348 J.D. Dana on the Volcanoes of the Moon. 
mountains, or have they been swept away by the changes of 
time? M. von Buch has described a circular area on the island 
of Palma, one of the Canaries,* six miles in diameter, which has 
been compared to a lunar crater, with some appearance of reason. 
It is in fact hardly twice the diameter of Kilauea, which it oth- 
erwise resembles. On Mauritius there is a similar area fifteen 
miles in diameter, surrounded by precipitous walls composed of 
the edges of strata dipping outward.:+ Either it is a volcanic 
mountain whose centre has fallen in, as suggested by M. Bailly, 
or it is the remains of a great pit-crater.. I merely state the fact 
without expressing an opinion. Other instances might be men- 
tioned, but this will suffice. At the present period, few active 
boiling pits remain, and Kilauea is’ the only one whose charac- 
ters have been well determined.’ The surface fires of the globe 
have so far eee in meee: that in nearly every existing vol- 
cano, cinder-ejectio the action at summit, and erup- 
tions of lavas in streams are confined to fissures nee the sides 
and flanks of the mountain. 
Bic ig.cnaled by the facts avinped,, to remark also. on-the 
origin of the mineral constitution of igneous rocks. 
It has been a difficult problem for solution, why volcanic re- 
gions should have.a centre of solid feldspathic rocks, unst 
and compact, while the exterior consisted mainly of. ‘olin 
lavas. .Scrope, Von Buch, and. other writers on voleanoes, have 
mentioned instances of this structure; and it seems to charac- 
terize generally the large voleanic mountains. It is-well exhib- 
ited when the elevations are cut through by gorges; and when 
not, the clinkstone appears often atthe summit of the cone oF 
dome. . The explanations we here venture, proceed on two 
principles : 
1. The motion which belongs toa boiling fluid. _ 
2. The less fusibility of hidenin than the other ingredients. 
Inthe great boiling pools, there will necessarily be a rising of 
the fluid, in the hotter part, and a flow away towards either side, 
producing a kind of circulation, This is no hypothesis, as the 
fact may be witnessed in any boiling cauldron; and the lavas of 
Haonen, are a visible example of it, The ebullition, in lavas 
_ Aa Saabs Canaries., noe = - 281. 
