of a number of ill-characterised genera but assigned to them only a fraction of the known species”. Burr restricted the genus Anisolabis to six species, of which A. littorea is one, but this has not been followed entirely by later workers, nor is it here.

White's original description of A. littorea is as follows:

“Deep blackish-brown, with fulvous legs; head somewhat triangular, the sides behind the eyes rounded, very deep blackish-brown; labrum, cibarial organs and antennae fulvo-testaceous, two fulvous spots on head, one close to inside of each eye, a short fulvous line on middle of hind part. Antennae with at least nineteen joints, first joint the longest, second very short, third three times the length of second, fourth a little longer than second, the others gradually increasing in length. Prothorax square, fulvous in front, with a short impressed line in the middle. Abdomen widest about the seventh joint, deep blackish-brown, margins slightly fulvous, last segment of the abdomen large, with some wide longitudinal lines above; forceps short, slightly hooked at the end, with two or three sinuations on the inner edge, legs fulvous, tarsi without apparent pads. Apterous.”

Further data may be added to this account. Adults are about 2 7 cm. long, but due to the structure of the abdomen specimens are capable of fairly considerable temporary change in length and there is much variation within the species itself. The forceps of the male are very asymmetrical, the right arm being more incurved than the left, while those of the female are almost symmetrical, the right being just slightly more bent Occasional males are found possessing forceps much less curved than is usual. The head capsule width in the male has a range of 2 64 mm. to 3·51 mm (mean 3 20 mm.) and in the female 2 64 mm. to 4 04 mm. (mean 3 26 mm.). Variation in the number of antennal segments as the result of damage is very great and counts of from 6 to 22 have been made, but the total number of those from 19 to 21 exceeds that of all the others However, the number of annuli in the middle region varies only from 16 to 20. A full consideration of the data from head widths and antennal segment counts is given by Giles (1952).

The colouration of the species is a very variable feature and always shows considerable darkening in mounted specimens, which no doubt White employed. From his characterization of the forceps as “short, slightly hooked at the end …” it would appear that the description is based on a female. His figure of a male shows the left arm of the forceps more incurved than the right, yet all specimens collected by the writer have the right more flexed than the left. Lateral inversion brought about by the lithographic printing process used for White's figures would account for the reversed appearance. Mr. R. B. Benson again says in a letter. “I have examined the type series in the B. M. collection and they agree with what you say in that the right arm of the forceps is more bent than the left.”

Kirby (1891) states that A. littorea “is allied to” Anisolabis xenia Kirby from Norfolk Island, while Caudell (1927) holds that A. littorea “is very near to Anisolabis maritima, in fact probably no more than a variety of that cosmopolitan insect. I have specimens from New Zealand determined as maritima by Burr and by Rehn” The writer has not yet examined A. xenia, but A. littorea is readily distinguishable from A maritima, A littorea is larger than A. maritima and has the posterior edge of the tenth tergum more strongly toothed above each

and late afternoons over a period of some months showed a range of 66°F. to 73°F.

Copulation

This has been observed to take place at any time of the year, usually in the afternoons. The relation between the times of copulation and oviposition will be discussed below (p. 390).

Prior to copulation the male engages in what might be termed a courtship, which usually lasts about an hour and begins by the pair passing their antennae over each other. At frequent intervals the male passes the hind tibia over the dorsal surface of the abdomen of the female and at times they will interlock their forceps. This stage is followed by the former grasping the abdomen of the latter or passing the forceps along the female's upper surface. The male next slowly moves until he is behind and in line with the female, but pointing in the opposite direction, and the two pairs of forceps are then interlocked.

After some time the forceps are disengaged and the male and female simultaneously raise their abdomens from the substratum, the former rotating his through 180° and moving slightly backwards. The ventral surfaces of the two pairs of forceps are now in contact and the male intromittent organ is inserted between the female's subgenital plate and tenth sternites. A slight movement of the abdomen of each copulant ensues, and, if the hinder part of the male's body is not in contact with the ground, it falls within a short time. A narrow portion of the basal section of the aedeagus is usually visible between the apices of the male and female subgenital plates. The pair then remain motionless while in copula (Plate 84, fig. 4) and after fifteen to twenty minutes move apart.

Copulation is resumed again after a short interval. It would appear that repeated copulation (two or three times) with only a very short period between is the usual procedure, but on one occasion a pair were observed to copulate four times in one and a half hours. In the laboratory it was found that copulation occurred quite indiscriminately when several males and females were placed in the one container or when one male was placed separately with different females.

Although some variation in the behaviour pattern does occur, the above account describes the course generally followed.

These observations were made in the laboratory, but there is no reason to believe that the earwigs would act differently in the field. Individuals in the field were taken in copula at various times during the course of the work and it was noticed that they were usually not found along with the dense populations.

Goe (1925), working with F. auricularia, records the same sort of courtship as in A. littorea, but in copulation the male uses the forceps as claspers round the female's abdomen. He also reports that “in pairing earwigs, we often had to change them three or four times before finding two that were congenial.”

Examination of the spermatheca of large numbers of females taken in the field at intervals throughout the year revealed in every case the presence of sperms. The spermathecae were detached from the median oviduct, mounted on a slide in 1 per cent. saline solution and crushed beneath a cover slip. The long, coiled cuticular tube was thus ruptured and sperm extruded at the breaks. The spermatozoa, which are not arranged in bundles, are very long with a

cylindrical head, remain quiescent within the tube and are disposed along its length. In contact with the saline, however, they become active, the tail vibrating slowly. As the solution passes along the tube, those within it are also stimulated to activity.

Females placed in the hot box at any time of the year could be induced to lay a full complement of fertile eggs. Two females taken on 4th March, 1950, and retained at outdoor temperatures until 11th September (when they were put in the hot box) commenced on 4th October to lay a normal number of eggs, all of which were fertile. It is thus probable that some of the sperm is retained in a viable condition in the spermatheca for about eleven months before oviposition. Nutrition of the spermatozoa would appear to be carried out by the secretory epithelium surrounding the cuticular tube of the spermatheca.

Oviposition, Hatching and Parental Care

After 12th March, 1949, no eggs were found, but first instar nymphs were present in large numbers. Unless the late egg clusters had been destroyed, it is apparent from the incubation period (19 to 24 days) that no eggs had been laid for at least three weeks before that date.

In 1950, the first eggs were taken in the field on 23rd January and thereafter irregularly until 2nd April. As the former series hatched in nineteen days, they were most probably laid only two or three days before collection. The latter group failed to hatch in the laboratory, and therefore it cannot be estimated when they were laid. Six females kept in the laboratory at room temperatures started to lay eggs on the following dates: 28th January, 9th, 14th, 18th, 26th February, and 4th March. Dr. Woodward took a female and eggs at Ngakengo Beach under a pile of driftwood on 28th January. 1950.

Although these observations were necessarily and unfortunately restricted to one season's work, confirmed partially by a second, it is apparent that A. littorea lays eggs normally during late summer and most probably only within the period late January to early March. It seems to the writer that the time of egg laying could be changed by unusually high or low summer temperatures. Early in 1951, Mr. R. R. Forster informed the writer that he had observed on 12th December, 1949, at Okarito, Westland, a female brooding over newly hatched nymphs. Mr. Forster pointed out that Okarito has a very warm climate. This may account for the early appearance of the nymphs.

In the field, the eggs are laid in damp depressions under stones or debris, somewhat removed from the large colonies. The cavity is cleared of rubbish, and the floor is always smoothed. The female alone was seen either close beside the low egg pile or actually “brooding” over it. Although the eggs stick together slightly, they are always clean and no sand grains adhere to them. On being disturbed, she will either adopt the characteristic defensive attitude of the species or carry the eggs singly between the tips of the mandibles and maxillae, scattering them over the floor of the depression. This is doubtless done for protective purposes.

The eggs are ovoid in shape when first laid and creamy white in colour. The chorion is thin and transparent, has a dull lustre, and is quite smooth. Measurements taken of fifty recently laid eggs selected at random from field and laboratory material gave the following ranges.

Length: 1·44 mm. to 1·92 mm.

Breadth: 1·28 mm. to 1·52 mm.

High values for the length were not necessarily accompanied by similar values for the breadth. Chopard (1949), dealing with the dimensions of dermapteran eggs says that “chez une grande espéce de Polynesie (Anisolabis littorea White), ils peuvent atteindre 3 millimètres” None of this size, which would have been immediately noticeable, occurred in the large amount of material considered in the course of the present study.

Data on the number of eggs laid by field and laboratory specimens are given in Table I.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Eight females placed in the hot box at intervals during winter laid a full complement of eggs after twenty-one to twenty-six days. High temperature is thus a necessary and probably one of the main factors providing the stimulus for egg production.

It may be pointed out again that A. litorea is a member of an essentially tropical genus. As it breeds in New Zealand in the hottest part of the summer, it may be assumed that the maximum temperatures in this country are required to raise the physiological activity of the female sufficiently to induce oviposition. Also it would appear that very many applications of the summer maximum daily temperature are needed before their cumulative effect induces egg laying. Possibly the species just manages to survive in New Zealand and might not breed during a cold summer.

Two females which were retained in the hot box after the hatching of their eggs laid a second batch (thirteen and twenty-seven in number respectively) about ten weeks after the first, but all these proved to be infertile. The other females which were also kept became so heavily mite infested that it was necessary to kill them. It seems almost certain that in the field there is only one batch of eggs per season, for by the time even early nymphs left the female's care, the air temperature would be so low as to inhibit further oviposition. The writer has no direct evidence that females will lay in a second summer.

The account given below is based on laboratory material.

Each female cleared the wood wool from part of the surface of the sand in the stack dish about a week before oviposition Approximately half the eggs are laid during the first day and the remainder irregularly on the succeeding two or three days. The female heaps them into a low pile and “broods” over them until they hatch. Even when left alone, she moves the pile about from day to day, but this does not appear to be necessary for the hatching of the eggs, because in two cases where the females died a full complement of nymphs later emerged at the normal time. When disturbed, all females reacted similarly to those in the field.

After about twelve to fifteen days the embryo becomes discernible through the chorion and from the fifteenth to the seventeenth day the body segments (including an abdominal series of large size) appear. By the sixteenth to eighteenth day the dorsal vessel is clearly seen pulsating, and after the seventeenth to nineteenth day the eyes become visible as paired black spots. During the remainder of the embryonic period the abdominal appendages disappear, the pulsation of the dorsal vessel becomes less apparent, due to the thickening of the cuticle, and the eye spots enlarge. On the nineteenth to the twenty-fourth day, the nymph finally emerges. At first it is white or very pale cream in colour, but after about two hours it changes through grey to brown. Some variation in the incubation period (19 to 23 days) was observed in those batches raised in the hot box and a variable but generally longer time (21 to 24 days) was taken for the eggs to hatch at room temperature in the laboratory. It seems likely that the extremes of temperatures experienced in the field would lengthen the time taken for the eggs to hatch.

During the final week of incubation the chorion is distorted and under strain as the embryo develops It would appear that rupture of the shell is due to the pressure and movement of the enclosed embryo. Several chorions were examined after the hatching of the nymph and they were all irregularly split.

Young first instar nymphs rely on the female for protection and are “brooded” by her for the greater part of the stage. A female with very young nymphs when disturbed will endeavour to disperse them by gripping them singly in her mouth parts and depositing them a short distance from the main group Later, however, the nymphs become more active and many will move away for short distances, but some seem always to be near or under the female. On danger threatening, the former will either hastily return to the female or cluster together as if for mutual protection This gregariousness and dependence on the female is lost as the life cycle advances, and by the time the second instar is reached she will turn cannibal, even though apparently sufficiently fed.

On four occasions another earwig was placed with a female with eggs or first instar nymphs and in every case it was attacked and killed within a short time.

Parental care is known from a wide range of species of earwigs and was first reported as long ago as 1773 by De Geer, whose account is translated by Lucas (1920) The latter author mentions that de Kerville recorded much of the information on the subject available up to 1907. Parental care by F. auricularia has been dealt with by many authors—Sharp (1901), Jones (1917). Lucas (1920), Goe (1925), Carpenter (1928), Crumb, Eide and Bonn (1941) and Imms (1948). It is also noted by Terry (1905) for E. annulipes and Chelisoches morio, by Lucas (1920) for Labidura riparia and by Hincks (1948) for Prolabia arachidis. Chopard (1949) says that “on a observé les soins maternels chez la plupart des Forficules d'Europe et aussi chez quelques espèces exotiques, un Diplatys de Ceylan, en particulier.” However, Main^* (1927) states that the female of A. maritima does “not brood over her ova” but occasionally moves them about

and “carries the young nymphs between her mandibles into the burrow which she had made.”

The Nymphs

Except for their size, of course, the nymphs (Pl. 84, fig. 1) are strikingly similar to the adults. However, the ecdysial cleavage line of the head is more prominent than on the latter and the young nymphs are much lighter in colour, though later instars darken up considerably. Sexual dimorphism is not exhibited by the immature stages, which possess the male-type abdomen of ten visible terga but the female-type forceps. Caudell (1927) records taking twelve male and two female nymphs at Horuhoru, but it seems to the writer that these so-called nymphs were in effect small adults which occur in all populations of A. littorea. The “two immature females” of Lysaght (1925) are probably in reality nymphs.

Nymphal Growth

A. littorea passes through five nymphal instars during development. This was shown by following the growth in the hot box of a group of nymphs from their hatching from eggs until they reached maturity. Analysis of the head capsule widths and the growth of the antennae of 282 field specimens (Giles, 1952) confirms these breeding experiments.

A wide range of other Dermaptera all have four nymphal instars.^* according to many authors. Jones (1917), Crumb, Eide and Bonn (1941), Lhoste (1942), Henson (1947), and Hincks (1949a) state that F. auricularia has four nymphal stages, but Chapman (1917) holds that the number is six; Lhoste discusses the errors made by the last writer. The following authors also give other species as having this number of immature stages: Lhoste (1942)—Chelidurella acanthopygia; Terry (1905, 1906) and Hincks (1948, 1949a)—Chelisoches morio; Hincks (1948, 1949a)—Euborellia annulipes, Labia currvicauda and Prolabia arachidis.

During the present study no attempt was made to find the duration of each stadium under field conditions. This must be very variable, as development ceases during the winter (p.). With this in mind, it seemed preferable to compare the length of the stages at high and fairly uniform temperatures.

The nymphs are cannibalistic when kept together in a confined space (even though apparently adequately fed) and it was impracticable to keep large numbers in separate containers in the heating device used. As a result, only four were raised to maturity. Table II shows the mean duration of each stadium under the high temperature conditions used.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

It is felt that these figures give a general indication of the length of the life cycle at fairly constant high temperature. The numbers of individuals involved, however, except for the first and second instars (23 and 16 respectively), were too small to give reliable results and the figures can be regarded as only tentative.

Due to the telescopic nature of the abdomen, the lengths of the nymphs are highly variable. In Table III is given the average length, including the forceps, of six specimens of each instar; as far as could be determined the abdomens were in a normal state of distension.

The range of the head capsule widths and of the numbers of antennal segments of each instar are given in Table III from the data of Giles (1952). For a full analysis of the growth of the head and of the antennae, reference should be made to that paper. The body length, if the abdomen is not unduly enlarged or contracted, provides a fairly reliable guide to the determination of an instar, but the most dependable is given by the width of the head. The least adequate key, owing to the very great amount of damage they suffer, is furnished by tile whole antennae, but the number of segments in the middle region is very useful in this respect.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Moulting

In a nymph about to moult, the body becomes distended and wide bands of intersegmental membrane are visible between the sclerites. At the initiation of the moult, the head is deflected until the labium is in contact with the neck and prosternum, while the antennae are stretched along the ventral surface of the abdomen. Splitting of the cuticle then occurs along the cleavage line of the head and thorax, so that the anterior portion of the exuviae is widely open and the frontal apotome (of Snodgrass, 1947) is bent far forward.

The animal then wriggles its way forward out of the old skin. Periods of resting alternate with periods of activity and eventually the nymph crawls away to allow the new skin to harden, which takes about two hours. Cast tracheal linings are drawn forwards as the insect emerges.

There appears to be fairly rapid regeneration of lost members—half of a forceps arm after two moults and part of a tarsus after three. In a few cases the exuviae of a nymph disappeared after moulting, but generally it is not eaten and may be recovered.

The Overwintering of Nymphs and Adults

In the field all stages of nymphs and adults are found together throughout the year (subject to the conditions noted below). Individuals overwinter at whatever stage of the life cycle they reach by the time reduced temperatures

inhibit further growth. The writer has never taken eggs in the field during winter.

It was demonstrated in the course of this study that nymphs do not moult in winter. From mid-March onwards nymphs of all stages were kept under observation at outdoor temperatures. About half a dozen of the same instar were kept in a culture dish and none of them moulted by early September, when, owing to very heavy mite infestation, the nymphs were all killed.

During winter, certain trends were noted in field populations. Whereas first and second instar nymphs were very common during March and April, their numbers declined until from July onwards they were not collected by the writer, and did not reappear until after the next season's eggs had hatched. Occasional specimens, of course, probably survived the winter, but careful search in many localities did not reveal any. This reduction in the numbers of the smallest nymphs is undoubtedly linked with the life habits of the species in circumscribed habitats harbouring numbers of predaceous animals, among which are included the larger nymphs and adults of A. littorea. It seems that, in order to have a reasonable chance of surviving, nymphs must hatch early enough in the year to reach the third instar before a reduction in temperature brings about a cessation of growth.

The writer noticed a distinct tendency for males to be less common in the field than females. In 1949 it became apparent that the number of males diminished noticeably as the winter progressed and by August-September they became scarce and continued so until December. This trend was again noticed in 1950. Owing to the shortage of time, this could not be verified by direct counts and it may even have been a chance characteristic of the collecting localities. Nevertheless, it is interesting to recall that fertilization of the females (with their retention of viable sperm) takes place many months before oviposition and thereafter the presence of males is redundant. In addition, the fertilization appears frequent and indiscriminate and the general activity of the species in dense populations precludes the chance of some females remaining unfertilized. It would be considerably to the species' advantage if the females, during the greater part of the winter, had little competition from the males for the food, shelter, etc., offered by the environment. It seems that this is so. Callan (1941) points out that when the counts of F. auricularia given by Brindley (1912) are summated, “the male : female ratio is approximately 45 : 55 (total count = 31820)”. Dealing with the same species, Jones (1917) states that of the males, “nearly all of them die after copulation, which occurs in the fall”, but Crumb, Eide and Bonn (1941) hold that the males occur “until the last of April or early in May, when they disappear rapidly”. (Both are northern hemisphere observations.) A population study of A. littorea over a period of some years and in different localities would certainly yield much of interest.

The females found during the winter would apparently comprise two types: those too immature to breed before the onset of cold weather and those which have laid eggs the previous season Observations on the ovaries and fat-bodies of a series of females taken from April to October in 1949 showed the presence of both type. The former had ovaries in all stages of development and the fat-body was always large This would seem to indicate, particularly in those cases where the ovaries are large, that sufficient time had elapsed since their reaching