i9«] ON VOLCANOES 317 
prehistoric examples Deccan (India) and Oregon 
(U.S.). 
There is small ground for supposing relation between 
adjacent volcanoes — activity in one is rarely accompanied 
by activity in the other. It seems most likely that vent 
tubes are entirely separate. 
Products of Volcanoes. — The lecturer mentioned the 
escape of quantities of free hydrogen — there was some 
discussion on this point afterwards ; that water is broken 
up is easily understood, but what becomes of the oxygen ? 
Simpson suggests the presence of much oxidizable material. 
C0 3 as a noxious gas also mentioned and discussed — 
causes mythical 1 upas ' tree — sulphurous fumes attend 
final stages. 
Practically little or no heat escapes through sides of 
a volcano. 
There was argument over physical conditions influ- 
encing explosions — especially as to barometric influence. 
There was a good deal of disjointed information on lavas, 
ropy or rapid flowing and viscous — also on spatter cones 
and caverns. 
In all cases lavas cool slowly — heat has been found 
close to the surface after 87 years. On Etna there is lava 
over ice. The lecturer finally reviewed the volcanicity of our 
own neighbourhood. He described various vents of Erebus, 
thinks Castle Rock a 'plug* — here some discussion — 
Observation Hill part of old volcano, nothing in common 
with Crater Hill. Inaccessible Island seems to have no 
connection with Erebus. 
Finally we had a few words on the origin of volcanicity 
