186 LIMITS. OP VISION. 



course be little, if any, doubt, that bees are capable of 

 distinguishing colours; and I have proved experi- 

 mentally that this is the case. Under these circum- 

 stances, I have been naturally anxious to ascertain, 

 if possible, whether the same holds good with ants. 

 I have, however, found more diflBculty in doing so 

 because, as shown in the observations just recorded, 

 ants find their food so much more by smell than by 

 sight. 



This being so, I could not apply to ants those 

 tests which had been used in the case of bees. 

 At length, however, it occurred to me that I 

 might utilize the dislike which ants, when in their 

 nests, have to light. Of course they have no such 

 feeling when they are out in search of food ; but if 

 light is let in upon their nests, they at once hurry 

 about in search of the darkest comers, and there they 

 all congregate. If, for instance, I uncovered one of 

 my nests and then placed an opaque substance over one 

 portion, the ants invariably collected in the shaded part. 



I procured, therefore, four similar strips of glass, 

 coloured respectively green, yellow, red, and blue, or, 

 rather, violet. The yellow was rather paler in shade, 

 and that glass consequently rather more transparent 

 than the green, which, again, was rather more trans- 

 parent than the red or violet. I also procured some 

 coloured solutions. 



Prof. Dewar was kind enough to test my glasses 

 and solutions with reference to their power of trans- 



