Two gear-wheels cannot run properly together unless they are proportional in size to the number of teeth. Now, the stretching of a sprocket-chain alters the pitch in a manner similar to that of retaining the same number of teeth in each of two intermeshing wheels, and then altering the size of one. A sprocket-chain acts substantially as an idle wheel; when it stretches we have, as it were, this idle wheel made larger while the size of the others and the number of teeth in each remain the same. Increasing the distance between the centres does not affect the size of the wheels, and when a sprocket-chain stretches or becomes longer by wear the wheels should either be larger or else the number of teeth diminished. It is a general idea among mechanics that chain gearing is about the most undesirable of all means of transmitting power we have. This is perhaps an exaggeration, and I think the cycle art has proved it to be so; but the idea no doubt is fostered by this constant tendency of the chain to stretch, and when this stretch takes place a very considerable amount of friction must result. There is another annoyance felt by patrons of the small wheel: the chains being low down and well oiled, as they should be, especially if once they become stretched, have a superlative capacity for accumulating and holding dirt, causing a grinding second only to that of a finely-set quartz-crusher. This feature is not so much to be deplored if the dirt can be kept out of the chain-link bearings, since it is not the wear of the link against the tooth of the wheel, but that within the link, which makes it longer, alters the pitch, and causes great friction. We shall, however, have to accept this chain arrangement for the present in Safeties, as it cannot be helped. Some ingenious inventor will no doubt ere long come to our assistance; but until then we can tolerate it with a good grace, since it is a necessary concomitant of so valuable an acquisition to our assortment of mounts. There is apparently little difference in the construction of the crank Rover Safeties, yet there is more than a cursory glance would lead us to suspect. To begin with, there is quite a variation in the slant of the neck or front fork, many makers giving a considerable curve to the fork, thus throwing the neck much straighter up. Then we have the telescope head, where the front fork revolves inside the tubular front extension of the main frame; and lastly, the swing-joint or Stanley head.