We will now consider the methods of mounting specula of larger size, and will take as an instance the mounting of some of the largest specula in existence which must act so as to prevent flexure in any position of the speculum. The speculum is, in the case of the Melbourne telescope, of the weight of something like two tons. When it is inclined at any considerable angle to the horizon, it is apt to bend over at the top, and thus destroy its proper curvature; and when horizontal, if not equally supported, it will also bend, and unless some measures are taken to prevent this flexure it will so entirely alter its figure by its own weight as to render minute observations of any delicate stars absolutely impossible. Mr. Lassell was the first to suggest an arrangement for preventing this flexure. Through the back of the speculum case—the case which holds and supports the speculum, which we shall have to speak about presently—he inserts a large number of very small levers, the centres of which are fixed to the exterior part of this case, the forward part of each resting against a small aperture made in the back of the speculum. The ends of the levers furthest from the speculum are crowned with small weights, the weights varying on different parts of the speculum. Now so long as the speculum is perfectly horizontal, _i.e._ so long as the zenith is being observed, these levers will have no action whatever; but the moment the reflector is brought into any other position, as, for instance, when we wish to observe a star near the horizon, the more the mirror is inclined to the horizon the greater will be the power of these small levers, and at length their total effect comes into action when a star close to the horizon is being observed. Then the whole weight of the mirror is carried by these levers acting at points all over its back. In the Melbourne reflector, which has recently been finished, Mr. Grubb manages this somewhat differently, as will be seen by Figs. 73-76. In Fig. 73 the speculum is in a vertical position. It is supported in a frame, B B, all round it, which consists of a slightly flexible hoop of metal a little larger than the speculum. This in its turn is supported by a large fixed hoop, A A, having a hook-shaped section. This hoop is attached to the tube of the telescope C C. The hoop, B B, is rather larger than the part of A on which it hangs, so that it can adjust itself to the form of the mirror; and not only is the mirror supported in the hoop B B, like as in a strap in the position shown, but in every other position of the tube the speculum still hangs evenly supported. [Illustration: FIG. 73.—Support of the mirror when vertical. ]