Mr. G. J. Stoney on Polarization Stress in Gases. 411 



within the slice may evidently be conceived of as arising from 

 the coexistence of two streams travelling in opposite directions 

 along the tube, and each consisting of gas which is less pola- 

 rized (i. e. which deviates less from the condition of ordinary 

 gas) than the gas that results from their coexistence. Each 

 stream is exposed within the slice to the mutual jostling of its 

 own molecules ; and it is also attacked by molecules of the 

 other stream. The mutual jostling of its own molecules tends, 

 as explained in section 10, both to maintain the onward velo- 

 city of the stream and to reduce the gas of w T hich the stream 

 consists still more towards the condition of unpolarized gas. 

 These encounters then, taken by themselves, tend to bring 

 about the state of the gas described in section 10. But the 

 interference of the two streams with one another counteracts 

 this. This interference modifies the effect of the encounters 

 within the streams, but it is incompetent to annul it; for 

 the two streams do not by their mere coexistence constitute 

 stationary unpolarized gas, and hence they would need time 

 before they could by their action upon one another reduce the 

 gas to this condition. It is, however, plain that whatever 

 action they exert is a step towards bringing about this condi- 

 tion; for the gas would become depolarized if the cross sec- 

 tions which bound the slice could be rendered impervious both 

 to energy and molecules, so as to leave the two streams time 

 to act fully on one another. In reality, however, sufficient 

 time is not allowed to them, because the streams pass one ano- 

 ther, and the struggle is continually renewed within the slice 

 by fresh portions of the streams which come up in the same 

 state as those that had been obliged to pass on. These fresh 

 portions keep in the same state because a sufficient supply of 

 swift molecules is without intermission being thrown back 

 along the tube from one end by the heater, and a correspond- 

 ing supply of slow molecules from the other end by the 

 cooler. 



13. The two streams, however, though not annulled, are 

 different from what they would have been if they had been 

 without influence upon one another. They do not consist of 

 the same molecules from one instant to another ; for there is 

 such a perpetual shifting of molecules between them, owino- 

 to the vast number of encounters that take place, that no one 

 molecule is likely to remain long in one stream. Again, after 

 an encounter between molecules of the two streams, both of 

 the colliding molecules will sometimes join the same stream ; 

 and it will most frequently happen that the stream so joined 

 is the hotter and swifter stream. Hence the stream from the 

 heater to the cooler receives an accession to the number of its 



