10 Transactions of the South African Philosophical Society. 



African volcanoes — The connection of volcanoes with structural lines— The 

 fusion of sedimentary rocks and the older crystalline series to form lavas — 

 The nature of the earth's centre— The nature of the earth's original crust — 

 Professor Vogt's researches in acid and basic types of rock representing rocks 

 of the original crust — The magmatic separation of rocks into acid and basic 

 series not due to gravitation — Conclusion — Petrography of rock specimens 

 from Tristan d'Acunha, Inaccessible Island, and Nightingale Island. 



Eecently, when the annual vessel went down with mails for 

 Tristan d'Acunha, I applied, through the courtesy of Sir David Gill, 

 to the Admiralty at Simon's Town, to be allowed a passage on 

 H.M.S. Odin. I was, unfortunately, unable to obtain the 

 required permission, but I received a certain number of rock speci- 

 mens which were collected by Commander H. L. D. Pearce, of 

 H.M.S. Odin, Mr. Hammond Tooke, and Mr. Bonhomie, of the 

 South African Museum. 



When I was agitating about the matter, I was frequently asked, 

 " What is the good of going there ? The islands are purely volcanic, 

 and there was not likely to be anything of geological interest there, 

 except what has been recorded again and again in such islands." I 

 could only reply I was unable to say what I could find till I 

 had actually been there, and the specimens brought back show how 

 much there is to be done there. I will endeavour in the present 

 paper, in the light of the evidence from the specimens brought 

 back, to show in what direction research must be carried on. 



The specimens come mostly from Tristan d'Acunha, but there 

 are some from Nightingale Island and from Inaccessible Island. 



The Tristan group lies 1,550 miles from the Cape of Good Hope, 

 2,000 miles from Cape Horn, and 1,320 miles from St. Helena. 

 The history of the group is given in the first volume of the " Narra- 

 tive of the Voyage of the Challenger." The officers of that ship also 

 collected specimens from Tristan d'Acunha and the other two main 

 islands, and the following description of the localities from which 

 theirs were obtained will hold good for the ones which are now 

 deposited in the South African Museum. The Challenger specimens 

 from Tristan consisted of large grained felspathic basalts, some- 

 times bordered with layers of black basaltic glass passing to pala- 

 gonite, basaltic tuffs, augite andesite, pyroxenite, and amphibolitic 

 andesite containing sanidine. Streams, or rather cascades, which 

 come dashing down to the sea during the constant heavy rains, have 

 eaten their way into the cliffs, and the beds form conspicuous 

 features in the view as narrow gullies descending the rocks in a 

 series of irregular steps. At the foot of the cliffs, immediately 

 opposite the anchorage, there are debris slopes and irregular rocky 

 and sandy ground, forming a narrow strip of low shore land. The 



