732 DEPARTMENT OF TEE INTERIOR 



2 GEORGE V., A. 1912 



tention is sought in the facts of the internal contacts, at which magmas are 

 sometimes seen to be modified by the incorporation of the country-rock. This 

 second view doubtless appeals the more strongly to the majority of those 

 geologists who have actually to do with granitic bodies in the field. In fact, the 

 impression has prevailed among some of them that the ' laccolithic theory ' is as 

 widely held as it is because of its apparent necessity in the prevailing theory of 

 rock differentiation. Yet it must be considered as conclusively proved for the 

 great majority of stocks and batholiths investigated, that analysis has not yet 

 shown that the second or ' assimilation ' theory really meets its own crucial test, 

 the chemical and mineralogical blood-relationship between the average intrusive 

 rock and its country-rock along their mutual contact. Currents within the 

 magma would, of course, tend to remove and diffuse products of assimilation 

 from molar contacts; but it is extremely doubtful that they could so completely 

 mask the expected results of the process as is over and over again illustrated in 

 nature. No single fact concerning granite, for example, is more striking than 

 its astonishing homogeneity in contact with argillite, limestone, crystalline 

 schist, or basic igneous formation — a homogeneity that persists, too, from contact 

 to center of the eruptive. In the very common case where the assimilated 

 product is more acid than the original magma, it would tend to rise through 

 the latter, slowly diffusing in the journey. The upper part of the magma basin 

 should, for that reason, become filled with mixed magma more silicious than 

 the original. Heterogeneity, even stronger vertically than horizontally, would 

 be expected in a diorite or gabbro magma cutting crystalline schists, or in a 

 granite magma cutting heavy beds of sandstone or quartzite. True thermal 

 convection currents must, under these conditions, be greatly weakened by the 

 strong differences in density of the original magma and the magma diluted, so 

 to speak, by more silicious material. In the absence, then, of the only kind 

 cf current likely to be set up in the process of cooling and mere caustic solution 

 on molar contacts, the diffusion of the diluted magma would take place only with 

 extreme slowness.f Yet, up to the present time, this consequence of con- 

 siderable vertical heterogeneity under the stated conditions has not been demon- 

 strated in nature. The recorded field discoveries point, on the contrary, to a 

 distinct failure of the known facts to match the deduction from the theory. The 

 few proved instances of endomorphic changes of magmas by assimilation (e.g., 

 the granite of the Pyrenees described by Lacroix) serve, by their conspicuous 

 and exceptional nature, to show that granitic magmas, if they have ' made their 

 own way ' at all, have usually done so in some manner different from merely 

 assimilating the invaded formations on molar contacts. 



A main feature of the explanation by marginal assimilation is the immense 

 superheat demanded in the magma. If the solution of invaded formations is 

 directly performed by the liquid rock, the available superheat must be speedily 

 exhausted in supplying the latent heat represented in the solution, to say noth- 

 ing of the loss by conduction through the earth's crust. The latent heat, accord- 

 ing to Vogt and others, is at least 20 per cent of the total melting heat in 



t Cf. G. F. Becker, Amer. Jour. Sei., Vol. 3, p. 30, 1897. 



