74 THE EiNTOMOLOGIST. 



Further, so far as the genus Acraa is concerned, the struc- 

 ture cannot be used for jerking the eggs away, as suggested by 

 Dr. Chapman for Paniassius, for, as stated above, the eggs do 

 not touch it ; and, besides, in the six species of Acraa which I 

 have seen laying their eggs, these were always deposited on the 

 plant, and not jerked away. 



I am thus inclined to believe that in Acrcea the sac has no 

 connection with oviposition, but that its use consists in protect- 

 ing the gravid females from the advances of unattached males. 

 But the question then arises, why the female Acrcea should 

 require this special protection. In reply thereto I may say that, 

 so far as my observation goes, in all our South African butter- 

 flies, " marriage by courtship " is the rule, with the exception of 

 the Acrseas, among which "marriage by capture" is certainly a 

 frequent, and probably the general, method ; at least, I have 

 observed its occurrence in no less than eight different species of 

 the genus. The male Acrcea does not indulge in the graceful 

 fluttering and " showing-off " which is so characteristic of the 

 courtship of most butterflies, but his chief idea seems to be to 

 seize the female with his legs. In the case of the sylvan 

 A. petrcBa and A. horta I have frequently seen a male capture a 

 female in mid-air by grappling the costa of her fore wing, thus 

 bearing her to the ground, when, after a few seconds, they would 

 fly away paired. In the open veldt species, as doubledayi, 

 nohara, violarum, &c., the female is more often pounced upon 

 while settled on the ground. 



The importance of some protection for the female while 

 laying her eggs under these cirucmstances is sufficiently evident, 

 and the development of the pouch has therefore been probably 

 fostered by natural selection. However, there is evidence to 

 show that the protection afforded by the secretion of this process 

 is not absolute, for I have taken three females in which the sac 

 has been duplicated, which I can only attribute to the result of 

 a second pairing. In every case both sacs were more or less 

 distorted or abnormal in shape, and it is therefore probable that 

 the second pairing must have taken place almost immediately after 

 the first, that is, whilst the first pouch was still soft or pliable. 

 If this be so, these exceptions need not invalidate the supposition 

 that that the pouch, when hardened, would offer a sufficient 

 obstruction to the use of the complicated male claspers. It may 

 be mentioned, that after a certain time the sac is more easily 

 detached from the ventral plate, and it is thus frequently want- 

 ing in old and battered specimens. 



In those butterflies in which courtship precedes pairing the 

 male soon ceases his attentions after persistent refusal ; but 

 among many of the Pieridse a characteristic attitude is adopted 

 by the female for this purpose. The wings are laid flat on the 

 ground, but directed so much backwards that the fore wing 



