174 EXPERIMENTS ON VISION OF INSECTS, 



equal, the opening on the one side should leave a clear 

 passage, while that on the other should be divided by- 

 bars large enough to be easily visible, and sufficiently 

 close to prevent the insect from passing. 



His experiments were conducted in a room five 

 metres square, lighted by two similar windows looking 

 to the west. It was on the first flour, and looked out 

 on to fields. Moreover, he had ihe glass of tiie windows 

 slightly ground, so that, while the light penetrated, 

 nothing outside could be seen. He then covered up 

 the windows, leaving only two orifices, one of which 

 was simple and square, while the other was divided 

 by cross-bars. To secure equality of light, the latter 

 was left somewhat larger than the other, and the 

 equivalence of the two was determined by a Rumford's 

 photometer. The insects were set free on a table at 

 the back of the room, exactly between the two open- 

 ings, and at a distance of four metres. He states that 

 a very slight difi'erence in the intensity of the light 

 determined the flight of the insect to either one or the 

 other opening; nliile, if the amount of light was as 

 nearly as possible equal, they flew as often to the one 

 as to the other. 



Omitting the ca-!es when the light was not equal, the 

 numbers wqre as follows: — ■ 



Clear Trellised 

 opening, opening. 



Musca vomitoH a (ihe hluobotUe') 8 ... 7 



On the other hand, tliey were — for 



Eristalis tenux (the hee &y) i ... 8 



Vanessa urtiiee (tortoiseshell buttevfly) ... ... 1 ... 5 



13 20 



In fact, then, the insects seem to have gone more 



