200 



around at an angle of i8o degrees, so that its head pointed in 

 the direction from which the males came, all of which followed 

 the same course. It could be observed that, after hovering 

 for a little while above the supposed female in a reversed 

 position, they would suddenly turn around as if they were 

 aware of their mistake, so that we are led to assume that they 

 plainly recognised what was head in the model and what was 

 tail. It must be added, however, that by no means all the 

 males that arrived acted thus, but only those which were 

 unusually active and insistent, and that the distance at which 

 recognition took place never exceeded 6 to lo cm. (2 to 4 in.), 

 whereas in the case of the wrong colour it was more than 2 metres 

 (6 ft.), and abnormal size was perceived approximately at a 

 distance of from ^ to i^ metres (ij to 4I ft.). 



The amount of assistance derived from the sense of touch 

 was tested in the following way : It was noticed that if freshly 

 captured specimens, which had been killed with cyanide of 

 potassium, were fastened in the right position, copulation took 

 place in a normal manner, as far as could be observed ; even 

 females that had been taken the day before, and were entirely 

 dry, exerted the same attraction as freshly captured or living 

 specimens. The paper model, however, offering a smooth 

 surface affording no foothold for the males, indicated that the 

 males were not materially aided by touch. As soon as the 

 abdomen of the very accurately cut paper model was turned 

 slightly upwards, it was observed that the males endeavoured 

 to touch it ; but although they were not at all shy, and hovered 

 close around the model when it was held in the hand, the quick- 

 ness of their movements, aided by the blinding sunhght, made 

 it impossible to see whether contact actually took place. 



Another observation also proved that in butterflies the 

 sense of touch is not greatly developed. The wind, which on 

 the tops of those hills always blows pretty strongly, in striking 

 the edges of the paper communicated to the wings of the paper 

 models a fluttering or vibrating motion : each time this took 

 place the males present were visibly affected, and renewed their 

 efforts with increasing energy. Having settled on the back of 

 the model, they were so far from resenting being hit by the 

 paper wings which oscillated in the wind, that they became 



