The Bocks of Tristan cVAcunha. 27 



islands constituting the Fiji group. These last rise from the 2,000 

 fathom-deep base. On the very edge of this same platform, and 

 facing the abysmal depths of the Pacific, rise the Tonga or Friendly 

 Archipelago, distant over 2,000 miles from the mainland of Australia. 

 One of the group, named Eua, has yielded among the volcanic 

 ejactamenta a boulder of uralitized gabbro, and in the tuffs, crystals 

 of red garnet and tourmaline. Lester very rightly points out that 

 these minerals indicate the near presence of crystalline schists. * 



This observation raises a vast subject of inquiry, namely, the 

 ability of volcanoes to bring specimens of the underlying rocks to 

 the surface in the form of small fragments of dust and mud, and 

 by the study of these to allow us to investigate the nature of the 

 rock underlying the vent, much in the same way as when on a 

 smaller scale, the geological surveyor utilises the burrows of rats 

 and rabbits in a district covered with soil to see what solid rocks lie 

 beneath. 



From the earliest times it has been known that volcanoes tear off 

 pieces of rock from the throat, deep down below the earth's surface, 

 and eject them along with the lavas and ashes. As a rule, however, 

 the general conception of volcanic activity is that of an outpouring 

 of molten material, or of the ejection of this material which has been 

 blown to powder by the explosive action of gases imprisoned in the 

 liquid magma. It was this idea that obscured much of the earlier 

 writings on volcanology, and still renders it difficult to get people to 

 dissociate this idea of the effusion of liquid lava from the popular 

 misconception of a liquid interior of the globe. The whole trend of 

 recent work in this field is to prove that the interior of the earth is 

 not liquid, and that a very large number of volcanoes eject solely 

 and exclusively solid material, that is, rocks that have been torn 

 from the throat of the volcano by the action of expansive gases, and 

 which have not had time, or have not been subjected to sufficient 

 heat, to get melted. 



Such volcanoes have been found in the Eifeland Swabia,in Germany, 

 and on the Firth of Forth in Scotland ; but it must be remembered 

 that even in the ordinary types of volcanoes in Italy that have been 

 studied from the earliest times, granite and gneiss blocks, unmelted 

 and unaltered, are occasionally brought to the surface in the 

 eruptions.} 



Unfortunately we have none of the cold volcanoes acting at the 

 present moment, unless we except the Javanese and Andean mud- 

 volcanoes. In South Africa the type has long been known to be repre- 



* J. J. Lester, Q.J.G.S., 1891, p. 600. 



f Branco, " Swabiens Vulcan Embryonen," Stuttgart, 1894, p. 760. 



