The Bionomics of South African Insects. 313 



was still scared and would not go near 

 them. 

 Sept. 8. Mantis died. There were no signs of ill- 

 health as in former experiments, and the 

 characteristic blotch in the eye was 

 absent. Death was probably due either 

 to hunger or natural causes. I only 

 wonder at its lasting so long^ considering 

 its long fast and unhealthy food. 

 [The above experiments upon Mantida; of four different 

 genera are summarized as follows : — 



Mantis I., in the Karkloof. Evident intense dislike, 

 after trial, oi Acrxa horta. 



Mantis II., male, in the Karkloof. Evident dislike of 

 A. horta, although one specimen out of three was almost 

 entirely eaten. Two A.echeria were eaten, and one partially. 

 One L. chrysi'piyus, one B. severiiia, and one P. demodocus 

 were eaten. 



Mantis III., female, probably Polyspilota caffra, at 

 Malvern. Ate one A. 'pctnva, but rejected A. serena after 

 trial. Ate one male H. inisippus, but after this refused all 

 butterflies, and exhibited signs of weakness. 



Mantis IV., female, probably the same species, at 

 Malvern. Ate Papilios and NymphalinX' freely, including 

 the probably aposematic genus Neptis, and the probably 

 aposematic Lycsenid ger^Qxa, Alx7ia and Pcntila, and Pierine 

 genus Mylothris. Ate L. chrysippus with hesitation, and 

 partially in one case, freely in two cases. Hence the Mantis 

 appeared to be a very general feeder on all butterflies 

 except the genus Acrxa, the species of which (cahira, 

 serena, cncedon) were rarely eaten until after one or more 

 trials, and were sometimes finally refused. Natalica was 

 only offered once, and rejected after trial. It is interesting 

 to note that immediately after trial of three different 

 Acrseas, the Mantis refused species which she freely ate 

 at other times. The final weakness without power of 

 recovery was a probable result of the diet. 



Mantis V., male, Pscndocreobotra ivahlhcrgi, on the Lower 

 Umkomaas River. Ate M. safitza, T. omphalc, and Neptis 

 agatha freely, the first-named on three occasions. 

 Acreea cahira, refused twice ; encedo7i, refused twice after 

 trial and accepted once ; serena, refused on five occasions, 

 eaten on four. It is probable that the weakness and loss 

 of sight was due to the Acrs&a diet. 



