An interesting form of drier has its own motor, a little steam engine, attached to the frame of the machine. See Fig 24. This of course demands fixed bearings. The engine is very small. One size used is 3"x4". When a higher velocity of basket is required, we have the arrangement in Fig. 26. [Illustration: Fig. 26.] MOTORS.--This naturally introduces the subject of motive power. We may have the engine direct acting as above, or the power may be brought on by belting. Fig. 27 shows a drier with pulley for belting. Fig. 28 (W.H. Tolhurst) shows a very common arrangement of belting and also the fast and loose pulleys. When the heaviest part of the engine is so far from the vertical shaft as to overhang the casing on one side, there is apt to be an objectionable tremor. To remedy this, it is suggested to put these heavy parts as near the shaft as possible. It has been suggested also to use the Westinghouse type of engine, although the type shown in Fig. 24 works faultlessly in practice. [Illustration: Fig. 27.] One plan (Pat. 346,030), designed to combine the advantages of a direct acting motor and an oscillating shaft, mounts the whole machine, motor and all, on a rocking frame. The spindle is of course in fixed bearings in the frame. However, the plan is not practical. [Illustration: Fig. 28.] In driers the direct acting engine has many advantages over the belt. The atmosphere is always very moist about a whizzer, and there are frequently injurious fumes. The belt will be alternately dry and wet, stretched and limp, and wears out rapidly and is liable to sever. In all machines in which the shaft oscillates, if the center of oscillation does not lie in the central plane of the belt, the tension of the latter is not uniform. This affects badly both the belt and the running. A reference to the various figures will show the best position for the pulley. The greatest difficulty experienced with belting is in getting up speed and stopping. The basket must not be started with a sudden impulse. Its inertia will resist and something must give way. A gradual starting can be obtained by the slipping of the belt at first, but this is expensive. The best plan is to conduct the power through a species of friction clutch--an iron disk between two wooden ones. This has been found to work admirably. BRAKES.--The first centrifugals had no brakes. They ran until the friction of the bearings was sufficient to stop them. This occasioned, however, rapid wearing and too great a loss of time. The best material for a brake consists of soft wood into which shoe pegs have been driven, and which is thoroughly saturated with oil. The wooden disks referred to just above are of the same construction. The center of oscillation ought to be in the central plane of the brake as well as that of the pulley, but the preference is given to the pulley.