MIMETIC ASSOCIATIONS 



95 



form, which produced fifteen males and seventeen females, one of which was hippocoon and 

 the rest cenea. The last family arose from a hippocoon female, and contained seventeen males 

 and thirteen females, all the latter being of the cenea type. Perhaps the most obvious point 

 elucidated by these breeding experiments is the fact that the males do not appear in any- 

 thing like the overwhelming proportion in which they are observed in nature ; in fact, 

 excluding the case in which only five specimens were reared, in only one instance did the 

 number of males even exceed that of the females, and then only in the ratio of seventeen 

 to thirteen, whilst the total sum of individuals shows eighty females to seventy-three males, 

 thus showing that the scarcity of females in the field is apparent rather than real. The 

 next point to strike the observer is the great preponderance of cenea and the rarity of 

 trophonius. We should expect the cenea form to be plentiful in view of the predominance of 

 Amauris echeria. On the other hand, chrysippus is also plentiful, and we should expect the 

 corresponding form trophonius to be at least as common as the others. Such, however, is not 

 the case and the trophonius form is everywhere rare. This fact has been remarked upon by 

 Trimen (Trans. Ent. Soc, 1904, p. 688), and the subject has been still more fully dealt with 

 by Professor Poulton (1. c). In discussing the predominance of the forms of dardanus, that 

 author points out that, judging from a large collection received from the Chirinda Forest in 

 South-Eastern Rhodesia, the predominating form appears to be hippocoon^ which was found 

 in the proportion of twenty-two to one of cenea and one of trophonius. In spite of this fact, 

 Amauris dominicanus is by no means the predominant model in the district in question ; 

 indeed, judging from the numbers actually collected, Amauris lobengula^ and albimaculata 

 greatly outnumbered the black and white Amauris, as did also D. chrysippus. The con- 

 clusion arrived at is, that bearing in mind that A. dominicanus is nevertheless a common 

 butterfly at Chirinda, whereas in Natal it is rare, it may be that the black and white Amauris 

 confers a greater benefit on the mimic by its greater conspicuousness, especially in the 

 forest. The question of the relative efficiency of the different Danaine models is, however, 

 a subject which requires considerable further study. Dealing with the rarity of trophonius, 

 the author points out that whilst dardanus is a forest insect, D. chrysippus frequents more 

 open regions, and thus the influence of the latter as a model would be to a great extent 

 limited to those areas where forest and open country were contiguous. Hitherto we have 

 been considering three forms of the female dardanus, but these by no means complete the list 

 of known forms of this protean butterfly. The form dionysos described by Doubleday and 

 Hewitson in 1846 is, so far as we know at present, confined to the west coast. As will be 

 seen from a reference to the plate, it resembles the hippocoon form, but has the ground- 

 colour of the hind-wings yellow. It is decidedly rare in collections and can only be regarded 

 as either a primitive intermediate form or a reversion occurring sporadically. Figs. 4 and 8 

 represent two extremely interesting varieties, niavioides and ruspinae, of the Abyssinian 

 subspecies antinorii taken at Lake Tana by the late Dr. Stecker. These two butterflies 

 are at present absolutely unique, no other examples ever having been received. I am 

 indebted to the publishers of Haase's work on mimicry for permission to copy their 

 figures of the insects which I understand are now at Prague. They were described 

 and figured by Kheil in 1890. As will be seen, niavioides presents the hippocoon form, 

 but retaining the tails of the hind-wings, whilst ruspinae is distinguished by having 

 the whole of the basal area in both wings reddish-yellow, as in the trophonius variety of 

 the other subspecies. These butterflies remain something of a mystery. They may be 

 an ancestral form of hippocoon and trophonius, and may have occurred by reversion. 



^ A form nearly resembling Amauris echeria. 



