But the fact is, that this motion would not at all account for the blue veins, for they are not parallel to the sides, but _oblique_ to them. This difficulty revealed itself after a time to those who first propounded the theory of differential motion, and caused them to modify their explanation of the structure. Differential motion is still assumed to be the cause of the veins, but now a motion is meant oblique to the sides, and it is supposed to be obtained in the following way:--Through the quicker motion of the point _c'_ the ice between it and _n_ becomes distended; that is to say, the line _c' n_ is in a state of strain--there is a _drag_, it is said, oblique to the sides of the glacier; and it is therefore in this direction that the particles will be caused to slide past each other. Dr. Whewell, who advocates this view, thus expounds it. He supposes the case of an alpine valley filled with india-rubber which has been warmed until it has partially melted, or become viscous, and then asks, "What will now be the condition of the mass? The sides and bottom will still be held back by the friction; the middle and upper part will slide forwards, but not freely. This want of freedom in the motion (arising from the viscosity) will produce a drag towards the middle of the valley, where the motion is freest; hence the direction in which the filaments slide past each other will be obliquely directed towards the middle. The sliding will separate the mass according to such lines; and though new attachments will take place, the mass may be expected to retain the results of this separation in the traces of parallel fissures."[A] Nothing can be clearer than the image of the process thus placed before the mind's eye. One fact of especial importance is to be borne in mind: the sliding of filaments which is thus supposed to take place oblique to the glacier has never been proved; it is wholly assumed. A moraine, it is admitted, will run parallel to the side of a glacier, or a block will move in the same direction from beginning to end, without being sensibly drawn towards the centre, but still it is supposed that the sliding of parts exists, though of a character so small as to render it insensible to measurement. [Sidenote: STRUCTURE CROSSES LINES OF SLIDING.] My chief difficulty as regards this theory may be expressed in a very few words. If the structure be produced by differential motion, why is the large and _real_ differential motion which experiments have established incompetent to produce it? And how can the veins run, as they are admitted to do, _across the lines of maximum sliding_ from their origin throughout the glacier to its end?