Volcano Ôsliima, Ll/.u. ll'> 



of the i-esiJual liquid after the mugma has been extruded as lavas. 

 At such a very rapid rate of coohng as prevailed at the time of 

 consohdation of the groundiuass, the liquid crystalUzes quickly 

 uninfluenced by the crystals in contact with it, there being Uttle 

 •opportunity or none for the adjustment of equilibrium between 

 liquid and crystals and for the sorting of the crystals separating. 

 The composition of the groundmass as a bulk must therefore 

 -correspond to that of the residual liquid. 



This residual liciuid in the magma basin is thought to have 

 been practically constant in composition, since its temperature is 

 supposed to have remained nearly the same throughout the whole 

 volcanic history so far as represented by the observed lavas (p. Ill) ; 

 accordingly it is expected that no very appreciable variation will 

 be seen in the composition of the groundmass of our rocks. 



What has been actually observed is in harmony with the 

 above expectation as to the composition of the groundmass, though 

 there is a rather wide variation in its texture— cry stall inity, 

 granularity, and fabric—, which may be merely due to varying 

 conditions at the time of extrusion. 



A difference has been seen, however, in the composition of 

 the plagioclase microlites in the groundmass, from sodic labradorite 

 to sodic bytownite. This would appear to be at variance with 

 what has been stated al)Ove, but it must be recalled that, on the 

 separating out of mix-crystals from the liquid, the composition of 

 these differs according to the proportion of the whole liquid that 

 consolidates as crystals, even if the composition of the original 

 li(iuid was the same. This may account for the diflference in the 

 composition of the plagioclase crystals in the groundmass. 



