A SPHIiEULlTIC DOLERITE FRO.AI YRYHEID, NA'l'AL. o91 



tills prevention of crystallisation by the increasing viscosity of 

 u rapidly cooling* magma. 



It is very difficult to say how far one is justified in drawing 

 an analogy between the aqueous solutions studied by Miers 

 and Miss Isaacs and the complex solutions represented in a 

 rock magma. However, the facts recorded above seem to 

 show that supersolubility curves and " hypertectics" may be 

 of much more importance in controlling crystallisation than 

 has been supposed. 



In conclusion I wish to acknowledge my indebtedness to 

 both Dr. G. T, Prior and Mr. L. J. Spencer for much kind 

 help in the preparation of this paper. 



EXPLANATION OF PLATES XXYJII-XXX, 



Illustrating Mr. W. Campbell Smith's paper, " On a Spheru- 

 litic Dolerite from Vryheid, Natal." 



PLATE XXVIII. 



X f . Weathered spherulitic dolerite from the upper sill (left liand 

 side) and lower sill (right hand side). 



PLATE XXIX. 



Fig. L— X 26. Section near tiie centre of a spherulite, showing the 

 anungement of the felspars. 



Fig. 2. — X 26. Section across the plane of contact of two spherulites. 



These two figures also show the characteristic habit of the augite 

 phenocrysts and their secondary zones. A large crystal of enstatite, 

 liroken in two, is seen in fig. 1. 



PLATE XXX. 



Fig. 1. — X 140. Cross-sections of felspar crystals showing the way 

 in which they enclose cores of angite which can hardly be distinguished 

 from the augite growing between sepai-ate felspar crystals. 



Fig. 2.— x 140. A long thin crystal of augite stretching right aci-oss 

 the field shows its secondary zone and the peculiar manner in which 

 this zone grows in between the felspars. 



