The introduction of railways involved the making of a large number of tunnels, perhaps more so in the beginning, when it was thought that the use of the locomotive would be confined to very moderate gradients, and when engineers hesitated to adopt the steeper inclines and sharper curves which form the practice of modern times. Another element of consideration also consisted in the fact that the first railways were designed to connect the most populous and busiest districts, where the prospects of heavy traffic would appear to warrant a large outlay for works of construction. As the system spread and railways extended further away from the important centres, the 164 probabilities of traffic would become less promising, and efforts would be made to keep down cost of construction, and avoid tunnel work as much as possible. It is not easy to define where cutting should end and tunnelling begin. There is no practical difficulty in making a cutting 50, 60, or 70 feet deep, with slopes to suit the material excavated, and the estimated cost per yard forward may even compare favourably with the cost of average tunnel-work. But there are other questions which must be kept in view--the time required to form the cutting, the space to be obtained on which to deposit the enormous quantity of excavated material, and the probable difficulty in obtaining the large area of land necessary for the cutting. Before deciding the actual position of a tunnel, both as to line and level, it is necessary to obtain the most reliable information possible regarding the strata through which it has to pass. In addition to the geological indications on the surface and in the locality, borings should be made, and trial holes or shafts sunk along the proposed centre line of the work, and from these an approximately accurate longitudinal section can be laid down on paper, showing the respective layers of material to be cut through, and the angle at which they lie. With these particulars before him, the engineer may, in some cases, consider it more prudent to change the position of the tunnel in preference to incurring specially difficult or tedious work in dealing with some recognized unfavourable material. Occasionally the route may be slightly varied and better material obtained, but very frequently there is little to be gained except by a wide deviation from the original line. Solid rock, except for the slow progress, is perhaps the most favourable material for tunnelling, as the timbering, side walling, and arching can be almost, if not entirely, dispensed with. Loose rock, although more readily removed, necessitates strong timbering to prevent large masses breaking away and falling into the tunnel. Some clays are very compact and tenacious, and will stand well with moderate timbering, but even these should not be left long before following up with the side walls and arching.