PERPETUAL MOTION 59 



point D, pass into the right leg (containing air), and fall to 

 valve No. 5, strike and open the small valve in its center, 

 then open the large one, and pass through ; this valve will 

 then, by appropriate weights or springs, close ; the ball will 

 roll on through the bent tube (which is made in that form 

 to gain time as well as to exhibit motion) to the next valve 

 (No. 6), where it will perform the same operation, and 

 then, falling upon the four balls at E, force the bottom one 

 round the corner at C. This ball will proceed as did No. 

 8, and the rest in the same manner successively." 



That an ordinary amateur mechanic should be misled by 

 such arguments is perhaps not so surprising, when we re- 

 member that the famous John Bernoulli claimed to have 

 invented a perpetual motion based on the difference be- 

 tween the specific gravities of two liquids. A translation 

 of the original Latin may be found in the Encyclopaedia 

 Britannica, Vol. XVIII, page 555. Some of the premises 

 on which he depends are, however, impossibilities, and 

 Professor Chrystal concludes his notice of the invention 

 ' thus : " One really is at a loss with Bernoulli's wonderful 

 theory, whether to admire most the conscientious state- 

 ment of the hypothesis, the prim logic of the demonstra- 

 tion so carefully cut according to the pattern of the 

 ancients or the weighty superstructure built on so frail 

 a foundation. Most of our perpetual motions were clearly 

 the result of too little learning ; surely this one was the 

 product of too much." 



A more simple device was suggested recently by a cor- 

 respondent of "Power." He describes it thus : 



The J-shaped tube A, Fig. 14, is open at both ends, 

 but tapers at the lower end, as shown. A well-greased 

 cotton rope C passes over the wheel B and through the 



