The BocJcs oj Tristan cVAcunha. 35 



depths. These residual masses of rock are not composed of materials 

 of great specific gravity such as we would expect if the volcanic 

 vents really did go down to enormous depths, but are precisely similar 

 in average density to the ordinary sedimentary rocks of the surface. 

 In South Africa we have had our attention riveted to two classes of 

 igneous action : the intrusion of basic rocks in strata already exist- 

 ing and the extrusion of lavas on the surface. Both are probably 

 accompanied with the formation of vents filled with non-volcanic 

 material, these tuff- and boulder-filled pipes of the former group 

 being connected with the great dolorite intrusions to which, however, 

 they are subsequent ; while the lavas are connected with the boulder- 

 and tuff-filled necks which were formed both previously to, and con- 

 temporaneously with, the extrusion of liquid rock. 



There is a third kind of volcanic activity which finds its expres- 

 sion in the vast masses of amygdaloidal melaphyre that apparently 

 welled up through great fissures without the accompaniment of vol- 

 canoes. In the Colony Mr. Eogers and I have studied them in the 

 Prieska Division,'" and they are in all respects similar to the Bush- 

 veld amygdaloid in the Transvaal. These are probably connected 

 with vents filled with rhyolite and acid breccias which occur in the 

 Hope Town Division, but they are too little understood to contribute 

 to my argument. All we know about them is that they are exceed- 

 ingly ancient, being found intruded among rocks which, if litho- 

 logical similarity can be taken as a guide, are Archgean, the Indian 

 Vindhyan rocks and the Lake Superior iron-bearing series, being the 

 same not only in composition but in the order of the superposition 

 of the several members. 



The first two groups are connected with great movements in the 

 earth's crust. 



In Willowmore we get masses of quartzite belonging to the Table 

 Mountain series crushed and ground to powder as if caught in the 

 jaws of an enormous press. The brecciated rock occurs at the 

 crossing of two systems of folds. The cubic content of one of these 

 masses in Baviaan's Kloof, which could be actually seen and 

 measured, was over a cubic mile. I have referred to this pheno- 

 mena already,! and have shown that the enormous force requisite 

 to crush such a vast mass of the most resistant rock would, had 

 there been a flux or had the material been of a less refractive 

 nature, resulted in the fusion of the mass. 



Mallet held much the same views. He says : " The result of the 



* Ann. Kep. Geol. Comm., 1899, Cape Town, 1900. 



f Hot Springs, Geol. Mag., 1904, p. 252; Ann. Kep. Geol, Comm., 1903, Cape 

 Town, 1904, p. 133. 



