The change of elasticity and density, between the rock of one formation and another, does not, so far as my observation has informed me, produce at all an equal extent of superficial modification, of the wave of shock, to that which is effected by these four causes.
Breach of continuity of medium, as I have long since shown, by the loss of vis vivâ that takes place at every surface of contact, is the condition, of all the most eflfective, in limiting the progress of the wave, and robbing it of its volume, up to extinction.
From the nature of the building up, of every mountain chain, of anticlinals, with strike more or less parallel to its axial line, it follows, that the mass of material, is most solid, the breaches of continuity, fewest and least, in the latter direction. The wave is therefore best transmitted "end on," by mountain chains; as well as by the formations reposing on their lowest flanks, and filling the valleys between parallel ranges.
Humboldt mentions, that the course of earthquakes in the north of the Asiatic continent and elsewhere, has been observed "to follow the courses of the great rivers," but he offers no explanation of the fact. It is not, however, in consequence of any connection with a river course, that the
