Travees. — On the Causes of Volcanic Eruptions. 835 



the validity of such an assumption, if we are to admit that the 

 increase in temperature found to obtain as we penetrate the 

 crust below the stratum of invariable temperature continues 

 beyond the depths to which our observations have extended, for 

 it is clear that, in such case, the temperature reached at the 

 depth of 25 miles would be sufficient to fuse nearly all the rock- 

 material with which we are acquainted, whilst at the depth of 

 150 miles all such material would be reduced to a state of 

 vaporiform incandescence. 



Now it is very singular that, notwithstanding the admitted 

 connection of this internal heat with the phenomena of earth- 

 quakes, volcanic disturbances, upheavals and subsidences which 

 affect the outer crust of the earth, some very important inves- 

 tigations made by Messrs. Nasmyth and Carpenter, in connection 

 with their long-continued and exhaustive examination of the 

 surface conditions of the moon, appear to have been entirely 

 overlooked by geological writers, although Nasmyth and 

 Carpenter distinctly pointed out that the results of their inves- 

 tigations would most probably be found to have an important 

 bearing on the origin of the phenomena referred to, and tend 

 to show that the thickness of the solidified, and especially of the 

 rigid, portion of the crust, must be very much less than that 

 which has been generally assigned to it. 



1 will now proceed to give some idea of the nature of those 

 investigations, and of their suggested bearing upon the matters 

 referred to. 



Messrs. Nasmyth and Carpenter were induced, as one of 

 the results of their long and careful observations of the sur- 

 face of the moon, to inquire into the relative densities of 

 fusible matters in the fluid and solid conditions. They found 

 that, with few exceptions, — exceptions having no influence 

 upon the questions at issue, — all fusible substances solid at 

 ordinary temperatures are densest when molten. They found 

 that solid gold, silver, iron, copper, and other metals floated 

 upon the same substances in the molten state ; that solid 

 slag floated on melted slag, and so forth ; thus account- 

 ing, in part at all events, for the greater density of the deeper 

 portions of the globe's mass, assuming those portions to be 

 still in a fluid condition from heat. They pointed out, what is 

 indeed a corollary to the first proposition, that molten material, 

 solid at ordinary temperatures, expands to and attains its 

 minimum density in the act of solidifying, and that this expan- 

 sion is followed by contraction as the solidified matter after- 

 wards parts with its heat by radiation. Thus, if the tire of a 

 wheel has to be formed as a casting, the fused metal must in 

 the first place be poured into a suitable mould, in which 

 provision has been made for the expansion of the solidifying 

 matter. After it has cooled sufficiently to become rigid, and to 



