Jan. lo, 1878] 



NATURE 



203 



transfeience of this motion to visible masses is called 

 '* luork," so therefore the conversion of heat into work is 

 no moe than the transference of motion from small to 

 large portions of mattf r, z>., the transference of motion 

 between portions of matter oi dijfferent dimensions. The 

 tnechantcal equivalent 0/ /leai ititre^oxQ simply represents 

 the equivalence in energy between the vtotiotts of portions 

 of matter of dijfferent dimensions (molecules and visible 

 masses). To deny, therefoie, that the heat possessed by 

 matter at normal temperature could be converted into 

 work would be to assume that by a certain difference in 

 dimensions the conditions are such that motion can no 

 longer be transferred from the smaller portions of matter 

 to the larger. This would evidently, d, priori, be by no 

 means a necessary assumption ; indeed it would appear, 

 perhaps, rather strange that by no device could such a 

 thing be done. 



5. At the first sight one difficulty in the way of utilising 

 this motion that surrounds us on all sides is that the 

 larger scale portions of matter (visible masses) are im- 

 mersed among the smaller scale portions of matter 

 (molecules) which surround the vi> ble mas es on all 

 sides (as the molecules of the surrounding air, &c.), so 

 that a perfect equilibrium of motion exists on all side> ; 

 so that it becomes impossible to trans er the motion to 

 the larj^er scale mass in the one direction or in the other, 

 and we cannot lay hold of ea h moving molecule indi- 

 vidually, on account of its minute s ze, 



6. It is an observed fact (and demonstrated theoreti- 

 cally) that poitions of natter in motion among themselves 

 tend to acquire the same entrgy of motion (called " tem- 

 perature" in the case of molecules). In accordance with 

 well-known facts, whenever the energy of this system of 

 small moving portions of matter is greater in one part 

 than in another, i.e., whenever the equilibrium of energy is 

 upset, then we can transfer some of the energy to larger 

 scale masses (convert heat into work). Is there, however, 

 no other means of converting heat into work but through 

 inequality of energy f It was pointed out in the last 

 article that inequality of velocity (by the mechanism of 

 diffusion) will serve the same purpose. The portions of 

 matter (molecules) which by equal temperature possess 

 equal energy, possess, when their masses are unequal, 

 unequal velocities. This inequality of velocity can be 

 utilised for work as well as inequality of energy. 



7. Since size is only relative., or there is nothing abso- 

 lute in size, it will be quite legitimate to suppose molecules 

 magnified up to a larger scale so as to be visible, and we 

 do this as in dealing with the mechanism of a process, it 

 is almost impossible to visualise or conceive clearly the 

 results without this condition, and it is our object, on 

 account of its practical bearing, to exhibit the process 

 involved in a clear light. Suppose, therefore, the mole- 

 cules of two diverse gases (oxygen and hydrogen) to be 

 magnified up to visible dimensions, and as we are not 

 concerned with the shape or form of the molecules, we 

 may simply represent the molecules of the two gabcs by a 

 number of spheres, those representing hydrogen possessing 

 each one-sixteenth of the ma-s of those representing 

 oxygen, and also possessing a normal velocity four times 

 as great. This is known to be the fact in the case of the 

 two gases when at the same temperature. We will further 

 suppose the spheres inclosed in the two separate halves 

 of a cylinder with a piston between them. The spheres 

 may either be supposed perfectly elastic or their motion 

 kept up artificially in some way ; just as in the case of a 

 gas the motion of its molecules is kept up by the mole- 

 cular vibrations of the sides of the cylinder. 



8. The spheres of the two sets possess equal energies 

 of motion, the one set making up in mass lor what they 

 want in velocity. The coUiding sphetes in each compart- 

 ment will arrange themselves (according to a known 

 principle) so that ihe number of spheres in unit of volume 

 of each set is the same, ar.d therefore the pressure exerted 



by their impacts on opposite sides of the piston will pro- 

 duce perfect equilibrium, so that the piston remains 

 immovable. Now the question is, supposing that (as 

 in the case of molecules) we cannot lay hold of each 

 of these spheres separately, is there any means of 

 utilising ^he inequality of velo:ity for the performance of 

 work ? [This is the question we hwe to make in the case 

 of two gases at the same temperature, wnose m lecules 

 we cannot grasp, and which poss:;ss unequal velocities.] 

 If we could by any device get a number of the spheres 

 from one compartment into the other without changing 

 their velocities in the act, then the pressure would evi- 

 dently rise in one compartment, and we should thus 

 obtain a capacity for work without the performance of 

 work. It is evident that this could be done by making 

 several perforations in the piston, about the S'ze of the 

 spheres themselves, so that the spheres, in impinging 

 ai^amst the piston would sometimes happen to encounter 

 the void space of a hole, and thus p tss through with 

 unchanged velocity into the op.iosite compartmf;n\ If 

 the spheres of the two sets were moving with equal 

 velocities, it is evident that as many on an average wo lid 

 pass through one way as the other, and there would 

 there'ore be n > disturbance of the equilibrium of p'-essure, 

 and consequently no work to be derived. But from the 

 fact that the spheres a^e moving with «;/^^«^^/ velocitii s, 

 a different result will ensue It will be evident that the 

 number of spheres passing through the hole will b-; pro- 

 portional to the r umber ot times they strike again«t the 

 piston, for the ctances that a 'phere will encounter a 

 hole will be proportional to the number of iss impacts 

 asainst the piston, i.e. to the velocity of the sphere. So 

 the velocity of the spheres in one compartment beint; four 

 times that in the other, four times as many lighter spheres 

 pass through one way, as heavier spheres pass through the 

 other. The number of spheres in one compartment will 

 therefore rapidly augment, and thus the pressure against 

 the piston will rise, and the piston will be finally driven 

 towards the opposite end of the cylinder, and in this act 

 energy will be transferred from the spheres in the one 

 compartment to those in the other ; or part of the energy 

 could be transferred to an outside mechanism in a self, 

 acting manner if desired, by simply connecting the piston 

 to the mechanism. 



9. Now if precisely the same thing can be done in the 

 case of two gases, it is evident that here the energy being 

 heat, we have in the result attendant on the motion of the 

 pi ton, the transference o* heat from one portion of gas to 

 another at normal trmjerature, i e. the transference of 

 heat in a self-acting manner from a colder to a hotter 

 portion of matter ; and if desired, a conversion of a part 

 of the heat of the gas (at normal temperature) into work 

 by cooling it down below the temperature of the coldest 

 of surroundiuii objects. 



10. In the case of a gas it is clear that we cannot make 

 perforations analogous to the above sufficiently S'nall to 

 suit molecules, but to attack molecules we must have re- 

 course to molecular mechanism, or to attempt to handle 

 them like the spheres we must have recourse to mechanism 

 on a suitable scale. We have such a mechanism in a 

 porous diaphragm (such as of pipeclay or plumbago) 

 which represents a piston with molecular perlorations. 

 Such a diaphragm, if fitted as a piston into a cylinder 

 will exhibit, with the molecules of two separate gases 

 possessing different molecular velocities (such as mole- 

 cules of oxygen and hydrogen), precisely the same phe- 

 nomena as those exhibited, simply on a magnified scale in 

 the case of the spheres ; or the above description applies 

 word lor word. We have by the motion of the purous 

 piston the conversion of the heat-motion of the gas at 

 normal tetnperuiure into work, the transference of heat 

 au omatically from the colder portion of gas to the warmer. 

 The second law of thermodynamics only holds when the 

 molecules brought into contact happen to be of the same 



