but in the laboratory, from samples of all these plants only Melicytus and Macropiper were eaten. M. filifer is found in a dim twilight from five to fifteen yards, from the mouth of the tunnel, occurring on the walls and roof and often hidden in groups behind out-jutting rocks or in crevices (Plate 31). Usually the young congregate nearer the mouth than the adults. The floor of the cave in the region in which the wetas are found consists of two large pools of stagnant water which maintain a high degree of humidity in the atmosphere. Other animals sharing the habitat with the cave-wetas consist of a few spiders, opiliones, glow-worms, and a member of the Tipulidae.

The tunnel at Percy's Reserve is smaller than that at Karori. On either side of the mouth of the cave are large trees of Melicytus ramiflorus and one bush of Macropiper excelsum which probably form part of the cave-wetas' diet. Here the wetas occur in larger numbers, spread over a greater area, and yet further back in the cave than those at Karori. Unlike the Karori Cave there is a concrete floor, which prevents accumulation of water, and there is no percolation of water through the roof or walls. Although I have visited Percy's Reserve several times during rain, there has never been a trace of moisture inside it, and it is possibly because of this that there is no sign of lichen or moss. However the still atmosphere of the tunnel has a high humidity. The only other inhabitant apart from M. filifer is a smaller species of cave-weta.

It seems that the caves at Percy's Reserve and Karori must have been comparatively recently colonised by Macropathus, which presumably once inhabited the bush surrounding the caves, and became attracted to them because of the ideal conditions they offered for protection, darkness and humidity.

Cave-wetas also occur in a large greywacke cave surrounded by native bush at the side of the Eastbourne Road, Lowry Bay. Again there is no trace of water but the humidity is high. At Days Bay, M. filifer is found in the bush, and, according to Dr. R. A. Falla, he and his neighbours have found it on the walls of their garages and inside their houses. Two specimens of M. filifer from native bush at Pinehaven form another locality. I have been told by several people of a greywacke cave on the Makara-Johnsonville Road which contains a large colony of cave-wetas, but, although I have searched for it, I have been unable to locate it. Other places in the Wellington area visited and examined with no success were the Wainui-o-mata Pipe Tunnel, Koro-Koro, the Ngahauranga Gorge Tunnel and caves in the Tinakori Hills. Darkness, dampness and native bush occurred in each place, but no wetas—only spiders and a few glow-worms.

At Karori M. filifer is known to inhabit the basement of some of the houses if they are sufficiently damp and dark. In one house I discovered a small colony of a mature male and female and about a dozen offspring of various sizes. The outer wall of the basement is damp with a narrow opening at the bottom, which permits the insects to go out into the open whenever they please. They are found on the wooden rafters or hidden behind the loose folds of building paper which line the basement. In the United States of America both Banta (1907) and Hubbell (1936) record Ceuthophilus as occurring in cellars and even inside houses where one species is reported to have chewed holes in some lace curtains.

Spelaeologists exploring limestone caves in the Wairarapa discovered M. filifer at Dannevirke and Ruakokopatuna so, in May, 1952, a collecting trip was arranged to the latter place. The cave visited had a low, narrow tunnel at its mouth which widened out into a larger chamber with high funnels in the roof and,

although nymphs were collected near the floor, it was in the cracks and crevices of these inaccessible places that the mature M. filifer congregated. Blatchley (1920) writing on Ceuthophilus stygius observed that the adults seemed more or less gregarious. He says, “In one instance in Sibert's Well Cave, more than twenty were found in a small cranny in the wall. They were grouped in a circle, in a space about six inches square, with their antennae pointing towards the centre of the circle, and appeared to be holding a conference or cricket convention.” In the Saltpetre Cave they were never seen on the floor, but always on the sides of small projections and in small cavities of the walls or roof. This habit of the adult cavernicolous members of the Rhaphidophoridae to frequent the upper regions in caves seems common in all areas where they occur. Sharing the same habitat with M. filifer at Ruakokopatuna were very large opiliones, several species of spiders, numerous glow-worms and one solitary Peripatus. Running water and several deep pools maintained a constant humidity. Surrounding the cave was typical limestone country with many sinkholes and caves. Very little vegetation occurred except for grass, rotting tree trunks, and one solitary Hoheria Practically no light entered the cave because of the narrow entrance. This was the darkest place in which I have found M. filifer.

A trip to Trio Islands and Stephens Island in April, 1953, threw more light on the habitats of M. filifer. Middle Trio Island is the home of tuataras (Sphenodon punctatus), dove petrels (Pachyptila turtur), diving petrels (Pelecanoides urinatrix), fluttering shearwaters (Puffinus gavia) and mutton birds (Puffinus griseus). These animals all live in burrows, the tuataras during the daytime and the petrels and shearwaters at night. M. filifer was found to inhabit the old damp, unused burrows during the daytime and to come out at night to feed on the leaves and young branches of Melicytus ramiflorus, one of the main components of the flora on the upper part of the island. As there were numerous unused burrows, in many cases the wetas did not congregate in them in large numbers, but each burrow was found to contain a pair and sometimes one or two nymphs as well. However, several burrows have been unearthed containing large colonies living together. At night there was definite segregation among the wetas—only a solitary weta or perhaps a pair were found on each Melicytus tree. Numerous nymphs were seen, but these also were solitary. Searching for wetas was carried out over the whole upper area of the island, and M. filifer was found to occur in every direction, but confined exclusively to Melicytus.

On Stephens Island the habitat of M. filifer is inside large unused concrete water tanks attached to the lighthouse keepers' houses. Lining the walls and roof are hundreds of wetas, most of them in their characteristic attitude of hanging head downwards (Plate 31). Here there is no segregation into adults and nymphs, possibly because of the confined space. On Stephens Island Melicytus ramiflorus and Macropiper excelsum form a part of the main vegetation, and although no cave-wetas were observed on Melicytus trees at night, it is highly probable they feed on it as well as on grass, because in most localities where they both occur, Melicytus forms a part of their diet.

A female of M. filifer was recorded by Chopard (1923) from the bush on Mana Island. To-day Mana Island is almost entirely farmland with very little scrub, and in 1952 when a party landed there no trace could be found of M. filifer. However, in April, 1954, a headless male specimen was found at the southern end of the island at the mouth of a mutton bird burrow, but none were found

inside any of the burrows examined. As M. filifer inhabits the old petrel burrows on Middle Trio Island, it would seem probable that it does the same on Mana Island. The southern end of Mana Island has little vegetation, but a few Melicytus trees do occur at the more northern end of the island, and might form a source of food for the wetas.

At Waitomo M. filifer occurs in large numbers in the Aranui and Ruakuri Caves, but is absent from the Glow-worm Cave. This absence from the Glow-worm Cave may be accounted for because the Cave is closed by a wooden door which would cause conditions of total darkness and prevent the wetas from coming out into the bush at night. The Aranui Cave, however, is closed by a gate which permits a dim light to filter into the anterior portion of the cave, and also enables the wetas to come and go at will. As cave-wetas always occur where there is both a dim light and an opening to the outside world, they evidently form limiting factors in the distribution of the group, which are more important to the wetas than a constant high humidity and a still atmosphere.

In the South Island no records of M. filifer have been taken from either the far south, the east coast or the west coast, but it is possible they do occur in these areas. They have been found in Nelson in native bush near the Maitai and also in a greywacke waterworks tunnel near the Waimere Plains. At Cannibal Cove, in the Marlborough Sounds, they have been collected from among rocks near a stream, and they have also been caught from both Trio Islands and Stephens Island.

In the North Island, apart from the Wellington area, M. filifer has been recorded from limestone caves in two areas in the Wairarapa—Ruakokopatuna and Kaiparoro, and further north from Dannevirke. At Te Mata Peak, Havelock North, there is a large colony living in a limestone cave. In the centre of the Island they occur in a greywacke tunnel near the railway line at Ohakune. The Waikato area is riddled with limestone caves, and it seems highly probable cave-wetas inhabit them. They have been definitely recorded from Waitomo and Te Anga. Caudell (1927) while collecting at Roto Eho, Rotorua, found a medium-sized immature female and a very small male, both of which he doubtfully referred to M. filifer. This is the only record of M. filifer from Rotorua. North of this area M. filifer does not seem to occur.

Summarising these facts: M. filifer is to be found throughout the greater part of New Zealand, either in bush, house basement, greywacke tunnel, limestone cave or burrow. This wide range of habitat shows that the weta when seeking protection is not limited by the actual material of its home. It requires conditions of high humidity and is found in various degrees of darkness, but never total darkness; it occurs in varying abundance, from solitary to several hundred insects in one habitat, and is nocturnal in its habits. Although they may share the same habitat as spiders, opiliones and glow-worms, cave-wetas are frequently found alone, so it would seem there is no dependence upon other animals. As far as is known they do not occur in arid areas.

The description Hubbell (1936) gives of Ceuthophilus gracilipes is very similar to that of Macropathus filifer, showing the correlation between the two groups. He describes the habitat as being “In caves; under bark of dead trees and in hollow trees; under logs and stones; from foundations and cellars. It is characteristically a forest inhabitant which occurs in a variety of environments. Wherever it occurs it takes advantage of such shelter as is afforded by caves.

cellars, hollow logs and other natural cavities. Occasionally it may even penetrate some distance into the larger limestone caves, where it has been found 300ft. from the entrance. The species is neither a true troglobiont nor, apparently, a burrower in the soil.”

An interesting character in the distribution of a species is the geographic variation which sometimes occurs. In some cases among the Ceuthophilinae this is so great that superficially there does not appear to be the remotest resemblance, between the insects, and it is only by examining genitalia or other distinctive taxonomic characters that the true relationship is discovered. In discussing the geographic variation in Ceuthophilus gracilipes, in North America, Hubbell (1936) says, “The species exhibits local and regional differentiation in average size, proportions and colouration. In most instances these differences are of a minor nature and show gradual transitions from one area to the next; thus the average size gradually increases from north to south, and also from the Appalachian region to the Interior Lowland and Mississippi Basin.” With Macropathus filifer specimens from six localities ranging from the Marlborough Sounds area to Havelock North were examined, and it was found that there was a gradual increase in size from north to south Measurements of the length and width of the pronotum, length of the hind femur and hind tibia, and length of the ovipositor were taken and the results analysed. The same gradations in size were observed in each case. (See Table III on Geographic Variation in M. filifer.) The “t” test applied to the length of the pronotum showed that the six localities could be grouped into three pairs, as the tests between Trio and Stephens, Percy's Reserve and Karori, and Ruakokopatuna and Te Mata Peak respectively were non significant. In males from Trio and Stephens Island “t” was 1 16; in males from Percy's Reserve and Karori “t” was 0 03; in females from Ruakokopatuna and Te Mata Peak “t” was 0 16, all these results being non significant However, when any member of these pairs was compared with a member of another pair, the results were highly significant. Females from Trio and Percy's Reserve had a value of “t” of 96.6.

By combining the measurements of Trio and Stephens, Percy's Reserve and Karori, and Ruakokopatuna and Te Mata Peak respectively all “t” tests for both males and females were highly significant.

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

In all cases there was a slight difference in size between males and females. Length of body was not used as a measurement because it was greatly affected by the degree of telescoping of the segments, and attitude of the body.

Poikilothermal terrestrial animals generally attain larger size in warmer climates, and this is considered also to apply to insects. Peterson and Weber (1949) comparing growth ratio and geographic variation of Omocestus viridulus from six Scandinavian localities found that the size of most insects investigated decreased the further north they were collected, and that the relative size of extremities decreased in the same direction. This result is in agreement with

Hubbell's results so it is interesting to find the direct opposite to be the case with Macropathus filifer in New Zealand.

The largest specimens of Macropathus filifer were collected from Trio and Stephens Island. Combination of measurements from Trio and Stephens Island compared with the combined mainland measurements gave a value for “t” of 21.6 in the males, and 28 in the females, both of which were highly significant and agree with the theory that island fauna is larger than that on the mainland. This increase in size is perhaps an evolutionary response to island life with its absence of predators and competitors.

M. acanthocera has been recorded from Titirangi and the Waitakere Ranges, Auckland, and from no other part of New Zealand. It has been collected from a waterworks tunnel, Titirangi, and also from another old waterworks tunnel at Mackie's Rest, Waitakeres. The tunnel at Titirangi has patches of dampness on the walls and ceiling and the wetas are found closely associated with them. In December, 1952, when the tunnel was examined, about fifty wetas were present in groups of ten to twenty, on the walls of the tunnel. Males were in greater numbers than females, there being about five to eight males to one female. No nymphs were present. The only other inhabitant of the tunnel was a Neonetus. The vegetation surrounding the mouth of the tunnel consists of typical Waitakere rain forest, the most common plants being Fuchsia excorticata, Schefflera digitata, Nothopanax arboreum and Coriaria sarmentosa. Melicytus is absent, but is probably replaced by some of these plants in the weta's diet. In May, 1954, I visited the tunnel and found the population was much larger than that recorded for 1952, there being over one hundred and fifty wetas present. As in 1952 there was a larger proportion of mature males to females. More than half of the population consisted of nymphs at various instars, there being a gradation of five instars represented. The very smallest occurred about three yards from the mouth of the tunnel, and they gradually increased in size till the mature insects were found about ten yards within. They were all grouped on the walls and none on the ceiling. There was no water in the tunnel, but the humidity was very high. No specimens of Neonetus were seen, and it is possible that some of the nymphs of M. acanthocera may have been mistaken for it in 1952. Spiders and glow-worms occurred throughout the whole length of the tunnel, and at the far end was a single basidiomycete, the only one I have ever seen in a cave containing Macropathus, although I have found basidiospores in the stomach contents of M. filifer collected from the Karori Cave (Richards, 1954).

In contrast to Titirangi, the tunnel at Mackie's Rest is situated in open scrubby country according to Mr. R. K. Dell, who collected cave-wetas from there in 1949. The floor of the tunnel is filled with water so that a high humidity must be maintained in the atmosphere. Unfortunately I was unable to examine this tunnel, but the wetas collected from there are all M. acanthocera

M. delli has been recorded only from the southern part of the South Island It was first collected on the New Zealand-American Fiordland Expedition, 1949, at Stillwater River Base Camp, Caswell Sound. The wetas were found in the warm, moist interior of a rotten tree trunk, near a stream, and were the only occupants. A mature male and female, two immature males and two immature females were collected. Another mature female was found inside a rotten log. No Melicytus occurred in this locality, but Macropiper excelsum was common and may have formed part of their food. In January, 1953, on the Canterbury

Museum Fiordland Expedition, another mature female of this species was collected from the inside of a rotten log in the bush at Lake Te Au, near the southern arm of Lake Te Anau. Large numbers of this species are reported to occur in the limestone caves on the west of Lake Te Anau, but so far I have been unable to obtain any material for examination to confirm this.

Animals found in caves have been grouped in three categories—those which lost their power of reproduction in a cave environment (troglosseni), those which retained this power and could still live in daylight (troglophili), and those which were compelled to spend their whole existence underground, from birth to death, and could not survive in daylight (troglobi). It is important to remember that the troglosseni only arrive in caves by accident, whereas the troglophili actively seek out and prefer the underground dark. It is obvious from this that the Rhaphidophoridae belong to the troglophili. Typical cavernicolous animals commonly exhibit certain well marked characters. They are very often blind or have eyes much reduced in size or efficiency and frequently they are colourless. All these conditions to a greater or lesser degree are to be found in the family. Dolichopoda bolivari from Europe is pale in colour and its eyes are reduced in size and simplified in structure. Ceuthophilus pallidus from U.S.A. is another example showing loss of pigmentation. To compensate for these deficiencies organs of tactile sense are often highly developed. These consist of very elongated antennae which the insect waves about to detect any unusual movements. The normal stillness of the air within the cave creates an environment in which slight, small movements are more readily perceptible than in the open air. Another peculiarity among cavernicolous insects, which is well illustrated in the Rhaphidophoridae, is the tendency to develop long, thin, spindly legs, thickly covered with sensory setae, which enable the insects to beat a quick retreat when they sense unusual vibrations.

In his section on Insects in the Cambridge Natural History, Sharp (1901), discussing cave dwelling Locustidae says, “The species though found in the most widely separated parts of the world have a great general resemblance so that one would suppose the specimens found in the caves of Austria, in the Mammoth cave of Kentucky and in the rock-cavities of New Zealand to be one species although they are now referred to by entomologists to different genera”. Although this statement cannot be accepted to-day, it does draw attention to the superficial similarity in the whole family, especially with regard to their habitats and the conditions necessary for their existence. This similarity however, may be due to separate evolution along similar lines because of similar conditions, instead of being due to close relationship.

The distribution of the three sub-families of the Rhaphidophoridae covers a wide area, but with the exception of North America, is confined to-day principally to the Southern Hemisphere. The Macropathinae which are the most primitive, are typically paleantarctic, including New Zealand, South Australia, Patagonia and the Cape of Good Hope. They are most abundant in the Australian region where there are thirteen genera and about thirty species so far recorded (Chopard, 1949). The great majority of these are localized in New Zealand. The Rhaphidophorinae are found in the Indo-Australian region as far north as Japan, and also include some of the cave-dwelling forms of Europe; while the Ceuthophilinae are confined to North America.