558 Dr. J. W. Evans — Presidential Address. 



and 1,550° C, but they separate out from the complex magmas from 

 which our igneous rocks were formed at lower temperatures, rarely 

 much exceeding 1,200° C, and frequently considerably less. 1 



It has been found possible at the Geophysical Laboratory to 

 maintain a temperature of 1,000° C. or more under a uniform 

 pressure of 2,000 atmospheres for so long a time as may be desired, 

 and, what is equally important, the temperature and pressure attained 

 can be determined with satisfactory accuracy, the temperature within 

 2°C, and the pressure within 5 atmospheres. 



It has been ascertained that such uniform pressure as would 

 ordinarily be present at the depths mentioned does not directly affect 

 the physical properties of minerals to anything like the same extent 

 as the difference between the temperature prevailing at the earth's 

 surface and even the lowest temperature at which igneous rocks can 

 have been formed. It has, however, a most important indirect action 

 in maintaining the concentration in the magma of a considerable 

 proportion of water and other volatile constituents 2 which have 

 a far-reaching influence in lowering the temperature at which the 

 rock-forming minerals crystallize out, in other words, the temperature 

 at which the rock consolidates, and in diminishing the molecular 

 and molar viscosity of the magma, thus facilitating the growth of 

 larger crystals and the formation of a rock of coarser grain. They 

 must also be of profound significance in determining the minerals 

 that separate out, the order of their formation, and the processes of 

 differentiation in magmas. 



It is, therefore, obvious that any conclusions derived from the 

 early experiments which were carried out with dry melts at normal 

 pressures must be received with very considerable caution. Nor does 

 much advance appear to have been made, even at the Geophysical 

 Laboratory, in experiments with melts containing large amounts of 

 volatile fluxes, and yet, if we are to reproduce even approximately 

 natural conditions, it is absolutely necessary to work with magmas 

 containing a proportion of these constituents, and especially water, 

 equal in weight to at least one-third or one-half of the silica present. 

 This will obviously present considerable difficulties, but there is no 

 reason to doubt that it will be found possible to surmount them. 



A much more formidable obstacle in realizing the conditions under 

 which rocks are formed is the small scale on which our operations 

 can be carried on. There are important problems connected with the 

 differentiation of magmas, whether in a completely fluid or partly 

 crystallized state, under the action of gravitation, for the solution of 

 which it would seem for this reason impossible to reproduce the 

 conditions under which nature works. Instead of a reservoir many 

 hundreds of feet in depth, we must content ourselves in our 

 laboratory experiments with a vertical range of only a few inches. 



1 It is probable that the temperatures recorded in some lavas higher than 

 the melting-point of copper, which is well over 1,200° C, are due to chemical 

 reactions, such as the oxidation of hydrogen, carbon monoxide, ferrous oxide, 

 and perhaps sulphur. See Day & Shepherd, Bull. Geol. Soc. Am., vol. xxiv, 

 pp. 599-601, 1913. 



2 Johnston, Journ. Franklin Inst., January, 1917, pp. 14-19. 



