254 DE. H. H. THOMAS OlS" XEIfOLITHIC [vol. Ixxviii,. 



We now come to another and obviously later phase of meta- 

 morphisin. that has been superinduced upon the buchite (h) and 

 the cr3^stalline zone (c). Disruption of the already cr3'stallized 

 anorthite-corundum-spinel assemblage of the hybrid zone, by 

 stresses acting within the basin prior to the upward intrusion of the 

 magma, allowed the locall3''-modiiied magma mentioned above to 

 make fresh contacts with the buchite, and wdth the more aluminous 

 portion of the anorthite-corundum-spinel zone adhering to the walls 

 of the magma-hasin. This partial magma being relatively rich in 

 magnesia, iron, and alkalies, the new contacts resulted in a fresh 

 set of concentrations. Cordierite (p. 245) under the new con- 

 ditions was one of the chief crystalline substances to separate out ; 

 while the rapid enrichment of the melt by alumina from the re- 

 sorbed sillimanite, taken in conjunction with the excess of magnesia 

 be3"ond that required for the formation of cordierite, produced an 

 almost simultaneous precipitation of corundum and spinel. The 

 melt, being of a more alkaline character than previously, caused 

 the resorption of much of the earlj^-formed anorthite, and onh' 

 permitted the subsequent sepai*ation of a plagioclase of increased 

 alkalinit}^ (p. 245). 



Reviewing the temperature-conditions, as suggested by the crys- 

 talline phases of the accidental and cognate xenoliths, we are led 

 to the conclusion that the tempera tui-e-interval during which the 

 metamorphism of the aluminous material was accomplished was 

 small, and probably lay between 1400° and 1250° C. On the other 

 hand, the time-interval was great, for extremely slow cooling and 

 no considerable change in the viscosity of the melt are indicated. 



Towards the end of this period, a limit only imposed by the 

 upward intrusion of the magma, that magma had commenced to 

 crystallize on its own account, with the formation of the cognate 

 xenoliths at a temperature that did not exceed 1300^0., and was 

 probably still lower. At the time of upward intrusion the tem- 

 perature was sufficient to keep molten the glassy matrix of the 

 sillimanite-buchite, and to produce tridjanite-fringes upon the 

 quartz-grains of the Carsaig Sandstone. 



Experiments on the glass}^ matrix of the sillimanite-buchite 

 prove that it melts at a temperature near that of fluorspar 

 (1250° C), so that we may suggest that the temperature of the 

 magma at the time of its intrusion approximated to this value. 



It will be seen from the above that the probable range of tem- 

 perature covering the metamorj)hic processes that operated within 

 the basin, and the subsequent upward intrusion of the magma, is 

 surprisingly small, possibly less than 200° C. 



At the time of upward intrusion all further transfusive action 

 between magma and xenoliths would be checked by rapid cooling, 

 and this would account for the sharp boundaries presented by the 

 xenoliths to the surrounding tholeiite. 



