124 PROCEEDINGS OF THE CAXADIAX INSTITUTE. 



of a fluid magma, a distinctly gneissic striictui'e may be developed, as 

 I have shown to be the case in the instances cited, then the argument 

 that gneissic foliation is an indication of bedding is terribly shaken ; 

 and if, further, by extending observation, the foliation of gneisses gen- 

 erally can be shown to be due to similar forces operating on similar 

 material, the metamorphic theory must go to the wall as a needlessly 

 far-fetched and complicated hypothesis. I am well aware that the 

 metamorphic theory is not altogether based on the structure of 

 gneisses, nor do 1 deny that there is a metamorphism to which rocks 

 are susceptible, but as an explanation of the origin of the Lauren- 

 tian gneisses it seems to be very weak, [)articularly where it seeks 

 for support in the assumption that foliation in these rocks is proof of 

 sedimentation. 



If we consider for a moment the probable causes of the differentia- 

 tion of a granitic magma into a foliated gneiss, such as occurs in 

 d^^kes of gneiss, we shall find that pi-essure was in all likelihood the 

 most important force at work ; further, that this pressure was due 

 not to external disturbance but to internal changes ; that it was, in 

 fact, the pressure due to the expansion of the injected material during 

 the process of solidification within confining walls practically rigid. 

 Wadsworth has pointed out in his Lithological Studies the high 

 probability of rock magmas following the rule found to hold in many 

 of the economic metals. These expand rapidly during the actual 

 process of solidification, so that when solid, but at a temperature 

 only slightly below the point of fusion, they have a lower specific 

 gravity than when in the fused state ; though in cooling from this 

 temperature the solid contracts so much as at the normal temperatuie 

 to have a specific gravity higher than that of the molten fluid. An 

 expansion of this kind in a dyke solidifying by crystallization can be 

 readily imagined to have had a very potent influence upon the 

 arrangement of the nascent crystals. The pressure would of course 

 be equal in all directions if the magma weie a perfect fluid. As 

 crystallization proceeded, however, the pressui-e would act more and 

 more in directions pei-pendicular to the planes of resistance, i.e. to 

 the walls of the dyke, and as a consequence the crystals would be 

 forced to assume positions in which their longest diameter would be 

 pai'allel to the lines of least resistance. Such an explanation of the 

 development of a gneisf,ic structure in dykes may be extended with 

 some probability of truth to the gneisses of the Laurentian, par- 



