VARIATION. 103 



cites several species as having red in the costal lunule, and adds, " It 

 often happens that the second lunule has a trace of the reddish or 

 orange colour, this may be occasionally found in Papth'o machaon." 

 The species he cites as having red in the upper lunule or lunules 

 " belong to at least three different but more or less allied genera," and 

 " in all the species dealt with there is at the anal angle of the hind 

 wings, a more or less distinct ocellus with a red pupil or patch : tliis is 

 very well marked in Papilio machaon.'" Mr. Jenner Weir then goes on 

 to say : " Arguing from a priori considerations, one would be inclined 

 to consider that the reddisli or orange suffusion in question was the 

 nearly obsolete trace of an ocellus -which had once existed in the 

 ancestors of these butterflies. The question then arises as to whether 

 any facts can be adduced in support of such an hypothesis. Africa is a 

 continent which has not undergone those stupendous geological changes 

 which have affected so profoundly the pala?arctic and nearctic regions, 

 from which my illustrations have been draAvn, and in which, therefrom, 

 some archaic forms of Papilionina'. might be expected to be found. 

 Orpheidea demoleus is a well-known and very common African butterfly, 

 whicli has an ocellus both on the costa and on the anal angle of the 

 lower Avings ; but it is only the lower ocellus that has mucli red aroimd 

 the pupil ; still, I have a specimen Avith the costal ocellus well marked 

 -•with red, exactly in the same part of the wing in Avhich a similar orange 

 spc^t is found in Enpho'ades troilus, and, in all cases, the dark centre of 

 the jiupil is suffused Avith red. In the \-ery closely allied Indian 

 species, Orpheides erithonius, I cannot detect any red around the upper 

 ocellus ; the loAver has a large ]:)atch of that colour, but, regarded as an 

 ocellus, it is indistinct, especially in the males ; indeed, in both these 

 species the costal ocellus is much better defined than that at the anal 

 angle. Both are Avithout tails, and it Avould seem that the dcA'cloiiment 

 of tlie tails in Papilio, JasoniadcH and EnpJia'ndes, destro3'ed the original 

 bilateral symmetry of the under Avings, and reduced the costal ocellus 

 to a mere A^estige ; and in Iphidides, Avhere the tails are more dcA-eloped, 

 CA'en the slight vestige has disappeared. The shape of the Aving in 

 this last mentioned genus is cpiite different from that of Papilio ; tlie 

 length of the tails seems to haA'e l)een effected at the expense of the 

 breadtli of the Aving ; indeed, in this res})ect Iphiclides apj)roaches 

 more nearly to the contour of the Aving of Leptocirciis. A similar re- 

 mark applies to the Indian genus of Swallow-tails, Pathisa ; in P. agetes 

 there are red spots above an obsolete ocellus in the anal angle of the 

 wing, but no red at the upper angle ; in some of the other species even 

 the red at the anal angle has faded into yelloAv, e.g. P. paphiis and 

 P. gli/cerion, and in P. antiphafes the yelloAv aboA'e tlie pseudo (tcellus 

 has disappeared. There is another genus of African PapiJioninte 

 possessing large tails, and in tlie male sex, on tlie up})er Avings, tliere 

 are androconia on the sub-median iierA'-ure and median nervules clothed 

 Avith short, cotton-like scales. I do not think this genus has ])een 

 named, but it appropriately might l^e called Erioptera. In E. ophidi- 

 cephalm, although a tailed species, the upper ocellus is develojied in the 

 lower wing ; but it may be added that, in this species, the ocellus at 

 the anal angle is very largely developed, and that there are indications 

 of a second one above the third sub-median nerAaile. In the very 

 closely allied sj^ecies E. meneslhem, the upper ocellus is absent, and in 

 its place there is a costal red spot, exactly in the same position as the 



