508 SCIENCE PROGRESS 



At absolute zero, d equals do, the exponent becomes 6, and t t = Gs"', where s is 

 the density of the substance at absolute zero. 



When dido is large, we get the ordinary Newtonian formula. 



For a pair of iron molecules, the tractation according to this formula falls from 

 about io _s dynes when in contact (compare value given above for hard steel) to 

 io -8 when there is 10 per cent, separation. 



The ratio of this tractation to simple Newtonian falls from 17 x io 50 for 

 contact conditions down to unity for a separation of a millimetre. 



As soon as sufficient investigation shall have shown that this is generally 

 consistent with the properties of matter (if it does so) the magnitude of the 

 pellation can be studied. It appears to require that the pellation shall follow 

 some such rule as 



/, =» Kd - « 



where n is a very high number (about 70 in some cases that have been examined). 

 K is necessarily very small. 



Assuming that the pellation is wholly a kinetic energy effect, some difficulties 

 still arise. Low-temperature experiments indicate great increase of cohesive 

 strength with decrease of temperature, and there seems no reason to suppose that 

 even at absolute zero compressive strength has ceased. We may, however, suppose 

 that the tractation increases, as would be indicated by the above formula, and that 

 resistance to compression at absolute zero arises from the kinetic energy of the 

 electron fields as distinguished from that of the atoms and molecules themselves. 

 A problem also occurs in connection with the storage of strain energy in com- 

 pression. Compressive strain reduces the tractation potential but increases the 

 pellation energy. A temporary increase is shown by a measurable rise of 

 temperature, but this is only a small part of the strain energy if the load is 

 applied slowly. The amplitudes of molecular oscillation having decreased, we 

 must either suppose that the molecular frequency remains higher (even after 

 radiation has restored the temperature balance) or that the electron field becomes 

 more energetic. 



If, as has been supposed, cohesion is a residual effect of electrostatic tractation, 

 it would be natural to expect that some more definite relation between cohesion 

 and electrical phenomena should be apparent. The writer suggests that such 

 relations do exist, and while he does not press this point too strongly, seeing that 

 it is frequently difficult to know whether there are or are not free ions or electrons 

 participating, he cannot but feel there is a considerable amount of evidence in 

 favour of such an hypothesis. 



In the first place, the production of static charge by friction indicates that the 

 disturbance or rupture of cohesion bonds liberates electrons. Secondly there are 

 several direct relations between electricity and cohesion. According to Silvanus 

 Thompson : 



" Metal conductors, when subjected to the prolonged action of currents, 

 undergo slow molecular changes. Wires of copper and brass gradually become 

 brittle under its influence. During the passage of the current through metallic 

 wires their cohesion is temporarily lessened, and there also appears to be a 

 decrease in their coefficient of elasticity."— Electricity and Magnetism, 249. 



Again it has been observed by Beccaria that an electrified liquid evaporates 

 more quickly than one not electrified, and the capillary electrometer shows how 

 considerably surface tension is modified by difference of potential. 



