The Valence of Chlorine 259 



of determining the valence from the cohesion, as is shown in 

 columns 5 and 6 of Table I. With three possible exceptions, 

 chlorine is seen to be everywhere trivalent. The possible ex- 

 ceptions are hydrochloric acid, propyl chloride and one of 

 the chlorine atoms in germanium tetrachloride. In the first, 

 association renders the valence of the chlorine somewhat 

 doubtful; in other words, here also chlorine may be trivalent. 

 In the second, propyl chloride, the critical data may be wrong. 

 They were determined in 1886 by Vincent and Chappuis, 

 and the other determinations by these authors give generally 

 a value for "a" slightly lower than is to be expected. It is 

 not impossible, therefore, that a redetermination of the crit- 

 ical data for this substance will make M 2 K sufficiently high 

 to make the chlorine trivalent. For the third substance, 

 germanium tetrachloride, I can find but one determination 

 of the critical data in 1887. It is not unlikely, therefore, that 

 the data will need some revision. This method of deter- 

 mining the valence of chlorine confirms, therefore, the con- 

 clusions of Pascal, based on the study of the dispersion and 

 the magnetic properties, that chlorine is polyvalent, and, 

 further, this method shows it to be beyond doubt generally 

 trivalent. 



Moreover, since nearly all these compounds are normal 

 and not associating, the valences of the chlorine are shown to 

 be not resting, or in reserve, and not dissociated active val- 

 ences, but actually extending between the atoms of the 

 molecule. I have indicated in column 8 of Table I, some 

 possible structural formulae showing how these valences may 

 be extending in the molecule. Where there are two or more 

 chlorine atoms in the molecule, no serious reconstruction of 

 the graphic formulae is required, since the extra valences may 

 be pictured as reaching between the chlorine atoms; but where 

 there is an odd number of chlorine atoms in the molecule, 

 or where there is but a single one, then a fundamental change 

 must occur in the graphic formula. For example, in ethyl 

 and methyl chlorides, the formula must be written as I have 

 indicated, with the chlorine joining the carbon by two bonds 



