ROTARY ENGINES. 343
been practically used for both of them. At the same time there is perhaps no department of practical work about which more indefinite and certain ideas have been held than this. The method of dealing with it has been little better than a groping in the dark, without the guidance of any principle, without knowledge of its course, without understanding of its materials. Such a multitude of arrangements have been devised and are still being devised for the carrying out of one and the same purpose, that it appears almost impossible for a single individual to reduce them all to order, or even to find out their existence. The irresistible tendency towards the invention of "rotary" steam-engines has contributed greatly to increase the number of these arrangements. This tendency has given us many useless, or apparently useless, machines, and has been the means of wasting much thought and capital. Would-be inventors have again and again been warned of it, but the warnings do not seem to have had any effect. From a kinematic stand-point the warning cannot be unconditionally repeated. In the first place the class of inventors here concerned are exactly those whom our warning will never reach, or who will not listen to it. Then again the attempted combinations may not in themselves be bad, although their practical usefulness may not correspond to the hopes of their inventors. And lastly, it does not seem justifiable to check empirical experiments so long as theoretical investigations into the matter fail to furnish the means of doing better or as well, or even to say with certainty what is the real worth or worthlessness of the results obtained.
We have now, however, got a most important auxiliary in the matter, for in kinematic analysis we have the means of deter- mining the real nature even of the most disguised mechanisms, and we shall proceed at once to the consideration of this problem so far as it concerns the machines just mentioned, in which crank trains are formed into pressure-organ machines.
The process of designing such a machine divides itself into two parts. There is, first, (a) the making of one of the links into a vessel or chamber, and (6) the forming of another (or two others) into a movable diaphragm or piston, the relative motions of these parts being so arranged that the pressure-organ alternately fills and is driven out from the space between them. The second part of the process consists in the addition of kinematic arrangements
