REPORT OF TEE CHIEF ASTRONOMER 753 



SESSIONAL PAPER No. 25a 



This surplus heat energy is available for the fusion and assimilation of 



eountry-roek. There are good reasons for believing that the average wall-rock 



of granite batholiths has the composition and crystallinity of a granitoid gneiss. 



For purposes of calculation this will be assumed to be the fact. The average 



temperature of the wall-rock before an abyssal intrusion may be conservatively 



estimated from the normal temperature gradient to be 200° C. In order to raise 



the gneiss to the temperature of 1200°, where it is just molten, about 410 calories 



(assuming latent heat at 90 calories — a value estimated by Vogt for the silicates) 



per gram must be supplied from an outside source. If all the superheat of the 



55 

 basalt were available for melting (not dissolving) gneiss. — — of mass-unit of 



gneiss would be melted by mass-unit of the superheated basalt; or about 7-5 

 mass-units of the basalt would melt a mass-unit of wall-rock. 



Such simple melting would, however, not occur. There are plenty of field 

 and laboratory proofs that molten basalt, even slightly superheated, will dissolve 

 fragments of gneiss and allied rocks. The mutual solution of two contrasted 

 silicate mixtures takes place at a certain temperature which is lower than the 

 melting point of either one. The simple contact of two such materials suffices 

 to cause their mutual solution at that lower temperature.* This fundamental 

 law of physical chemistry has been experimentally demonstrated for silicates by 

 Vogt and by Doelter and his pupils, although the last mentioned authors have, 

 perhaps, not sufficiently regarded the fact that it takes considerable time for the 

 mutual solution to take place.f 



Petrasch has experimentally shown that, when two parts of limburgite and 

 one part of granite are mixed and heated, they melt together at 950° C. and 

 the solution remains fluid down to 850° C.% Predazzo granite softens at 1150° C. 

 and the limburgite at 995° C.§ In this case, there is a lowering of 200°-300° 

 below the melting point of granite and. 45°-145° C. below that of limburgite. 



It seems highly probable, thus, that gneiss-xenolith and basalt would form 

 a solution or syntectic film which is molten at a temperature at least 100° C. 

 below the fusion-point of basalt at the average depth of ten kilometres or less 

 below the earth's surface. At those depths basalt melts at about 1150° C. ; the 

 syntectic would be molten at or below 1050° C. If the syntectic film were con- 

 tinuously removed during the sinking of the block or by the currents inevitably 

 set up during stoping, nearly all of the superheat of the basalt might be used 

 in dissolving the gneiss. The total melting-heat of gneiss, if molten at 1050° C, 



*Cf. O. Lehmann, Wiedemann's Annalen der Physik, Vol. 24, 1885, p. 17. 



t See J. H. L. Vogt, Christiania Videnskabs-Selskabets Skrifter math.-naturv. 

 Klasse, 1904, No. 1, p. 191; and Tscherm. Min. u. Petrogr. Mitth., Vol. 24, 1906, p. 473. 



JK. Petrasch, Neues Jahrb, fur Min., etc., Beil. Bd. 17, 1903, p. 508. Petrasch 

 mixed the powders of one part of granite (softens at about 1150°C.) with two parts of 

 hornblende andesite (softens probably about 1050°C.) and found the mixture to become 

 molten at 900°C, proving again an important lowering of the melting-point below that 

 of either rock. Basic rock thus acts as a flux for granite (or gneiss) to an extent 

 comparable with that proved by Petrasch and others for lithium chloride, calcium 

 fluoride, ammonium chloride, and sodium tungstate. 



§C. Doelter, Tscherm. Min. u. Petrogr. Mitth., Vol. 20, 1901, p. 210. 



