This leads me to the point at which what I consider to be the true cause of the structure may be referred to. The theoretic researches of Mr. Hopkins have taught us a good deal regarding the pressures and tensions consequent upon glacier-motion. Aided by this knowledge, and also by a mode of experiment first introduced by Professor Forbes, I will now endeavour to explain the significance of the fact referred to in the last paragraph. If a plastic substance, such as mud, flow down a sloping canal, the lateral portions, being held back by friction, will be outstripped by the central ones. When the flow is so regulated that the velocity of a point at the centre shall not vary throughout the entire length of the canal, a coloured circle stamped upon the centre of the mud stream, near its origin, will move along with the mud, and still retain its circular form; for, inasmuch as the velocity of all points along the centre is the same, there can be no elongation of the circle longitudinally or transversely by either strain or pressure. A similar absence of longitudinal pressure may exist in a glacier, and, where it exists throughout, no central structure can, in my opinion, be developed. But let a circle be stamped upon the mud-stream near its side, then, when the mud flows, this circle will be distorted to an oval, with its major axis oblique to the direction of motion; the cause of this is that the portion of the circle farthest from the side of the canal moves more freely than that adjacent to the side. The mechanical effect of the slower lateral motion is to squeeze the circle in one direction, and draw it out in the perpendicular one. [Sidenote: MARGINAL STRUCTURE.] [Illustration: Fig. 40. Figure explanatory of the Marginal Structure.]