334 TlMEHRI. 
The very curious configuration of the top of Roraima 
as described, and which is evidently due, as Mr. IM THURN 
points out, to extraordinarily active aerial denudation, is 
not difficult of explanation with regard to the variety of 
conformation. The effe£t of marked alternation, though 
possibly for variable periods, of the dry and rainy 
seasons, on such an exposed and unsheltered area, 
and on such a yielding substance as sandstone, 
presents circumstances sufficient for the explanation — 
though real confirmation could only be obtained by obser- 
vations on the spot extending over a considerable period 
of time. The effe6t of continued dry weather on the 
surface, with the resulting cracks or fissures, but opens 
out the road, so to speak, for the march of the effeft 
of the rainy season. The prevailing dire6lion of the 
winds and rains, on an unsheltered surface, gradually 
affefts most the surfaces opposed ; while other surfaces are 
modified, in the long run, by changes in the direction of 
these aerial forces. Abounding channels are formed in 
the sun-cracks, pillars are left standing and basins 
hollowed out ; and these, according to the shapes and 
direftions of these cracks or fissures, and the opposed 
forces of wind and rain at varying periods on the 
hardened upper surface and the softer revealed surface, 
work out the various marvellous shapes present — the 
most marked effefl being noticed where an unequal play 
results from the greater constancy in any one direftion 
of the denuding forces. Such a formation, resulting from 
such denuding forces on the top, fully bears out the 
comparison made with what, Mr. IM THURN tells us, are 
called Eppellings by the Indians ; but the conformation 
of the mountain group itself and its neighbouring parts, 
