundant elsewhere in the rock, which 

 suggests that they represent the iron rejected from the biotite 

 used up in forming the andalusite. 



In another schist from close to the syenite contact the 

 development of andalusite with sieve structure, of a very 

 similar type to that just described, has very plainly followed 

 the former lamination of the rock, and is possibly due to the 

 presence of strongly aluminous layers at intervals in the 

 original sediment. Though the schistosity is here at 40° to 

 the bedding planes it is not distorted in the neighbourhood 

 of the andalusite, and likewise there is no sinuous arrange- 

 ment of inclusions. There is a tendency for the development 

 of little knots or segregations of sericite in parts of the slide, 

 which in shape and general arrangement are very similar to 

 the andalusite. 



Biotite is slightly bleached and tiny zircon inclusions are 

 fairly numerous. Some chlorite in crystals of quite distinctive 

 blade-like habit is found, generally associated with the 



