10 THE TRUE VALUE OE a OF VAN DER WAALS' 



gases are actually octovalent ! That valence electrons may be attached 

 to the atoms of these gases is, I believe, a fact and not a theory. 

 How are we to explain the conduction of electricity in neon and 

 argon and the existence of positive helium and krypton atoms, 

 unless their atoms have the power of picking up and losing elec- 

 trons, without decomposing? It is the characteristic of the valence 

 electrons, in contrast with the electrons in the nucleus of the atom, 

 that they are reversibly displaceable without disrupting the atoms. 

 Neon and argon conduct the current un usually well and Sir J. J. 

 Thomson has shown that helium may lose one or two electrons 

 and occur as singly or doubly charged atoms and krypton may 

 lose at least four. The fact that these atoms have little combining 

 power is no proof that they have no valence. Chemists have not 

 given up hunting for compounds of these- gases, and 1 cited the 

 fact that the solubility of xenon in water is so great as to indicate 

 some kind of union between the water and the gas. That the 

 n limber of valences found from the cohesion should be fractional 

 is easily explicable if the number of valence electrons attached to 

 the atoms is not fixed. Some atoms may have none. Perhaps in 

 the case of helium only a few atoms are charged and as a result 

 the cohesion is so reduced that the computation gives a fractional 

 number. Flow otherwise shall we explain the fact that while in 

 hydrogen, oxygen, nitrogen, and carbon dioxide there is a steady 

 advance in a with the molecular weight, yet in helium, which is 

 twice as heavy as a hydrogen molecule, cohesion is so vastly less that 

 this gas is the most difficult to liquify of all that are known? Or 

 consider on the other hand metallic mercury. The cohesion and 

 affinity are so low that mercury exists in the vapor as separate 

 atoms. How naturally this is explained by the law of cohesion I 

 have found, where the cohesion is shown to depend upon the 

 ratio of the number of valences to the molecular weight ! In mercury 

 we have one of the heaviest of the elements with but a single valence 

 so that the ratio of Val I iff is very low and the cohesion is also low. 

 With this brief reply to some of the points raised by van Laar, 

 we may proceed to discuss the real point at issue, and to show 

 what the real value of a is. This will show the real basis 'of 

 fact of the law I have found. 



* 



I. THE REAL VALUE OF a. 



The first step to be taken in the solution of the problem of the 

 nature of cohesion is the determination of the real value of a 



