The Bionomics of South African Insects. 811 



March 27. She ate one Catopsilia florcUa ; I then put 

 in a Bclenois onescntina, and she be- 

 came much excited, running about after 

 it, and making several futile snatches at it 

 on the wing. At last she gave a vigorous 

 stroke, and missing the butterfly caught 

 the gauze with which the box was 

 covered. Imagining apparently that she 

 had caught her prey, she began trying to 

 eat the gauze, in spite of my attempts to 

 drive her away, for fully two or three 

 minutes. At last she desisted and soon 

 caught and devoured the butterfly, 

 eating a B. scverina and Axiocerces harpax 

 immediately afterwards. 

 „ 28. One Precis sesamus and one B. mcsentina. 

 29. One Hamauumida dmlalus and one B. 



severina. 

 .30. One Pyramcis cardui and one B. mcsen- 

 tincf. 

 „ 31. One J. cehrenc and two B. scverina. 

 April 1. One J. cehrene and one C.floreUa. 

 „ 2. Two G. ftoreUa and one P. cardui. 



3. Mantis escaped. She was fully as vigorous 

 and healthy on the last day as when first 

 caught. 



XI. Third Experiment with Mantis. Salisbury. 

 1898. 

 April 3. Captured a female Mantis closely allied to 

 those of preceding experiments, being of 

 same size and colouring, but having a 

 much broader thorax and the moutli 

 pink. This insect I submitted to a 

 purely distasteful diet, combined with 

 periods of starvation, as follows : — 

 ,, 4. Three Acricct caldarcna eaten. 

 „ 6. Two 



„ 7. Two „ (Lvina „ 

 8. One „ caldarcna „ 

 „ 9. Tw^o Liiiinas clirysippus „ 

 >. 10- One 

 » 11- One „ 

 „ 12. One 



