138 NATURAL SCIENCE. Feb.. 



But it will be argued that this segregation will occur in the 

 primitive igneous material after it has liquefied by diminished 

 pressure, and is on its way towards the surface. Then, it is true, 

 it frequently occupies tortuous fissures where differentiation and 

 segregation or sedimentation of crystals of the constituents might go 

 on ; but this could hardly explain the regular flow of definite but 

 very different types of rocks from numerous points along a line of 

 active vents, such, for instance, as the Lipari Islands. Here we 

 observe, along a stretch of some 50 miles in length, a series of numerous 

 vents from which outpours of basic dolerites and andesites simul- 

 taneously took place, and overlapped effusions of intensely acid 

 lavas. In Vulcano for 2,000 years eruptions of an obsidian lava and 

 a dolerite have been going on at vents only a mile apart. Is it con- 

 ceivable that differentiation could progress, in the great fissure 

 supplying these volcanoes, to such a complete extent that, without 

 a difference of time, and in such a limited area, two extreme rocks 

 could be poured forth and almost practically none of intermediate 

 composition ? What is said of this region is no less applicable to 

 many others. 



There yet remains one other possible cause, foreshadowed as far 

 back as 1876 by the late Mr. J. C. Ward; that is, the interaction of 

 the primitive igneous paste and the rocks it traverses or comes in 

 contact with in its way into and through the earth's crust. Mr. 

 Ward did not state clearly whether he meant by simple fusion of the 

 walls of a volcanic canal or by a process of osmosis. 



In my paper on Vesuvius and Monte Somma,7 I drew particular 

 attention to the subject, and in later writings those views have been 

 extended and confirmed, but it will be advisable here to discuss the 

 question in some detail. When we observe the denuded roots of 

 volcanoes we find a certain amount of rock that has disappeared and 

 has been replaced by the igneous intruder. We have little evidence 

 in most cases to tell us whether the removal has been by mechanical 

 export, fluxion, or fusion, though in many cases evidence points to 

 the first as the principal means. I think it may be stated as a 

 general law that contact metamorphism increases in intensity with 

 the bulk of the igneous intrusion and the coarseness of its crystal- 

 lisation. In other words, it is greater the larger is the amount 

 of the heated medium, and the longer time this takes before 

 solidifying, and, therefore, before its subsequent cooling. In con- 

 tact metamorphism in regions which show evidence of being for a 

 long time active volcanic centres, very marked chemical changes have 

 been wrought in the solid rock — elements have been introduced and other 

 ELEMENTS HAVE BEEN REMOVED. All the elements introduced are 

 derived, I presume no one will deny, from the neighbouring paste, 

 which will have been modified in composition according to the 

 percentages of those elements that passed into the cavity walls. 

 "^ Quart. Journ. Geol. Soc, vol. xl., p. 54. 



