( 606 ) XXV. Pseudacraea eurytus hobleyi, Neave, its forms and its models on Bugalla Island, Lake Victoria, with other members of the same combination. By G. D. Hale Carpentee, D.M., Oxon., Member of the Royal Society's Sleeping-sickness Commission. [Read November 5th, 1913.] Plates XXXIV-XXXVI. The following is a complete account of all the forms of Pseudacraea eurytus hobleyi, their Planema models, and other mimics in the same group, which I caught on Bugalla Island in 1912 and January-February, 1913. I wish, firstly, to express my indebtedness to Prof. Poulton for the great help he has given me in the preparation of this paper, especially in the preparation of the plates, the arrangement of which is entirely due to him. It seemed best to pubhsh the results in tabular form, in spite of the greater bulk of such a paper, because by such means a graphic representation of the numerical differences between models and mimics is brought home to the reader as he sees the long array of blank spaces under the headings of the models. I have taken the opportunity of figuring, on Plate XXXIV, some of the most interesting transitional forms of Ps. eurytus hobleyi from Bugalla Island, and of showing the close relationship of a single female (fig. 11) to a typical West African female of eurytus, L., from the Lagos dis- trict, represented in fig. 12, with its model Planema epaea, Cram., in fig. 13. On Plate XXXV I have figured three of the most interesting of the famihes of Ps. eurytus hobleyi bred from known female parents captured on Bugalla Island. An account of two of the families, B and E (figs. 1-8), together with other synepigonic groups from the same locahty, has already been pubhshed in these Transactions (1912, pp. 706-16). The third family, J (figs. 9-16), is recorded in Proc. Ent. Soc. 1913, pp. ix-xi. These breeding experi- ments conclusively prove that all the forms of eurytus hobleyi tabulated in the present paper form a single interbreeding community. trans, ent. soc. lond. 1913. — part iv. (mar. 1914) Dr. G. D. H. Carpenter on Pseudacraea eurytus hobleyi. 607 Plate XXXVI represents, in figs. 14-17, some of the intermediate forms of eurytus hobleyi captured by me in Damba Island (1911) — a part of the series of which nearly the whole was described by Prof. Poulton in our Proceedings (1911, pp. xci-v; 1912, pp. xix-xxiii). Figs. 1-7 represent typical Planema models, and figs. 8-13, typical Pseudacraea mimics captured by Mr. C. A. Wiggins, D.P.M.O. of the Uganda Protectorate — fair examples of his great collection of these forms of which a part is pub- lished in " I. Congr. Internat. d'Ent.," 1910, vol. ii, p. 483. Fig. 10 represents the male-hke female, poggeoides, of Ps. eurytus hobleyi, rare on Bugalla and Damba, even rarer near Entebbe, but common to the E. of the Nile, where PI. poggei is found, but PL macarista absent (Pro- ceedings, 1912, pp. Ixx-lxxi). Plate XXXVI illustrates the intermediate forms of Pseudacraea that are relatively common on Damba as compared with the mainland — those intermediates that will be here shown by a much larger mass of evidence to be also characteristic of Bugalla. In correspondence with this resemblance between the Pseudacraea mimics of the two islands. Prof. Poulton has shown [1. c.) that the Planema models are relatively rare on Damba, and they are shown in the following tabular statement to be relatively rare on Bugalla. It must furthermore be borne in mind that the 127 Bugalla Planemas include 75 epaea paragea, and that special reasons for this large proportion are given later (p. 611). Mr. C. A. Wiggins' collection, between May 23 and Aug. 31, 1909, is analysed in our Proceedings, 1912, p. xciii, where it is shown that 244 Planemas and 82 forms of eurytus hobleyi were taken. What a contrast to the respective figures — 127 and 356— for Bugalla ! In the tabular statement on p. 608 the numerical relations between the various Planema models and their mimics on Bugalla can be seen at a glance. There were also taken during this period 17 Mimacraea poultoni, Neave, of which one specimen might be considered to be an outlying member of Combination Ib, as it had the orange of the hind- wings replaced by white. There was considerable variation amongst these Mimacraeas : one being of a paler yellow was a beautiful mimic of Acraea viviana, Staud. In the locality where the mimetic Ly- caenids were taken, the model for the normal form of poultoni appeared to be Acraea alicia, E. M. Sharpe. 608 Dr. G. D. Hale Carpenter on Pseudacraea eurytus hobleyi. 609 Combination Ia. The model, Planema poggei nelsoni (Plate XXXVI, figs. 3, 4), was the scarcest of all the Planemas on Bugalla Island : I only succeeded in taking 2 males and no females during the 14 months I was there. Its chief mimic, Pseudacraea kuenowi hypoxantha, was not so uncommon, 3 males and 6 females having been taken. This fine Pseudacraea seemed to vary very httle indeed : a marked contrast to the protean Ps. eurytus hobleyi. I found it quite easy to distinguish the hving kuenowi from the ^ hobleyi : its flight is very much bolder, it seems even more alert, with brisker move- ments ; and when seen on the flowers of bushes which it frequents, it appears to carry the wings in a shghtly different manner. I have never been deceived by its likeness to the model as in the case of hobleyi. Two females of eurytus hobleyi fall into this combination. They belong to the form poggeoides (Plate XXXVI, fig. 10), with a yellow band across the fore-wing. This is not quite of the same tint as the orange band of the male, and corresponds with the band of poggei rather than of the ^ tnacarista. A single specimen of the planemoides female of Pap. dardanus was obtained. It is worth noting that, although I had been collecting for two years before I came across this interesting butterfly, I was completely deceived by it. It was flying slowly in front of me in an open space in the forest belt, and my first thought was " What an enormous Planefna ! ", so much did its general appearance and flight resemble its model. When captured, it lay perfectly still in the net, as does its model, instead of fluttering wildly like so many Papihos. This specimen is of particular interest, for from ova obtained from it I was able to rear the family exhibited at a meeting of the society (Proceed- ings, 1913, p. liii) and figured on Plate XXXIX of the present volume. It is worth mentioning that I only col- lected one other dardanus female on Bugalla — of the form hippocoon, F. Combination Ib. Of the model, the male PI. macarista (Plate XXXVI, fig. 2), 10 were taken. Acraeine mimics are represented by the female of Acraea alciope — ^synaposematic with the male macarista ; 610 Dr. G. D. Hale Carpenter on though many of these females have such a broad brownish border to the white band on the hind-wing that they are to some extent intermediate between the typical eastern Uganda $ form aurivillii, and the typical western female. Of the 17 alciope which were taken 5 were males, 11 were of the eastern form of female {aurivillii), and one transi- tional towards the typical western form, which closely resembles the male PI. alcinoe mentioned below. The single transitional $ resembled this model so closely that I was quite deceived by it even after I had seen the speci- mens in the cabinet. The male Acraea alciope stands by itself, and does not mimic anything in either E. or W. Africa. The scarcity of this species on Bugalla Island is remarkable — and probably due to scarcity of its food-plant, which I never saw there. On another island, where I was previously (Damba), the food-plant was abundant, and alciope was extremely common. The Pseudacraea mimic is the male of the mimetic form hobleyi, of Ps. euryius hobleyi (Plate XXXV, fig. 12 ; XXXVI, fig. 9). This mimic is abundant on Bugalla, 28 having been captured. The resemblance is so close that I was often deceived until I had learnt to distinguish them. Another Nymphaline member of this combination is Precis rauana, whose male is non-mimetic, but the female mimics well the male macarista : 24 males and 23 females were taken. This species is not often seen actually withiro the forest, but is to be found along the border-hne between the forest and the open grass-land, or at the edge of the forest on the shore. In both these localities may be found beds of a thick-leaved aromatic Labiate herb, which may be the food-plant of the larva of this species. It is to be found also on the flowers of the " Gamboge " tree, Haronga madagascariensis, Chois. (Hypericineae), which particularly favours such localities, and attracts numbers of all the butterflies mentioned in this paper. Precis rauana has the typical, very dashing and rapid flight of the genus to which it belongs, and is unlike that of its model. But the general impression gained from its appearance leaves no room for doubt that the female is a mimic of the male PI. macarista. The male, having no white on the hind-wings, is not mimetic of this Planema, but on the other hand a fresh specimen often has such a rich crimson suffusion over the hght-coloured band on the wings that I think it presents a decidedly Acraeine Pseudacraea eurytus hobleyi. 611 appearance when one gets a glint of crimson as it flashes past. Another interesting point is that this butterfly- seems, at first, to rely for its protection upon the appear- ance of its upper surface. It is not always an easy species to catch; and if one strikes at it and fails, it will settle again and open and close its wings, displaying the colouring of the upper sides. If, however, one follows up and strikes again so that it is really alarmed, it will fly off and make use of the markedly procryptic, dead-leaf like appearance of the underside, sitting motionless with the wings brought together over its back. I endeavoured to obtain ova from captive females, putting them with branches of the aromatic herb before mentioned, but was unsuccessful. The early stages are, I believe, not known. Combination II. Seven examples of the principal model, the female of PI. macarista (Plate XXXVI, fig. 1), were taken. A second model is provided by the female of PI. alcinoe camerunica, of which 2 were captured, together with 8 males. The resemblance between these two female Planemas is extraordinarily close, so that it was a very long time before I was able to differentiate them. The male alcinoe is totally different and is of a type common in W. Africa, but comparatively rare in Uganda. The black-and-white female of Acraea jodutta — the jodutta form of female — is beautifully synaposematic with the two Planema models. Of this mimic 3 were taken. The models are closely mimicked by the abundant female of the form tirikensis (Plate XXXV, figs. 5, 6, 9 ; XXXVI, fig. 8) of Pseudacraea eurytus hobleyi, of which 40 were taken. Combination III. The model is the eastern form, paragea (Plate XXXVI, fig. 7), of the western species Planema epaea. Of this 41 males and 34 females were taken, so that it seems not uncommon. But these figures give a quite dispropor- tionate idea of its relative abundance in the forests. I happened to hit upon a locality at the edge of the forest where the species seemed to collect in numbers owing to the attractiveness of certain flowers, and I naturally made a point of visiting this locality every evening, since I wanted as many specimens as I could obtain. Had I 612 Dr. Gr. D. Hale Carpenter on merely caught what I saw in the forests, it is doubtful if a dozen specimens would have been obtained. The Bugalla specimens are interesting as they are all very hght-coloured, hke the lightest forms obtained by Mr. Wiggins at Entebbe and presented by him to the Hope Department. They contrast very markedly with the 4 specimens which were all that I obtained in the forests of Damba Island, and were very dark indeed (Proc. Ent. Soc, 1912, pp. xxiii, Ixxxvi). The form of eurytus hobleyi mimetic of paragea, namely obscura (Plate XXXV, figs. 1, 10, 11, 13-16; XXXVI, fig. 13), was the least abundant of all the mimics into which this Pseudacraea subdivides, only 7 fully mimetic males and 19 such females being obtained. The only other known mimic of paragea, namely the form peculiaris of Papilio cynorta, I did not obtain, much to my disappointment. The species does occur on the island, however, for I caught a single male, which is totally different in appearance from the female. It would be extremely interesting to ascertain whether the island female is also much paler than usual, following the model. Combination IV. The model is PI. tellus eumelis {platyxantha), of which the male and female are aUke (see Plate XXXVI, figs. 5, 6) : 24 males and 9 females were captured. This species exhibits in a marked degree the nonchalance of a typical model. I spent a long time one evening trying to get a photograph of this butterfly on a clump of mauve Com- posite flowers, Erlangea tomentosa, S. Moore, which were extraordinarily attractive to all these butterflies ; and although it frequently took alarm and flew away, it as frequently returned after a very short time. Indeed, I could almost have caught it in my hand. There is one synaposematic Acraea in this Combination, namely A. jodutta, of which 3 males and 6 females of the dorotheae form were taken. The resemblance of this latter female form to PI. tellus is extremely close, and until I had learnt the generic differences between Acraea and Planema I was always confusing the two. The specimens showed some variation : in one or two cases the black bar between the subapical and inner marginal tawny areas on the fore-wing is broken through, forming a variety comparable to those of Ps. terra, described on p. 613, Pseitdacraea eurytus hobleyi. 613 The form of Ps. eurytus hobleyi, mimetic of PI. tellus, namely terra (Plate XXXV, figs. 2, 4 ; XXXVI, figs. 11, 12), was the most abundant of all the forms, 104 being taken altogether. Of these, 39 males and 26 females corre- sponded with the type, while 6 males and 20 females diftered only by having the tawny subapical area on the fore-wing suffused with white scales to a greater or less extent. In 1 1 males and 1 female the black bar between the subapical and the inner marginal area was thinned or broken through, so that, in the most completely developed variety (No. 33 in hst : Plate XXXIV, fig. 7) there is one large tawny area on the fore-wing of irregular shape, and bordered with black. An even more extreme form from Damba Island is represented on Plate XXXVI, fig. 16. To this variety Griinberg has given the name impleta. Transition in Bugalla Island between the mimetic FORMS OF Ps. eurytus hobleyi. I now come to the most interesting points, which this paper is intended to demonstrate. It will be seen in the tabular statement (pp. 618 et seqq.) that there are very many forms of Pseudacraea eurytus hobleyi not belonging to any of the types, but described as transitional. (1) Between (^ hobleyi with $ tirikensis and obscura there are 45 of these intermediates, (2) between obscura and terra 37, and (3) between terra and ^ hobleyi with $ tirikensis 74. Classes (1) and (3) are principally shown to be inter- mediate by the development in various degrees of the umber basal patch on the under surface of the hind-wing, a feature that is characteristic of the ^ hobleyi and its ^, tirikensis, but is absent from the typical terra and very faintly represented, and of a yellowish tint in the typical obscura. In (3), the umber triangle may be bordered, on the site of the white band of hobleyi and. tirikensis, with whitish yellow, much paler than the rest of hind-mng under surface of terra. Furthermore the transition towards the $ pattern tirikensis in (1) and (3) is shown upon the upper surface by the whitish or whitish grey tint of the pale areas, especially the subapical bar, and, although to a less extent, the inner marginal patch of the fore-wing (Plate XXXIV, fig. 10 ; XXXV, figs. 3,7,8; XXXVI, figs. 14, 15). A shght tendency towards transition between terra and hobleyi is also sometimes seen in an orange 614 Dr. G. D. Hale Carpenter on suffusion at the costal end of the white bar crossing the hind-wing, a tendency which is feebly developed in the specimen figured on Plate XXXV, fig. 12, and is only strongly marked in a single specimen from Bugalla (Plate XXXIV, fig. 9, No. 57 on the hst). This interest- ing example is a male with fore-wings like the typical hobleyi, but hind-wings above of the terra form. Below, the hind-wings show the umber triangle of hobleyi well developed. There is little doubt that this specimen is a blend of terra and hobleyi, but, as regards the former examples, with sUght orange suffusion, it must be remem- bered that the (^ PI. macarista itself often exhibits the same coloration. Indeed, in W. Uganda, Mr. Neave collected 2 examples of PI. psevdeuryta, Hew;, with the pattern of macarista, but the hind- wing bar on the upper surface entirely orange ; and one of these was accompanied by a (^ hobleyi with the same colouring. It is therefore probable that the forms here referred to are a mimetic modification of the (^ hobleyi. Class (2), the intermediates between obscura and terra, form a far more perfect transitional series. Commencing with a terra which shows merely a slight dusky suffusion at the margins of the orange areas, and a little dark colour along the nervures, one can trace the gradual increase of the obscura dark colour until one reaches a point midway between the two forms (e. g. Plate XXXVI, fig. 17) ; beyond this the terra colour is more and more swamped until one gets to specimens of obscura showing only a sprinkling with orange scales on the inner margin of the fore-wing. S. A. Neave's type of obscura, in the Hope Department, is really one of these intermediate forms. What may be considered the real obscura has no orange colouring on the upper surface, and it is a much better mimic of its model, Planema epaea paragea. The commonest form, of all those on the island, is terra, the least common, obscura. The latter appears to be the least stable : it is, in fact, quite difficult to find one which shows no transition towards terra, hobleyi or tirikensis, and even those not transitional exhibit considerable variation. On the other hand, the forms hobleyi and tirikensis appear to be the most stable : they are very true to type and show extraordinarily httle variation. It has already been shown that they very strongly impress their most characteristic feature, the umber basal triangle, on the hind-wings of Pseudacraea eiirytus hobleyi. 615 both terra and obscura, but it is almost impossible to find a specimen which one could describe as hobleyi or tirikensis influenced by terra or obscura. The specimen mentioned on p. 614 (Plate XXXIV, fig. 9), with fore-wings of hobleyi pattern and hind-wings of terra pattern, is the only exception to this which I have caught on Bugalla, out of the 356 Pseudacraeas. It has been pointed out on p. 614 that the ^ hobleyi with an orange suffusion on the hind- wing are probably mimetic rather than transitional. I would suggest that, in Uganda at any rate, hobleyi and tirikensis are the most stable forms, and from them the others have been developed, namely terra and obscura. The extraordinary number of transitional forms on Bugalla Island contrasts markedly with their scarcity on the mainland. In the very large collection presented by Mr, C. A. Wiggins to the Hope Department, which has been made in the neighbourhood of Entebbe on the mainland shore of the lake, only 25 miles or so to the N.E. of Bugalla Isle, there are relatively very few transitional specimens, and three out of the four mimetic patterns, viz. hobleyi, tirikensis, and terra, seem to keep very true to type. An account of the transitional forms observed in an examina- tion of the 1909 material from Entebbe is pubhshed in " I. Congr. Internat. d'Ent.," 1910, vol. ii, p. 497. Among them was a form somewhat similar to that represented on Plate XXXIV, fig. 9, but much nearer to terra than this Bugalla specimen. Obscura appears to be an exception and to be variable on the mainland, but this form seems to be rare in the neighbourhood of Entebbe, and Mr. Wiggins' collection contains only a few specimens. Mr. Neave's much longer series from many localities in Uganda show great variety. The explanation of this relative variabiUty of the forms of P. eurytus hobleyi on Bugalla, and on Damba too, seems to be as follows : — The various Planema models which abound on the main- land, are relatively extremely scarce on these islands. The figures for Bugalla and for a part of the Wiggins collection have been given on p. 607, and it was also pointed out on p. 611 that the number of Plane?na epaea paragea was not a correct measure of their true relative abundance. I beheve this scarcity on the island is due to scarcity of food-plant. I know the food-plants of both macarista and poggei — creepers which I never saw at all on Bugalla Island. 616 Dr. G. D. Hale Carpenter on Now on the island it is quite conceivable that an enemy of the Pseudacraeas might never see a Planema at all : at any rate the latter are so extremely scarce that they can have little protective value, and the Pseudacraeas would gain little by resembhng models that are much less common than themselves. Consequently any form of Pseudacraea that is produced will have as much chance of surviving as the most perfect mimic, and the transitional forms appear almost as abundantly as the types. On the mainland, however, conditions are very different. Owing to the abundance of Planemas, their presence is of definite pro- tective value to the Pseudacraeas, and varieties that are produced which do not conform rigidly to the types of the models are put at a disadvantage in the struggle for exist- ence, and are destroyed by enemies in preference to the types. On the mainland the mimics are kept rigidly up to the mark, and transitional varieties between hobleyi, tirihensis and terra are by comparison rarely to be found. It may perhaps be argued that there is some condition productive of greater variabihty on the island, but not on the mainland. But though intermediate varieties are scarce on the mainland, yet they do occur, and it is difficult not to beheve that they are rarely caught by collectors because they are so much more destroyed by enemies than are those which more closely resemble the models. If, as I beheve, this explanation be the correct one, it supplies the strongest possible proof of the reality of mimicry and of the power of natural selection to preserve it — indeed it is a crucial test. Localities referred to in the following Tables. Bugalla is a large island made up of broad northern and southern portions connected by an intermediate and com- paratively narrow section. A narrow arm runs eastward from the northern part, Buninga, and meets at a right angle a less narrow northward extension from the southern part. At the angle of Kerinya, as this isthmus is called, and near its N.E. shore, my camp was situated on a forest- ringed grassy hill about 150 ft. above lake level. The place is known as Lutoboka or Fort Stanley. Kerinya itself is bordered right down to the shore with forest, behind which grassy downs rise to a height of about 350 ft. The forest belt is in some places very narrow, not more Pseudacraea enrytus hobleyi. 617 than 20 yards through. The locaUties indicated by letters in the tables are as follows : — A. A narrow hippopotamus track through the forest belt which is here about 300 yards wide. There were no open spaces in its course. The butterflies were chiefly captured at the two ends. B. Another path to the N.W. of A. The forest is here so narrow that the path is only about 20 yards long. C. The sandy beach at the edge of the forest to the E. of my camp. " On shore." A similar locality to the N. of camp. D. At the landward edge where the forest is replaced by grass near the end of track A. " At edge of forest." These words are used for the continuation of the forest edge N.W. from D to the end of track B. E. The continuation of the forest edge S.E. from the landward end of track A. While all the other localities hitherto mentioned are only a few feet above lake level, the forest edge at E rises south-eastwards up to about 150 ft. 618 Dr. G. D. Hale Carpenter on Psevdacraea eurytus hobleyi. 619 TEANS. ENT. SOC. LOND. 1913. — PART IV. (mAR. 1914) S S 620 Dr. G. D. Hale Carpenter on Pseudacraea eurytus hobleyi. 621 622 Dr. G. D. Hale Carpenter on Pseudacraea eurytus hobleyi. 623 624 Dr. G. D. Hale Carpenter on Pseudacraea eurytus hobleyi. 625 626 Dr. G. D. Hale Carpenter on Pseudacraea eurytus hobleyi. 627 628 Dr. Gr. D. Hale Carpenter on Pseudacraea eurytus hobleyi. 629 630 Dr. G. D. Hale Carpenter on Pseudacraea eurytus hobleyi. 631 632 Dr. G. D. Hale Carpenter on Pseudacraea eurytus hohleyi. 633 634 Dr. G, D. Hale Carpenter on Pseudacraea eurytusjiohleyi. 635 Nymphaline Mimics. Pseuda- craea kuenowi hypo- xantha 6 9 mimics lA Forms of Pseudacraea eurytus hobleyi 9 " poggeoides " mimicking Ia