418 Mr. 8. Lupton on the Reduction of 



has never been done in a single case with salts. There conse- 

 quently exist no uncharged atoms in a salt solution. This 

 purely empirical proof can evidently not be extended to the 

 case of I 2 : but there is a great difference between the two 

 instances. It is easy to imagine that the two ions of a salt 

 (Na and CI from NaCl) , which are in other respects so differ- 

 ent from each other, should also be the bearers of different 

 kinds of electricity ; but it is difficult to admit that atoms 

 otherwise completely identical (I and I from I 2 ) should 

 assume, or possess, opposite electrical charges ; for complete 

 symmetry prevails in the molecule. Prof. Thomson's experi- 

 ments on the feeble conductivity of iodine vapour compared 

 with mine on caesium chloride (CsCl), yield, besides, sufficient 

 proof that the iodine atoms are not charged in measurable 

 degree according to Faraday's Law. 



The second question drops of itself when the first has re- 

 ceived the above answer. It runs: " If we admit the existence 

 of uncharged atoms (in solution), why should there be any 

 connexion between electric conductivity and osmotic pres- 

 sure?" Since our experience shows us that there are no 

 uncharged atoms in salt solutions, the connexion Prof. 

 Thomson mentions must obtain. With regard to Cl 2 , Br 2 , 

 and I 2 , we have no such experience as for salt solutions ; and 

 Prof. Thomson's assertion that the conductivity of these gases 

 stands in connexion with their dissociation at high tempera- 

 tures, i. e. with their gaseous pressure, is as yet entirely with- 

 out foundation, either theoretical or experimental. 



I share with Prof. Thomson the opinion that, after his two 

 questions are disposed of, my own query is so likewise, and 

 trust that the facts brought forward in the preceding pages 

 will supply sufficient material to any one for its unequivocal 

 answer. 



Upsala, 23rd February, 1891. 



L. On the Reduction of the Results of Experiments with special 

 reference to the Hydrate-theory of Solution. By Sydney 

 Lupton, M.A.* 



FT1HE problem of finding a formula to express quantitative 

 A experimental results is, partly owing to unavoidable 

 errors of observation, by no means an easy one, and has 

 severely taxed the powers of mathematicians. Experiment- 

 alists are very apt to attribute impossible accuracy to results 

 on which they have expended much care and time, and in the 



* Coniniuuicated by the Author. 



