284 Prof. H. E. Armstrong. Electrolytic [Mar. 25, 



pronounced minimum is reached, corresponding to the pronounced 

 maximum attained on addition of a moderate amount of water to 

 sulphuric acid, and at a point moreover corresponding to a definite 

 compound, SnCu 4 ; a further slight addition of tin develops another 

 minimum less pronounced than the first, but also corresponding to a 

 definite compound, SnCu 3 ; the curve then falls slightly and exhibits 

 a third minimum, its course being analogous to that of the sulphuric 

 acid and water curve near to sulphuric acid. If one or the other 

 curve be inverted, the general similarity in form is especially striking. 

 It is obviously important that alloys intermediate between those 

 studied should be examined; a comparison of Lodge's with Matthies- 

 sen's curve shows how much may be missed ; this remark applies to 

 alloys generally. Whatever may be the explanation,* it appears to 

 me to be clear than in alloys as in composite electrolytes the con- 

 stituent members of the system influence each other, and thus 

 mutually contribute to the final result. The marked diminution in 

 the conductivity of copper produced by very small quantities of 

 oxygen, of phosphorus or of the metalloid arsenic is well known. 

 It would appear probable that this is in some way due to the occur- 

 rence of an electrolytic change, which at least in part is opposed in 

 direction to that taking place in the pure metal during conduction. f 



Valency — Chemical Change. 



Notwithstanding the fierce controversy which has been waged 

 between the advocates of the doctrine of' fixed valency and the 

 advocates of the doctrine of varying valency, our views on the 

 subject are still in an unfortunate degree unsatisfactory and 

 indefinite. Even those — and they probably form a large majority — 

 who regard valency as a variable, dependent both upon the nature of 

 the associated radicles and the conditions — especially as to tem- 

 perature — under which these are placed, often hesitate to attribute a 

 valency sufficiently high to account for every case of combination ; in 

 fact both parties agree in distinguishing " atomic " from " molecu- 

 lar " compounds, and differ only as to where the line shall be drawn. 



* It is very remarkable that not only do the heat conductivity and the induction 

 balance curves for the tin-copper alloys correspond (Chandler Eoberts, " Phys. Soc. 

 Proc," 3, p. 156) , but that the curves given by Thurston as representing the strength 

 of these alloys ("Materials of Engineering," Part III, p. 412) also exhibit a marked 

 similarity to the electrical conductivity curves. 



f The change produced in gold by a very small quantity of lead is most astonish- 

 ing : its conductivity is reduced almost to that of lead and it becomes as brittle as 

 glass. It is difficult to understand this change unless it be that opportunity is given 

 for the gold itself to assume a different molecular state, owing to continuity becoming 

 disturbed. The effect \ roduced appears to be strictly comparable with that observed 

 on lowering the temperature of silver iodide from above about 150°, and in the 

 passage of liquid water at 0° into ice. 



