5 o6 SCIENCE PROGRESS 



Hence it appears that cohesive tractation enormously exceeds gravitational 

 force but does not quite reach electrostatic (chemical) values. 



Before proceeding further it must be clearly pointed out that in the solid state 

 the cohesive force which provides tensile strength to the material is only the excess 

 of tractation over pellation due to the enlargement of the molecular field. Similarly 

 compressive resistance is the excess of pellation over tractation when the molecular 

 field is contracted. In the condition of no stress, tractation balances pellation, 

 and at first sight it might seem that the two forces might have any value whatever 

 so long as their differences satisfied the space conditions. Experiments on tensile 

 stress show, however, that it reaches a maximum when the separation of the 

 molecules is from 01 to 05 times the original interval, and analogy suggests that 

 in this condition of ultimate stress the pellation is almost zero, so that in the 

 position of equilibrium it is improbable that the gross tractation (which is there 

 equal to the pellation) can be very much greater than the maximum tensile stress 

 in the position of rupture. In a paper on the " Cohesion of Solids" contributed to 

 the Physical Society in 191 5, this point is discussed by the present writer, and it is 

 shown {Proceedings, vol. xxvii. Part V., p. 450) that a stress-strain diagram is 

 the algebraic sum of tractation-strain and pellation-strain diagrams, and that the 

 space rate of decrease of the pellation exceeds that of the tractation. The fact 

 that repulsion is associated in some way with contact (probably of enormously 

 energetic fields of corpuscles) certainly suggests that its value will decrease very 

 rapidly as the propinquity decreases. 



In the case of the liquid state appreciable tensile stress only occurs when the 

 liquid is free of exceptionally active molecules (i.e gas), and for general purposes 

 it may be assumed that the liquid condition chiefly depends on a very slight 

 excess of pellation over attraction which is balanced at the surface by gas pressure. 

 At the surface the lack of upward tractation (due to the wider spacing of the gas 

 molecules) enables the downward tractation to there exceed the pellation, so 

 producing surface tension effects. 



In gases the tractation is only temporarily potent between colliding molecules, 

 and is almost wholly neutralised by the impulsive forces arising from the reaction 

 momentum of collision. 



Reverting to the question of the solid state, the very minute range within 

 which tractation so exceeds pellation that the net cohesion increases with 

 separation (a condition necessary for the stability of the solid) is shown by the 

 fact that very great pressure is required to make solid substances cohere, and 

 also by the fact that rupture occurs in all true solids with a linear expansion of 

 less than 50 per cent., indicating that the maximum stress occurs at or below that 

 degree of separation. 



The case of liquefaction of solids is a little different, since it seems clear that 

 the pellation is chiefly if not wholly due to the kinetic energy of the molecules, 

 which, being increased by temperature, rises at the melting temperature to a value 

 at which it exceeds the tractation. This may be shown graphically by raising 

 the pellation curve vertically so that it ceases to give intercepts on the tractation 

 curve, and gives a stress diagram entirely corresponding to expansion which can 

 only be neutralised by an externally applied pressure. 



The space rate of change of stress in solids (which is termed the "modulus of 

 slasticity " so long as it remains sensibly constant) is the difference between the 

 space rates of change of tractation and pellation and becomes zero at the point of 

 maximum tensile stress, and is indefinitely large for pure compression (i.e. without 

 shear or lateral bursting tension). 



