EXPERIMENTS ON VISION OF INSECTS. 175 



often to the trellised opening. M. PJateau concludes 

 that insects do not distinguish differences of form, or 

 can only do so very badly ("lis ne distinguent pas la 

 forme des objects on la distinguent fort mal "). 



I confess, however, that these experiments, ingenious 

 as they are, do not seem to me to justify the conclu- 

 sions which M. Plateau has deduced from them. 

 Unless the insects had some means of measuring 

 distance (of which we have no clear evidence), they 

 could not tell that even the smaller orifice might not 

 be quite large enough to afford them a free passage. 

 The bars, moreover, would probably appear to them 

 somewhat blurred. Again, they could not possibly 

 tell that the bars really crossed the orifice, and if they 

 were situated an inch or two further off they would 

 constitute no barrier. 



I have tried some experiments, not yet enough to be 

 conclusive, but which lead me to a different conclusion 

 from that of M. Plateau. I trained wasps to come to 

 a drop of honey placed on paper, and, when the insects 

 had learned their lesson, changed the form of the paper, 

 as I had previously changed the color. It certainly 

 seemed to me that the insect recognized the change. 

 M. Forel has also tried similar experiments, and with 

 the same result. 



We know, however, as yet very little with reference 

 to the actual power of vision possessed by insects. 



On the Function of Ocelli. 



Another interesting question remains. What is the 

 function of the ocelli ? Why do insects have two sorts 

 of eyes ? 



