Studies in the Parthenogenesis and Sexuality of Saw-Flies (Tenthredinidæ): II.—A Gynandromorph of Pteronidea (Nematus) ribesii Scop. : with Observations on Gynandromorphism and Saw-fly Sexuality

A number of wild ribesii caterpillars, nearly full-grown, was collected from a gooseberry bush in the moorland locality of Roughside, Co. Durham, on 2nd September 1923. The first yearly brood of adults began to emerge on 31st March 1924, and a female obtained on 7th April was fed on sugar syrup, provided with a twig of gooseberry and placed captive beneath a glass jar in the manner usually adopted in experimentation (Chawner and Peacock, 1923). The 20 eggs obtained parthenogenetically from the female hatched normally, and from them 6 males and an abnormality emerged, 4 males appearing on 14th June, 1 male and the abnormality on 17th June, and 1 male on 21/22 June; 1 larva still remains alive (February 1925).

After it had laid parthenogenetically, and six days after its emergence, the female was paired with a male ribesii. Then, probably because of dull, cold weather, no further laying occurred until three more days had elapsed ; but during the subsequent seven days, brighter weather supervening, 83 sexually-produced eggs were obtained. From these there emerged 23 males and 5 females, all normal. From this experiment arise many biological questions such as egg-laying, viability, sex production and sėx ratio, which are worthy of discussion, but such must be deferred till a more appropriate occasion.

Hence, just sufficient facts are given in order to show the circumstances under and in which the abnormality was obtained. This specimen had 6 parthenogenetically produced brothers, and 23 brothers and 5 sisters sexually produced, all to appearance perfectly normal. No further evidence of sexual abnormality could be obtained by breeding experiments with this generation, as such experimentation was precluded by my departure abroad.

This section may best be introduced by giving the facts in chronological order. The points of most importance have been italicised.

It is a little unfortunate that it is impossible to answer the initial question as to whether the specimen had paired with the brother with which it hatched, but, as will be seen later, no vital issue is thus affected. In regard to its sexual behaviour we see that, first of all, it showed a male reaction in the presence of males. Sexual excitement among a batch of male insects in the absence of females is by no means unusual. I have frequently observed this among male saw-flies as well as among male lice and the same is well known for other insects. At the same time it must be stated that among the hundreds of saw-fly males observed, the writer has never witnessed a male copulation. (Male copulation has once been recorded and is quoted later.) Next, in the presence of a female, the sexual excitement reached the stage at which attempts were made at pairing. These attempts were naturally foredoomed to failure owing to the fly’s lack of male genitalia, but the cause of this physical impotence was not realised at the time of the experiment because the specimen, deliberately, had not been subjected to an exacting examination likely to cause it harm. When a male did copulate with it the physical act to all appearance was perfectly normal. That the pairing was not a quiet one is no argument that the specimen did not behave as a normal female ; such sexual struggles often occur in saw-flies and the like has been observed by the writer in Dysdercus superstitiosus, the West African Red Cotton Stainer Heteropteron. Internal examination revealed the presence of spermatozoa within the spermatheca so the specimen had been inseminated, either from the male with which it was first discovered, or the male with which it was observed to pair, or from both. The bearing of these facts will be discussed later.

After the specimen had been six days under observation it was killed by chloroform and the body severed at the junction of the abdomen and thorax to facilitate fixation by Carnoy’s fluid. After the usual period of fixing and dehydration the parts were retained in 70 per cent, alcohol. The external characters were noted during fixation and careful drawings of the external genitalia in their natural position were made from various aspects. Further drawings were made when the less easily observed features could be inspected after minor dissections. These preliminaries finished, the abdomen was carefully eviscerated so that all important organs, especially the internal genitalia, were in one mass and could be properly fixed and preserved for histological examination. This tedious procedure was adopted as previous experience in the micro-dissection of small and unique specimens had shown the risk of losing, damaging, or displacing the small organs such as testes, the consequent result being that one’s observations were rendered incomplete. Further, with such a sexual abnormality the condition of the gonads demanded careful investigation for any evidence of abnormality. The visceral mass was coloured in toto with safranin to facilitate handling, cut in sections 6 M thick and stained in iron hæmatoxylin using safranin as counter-stain. Practically all the sections have been sketched thus enabling the characters and relationships of the various genitalia to be accurately ascertained.

Of course the process of dissection may have displaced certain of the parts but to what extent it is impossible to state.

The general appearance at first sight was that of a bilateral gynandromorph, the left half in shape and colour being male and the right female, but closer examination revealed several important divergences from a strict “half and half” condition. The drawings and the following table will make this clear. The comparisons have been made with normal insects from experimental stocks and the nomenclature of the segments is Zander’s (1890), which is usually accepted.

Table showing the Detailed External Structure of the Specimen.

A view of the right side of the posterior segments 8 to 10 is typically female in all its details (Plate I., fig. 6), except that the black edge of the moiety of the male tergum is just visible. The left side aspect of the same segments, on the contrary (Plate I., fig. 5), is not distinctively male ; tergite 8 is certainly male and so also is the large posterior sclerite, sternite 9, which is almost certainly the left half of the male hypopygium and forms the left companion and homologue (?) of the female saw-sheath, also sternite 9. Its position, shape, and lack of setæ exclude the possibility of its being a clasper valve. Of less certain homology is the small sclerite near and dorsal to the hypopygial sclerite (Plate I., fig. 5, S8 (?)). From its size and position, most probably it is the small left sclerite of the male sternite 8, this region in the normal male being reduced to two small plates which are usually hidden. Another somewhat doubtful plate is the sclerite which immediately succeeds the male tergite 8 (Plate I., figs. 2 and 5, T9 ♀ (?)). This sclerite extends upwards to meet the female half of tergite 9 in the mid-dorsal line (Plate I., fig. 4) and to it runs the bow of the left saw-guide. It might conceivably be the left representative of the male tergite 9, the tergite of this segment in the male being reduced to two small plates ; it is too large to be typical, however, and, from its size, the configuration of its posterior edge, its relations with the right female half of tergite 9 and with the bow of the left saw-guide, it most probably is the left portion of the female tergite 9 reduced and modified. The external genitalia proper are typically, completely, and exclusively female (Plate I., fig. 2) ; they comprise saws and saw-guides, the only deviation from the normal being that the laterally paired structures do not lie in close apposition along all their length but cross near the distal extremities, the right moieties being dorsal. The anal segment appears typically female. With the possible exception of the small sclerite of the male (?) sternite 8 there are no additive parts.

Generally speaking, therefore, the head is female, the main mass of thorax and abdomen is male on the left side and female on the right, whilst the genital and anal segments, except the half hypopygium, are exclusively female.

Turning to the internal genitalia we find quite a different state of affairs. The gonads are male with efferent ducts (Plate II, fig. 1) and there is present also a female receptaculum seminis (spermatheca) complete with duct, and accessory glands leading to a large reservoir (Plate II., fig. 4). But a most interesting feature is this—whereas the testicular follicles on the left or male side are evidently typical in number, i.e. 8 to 9, those on the right or female side are atypical and fewer in number, namely 3.

The gross anatomy of the male organs, which could only be worked out by reading the series of sections, shows that the 66 follicles seem normal in structure. There seem to be vasa deferentia and seminal vesicles but the latter do not appear typical. These are the parts mesodermal in origin. No accessory glands, usually so conspicuous, have been distinguished. There follow dilated ejaculatory ducts, which in turn lead to the common ductus ejaculatorius, all ectodermal.

The germ cells consist of undifferentiated spermatogonia with a few spermatocytes showing the usual meiotic stages in the proximal region of the follicles and spermatozoa in the distal lumina. The latter condition is abnormal as in the sexually active male, according to the writer’s experience with other species, the spermatozoa are stored in the seminal vesicles. Nurse cells are also present. A detailed cytological consideration of the specimen is deferred for a future paper.

It is important to note that there is no cytological evidence whatever of sex-change, pathological change or bisexuality in the germ cells ; nor is there a trace of ovarian tissue to be found. There certainly co-exists, however, gonadial abnormality with somatic abnormality, as witness the diminution of the amount of male gonad on the female side of the body.

The female genitalia are all of ectodermal origin and consist of a normal receptaculum seminis (spermatheca) complete with its duct, the former containing spermatozoa received from another male.

As to the character and relationship of the genital orifices of the internal male and female organs there is some uncertainty. In eviscerating the specimen so that the external and internal genital parts could each be preserved, in toto but separately, the details of the terminaba of the unpaired male gonoduct and the female spermathecal duct could not be distinguished. But from the external appearance of the region around the genital orifice—typically female—and from the evidence of the sections, it is possible that each separate duct opened separately at a common vent on, or just behind, sternite 8, the normal position of the female orifice. The presence of the left half of the male hypopygium (sternite 9) may indicate, however, that the male opening occupied a position on that sternite.

Other female structures comprise an accessory gland or glands (it is impossible to distinguish the number from the 68 sections), leading to a large, thin-walled reservoir which empties by a duct opening separately behind the genital orifice. The glands and reservoir are full of the same secretion which presents a coagulated appearance and is stained bright pink. These parts are evidently homologous with the glands, etc., associated with the “stings” of bees and ants (Janet, 1898); but in the saw-flies, most probably, serve for the supply of a lubricant for the descending eggs and/or an adhesive for the attachment of eggs to their nidus.^*

Some little discussion is necessary here in order to settle, if possible, whether any of these parts are additive or substitutive and, as a corollary to this, whether there is any correspondence between the external and internal somatic features. Fortunately, an excellent recent paper on homopterous insect genitalia by Hem Singh-Pruthi (1924) is of assistance in these regards, and it would appear that the origins of the internal male genital ducts in the homoptera and the hymenoptera can be homologised. Utilising this work on male insect homologies and the results commonly accepted on the female saw-fly (Zander, 1916), the following tabulation (p. 71) is a convenient start for our argument.

Though not vital to the main subsequent conclusions of this discussion, but largely as a point of interest and to clear it out of the way, the suggestion is here interpolated that the male paired ejaculatory ducts and their accessory glands are homologous with the female unpaired terminal portion of the oviduct (uterus of ant, according to Janet), the sperm duct, and spermatheca.

Turning to the facts relating to the specimen and comparing them with the above table, we see that the female internal genitalia of ectodermal origin, all pertaining to sternite 8, are complete except possibly the lower portion of the oviduct, i.e. the uterus, in the region of the spermathecal duct. Further-more, this condition corresponds with that of the external genitalia.

What, now, is the nature of the paired and unpaired ejaculatory ducts of the gynandromorph ? The former are certainly abnormal in shape and they lack the accessory glands. The questions which arise out of this are: (1) Are they really imperfect male parts ; or (2) are they really the terminal portion (uterus) of the female oviduct inordinately developed ; or (3) are they tissue of neutral character affected more towards maleness. If they come under the heads of (2) or (3) they may be looked upon as substitutive parts, but there is no proof that they should be so classified. The idea that they are essentially male is assisted by the fact of the presence of the little sclerite which (vide supra) we have considered as being, in all probability, a fragment of the male sternite 8 on the left side. Further, the meagre amount of male tissue would preclude the complete development, during pupation, of the ducts and accessory glands pertaining. Hence there would be correspondence of sorts between external and internal members and, moreover, both of these would be additive somatic parts in a specimen whose external and internal somatic genitalia are female.

As already noted, the precise relationships of the terminalia of the single common ejaculatory duct and the spermathecal duct were not determined, though the probability is that this region was “uterine” and belonged to sternite 8. This would mean that there would not be correspondence between the internal structure and the external male sternite 9, to which segment, according to Singh-Pruthi, the single ejaculatory duct of the normal male pertains. If, however, the duct of the gynandromorph did really reach to the male sternite 9 (strictly speaking, the left half in the specimen) it would be additive to the complete female internal genitalia of ectodermal origin. Comparison with normal specimens strongly suggests that the female accessory glands and their reservoir belong to the ninth segment, and therefore are in correspondence with the externalia of this predominantly female segment.

To return to the fundamental matters under this heading it may be concluded (1) that the only parts of the internal genitalia of ectodermal origin which are additive to the general female make-up are probably the abnormally modified paired ejaculatory ducts and the common ejaculatory duct ; (2) that, in all probability, there is a strict correspondence between the external genital segments which are almost entirely female in character and the primitively ectodermal internal genitalia of female nature.

To summarise : the specimen is a gynandromorph which (1) somatically is predominantly female, the female parts being the head, the right half of the thorax, including the fore wing and most probably the right hind wing and legs, the right half of the abdomen, the anal segment, all the external genitalia except the left saw-sheath, the receptaculum seminis with its duct, and the accessory glands with reservoir and duct, while the remaining parts and the left half of the hypopygium, the additive left sclerite of sternite 8 possibly, and the additive ejaculatory ducts, are male ; (2) gonadially, is male with the testes of the left (male) side apparently normal—eight or nine follicles—and that of the right side reduced to three follicles, but both gonads bearing ripe spermatozoa and, apparently, with no seminal vesicles ; there appears to be strict correspondence between the external genitalia and internal genitalia of ectodermal origin, with few additive parts, and there are no male accessory glands.

At first sight an animal which attempted to pair with another of the same kind and then, in turn, accepted copulation from a male, could be regarded as bisexual in its instincts. But, in dealing with insects whose sexual instincts are but imperfectly known, the case is not so simple and in the pre-sumably abnormal behaviour of abnormal specimens care must be taken to make comparisons with what occurs in normal specimens of both sexes. We must therefore first marshal the facts of sexual behaviour in saw-flies before coming to any judgment on that of our specimen.

Male saw-flies demonstrate sexual excitement by half spreading the wings, fluttering them rapidly in this position, and by sidling toward the female. Stimulation is not necessarily due to the presence of the female because the same demonstrations occur when in company with males only. Further, there seems no specially striking sensitiveness, or play of the antennæ or other organs, as is so well known in other insects. If a male manifests sexual excitement in the presence of males only we are forced to suppose that this behaviour is due to one, or combinations of the following stimuli :(a) the males as well as females produce and are sensitive to an erotic ; (b) the general “mass “excitement due to the presence of a number together in a confined space ; (c) the general physiological condition of the male ? Of these we are totally ignorant.

The female in the presence of males may be passive or even, to all appearance, resistant to male overtures. There seems nothing in the nature of reciprocal courtship and pairing is over in about a minute. Alone or in the presence of their own sex the females never betray sexual excitement. The writer’s observations indicated such, and in this connection the following evidence, kindly supplied by authorities on saw-flies, is of interest and importance.

From Miss E. F. Chawner, Lyndhurst, Hampshire.—“ You ask if I have ever seen females betray sexual excitement when in the presence of their own sex only. Frankly I am not sure. I have known females display great excitement under such conditions ; they sometimes fight furiously and will bite off each other’s legs and antennæ : but is this sexual excitement ? I do not feel sure that it is. However, it is the nearest approach to it that I have seen. It usually takes place among the larger species, eg. Tenthredella arcuatus, lívida and Lophyrus pini, but whether it is merely the result of confinement, I cannot say. It does not occur so frequently if the females in question are provided with material in which to lay their eggs, which rather points to exasperation produced by interference with their normal instincts and by confinement under unnatural conditions.”

From Dr E. Enslin, Fiirth, Bavaria (Writer’s translation). — “I have never observed a female saw-fly attempt to copulate with another female. The females certainly usually play a tolerably (ziemlich) passive role in copulation and I consider that the females, if no males are present, proceed without more ado to parthenogenetic egg-laying.”

From Rev. F. D. Morice, Woking, Surrey.—“ The only thing I have noticed about female saw-flies in captivity is that certain kinds seem very quarrelsome and especially are inclined to destroy each other’s antennæ. I never saw in them any indications of a desire to pair, but then the species I have observed are mostly parthenogenetic, and lay eggs almost as soon as they emerge from the pupa-state.”

From Dr Runar Forsius, Helsingfors, Finland.—” I have never observed female saw-flies sexually excited in the absence of the other sex.”

All the evidence of behaviour therefore goes to show that males do give positive signs of sexual excitement while the females do not and the writer considers that, in saw-flies, the female feels merely toleration, if not indifference, towards pairing. This is not infreqüent in insects but it may be speculated that the feature has become accentuated in saw-flies in correlation with their parthenogenetic proclivities.

In order to pair, the male approaches the female sideways and places the posterior of the abdomen beneath that of the female. According to the relative positions of the genitalia copulation may be quickly effected, but at times the male has to draw its genital parts backwards along the venter of the female before success is achieved. Copulation may be quiet, in which case the male and female face away from each other in approximately the same axis ; but just as often there is a vigorous but haphazard fencing of the hind legs of both individuals and a pushing, rather than stroking, of the opposed abdomens by the same limbs. Activity may even be so great that the act resembles a wrestling match and at times the male may be dragged around by its mate.

The gynandromorph, despite its mixed somatic character and the absence of male copulatory organs, evinced all the symptoms of male sexual excitement once in the presence of males and once in the presence of females. That it later accepted a male and that the copulation to all appearance was perfectly normal does not decisively prove, however, that it behaved really as a female. As we have seen, the female in the presence of its own sex, and even in the presence of the male, remains passive. But such a negative sign as passivity in our specimen cannot be accepted as proof of femininity. Further, the activity of legs and abdomen, shown during the act of copulation, do not necessarily prove femininity either because we cannot exclude the possibility that two males would behave similarly in an abnormal male pairing. While the writer has never observed two male saw-flies in coitu one would expect, in such events as a male attempting to pair with another male, or a male actually coupling with another, just such struggling activity as occurred in the case under consideration. From the evidence of behaviour it is therefore an open question as to whether the specimen was female to any degree, and we may have to regard the pairing and the insemination as due to a “fortunate” accident, the consummation of these acts being permitted by the genital morphology rather than by feminine behaviour on the part of the specimen.

The data concerning the sexual behaviour of saw-fly abnor-malities are few and scanty. The writer has record of fifteen such forms, but only in five cases is there any note on their sexual reactions,^* and of these three relate to specimens from the writer’s experiments. The four other examples are as follows :—

1. Abia sericea, L. Cited by Gerstäcker (1872). Male and female parts scattered but specimen partly a dorso-ventral gynandromorph though with a male facies ; externally really predominantly female. Female external genitalia. Gonads female.

   Found in company of males, as well as numerous females, but the receptaculum seminis was empty of sperm. Gerstacker suggests, therefore, that it escaped the attentions of the males by reason of its general male facies.

   Comment.—These observations prove nothing concerning sexual behaviour, though, as will be seen from this paper, Gerstäcker’s suggestion is most probably correct.

2. Macrophya rustica, L. Cited by Krieger (1892). A “criss-cross” gynandromorph mostly ; male sexual genitalia almost perfect ; female external genitalia of the left side only developed.

   Found in copula with a normal male rustica; a true pairing.according to Krieger can hardly have occurred owing to the nature of the external genitalia, “but one similar to that of two male cockchafers which have been observed in copula.”

   Comment.—Again this example proves nothing concerning sexual behaviour for, as will be seen from this paper, the specimen could have copulated in this fashion while reacting as either male or female.

3. Pristiphora pallipes, Lep. Cited by writer (1924). General shape like male ; certain external female characters. Both sets of external genitalia almost perfect. Gonads cytologically male but shaped like rudimentary ovaries ; relatively few spermatozoa. Possibly artificially produced by chemical action on developing egg.

   On being placed with females no male reactions evinced. Impossible to test female reactions as the specimen is thelytokously parthenogenetic, the males being excessively rare. (Writer’s unpublished data.)

   Comment.—The absence of the positive reactions of the male, which, however, in this species do not appear to be so marked as in other bisexual types, shows that the gonads are not the stimuli to sex behaviour. On the other hand its passive behaviour does not prove it had female instincts.

4. Pristiphora pallipes, Lep. (From writer’s unpublished data.) General shape and size like male ; antennæ and legs female. External genitalia female, but distorted, and one large male clasper on the left side. Gonads absent, probably through disease.

   Evinced no sexual reactions.

   Comment.—None necessary.

Whiting’s experiments with hymenopterous gynandromorphs of the parasitic wasp Habrobracon may now be quoted (1923). One specimen with very minute ovaries and aborted testes behaved as a normal female thus showing that fully developed ovaries are not necessary for all normal female reactions. This case, and two others in which no account is given of the gonads, indicate to Whiting that the head is the dominant factor in sexual behaviour. In one other example (no account of gonads) some other centre than the head acted as an inhibitor to male behaviour.

The evidence of gynandromorph behaviour in Drosophila is also of interest and importance. The observations of Sturtevant, Duncan, and Morgan (1919) clearly show that, however mixed the somatic constitution, sexual behaviour was not determined by the gonads. Further, though the number of cases is few, it does not follow that there is any direct correspondence between the somatic makeup and sexual behaviour : e.g. such somatically mixed specimens do not necessarily show both male and female behaviour.

Sufficient has been cited, therefore, to show that there is nothing unexpected or exceptional in the sexual behaviour of our specimen when compared with that of other insect gynandromorphs. The gonads having no influence on sexual behaviour, we must conclude that in our specimen the strongly demonstrated male reaction was still possible with a male soma consisting only of half the thorax and abdomen, half a hypopygium, and sperm ducts, and despite the presence of a female head, female half of thorax and abdomen and female external and internal genitalia. But, when we push the argument further and look for special organs of psycho-sexual significance in this specimen’s male soma, we are baffled, for we do not at present know of any such structures.^*

Therefore, on the above reasoning and from the behaviour evidence already adduced, and bearing especially in mind that the specimen’s head, together with sensory palps and antennæ, were female, we would be justified in concluding that most likely the specimen really did behave in female fashion.

In the previous paper of this series (1914, p. 400) the sex-chromosome constitution of saw-flies is discussed and need not be repeated here. The females are homogametic, their sex-chromosome nature being of the XX order and the usual maturation reduction would yield haploid eggs having one X-chromosotne. These haploid eggs in ribesii and, for that matter, in the many arrhenotokous saw-flies, are capable of parthenogenetic development and become males. This fact, together with the experimental evidence that the specimen was parthenogenetically produced, had only brothers of similar origin and was gonadially masculine, renders it most probable that the gynandromorph was a modified male. Evidently, then, the explanation of its somatic morphology is as follows : the specimen started as a male egg and, during early segmentation, the chromosomes of one of the blastomeres of X constitution behaved irregularly so that, instead of being shared by two daughter cells after mitosis, were retained in the original blastomere, i.e. the blastomere became diploid and therefore female. This diploid blastomere, reproducing its kind, gave rise to the female region of the gynandromorph. As male haploidy and female diploidy are well known, and are constantly being revealed by new work in Hymenoptera, the only principal unusual features in our example are, firstly, the combination of both haploidy and diploidy in one and the same insect of a parthenogenetic arrhenotokous species which yields haploid eggs and, secondly, that a female blastomere, of double the chromosome constitution of the haploid parent ovum, can so successfully reproduce its kind in the “egg” of a parthenogenetic arrhenotokous species.

As no intra- or interspecific crossing took place and no similar abnormalities were produced by the same female, whether asexually or sexually, we may exclude heredity from being responsible for the mitotic disturbance in segmentation. Hence we must look for an environmental etiological factor, but of such nothing is known.

According to present standards the specimen is a gynándro-morph. The criterion of intersexuality is sex-change in the germ tissue and of this nothing has been discovered in the insect. The absence of seminal vesicles on both sides and the quantitative reduction in the amount of testis on the right (female) side in such a somatic abnormality require explanation, and this will be attempted presently after the consideration of the general question of gonadial teratology in gynandromorphs. On this point the Drosophila gynandromorphs (Morgan and Bridges, 1919) are of value, for in no other group of animals is there available so much experimental and observational information. Yet even this accumulation of data is still too small to yield anything but the very general suggestions cited below. The writer has gone carefully through these gynandromorph dossiers and prepared a list of comparable cases, whose detailed pedigrees and histories show no unusual environmental and hereditary features (except the sex-linked characters used as experimental indicators) and whose somatic and gonadial nature have been investigated.*

In all, 65 gynandromorphs are discussed, comprising specimens classified under the above and other headings. Of these cases there are 44 whose gonads were investigated, 11 (25 per cent.) of them proving affected. But as many of these 44 had special circumstances relating to their life-histories, e.g. XXY mothers, ancestors of high non-disjunction, etc., so that fair comparisons cannot be made, they are rigidly excluded from the cases treated above.

The general developmental condition of the gynandromorph sex glands, therefore, could not be attributed to any hereditary or environmental influence which would not operate likewise on normal specimens of the sexes. The generalisation which results is : nearly 27 per cent, of the gynandromorphs were affected in regard to the development of their gonads, a percentage which we may regard as above that found in “normal” specimens of the sexes. Hence we seem justified in suggesting that the gynandromorphism and retardation of gonad development were correlated. Whilst it is recognised, after Meisenheimer and others, that the soma and gonads are self-determining in respect to sex and development, it should nevertheless be remembered that for successful consummation both depend upon the general physiological condition of the body. Both will succeed if the general physiological condition is sufficiently high, one or both will suffer if it be too low. If only one succeeds it will be at the expense of the other and this success would probably depend upon which had the start in development but, should the enfeeblement of the soma be too great, crippling or death would supervene. A gynandromorph, an organism composed of two types of somatic tissue, is bound to be physiologically different from the normal, which is composed of purely one type of soma. Hence, if the under-development of important organs—in this case the gonads—be taken as an index of general physiological weakness, we must conclude from the high percentage of under-developed gonads in Drosophila gynandromorphs that a high percentage of them were physiologically sub-normal. But in the present state of knowledge as to the metabolic levels of the sexes, as well as of the different regions of the individual soma, and of the interactions of these in abnormalities composed of two somas, we can only speculate as to how this physiological depression in gynandromorphs comes to pass.

Applying the above to our saw-fly we suggest, therefore, that the under-development of the right testis was due to a physiological depression which was consequent upon the presence of two kinds of somatic metabolism in one and the same insect. To suggest that the condition of this male gonad on the female somatic side was a positional matter and due to the “milieu” in which it occurred would be to go too far ; such would assume that the Tenthredinidæ behave differently from Lepidoptera and Díptera which, up to now, provide the critical evidence as to the independence of soma and gonad.

For the very species under discussion Siebold (1871 and 1884) proved that whilst it was parthenogenetically male-producing, there did arise occasionally an asexually produced female. The writer’s recent work (unpublished) corroborates that of Siebold. And now this specimen adds another proof to the latter’s thesis, for we may look upon this gynandromorph as a “half-way” or a “connecting” type. Bearing in mind the facts and argument of the previous section, there seems no reason for doubting that a haploid, i.e. male, ovum may become diploid, i.e. female, in response to a stimulus similar to that which altered the chromosome constitution of the haploid soma-producing blasto-meres. Such a diploid egg on development would yield a form purely female.

1. A gynandromorph of Pteronidea (Nematus) ribesii, Scop., was obtained parthenogenetically and has been studied in detail.

2. It was somatically mainly female ; the female parts were the head, the right half of the thorax—including, most probably, the hind wing and legs—the right half of the abdomen, the anal segment probably, all the external genitalia—except the left saw-sheath—the receptaculum seminis and its duct, and the functional accessory glands with reservoir and duct ; the male parts were the left half of the thorax and abdomen, the left half of the hypopygium, the left sclerite of sternite 8 probably, and the paired and common ejaculatory ducts ; gonadially it was male, with the testis of the left (male) side apparently normal—eight or nine follicles—and that of the right (female) side reduced to three follicles, but both gonads bore ripe spermatozoa and had, apparently, no seminal vesicles ; there appeared strict correspondence between the external and internal genitalia of ectodermal origin, with few additive parts, and there were no male accessory glands.

3. Sexual behaviour in normal ribesii, other abnormalities and this gynandromorph is discussed ; males play an active rôle, females a passive one. The gynandromorph, despite the presence of a female head and external genitalia, behaved as a male and almost certainly as a female also ; its external genitalia precluded its pairing with a female ; it was inseminated by another male.

4. The presence of female soma in the gonadially male insect is explained by mitotic irregularity in an early blastomere resulting in the production of diploid cells of XX constitution instead of haploid X cells.

5. The etiological factor remains unknown.

6. The teratology and physiology of gynandromorph Drosophilæ and saw-flies are discussed, and the high percentage of under-developed gonads, in such, is attributed to a general physiological depression consequent upon the presence of two types of soma at different physiological levels.

7. The appearance of this “half-way” specimen affords additional proof that occasional females do appear in parthenogenetic arrhenotokous species.

In addition to those workers whose names and good offices are mentioned in the text the writer’s thanks are due to the British Association for the Advancement of Science for a grant, the Research Committee of Armstrong College for the use of a new Leitz microscope, Professor A. Meek for laboratory facilities, which include the able services of Mr D. C. Geddes, the steward who cut the sections, and Professor J. H. Ashworth, F.R.S., for assistance in obtaining literature.

Sections 6μ thick. All the figures, except 3, drawn in outline with the aid of a Spenser Drawing Apparatus and a Leitz Microscope using its Ocular No. 10 and Objectives 16 mm. (figs. 1 and 4) and 6 a (fig. 2). Details were obtained by using a Leitz 2 mm. oil immersion objective.

Corrections.—In the light of further opportunities for the dissection of Pristiphora pallipes, Lep., particularly of the rare males, and of greater access to literature, I take this, the first opportunity, of correcting certain mistakes in the first contribution to this series. “On the Males and an Intersex-like Specimen of the Parthenogenetic Saw-fly Pristiphora pallipes, Lep.,” Brit. Journ. Exp. Biol., vol. i., April 1924. The corrections, however, do not invalidate the conclusions adduced.

• (1) The descriptive terms “dorsal’’and “ventral,” used in reference to parts of the male genitalia, e.g. Plate I., figs. 4 and 10, should be interchanged in both text and drawings, the reasons being that fig. 4 is really a ventral view and fig. 10 a dorsal. A technical note may not be without service here : unless one is forewarned, and pays especial regard to the matter, it is very easy during dissecting manipulations to rotate, unwittingly, the whole male genital mass through 180° and so deceive oneself as to its proper orientation.

• (2) The term “receptaculum seminis” in text and drawing, e.g. Plate II, fig. 3, should be altered to “accessory gland reservoir.”