VANESSA ANTIOPA LINN. 457 



butterfly rests with its head away from the source of light, we should expect that on 

 a vertical surface, such as a tree-trunk, it would rest vertically, with the head directed 

 downward. For the inverted position assumed on tree-trunks and on other vertical 

 surfaces brings the butterfly into the same relation to the source of light as the position 

 taken on the level does. 



A second position, which was often observed and which for some time puzzled 

 me, was one in which all evidence of orientation disappeared. A butterfly, that 

 on several previous occasions had alighted with perfect orientation, came to rest 

 entirely out of position and, after turning once or twice, it remained quiet in what 

 seemed to be an entirely irregular position. On watching such instances carefully 

 I found that they often occurred on rocks, and further that they happened only when 

 the rock surface was approximately at right angles to the sun's rays, so that any posi- 

 tion assumed by a butterfly alighting on this surface was like every other one so far 

 as the direction of the rays was concerned. That such butterflies were entirely 

 capable of orienting in the usual way was proved in half a dozen instances by driving 

 the insects to a level situation where all were found to orient with perfect accuracy. 



A factor that in all cases overruled phototropism was the chemotropic response 

 to food. When a butterfly alights on a bough, it orients in the sunlight with the 

 usual precision. Should the sap be running from a near stem, the insect is very soon 

 attracted to the spot, begins feeding, and moves about from that time on with no 

 reference to the direction of the sun's rays. Thus when feeding or near food the 

 butterflies do not respond phototropically. When, however, food is absent and the 

 butterfly alights in bright sunlight on a surface where orientation can be accomplished, 

 it places the axis of its body as nearly as possible parallel with the rays of light, with 

 the head away from the source of light, its negative phototropism being of a pro- 

 nounced and unequivocal kind. 



That similar conditions obtain among some other species of butterflies seems 

 probable from the fact that some members of the closely allied genus Grapta orient 

 with as much precision and in the same way as Vanessa antiopa. This peculiarity 

 may also occur among some of the hesperids, but such of the papilios, pierids, and 

 lycsenids as I have observed have given no sign of this form of phototropism. 



So far as I am aware, Radl (: 01, p. 457) is the only observer who has recorded 

 similar observations on the orientation of butterflies. According to him, one butter- 

 fly that he observed, an undetermined satyrid, oriented almost invariably with the 

 head to the east when the sun was in the west, but in other instances so much irregu- 

 larity was shown that Radl refrained from drawing any general conclusions. Had 

 he paid attention to the exact direction of the sun's rays in relation to the surfaces 



