1964 Jaltrnal of the Lepidopterists' Saciely 91

THE MAINTENANCE FOR EXPERIMENTAL PURPOSES OF FORM "SULPHUREA" OF NAPI ()

by S. R. BOWDEN 53 Crouch Hall Lane, Redbourne, Herts., E.'J'CLAND

It may be guessed that when in 1909 Mr. H. W. Head (as he related 30 years later) received his "female specimen of P. napi of a pale yellow colour" from the teacher at Tullybeg School in Donegal, its primary importance to him was as a possible source of income; since he was a commercial breeder, he cannot be reproached for that. The bright yellow variety, previously almost unknown, that he ultimately bred from this Irish female, became known as "ab. 'citronea' Frohawk". It has given pleasure to European collectors ever since and for a long time this was the chief function that it fulfilled. However, more recently this form, properly known as "sulphurea" Schoyen, has provided a means of attacking a number of problems, both those that are special to the Pieris napi - bryoniae group of and those of a more general character that can be studied in that group. Its merits for these purposes arise primarily from the certainty with which it can be determined, its simple recessive inheritance and its nevertheless robust constitution. A particular advantage is its genetic control from the locus which also governs the characteristic parti­ coloured wild-type pattern of napi and related butterflies.

EXPERIMENTAL USES OF FORM "SULPHUREA" (1) It may seem obvious that experiments involving the breeding of any species will be carried out most conveniently where that species is native. This is far from being the case. Though the food-plant will certainly be available, and fresh stock can be obtained with little difficulty, in default of quite exceptional precautions there will always be some risk of waifs' entering the experimental stock with gathered food. In some cases this may vitiate the experiment or, if it is unsuspected, even lead to incorrect conclusions. If, however, the whole of the ex­ perimental material can be marked with a bisexual recessive variation, this risk is eliminated: if the stock is homozygous, the intruder is recog­ nised when the adult emerges; if it must be heterozygous, the suspected intruder is detectable only by mating with the corresponding homozygote. ( 2) Successful pairing of a homozygous rece:;sive female with a like or an unlike male is demonstrated by raising her eggs to the ima­ ginal state. This was the means of disproving a statement that a female Pieris napi would pair successfully only once, and of showing that after the second pairing the first became ineffective (Bowden, 1951). Even 92 BOWDEN: Pieris napi VaLlS: no.2 the forced-pairing techniques of Lorkovic (1952) fall short of artificial insemination, but if the latter is ever applied to Pieris we may be sure that "sulphurea" will again prove useful. (3) The present author has used it to supply a "visible" napi gene which could be traced in hybrids with P. bryoniae Ochsenheimer for many generations (Bowden, 1962). Although there are many ban-iers, unsuspected at first, between napi and bryoniae, it appears to be possible to backcross the hybrids indefinitely to bryoniae. As long as the presence of "sulphurea" can be demonstrated, it remains clear that napi has not been completely "bred out" by any automatic selective process. ( 4) Petersen and Tenow (1954) described experiments in which they investigated the supposed sexual preference ban-iers between Pier is napi and bryoniae. They found, rather surprisingly, that P. hryoniae males were more attracted to napi females than to their own, the white colour being apparently the deciding factor. We do not presume to suggest what the result might have been, but we should have liked to see these experi­ ments extended to include females of form "sulphurea". (5) The question of selective pairing may arise, of course, within a species. It has impOltance as one of the mechanisms which may con­ tribute to the maintenance of a polymorphjsm (Williamson, 1958). Pair­ ing selection, if proved to exist, is difficult to interpret: it may be the result of differing visual stimulus, or it may depend on some apparently unrelated invisible character which has become con-elated with the visible one. Assuming that the basis is visual, "sulphurea" would provide good experimental material. ( 6 ) It is quite possihly true that, as long as good "resting" camouflage is retained, white colour is advantageous to Pieris. The disadvantage that we are inclined to associate with recessive inheritance probably lies, in the case of "sulphurea", in the extended yellow pigmentation. The breeder would judge that any invisible correlated disadvantage must be slight indeed. He might be of a very different opinion about the "albino" variety (with decolorized "black" markings) which turns up rather frequently in the Pieridae: here there may well be an accompanying weakness. Quasi-"natural" selection experiments using these two re­ cessives and wild-type might give interesting results. (7) A project that we have toyed with for some time is an attempt to detect "sulphurea" alleles in natural populations of Pieris napi. The heterozygotes of any rare recessive will be velY much less rare than the visible homozygotes; it is the heterozygote therefore that we may hope to find. Though there are several "sulphurea" alleles (Bowden, 1961), Head's "citronea" is perhaps the ultimate recessive. Captive virgin females of this form, mated with wild males collected at random, would 1964 Journal of the LepidoT'terists' Society 93 show by their offspring if any of these "type" males was in fact the heterozygote of a "sulphurea" allele. A further generation would have to be bred only to characterize fully all' allele already detected in this way. Although the best-known British "sulphurea" specimen came from Nor­ folk (Barrett, 1881), and Thompson's pale yellow had apparently a Flintshire origin, it is still the general opinion that Ireland would be the best place to search. Nevertheless, if "sulphurea" is maintained by recurrent mutation, it must surely exist (even if at very low frequency) in Asiatic and American subspecies. ( 8) The dominant white-underside morph "subtalba" Schima which is maintained in balanced polymorphism in Pieris bryoniae neobryoniae Shelj. has proved to be very closely linked (if not indeed allelic) with the recessive "sulphurea" (Bowden, 1963). "Subtallba"-like forms which occur in such distinct subspecies or species as dubiosa Rober or virgin­ iensis Edwards are probably also dominant, but there is no obvious way of determining experimentally whether the responsible genes are identi­ calor homologous. However, successful hybridisation and back-croSSing with napi f. "sulphurea" should allow a decision whether or not they are linked. (9 ) Pieris napi provides exceptionally favourable material for the study of pterin pigments and their genetic control, apparently surpassing in this respect any species of Colias. The sex-limited ochre-yellow of the various bryoniae subspecies is from many points of view of outstanding interest; this is increased by its relation to the lemon-yellow and white pigments controlled by the "sulphurea"-"subtalba" alleles, some of which (but not all) operate independently of sex.

BREEDING FORM "SULPHUREA" A few notes on the maintenance of stocks of this useful and attractive variety of napi may be of some assistance, in spite of their European bias. First, we offer some remarks which apply to any subspecies of P. napi or bryoniae. The larvae may be reared by the usual techniques, which it is not necessary to describe in detail. They are less exacting than most species, and can sometimes even be raised successfully in the dark, with little ventilation, provided that paper or the like is supplied to absorb moisture. Crowding also seems to be harmless in itself, as long as it does not lead to excessive condensation, but of course if disease does appear congestion results in disaster. Pupating abo may be more fre­ quently disturbed by wandering larvae, and the chance of cannibalism is probably rather increased. If it is intended to breed napi continuously, it is unsafe to rely entirely on wild food. In the late summer clean leaves in good condition may 94 BOWDEN: P'ieris napi VoLlS: no.2 not be plentiful on wild plants, just when stocks of larvae are at their highest. In England, the most satisfactory food to grow is Dame's Violet, Hesperis matronalis L.; it has the advantage, too, of possessing decorative, sweet-scented flowers. Very "sappy" leaves are bcst avoided, but the plant has a very long productive season, certainly well into December in normal years, and being cut down will spring up again. The flowering shoots have strong stems and as they last well in water can be put in the laying cages for eggs. Some individual females lay on the flower petals (which drop before the eggs hatch); if this is found to be happening, the flowers and flower-buds must be removed from the shoot for that female. Subject to this same contingency, the flowers when available are a good source of nectar for the caged butterflies. Eggs and larvae of wild P. napi, as well as occasional EuchlOe cardamines L., may be found on H. matronalis; it should therefore be searched before it is put in the breeding cage or used as fodder. More troublesome on H esperis are microlepidoptera, particularly PluteZZa porrectella L. When an infestation is discovered, affected leaves should be destroyed and the plants sprayed twice with derris at a fairly high concentration. After thorough washing by rain or hosing with water, to remove insecticide residues, the plants are soon available again and usually remain free for some time. The most useful wild plant is Jack-in-the-hedge, Alliaria petiolata Bieb., if it is to be found locally. It keeps for some days after picking, in a closed tin, without water. It is not always easy to detect waif larvae on the underside of Alliaria leaves; as well as looking over them carefully in the usual way, one should hold the leaves up to a strong light. In some districts Alliaria is absent or exceedingly scarce; in many such places, however, Lady's Smock, Cardamine pratensis L., can be found in quantity on marshy ground. Cardamine hirsuta L., growing by upland streams and elsewhere, is more difficult to obtain in quantity but is much favoured by the wild females for egg-laying. In the late summer Horseradish, Armoracia rusticana Gaertn., may be resorted to, but it does not keep well in water and it frequently bears larvae of Pieris rapae L., which are not easily seen and may perhaps carry disease. Shoots of Watercress, Nasturtium officina Ie R.Br., are sometimes useful but have so high a water-content that they are difficult to deal with in closed tims. "Shop" Watercress is often too stale. In the breeding cage a Watercress shoot attracts eggs well from laying females, but only too often proves to carry undetected eggs of the Mustard Beetle, Phaedon armoraciae L., whose black larvae proceed to destroy the leaves before the napi hatch. Many other Cruciferae are useful if leaves in good condition are available. In the depth of winter even Brussels Sprouts can be used to feed up a 1964 Journal of the Lepidopterists' Society 95 late brood. We have not tried frozen-packed sprouts! The over-wintering pupae of the napi - bryoniae group can be stored till the following October at l°C, though deterioration is sometimes rapid after September. Cold storage makes the approximate synchron­ ization of emergences a simple matter; moreover batches from the same brood can be withdrawn from cold at different times and used for differ­ ent purposes. Very close synchronization of emergences is possible by a refinement of the cold-storage method, due to Petersen (in litt., 1961). But once imaginal development in the pupa is well under way, a pro­ longed return to low temperature is not usually tolerated well, and crippling results; the males are usually the worse affected. The butterflies pair readily in warm weather in net-sided outdoor cages of 6 cu. ft. or less; larger cages are not as successful. The roof of the cage, while transparent, must be capable of breaking the force of heavy rain; in England, a light wire-reinforced plastic material called Windolite seems most convenient. In cold weather, well illuminated cages indoors can be used with good results, but it is less troll blesome to complete the desired pairings outdoors if possible. Again, although Lorkovic and others have used forced pairings among Pieris, such assistance seldom seems to be required within the napi - bryoniae group. Unless the cages are individually -proof, they should be placed on a wooden plat-

Fig. 1, Outdoor net-sided cages used for pairing Pieris napi. 96 BOWDEN: Pieris napi VoLlS: no.2

form supported off the ground in such a way that earwigs and ants are excluded (Fig. 1). For years the writer has used four cast-concrete supports which he barbarously calls "isolegs": these are shaped like lidless boxes and have a central pillar, also of dense concrete, fixed in the centre of the bottom and projecting two inches above the rim. The space around the pillar, being kept full of water, acts as a defensive moat (it is also much visited by thirsty birds and bees). Since the arrangement raises the platform only eight inches off the ground, the grass must not be allowed to grow too long beneath it. It may be necessary to provide some protection from the curiosity of youthful cats. The butterflies can certainly be fed from sugar-pads and other types of drinking-fountain. But these require regular attention; if it is doubtful whether this can always be given, it is safer to place in the cage jam­ jars of fresh flowers in water. The choice of flowers is governed by what is available, but it is soon clear that some are useless, either because they contain no accessible nectar or because they droop quickly in water. It will probably be necessary to draw on both wild and cultivated flowers. Some of the very best are Dandelion (Taraxacum officirwle agg.), Crepis taraxacifolia Thuill., Sweetwilliam, Lavender, Mint, Golden-rod and Michaelmas Daisy. While eggs are appearing in the cage it is best to take them indoors every few days, so that the larvae on one leaf hatch more or less together. Eggs brought in usually begin to hatch about five days later.

BREEDING SYSTEMS FOR STOCK MAINTENANCE Pieris rwpi seems to be sensitive to inbreeding. In consequence, the breeding procedure used for maintaining a recessive variety has to provide for frequent outcrossing. In general, a brood wholly homozygous for "sulphurea" should not be inbred further without initiating also an outcrossed line which can be taken up later if necessary. The simplest system is (Fig. 2): (1) Outcross to wild-type: offspring will all be white. (2) Inbreed these: one-quarter of offspring will be yellow. ( 3 ) Pair these yellows and discard the whites: offspring will all be yellow. ( 4) Outcross to wild-type, etc. In the diagram s is the "sulphurea" gene, + is its wild-type allele. If J each brood produces 100 butterflies, one rears 300 to obtain 125 yellows. A rather more elaborate system, offering greater security, is' (Fig. 3) : (1) Outcross twice to unrelated wild-type: two separate all­ white broods will result. (2) Inbreed these separately: one-quarter of offspring will be yellow. 1964 Journal of the Lepidopterists' Society 97

( 3 ) Pair (twice) a yellow from (~!) with one from the non-parental white brood (1). (This is possible if, as usually happens, part of a brood emerges at once and part lies over the winter; it can also be done, using over-wintering pupae, by withdrawing them in separate batches from cold stonige. ) Half of offspring will be yellow in each brood. ( 4) Pair yellows of one brood with yellows of the other: off­ spring will all be yellow. (5) Outcross twice to wild-type, etc.

"sulphurea" wild 38 ++ s+, 38,(8+,++)

s Is ------,-_.____ ++

etc. Fig. 2, Diagram of system for outcrossing "sulphurea" line.

wild "sulphurea • wild ++ 88 ++

81+~ -T ( __ ,,_),88 ~8'(8_' __ )

(8+),88 88,(8'..)

++ 88 ++

etc. etc. Fig. 3, Diagram showing more elaborate system for outcri:;sing "sulphurea" line. 98 BOWDEN: Pieris napi Vo1.l8: no.2

If each brood produces 100 butterflies, one rears 700 to obtain 250 yellows. In practice, one reduces the size of those broods which cannot be expected to give yellows, and duplicates all-yellow broods when large numbers are required. It is possible to maintain two recessive "sulphurea" alleles together, as easily as one, if the double heterozygote is acceptable as one of the fon11S. Writing "citronea" as ShSh and Thompson's pale pellow as sPsP and SPSh (Bowden, 1961), the scheme is as in Fig. 4. If by mischance the initial pale yellow is sPsP, its pairing will produce no "citronea". All the off­ spring will be SPSh, which should then be paired with any available "citronea" and the system restarted. One then rears 300 to obtain 150 of the yellow fonns. Two unlinked recessive genes can be maintained together by a scheme such as that in Fig. 5. The yield here is 150 homozygous fonns (including 50 double homozygotes) in 300 insects reared. In breeding of this kind it is important to base the system on forms whose genotype can be identified with certainty. For example, the scheme shown in Fig. 4 would break down if the pale yellow, instead of the "citronea", were outcrossed to wild-type, since the offspring would then be a mixture of indistinguishable Sh+ and sP+. Finally, it must never be forgotten that, however careful one may be, some unforeseen disaster may destroy all one's stock of an insect. If the species is napi, one can do a great deal to guard against the risk, but that may not be sufficient. So it is wise to share stocks with others.

SUMMARY Head's bright yellow variety of Pieris napi is not only beautiful, but also experimentally useful. It has been used in experiments on repeated pairing, as a "visible" napi gene in bryoniae hybrids, and as a means of excluding waifs. It might be used in investigations of selective pairing, both inter- and intra-specific. With the less robust "albino" form, it would be suitable for experiments on quasi-"natural" selection. It could be employed to detect the heterozygotes of other alleles at the "sulphurea" locus. Various napi-group butterflies which lack bright yellow on the underside can be investigated conveniently by hybridizing with f. "sulphurea." Finally, the various "sulphurea" allelomorphs pro­ vide favorable material for the study of genetic control of the pterin pigments. Breeding follows the straightforward techniques applicable to all the green-veined white butterflies, but should be planned in such a way that a wholly homozygous brood is never inbred further, without also outcrossing to wild-type to provide subsequent generations. Two 1964 Jomnal of the Lepidopterists' Society 99

"sulphurea" alleles can be maintained as easily as one, and two unlinked recessives are no more difficult, but in all cases the breeding system should be based on forms of identifiable genotype.

/lei tronea II pale yellow sh sh sP sP wild I ++ I ,/I sh, sF ,/1-7

sh + ------r------/

.h h h I h \r----',.. s , {s" +,sF +),s '~

I ++ ----T""--- sh sh ,sh aP

etc. Fig. 4, Diagram showing system to maintain two recessive "sulphurea" alleles together.

albino "sulphurea' wild ....-___ aa ------r------++ 88 ++

'-____,- ___ a+ 8+

a+ ,(a+ ,aa) ,aa ___...... (Several alternatives 88 8+ 88 8+ at this stage)

etc. Fig. 5, Diagram showing scheme wherein two unlinked recessive genes can be maintained. 100 BOWDEN: PieTis nap; Vo1.l8: no.2

References Barrett, C. G., 1881. Singular variety of Pieris napi. Ent. man. mag. 18: 110. Bowden, S. R., 1951. Results of repeated pairing in Pieris napi: use of a recessive variety as a marker. Entomologist 84: 278 . ... , 1961. Pieris napi L. abo sulphurea Schoyen: experiments with a "new" pale yellow form. Entomologist 94: 221-226...... , ] 962. Ubertragung von Pieris napi-Genen auf Pieris bryoniae durch wiederholte Ruckkreuzung. Zeits. Arbeitsgemeinschaft osterr. Entomologen 14: 12-18 . ... , 1963. Polymorphism in Pieris: forms subtalba Schima and sulphurea Schoyen. Entomologist 96: 77-82. Head, H. W ., 1939. The history of Pieris napi var. citronea. Entomologist 72: 179-180. Lorkovic, Z., 1952. L'accouplement artificiel chez les lepidopteres ...... Trans. 9th into congr. ent. 1: 223-224...... , 1962. The genetics and reproductive isolating mechanisms of the Pieris napi-bryoniae group. Journ. lepid. soc. 16: 5-19, 105-127. Petersen, B., & O. Tenow, 1954. Studien am Rapsweissling und Bergweissling: Isolation und Paarungsbiologie. Zool. bidr. Uppsala 30: 169-198. Williamson, M. H., 1958. Selection, controlling factors and polymorphism. Amer. naturalist 92: 329-335.

INTERNATIONAL COMMISSION ON ZOOLOGICAL NOMENCLATURE - NOTICE OF USE OF PLENARY POWERS In accordance with a decision of the 13th International Congress of Zoology, 1948, public notice is hereby given of the possible use by the International Commission on Zoological Nomenclature of its plenary powers in connection with the following case, full details of which will be found in Bulletin of Zoological Nomenclature, Vol. 21, Part 1, pub­ lished on 25 March 1964. Designation of a type-species for Hypercompe HUbner, [1819] (Insecta, ). Z. N. (S.) 1611. Any zoologist who wishes to comment on this case should do so in writing, and in duplicate, as soon as possible, and in any case before 25 September 1964. Each comment should bear the reference number of the case in question. Comments received early enough will be published in the Bulletin of Zoological Nomenclature. Those received too late for publication will, if received before 25 September 1964, be brought to the attention of the Commission at the time of commencement of voting. All communications on the above subject should be addressed as follows: The Secretary, International Commission on Zoological Nomenclature, c/o British Museum (Natural History), Cromwell Road, LONDON, S. W. 7, ENGLAND W. E. CHINA Acting Secretary to the International Commission on Zoological N omeHdature 1964 Journal of the Lepidopterists' Society 101

AN UNUSUAL ABUNDANCE OF AMATHU:SIA PHIDIPPUS (AMATHUSIIDAE) IN CEBU, PHILIPPINES

by JULIAN N. JUMALON

Univ. of San Carlos, Cebu City, PHILIPjPINES

Few are the hamlets and countrysides in the Philippines where Amathusia phidippus (Linne), a large brown butterfly, is not a familiar sight at twilight. Its hasty, jumpy flight, large size, and habit of gamboling if several are around, sets it apart from other crepuscular fliers such as species of the satyrid genus M elanitis. Likewise, one seldom comes across banana and coconut groves which do not have a colony of a form of phidippus. Cebu's subspecies is Amathusia phidippus pollicaris Fruh. In past years, there was no recorded or reported instance of pronounced abundance of this butterfly, that is, inside of this somewhat limited area, although in February 1954 about twenty pupae were brought indoors by children for hatching, all taken at the area under treatment in this article. In respect to their crepuscular activity, and frequency of encounter with immature stages in recent years, no sign of abnormalcy was noted. The same is true with observations made in parts of western Leyte and the provinces of Agusan, Surigao, and Davao in Mindanao in 1960-62. From the latter part of December 1961 to the latter part of yIarch 1962, this twilight beauty has multiplied in numbers which to perennial ob­ servers is beyond normalcy. The area of infestation in the suburb of Cebu City is less than a square kilometer, and involves the districts of Labangon, Mambaling, and Punta Princesa. There is a possibility that adjacent areas where investigation was not extended, may have received a share of the infestation by the destructive larva of the butterfly. When this was first observed and reported by a team of high school students headed by Osman, the writer's son, sometime in January 1962, the density of the population was already slightly above normal, but subsequent week-end visits to the place revealed a steady crescendo until the peak was reached somewhere between February and March. The increase in numbers especially of males was particularly noteworthy in an area bordering a small slightly watered creek profusely vegetated with a tangle of shrubbery, with a predominance of coconut and bananas, and a sprinkling of betel nut, citrus, and other fruit-trees. Larvae in various stages were noted on all the monocotyledons mentioned. On these visits pupae were collected for indoor hatching. Banana groves left uncleaned by their owners provide good hiding places of imagines, since their tangle of dried leaves affords perfect camouflage background when the insect alights with folded wings in 102 JUMALON: Amathusia abundance VoLlS: no.2 daylight. Here, by shaking tbese clusters of dried leaves, scores of sleepers are flushed out. Usually the insects do not fly far and long. They immediately settle on nearby hideouts, although as a rule most usually return to their favorite hideouts in due time, unless the dis­ turbance is prolonged. At the height of infestation it was easy for two or three boys to pick in few hours well over a hundred specimens. Visits were usually made in the morning at which time the flight of the disturbed sleepers is brief. At twilight, or close to twilight, when they become active, col­ lecting becomes brisk, and the scores of flying dark forms offer good exercise to one's limbs and eyes. On about the last week of February, two boys took 68 fresh specimens in less than two hours in the later part of the morning, collecting at the same time live pupae for indoor hatching. Collecting imagines, however, was limited to those which alighted on low or young coconuts and bananas. Those which sought the tops of full-grown coconuts escaped molestation. On going over the sexes, we found that the males outnumbered the females by five to one. There is considerable uniformity in the stripes and colors of males; on the other hand, the sub-apical light brown area on the dorsum of the females' wings, varies in size and intensity of shade although not to a marked degree. Attack by parasitic Diptera is about 250/0 on all pupae brought indoors. Those which emerged were perfect, beauti­ ful specimens with only about 1% abnormality in one form or another. Also, out of about a hundred specimens, only one female and three males are undersized. (Of the 1954 batch, 90% were parasitized by 2 species of flies.) All the recorded foodplants of this amathusiid form of Cebu are monocotyledons. They are: Cocos nucifera. Linne ( Coconut), Areca catechu Linne (Betel Nut), Corypha elata Roxb. (the Buri Palm - source of sago flour, basic food of Melanesia), Nipa fruticans (Nipa or Sasa Palm, widely used for thatch), and M usa sapientum (Banana). It is likely that other members of this division of angiosperms are also favored by the larva, especially at places where the above-mentioned are less abundant or absent. Individuals were found by our expeditions to Leyte and Mindanao on mountainous areas of not over 1,000 feet elevation; the only possible foodplants present were wild bananas, Abaca, Rattan, and wild species of palms. Among these plants, we also flushed out species of Zeuxidia and Discophora., known feeders on similar or the same plants. Amathusia phidippus is a familiar twilight flier in most countrysides of many provinces here. It is however more numerous in the large coconut plantations where borders are usually lined with bananas. In November 1964 lournal of the Lepidopterists' Society 103

l' UNUSUAl..

Immature stages of the Cebuan form of Amathusia phidipptts, Larva feeding on coconut leaf, and dorsal and lateral views of the pupa. Tbe end of the pupal head is occasionally bifurcate; it is nonnally produced to a single point. of 1952, the writer spent nine days in a sizable coconut plantation near the Maria Cristina Falls in Lanao province. Here, from the eaves and under the floors of stilted huts of tenants, a good number of sleeping adults were collected. Sometimes over a dozen would appropriate dingy corners. Tuba (coconut wine) gatherers frequently find drowned or drunk individuals inside bamboo tubes, together with bees and coconut beetles. In August and September 1961, a similar opportunity was en­ joyed by the writer in southern Davao. For nearly two months, the coco­ nut plantations with their complements of bananas and bamboo thickets provided daily observation of the Mindanao form of this amathusiid. At the American-run Lais Plantation along the Pacific coast at southern Davao, these butterflies favor for their day-hiding large mango trees with spreading low branches. They sleep upside down under branches and on dense clusters of leaves. The Amathusia is subject to attack by predators. In Davao, specifically, the Malita-Talaud-Kinangan sector, is found a goodly population of several species of Robberflies (Asilidae), and I cannot think of another place which is more spiderous. All these take their toll of butterflies and moths, especially the Robberflies which were frequently seen grasping a small butterfly. Crumpled wings of an Amathusia were seen on several webs of spiders, especially one with a strong, sticky web which sometimes can even snare a sunbird. Many specimens collected showed wings neatly clipped by either lizards, 104 SEVASTOPULO: "TOXIC" Lost plants Vo1.l8: no.2

birds, or perhaps tarsiers. Actually, a Hestia which soared up the coconut tree when chased was seen to become entangled in a strong web and was immediately stunned by the spider. Because of their large size, many an Amathusia escaped from the jaws of lizards, suffering at most about a fifth of their wings lost to the predators, which missed the vital part of the body. Uniformity of these bite marks On their wings showed that the attacks were mostly made when the prey were at rest with folded wings.

LEPIDOITERA OVIPOSITING ON PLANTS TOXIC TO LARVAE I can quote two East African examples analogous to those quoted by Mr. Straatman (Journ. lepid. soc. 16: 99-103; 1962) . Charaxes brutus Cr., and its subspecies natalensis Staud., with a considerable number of recorded food-plants, now lays freely on an introduced plant Melia azedarach L. ( Meliaceae ), commonly called Persian Lilac, but such larvae invariably die in their first instar despite the fact that they feed freely. On the other hand I have transferred last instal' larvae from other food-plants to M. azedarach and they have oompleted their metamorphosis successfully. Van Someren records Charaxes lasti Gr. Sm. as laying on Afzelia quanzensis Welw. (Caesalpinaceae), and I have obtained ova freely from captive females caged over the same plant. My larvae have all, however, died in the first instar and Van Someren also states that he has failed to rear the species through. The case of the Australian subspecies of Papilio demoleus L. is most surprising. In India, where I have bred it in large numbers, it has a fairly wide range of food-plants belonging to the Rutaceae, and its near ally P. demodocus Esp. in East Africa is the same. To have diverged from the normal food-plants of the group to the extent that it cannot develop on them seems to indicate a very wide separation from the parent species. I would like to query Mr. Straatman's use of the term "toxic" in this oontext; to me "toxic" implies something active or positive, but my impression is that in these cases the trouble is more passive or negative, the plants in question lacking something essential to the larva's de­ velopment. D. C. SEVASTOPULO, P. O. Box 5026, Mombasa, KENYA