R. A. Daly — -Mechanics of Igneous Intrusion. 273 



the foreign inclusions to become rapidly rarer, until, in the heart 

 of the eruptive area, one may go hundreds of yards or even 

 several miles without discovering any such inclusions. If there 

 are hundreds of them in a given part of the contact-belt at the 

 present surface (evidently a chance section exposed by ero- 

 sion), the natural inference that there are thousands or millions 

 of others enclosed in the eruptive below the level of the vis- 

 ible contact, is clearly permissible. Another legion of them 

 has been destroyed along with their matrix in that part of the 

 igneous body removed by denudation. It is manifest, further, 

 that the rifting of the blocks has so far enlarged the chamber 

 occupied by the eruptive. That is, the walls are, on the aver- 

 age, farther apart because of the rifting. The question arises 

 as to whether the chamber may owe the greater part of its 

 present size to a long continuation of the selfsame process, with 

 a simultaneous removal from the visible chamber of the blocks 

 formerly rifted off. The affirmative answer to this question 

 is the kernel of the hypothesis to be proposed. 



Strangely enough, the explanation of the presence of for- 

 eign blocks within igneous bodies along the molar contacts and 

 the equally conspicuous rarity of such fragments toward the 

 centers of the bodies, has not been, so far as the writer is aware, 

 adequately discussed anywhere in geological literature. How 

 blocks still close to their former homes in the country rock 

 could be suspended in the magma until crystallization of the 

 latter was complete, and whether the normal effect of their 

 complete immersion would be to permit of their floating 

 upwards or sinking downwards in the magma, are questions 

 not as yet properly answered. It will be seen that these 

 queries are of prime importance to the ensuing hypothesis and 

 that the attempt has been made to answer them by correlating 

 experimental and other data acquired for penological science 

 within recent years. We may, for the present, assume the 

 generally accepted high liquidity of normal plutonic magmas, 

 though in the sequel certain arguments therefor will be given 

 in brief outline. 



The relative densities of holocrystalline, glassy, and molten 

 rock at ordinary atmospheric pressure. — As the foundation of 

 the argument, the very careful and fruitful experiment in the 

 dry fusion of diabase by Barus must be briefly described.* 



With a much greater refinement and accuracy of method 

 than that practised by any of his predecessors, Barus has meas- 

 ured the density of a sample of normal diabase at normal pres- 

 sure (one atmosphere) and temperature (20° C), and compared 

 therewith the densities of the completely melted rock and of 



*Phil. Mag. (5), xxxv. 173 (1893), andU. S. Geol. Surv. Bull. No. 103(1893). 

 Of. Joly, ou the fusion of basalt, Trans. Boy. Dublin Soc. (2), vi, 298 (1897). 



