250 Mr. S. Bidwell on some 



There are few more beautiful experiments than to watch 

 these changes. Purity of colour, however, depends much on 

 the purity of the paper employed. Want of this purity will 

 often cut short the changes at the pale blue of the second 

 order. 



I have endeavoured to give some idea of these colours in 

 the Plates which accompany this paper, but it has proved to 

 be a most difficult task. The colours represented are — 



Plate L, purple and blue of the second order. 



Plate II., purple of the second order and green of the third 

 order. 



Plate III., brown-red of the first order. 



It has not, however, been found possible to correctly re- 

 produce the brilliancy and depth of colour of the originals. 



Explanation of the Plates. 



In each pair of figures the upper one represents the effect 

 of exposing allotropic silver to some form of energy. The 

 changes are in all cases the same ; the gold-coloured silver is 

 converted by all forms of energy to ordinary silver. 



In the lower of each pair of figures the effect is repre- 

 sented which would be produced by immersing the upper one 

 in a solution of potassium ferricyanide. This affords proof 

 of the completeness of the change by showing that gold- 

 coloured silver in passing into the ordinary form has lost its 

 power of reacting with a ferricyanide. In these lower figures 

 an attempt has been made to show some of the colours pro- 

 duced in this way. But they fall short of the originals in 

 brilliancy and intensity. These last are so remarkable that 

 the lithographer w T ho executed the work remarked that even 

 an artist with a brush and palette of colours could not imitate 

 them, and that, therefore, it was hopeless to expect to repro- 

 duce them by lithography. In other respects the Plates 

 represent fairly well the changes that take place. 



Philadelphia, Jam 24, 1891. 



XXVII. Some Experiments ivith Selenium Cells. 

 By Shelfokd Bidwell, M.A., F.R.S* 



IT is well known that selenium, like sulphur, which it 

 closely resembles in its chemical properties, is capable 

 of existing in several distinct forms, differing greatly in 

 appearance. The modifications most commonly met with are 



* Communicated by the Physical Society : read December 12, 1890. 



