BUTTERFLY VISION. 1671 



made, obviatinn: all these objections, and has caniod mit tliesc t'xpcriinents 

 even luuro (.•aret'ully and extensively tliaii tlic |ii\'vioii>!, instituting, more- 

 over, conipaiative investigations witli tiie vision of vertebrates under 

 precisely similar circumstances. To do this he constructed what he has 

 termed a labyrinth, a table covered with a large number of vertical l)arricrs 

 placed in concentric series in sucii positions that the creature must lake a 

 very circuitous course among thcni to escape. Of course onlv crawling 

 insects could be experimented on, but the structure of the eyes is essen- 

 tially the same in ail. Care was taken tiiat tlic ctpliiratidn of the surface 

 should be of a neutral tint and the vertical barriers were painted of various 

 colors, white, brown or black; also that the animal should begin its move- 

 ments without excitement, by the quiet lifting of a glass cover under 

 which it had been placed, when it would find itself surrounded by walls be- 

 tween wiiieh were abundant passages. Where the vision was really good, 

 the animal would be ex|)ected to move in a serpentine course between the ob- 

 stacles, never striking them, usually moving toward the opening which 

 was largest or nearest. When, however, the power of vision was to some 

 extent defective, so that the animal seemed to be aware of the existence of 

 an olistacle only when it had almost reached it, then the movements should 

 be made in a series of zigzags which would change direction somewhat 

 slr.ir[)ly, shortly before the barrier v,-as reached. Or if the vision were 

 absolutely defective, the creature would be likely to find its way only by 

 first striking the objects and then moving around them. 



The result of his experiments proved that vertebrates had complete 

 vision, directing their movements with ease without striking any of the 

 obstacles, moving in the nearest path ; wiiile insects acted in all cases 

 as if they had a veil before their eyes, their change of direction before 

 reaching the I)arrier being such as to indicate that it was only when they 

 reached the shadow of the obstacle before them, when they could distin- 

 guish some difference in the intensity of light, that they turned aside 

 to avoid such an obstacle. The result of these experiments has been so 

 uniform and so clear that, however the actual anatomical structure of the 

 eyes of insects niiiy be regarded, there can be little doubt that their vision 

 is 80 extremely imperfect that they perceive sharp images of nothing im- 

 mobile, and, therefore, do not distinguish the precise form of objects, 

 thfiugh they can readily distinguish objects in motion ; indeed they are 

 particularly keen in this sort of vision, so that their sight must be best 

 while upon the wing ; and this at once explains the tremulous movement of 

 the Lycaenidae before alighting on an object, for of all buttei-flies these, as 

 Plateau's special experiments show, have the poorest vision, being dis- 

 turbed by a moving object only when at half the distance at which a Nym- 

 phalid would take alarm.* So too, they can distinguish masses of color, 



• It sh'julij also be noticed that the Lycaeni- smallest visual surface on the oeelhirglobe, 

 dae have, of all butterflies, relatively the being hardly half that of some. 



