SUBOCEANIC STRUCTURE 297 



underlies a mass of gneiss and is subject to heating from below, the heat 

 will not be conducted away by the gneiss as fast as it is supplied through 

 the basalt. It must therefore accumulate and raise the temperature of 

 the basalt until the basalt melts. Since pressure tends to restrain the 

 basalt from melting, the critical temperature will first be reached where 

 the pressure is least — that is, on top. 



(2) If a body of basalt lies above a mass of gneiss or is isolated in it, 

 it can receive heat only as fast as heat is conducted through the gneiss, 

 and it will melt sooner than the gneiss only because its melting point is 

 lower. 



(3) A dike or dikes of basalt extending from a basaltic mass upward 

 through gneiss would serve as heat conduits and might conduct the heat 

 away so rapidly that the gneiss would not reach the melting point. 



These considerations bring us to the question of structure. It is postu- 

 lated that holocrystalline rocks exist in the lithosphere only in so far as 

 they have, because of youth or peculiar local conditions of stress and 

 temperature, escaped the effects of metamorphism. The prevailing struc- 

 ture of deep-seated rocks is believed to be a foliated one; hence it is 

 appropriate to use the term gneiss rather than granite to designate the 

 acid rock of the lithosphere. The attitude of the foliation may be con- 

 ceived to be vertical, horizontal, or inclined. 



In addition to foliation, shearing planes or shearing zones demand 

 recognition. They must necessarily develop in all parts of the lithosphere 

 in which the ratio of strength of rock to superincumbent load is less 

 tlian 1. They must develop in response to gravitative stress, and their 

 general attitude should be approximately at 45 degrees to the vertical. 



In case of intrusion of the gneiss by molten magma, the liquid rock 

 should be guided in its course by the foliation and by the shearing planes. 

 There would result a structure which I propose to call a bonded structure. 



The term bonded is borrowed from architecture in the use which signi- 

 fies tied together by interpenetration of integral parts. The intrusive 

 and the intruded rocks are interbedded along the foliation, and the lenses 

 of the intrusive are connected by cross-cutting bodies, dikes, along the 

 shearing planes. 



The origin of a bonded structure is obvious. Hypothesis presupposes 

 a large mass of basic rock underlying a mass of gneiss. The basic rock, 

 being subjected to a rising temperature, melts at and immediately below 

 its contact with the gneiss and thus becomes lighter than the solid gneiss. 

 When the molten mass attains sufficient extent and depth, the gneiss will 

 tend to bow down into it, pressing the liquid magma out at the margins 

 between the foliation planes. Escape of the magma must gradually occur 



