294 Mr. Norman Campbell on Delia Rays. 



the curves with different metals are compared. I£ different 

 metals have different reflecting powers, the various forms of 

 curve might well be due to these differences, and not to dif- 

 ferences in the speed of the rays. Now v. Baeyer has found 

 it impossible to prove any clear difference between the re- 

 flecting powers of different polished metals, except possibly 

 in the case of aluminium, which appears to reflect somewhat 

 less than other metals. His results do not lead to the anti- 

 cipation of such a difference as has been found between 

 silver and gold, and the difference found between aluminium 

 and gold is in the wrong direction. For it has appeared 

 that reflexion should on the whole make the curve steeper : 

 the curve for gold should on account of reflexion be steeper 

 than that of aluminium. As a matter of fact it is flatter. 



15. It appeared that useful information might be obtained 

 on this important point if the curve were examined when 

 one of the electrodes A and B was covered with one metal 

 and the other with another. For if the curves previously 

 obtained with different metals differed owing to a difference 

 in the rays emitted from them, then if an electrode of copper 

 were opposed to an electrode of aluminium, the part of the 

 curve belonging to the aluminium electrode ought to show 

 the quality of the aluminium rays and vice versa. But, if the 

 curves differ owing to a difference in the reflecting power, 

 then, since the aluminium rays are reflected at the copper, the 

 part of the curve corresponding to the aluminium electrode 

 ought to show the properties of the copper rays. The 

 experiment is easily tried, but the results are rather difficult 

 to interpret. 



In order to explain the nature of the difficulty two of 

 the curves obtained are plotted in figs. 4 and 5 ; they should 

 be compared with the aluminium curve of fig. I. It will be 

 observed that in fig. 1 the change of curvature of the curve 

 takes place at, or very near, the axis of zero potential dif- 

 ference: on the other hand, in figs. 4 and 5 the change of curva- 

 ture is certainly not on that axis. In fig. 4 (where + refers 

 to an aluminium electrode and — to a copper electrode) the 

 change is certainly on the negative side of that axis; in fig. 5 

 (where -f- refers to a silver electrode and — to an aluminium 

 electrode) the change is less obviously, but certainly, on the 

 positive side. Similar results were observed for the combi- 

 nation gold-graphite, but the results are not included, for no 

 success has been attained so far with graphite- or soot- 

 covered electrodes ; the difficulties are doubtless connected 

 with an adhering film of air or moisture, which will be noted 

 below. 



