proportional to them, and therefore determines the relative lengths (parallel with the axis) of the blade at the different points. These half lengths are represented by v z 3, u 14, t I 5, s z6, r I7 and q z8, and, on being transferred to Fig. z , produce the curved line a p bt d d mtt d mt fro1, which, in order to meet the hub, I change to d b t d d t d ft. I am not a mathematician, but I am perfectly aware that this last demonstration of mine is not mathematically correct, but I believe it to

more »

... be sufficiently so for all practical purposes. In conclusion, I frankly acknowledge that I have neglected to take into consideration several very important" matters, namely, the shape of the vessel, the thickness of the stern and rudderposts, and the diameter of the hub of the sternpost. At present I am unable to say in What way and to what extent they would tend to modify my system of constructing screw propellers. Neither did I take any notice of the friction caused by a screw in passing through the water, as I believe that by designing a screw propeller otherwise rationally, more will be gained than can possibly be lost by friction.