PHASES OF NATURE AROUND PRETORIA. 65 



increase was curtailed; but though generally unsuccessful, 

 I did discover what I believe up to the present to be its 

 only recorded enemy. This is a moderately large or- 

 thopterous insect (Hemisaga prcedatoria, n. sp.), which I 

 found lurking among the tops of tall flowering grasses, 

 to which it has a considerable assimilative resemblance 

 and which in this case enables it to secure its prey. 

 The Danais hovers about, or partly settles on, the flowers 

 and is then secured by the Hemisaga, which, in one 

 instance, I found dismembering a freshly-caught speci- 

 men *. It is just possible, during the dry season, 

 Avhen insect-life is very scarce, that some insectivorous 

 birds may, in a somewhat famished condition, make an 

 experimental dash at a Danais. At that season I 

 captured a specimen which was certainly mutilated 

 as though by the bill of a bird, for the wings were 

 not bitten symmetrically, as is the case when the 

 attack takes place by a lizard or mantis, whilst the 

 butterfly is reposing with its wings vertically closed f . 



As is well known, the female of Hypolimnas misippiis 

 is a wonderful mimic of this butterfly. To an expe- 

 rienced eye the Hypolimnas may be distinguished from 

 the Danais by its flight ; but this is scarcely noticed 

 without both species are known to be present and 

 attention is thus directed. So close is the resemblance 

 that well knowing both insects, I was not aware of the 

 female Hypolimnas being present with the Danaids till 

 I observed one in copula with its dark blue male. A 

 purely English lepidopterist, not knowing these facts in 

 mimicry could cross the veld and merely observe that 

 D. chrysippus was very abundant. But these mimicking 

 resemblances, by which the fema\eHypolimnas has found 

 protection by being mistaken for the uneatable Danais 

 and avoided accordingly, are even still more complicated. 

 D. chrysippus has two varietal forms, alcippus, Cram., 

 and dorippus, Klug, both of which occur in South 

 Africa ancl both of which I found in the Transvaal. 



* When in Natal that old lepidopterological veteran, Col. Bowker, informed 

 me that he had frequently observed the Mantid(c to prey on butterflies. 



t I am bound to affirm that this view, formerly advocated by rny 1'riend 

 Prof. Meldola, was at the time contested by myself. 



F 



