generally in relation to that of the reproductive system. In a general discussion of neoteny in this family (pp. 289–293), he points out that the extremely “reduced” condition attained in Phthirocoris is due- to an extreme degree of retardation in development of a kind that occurs intraspecifically in other genera of Enicocephalidae, as well as in other families, i e. greater or lesser degrees of brachyptery which are commonly associated with corresponding modification and “reduction” of the pro-notum and usually also of the eyes and ocelli. It further seems that the retention, throughout the Enicocephalidae, of the nymphal type of abdominal scent-gland openings in the adult is an example of a neotenic tendency unique to this family.

There remains the question of the taxonomic relationships of antarcticus and the two related New Zealand species. They cannot now be placed where Breddin and Bergroth relegated the nymphal types of antaicticus, in the genus Enicocephalus Westwood, in its present restricted sense, being immediately excluded by their two-clawed front tarsi and the form of the ventral apophysis of the male pygophor. (These authors, and some others writing at about the same time, used Enicocephalus in a very wide sense, to include a large number of already established genera, though, as Usinger points out (1945, p. 322), Bergroth later changed his views in this respect.) In a group the species of which have become entirely apterous or with reduced wings lacking veins, it is difficult, if not impossible, to relate the species with certainty to genera in which the wing venation is a critical diagnostic feature, as it is in the alate Enicocephalidae On the basis of other structural features, the species concerned appear to come nearest to the tribe Systelloderini Jeannel, but on these grounds could, it seems to the writer, be placed with equal justification in either of the genera Systelloderes Blanchard or Henschiella Horvath These species have irrevocably diverged as a flightless side-branch with the neotenic characteristics listed above and also as seen in the “reduced” posterior pronotal lobe and scutellum. It seems that, at least with our present knowledge, it would make for an uncertain and possibly unnatural taxonomy to include them in any of the alate genera. Additional and very important evidence for the retention of Phthnocons was provided by Jeannel's study (1942) of the male genitalia of antaicticus. The genitalia of this species are certainly strikingly unique, but some of their most distinctive features are not found in the two undoubtedly related New Zealand species, and these particular characters must now be regarded as specific only. However, at least one other remains, as seen below, which might be considered to support generic or even tribal separation For these reasons the author has followed other recent workers, including Womersley (1937), Jeannel (1942), Usinger (1946), Poisson (1951), and Gourlay (1952), in accepting Phthirocoris Enderlein as a valid genus. The redescription of the genus, given above, is necessitated by the incorpotation of two additional species since the descriptions of Enderlem and Jeannel.

The three known species of Phthirocoris differ very considerably in the form of the ventral apophysis of the male pygophor. The spine-like extension of the apophysis and the long epiphallic process of the median lobe of the pseudosternite above it, as described and figured by Jeannel (1942) for antarcticus, are entirely lacking in the two New Zealand species. These two peculiar structures of Ph antarcticus, which were important diagnostic features of Jeannel's monotypic tribe Phthirocorini, must now be regarded as of specific value only. But despite the very different form of the apophysis, in all three species it is unperforated, whereas, according to Jeannel, in all other members of the Enicocephalinae (including the Systelloderini) it is perforated to act as a “guide” for the aedeagus. The unperforated condition of the apophysis might thus be considered strong grounds for the tribal separation of the Phthirocorini.

There are two possibilities for the origin in this group of the imperforate apophysis, which could be either secondary. due to closure of a Systelloderes-type apophysis, or primitive, being a character retained from the common ancestral stock

of the Enicocephalinae, before the evolution of the apophysis as a “guide”. The writer agrees with Jeannel (1942, p. 284) in considering the former the more probable. On present evidence it seems we may best follow Jeannel in considering the Phthirocorini as a separate tribe having closest affinities with the Systelloderini. In Phthirocoris mirabilis the general form of the male genitalia is very close to that in Systelloderes, apart from the imperforate apophysis. The more primitive Aenicto-pechys also have an unperforated apophysis, but this is of a very distinctive and otherwise specialised form and apparently functions as a sort of sucker during copulation (Jeannel, 1942, p. 284; Fig. 23 e, f). The possession of true harpagones separate this and allied genera as a distinct subfamily; the other known genera have the apophysis reduced or absent. Since the evidence suggests an unperforated apophysis to be a primitive condition in the family, and since it would seem rather unlikely that the specialised function as a “guide” for the aedeagus, once acquired, would be subsequently lost, we may apparently regard the Phthirocorini, at least in this respect, as a relatively primitive section of the Enicocephalinae. They are perhaps best considered as neotenic off-shoots from a winged stock that was derived from the Aenictopechinae and retained the unperforated apophysis of that subfamily but had already lost the gonapophyses, and that gave rise in turn to the Systelloderini and the Enicocephalini. An unperforated apophysis is adaptable to a wide range of morphological and functional variation, including both hypertrophy and reduction or loss; such variation is well displayed among the known species of Aenictopechinae and Phthirocorini. But variation of a perforated apophysis, committed to act as a guide, is restricted to details of shape; it is this difference in shape that has been used by Jeannel as a major criterion in separating the Systelloderini and Enicocephalini.

The neotenic features of the Phthirocorini, already discussed, while by themselves scarcely sufficient to separate a tribe, supplement the genitalia as tribal characters. The Phthirocorini are the only known members of the subfamily Enicocephalinae in which such an extreme degree of neoteny exists.

According to Jeannel (1942), Ph. antarcticus has no trace of a connexivum, but its absence may well depend on the condition of the abdomen of the specimens examined. In both New Zealand species the presence or absence of a visible connexivum depends on the degree of distension of the abdomen, a condition that, as Jeannel has remarked, holds through the other genera of Enicocephalidae. This may be associated with the blood-sucking habits of these bugs.

The type species, Ph. antarcticus Enderlein, 1904, was described from the Crozet Islands. Jeannel (1942), and Gourlay (1952) in describing Ph. mirabilis from New Zealand, pointed out that additional species might well be expected from the islands in and near the Antarctic area. The discovery of the new form from the South Island of New Zealand and the Auckland Islands would seem to be only a further step towards filling in the probable distributional gaps foretold by these writers.

Phthirocoris mirabilis Gourlay. Figs. 21, 22.

Phthirocoris mirabils Gourlay, 1952, Trans. R. Soc. N.Z., 79 (3–4): 363; pls. 68–69.

Numbers of adult specimens from a wide range of localities are to hand which are considerably smaller than the type material, but which are similar in all details of structure (including, so far as can be seen, the form of the male genitalia). A large series of measurements of all major sclerotised parts show no significant differences in linear ratios. There is, moreover, no question of two size groups being involved; there is a continuous range from smallest to largest. All these specimens are therefore considered as belonging to mirabilis, and the measurements given below for this species include those of the smaller adults.

The following additional specific features may be noted, in which this species differs from antarcticus and magnus:

Eyes. The pigment spots of the two ommatidia are usually closely apposed in the adult (occasionally separated), the anterior one larger than the posterior.

Rostrum. Segment II about half as long again as I; III about twice as long as II and about half as long again as IV.

Pionotum. Greatest width (near middle) about half as much again as across anterior lobe. Legs Spination as described for magnus. Pecten of anterior tibia with about 24 fine hairs. Length Exclusive of rostrum, ♂ 2.4–3.4 mm; ♀ 2.5–3.9 mm Including rostrum, ♂ 2.8–4.0 mm; ♀ 2.9–4.5 mm.

Width of head across eyes: ♂, 0.21–0.30 mm.

Ternalia. ♂ : 9th abdominal segment a capsule-like pygophoi open, not extensively, postero-dorsally; no epiphallic spine. Ventral apophysis plate-like, transverse, imperforate, not acurminately produced, upper (apical) margin broadly convex, reflected forward Pseudo-sternite with margins much thickened, sclerotised. convoluted, divided into a posterior median.

Fig 21—Phthirocoris mirabihs Gouilay a, head, lateral(paratype ♂. slide specimen); b, antenna of nymph last instar (v): c, antenna of nymph probaly penultimate mstan(IV).

Fig 22—Phthirocoris mirabihs Gouilay a. terminalia of ♂, postero-ventral, b, terminlia of ♂, dorsal (holotype, slide specimen); c, pygophor of ♂, dorso-lateral (paratype, slide specimen) a, anal tube. ap, ventral apophysis, l.p. lateral plate, ps l. ps m. lateral and median lobes of pseudosternite pyg, pygophor, VIII. 8th abdominal tergum 8.8th abdominal sternum

and two lateral lobes. A pair of plate-like sclerites are present at the sides of the anal perforation and plate, separated from the walls of the pygophor by well-marked sutures; it seems that these plates should be interpreted as the basal portions of the pseudosternite, and if so differ from the pseudosternite of other Enicocephalids, so far as known, in bearing a group of hairs (Fig. 22). However, it seems possible that the plates are later sclerotised additions to the pseudosternite; they occupy similar relative positions to those of entirely membranous and hairless areas in Ph magnus (Figs. 25–27).

♀: 8th abdominal sternum (subgenital plate) with a single median tubercle-like process near the convex posterior margin (in this, resembling magnus and differing from antarcticus)

Associated with the neotenic development of this species is a variation in the degree of resemblance between adult and last instar nymph in certain characters In the adult the two pigment spots of the eye are usually contiguous but clearly distinguishable; rarely the two spots are fused so that their limits are scarcely to be determined; occasionally the nymph-like condition is retained of two quite separate spots. In the adult antennae, a considerable gradation occurs between relatively long and slender and relatively short and stout segments. In Ph. antarcticus the antennal segments of the adult, as figured by Enderlein and by Jeannel, are much shorter and more nymph-like than in any adult specimens of the two New Zealand species. The most reliable characters for separating nymphs from adults are the single-segmented middle and hind tarsi and the absence of the adult type terminalia.

The known range of Ph. mirabilis is now considerably extended to include West-land, northern Marlborough, and north and west Nelson in the South Island, and the south-west part of Auckland province in the North Island.

Ovarian Eggs. Oval, length ca 0.45, width ca 0.32 mm.

Nymphs. The nymphal instars of Phthirocoris can be distinguished from nymphs of other New Zealand Enicocephalids by the general structural resemblance to their adults. From the later instars of the New Zealand species of Systelloderes and Maoristolus they differ conspicuously in their smaller size, paler colour, absence of wing rudiments and well developed scutellum, shape of pronotum, spination of leg, and the smaller eyes with fewer ommatidia. The last two characters also separate them from the earlier instars of these species, which in addition are larger than the earlier instars of Phthirocoris. The nymphs of Nymphocoris maoricus are unknown, but the spination of the legs and the form of the rostrum of this species are very different from those of Phthirocoris. The eye of at least the last two instars of Phthirocoris has two ommatidia (against only one ommatidium even in adults of Nymphocoris).

Nymphs of Phthirocoris mirabilis differ from their adults in the following characters : tarsi of middle and hind legs of only one segment, ♂ without sclerotised pygophor, apophysis and pseudosternite, ♀ with the last (eighth) abdominal sternum not darkened and relatively sclerotised, concolorous with the others; integument softer and paler; antennal segments generally shorter in relation to their width, especially segment III (in which ratio of length to greatest diameter is 4—5 in fifth instar nymphs, 3 in earlier instar (probably fourth), 6–8 in adults); antennal segments with basal subsegments not or only incompletely separated; antennal segment IV longer relative to III, especially in earlier instars (subequal in adult, IV about ¼ longer in fifth instar, and about ½ as long again as III in (?) fourth instar), the two ommatidial pigment spots usually separate (occasionally contiguous in the last instar); width across eyes in specimens examined: instar V, 0.21–0.24 mm, instar (?) IV, 017 mm The spination of the legs in at least the last two instars is similar to that of the adult.

Because of the great size range in Ph. mirabilis and the absence of wing pads in the nymphs, it is difficult without rearing and with a small number of specimens to determine accurately the nymphal instars. In the following list of specimens the probable instars are included in brackets. Those noted as last (fifth) instar are almost certainly such, having head-capsule widths within the adult range.

Range of head-capsule width most probably decreases, as usual, in the earlier instars; this measurement and the antennal ratios seem the characters most likely to be useful in distinguishing the nymphal instars; in the earhei stages it is possible reduction will be found in the number of pectinal hairs.

Specimens Examined (from leaf mould except where otherwise stated): Holotype ♂, paratype ♂ and ♀, 2 paratype nymphs, Upper Maitai R. Valley, Nelson, S.I., 3.4.1950, E. S. Gourlay (C I), (see Gourlay, 1952) 1 ♂, 1 ♀, L. Paringa, West-land, S.I., 23.10.1950, G. P. Hughson (C M.) 2 nymphs (V), Camerons, Westland, S.I., 5.9.1950, R. Chapman (C M.) 1♂, 1 ♀, L. Mahinapua, Westland,. SI, 6.1.1951, R. Jacobs (CM) 4♂, 1 nymph (? IV), Okarito, Westland, S.I. (moss),

26.4.1951, R. R. Forster (C.M.). 1 ♀, Haast R. (S. bank), S. Westland, S.I., 31.12.1951, I. Wheeler (C.M.). 1 nymph (V), Bruce Bay, 10.1.1954, W. Clark (C.M.). 2 ♀, Leslie Valley Track (beech forest), 23.1.1948, R. R. Forster (C.M.). 1 nymph (V), Mt. Arthur Track, Nelson, S.I (moss and lichens, 3,400ft), 22.1.1948, R. R. Forster (D M.) 1 nymph (V), Salisbury's Opening, Mt. Arthur Tableland, Nelson, S.I. (moss and lichens), 23.1.1948, J. T. Salmon (D.M.). 1 ♀, Canaan Track, Nelson, S.I., 25 10.1948, R. A. Cumber. 1 ♂, 1 ♀, Oparara, W. Nelson, S.I., 18–20.1.50, R. R. Forster (C.M.). 1 ♀, Moana, L. Brunner, S.W. Nelson, S.I., 10.3.1950, R. R. Forster (C.M.). 1 ♂, Bluemine I., Queen Charlotte Sound, Marl-borough, S. I, 11.9 1948, J. T. Salmon (D.M.) 2 ♂, 1 ♀, 2 nymphs (V and? IV), Pelorus Bridge, Marlborough, S. I. (ca. 900ft), 17 12.1951, R. Pilgrim (C.M.) 1 ♂, 2 nymphs (V), near Waitomo, S.W. Auckland, N.I., 4.12.1947, A. J. Healy (D.M.). 5 nymphs (V), Taumatatotara, Kawhia County, S.W. Auckland, N.I., 5.12.1947, A. J. Healy (D.M.) 1 ♂, 1 nymph (V), Oparau R, Kawhia County, S. W. Auckland, N. I., 6.12 1947, A. J. Healy (DM) 2 nymphs (V), Pakoka R, Kawhia R., S.W Auckland, N.I., 11 12 1947, A. J. Healy (DM.). 1 ♂, Raglan, S.W. Auckland, N.I., 13.12.1947, A. J. Healy (D.M.). 1 ♂, 1 ♀, Waiharakeke, Kawhia Hbr., S.W Auckland, N. I. (Weinmannia forest), 1.1951, J. W. Ronaldson.

Phthirocoris magnus sp. nov. Figs. 23–31.

Surface shining, with a covering of fine, pale brown hairs, impunctate except for obsolete punctures on posterior pronotal lobe and sometimes a few scattered elsewhere on thorax.

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

Head (excluding “neck” and rostrum) ¼-⅕ as long again as pronotum; anterior lobe 14/5 to twice as long as wide across eyes and equal or very nearly equal in width to posterior lobe. about twice as long as posterior lobe (39:20); posterior lobe separated by strong transrerse constrictions from both anterior lobe and “neck”, rather elongate subglobular, but, owing to concavity of anterior constriction, median length only equal or closely subequal to width, dorsal surface moderately convex, sides weakly convex, widest near basal third, ventrally gradually deepening towards base, where considerably swollen beyond level of “neck”. Eyes small, each with four relatively large ommatidia (represented by black pigment spots), two anterior ones closely opposed and two posterior ones which are often more separated, some times a smaller fifth ommatidium below the anterior pair. Ocelli absent. Antennae subequal in length to head and pronotum together, with a covering of erect and suberect hairs longer than width of the segments; first stout, narrowed at base; second and third subcylindrical, thickening towards apex, subequal in width, fourth fusiform, about as wide in middle as first; relative length of segments I-IV, 13:25.25:24 Rostrum with third segment moderately stout and only moderately incrassated below; relative length of segments I-IV, 8:10:20:11; all segments with fine pale hairs.

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

Pronotum with anterior margin very shallowly concave, posterior margin nearly straight; posterior constriction better defined than in antarcticus or mirabilis and posterior lobe longer and more distinct, its sides nearly straight, middle lobe with sides sinuately convex, posteriorly gently incurved, anteriorly forming broadly rounded shoulders; measured from these shoulders, middle lobe about twice as long as anterior lobe and over 2½ times posterior lobe (27:13:10), middle lobe 2/3 as wide again as anterior lobe (46 28) and 1/101/6 as wide again as posterior lobe, middle lobe with sublateral foveae distinct and median longitudinal incision deep, the latter behind with two diverging shallower arms demarcating a median triangular area before posterior lobe, in front median incision extending more shallowly on to anterior lobe, where it separates a pair of moderately raised transverse callosities occupying most of dorsum. Mesonotum with base depressed and finely granular, posteriorly convexly raised and shining, widest behind, posterior angles obtusely rounded, sides straight, converging anteriorly Scutellum short, finely granular, not at all raised, transversely subtriangular, with base concave, apex obscurely delimited. Metathorax, as seen from above, widest anteriorly, sides straight, posterior angles acutely rounded, projecting backward. Pronotum nearly ½ as long again as meso- and metanotum together, (50:35) and scarcely (about 1/10) wider across middle lobe than metanotum.

Wings vestigial, fore wings in form of elongate, narrow, pale ribbons about 6 times as long as wide (35:6), widely separated and nearly parallel to each other, applied sublaterally to surface of mesonotum and metanotum, reaching posterior ¼ of the latter, without veins and with a covering of rather long, pale brown hairs.

Legs Relative length of coxa, femur, tibia, tarsus, 25:55:40:13 (fore leg), 28:50.48:20 (hind leg) Front femora, length:depth in middle.: 55:25; hind femora, length.depth at basal ⅓ 50 17 Tibiae, length depth at apex :: 40:16 (front), 48.9 (hind) Hind tibiae

Figs 23–30—Phthirocoris magnus sp. nov Fig. 23.—♀, dorsal. Fig. 24.—Head of ♂, lateral. Fig 25.—Terminal of ♂, lateral. Fig. 26.—Terminalia of ♂, dorsal Fig 27—Terminalia of ♂, posterior Fig 28—♀, apex of distended abdomen, ventral (anal tube exserted) Fig. 29.—Hind leg. Fig. 30.—Front tarsus and apex of tibia, ventral

a, anal tube, an, antennifer; ap.a, ap.l, ap.m, apical, lateral, and median sclerotisations of ventral apophysis; ap mem, membrane of ventral apophysis (in undistended condition); c, connexivim; g, opening for aedeagus; n, “neck” (posterior cephalic collar); o, orifice of scent gland, p, tibial pecten, ps.l, ps.m, lateral and median lobes of pseudosternite; py, pygophor; ta p, anterior subapical process of tarsus (“Sinneskolben” of Enderlein); ta 1, basal segment of tarsus; tr, trochanter; VIII. IX. 8th, 9th abdominal terga; 8. 8th abdominal sternum.

slightly narrower at apex than at middle (9:10) Posterior (outer) claw of fore tarsus as long as, and anterior (inner) claw longer than tarsus (16:13). Claws of hind tarsus subequal, nearly ½ as long as tarsus (9:20). Legs with spination as described for genus; spines of fore tibiae arranged in three transverse rows of 2, 3, and 2, those of middle row longest, the upper most pair shorter, stout, thorn-like, recurved, directed obliquely upward toward ventral surface of tarsus; anterior subapical process of fore tarsi tubercular, rounded.

Abdominal Terminalia. ♀: 9th sternum a large subgenital plate (width:median length :: 50:30), posterior margin broadly convex, with a very small median tubercle-like projection. 9th tergum and posterior margin of 8th tergum with hairs longer than on rest of abdomen. 9th tergum with posterior margin broadly and simply convex; basal width:median length :: 36:11. Anal tube when completely extended a pale cylinder (length:width :: 8:6), when not, appearing as a short hemisphere, projecting from posterior face of 9th tergum above the genital orifice, a transverse slit between 9th tergum and 8th sternum.

♂: 8th sternum with posterior margin, in direct ventral view, nearly straight; curving forward at sides, when segment extended, basal width:median length :: 46:20. 9th segment and posterior part of 8th tergum with posteriorly directed hairs longer than on rest of abdomen, especially long on postero-ventral margins and postero-lateral angles of 9th segment, where they are up to more than twice as long as the 9th sternum. 9th segment forming a rather short, ring-like pygophor widely and deeply open behind. Posterior margin of 9th tergum widely excavated at sides, apically truncate between emarginations; the anal tube pale, sub-cylindrical, projecting behind the truncate margin. Basal width of 9th tergum:median length of tergum:length of anal tube:. 36:10:5. 9th sternum strongly convexly arched from side to side.

Ventral apophysis of pygophor (Fig 25) long (about three times as long as wide), erect; sides nearly parallel, forming a nearly oblong plate with apical margin convex, apex marginally sclerotised and of a rounded arrow-head shape; sides narrowly dark and sclerotised; with a medium sclerotised strut, between this and the margins the apophysis membranous, the membranes capable of considerable distension (Fig. 27). Lateral lobes of pseudosternite (Jeannel's

Fig 31.—Phthirocoris magnus sp. nov, fore leg, anterior (inner) aspect, p, tibial pecten, ta.p, innei tarsal process, tr, trochanter.

terminology) stout, clasper-like, but without hairs, fitting against postero-lateral emargnations of 9th tergum, upper surface of each depressed just before apex, which is truncate, marginally dark and heavily sclerotised, and separated by a deep, narrow cleft from the posterior lobe, the clefts are formed by the abrupt vertical folding of the pseudosternite at these points, the posterior lobe a vertical sclerotised plate, with upper margin broadly concave, set between anal tube and ventral apophysis. Jeannel (1942, p. 284) does not consider the three-lobed structure as the homologue of the claspers (harpagones) found in the related Reduviidae and the primitive Enicocephalid subfamily Aeictopechinae, but comparies it rather with the “pseudosternite” of E. M. Walker in the Orthoptera, which is regarded as an intersegmental “chitnisation”; in addition, in support of this interpretation, Jeannel points out that the pseudosternite in Enicocephainae does not bear hairs (which are present on the claspers of Aenictopechinae). An apparent exception is in Ph. mirabilis (p. 412 and Fig. 22).

In the absence of a perforation in the apophysis (which Jeannel shows as functioning as a “guide” for the aedeagus in most Enicocephalidae) the aedeagus in Phthiracor is presumably extruded between the apophysis and the posterior lobe of the pseudosternite.

Colour. ♀: Head and pronotum dark, more or less ferruginous testaceous, legs and rest of thorax rather paler testaceous, fore legs darker than mid and hind; ommatida black, antennae and rostrum yellowish brown, abdomen pale, duller yellowish or gieyish brown, relatively unsclerotised except for the testaceous first and second terga, the 8th tergum, except at apical margin, the 9th tergum and the 8th sternum.

♂: a shining, more or less ferruginous testaccous, including all abdominal segments except for their paler posterior margins Antennae, rostrum, and middle and hind legs paler testaceous.

Length (excluding rostrum). ♂, 3.3–4.1 mm, ♀, 3.8–4.5 mm.

Length (including rostrum). ♂, 3.9–47 mm, ♀, 4.5–52 mm.

Width across eyes: ♂, 0.26–0.27 mm; ♀, 0.27–029mm.

Width across middle pronotal lobe: ♂, 0.56–0.59 mm, ♀, 060–0.68 mm.

The most striking difference between magnus and the other two known species of Phthirocoris is the possession of the extraordinary ribbon-like vestiges of the fore wings. Other differences from mirabilis are the generally greater size, the larger number of ommatidia, the pronotum wider in the middle in proportion to anterior lobe, the more distinct posterior pronotal lobe, the third rostral segment longer relative to fourth, the darker and more sclerotised 8th and 9th abdominal terga of the ♂, and the differently formed ♀ terminalia, the ventral apophysis of the pygophor being particularly distinctive. From antarcticus, magnus also differs in the relatively longer antennae, the more strongly constricted postocular and anterior pronotal transverse impressions, the deeper median longitudinal impression of the middle lobe of pronotum, with its diverging posterior arms, the larger and more distinct posterior pronotal lobe, the presence of a reduced and poorly defined scutellum, the more heavily pigmented body, the tubercular anterior subapical process of fore tarsi (sense-organ of Enderlein; strongly curved horseshoe shaped in antarcticus), the presence of a single median apical tubercle near posterior margin of 8th sternum of ♀ (a premarginal pair in antarcticus), and the very different ♂ terminalia, particularly the form of the ventral apophysis of the pygophor and the absence of a long, upwardly and forwardly curved epiphallic spine.

Type Specimens. Holotype ♂, allotype ♀, Auckland Is (leaf mould),—.12.1944, E. G. Turbott (D M. coll.). Paratypes. 2 ♀, Musgrave Pen., Auckland Is. (“trunks of rata trees”), 19.4.1947, J. H. Sorensen (D.M.); 2 ♂, 1 ♀, Leslie Valley Track, (leaf mould), 23.1.1948, R. R. Forster (C.M.).

Small form of magnus Two females (D M.) from Musgrave Pen, Auckland Is (coll. J. H. Sorensen, 1941947, from debris, forest floor), apparently represent a small form of this species In size they fall within the range for mirabilis and resemble this species and differ from typical magnus also in having only two omma-tidial spots to each eye, the width across the middle of the pronotum only about half as much again as across the anterior lobe, the posterior pronotal lobe and constriction poorly defined, and the 8th and 9th abdominal terga concolorous with the rest and not more strongly sclerotised. However, all these features are the effects of a greater degree of neoteny correlated with smaller size The specimens are regarded as magnus because of the presence of linear wing rudiments (more or less as for typical magnus on the mesonotum, but becoming narrower and ridge-like on the