190 EXPEKIMENTS WITH COLOURED GLASSES. 



had scarcely any tendency to congregate under it. 

 There were nearly as many unaer the same area of the 

 uncovered portion of the nest as under that shaded by 

 the violet glass. 



Lasius Jlavus also showed a marked avoidance of 

 the \'iolet glass. 



I then experimented in the same way with a nest 

 of Formica fusca^ in which there were some pupag, 

 which were generally collected in a single heap. I 

 used glasses coloured dark yellow, dark green, light 

 yellow, light green, red, violet, and dark purple. The 

 colours were always in the preceding order, but, as 

 before, their place over the nest was changed after 

 every observation. 



To our eyes the purple was almost black, the violet 

 and dark green very dark and quite opaque ; the pupae 

 could be dimly seen through the red, rather more 

 clearly through the dark yellow and light green, while 

 the light yellow were almost transparent. There were 

 about 50 pupae, and the light was the ordinary diffused 

 daylight of summer. 



These observations showed a marked preference for 

 the greens and yellows. The pupae were 6^ times 

 under dark green, 3 under dark yellow, 3^ under red, 

 and once each under light yellow and light green, the 

 violet and purple being altogether neglected. 



I now tried the same ants under the same colours, 

 but in the sun ; and placed a shallow dish containing 

 some 30 per cent, solution of alum sometimes over 



