r.VMILY SATURNIID.E. 79 



almost in their primitive condition, and prove that the beautifully shaped ocelli arc the result of the 

 iiiodilication of three of the discal bands witii their interspaces ; tlie central band is widened and toned 

 down in colour, and the irregularities of the two outer bands modified into a regular arch, and then wc 

 get ocelli such as the fourth on the frontwings ; now comes tiie most difficult point — the introduction 

 of light, but looking at the third ocellus of the hindwings we see the internervular fold of a whitey- 

 brown colour which Ixjcoraes white in the fourth ocellus, is modilied into a small white spot in the 

 hfth, and into a large elongate spot in the sixth ; thus we see that, if necessary, the three black bands 

 with their interspaces, immediately following the central white band, could as c;jsily be modilied into a 

 second scries of oviform ocelli. 



But now the question arises — Is there any Bra/nnivn entirely destitute of these remarkable 

 ocellations ? To this I can only reply that at present I do not know of any ; it is however an 

 imjjortiint fact that li. Swanzii and li. Litciwi are the only species that have brought them to 

 perfection, wliilst in Brahnuca luniilafft they are entirely wanting to the hindwings ; in li. Orthia and 

 B. W/iilvi they appear to have been produced by the modification of two only of the discal bars, and 

 therefore more nearly resemble drops of dirty water than oviform ocelli. 



Supposing that the above ocelli have been produced by sexual selection, as I am inclined to 

 believe, from the fact that there are slight differences between those of the male and female in Ji. 

 Certhi'i (the only species of which I have seen both sexes), I can conjecture how this action may have 

 come into play : it will be clear, at least to every artistic eye, that the ocelli upon the opposite wings 

 are lighted up from opposite directions, and therefore, taken together, are incorrectly shaded ; it is also 

 manifest that, not only would some of the ocelli be hidden when the moth was at rest, but in that 

 condition it could never charm its partner ; however, the habits of many of the day-flying moths, .to 

 which this genus in all probability belongs, have guided me to what I think may be the correct 

 solution of the difficulty — the females of many of these moths when fresh remain motionless, whilst 

 their admirers, sometimes in numbers, fly round them ; very many moths when approaching the female 

 swing up and down with a pendulum motion behind her, gradually shortening the movement as 

 they near the prize ; in this kind of flight the female would only see two wings perfectly at the end of 

 each swing, and it would be desirable that these wings should be correctly shaded, as wouM be the 

 case with Bntlimiva Sicniizii. 



I am doubtful about the accuracy of the white markings in the central black patch of the front- 

 wings, the specimen from which I took my figure being somewhat rubbed. 



A curious coloured ocellus in course of formation may be seen in fig. 3 of the same jilate, whilst 

 on pi. xxxi. fig. 1, an interesting alierration of Roiifikosonui CerM, in which most of the usual black 

 spots are converted into glossy blue, has retained two of these spots in a transitional state, and thus 

 given us a clue to the probable development of the ordinary ocelli so common on S'lii/ridiaii butterflies, 

 and to the origin of the ring-like markings in the discoidal cell so universal in the allied genera 

 £uri/j)JieHe Aterica, Ilnrmn, Ailoli'ts, &c. 



