ANTARCTIC EXPLORATION. 295 



from elevation or erosiou. They consist of debris piled round a vent. 

 Lava and ashes surround the crater in alternate layers. But in this 

 polar region the snow-fall must be taken into account as well as the 

 ash deposit and the lava tlow. It may be thought that any volcanic 

 ejecta would speedily melt the snow upon which they fell, but this does 

 not by any means necessarily follow. Volcanic ash, the most wide-spread 

 and most abundant material ejected, falls comi)aratively cold, cakes, and 

 then forms one of the most effective nonconductors known. When 

 such a layer a few inches thick is spread over snow even molten lava 

 may How over it without melting the snow beneath. This may seem 

 to be incredible, but it has been observed to occur. In 1828 Lyell saw 

 on the flanks of Etna a glacier sealed up under a crust of lava. Now, 

 the Antarctic is the region of thick-ribbed ice. All exposed surfaces 

 are quickly covered with snow. Snow-falls, tishfalls, and lava-flows 

 must have been heaping themselves up around the craters during 

 unknown ages. What has been the result ? Has the viscosity of the 

 ice been modified by the intercalation of beds of rigid lava and of 

 hard-set ash ? Does the growing mass tend to pile up or to settle 

 down and spread out ? Is the ice wasted by evaporation, or does the 

 ash layer preserve it against this mode of dissipation ' These inter- 

 esting questions can be studied round the South Pole, and perhaps 

 nowhere else so well. 



Another question of interest, as bearing upon the location of the 

 great Antarctic continent, whi(;h it is now certain existed in the Sec- 

 ondary period of geologists, is the nature of the rocks u])on which the 

 lowest of these lava beds rest. If they can be discovered, and if they 

 then be found to be sedimentary rocks — such as slates and sandstones, 

 or plutonic rocks — such as granite, they will at once aftbrd us some 

 data to go upon, for the surface exposure of granite signifies that the 

 locality has been part of a continental land sufficiently long for the 

 weathering and removal of the many thousands of feet of sedimentary 

 rocks which of necessity overlie crystalline rocks during their genesis ; 

 whilst the presence of sedimentary rocks implies the sometime prox- 

 imity of a continent from the surfaces of which alone these sediments, 

 as rain-wash, could have been derived. 



As ancient slate rocks have already been discovered in the ice-clad 

 South Georgias, and as the drag-nets of the Erebus and the Challenger 

 have brought up from the beds of these icy seas fragments of sand- 

 stones, slates, and granite, as well as the typical blue mud which 

 invariably fringes continental land, there is every reason to expect 

 that such strata will be found. 



Wherever the state of the snow will permit, the polar mountains 

 should be searched for basaltic dikes, in the hoi^e that masses of spec- 

 ular iron and nickel might be found, similar to those dis(;overed by 

 Nordeuskiold, at Ovilak, in north Greenland. The interest taken in 

 these metallic masses arises from the fact that thev alone, of all the 



