'390 w. (L-^iirnKLL smith. 



Avliicli plienocrysts have been pushed aside by the outgrowing 

 felspars shows that, at the time of tlieir crystallisation, the 

 viscosity of the magma was already considerable. 



Now this development at one moment of crowds of crystals 

 with a long-prismatic habit has been shown by Miers^ to take 

 place when a solution is brought into tlie labile state by 

 reason of a fall in temperature or increase in concentration. 

 It appears that the rapid cr\'stal]isation of the augite plieno- 

 crysts carried the composition past the eutectic point and 

 made the magma supersaturated with respect to labradorite. 

 If the cooling were fairly rapid the supersolubility curve for 

 labradorite would be reached before any felspar crystallised 

 out. There would then be a sudden development of long thin 

 crystals of labradorite, the fall in temperature would be 

 checked, and, the composition of the magma returning to- 

 wards the eutectic, the remainder of the augite Avould com- 

 mence to crystallise out, filling up the spaces between the 

 felspars while these were still growing. 



It is not known Avhether any parts of the Vryheid dolerite 

 show a glassy development; but it will be observed that in 

 all other cases of spherulitic basic rocks they are always 

 associated with rocks which were once glassy. 



In many spherulitic tachylytes the spherulites have been 

 developed in a zone lying between the pure glass and the 

 completely crystallised rock. The temperature conditions 

 favourable to spherulitic growth are evidently intermediate 

 between those obtaining for the glass and for the holo- 

 crystalline rock. 



It seems that a magma approaching to a eutectic composi- 

 tion will give an ophitic or even micrographic structui-e if 

 enabled to crystallise from the metastable state. If it reaches 

 the labile state there will be a sudden development of spheru- 

 litic structure. Finally, if the temperature falls so rapidly 

 that the magma becomes too viscous to allow of rapid 

 crystallisation a glassy rock w411 result; the very small 

 dimensions of spherulites in many glasses is no doubt due to 

 ' H. A. Miers. • Science Progress," ii (1907). pp. 121-134. 



