196 EXPEEIMENTS WITH COLOURED SOLUTIONS. 



taining 20 pupse, I placed a saturated solution of 

 bichromate of potash, a deep solution of carmine, which 

 let through scarcely any but the red rays, and a white 

 porcelain plate. 



Obs. 



1. 



2. 



3. 



4. 



6. 



6. 



7. 



8 



9. 

 10. 

 II 



Under the bichr. of potash were pupte, carmine 18, porcelain 2 



Total 



18 



81 



124 



I then put over another nest of Formica fusca 

 four layers of red glass (which, when examined with 

 the spectroscope, let through red light only), four 

 layers of green glass (which, examined in the same 

 way, transmitted nothing but a very little green), and a 

 porcelain plate. Under these cii'cumstances the ants 

 showed no marked preference, but ajipeared to feel 

 equally protected, whether they were under the red 

 glass, the green glass, or the porcelain. 



Thus, though it appears from other experiments 

 that ants are affected by red light, still the quantity 

 that passes through dark red glass does not seem greatly 

 to disturb them. I tested this again by placing over a 

 nest containing a queen and about 10 pupa3 a piece of 



