192 CORRELATION OF PHYSICAL FORCES. 



lance (and in such sense i have used it), to a perpetual recur- 

 rent motion, e.g. a weight which by its fall would turn a 

 wheel, which wheel would, in its turn, raise the initial weight, 

 and so on forever, or until the material of which the machine 

 is made be worn out. It is strange that to common appre- 

 hension the impossibility of this is not self-evident : if the in- 

 itial weight is to be raised by the force it has itself generated, 

 it must necessarily generate a force greater than that of its 

 own weight or centripetal attraction ; in other words, it must 

 be capable of raising a weight heavier than itself: so that, 

 setting aside the resistance of friction, &c., a weight, to pro- 

 duce perpetual recurrent motion, must be heavier than an 

 equal weight of matter, in short, heavier than itself. 



Suppose two equal weights at each end of an equi-armed 

 lever, there is no motion ; cut off a fraction of one of them, 

 and it rises while the other falls. How, now, is the lesser 

 weight to bring back the greater without any extraneous ap- 

 plication of force ? If, as is obvious, it cannot do so in this 

 simple form of experiment, it is d fortiori more impossible if 

 machinery be added, for increased resistances have then to 

 be overcome. Can we again mend this by employing any 

 other force? Suppose we employ electricity, the initial 

 weight in descending turns a cylinder against a cushion, and 

 so generates electricity ; to make this force recurrent, the 

 electricity so generated must, in its turn, raise the initial 

 weight, or one heavier than it, i. e. the initial weight must, 

 through the medium of electricity, raise a weight heavier than 

 itself. The same problem, applied to any other forces, will 

 involve the same absurdity : and yet simple as the matter 

 seems, the world is hardly yet disabused of an idea little re- 

 moved from superstition. 



But the importance of the deductions to be derived from 

 the negation of perpetual motion seems scarcely to have im- 

 pressed philosophers, and we only find here and there a scat- 

 tered hint of the consequences necessarily resulting from that 



