I am a stranger in the field of speculation, and am quite unacquainted with the intricacies of its authorized boundaries. It is therefore with some hesitation, lest I should tread upon forbidden ground, that I venture to offer a suggestion on one point in Professor Sollas's paper on “The Figure of the Earth.”

It has long been observed that mountain ranges and chains of islands (which, indeed, are only mountain ranges partially submerged) are generally curvilinear in form, but Professor Sollas is, I believe, the first to show clearly that the curve often coincides almost exactly with an arc of a circle. Such a mountain chain is frequently defined along its convex margin by a great reversed fault over which the mountain mass has slid forward; and in these cases, at least, we may safely adopt Suess's conception, and look upon the chain as the crumpled edge of a ‘scale’ of the earth's crust which has been pushed forward over the part in front of it. The surface along which the movement has taken place is called a thrustplane. If this surface really is a plane, then the edge of the ‘scale’, that is the mountain chain itself, must necessarily be circular in form; for if any plane cuts a sphere, in any position whatever, the outcrop of the plane on the surface of the sphere will always be a circle. There can be no deviation from the circular form unless the ‘sphere’ is not truly spherical, or the ‘thrust-plane’ is not a true plane.