sjiecies of Western Pacific Rhojxdocera. 361 



ants, who had managed to bore their way into it, and 

 its violent writhing attracted my attention. When I 

 passed that way again a few hours after there was 

 nothing left but a little piece of shell attached to the 

 anal hooks. If this chrysalis had had time to harden I 

 do not believe the ants could have harmed it. 



The chrysalis is from 80 mm. to 85 mm. long, angu- 

 lated ; head strongly bifid ; back considerably humped ; 

 sheath of haustellum prominent; body much pinched 

 in at centre, where the suspending silken girdle supports 

 it ; ridged subdorsally, and with two small tubercles 

 each side of lower })art of abdomen ; all the upper parts 

 of a beautiful deep bluish green ; costal portion of wing- 

 sheath streaked with silvery white ; a spiracular silvery 

 white abdominal streak; a short ventral silvery stripe 

 near the anal point ; central portion of the abdomen 

 bright golden green, forming a conspicuous triangular 

 mark ; sheath of haustellum and other parts slightly 

 tinged with golden green. Shortly bcifore the butterfly 

 emerges the chrysalis changes to a deep leaden green 

 colour, and then to black, and the butterflies generally 

 appeared between seven and eight in the morning. 



Pwpilio Godefroyi, Semp. 



This is, I believe, the only species of Papilio (and the 

 largest butterfly) occurring in the Samoan islands. It 

 was common near Apia, on the island of Opolu, and at 

 Pango-Pango, on the island of Tutuila, in June, 1884, 

 and is probably to be found upon all the islands of the 

 group. I was fortunate enough to take it in all its 

 stages. Its habits are almost similar to those of Papilio 

 Schmeltzi. The perfect insect flies rapidly among the 

 forest trees, and often ascends to a considerable height, 

 being attracted by the flowers on the topmost branches. 

 I only succeeded in capturing one, although I had 

 several other chances. They fly in a wild irregular 

 manner, and arc difiicult to net. The larvae were not 

 uncommon, and I took them of all sizes. They fed 

 perfectly exposed upon young stunted plants of Andia 

 growing in shady and sheltered places. This Andia 

 differed slightly from the Fijian, the leaves being larger 

 and the shrub altogether more robust ; but it was 

 evidently merely a climatic variety. 



The ova were attached to the under sides of the leaves 



