BISULPHIDE OF CARBON. ' 217 



I then took the light- yellow glass instead of the dark. 



Obs. 1. The ants were all under the violet glass and 



sulphate of quinine. 

 „ 2. Eather more than half under the yellow glass. 

 „ 3. Almost all under the violet glass and sulphate 



of quinine. 



J5 ^* -^^^ 5? ?J 55 55 



I then took the green glass instead of the yellow. 

 Obs. 1. They were under the violet glass and sulphate 

 of quinine. 

 2 



55 -" • 55 55 55 



„ 3. Abovit half under each. 

 „ 4. About three-quarters under the green glass. 

 „ 5. Almost all under the violet glass and sulphate 

 of quinine. 



Thus, then, while if the ants have to choose 

 between the violet and other coloured glasses, they 

 will always prefer one of the latter, the effect of 

 jjutting over the violet glass a layer either of sul- 

 phate of quinine or bisulphide of carbon, both of 

 which are quite transparent, but both of which cut off 

 the ultra-violet rays, is to make the violet glass seem 

 to the ants as good a shelter as any of the other glasses. 

 This seems to me strong evidence that the ultra-violet 

 rays are visible to the ants. 



I then tried similar experiments with a saturated 

 solution of chrome alum and chromium chloride. These 



