616 F. Finn — Experiments with various Birds. [No. 4, 



IV. The Delias eucharis pnt in the previous day was still alive 

 and not much torn, while of the other butterflies I found only wings 

 left. There was no fruit in the cage but a partly eaten orange which 

 I had put in at the same time as the butterflies. I put in some cock- 

 roaches, which, like the D. eucharis, were not attacked. Later on in the 

 day I found the D. eucharis dead, but uneaten. The cockroaches got 

 away under the drinking vessel. 



I put into the aviary a large insect-cage containing various butter- 

 flies, mostly warningly-coloured ones ; a Limenitis and two Danais 

 {chrysippus and genutia ) first came out, and the former was seized by 

 a Liothrix, which could just as easily have taken a Danais. The 

 Liothrix did not take any insects from the cage, but one then caught, 

 tore, and apparently ate the D. genutia which had come out. They 

 tried, however, to get at the insects through the glass, and then one 

 came to the entrance and took a non-warningly-coloured butterfly. 



I then took out the cage and put its contents into the aviary. I 

 did not see the Liothrix eat any more Danais — on the contrary, I several 

 times saw them take a non-warningly-coloured butterfly when they 

 could easily have had one of these. 



I saw a Bulbul swallow a D. chrysippus, and two D. chrysippus 

 and a D. limniace taken, beaten, and dropped by this species. There 

 was plenty of fruit in the cage. 



When the birds were going to roost the only non-warningly- 

 coloured butterflies left out of about a dozen were one Blymnias undu- 

 laris <3 (dead), and a Nepheronia hippia d*. These two I took out and 

 reserved, with two Danais chrysippus. (I could not find them, however, 

 when I wanted them next day ) . 



V. The Delias eucharis mentioned in the account of the experi- 

 ments of the previous day, was still uneaten, though the other butter- 

 flies seemed to have been devoured. 



I put in some more butterflies, and saw that though the Liothrix 

 readily attacked Danais at first, they took non-warningly-coloured 

 butterflies when they had the choice. 



I now added to the collection two common Bulbul s (Molpastes 

 bengalensis), and four Button-Quails {Turnix taigoor), and put in some 

 more butterflies, including a Papilio aristolochise. 



This time again the non-warningly-coloured species were obviously 

 preferred by the Liothrix. 



At some time or other to-day I saw a D. chrysippus swallowed whole 

 by a Red- whiskered Bulbul, and another eaten by a Liothrix, which 

 species also ate a Danais limniace. I also saw a common Bulbul reject 

 a D. chrysippus, many of which species had been given. 



