VOL. 92 DECEMBER, 1968 TRANSACTIONS OF THE ROYAL SOCIETY OF SOUTH AUSTRALIA INCORPORATED : ADELAIDE PUBLISHED AND SOLD AT THE SOCIETY'S ROOMS STATE LIBRARY BUILDINGS | NORTH TERRACE, ADELAIDE Price: Six Dollars Thirty Cents FIELD NOTES ON RABBIT BANDICOOTS, MACROTIS LAGOTIS REID (MARSUPIALIA), FROM CENTRAL WESTERN AUSTRALIA BY D. R. SMYTH*} AND C. M. PHILPOTT* Summary The numbers and range of the rabbit bandicoot have greatly diminished in recent years. The results of a natural history study are presented to act as a basis for future research, and to facilitate conservation of bandicoot populations. At least two bandicoots were living in one square mile of scrub containing mulga, spinifex and tussock grasses, near the Warburton Range, Western Australia. Their burrows, scratchings, tracks and faeces are described. Fifty-eight burrows were mapped and 55 of them found to be distributed at random within two areas delineated by presence of bandicoot scratchings. The main food species of the bandicoot were the termites Hamitermes (Drepanotermes) rubriceps (Froggat) and Eutermes tumuli Froggat. On each of 8 consecutive nights from 2% to 27% of the burrows were used. Out of 35 representative burrows, 14 were not used on any of the 8 nights, and none were used on more than 4 nights. FIELD NOTES ON RABBIT BANDICOOTS, MACROTIS LAGOTIS REID (MARSUPIALIA), FROM CENTRAL, WESTERN AUSTRALIA By D, RB, Smyru*f and C. M, Paiwrorr* [Read 2 October, 1967] * Royal Zoological Soviety of South Australia, Frome Road, Adelaide, South Australia 5000; f Present address: Research Sehool of Fhological Sciences. Australian National University, Qynberra, A.CLT. 2600, SUMMARY The munbers and range of the rabbit bandicoot have greatly dirinished in receut yeurs. The results of a natural history study are presented to act as a basis for future research, and to facilitate conservation of bandivoot populations. At Jeast two bandicuots were living in one square mile of sernh containing mulga, spinifexs and tussock prasses, near the Warburton Range. Western Aus- tralia, Their burrows, scratchings, tracks and foezes are desertbed. Fifty-eight burrows were mapped and 55 of them found to be distributed at random within two areas delineated by presence of bandicoot soratehings. The main food spe+ cies of the bandicnut were the termites IJamitermes (Drepanotermes) rnebriceps (Froggat) and Hutermes tumuli Froggat- On cach of 8 consecutive nights from 2& to 27% of the burrows were used. Ont of 35 representative hutrows, 14 were not used on any of the 8 nivhts, and none were used on more than 4 nights, INTRODUCTION Since the coming of European man to Australia, the range and population size of many marsupials have been greatly reduced. In some species it is probable that this reduction has reached its limit, and that the species are extinct, Calaby (1963) suggests that there are six such species. If so, then further study of them is restricted to museum specimens and fossils, Calaby also lists other marsupials close to extinction. The biology of most of these is little-known, and opportunities to record it imcommon. Calaby includes the rabbit bandicoot, Macrotis lagotis Reid, shown in PI, 1, Fig, 1, amongst these species, Recently, we studied some of the natural history of a small population of rabbit bandicoots near the Warhurton Range, Western Australia. Previous work of this uature in the genus Muerotis has mostly been in the form of brief observa- tions. made during collecting trips. Wood-Jones (1924) made many such obserya- lions. He diseusses notes recorded by other authors, and also describes aspects of behaviour of captive rabbit bandicocts, Our aim in recording and discussing our results is to provide a background on which other studies on rabbit bandicoots may be based, or with which they may be compared, If these handicoots were to become extinct, we hope our study will give future workers some idea of how they lived. METHODS During 1866, we surveyed the present range of rare marsupials in parts of arid Australia, the results of which have been reported (Philpott and Smyth, Trans. Roy. Soc. $.A. (1968), Vol. 92. 4 D, RK, SMYTH ann C. M. PHILPOTT 1967). During this survey, we located an area near the Warburton Range where rabbit bandicoots were living, and used it as our study arca. We carried out the work recorded here during November and December, 1966. We were taken to the areca by semi-nomadic aboriginal people of the Ngu- ‘nyatjara and Pitjantjatjara tribes, These people have an intimate knowledge of the signs and habits of the rabbit bandicoot, because they still hunt it occasionally for food. They were happy to communicate this knowledge to us. This communi- cation was facilitated by use of the aboriginal names of the rabbit bandicoot. Near the study area, they call it.the “ninu”. East, in the Musgraye and Everard Ranges, it is called the “talku”, while south-west towards Laverton it is called “matura”, All aboriginal names are spelt phonetically following Douglas (1964), We mapped the burrows in the study area by triangulation using a prismatic compass. To record if burrows had been used overnight, we inspected them daily, swept away fresh tracks from the mouth, and placed a twig in the single entrance such that it would be dislodged by a bandicoot entering or leaving. We took care to minimize our scent on the twig und around the entrance of the burrow. We scored a burrow as being used by # rabbit bandicoot if the twig was knocked down and characteristic tracks and tail marks left near the cntrance. Faecal contents werc examined microscopically after crushing in water con- taining a drop of detergent. We identified plant species in the field, and checked several in the State Her- barium of South Australia. Ants were identified at the C.S.I.B.O, Division of Ento- mology, Canberra, and termites and insect larvae by staff at the South Australian Museum. RESULTS (a) General description of the study area The study area is about 7 miles (11 km) north of Warburton Mission, and its latitude and longitude are 26°02’ S. and 126°34' E. respectively. Its position. is shown in Fig, 1, It is an irregular shape, with a maximum length of 1-4 miles (2-3 km) and is 1-0 mile (1-6 km) across at its widest point. Fig. 2 is a sketch map of the study area, which has a total area of 1-0 square mile (2-6 sq. km), STUDY AREA : : aXe Fig. 1, Location of the study area near Warburton Mission, Central Eth Australia. with the general area shown om an indented map of Australia. FIELD STUDY OF THE RABBIT BANDICOOT ie MULGA-TUSSOCK — MULGA- SPINIFEX OPEN SPINIFEX a Fl OPEN ROCKY SAVANNAH + BANDICOOT BURROW #f | OPEN MULGA = RABBIT ae | IG, wet - LIMIT OF SCRATCHING Vig. 2, Sketch map of the study wea, showing approximately the positions of 58 bandicont burrows, the limits of bandivoot seratchings, and the areas covered by the 5 yegetation groups. The following general description is based on observations and on informa- tion from the Atlas of Australian Resources (1959). The area has a dry continental climate, with irregular rainfall, but with the maximum amount falling in summer. The average annual rainfall is 5-8 inches (13-20 em). The average daily maximum température in January (summer) is 95-100" F (35-39° C), In July (winter), the average daily minimum is in the range 40-45° F (4-7° Q). The study area is bounded by lew rocky ridges on the north and north-west. From the low pass between them, the land slopes gradually to the south-west. Skeletal soils cover the hills, while the flats and plains are of deep, desert loams, with small rocks and pebbles distributed through all Jevels. Permanent surface water is absent from the study area, and uncommon nearby. Tor the purpose of description, the vegetation of the arca has been divided into five classes. These are not discontinuous, and are based on subjective observa- tions. (i) Open rocky savannah. This type occurs on the skeletal soils of the ridges and hills. Perennial hard-leafed grasses and occasional shrubs such as. Cassia and Piiletus obovatus provide a sparse coyer. Mulga (Acacia. aneura) also occurs at very low density, (ii) Open mulga scrub, Pl. 3, Fig. 1—Mulga occurs in stands of low to medium density in certain areas. There are no shrubs or perennial herbs under the open mulga canopy. Ephemeral crucifers, grasses, composites and other herbs spring up on the bare ground after rain, The soil is often shallow. 6 D, A. SMYTIT ann C. M. PHILPOTT (iii) Mulga-tussock scrub, Pl, 3, Fig. 2—This is similar to the second type, except that the mulga is usually more dense, and the ground has a moderate cover of perennial tussock grasses, The soil is probably deeper than in the previous class, (iv) Mulga-spimifex scrub, Pl. 3, Fiy, 2—On the deep desert loams, mulza and spinifex (Triodia bysedowii) often occur together at medium densities. Many other specics of trees and shrubs grow as scattered solitary individuals. Such trees as the bloadwood (Eucalyptus terminalis) and the corkwood (Hakea sub- erea) occur, The shrubs present include the dead finish (Acacia tetragzonophylla), the witchetty bush (Acacia kempeana), aud various Cassia and Eremopliila species. Some herbs and grasses also occur, (v) Open spinifex glade, Pl. 3, Fig. 4—In some areas, the major plant cover is spinifex, Occasional individuals of mulga and witchetty bush occur. Ephemerals grow in the spaces between spinifex clumps after rain, The approximate limits of these classes in the study area are shown in Tig. 2. (by) Signs We did not see any rabbit bandicoots above the surface of the ground during daylight hours, They are probably strictly nocturnal. Their presence in an area seatched during the day can only be proved by capturing a specimen. However. various characteristic signs of their presence can be readily secn. (i) Burraws—The most obvious sign is the hurrow. We located 58 bandicoot burrows in the study area, and completely dug out six. A sketch diagram of two representative burrows is shown in Fig, 3. Fig. 3. Sketch diagrams of two burrows. which were dug ont com- pletely. (1) and (2) are the side elevation and plan respectively of «burrow near the unrth-east corner, (3) and (4) are the side eleyation and plan of a burrow an the central western edge, X indi- cates the position from which 2 temale rabbit bandivoot was taken. ENTRANCE A small cireular mound of soil was present outside the entrance of all bur- rows. The tunnel cross-section was usually cirenlar or slightly oyal, and about 6 inches (£5 cm) in diameter. Burrow entrances were often partially hidden by dense spinifex, tussock grass or juvenile mulga plants. Occasional burrows were underneath or near fallen logs. Eight of the burrows were conspicuous by being dug into the large mounds of loose soil frequent in the area. Two representative lurrows are shown in Pl, 2, Figs. 2 and 3. - FIELD STUDY OF THE RABBIT BANDICOOT ' In two burrows which contained a solitary rabbit bandicoot, the tunacls were blocked in from one to three places along their length by loose, frequently dug soil. There were no faeces in, or close to the burrows. In one of the burrows con- taining a bandicoot, an area of the floor near the end was damp, possibly from recently voided urine. The two bandiceots were at the extreme ends of their res- pective burrows, und were rapidly lengthening them when captured. Rabbit bandicoot burrows were different from other types of burrows present in the area. Rabbit ( Oryctolagus cuniculus Line) burrows are often built in war- rens, are of a different plan, and are often marked by facees, We located two bur- mows which were probably used by rabbits in the study area, Goanna (Varanus sp.) bnrrows which were also present, were smaller, semi-circular in cross-section, and much shallower. (ii) Scratchings—Diggings and scratchings covered the soil over inuch of the study area, They were of distinct types. (1) The most frequent type was a shal- low, cylindrical pil, 2-8 inches (5-20 cm) deep, and 2-6 inches (5-15 cm) wide, an example of which is shown in PI. 1, Fig. 2. The soil from them was piled up in all directions around their mouths. (2) There were also smaller seratchings of irregn- lar shape, and areas of looseued topsoil, often near mulga roots. (3) A further tyne was a conical pit under certain tussock grasses, as illustrated by Pl. 1, Fig. .3, Chit of a sample of 107, 62% were under Eragrostis erivpoda and EF. luniflora, 18 under Denthonia bipartita, and the rest under Aristida contorta. and Eriachne mucronata. These pits were mostly in mulga-tussock scrub. They were about 4-8 inches (10-20 cm) deep, and descended at an angle, We helieve that all these diggings were made by rabbit bandicoots. The evi- dence which suggest this is: (1) the diggings were distributed in close association with the handicoot burrows, as shown in Fig, 2, and were never more than 600 feet (180 m) From one; (2) the density of seratchings was frequently higher in the vicinity of a burrow; (3) faeces, later shown to be closely similar to those of recently captured bandicoats, were occasionally found on or near the freshly dug soil; and (4) insect species, fragments of which were found in newly captured bandicoot's faeces, (see Kesults, Part [d]), were often found in the seratch- ings, (iii) Tracks—Characteristic tracks were often left in burrow entrances and on fresh scratchings. They were also left on the pebbly surface between some bur- rows in the area. A sketch diagram of a set of tracks entering the mayth of a rubbit handicoot burrow is shown in Fig. 4. Although we have only aboriginal opinion and circumstantial evidence, we believe these charactcrishe tracks were those of a rabbit bandicvot. We observed charactcristic grooves on loose soil near scratchings and also near burrow entrances, as shown in Pl]. 2, Fig. 3. Captive bandicoots left stmilar Fig. 4. ‘Sketch diagram of a series of presumed, rabbit byn- dicont tracks leading to the entrance of a burrow. The probable direction of travel is shown with an ariiow, § D, R, SMYTH asp C. M. PHILPOTT 4 Fig. 5, Frequency distribution of 58 burrows, showing the dis- tance of the nearest neighbour- ing burrow to each of these burrows. The mean distance is 249 feet with a standard error 4 of 14-1 feet, NO. OF BURROWS o i) 100 200 300 400 500 600 DISTANCE OF NEAREST NEIGHBOURING BURROW (FEET) * Fig. 6, Frequency distribution ie of 85 burrows, showing the number of nights each of these g burrows was used over a period \ of 8 consccutive nights. The ae ~~. mean number of nights is 1-14, and its standard error is 0-036 nights, OF BURROWS No. ——— 2 « —__—s 0 1 2 3 4 5 6 7 2 NO, OF NIGHTS EACH BURROW USED marks while digging in loose soil on the cage floor. The marks were left by the tail, which was used as a strut while the hind legs were in use. (iv) Faeces—I'aeces of captive bandicoots were similar in shape and size to those found occasionally on or near scratchings. Four representative groups of pel- lets are shown in Pl. 2, Fig, 1. (ce) Distribution of burrows The distribution of the 58 burrows over the study area is shown in Fig. 2. Most burrows were in mulga-tussock scrub and mulga-spinifex scrub. Two were in each of open mulga scrub and open spinifex glade. There were none in open rocky savannah. FIELD STUDY OF THE RABBIT BANDICOOT 4g The burrows were distributed in two clusters bounded approximately by the limit of bandicoot scratchinys. In order to test if the burrows were distributed at random within the scratched arcas, a grid of quadrats with sides equivalent to 340 feet (105 m) was laid over a map of the area. We considered only those quadrats within the scratched areas, and those intersected by the limit of scratchings. The distribution of burrows within these quadrats was tested for goodness of fit with a Poisson distribution, as shown in Table 1. TABLE 1 Distribution of burrows m quadrats covering all of the seratchod area, showing goodness of fit test Lo a Poisaon distribution. No, of Observed Expected (O—f)* burrows No, of No. of —-- per quadrats quadrats EH qitartrat (QO) (BE) | 0 60 | 61-28 027 | Mean = 0-529 ot u2e41 | O07 | Wee = 0-458 2 , 9 $57 x o> 1d | , 1 1-5] L b-0n0 70%, p= 80%), tor iuicre | i ; 0-25 Total | 14 104-00 ees The result (x7, = 0-114, 70%4< p < 80%) strongly suggests that the burrows were randomly distributed over the area, This indicates that the position in which a bandicoot dug a burrow was independent of the position of other burrows in the area. It is probable that a rabbit bandicoot uses a burrow as a refuge into which to flee when danger threatens. If so, then some description of distances between bur- rows is useful. Fig. 5 shows a histogram of the number of burrows and the dis- tance of the nearest neighbouring burrow to each of these 58 burrows taken in turn. The spread of the nearest neighbour distance is small, with no burrow more than 550 feet (170 m) from any other one. There is a small peak in the histogram composed of burrows with nearest neighbour distance greater than 350 feet (105 m). These burrows were distri- huted over the whole study areca, and the presence of this peak is therefore per- haps fortuitous. (d) Food We collected all the faeces voided by two bandicoots during their first night in captivity, One pellet from each was examined microscopically after crushing, They both contained more than 50% by volume of soil and grit. They also con- tained a small number of rabbit bandicoot hairs, However, the most striking con- tents. were the hard, keratinised mandibles and whole heads of certain insects, We collected insects from the study area which were possible food species, dissected their mandibles, and compared these mandibles and heads with those found in the faecal pellets. By this method we were able to identify remains of both soldier LO D, R, SMYTH snp C. M, PIITLPOTT and worker castes of two termite species, Hamitermes (Drepanvtermes) rubriceps (Froggat) and Eutermes tumuli Froggat, and workers of one species of ant, Camponotus sp. Counts of the numbers of mandibles, and in two instances twice the number of heads (for comparative purposes), of these species are given in the first two rows of Table 2, ‘ TABLE 2 Frequency of mandibles of yarivus insect specics in facenl pellets from recently captured rabbit bandicoats and from near scratchings made by bandicoots in the study area, ! Nowak. Termites | Ant Various Origin of pellets Tridomyr- | unknown pellots examined) | Hamitermes rubriceps Budermes tumuli ned apecins {Total No. | Worker — Soldier Worker Soldier* — species* Bandieont I 1/5 S 1 TO 34 { 0 OF Baucheoot 2 1/12 4 0 60 1k 0 OF xtreme 13 5/5 76 1 a. 4 V) cf) odge 57 1 1 {) | i} ) 72 2 ! £ H 0 fi | 18 4 a 2 0 0 | 120 ih 6 2 j () } 4 SW Centre 2/2 70 4 0 0 2 1 59 ‘) H 12 0 ‘) Extreme Ho! 2/2 87 i i) 2 174 i) edge S85 f) 0) 2 140 Of H eentre 1 1/4 221 ti) u 2A i) 0 Central E 0-5/4 HA ‘) 2 iE 4 i) edge } Centre O-a/2 26 0 i) 0 8 2 * No. of heads » 2. F +1 mandible of Camponotus species. ~ +2 mandibles of Camponotus specios. § $2 manilibles of larval Lepidaptera. The counts may not represent the relative numbers of individuals eaten. Cap- tive bandicoots masticated their food very thoroughly, and it is possible that a dif- ferent proportion of mandibles and heads from different food species were crushed beyond recognition. However, most mandibles from pellets were whole, and little changed in shape or colour. There is probably some simple relationship between numbers eaten and fragments in faeces. After working with faeces known to be bandicoots’, we analyzed contents from presumed handicoot faeces collected near scratchings. These results are given in the remainder of Table 2. They closely resemble those from. known ban- dicoot faeces. In these other pellets, we identified remains of a small black ant, Tridomyrmex sp., and a white Lepidopteran lJarya, as well as species found in the. earlier analyses. The two species of termites were abundant ever much of the study area. They were found in vast underground colonies, with domes and mounds raised above the soil surface in places. Bandicoot seratchings often intersected the under- ground passages, bit there was no sign of interference with the mounds. From the numbers of mandibles in the pellets, and from the numbers of scratchings inter- secting colonies, it is likely that these termite species were a primary source of food for bandicoots in the study area. MIFIM STUDY OF TIE RABBIT BANDICOOT i The species of Lridomyrmex identified from faeces was also extremely abun- dant over the area, Being highly active and pugnacious, they readily attacked and carried off termites when given access to their colonies, Because of this, and their ubiquitousness (therefore being readily available where a bandicot was feeding on termites), as well as the fact that their remains only occurred in large numbers in one group of pellets, it is possible that they are ouly a subsidiary food sonree, or perhaps even a “contaminant” of the usual diet. The larger ant, Camponotus sp., was less abundant than Iridomyrmex, Ts sporadic appearance in faeces suggests that this is also a secondary food species. We found the large white grub (larval Lepidoptera) under fwo specees of tussock grass, Eragrostis eriopoda aud FE, laniflora. Of the five grasses scratched imder in the study area, these two were the most frequent. Although mandibles of this particular larva were tound only twice in a safaple of 14 Keller the large size of the grub (up to 20 mm long), and the large number of tussocks scratched wnder, suggests that larvae of this species and perhaps other unidentified ones are an impertant food. Faeces oecasionally comtuined other materials. Small amounts of plant tissue such as undigested seeds and plant hairs were sometimes present. Unidentified thickened tissue also oceureed in moderate amounts. These objects were not eon- sistently preseut, (oc) Activity For eight consecutive days we kept a record of the burrows that had been used during the previous night, Thirly five burraws were scored on all eight nights, and more were added ta tlic list as we found them, For the last two days we scored 52 burrows, and considered thal we had located at least 90% of all the lirrows present. Our method of scoring did not differentiate between a single entry or exit by a bandicoot, nor did it dillerentiste either of these from inultiple entries and exits during one night. We must assume firstly that our working in the area did not have any effect on the behaviour of the bandicoots present. Although we have no concrete evi- dence to suggest that this is a valid assumption, we noted that recently captured bandicoots showed little fear of hurnans or human scent, Secondly, we will assume that the burrows we observed for the longest period were a representative sample of all burrows with respect to use by a bandicoot. We could not detect any visible difference between these burrows and the remaining ones. For each night we calenlated the percentage of burrows which were used oul pf all those we scored. The mean percentage and its standard errov were 15-5 + 2-6%, with a range of from 2-0% to 265%, Tho fact that only one burrow was used om one night (2:0%), indicates that all the handicouts present did not forage every night, but spent at least one underground. We recorded overnight minimum tem- perature during the period, but could not deduce any relationship between it and percentage of burrows used, When considering the number of times each indinidiual burrow was used, it is convenient to take only those hurraws scored for all eight nights. The frequency élistribution of these 35 burrows showing the mumber of nights each of thein was used is shown in Fig, 6. Fourteen burrows (40%) were not used on any night. These burrows did not appear to he distributed at random oyer the area, but a statistical test is impracti- cal because of the small numbers involved: They tended to be on or near the east- ern edge, where habitats with spinifex predominated, From Fig. 6, we can also \2 Dy H. SMYTIT ann ©, My PHILPOUT note that more burrows were used on two cecasions than on one. This can be explained when we consider that a bandicvot sponding # day in a burrow would use it on entering one night and leaving on a subsequent night, It would be scored both times. We know that there were at least two female bandicoots in the study area, because we captured two from burrows about 1,400 feet (425 m) apart on the western edge of the area, This was within two days of completing the observa- tions on activity. Outside the study area, the nearest locality we found where rabbit bandicoots were living was about 3 miles (5 km) away. Even assuming no migration to or from the study area, it is impossible to estimate the total sumber af handicoots present. Recause there were no signs of activity beyond 600 feet (180 m) from the nearest hurrow, we may assume that for both female bandicouts captured, their home range was completely covered by the study area. (f) Notes on -caplite specimens The twe adult females we caught in the study area did not have pouch young whon taken in mid-December, nor when examined 2 month later. When kept unavoidably at an air temperature: of 100° F (39° C) for a short period, they lapped the water which was supplied. There was no permanent water available in or near the study arca, and it is milikely that bandicoots experience such high air temperatures in the wild. At these temperatures. they slept on their sides, with tail, hind-feet, and head all extended. However, at lower temperatures they slept crouched ou all four feet, with the nose tucked under the chest. the tail curled around the body, and the ears folded forward over the eyes, ‘The captives started to eat chopped, tinned beef and to drink milk about # week after being taken into captivity. Later, they ate minced fresh meat und live mealworns. DISCUSSION Where ony study overlaps. with those of other recent authors, the Beucral results agree. Marlow (personal communication) caught two rabbit bandicoats in an area of similar habitat on April 5 and 6, 1965. At latitiide 20°49 S, and longitude 130°195 E., (about 25 miles (40 km) NW. of The Granites, N.T.), the bandicouts were living in an area of spinifex-Melaleuca scrub, wilh sandy soil and numerous termite mounds. One of the bandicoots, a female, is preserved in the Australian Museum, registered number M8620. A brief survey of the distribution of rabbit bandicoots neur Warburton Mis- sion indicated thit they lived in small isolated groups over a large area. Newsome (1962) says he found that rabbit bandicoots were living in at least four and per- haps six localities in a rectangular area 50 miles long by 30 miles wide (80 km hy ‘50 km) in Central Australia. It is probable that rabbit bandicoots utilize only a small percentage of the large homogeneous areas which could support them. Tnsect larvae and small vertebrates have both heen the most frequently noted types of food of the rabbit bandicoot, ¢.. Marlow (1962), Finlayson (1933) found rodent fur in the gut of a closely related species, Macrotis minor (Spencer). Although house mice, Mus musculus Linne, were present in the study area, their fur did not appear in any of 50 pellets cursorily examined, Aboriginal reports clearly indicated that the rabbit bandicoot eats the large witchetty grubs, (larvae FLELD STUDY OF THE RABBIT BANDICOOT \4 of buprestid beetles}, when they are present in the roots of the witchetty lush, Acacia kempeena, Gould (1645) made a similar observation in South-Westeru Australia. At the time of our study, witchetty grubs were absent from the sludy area, although ahoriginals elaim that they had been plentiful in former times. The food of the rubbit bandicwot reported here is remarkably similar to that of the marsupial, Myrmecobius fasciatus Waterhouse, (banded ant-eater), which Calaby (1960) studied in South-Western Australia, He fommd that termites of many species, all different from those recorded above, were the principal food, and that remains of ants were also often Found in faeccs. Likewise, he argued that ants were ingested incideatally while termites were being caten. However, he fonnd no traves of Lepidopteran or other insect larvae, There is little concrete data on the breeding of the rabbit bandicoct, Woaod- Jones (1924) comments that breeding of Macrotis in Central Australia is regu- lated by rains and abundance of food, an observation made om many species which live ina climate without regalar seasonal variation. Two other species of bandicoots haye heen shown to breed throughout the year, Lyne (1964) working with Perameles nasuta Geoffroy, and Mackerras and Smith (1960) with Isoedon macrourus (Gould), have heth recorded breeding during all seasons. Observations of activity of the rabbit bandicoot are limited, but Wood-Jones (1924) has noted that captives nsually emerged about an hour after dusk, Stodart (1966) observed the long-nosed bandicoot, Perameles nusuta Geoffroy in a M-acre enclosure. She found that they usually emerged from their vests at dusk, and most of their activity was completed after about two hours. They spent nearly all of this Lime searching for and, Although we found that a group of rabbit bandicoots lived in a small propor- tion of an area with the potential ta suppert them, we have no evidence concern- ing territorial behaviour within this small area. Two male and four femule long- nosed handicoots in am enclosure did not have territories (Stodart, 1966) and were solitary feeders, with intentional contact between individuals restricted to actual inating. The rabbit banchicaot was moderately plentiful and widespread in temperate and arid Australia during the last ceritury. owever, a rapid decrease in range andl numbers has been the pattern im recent years, especially in Lhe more termper- ate regions. It has not been recorded frown New South Wales since 1912 ( Marlow, 1988.) and from Victoria since at least 1866 ( Brazenor, 1950). In 1934, tt was said to he widely distributed in Western Australia south of the Kimberley region (Glaucrt, 1934), but twenty years Jater he reported that there were “no reports of its pres- euece anywhere (in Western Australia) within recent years”, (Glanert, 1954), In arid South.and Central Australia, Finlayson (1961) claimed that it was rapidly bein reduced loa rare form, but Newsome (1962) indicated that it had recently been recorded from 36 loculities in the Northern Territory. Rabhit bandicnats were ttken near Birdsville i South-Western Queensland during 1957 ta 1954 (Mack, 1961). but while in the area recently; we obtained good evidence that the bandicoots disappeared there about five years later. Thus, a steady reduction 10 rattwe towards parts of arid Central Australia seems to be taking place, with no evidence that this reduction is slowing down or stopping, An attempt to conserve some natural populations of rabbit bandicants 14 probably necessary now if they are to continuc as one of Australia’s naturally occurring species. Suecessfal canservation of a species depends largely on.a know- ledge of its distribution and eeolugy. We hope our study will be useful to those who plan to conserye the rabhit bandicoot in its natural habitat. 14 D. R, SMYTH ann C. M. PHULPOTT ACKNOWLEDGEMENTS This work was supported by a Nuffield Foundation Research Grant to the Royal Zoological Society of South Australia, and was administered by the Society's President, Dr. P. S$, Watts. We wish to thank Professor IE. G. Andrewartha of the Department of Zoology, University of Adclaide, for encouragement, guidance and for many useful discus- sions. The Council and Staff of the Royal Zoological Society of S.A. gave us unlimited help. We received excellent co-operation from the Western Australian Fisheries Department, and Mr, A. J. Fraser, Chief Warden of Fauna. Finally, we wish to thank the Western Australian Department of Native Welfare; and the supeciteadent, stall and people at the Warburton Ranges Mission for generous help. REFERENCES, Atlas af Australian Resources (1959), (Department of National Development: Canberra. ) Brazenon, C. W, (1950). “The Mainmals of Victoria”. (Brown, Prior, Anderson; Melbourne. } Carapy, J, H. (1960), Observations on the banded ant-eater, Myrmecobius f, fasciatus Water- Iynase (Marsupialia). with particular reference to its food habits. Proc. Zoul. Soc. Lond. 138; 183-207. Carapy, J. A. (1963). Australia’s threatened mammals, Wildlife, Brisbane J: No. 1. 135-8. Doucras, W. H, (1964), An introdnetion to the Western Desert Language. Oceania Linguistic Monogr, No. 4 (revised). (University of Sydney: Sydney.) Finnaysox, H, HL. (1935). On mammals from the Lake Eyré Basin: Part 11, The Peramelicdae, Trans. RB. Soc, S, Aust. 59: 227-36, Wincayson, FL. H. (1961). On Central Australian Mammals: Parl IV, The distribution and status nf Central Australian species, Ree, $, Aust. Mus. 14: 141-91. Gravenr. L. (1934), The distribution of inarsupials in Western Australian. J. Proe, R. Soc, West, Aust. 19; 17-32. Guavert, L. (1954). The recent increase of the rarer native mammals. If Records fron the W.A, Museunt. W, Aust. Nat, 4: 129-82. Goun, J. (1845). “The Mammals of Australia, Vol, 1° (The Author: London.) Lyxz, A, C (1964), Observations on the breeding and growth of the mursupial Perameles nasuta Geolltroy. with notes on other bandienots, Alist, J, Zool. 12; 322-39, Mack, G, (1961). Mammals from South-Western Queensland. Mem. Od. Mus. 13: 213-29. Macxerras, M, Josepnink, and Sacre, Ruta Uf, (1960), Breeding the short-nozed bandi- coat, Isoodon macrourus (Gould). in captivity. Aust. J. Zool. 8: 371-82. Mannow, B, J, (1958), A survey of the marsupials of New South Wales. C.S.1.R.0. Wildl. Res. 3: TLld4 Mantow. B. J, (1962). “Marsupials of Australia.” (Jacaranda; ‘Brishanw.) Newsome, A. E. (1962), Rabbit-cared. bandicoots or bilbies.. Aust. Nut. fist. I4: 7-8. Parteotrr, CG. M., and Ssavru, D. R, (1967), A contribution to our knowledge of some natiye mammals fram inland Australia, Trens, RA. Soe. S. Aust. 91: 115-34, Stoparr, BE, (1966). Management and behaviour of breeding groups vf the marsupial Perameles nasuta Geoffroy in captivity. Aust: J, Zool. 14: 611-23, Weae Tones F. (1924), “The Mammal; of South Australia, Part IL” (Gavernment Printer: Adelaide. } D. KR. Smevier and C, M. Prowreorr PLATE | PLATE | lig, 1. An immature female rabbit bandicoot from 120 miles east of the study area, Approx. 0-28 x natural size. to Vig, Cylindrical bandicoot scratching, with faecal pellets on the top edge. The measure is 1 foot (30 cm) long. io] Fig. 3, Conical bandicoot scratching under Eragrostis laniflora, The measure is 2 in. x UU in., (5-0 em x 8-5 em). PLATE 2 D. BR. Ssivru and C. M, Pintrorr Fs poe =) <=", = esi aril Piare 2 Fig. 1. Rabbit bandicoot faeces, showing groups of pellets voided together. The seale is in inches and centimetres, Fig. 2. Bandicoot burrow near a fallen mulga log. Fig. 3, Bandicoot burrow with tailmarks on the mound of soil at the entrance, The measure is 1 foot (30 em) long. D. R. Smeyvrn and C. M. Prirrorr PLATE 3 PLATE 3 Fig. 1. Open mulga serub, with a burrow at the base of « mulga in the foreground. This vegetation type contained only 2 burrows. Mulga-tussock scrub, with a burrow partially hidden by tussock grass in the fore- ground. 3. Mulga-spinifex scrub, with bandicoot scratchings in the foreground. Fig. 4. Open spinifex glade, showing low rocky hills in the background. There were only 2 burrows in this vegetation type. z # to “Tj z oo AN ANALYSIS OF VEGETATION ON STRANDED COASTAL DUNE RANGES BETWEEN ROBE AND NARACOORTE, SOUTH AUSTRALIA BY R. M. E. WELBOURN AND R. T. LANGE Summary The presence or absence of 80 species has been scored in 330 samples of sclerophyll forest on nine dune ranges, and the data classified by association analysis. The vegetation is a mosaic of groups, within which four main groups may be distinguished. These roughly correspond to orthodox vegetation societies, lint key species are determined solely by measured association with other species, without reference to physical prominence. For example, the species with the highest degree of association in this study, Phyllota pleurandroides, is a small shrub. The potential value of the groups in vegetation mapping is indicated. AN ANALYSIS OF VEGETATION ON STRANDED COASTAL DUNE RANGES BETWEEN ROBE AND NARACOORTE, SOUTH AUSTRALIA by R. M. KE. Wetvourn and BR. T, Lanon [Read 9 November, 1967] SUMMARY The presence or absence of 80 species his been scored in S30 samples of sclerophyll forest on nine dune ranges, and the data classified by association- anilysis, The veyetalion is a mosaic vt groups, within which four main groups may he distinguished. These roughly correspond to orthodox vegetation societies, hut key specics are determined solely by measured association with other species, without reference to physical prommence. For example, the species with the highest degree of association in this study, Phyllota pleurandroides, is 2 small shrub, The potential value vf the yroups in veyetation mapping is indicated, INTRODUCTION The purpose of this study was to show by measurement, the floristic composi- tion of the dune range vegetation and to discuss factors controlling its disposilion. A discussion of analytical methods used has been published clsewhere (Welbourn and Lange, 1967), The study was an M.Sc, project with the University of Adelaide, to whom, with the State Herbarium of South Australia, grateful acknow- ledgment is made. The habitats are distinct ranges considered to he successively younger toward the present coast. In the area studied there are at least twelve ranges subparallel to the. coast, about 2 km wide by 30 m high above the otherwise flat AR (PI. 1). Each range consists of two portions. The core is more or less consolidated caleareous beach sand, a rele of coastal dunes formed at various stages of the Pleistocene from about 600 to 200 thousand years B.P. (Sprigg, 1952). This material outcrops westward as acolianite limestone, parcnt material of the terra rossa soil carrying open woodland. Oycrlying this core are siliccous sands, of uncertain origin but considered to be windsorted, leached and of Recent age (Blackburn ez al,, 1965), These sands form the bulk of each range, and are the parent material of the podsol which supports the predominant vegetation. This is largely the Eucalypius baxteri association within the dry selerophyl!l forest formation (Crocker, 1944). Pattern, the degree of non-randommness in the spatial distribution of indivi- duals, was the vegetation feature inyestigated. Since pattern reflects habitat varia- tion, the patterns of ecologically similar species tend to coincide, or associate, thus indicating a vegetation group. ‘he variable commonly measured to reveal such groups is species frequency, the proportion of sample quadrats occupied. Associa- tins were determined with such data in this study. Trans, Roy. Soc, 5.A, (1968), Vol. 92, 20 R, M, E. WELBOURN anp R. T. LANCE 330 +59 Phyllota plevrandroidas “27 *63 Mongkoca scoparia -208 mS ql. 120 * 107 Isopegon ceratophyllus “101 0 3 if 6 A 2 be o Zz 2 of g i Y 9 5 a z < GO ie] Fig. 2. VEGETATION OF STRANDED COASTAL DUNES 21 METHODS The arew bounded by Kingston, Naracoorte, Penola and Rohe was selected for study because the dune ranges are most distinct here ( Fig. 1). Sampling areas were located within the remaining estimated 750 sq. kin of vegetation so that the variation of mean annual rainfall, about 200 mm, would be represented on each range. Most of the vegetation was unnatural in some way, so that, haying excluded areas obviousty disturbed, or burnt less than four years before, or dominated by bracken, 33 areas remained. At each of these, ten 20 sq.m circular samples were located randomly where feasible, otherwise systematically. All species likely to occur in upland sites in the area studied were peat also Banksia marginata and Calytrix tetragonda were divided into ad hoc torms to render the data more sensitive to habitat variation. A reference collection of relevant species was used throughout, The method selected to reveal vegetation groups was association-analysis as proposed by Williams and Lambert (1959), Them recommended association index is for cach apecies the sum of all the chi-squared. values obtained from contin- geney testing with each other species in turn, The procedure is to subdivide the 330 samples firstly into two groups, respectively with and without the appropriate kev species, The key species is that with the highest association index at any stage of subdivision, Within each of the groups thus formed, the process is repeated, tractionating the original single group of 330 samples into progressively more sub- slivisions, In this study, only species with a frequency greater than 1% were consid- ered; data were computed clectronically, and subdivision was stopped at 16 SOUPS, THE VEGETATION ‘Lhe analysis: ii Uhese terms reveals a serics of groups of deereasing importance, conveniently represented as a hierarchy (Fig, 2). The validity of any gcuup asa recognizable vegetation community is indicated by the range of associ- alion index over which it persists undivided; this is supported by its containing relatively few minor groups. For example, Group B subdivided on Monotnca is such a community, since it persists fram Index 155 to Index 20, and embraces only two minor groups as far as the data were analysed, On this basis four such groups may he distinguished from the complex mosaic of groups of yarying validity whicl: comprise the samples studied. These groups, A to D, exsist simultancously over a larger range, 125 to 60, than any other number of groups. Thus within the limitations of the analysis, they are communities approximating societies within the Eucalyptus baxteri association, They are identi- fied by 16 indicator species which vary in frequency between groups (Table 1), For example, vegetation containing Monotaca, Persoonia and Acacia spinesevns, birt nat Fudtonden, would be either Group A or B. lf furthermore it contained Phyllota, Xanthorrhoea and Boronia, but not Ac. oxycedrus or Styphelia, it would he Group A. On the other hand, the remaining 64 species analysed are relatively insensi- tive to the discontinuities that affect the key species, Of these, 22 have a fre- vency greater than 30% (Table 2), their similar group frequencies indicating their unbiased cistribntion throughout the area studied. The appearance of the vegetation is dominated by the five species of frequency greater than 75%, vis, the tree E. barteri, the small shrubs L. myrsinoides and X. australis, and the wider- shrubs ff. sericea and L. virgatus. Superimposed upon this picture of four groups Ww bt R, M. E. WELBOURN ano R. T. LANGE TABLE 1 INDICATOR, SPECIES Species used to identify the groups; relatively high frequencies are in bold faced type. | Frequency %%, Species Overall | - - ' A B C dD Monoloce scopuria (Sm.) R.Br. 22 4 100 0 0 Persoonia juniperina Labill. 8 17 21 5 i) Aciteia spinescens Benth. 2 10 2 0 0 Pultenaea prostrata Benth, ex Hook. f. ) 3 0 13 8 Phyllota pleterandratdea Fo ov. M- i 18 i 100 0) ) " Xanihorrhoen quadrangulota FL v. M. 6 | 82 0 {) ( Roroniu coerulescens F, v. M. 3 14 0 1 4) Hibbertia virgata R, Br, ea DC. var, crassifolia (Benth,) Black 4 5 27 3 () 0) Lhotskya alpestris (Lindl.) Druce 7 25 0 2 tt Acacia oaycedtrus Sieb. ex DO. 2 0 ll 0 0 Thomeasia petalocalys Ky. M- 15 29 2 15 14 Styphelin adacendens R. Br. 5 9 16 2 a Bucalyplus obliqua L' Herit, @ 8 3 28 4 Tsupayon ceratophyllus R.Br. , 57 77 68 100 0 Dodonaea viscoza Jacq. 4 3 0 1 IL Leucopogon collinus (Labill.) B.Br-. Nae ; 49 30 22 a TABLE 2 COMMON SPECIES Species of ovorall frequency greater than 30°, ; frequencies greater than 73%, are in bold faced type. Frequency %, Species Overall |-- : A | 2B co) 6D MYRTACEAE Eucalyptus baxter: (Benth.) Maiden et | j Blakely ex Black 89 55 i 73 74 Leptospermuin myrsinoties Schldl. 89 7 ts 4 7) Calytrix tetragona Labill, (glabrous form) 30. 16 AG 28 a4 EPACRIDACEAK | TLeucopogon virgatus (Labill.) R.Br. 80 61 86 82 44 Lstrolumu conoustephiailes (Sond.) Pov. M. ee Bonth, | 74 TW ta) 70 ; 74 Brachiloma eiliatum Bonth, Gs 42 70 65 65 Leucopoyon erieoides (Sm_.) R.Br. 52 mo, 60 45 7H Epacris impressa Labill. 5 47 | 4 58 BL Aeratriche serrutata (Labill.) R.Br. +7 a 52 GL au Astrolomn huminisum (Cav.) Ridr. 34 | 1 13 38 | fl A, humifusum vor. dentieulatum (R,Br,) Black 31 34 2h 23 44 PROTEACEAR Banksia ornata ¥. v. M. ex Meisn. AL 8&6 tb G4 BY Isopogon cerataphyllus R.Br. (see Tably J) BT fox x x x Banksta marginata Cay. (shrub form) ; 49 63 57 57 25 Bunksiu marginaia Cav. (tree form) a4 1 a2 55 47 LEGUMINOSAE j Acnvin myrtifolia (Sm) Wild. 30 4] A 26 13 Kennedia prostrate R.Br. ex Ait. 30 20 35 ab 2a OTHER FAMILIES | | Hibbertia sericea (R.Br. ex DC.) Benth. §4 80 860002 2 RS Nanthorrhoes musiralis R.Br. 80 Th sb 3 jeetil Hibbertiu siricie (DC.) F. vy. M. 62 83 54 7 | Ad Hypoluena fustigiata R.Br. 60 78 81 62 35: Tetratheca elfiute Lindl. 48 25 | 7 55 as Correa reflext (Labill.) Vont. var, refleace > 40 38 62 | 4g 18 VEGETATION OF STRANDED COASTAL DUNES BA contrasted with common widespread species, aré 42 ypectes with similar frequency to that of the Indicators, 2 to 29%, but occurring haphazardly in the cronps and thus distributed unpredictably throughout the vegetation. Superficially, then, the dey sclerophyll forest studied appears to comprise au open tree canopy oyer a layer of small shruhs dominated by relatively few species. The overall impression is of uniformity. [Mowever measurement revéals that, of the numerous Jow fre- ucney species, some cunsistently associate in socictios or groups. Since it is difficult to assess the validity of the groups other than by field experience, it is necessary tu state at least seme sources of crror inherent in the methods used, Firstly. there ure the hasic errors due to a sampling intensily ot iubout 0-001%, and to operator fallibility in naming and scoring species. Secondly, data based on frequency are liable to misrepresentation because quadrat size affects the results, For example, a large quadrut overemphasizes uncommon species, whilst a small quadrat underemphasizes the importance of species of patchy distribution. This defect can be overcome only with additional data on variables such as density, Thirdly, the analytical method was arbitrarily selected from several which might have been used (Welbourn and Lange, 1967). Its mast serious fault is that in subdividing on a single association index, this method does not dicate the extent or nature of any subordinate association which may exist. That is, there is no indication of the overall confidence with which a particular subdivision is made, Furthermore the assuciation index itself falls short of thearct- ical excellence. Finally, it should be recognized that any such method imposes definite cut-off points to groups hetween which there is some continuity. Never- theless such groups provide at least an abjective basis upon which to classify and mup vegetation variation, Fig, 3 is such a anap which reveals that the groups themselves are move or less geographically restricted. For example, Croup A tends io the north-central ares, B to the south-east, C to the west and south, and D to the cast and north. Sinee these appear not to be chance distributions, there are likely to be enviran= mental factors correlated with them. Several factors will be discussed, to illustrate the sort of hypothesis which may arise from such vegetation mapping. If time was the only factor involved, some groups could be expected to repre- sent stages in a succession, and so to predominate on ranges of similar age. Such trends are apparent; for example, between Naracoorte and Robe, Group D is gradually replaced firstly by A and B, then by C on the younger ranges, However, this evidence rests on the assumption that all other environmental factors aré held constant. This was not so [we the case of soil and topography which were tore Variable than expected, Thus the correlation of vegetation with habitat age canal be tested under the sampling conditions of this study. Similar remarks apply to 4 correlation with aspects of climate such as coastal influence and annual rainfall, From field descriptions of the sampling areas, it is apparent that variations in soil and topogriphy are correlated to same degree sith vegetation variations, For example Group A pceurs on sand ridges normal to the ranges, suggesting an immature profile; the Fucelyntis baxteri here is stunted, and E. diversifolia Bonpl, was observed nearby suggesting a gradation to the solodized solonctz soil common to the north, Crewp B occurs on deep sands in the higher rainfall area: the vewetation is profuse and undisturbed. Group C iy a diversified group which tends to occur in flat, shallow areas, particularly to westward where the ranges appear to have less siliceous sand covering the limestone. The next subdivision of this group (Vig: 2) is on E, obligua, a species known to ovcur on shallower soils. Group D occurs on the eastern sites of ranges well away from limestone, but with) some seasonal watcrtable influence, 24 R. M. E, WELBOURN ann R, T. LANCE Fig. 3. Clearly, profile measurements and more intensive sampling are required to substantiate any correlation of vegetation with soil and topography. Such correla- tion, once established, will be of value in habitat classification and mapping. REFERENCES Biacksurn, G., Bonn, R. D, and Crarxe, A, L. P., 1965. Soil development associated with stranded beach ridges in South-East South Australia. C.S.1.R.O. Aust. Soil. Publ. No, 22. Crocker, R, L., 1944. Soil and vegetation relationships in the Lower South-Kast of South Austrulia, A study in ecology. Trans. R. Soc. S. Aust., 68, pp. 144-172. Sprica. R. C., 1952. The geology of the South-East Province, South Australia, with special reference to Quaternary coastline inigrations and modern beach develoy:nents. Geol. Surv. S. Aust, Bull. No. 29. We tsourn, R. M. E. and Lancs, R. 'T., 1967, Subdividing vegetation on inter-specifie associa- tion. Vegetatio Acta Ceobotanica, 15, pp. 129-136. Witirams, W, T. and Lampert, J, M,, 1959. Multivariate methods in plant ecology. 1. Association-analysis in plant communities. J. Ecol., 47, pp. 83-901. R. M. I. Weteourn and R. T. LANGE PLATE | PLATE 1 General view of the area studied, looking eastward towards Naracoorte, Three ranues. each partly cleared of natural vegetation, are emphasized in the foreground and middle distance with a line along their crests. FURTHER TAXONOMIC NOTES ON THE SPECIES OF MILLOTIA CASSINI (ASTERACEAE) BY RICHARD SCHODDE* Summary One new species, Millotia inopinata, is described and chromosome numbers for it, M. myosotidifolia (Benth.) Steetz, and M. tenuifolia Cass. are reported. Attention is drawn to the reported tolerance to and accumulation of copper in M. myosotidifolia. FURTHER TAXONOMIC NOTES ON THE SPECIES OF MILLOTIA CASSINI (ASTERACEAE) hy Ricwanp Scuoppe* {Read 9 November, 1967] SUMMARY Gne new species, Millotia inopinate, is described and chromosome: rimbers for it, AL. myosotidifolia (Beuth,) Steetz, and Af, tenuifolia Cass. are reported, Attention is. drawn to the reported tolerance to and accumulation of copper in M. miyosotidifolia. Since the publication of my taxonomic account of the genus Millotia Cuss. (Schodde, 1963), further material and information have become available. The ulbbreviations AD, CANE, K, MEL, PERTH used im the text refer respectively to the State Terbarinm of South Australia, Adelaide, South Australia; the herharium. C.8.0H.0., Canberra, A.C.T.; the Herbarium, Royal Botanic Gardens, Kew, England; the National Ierbarium, Royal Botanic Gardens, Melbourne, Australia; and the Wester Australian Herbarium, Perth, Western Australia. {, Millotia inopinata Schodde, sp. nov.—Fig. 1 M. myosotidifolia (Benth. ) Steetz proxima, cujus invelucrum, formam floscu- lorum, et cypselas plerumque strigulosas habet, sed habitu latiore, foliis + asguste lineati-oblanceatis vel puene fliformibus, capitulis + depressis globosis vel ali- quantum turbinatis, bracteis involicralibus ad apices plerumdque ratundatis, corollis aureo-flavis, cypselis aliquantum longioribus in rostris deflexis, papillis cypselarum ad apices integris et + patentibus vix strigosis in rostris cypselarum, et pappi setis 5-8 grosse barbellatis vel sub-phimosis differt, flomer heads. Leaves relatively sparse and appearing confined towards stem hases inear or linear-oblanceate to ulmost * Division of Land Research, C.SJA.G.. Canberre, A.CT, Frans. Roy. Soc. S.A. (1968), Val, 92. SCHODDE R. 28 ant; B. mature Hower head; C, involucral bract; D, forets with maturing A, whole pl > and E, whole style and androecium. ig. L, ue 1 psel: ey TAXONOMIC NOTES MILLOTIA 20 Corollas rather natrow infundibuliform, 24-34 (— 34) mm Jong, het below the throat and deflexed over involucre when mature (central corollas as detlexed as ar somewhat more erect than peripheral corollas), bright golden vellow with the tube becoming brownish after anthesis; corolla limbs of 3 (exceptionally 4) spreading acute lobes $-3 mm long, Stanens 5, with apices becoming exserted to or a little beyond rims of corolla limb; anthers with oblong thecae 3-1 mm lon, the cannective tins extended }-3 mm beyond thecae, Style branches 4-1) mm long, dilated broadly at the apices with conoidal acute appendages, Cypselne lineyr, 6-9 (—10) mm Jong, sharply deflexed from near base of beaks over in- volucre and protruding 3-4 (—5) mm beyond at maturity, variously dark hrawn, sparsely hispidulous-strigillose; cypsela beaks (2k —) 3-5 mm Jong at maturity, compressed and most broadly so on area of deflexion; cypsela papillae = terete lo narrow cylindrical clavate, + 4 mm long, entire rather expanded chtuse ul the apices, colourless more-or-less treuspurent, somewhat appressed and diffuse on body of cypsela, rather spreading, denser, and confined to imuirgius of compressed faces on beak. Pappi of 5-8 twisting more-or-less erect setaz + MX as long as corollas, becoming sparsely coarse barbellate or semi-plumose towards the apices. Chromosome number; n= 8 (fide i. L. Turner). DISTRIBUTION AND ECOLOGY Known presently from three collections: leg. A. S. George 7978: PERTH; 29 miles west of Mt. Magnet, comprising 12 plants, leg. ©. H, Giffins 1540: PERTH, 7 miles south of Wannoo Roadhouse, comprising 7 plants, and the Lype collection, of 14 plants, fram 82 miles north of Murchison river, All localities are in the central west zegion of Western Australia, According to Turner (pers. comm.), the species was looally common in the type locality on coarse sandy soil in open areas with burned-over mulga (Acacia sp.); George records il in sandy loam at a granite outcrop, Flowering: August- September. CHARACTERS AND AFFINITIES M. inopinata is closely related to and sympatric with M. myosotidifolia. The inerphological similarities lie in the long compressed beaks and cylindrical mostly appressed papillae. of the cypselae, the form of the corolla, androvcium and style branches, the narrowly midribbed bracts, and the wholly lanate indumentum. The new species would be identified as M. myosolidifolia when my kev to the species of Millotia (Schodde, 1963) is used but may he distinguished by the following characters against which the contrasting characters of M. sayvsotidifolia are bracketed: eypselae with heaks conspictiously ceflexed and papillae obtuse at the upiccs (beaks erect and papillae minutely notched); corallas deep golden yellow (creamy white rarely creamy yellow); pappus bristles 5-S, + ¢ as long as corolla (bristles usually (15—) 18-25 (— 30), + as long as corolla, very rarely fewer and shorter); mature flower heads depressed globose (cylindrical-clobose ); apices of involucral bracts broad rounded obtuse. often finely or obscurely fim- briate, rarely wttenuated (caudate more-or-less entire, rarely to acuminate or acute); leaves narraw linear-oblanceate to almost filiform (4 —) 1-2 (—3) em long * (KM —) 41 (—1%) mm broad ( narrow to broad oblanceate or occasionally some- what spathulate (%—) 2-4 (—G@) cm long * (1—) 2-5 (—8) mm broad); habit broadly ascending (ascending to erect), The chromosome numbers so tar recorded. n = § in M, inop/nala aud n = 10 dr 11 in M. myasotidifolia (see below), also uppear to be consistent with differ- ences between the two species. Additional counts from the localised populations HI Kh. SCMODDE of M, inopinate and from the eastern Australian populations of MM. myosotidifulia are needed, however, to determine how far these numbers are characteristic at the species, It is noteworthy that a number of the characters distinguishing M. inopinata from M, myosotidifolia resemble those of the eremaean M. greevesii F, Muell., in purticular the habit, shape of leat and Hower head, depauperate pappus, and corolla colour, This may be a case of parallelism in so far as M, inopinata occurs towards the north-west and arid margins of the geographic range of M. myosotidi- folia, There appear to he similar convergences in the characters distinguishing M, macrocarpa Schodde from its nearest relative, M, tenuifolia Cass, There are also similarities between the new species and M. depauperata Stapf, which was reduced to M, myosotidifolia (Schodde, 1963), and to evaluate them Professor B. L. Turner (pers. comm,) has examined the type of M. ¢de- paliperate. This specimen, though like M. inopinata rather than M. myosotidifolia in its few bristled pappus, apparently obtuse-tipped cypsela papillac, and broad- Hpped involucral bracts, is apparently too young and depanperate in its various characters to permit certain identification. Turner considers that in gencral habit it resembles M, myosotidifolia and that its cypsela papillae and involueral bracts. thingl slightly different from much of the material of M. myosotidifolia lw examined, do mately certain determined collections of that species well enough, ww Pritzel 545. In the present circumstances, there seems to he Jittle point in treating, M. depenperata as anything else than a synonym of M. myoasoticdifolia- 2, Millotia myosotidifolia (Benth,) Steetz Chromosome nuinber: w= 10 or 11, reported by B. L. Turner. Vouchers: Turner 5283, 13 miles north of Norseman (1— ca lO); MEL; Turner 5442, 7 miles south of Three Springs (n = 10): MEL; Turner 5482, 30 miles south of Miawlirah (= 11); MEL, All localities are in sonthern Western Australia, Glissett (1966) has recorded tolerance to und accumulation of copper in plants of this species found growing exclusively and abundantly on strips of open ground directly below old copper-bearing dumps from the Ukaparinga coppe1 mine near Williamstown, South Australia. Spectrographie analysis of sample collections showed copper contents of approximately 40,000 p.pan. in the ash of gree plants and approximately 4% in the ash of dead plants, Voucher: |. Moisseeff su, Ukaparinga Coppermine, L miles cast of Williamstown: AD96647104. It is of Interest to mention that the species has also been observed growing abundantly on the rocky dumps from the old Blinman copper mine, at Blinman, South Australia. Vouchers: D, N. Kraehenbuehl 18, rocky hillslopes of the old Blinman Copper Mine at the north end of Blinman; AD 95909049; R. Schodede 997, hillslopes of the old Blinman Mine at the sorth end of Blinman: ADSS9080R4, 3, Millotia macrocarpa Schadde The shape of the corolla is more accurately described as yery narrow infiruli- buliform ef, the original description of the species in which the term “narrow cyathiform™ was intended to apply to the throat only, not the whole corolla. 4. Millotia fenuifolia Cass. Chromosome ntunber; 1 = 13, reported hy B. L. Turner. Vouchers: Turner 5271.5 miles north of Norseman: MEL, Turner 5315, 14 miles west of Southern Cross: MEL: Turner 5516, 53 miles north of Albany; MEL. The counts ure all from yr. tenvifolia and the localities are all in southern Western Australia, TAXONOMIC NOTES MILLOTIA 31 My attention has also been drawn to a recently collected specimen of this species from near Collie, in southern Western Australia (leg. P. G. Wilson 3750: CANB 161828) which is unusual in possessing predominantly 5-lobed corollas, and short laevigate cypsclae 3-4 mm long at maturity, only 3-2 as long as the involucre, and with short beaks 4-1 mm long. It extends the known variation of the species in these characters. ACKNOWLEDGEMENTS Thanks are duc to Professor B. L, Turner, University of Texas, Austin, Texas, U.S.A., and Mr. P. G. Wilson for their collaboration, particularly to Professor Turner for making available the chromosome counts and checking details of Millotia inopinata against the type of M. depauperata at Kew. Professor Turner’s investigations were assisted by a National Science Foundation Grant GBI1216. REFERENCES Burssetr, A. H., 1966. Copper Tolerant Plants from the Ukavaringa Copper Mine, Williams- town. Quart. Geol. Notes no. 18, pp, 1-3 inelL£., publ. Geol, Surv, S. Austral., Dept. of Mines, South Australia, ScHopve, R., 1963. A Taxonomic Revision of the genus Millotia Cassini (Compositae), ‘Trans. Roy. Soc. §, Austral. 87, 209-241, A NEW SPECIES OF PTILOTUS FROM SOUTH AUSTRALIA (Amaranthaceae ) By G. Bent, F.L.S.* (Communicated by Hj. Eichler) [Read 11 April, 1968] SUMMARY A. description and an illustration are given of a new species of Ptilotus, Pt. symonii from South Australia. The type specimens are cited and some critical notes are made on some characters of the new taxon which is compared with Pt. seminudus (J. M. Black) J. M. Black and other species. A recent examination of specimens of Ptilotus sent us from both the State Herbarium of South Australia (AD) and Mr. D. E. Symon (Herbarium ADW) revealed the existence of a hitherto unknown species which is here described as follows:— Ptilotus symonii Benl, sp, nov. (Fig. 1. a-c). Planta perennis, pluricaulis, in statu iuvenili leviter tomentosa, demum subglabra. Rhizoma lignosum adscendens, plusminusve tortuosum, fusiforme, in speciminibus examinatis usque ad 25 cm longum, superne 16 mm crassum. Caules 50 cm et ultra longi, 1-3 mm diametro (in basi lignosa 3,5 mm crassi), virgati, saepius curyato-adscendentes, visu cinerei dein pallido-virides; iuveniles teretiusculi per totam longitudinem pilis crispis crassiusculis nodulosis, ad 1,5 mm longis dense, adulti sulcati sparse induti. Rami valde ramulosi, ramuli tenerrimi subdensi, adscendentes, divaricati ve] patulo-erecti, ad 20 cm longi; basales 7-22 mm distantes, summi approximati, usque ad inflorescentiam dense foliati, Folia permulta minora quin etiam minima, alterna interdum specie secunda (Fig. a), 2-6 mm, raro ad 13 mm distantia, primo modice puberula denique glabra et laete viridia, plusminusve coriacea, integerrima, siccitate marginibus raro subsinuatis, nervo medio subtus vix prominente; omnia inferne attenuata, breviter vel brevissime petiolata (petiolo indistincto ad 2 mm longo), in apice mucronata (mucrone 0,1-0,2 mm longo); maiora 1-2 cm longa et 2-3 mm lata oblongo-lanceolata, minora saepe acicularia. Spicae haud amplae numerosae, hemisphaericae (1,5-2,5 cm diametro) vel subovoideae (2,5 cm longae et 1,8 cm latae), pedunculatae, solitariae ramulos terminantes, raro subsessiles laterales, rhachide breviusculo (0,7-1,3 cm longo) dense villoso, pilis plusminusve flexuosis nodosis, 1-1,5 mm longis. Flores (15-25, raro ultra) haud dense congesti, demum stramineo-flavescentes, albido-pilosi. Bractea bracteolaeque acutae integerrimae, pilosae, extus pilis articulatis, apicem attingentibus quin etiam paullulo excedentibus obsessae, uninervae, post lapsum perianthii superstites, inaequales: Bractea inferior rigida, subcordato- lanceolata 2,5-3 mm longa et 1-1,2 mm lata, fuscescens, nervum medium versus obscura, in dorso omnino pilosa, pilis articulatis, circiter 1 mm longis densius vestita. Bracteolae 2 distincte maiores subcarinatae, membranaceae, tenues, ovato-lanceolatae 4-6 mm longae et 2 mm latae, tantum nervum medium fusces- centem versus pilos circiter 1,5 mm longos gerentes, lateribus glabris generaliter incoloratis, hyalinis, nitentibus, perianthio adpressae. * Botanische Staatssammlung Miinchen, Germany. Trans. Roy. Soc. S.A. (1968), Vol. 92. 34 G. BENL Fig. 1. Ptilotus symonii Benl: Pedicelled spike (from AD 96131068) with the tiny upper leaves (Fig. a); perianth (from AD 968020399) with stamens and staminal cup spread open, inner view (Fig. b); pistil (Fig. c). ae a A NEW SPECIES OF PTILOTUS 35 Perianthium pentaphyllum elongato-erectum dein subcampanulato-patens, basim constrictam cartilagineam versus sensim indurescens, tubum angustum cylindraceum 0,8-1,2 mm longum, extus pubescentia plusminusve absconditum formans. Tepala elongato-linearia vel lineari-oblonga, anguste hyalino-marginata, trinervia—nervis lateralibus (superne saepius indistinctis) areolam medianam impellucidam, incrassatam, coloratam includentibus—, ecarinata, integerrima, truncata, apicibus inconspicuis, fere nudis 3-4 mm longis, rufescentia dein viridi- flavescentia, pilis dorsalibus strictis, rectis, articulatis, in articulis breviter verti- cillatis ad 9 mm (in ima basi circiter 1 mm) longis, imprimis dimidio inferiori areolae medianae tepalorum orientibus, sed apicem superantibus adpresse denseque obtecta, pilis marginalibus tenuioribus, brevioribus sparse ciliata, intus demum albido-laevigata, inaequalia: 2 exteriora 9-11 mm longa et 0,6-0,9 mm lata, marginibus (ca. 0,15 mm) in apicem paulo contractum, eroso-denticulatum, pilis dorsalibus longe (ad 2 mm) superatum transeuntibus (Fig. b); 3 interiora paullum breviora, sed angustiora (0,4-0,6 mm lata) et acutiora, marginibus superne involutis, inferne pilis crispis, nodulosis, uni sive duobus Jateribus tepali, praecipue autem margini tubi perianthii orientibus et introflexis, plusminusve copiosis munita. Stamina staminodiaque 5, in floribus examinatis semper 3 fertilia et 2 abortiva, basi modice (ad 0,2 mm) dilatata flamentorum applanatorum cupulam membranaceam, hyalinam, tubo perianthii adnatam formantia, anulo minimo (0,15-0,3 mm) libero integro, pseudostaminodiis nullis, filamentis ligulatis superne subulatis, interdum brunnescentibus, circiter 2 mm longis, abortivis rudimentis antherarum coronatis vel anantheris brevioribus (0,5-1,5 mm); antherae bilo- culares, flavae, lato-ellipsoideae 0,3-0,4 mm longae et 0,2-0,25 mm latae, dorso affixiae. Ovarium subclavatum, conspicue stipitatum 2,5-3 mm longum (stipite circiter 1 mm longo incluso) et 0,7-1 mm latum; stylus sicut ovarium regulariter glaber- rimum 1,3 mm longum et circiter 0,1 mm diametro, plusminusve excentricum; stigma inconspicuum haud distincte capitellatum, papillosum. Holotype of species—5 miles south of Koonalda Homestead (east of Eucla), south-western region of South Australia; D, E. Symon, No. 4684, 21.11.1967, AD No. 968020399. Isotypes—Idem, A, ADW_. CANB, K, LE, M, TI, UC, W. Further collection—Other material of this taxon had already been collected by P. G. Wilson (No, 1635) in North West Plains, ca. 40 km East of the Western Australian border, off Eyre Highway, 13.1X.1960 (AD 96131068), Our description is based on Wilson’s specimens, too, which, therefore, may be regarded as Paratypes. Habitat—Symon’s plants were growing “in open Mallee scrub” and “mostly found in the twiggy remains of dead or dying plants of Westringia rigida’. Wilson’s material had been gathered in Acacia woodland. Characteristics—The new taxon superficially approaches the South Australian Pt, seminudus (J. M. Black) J. M. Black as regards the general form and colour of the spikes. In this species, too, the outer perianth segments bear a truncate and denticulate apex, and the inner tepals are distinctly exceeded by dorsal hairs. The stems arising from a strong rhizome (see “Australian Plants” 4: 117, 1967) are pubescent in about the same way, when young. In Pt. seminudus, however, the stems and branches are constantly shorter and thicker, the leaves (especially the radical ones) considerably larger, the spikes richer, the bract and bracteoles nearly equally long and more acute; the bract being usually less hairy. Moreover, the tepals are longer, the inner ones more 36 G, BENL narrowed towards the apex, and the points of the outer segments are not so markedly overtipped by the erect dorsal hairs. In both species these hairs are articulate and of nearly the same length, but in Pt. symonii they primarily rise from the lower half of the tepals thus covering the perianth tube, while in Pt. seminudus the basal part of the segments looks naked, revealing the hirsute subglobular tube. In this plant we find inflexed hairs inside the perianth arising from the margins of the inner tepals above the tube, whereas in Pt. symonii a woolly pubescence of the inner segments takes its origin from the edge of the perianth tube, too. In addition to these characters, numerous details of the reproductive organs diverge: e.g. two stamens only are fertile in Pt. seminudus, its staminal cup shows a comparatively higher free ring, and between two filaments much broadened at their base you may find a small lobe, at times; the ovary being pilose in summit. Pt. seminudus, therefore, differs markedly from our taxon. As a striking particular feature of diagnostic importance are to be regarded the numerous and uncommonly small narrow leaves densely borne along the branches: and branchlets of Pt. symonii. Except for the bushy and extremely branched Pt. parvifolius (F. v. Muell.) F. v. Muell., no other representative of the genus is characterised by such leaflets. But these two plants are quite unlike one another. An additional trait of diagnostic interest is given by the relatively long and strict dorsal hairs of the intense pubescence in the perianth, distinctly exceeding and concealing the outer as well as the inner tepals, which look tapering because of this. To a lesser extent, we find a similar appearance in Pt, arthrolasius F. v, Muell., a subshrub with a yellowish pubescence, and still more in Pt. eriotrichus (W. V. Fitzg. ex Ewart & White) W. V. Fitzg., another frutescent species with a dense greyish tomentum. The shorter hairs in the small flowers of Pt. forrestii F. v. Muell. and the longer ones in Pt. villosiflorus F. v. Muell. are less distinctly articulate, and mostly lack the verticillate toothlets at the nodules, as is the case in Pt. arthrolasius, too. The tepals of the narrow-spiked Pt. lanatus Cunn. ex Mog. (including var. glabrobracteatus Benl) are surpassed by short bristly and thickish hairs, those of the long-spiked P#, leptotrichus Ben] by a tuft of relatively few articulate hairs. In Pt. albidus (C. A. Gardn.) Benl, Pt. brachyanthus (F. vy. Muell. ex Benth.) F. v. Muell. and Pt. petiolatus L. Farmar the perianth is more or less completely hidden among dorsal hairs forming an intricate wool. Each of the cited species has a dissimilar appearance of its habit, of its leaves, spikes or floral organs, and there is no doubt left as to the specific nature of our well distinguishable taxon: Pt, symonii does not at all agree with any of the species hitherto described. Name—The plant is named in honour of Mr. David E. Symon, B.Ag.Sc., Botanist at the Waite Agricultural Research Institute. Mr. Symon is one of the collectors of the new species, drew my attention to it, and supplied us with sufficient material. THE YELLOW-EYE MULLET AGE STRUCTURE, GROWTH RATE AND SPAWNING CYCLE OF A POPULATION OF YELLOW-EYE MULLET ALDRICHETTA FORSTERI (CUV. AND VAL.) FROM THE COORONG LAGOON, SOUTH AUSTRALIA BY J. A. HARRIS* Summary The "Coorong mullet" spawns once per year, from January to early April. Males begin maturing a few months before the females are running ripe. Seven stages can be recognized in the gonads of the female and measurements of the diameter of the ova in each stage are given. The smallest fish with mature gonads were found to be 21 cm (males) and 23 cm (females), measuring from the tip of the snout to the caudal fork. The ages of the fish were determined from a study of otoliths and this method was supported by the Petersen method of analysis of frequencies of lengths. The mean lengths attained by the "Coorong mullet" during their first four years are 14, 21, 26 and 31 cm respectively. Characteristics of the "Coorong mullet" are compared with both the eastern and western races of the yellow-eye mullet (See Thomson, 1957a; 1957b). It is concluded that the "Coorong mullet” has the characteristics of the eastern race. THE YELLOW-EYE MULLET AGE STRUCTURE, GROWTH RATE AND SPAWNING CYCLE OF A POPULATION OF YELLOW-EYE MULLET ALDRICHETTA FORSTERI (CUV. AND VAL.) FROM THE COORONG LAGOON, SOUTH AUSTRALIA by J, A. Hanats* [Read 9th May, 1968] SUMMARY The “Coorong mullet™ spawns onee per year, from January to early April. Males begin maturing a few months before the females are running ripe, Seven stages can he recognized in the gonads of the female and measurements of the diameter of the oya in each stave gre given, The smallest fish with mature gonads were found to he 21 em. (males) and 23 em (females), measuring from the tip of the snout to the caudal fork, The ages of the fish were determined frow a study of otoliths and this micthocd was supported hy the Petersen method of analysis of frequencies of lerigths, The mean lengths attained by the “Coorong mullet” during their first four years are 14, 21, 26 and 31 em respectively. Characteristics of the “Coorong mullet” are conypared with both the eastern and western races of the yellow-eve inullet (See Thomson, 19574; L957b). It is concluded that the “Coorong mullet” has the characteristics of the eastern race. INTRODUCTION The yellow-eye or fresh-water mullet Aldrichelta forsteri (Cuvier and Valen- ciennes) occurs in coastal waters of all Australian States except Queensland. Thomson (1957a) studied the yellow-cye mullet of Western Australia and com- pared it with those of Victoria and Tasmania. However, very little is known about the rate of growth and spawning cycle of the yellow-cye mullet of South Aus- tralia, especially from the Coorong lagoon (see Fig, 1). The yellow-eye mullet is very common in South Australia, particularly in coastal brackish waters, It is the principal species of mullet sold commercially in this State. The only records of the commercial fishery available during this study were very rough estimates of the total weight of mullet sold in South Australian markets from 1951-62: 1951-52 770,000 Yh 1957-58 560,000 Ib 1952-53 500,000. lb 1958-59 900,000 Ib 1953-54 500,000 Ib 1959-40 649,405 Ib 1954-53 431,220) Ib 1960-61 612,000 Tb 1955-56 550,000. Ib 1961-62 675,000 Ib 1956-57 650,000 Ib The total weight of rnullet handled by the Sonth Australian Fisherman's Co-operative Limited (S.A,F.C.O.L.) from July Ist, 1961 to June 30th, 1962, was 584,984 lb. Approximately 75 per cent of this came from the Covrong Iagoon, the other 25 per cent from the shallow coastal waters of South. Australia. The fish * Formerly Department of Zoology, University of Adelaide. Present address, Department of Zoology, University of Queensland, Trans, Roy, Soc. $.A, (1968), Vol. 92. 38 J. A. HARRIS \ | SOUTH ! AUSTRALIA lograths Flat Fig. 1, Location map of the Coorong lagoon. from the Coorong lagoon and those from the coastal waters are sold under different names at $.A.F.C.O.L. as “Coorong mullet” and “sea mullet” respectively. This investigation only deals with the “Coorong mullet”. The legal minimum length for yellow-eye mullet in South Australia is 7 inches (17-8 cm), total length. As so little is known about the commercial fishery in this state, the effectiveness of the present minimum legal length is open to uestion. The small amount of data pertaining to fish below 17-8 cm is a eficiency in the present investigation; however, the commercial catch (above 17-8 cm) has been covered as thoroughly as possible in the time available. THE “COORONG MULLET” FISITERY The name “Coorong” applies to a long, narrow lagoon and associated shallow lakes, paralleling most of the upper south-east of South Australia. This area is divisible into two sections, the zone of permanent lagoonal water to the north and the shallow, non-perennial lake to the south, In Fig. 1 the lagoon is enclosed by the rectangle. The “Coorong mullet” fishery is only concerned with the lagoon. YELLOW EYE MULLEW PROM Tif COORONG 39 The Coorong lagoon extends southwards from the mouth of the River Murray a distance of 68 miles and reaches a maximum width of 2% miles. The maximum depth reached in winter near Salt Creek is about 4 feet, but the average depth is only 6 feet. The mouth of the Murray provides the only connection between the lagoon and the Southern Ocean. The mouth is about 300 yards in width and benaiits ob a dal channel through the unconsolidated dunes of Younghusband Peninsula. Mesh and gill nets are the principal means of catching the “Coorong mullet” and regulations restrict the size of the mesh and the lengths of these nets, The influence that regulations on nets have on the yellow-eye mullet fishery in Western Australia has been diseussed by Thomson (1957a). Regulations during the period of this investigation restricted the length of any net to 60 yards, the length of rohit nets to 35 yards and the size of the mesh used in all nets to not less than 2 Aches. The “set net” is by far the most common method used for catching nvullet. Lengths of mesh or gill nets are anchored to stakes across the current. They are checked at regular intervals and the “gilled” fish collected. “Ring netting” is often used with great success on calm, still nights. The fisherman lays a ret, one end of which is attached to a buoy. around 2 school of fish. A few fishermen use beach seine nets from sandy beaches or sand bars. Unfortunately, these beaches are rare due to the recky nature of the bottom. During the summer months sand banks are exposed jn the centre of the lagoon and from these banks beach seining is very successfully carried out. Most of the mullet from the lagoon is sent te Meningie to he packed in ice prior to transport to $.A.F.C.O.L. SAMPLING METHODS Weekly samplex of “Coorong mullet” were collected from $.A.F.C.O.L. Adelaide market for one year from February, 1962 to January, 1963. Over 100 fish were selected at random each week: and their lengths were measured by punctur- ing, holes in & celluloid strip, mounted on a centimetre rule, with a hootmaker's awl (see Scott, 1954), From this random sample approximately 12 fish were chosen, representing the complete ranve of sizes of fish available, for further detailed examination in the laboratary. This detailed examination consisted of: (a) collccting scales and otoliths and stonng them in labelled envelopes, and (b) macroscopic examination of the gonads, after which they were fixed in 5 per cent formalin for microscopic exami- nation. On one occasion on October 7th, material was obtained from the Coorong lagoon by seine netting in shallow water with a net of 1 inch mesh. Only 31 fish tunging from 5-5 cm to 15-8 em were caught. During the year the Jengths of 6,054 fish were measured and 631 otoliths were examined in order to determine their age and rate of growth, In addition, the macroscopic and microscopic conditions of about 600 gonads were noted. All measurements OF lengths were taken from the tip of the snout to the eandal fork (I..C.F,) onless otherwise stated. DETERMINATION OF AGE BY MEANS OF OTOLITHS Otoliths were found to be a good index of age and were tasy to handle anil examine. Both otoliths in any one fish were found to be identical. No preparation prior t) the reading of them was required (see Dakin, 1939), Regular opaque bands and translucent bands are cbyious even in the largest otoliths. The two eto- liths collected from cach fish were read together using a Jow power binocular microscope. 40) J. A. HARRIS As the otolith bands are formed by regular coneretions (Hickling, 1931), the alternate opaque and translucent zones must at some stage actually constitute the margin of the otolith. All the otoliths collected from each weckly sample of fish were placed into two groups, those in which an opaque band constituted the margin and those in which a translucent band constituted the margin. Transitional or doubtful ones were discarded and these accounted for about 20 per cent of the otoliths. The occurrence of opaque margins was plotted as a percentage of all oto- liths examined during each month of the year from February, 1962 to January, 1963. The results are shown in Fig. 2. Fig. 2 shows that the lowest percentage of otoliths with opaque margins aceurs during August and September while the highest percentage occurs in January. Translucent bands occur during the winter months and opaque bands form during the summer months. A similar condition was observed for garfish in South Australia by Ling (1958), for plaice (Molander, 1947) and for the tiger flathead (Dakin, 1939; Fairbridge, 1951). Thomson (1957b), has shown that the annuli in the scales of the yellow-eye mullet from both Western and Eastern Aus- tralia are formed when growth recommences after the winter cessation. The first translucent band is Jaid down during the first winter, approximately 6 months after spawning. The first opaque band is laid down during the first summer after spawning. Hence, the number of opaque bands represents the actual 60 70 wy z iC) cc a =60 ld re] reg ? a o50 x = 5 ua E40 a rs) = fa) 630 ub 0) na Fig. 2. Monthly percentaye =20 incidence of otoliths with rs) opague margins for the 12- hi month period February, 1962 & to January, 1963. 10 0 FEB MAR APR MAY JUNE JULY AUG SEPT OcT NOV DEC JAM YELLOW EYE MULLET FROM THE COORONG Al age of the fish in years. For this reason the opaque zones were counted and the age groups denoted as 0+, 1+, 24+, etc., the numbers representing the age of the fish in years and the + — sign signifying an additional unknown number of months (but less than 12) above the preceding numbers. The results of the otolith readings are given in Table 1, the number of fish in each age group are arranged with their corresponding lengths. Ling (1958) discusses in great detail the ways of confirming the otolith method as a means of determining the age of fish. He discusses Graham’s original five tests to validate the scale method (Graham, 1929). In this investigation Peter- sen’s method of interpreting the modes of a graph of length distribution as the model lengths of successive age groups is teed te support the data obtained from otoliths. Monthly histograms of the frequencies of lengths of fish measured each week by the celluloid strip method (see Scott, 1954) are shown in Fig. 3. In the major- ity of histograms a definite mode occurs about the lengths of 20-22 cm, a less TABLE 1 Length frequencies of males and females with their ages, as Jetvermined from otoliths, for all fish examined in the laboratory from February, 1962 to January, 1963. Age groups Male Female Length (erm) ot | 14 | 2+ 34 oy | 14 | ot, pe | at | oe 4 1 i 1 1 2 25 . 1 | 10 36 ce) S 3 _ <7 ee ur an bo Total 2; 10 125 | 36 6 14 143 141 32 2 o “| AUG > - i] Le) - Z Z & Zoo 2 2 o g au uw m0 JuLy = 0 ue iw 49 J. A. HARRIS FIG 4 JAN © here) ig SEPT jag b o a eee kh hk LENGTH CCM) Fig. 3. Monthly frequency distribution of lengths of fish sampled each week for the 12-month period February, 1962 to January, 1963. 1 a n DEC SEPT JULY JUNE 30 450 OVA DIAMETERS (ju) Fig. 5. Monthly frequency distribution of diameters of ova examined from fish selected at random over the 12-month period February, 1962 to January, 1963. (Note: Hickling’s “reserve fund eggs” were only measured in February ), YELLOW EYE MULLET FROM THE COORONG A definite one about 24-26 cm, and the trace of a third mode about 30-32 em. By comparison with the different age groups obtained using otoliths, these modes cor- respond to the age groups 2+, 3+ and 4+ respectively. Theretore, agreement between the Petersen method and the determination of age by means of otoliths is quite good. It can be seen that the majority of fish are in the age group 24- with a smaller number in the age group 3-+ and very few in the age group 4+. ia FREQUENCY w Fig. 4. Frequency dis- tribution ot lengths of fish caught October 7th with a net of 1 in, mesh, a °o i] ’ LENGTH (CM) Fig. 4 shows the length frequency histogram of 31 fish caught on October 7th with a net of 1 inch mesh, Only one mode at 15 cm is obvious, but probably there is one at 8 em. These modes correspond to the age groups 0+ and 1+ respee- tively. In general, it may be stated that the determination of age by means of otoliths is supported by the results of the Petersen method. RATE OF GROWTII (a) Using otoliths and measurements of length Table 1 shows the frequencies of lengths of males and fernales with their ages, us determined from otoliths, for all fish examined in the laboratory from February, 1962 to January, 1963, An analysis of these results was carried out to test the two null hypotheses implicit in these results; (1) that there is no differ- ence in the lengths of the fish in the age groups 0-+ and 1-++, 1+ and 2+, 2+ and 3+, 3+ and 4+, and (2) that there is no difference in the lengths of a male and female in the age groups 0+, 1+-, 2-+ and 3+. These null hypotheses were tested by the “t” test (Snedecor, 1956). The first null hypothesis was disproved for all cases, but the second null hypothesis was not disproved for all cases. The differ- ences between the lengths of taale and female fish in groups 0-+- and 1+ were not significant, but those for groups 2+ and 3+- were highly significant. The mean lengths attained by the female “Coorong mullet” during their first four years are 14-2, 21-7, 26-1 and 30-9 cm respectively, Rapid growth occurs during their first lwo years, but then falls off during their third and fourth years. The mean lengths attained by the males during their first three years are 13-7, 21-1 and 24-5 cm respectively. Growth is again rapid during their first two years, but during their third ycar the mean length attained is 1:6 em less than that of the females. After the first two years, the rate of growth of males (using only length as the measure of growth) is much slower than that of females. 44 J. 4. HARRIS The two largest fish examined during this study were both females, approxi- mately 35 em Jong. Only one male fish exarnined was. found to be longer than 26 em and it was 30 em in length. These data indicate that the females grow tu a larger size than the males. (ly) Using the Petersen method As mentioned earlier, the histograms of frequencies of lengths given in Fig. 3 show two definite modes occurring about the lengths of 20-22 em and 24-26 cin with perhaps the trace of a third one about 30-32 cm. These modes are considered to correspond to the age groups 24-, 3+ and 4+ respectively, The evidence far mouul progression when following the progress of one particular mode in the ovurse of time is very good, If March is taken as the tine that spawning activity is at its peak, the histe- gram for March inay be said to represent the Frequencies of Jengths at the begin- ning (or end) of one or more complete yeurs in the life-cycle of the fish. Definite modes occur about the lengths 20-5, 24°5-26 and 30:3 um respectively in the March histogram. It is easy to trace the mode about 20-5 em throughout the year. In April and May, just prior to winter, a $mall amount of growth is apparent. In May the mode is obvious about 21°35 em. Over the winter months June to Sup- tember this mode about 21°35 em is maiutained, indicating no growth, or very little (as also indicated by the study of otoliths). From October to January the mode shifts rapidly from about 22 em in October, to 23-5 cm in November, 24-5 cm iv December and 24 -5-26 em in January, The histogram for February of the previous year shows a very good mode aboul 26 cm and supports the evidence for mudal progression over one year being approximately 20+5 to 26 om. This corresponds. ta a growth of 5-5 em which is very close to the annual growth of a 2-year-old “Cou- rong mullet” as sugeested by the study of their otoliths, Over the same time period possibly the mode about 24:5-26 em gives way to modes about 28:5 em in Qcto- ber and November, 29-5 em in December and 30°5 um in January. This corres- ponds toa growth of 4-5-6 em which is roughly equivalent to the rate of growth of ad-year-old fish as suggested by the study of otoliths. THE SPAWNING CYCLE ‘The methods by whicli the spawning cycle may be studied inelude a survey ol the seasonal variations in the conditions of the gonads. This involves a macra- scopiy examination of the appearance of the gonads and a microscopic examinn- lion of the ova. In most investigations (Clark, 1925; Fairbridge, 1951; Scott, 1954) the female eycle has heen studied in more detail than the male cycle. In the present investi- gation the male cycle was examined in less detail because it was found hard to distinguish the stages of the testes and because the condition of the ovaries changed much more noticeably than that of the testes, A month by month com- parson of the macroscopic canditions of the avaries was noted and this was later verrelated with a microscopic examination of the ova, including measurements of their size. (a) Macroscopic examination of gonads Hjort’s classification of the gonads of herring (Hjort, 1910) has heen adopted with cvertain modifications to suit this particular specivs. Seven stages in the ovaries of the “Coorong mullet” were distinguished while the testes were dis- tinguished only us immuture or riature. The classification of each of these seven staves of the ovaries is given in Table 2, YELLOW EYE MULLET FROM THE, COORONG 45 TABLE 2 Lengths and stage of maturity of 348 females sampled during January, 1963- Stages of ovaries Length (em) | Per cent Immature Adolescent Maturing Spent maturing | and spent, 19 2 | | 0 2) 2 2 | 0 2 2 2 | 0 22 3 2 i () 23 12 ls 2 1 z4 14 55 25 J 27 25 Li 43 a0 47 26 3 21 26 3 49 27 2 17 89 28 4 4 i 70 2a. 9 100 30) 5 10 al 3 100 Total 52 148 | idl 7 43 ! Stage I. Immature virgins. Ovaries small, thread-like and translucent. Stage ll. Adolescent, Ovaries larger, walls taut, extending half-way into the body cavity, Stage IIT. (a) Maturing virgins or (b) recovering or resting (spent) fish, Colour ranges from pink at early pigmentation to dark red at a later stage, In (a) walls of ovaries are taut but in (h) walls are ilaccid, Ovaries extend further into body cavity. Stage IV. Maturation continuing. Ovaries larger, red-yellow in colour. Walls taut and not quite transparent so that eggs can only be seen with difficulty afler cutting the wall. Stage V. Mature, Ovaries greatly enlarged from stage IV, occupying the whole of the body cavity, Yellow-orange in colonr, walls taut and transparent with eggs clearly visible, Stage VI, Running-ripe. Yellow-orange in colour with large eggs being extruded when slight pressure is applied to the abdomen, Stage VII. Spent. Oyaries dark red aud Haccid. This stage eventually passes into stage IIL from which it is hard to distinguish. The seasonal changes in the gonads may be summarized as follows, The testes of the immature fish are white, thin bodies, triangular in section. During maturation they become thicker and rounder, protruding further into the body cavity and milt being exuded if the walls are ruptured. During the months of May, June and July no mature testes were noticed. For the rest of the year there were always a few mature testes present and during the spawning season January to March, over 50 per cent of all testes examined were mature, (See Table 3 for the percentage of maturing males during January), When the inyestigation started in February, 80 per cent of the ovaries exumined were mature (stages IV, V and VI) with 25 per cent of these being in the “ripe” condition (stage VI). During March a similar set of conditions 46 }, A. HIARAIS continued with a few spent ovaries (stage VIL) becoming apparent. However. during April and May spent ovaries became more evident and the last mature ovary (stage VI) was found on April Eith, except for one ovary at stage V found on May 2nd which appeared to be abnormal, only one branch having developed, During the months of June and july, except for one fish on July 25th which was at stage IV, all ovaries were classified as belonging to stages II and III. Approximately 85 per cent of the ovaries examined during August, September and October were at stages IL and IlI. Only 15 per cent were more advanced than stage IJI and it is possible that they were carly spawners. During November and December, however, ovaries at stage FV became more evident with an occasional stage V until in Jannary ovaries at stage V were common with several at stage V1. (Sce Table 2 for the percentage of maturing females during January. ) (h) Microscopic examination of ovaries Several ovaries chosen at random, were set aside each week, preserved in 5 per cent formalin and stared in labelled vials. These ovaries were examine: under the microscope al a later dite. To measure the diameter of the ova a small piece of ovary was teased out onto a microscope slide, mounted in 5 per cent formalin and covered with a cover slip. The diameters were measured using a micrometer eye-piece and a micro- scope, ; All eggs were measured along whichever axis of the egg lay parallel to the cross-hair of the micrometer eye-piece. Clark (1925) has shown that this methoul gives a reliable estimate of the frequencies of diameters of the ova. Throughout the investigation group A eggs were not measured. except in February, as they were far too numerous in all of the oyarian stages. Consistent with the classification of Hickling (1930) and other workers, the ava of the “Coorong mullet” were classified into four groups. Group A — Immature ova. Very small, angular and colourless with nucleus elearly visible. They are Hickling’s “reserve fund eggs” and measure up to 126) », Group B — Maturing ova. The smallest eggs in this group are becoming round and opaque due to accumulation of yolk granules, while the largest ova are quite round and opaque. Size varies from 120 p to 255 py, Group C — Mature ova. Large round opaque ova, pale yellow. Size varies from 255 pe to 380 p. Group D — Ripe ova. Very large round opaque ova, yellow. Size varies from 380 » to 600 ». This stage merges with group C, but is distinct from group C by size and coloration, The microscopic characters of each ovarian stage were established as follows: Stage I. The ova are all typically group A, with none larger than 60 p. Stuge II. As for stage I, but an occasional ova of group B may be present No ova larger than 135 p, Stuze DT. Ova of groups A and B are very mimerous with an occasional group C. Usually ne eggs larger than 255 yu. Stage TV, Ova of groups A, B and © are mimerous but group B is usually nore numerous than group C. Stave V. Ova of groups A, B and C are numerous with a few of group 1). Group C is usually more numerous than groups B and D, Stage VI. As for stage V, but the ovaries are much larger with ova of croup D most numerous and lying free in the lumen. Groups A, B and C ere confined to a layer inside the oyary wall. YELLOW EYE MULLET FROM THE COOKONG AT Stage VII. Ova of groups A and B are both numerous. An occasional group D ovum is present undergoing breakdown. The frequencies of diameters of the ova examined were plotted as monthly histograms and are shown in Fig. 5. Throughout the year there is a residue of Hickling’s “reserve fund eggs” which were too numerous to measure every month and are shown only in the February histogram. From Fig. 5 it can be seen that numerous ova of group B also are present throughout the year. From May to October (except in August when the ovary examined may have been that of an early spawner) ova-of groups A and B are by far the most numerous. During November, however, an increase in the diameter of the ova present is obvious. As maturation continues through December, the diameters of the oya continue to increase until “ripeness” (stage D) is obtained. Ripe eggs are present in greatest numbers during March, with large numbers also present in February and April. Spawning is thus at its greatest intensity during late summer. A sharp drop in the overall diameter of the ova marks the spent and recovering stages, during April and May. It is followed by a period of quiescence from June to October. SIZE AND AGE AT FIRST MATURITY The gonads of 122 male and 348 female fish were examined; and their lengths taken, during visits to §,A.F.C.O.L. in January, 1963, The females were classified arbitrarily as immature (stages I and IL), adolescent (stage III), maturing (stages {V, V and VI) and spent (stage VIL). The males were classified under two divi- sions, immature and maturing. Tables 2 and 3 show the lengths and stages of maturity of these males and females. Table 2 shows that 15 fish with lengths of less than 22 em were examined and none showed signs of maturing, Sixteen per cent of the fish 23 cm in length were maturing and spent, The percentage rose to 100 per cent for all fish 29 cm and above. Females, therefore, do not reach maturity before 23 cm. Table 3 illustrates that only 3 males 21 cm long were examined and that one of these was maturing. Because so few fish of this length were examined it cannot be proved that males mature at 21 cm. However, 19 fish of length 22 cm were TABLE 3 Lengths and stage of muturity of 122 males sampled during Jamnary, 1963. Stages of testes Length (em) Immature Maturing Per cent raaturiag 20) | 4 0 21 ' 2 i 33 22 1 A 42 24 22 28 57 24 Il a4 6a 25 I 7 87 26 2 100 Total 51 71 61 examined and uf these fish 42 per cent were maturing, Males, consequently, are definitely mature by 22 em. Correlating the lengths at which the females and males first mature with the data obtained from the otoliths, it is concluded that the “Coorong mullet” attain maturity during their third year. 4h J, Av HARRIS DISCUSSION The “Coorong mullet” grows to a relatively small size of 35 cm. At first the growth, using only length as a measure of growth, is rapid. The mean lengths attained during their first and second years are 14 and 21 cm respectively, After this period, however, there is a marked slowing down in growth reaching a mean length of 26 and 31 cm during their third and fourth ycars respectiyely. This slow- ing down is more prouounced in ‘the males which do not grow as large as the females. TABLE 4 2 Lengths (em) attained cach winter average by the Western and Mastorn races of theYeollow eye Mullet (‘CThomson, 1957b) and the mean lengths attained by the “Cusrong mullet’? over aa equivalent time poring, l 1 Western i ! and Easter I i IL Lit TV ¥ VL Vil vares | | i ——— = = nl, ee ne “Coorong | 4 ' roullet™ | Of Af BOB Western fish | a 18.19 | 24.25 | 29-32 | 32-25 | 38 49 Eastern fish 5 | wus | i921 | 2427 9 a0 “Oourung rules” | T(t) 14 21 26006¢,~—C 3] Table 4 compares the growth of the western and eastern races of the yellow- eye mullet with that of the “Coorong mullet”, The data for the western and eastern races is taken from Thomson (1957b), The year-classes designated in Thom- son's paper refer to the length attained each winter average (the western stock being winter spawners). Over an equivalent time interval the “Coorong mullet” rows to a mean length of 7(?), 14-2, 26 and 31 cm by their first, second, third, ourth and fifth winters respectively. Tt can be seen that their growth rate agrees closely with that of the Victorian and Tasmanian yellow-eye mullet which com- prise the majority of fish referred to as “eastern” by Thomson (1957b), The “Coo- rong mullet” grow at a slower rate than the Western Australian yellow-eye mullet and do not become as large. The spawning period Jasts for 3-4 months from January to early April. This proves that the “Coorong mullet” have the characteristics of the eastern race rather than the western rave of yellaw-eye mullet (see Thomson 1957b, p, 12), Spawning probably takes place in the Coorong lugoon which would provide a very productive area in which the newly hatched fish can grow rapidly with little chance of a shortage of food occurring. The Jargest ova of the “ripe” fish were upproximately 600 , in diameter which is relatively small compared with whiting 980 » (Scott, 1954) and garfish 3,500 » (Ling, 1958). The smallest fish recorded with mature gonads were 23.cm for females and 21 em for males, They mature during their third year or by their third winter. The “Coorong mullet”, therefore mature a little earlier and at a slightly smaller size than the Western Australian yellow-eye mullet studied by Thomson (1957a). In addition, due to the spawning periods of the two raccs of fish occurring at differ- ent times of the year, the summer spawning “Coorong mullet” reach maturity faster than the winter spawning western racc, Comparing the commercial catch of “Coorong mullet” with the catch of yellow-eye mullet in Victoria, some interesting facts emerge. In Victoria the fish YELLOW EYE MULLET FROM TITE COORONG 49 reach the minimum legal length, 9% inches total length (21:3 em L:C.F.), from 2h to 32 months prior to spawning (Thomson, 1957a). In Sauth Australia the tinimum Jegal length is 7 inches total [length (15 em L.C.F.) and the fish attain this size approximately 12 months before they have spawned for the first time. South Australia’s minimum legal Jength is obviously too small and at first glance the “Coorong mullet” fishery appears to be in jeopardy. However, Fig. 3 shows that no fish below LT om were sampled at S.A-F.C.O.L. The majority of fish, in fact. were well above 20 cm. This biased sample was due to the size of the mesh of the nets used by the commercial fishermen. In South Australia the minimum legal mesh size during the study was 2 ches, compared with 2M inches in Victoria, Thomson (1957a) showed that, as far as vellow-eye mullet are con- cerned, any mesh from 2 inches to 34 inches allows the majority of the catch to be above the length of 20 crn. Thomson (19571) discusses the management of the yellow-eye mullet fishery heth in Eastern and Western Australia and concludes that the existing manage- ment policies are elective in maintaining stocks. Whetber this ix alsa tre fur the “Coorong mullet” fishery cannot be determined yet due to lack of information, both biological ancf stulistical. A study of the length/weight relationships 4s required together with estimations of fecundity and observations of the survival of fry and small immature fish. Statistics of the catch and fishing effort are also required hefore a more complete management programme can be suggested, ACKNOWLEDGMENTS The project was suggested and supervised by Mr, I. M, Thomas, Zoology Department of the University of Adelaide, prior to going overseas on sabbatical leave. I am greatly indeb(ed to Dr. S.J. Edmonds for supervising me during Mr, Thomas’ absence, for reading the original draft and offering much helpful advice. Thanks are due to Mr. T, D, Scott, formerly Curator of Fisheries of the South Australian Museurn for his guidance and to Prof. J. M. Thomson, Zoology Depart- ment of the University af Queensland, for critically reading this manuscript. Vinancial support for this investigation came from the Research Grants Committee uf the University of Adelaide. Finally, thanks are due to S.A.F.C.O.L. for enahling the sampling programme to be carried out on their premises. REFERENCES Crank, Frances N. (1923).—The life history of Leuresthey tennis, an atherine fish with tide controlled spawning habits, Fish. Bull., Sacramento No. 10: 1-51, Darin, W, J. (1939), The age determination of the tiger flathead. Neoplatycephalis (Calefrxia) mactradaw (Ogilbv), by means of otoliths. Rec, Aus, Mus. 20; 282-93, Famnmripce, W. 5, (£951). The New Sovth Wales tiger fathead Neoplatycephalusa macrodan (Ogilby). T. Biology and age determination. Aust. J. Mar, Freshw. Res, 2; 117-78, Granam, M. (1929). Studies of age determination in fish. 1, A survey of the literature, Fivh, Incest,, Lond, (2)11(3); 1-50, Thexiing, C, F. (1980). The natural history of the hake. LT, Seasonal changes in the condition of the hake. ish. Invest, Lond. (2)12(1); 1-78, Hiexiane, C, F, (1931) The stracture of the otolith of the hake, Quart. J. Mier, Sel. 74: 547-61 your, J. (1910), Report an herring investigations until January, L910. Publ, Cire. Cons. Explor. Mer, No. 53; 1-173. Linc, J. K. (1958),. The Sea Garfish, Reporhamphus melunachir (Cuvier & Valenciennes ) (Hemiramphidue). in South Australia; Breecling, age determination and growth rate. Aust. J. Mar. Freslaw, Res, 9; 60-105, Morannen. A. R, (1947), Observations om the growth of the plaice and on the formation nf annual rings in the otolith. Seenska Hydrogr—biol, Komm, Sk, 2: 1-9. Scatr, T, D. (1954). M.Sc. Thesis. University of Adelaide. Ssvepecon, G. W. (1956), “Statistical methods applied to experiments in agricultyre and hiology.” Fifth edition, (The Towa State College Press: Ames, Towa,) 50 J. A. HARRIS THomson, J. M. (1957a). Biological studies of economic significance of the Yellow-Eye Mullet, Aldrichetta forsteri (Cuvier & Valenciennes) (Mugilidae). Aust. J. Mar. Freshw. Res. 8; 1-138, Tnromson, J. M. (1957b). Interpretation of the scales of the Yellow-Eye Mullet Aldrichetta forstert (Cuvier & Valenciennes) (Mugilidae). Aust. J. Mar. Freshw. Res. 8; 14-28. THE DISTRIBUTION AND LIFE HISTORY OF THE SKINK HEMIERGIS PERONII (FITZINGER) BY MICHAEL SMYTH* Summary Hemiergis peronii occurs from south-western Western Australia to southeastern South Australia, but rarely inside the 12-inch isohyet and not in the Flinders or Mt. Lofty Ranges. Its northern limit is probably determined by aridity; its southern limit might be determined by the length of the winter. H. peronii bears two to five living young in February. Females are inseminated in the autumn, when they are two years old. They ovulate in spring and do not bear their first young until they are three years old. Males first come into breeding condition when two years old. H. peronii eats mostly arthropods and snails. THE DISTRIBUTION AND LIFE HISTORY OF THE SKINK HEMIERGIS PERONI (FITZINGER) hy MicuaAri, SmytTn® [Read 11 July, 1968] SUMMARY Hemiergis peronii ocours from sonth-western Western Australia to south- eastern South Australia, but rarely inside the 13-inch isohyet and. not in the Flin- ders or Mt. Lofty Ranges, Its northern limit is probably determined by aridity; its sonthern limit might be determined by the length of the winter. A. peronii bears two to five living young in February. Females are insemi- nated in the autumn, when they are two years ald. They ovulate in spring and do not bear their first-young until they are three years old, Males first come into breeding conclition when two years old. IL. peronii wats mostly arthropods and snails. INTRODUCTION Hemiergis peronii is a small, weak-limbed skink, very abundant in coastal dunes near Adelaide. Its reproductive cycle is unusual in that the females are inseminated in fhe autumn But do not ovulate until spring Gsesivth and Smith, 1968). I now report some other details of its life history. They will provide a background against which the adaptive significance of the unusual reproductive cycle will perhaps become clearer, and they will slightly diminish our great ignor- ance about our native reptiles, Some authors use Hemiergis as a sub-genus in the genus Lygosoma, H. per- onii is called L. (H.) quadridigitatum Werner by Glauert (1961), probably for reasons which are explained and dismissed by Loveridge (1934). Worrell (1963) calls it Lygosoma ( Leiolopisma) peronii. METIIODS The distribution of H. peronit was mapped from the records of the South Australian, Western Australian, and Australian Museums and the Department of Zoalogy, University of Melbourne, from the published records of Wermer (1910), Waite (1929), Loveridge (1934) and Mitchell and Behrndt (1949), and from my own collections and those made for me by a group of students of Naracoorte High School directed by Mr, D, Von Behrens. The natural history was described from samples of from four to 16 lizards taken at two to four-weekly intervals for a year at Port Gawler and Middle Beach, about 30 miles north of Adelaide. Most of them were taken from under dead clumps of the lily Dianella revoluta R, Br. on shell-grit dunes behind the beach. The lizards were brought hack to the laboratory and kept at 10°C until they were * Department of Zoology, University of Adelaide: Trans, Roy. Soc, S.A. (1968), Vol, 92, at G 52 MICHAEL SMYTH dissected. They were then measured with dividers, their reproductive organs removed and examined, their guts removed, the food taken out and identified, and the guts examined for parasites, he bodies were then preserved in buffered neu- tral formalin, The gut parasites are described hy Angel and Mawson (1968), DISTRIBUTION H. peronit is confined to southern Australia, from south-western Western Australia to south-eastern South Australia ( Fig. 1), It probably occurs in western Victoria as well, for according to Rawlinson (1966) there is a specimen in the National Muscum, Mclhourne, listed in the catalogue as from “Mallee district, Victoria’. But it does not occur in southern Victoria, Tasmania, or the Bass Strait islands ( Rawlinson, 1967). Lucas and Frost's (1894) claim that it has becn taken in the Dandenong Ranges near Melbourne is probably mistaken, and Weekes’s 4 SOUTH AUSTRALIA | WESTERN AUSTRALIA VICTORIA a S > TASMAN No Fig. 1 (1930) claim that she took Lygosoma (H,) quadridigitatum at Tanara and Jenolan, New South Wales, is probably based on a wrong identification. In Western Australia, I. péronii is found into the 50-60 inch rainfall belt near Northcliffe, but as far as we know at present its range does not extend right to the west coast. Loveridge (1934) does record a specimen from Perth, but this might have been the address of the collector rather than of the lizard, It has been taken as far inland as Fraser Range, in the 11-12 inch rainfall belt. In South Australia, H. peronii is very abundant around much of the coastline (Fig. 2), It also occurs on many of the offshore islands, including St, Francis, Franklin, Flinders, Pearson, Greenly, Price, Black Rock, the South Neptune, and Wedge Islands. It occurs inland on sandy or skeletal calcareous soils on Eyre and Yorke Peninsulas, Kangaroo Island, and in the south-east of the State. Its distribu- tion extends slightly inside the 12 inch isohyet only on the West Coast and near the head of St. Vincent Gulf (Fig. 2). It seems not to occur in either the Flinders or Mt. Lofty Ranges. Further collecting will no doubt extend the known range of H. peronii. For instance, it might be found all the way around the Bight, and further into the Murray Mallee of South Australia. HEMIERGIS PERONII 53 santana REPRODUCTION Male H. peronii come into breeding condition and inseminate the females in the later summer and autumn (Smyth and Smith, 1968). The sperm is stored in the female genital tract over winter; the females ovulate and their eggs are fertil- ized in the spring, between late October and the end of November. Two, three or four young are born at the end of February, larger mothers bearing on the aver- age large litters. 3 4 2 1 > A v 3 3 Fig. 3. The frequency distributions of 2 snout-to-vent lengths of female H, peronit z bearing two, three and four young. The wi | number of young was counted from either 5 large ovarian eggs (white) or eggs and O embryos in the oviduct (black), ur 3 2 a8 ] 5 6 7 LENGTH (cM) a4 MICHAEL SMYTII TABLE 1 Breeding condition of femulu Hemiergis peroni, Number of females with* = = , Total Date no sign of number developing ovarian oviducal = | reproductive follicles; verges eggs activity 16 November 1966 1 1 2 29 November é 1 4 14 December 6 2 8 § January 1967 8 3 25 January A ui ‘BI 16 February 4 3 7 1 March 5 5 14 March 2 2 28 March a 9 12 April 2 2 25 April 8 8 10 May ij 2 8 24 May + | sy 13 June 6 1 7 18 June 4 4 18 July 1 l 2 18 August <4 1 3 19 September 2 1 3 3. (ctober 4 | 6 17 October j 5 31 October 2 3 4 21 Nouyermber 2 2 4 *Femaics with no sign of reproductive activity included juveniles, ono-year-olds, and, in March and April, older ammals. Their ovarian follicles were small and transparent. Females were said to have developing follicles when the follicles berarne an apace white, though thoy had enlarged very little if at all, Later, after the follicles had obviously begun Lo enlarge, the females wore sail to have ovarian eggs. In H. peronii the right ovary is well anterior to the left in the body cavity, and the right oviduct is much longer than the left. About equal numbers of eggs are shed from each ovary but a high proportion migrates from the left ovary to the right oviduct. This is deduced from a comparison of the distribution of the corpora lutea between the wo ovaries with the distribution of eggs and embryos between the two oviducts. Thus of 90 eggs shed from the ovaries of 29 females, 48 came from right ovaries, but 61 embryos developed in right oviducts and only 29 in the left. In only one of the 29 females was there more than onc embryo in the left oviduct; in this case there were two in each. The most posterior egg was always in the left oviduct, GROWTII AND AGE STRUCTURE Figs. 4 and 5 show the distribution of snout-vent lengths of all individuals taken during the vear. It can be shown from these figures that males first come into breeding condition, and females are first inseminated, when they are two years old. Females therefore bear their first young when they are about three years old, Consider the snout-vent lengths of females (Fig. 4); in April and May there are clearly two year-classes in which there is no follicular development. These have been delineated in Fig. 4; the lower group is of young of the year, the HEMIERGIS PERONII 5D SV LENGTH Cem) N DJ F MA MJ JS A S ON Fig. 4. The snout-to-vent lengths of all female H, peronii caught from November, 1965 to December, 1966. @ pregnant; ® not pregnant, follicles undeveloped; © follicles visibly developing; + juvenile, sexes not distinguished. upper group is of onc-year-olds. Now consider the distributions for Noyember, 1966 to February, 1967, when the females were pregnant. It is apparent that those females not pregnant must have belonged to these two youngest year classes. A somewhat similar argument can be applied to the males (Fig. 5), From February to May, when the testes of most malcs in the samples were enlarged, there were always some males with quite undeveloped testes; these were young of the year and, as well, larger animals which must have been one year old. Neither the males nor the females, once they have reached sexual maturity, can he confidently separated into further year-classes, but it is likely that there are at least two year-classes of pregnant females in the summer, which by February will be three and four ycars old respectively. The larger females, probably approaching four years old, bear the largest litters (Vig. 3), Females grow to be longer than males. 80 SV LENGTH Com 8 N O JS FM A MJ J A S O N Fig. 5. The snout-to-vent lengths of all male H. peronii caught from November, 1965 to December, 1966, @ testis and/or epididymis con- tains sperm; © no sperm; + juvenile, sexes not distinguished. 36 MICHAEL SMYTH FOOD Tt was fairly easy to classify most arthropod foods to Order because at least the heads and usually other parts as well passed through H. peronii intact. But some other foods were probably missed. Land snails, for instance, were somctimes found without shells or with only a féw fragments of shell adhering, which made them hard to notice. Also, the shells of land snails were probably sometimes con- fused with the small marine shells, mostly of gastropods and foraminifera, which made up much of the ground on which the lizards were living, and which often appeared in the lizards’ guts. So the proportion of land snails in the diet was probably under-estimated. The abundance of various items of food taken by male, female and juvenile H, y:eronii is shown in Table 2, Obviously H. peroniié is almost exclusively insecti- yorous. The only plant food found more than once was the seeds of the lily D. revolita under which most of the lizards were caught. These seeds are small, black and shiny, and might have been mistaken for insects. They were largely undigested. TABLE 2 The dict, of Hemieripis peronit. Numbers of each item 9% wf botel Ttem . ; es i Males Females | Juveniles Males Females | Juveniles Beatles 100 133 6 28-3 299 7-1 Ants anc other Hymenoptera. 24 iG 2 6-8 12.6 24 Cockroaches 30 23 3 8-5 5:2 3-6 Jockroach oothecus 15 25 4-2 516 Moths | 25 22 7-1 49 Bugs (Hemiptera) 17 10 4-8 22 Unidentified insect larvae i 1} Bel 29 Collembola 2 1 0-6 02 Grasshoppers 2 0-6 Flies L ()-3 Earwigs 1 0-3 Ant-lions : 1 Og Mites 41 11 60 11-6 27:2 82-1 Spiders 13 1h 3-7 22 Psendoscovpinos 4 2. 0-8 24 Sluturs (Isapoda) 4 7 T-t 16 Centipedes 1 Q.2 Snails 20 6 1 8-2 13 1-2 Lizards 2 1 O+¢ Od Secda 21 ' 14 5-4 a+4 Unidentified 1 | 1 3-4 1-3 Total numbor of | individual items a3 445 | 84 i] The range of size of the foods taken was large, from small mites, mostly Ori- batids, up to adult skinks Ablepharus greyii (Gray), which are about 4 cm long. Juveniles took only small items: there is very little ditterence between the foods of males and females. Most of the animals caten were feeders in litter or on the surface. VEMIERGIS PERONI oT INJURIES TO LIMBS Several A, peronii had lost digits or parts of limbs. This would he no great hanclicap to them, since their movement is mainly by lateral undulations. Some of these members would have been lost in accidents: { found one H. peronid with an aut's head still firmly clamped onto its foreleg, though the limb distal to the ant’s mandibles had withered and would soon have dropped off. In other cases, the members might have been lost by discase or fights. Males and females apparently lost large portions of front and hind limbs with about equal frequencies, but in both sexes toes were missing much more often than fingers (Table 3), This suggests that digits and large portions of limbs are lost from different causes. TABLE 3 Limh injuries in Hemieryéx povewit. ——— rrr | Male | Fernale Na, feet from which one ue more digits Were lost | front 5 i | hinrl pm | | 20 No. linbs wholly ov partly last | front Ni 4 hind | 4 0 SO, animals with wll members intart 6 73 No. animals exumined te Ys One I. peronii was found to have five fingers on cach front foot, though the usual number is four, This anomaly is worth mentioning only because several spe- cies of Hemiergis are most easily distinguished on the basis of the number of their fingers and toes. Werner (1910) claimed that this was the only way he could separate them at all. Glauert (1961) mentions some other freak numbers of digits. and gives the impression that the mmmber of digits is not a good specific character. DISCUSSION Tt is likely that the northern limit to the distribution of H. peronii is deter- mined by aridity, for its powers to resist desiccation are relatively poor. Warbur (1966) has shown that its rate of evaporative water loss increases rapidly wit increasing temperature, and that it quickly dies in a dry atmosphere at tempera- tures as low as 35°C, The southern limits to ils distribution might be delennined by winter temper- atures. Where winters are cold, lizards usually become torpid, and will hecome active again only in warm, sunny weather, Even then, they probably do not feed, for torpid cold reptiles with food in their guts are in mortal danger because they cannot digest the food, which then putrefies (Regal. 1966), I found food in the gut of H, peronii throughout the winter, so, though it certainly becomes less active, it probably does not become torpid for long. Also, abdominal fat bodies, which in some lizards are known to be important in over-winter metabolism (Des- sauer, 1955), are absent in H. peronii, though they are prominent in some other local skinks. So H. peronii might need to feed over winter in order to survive. If this is 80, its pteferred body temperature should be low. There is no information available on this, but Licht, Dawson, Shoemaker and Main (1966) haye shown that the preferred body temperature of I. quadrilineatum is relatively low, and 38 MICHAEL SMYTH If. quadrilineatum, like II. peranii, has a very high rate of evaporative water loss (Licht, Dawson and Shoemaker, 1966). Why HH. peronii does not occur in the Flinders or Mt. Lofty Ranges is hard to say. Over much of its range it occurs on sand; in the south-east, for instance, its range seems to coincide with the relict coastal dunes there (Sprigg, 1952). On parts of Eyre Peninsula it occurs on soils which, though not yery sandy, are underlain by kunkar derived from old sand dunes. Perhaps H. peronii avoids soils prone to watcr-logging. But obviously a full analysis of its distribution reqnires much more detailed study. ACKNOWLEDGMENTS 1 wish to thank John Mitchell, Glen Storr, Harold Cogger, Peter Rawlinson, Dierk Von Behrens and students from Naracoorte High School, Michael] Braysher, Meredith Smith, Vaal Nielsen, Beverley Jones, Stuart Harris and Lewis Chinner for help and advice. REFERENCES ANGEL, J., M.-and Mawson, P. M. (1968). Helminths of some lizards from South Australia. Proe. R. Soc. §, Aust. 92; 59-72. Drssaurr, H, C, (1935), Seasonal changes in the gross organ composition of the lizard, Anolis carolinensis. J, Exp. Zool. 128: 1-12, Guaven?, L. (1961). “A Handbook of the Lizards of Western Australia”, (Western Australian Naturalists’ Club: Perth.) Lacirr, P., Dawson, W. BR. and Suommaxen, V. H, (1966), [leat resistance of soit: Australian lizards, Copéiga 1966: 162-169. Licut. P,, Dawson, W. R.. SHoemAxeR, V. H. and Mat, A. R. (1966). Observations ou the thermal relations of Wexteru Australian lizards, Copeia 1966: 97-110. Lovenipce, A. (1984). Australian reptiles in the Museum of Comparative Zoblogy, Cambridge, Massachusetts. Bull, Mus. Comp. Zool, Harvard 77; 243-383. Lucas, B Li. S. and Frost, C, (1894). The lizards indigenous to Victoria. Proc. R. Soc, Vict. 6; 24-92. MrrceEct, I. J. and Brarnnr, A. ©. (1949), Fauna and flora of the Greenly Islinds, Part 1, Introductory narrative and vertebrate fauna, Rec, South Aust, Mus, 9: 167-179. Rawniixson, P. A. (1966). Reptiles of the Victorian Mallec. Proc, R, Soc. Vict. 79: 605-619. Rawrrsson, P. A. (1967). The vertebrate fauna of the Bass Strait Islands; 2. ‘he Reptilia of Flinders and King Islands. Proc, BR. Soe. Vict. 80; 211-223. REGAL, e J. (1966). Thermophilie response following feeding in certain reptiles. Copcia 196A: 88-390. Syeytat, M. and Saar, M, J. (1968). Obligatory sperm storage in the skink Memiergis peronti. Science 161: 575. Spence, R, C. (1952). The geology of the Soxth-East Province, South Australia, with special reference to Quaternary coast-line migrations and modern beach developments, Bull, Ceol, Surv. §, Aust. no, 29: 1-20. Warez, % R. {1020}. “The Reptiles and Amphibia of South Australia”, (Government Printer: Adelaide. Wanbunc, M. R. (1966). On the water economy of several Australian geckos, agamids and skinks. Copeia 1966: 230-235. Werxes, H. C. (1980), On placentation in reptiles. UW. Proc, Linn. Soc. N.S.W. 55: 550-5764. Wenner, F. (1910). Reptilia (Geckonidue und Scincidae). In “Die Fauna Sidwest-Australiens, Ergebnisse der Hamburger siidwest-australischen Forschungsreise 1905". Ed. by Michacl- sen, W. and Hartmeyer, R. Bd. 2, pp, 431-93, (Fischer: Jena.) Worsert, &. (1963). “Reptiles of Australia’. (Angus and Robertson; Sydney.) HELMINTHS FROM SOME LIZARDS MOSTLY FROM SOUTH AUSTRALIA BY L. MADELINE ANGEL AND PATRICIA M. MAWSon* Summary An account is given of helminths from lizards from an area near Port Gawler in South Australia, including a table of their incidence. Records are also given of trematodes from two Queensland lizards. Species recorded (all from the Pt. Gawler region unless otherwise stated) are: Paradistomum. crucifer (Nicoll) (syns. Eurytrema crucifer Nicoll, Paragonimus trachysauri MacCallum, Cephalogonimus trachysauri MacCallum, Paradistomum muccallumi Johnston) from Hemiergis peronii, Trachydosaurus rugosus (Pt. Gawler area and from Murray Bridge S.A.), Tiliqua scincoides (Queensland) , and Varanus varius (Queensland) Oochoristica trachysauri (MacCallum) (syn. Oochoristica australiensis Spassky (from Trachydosaurus rugosus; Thelandros kartana Johns ton and Mawson from H. peronii, Phyllodactylus marmoratus; T. trachysauri Johnston and Mawson from Trachydosaurus rugosus; Skrjabinodon smythi n. sp. From P. marmoratus. Other helminths recorded are Microphallus sp. from T. rugosus; Trematoda, ? sp.., from Rhodona bougainvillii; Oochoristica sp. and Baerietta sp. from R. bougainvillii, H. peronii and P. marmoratus; Skrjabinelazia sp. From P. marmoratus; acanthocephalan cysts from H. peronii. HELMINTHS FROM SOME LIZARDS MOSTLY FROM SOUTH AUSTRALIA by L. Mapenrse ANGEL anp Patricia M. Mawson* [Read 8 August 1968] SUMMARY An aeeount is given of helminths from lizards froin an urea near Port Gawler in South Australia, including u table of their incidence. Records are also given of lrematodes from two Queensland Ligurds. Species recorded (all from the Pt, Gawler region unless otherwise stated) are; Paradistamum crucifer (Nicoll) (syns. Eurytrema crucifer Nicoll, Paragonimus trachysuuri MacCallum, Cephala- conimus trachysanri MacCallum, Paradistomum maccallumi Johnston) from Hemiergis peronii, Trachydosaurtis rugosus (Pt. Gawler area and from Murray Bridge, S.A.), Tiliqua scincoides (Queensland), and Varanus varius (Queens- land); Oochoristica trachysayri (MacQallum) (syn. Oochoristica australiensis Spassky (from. Trachydesaurus rugosus; Thelandros kartana Johnston and Mawson from MT. peronii, Phylloductyjlus marmoratus; T. trachysaurt Johnston and Mawson from Trachydosaurus rugasus; Skriabinodan smythi n. sp. from P. marmoratis. Other helminths recorded are Microphallits sp. fron T, rygosus; Trematoda, ? sp., from Rhedona bougainvillit; Oochoristica sp. and Baerietta sp. from RA. bougainvillit, H, peronii aud P. marmoratus; Skrjabineluzia sp. from P. marmoratus; acanthocephalan cysts from H, peranii. INTRODUCTION During 1967, Dr. Michael Smyth, of this department, undertook an investiga- tion of aspects of the ecology of lizards inhabiting a coastal strip north of Adelaide. The parasites of these lizards, and some from the same host species from different localities, are discussed in this paper. The occurrence of a trema- tode from two other species of lizards, both from Queensland, is also recorded. We are very grateful to Dr. Smyth, not only for giving us the viscera for examination, but also for the regularity and precision of his collection and records. An aceount of his work is in press. The study area is a short distance north of Adclaide and the two collecting centres aré Port Gawler and Middle Beach. The two areas are separated by two permanent salt water channels. The lizards from each area are listed in Table 1, with records of parasitism. It will be noted that trematodes were found (except in oné case) only at Middle Beach; cestodes and nematodes occurred in both areas. Juvenile lizards seem to be free from helminths. Records were kept of the sex of each lizard dissected, but this appeared to have no significance in relation to the infestation by helminths, and has not been indicated in the Table. Differences in the incidence of parasitism, as well as of the different groups of helminths, in the different species of lizards, are quite marked, and are discussed below. We are grateful to Dr. S. J, Edmonds, of this deparlment, for examining the acanthocephalan cysts, Our thanks are also due to Mr. John Mitchell, of the South Australian Museum, for the correct names of the lizards concerned. * Department of Zaology, University af Adelaide. Trans, Roy. Soe, S.A. (1968), Vol. 92, 60 L. MADELINE ANGEL ayn PATRICIA M. MAWSON TABLE 1 Incidence of helminths in the study ures. The figures refer to the number of lizards dissected or found parasitised. An asterisk mdicates that the alimentary canal only, not the gall bladder, of these specimens was examined a UE aE EEE Number yielding Lizard Locality Number | Number - - po dissected parasitiged | ‘Trema- | Cestoda | Nema- jAcantho- toda toda | cophala Trachydoxsauras | Pt. Gawler 3 3 a 3 = rugosa (Gray) Middle Beach! 6 i 1 2 ; | — I Ablepharus greytt Pt. Gawler 5* 0 (Gray) { | 0 Middle Beach) 2 | 9 Hemierqia peronit Pt. Gawler au* 16 os 7 12 (Fitzinger) Li) 26 — 7 eal] 2 ~ 10 juv. 0 Middle Beach 13* + — 4 il 37 15 2 84 3 § juv. A, devrestensis Pt. Gawler I 0 (Fitzinger) i Rhadona Pt. Gawler oth 14 iT Ii -— bougeinvéllii (Gray) : Phyllodactylus Pk. Gawler .P 1) a 2 1 nearmoratus (Gray ) 1 juv, 0 Middle Beach l [ -- i a PARASITES RECORDED, ARRANGED UNDER THEIR HOSTS Trachydosaurus rugosus Gray. Microphallus sp.; Paradistomum crucifer (Nicoll); Thelandros trachysauri Johnston and Mawson. Hemiergis peronii (Fitzinger). Paradistomum crucifer (Nicoll); Oochoristica sp. Baerietta sp.; Pharyngodon kartana Johnston and Mawson. Tiliqua scincoides (Shaw). Paradistomum erucifer (Nicoll). Rhodona bougainvillii (Gray). Trematoda, 2 sp.; Oochoristica sp; Baerietta sp. Phyllodactylus marmotatus (Gray), — Skrjabinodon smythi n. sp., Thelan- dios kartana Johnston and Mawson; Skrjabinelazia sp. Varanus varius (Shaw). Paradistomum crucifer (Nicoll). TREMATODA Paradistomum crucifer (Nicoll) (Figs. 1-6) Eurytrema crucifer Nicoll, 1914, 338, in gall-bladder, Dela fraseri. Paradistomum crucifer (Nicoll) Travassos, L919, 12; 1944, 256, Patagonimus trachysauri MacCallum, 1921. 173. im gall-bladder, Trachydosaturus rugosus (syn. Trachysauris rugosus ). Cephalogonimus trachysauri MacCallum, 1921, 176, in gall-bladder, Travhydosauras rugasus, Paradistoma trachysauri (MacCalltmm) Dollfus, 1922, 329. Paradistoma truchysauri (MacCallum) Dollfus; Johnston, 1932, 64, Paradistomum maccullumi Johnston, 1932, 64 (nom. nov. for Cephalogonimus. truchysauri (MacCalhm ). Paradistumum. trachysaurt (MacCallum) Dollfus; (syns. Paragonimus Wwachysauri MacCallum and Paradistomum maceallumt Johnston). Malan 1959, 87. Paradistomum trachysauri (MacCallum) Dollfus; Travassos, 1944, 262. TIELMINTHS FROM SOME LIZARDS &1 Hosts and Localities: Hemiergis peronii, Middle Beach, South Australia, from gall-bladder, occasionally in sma!l intestine; Trachydosaurus +tigosus, Murray Bridge district, South Australia, from gall-bladder and liver; Tiliqua scincoides, Iyueing Island, near Gladstone, Queensland, apparently collected by Professor T. Harvey Johnston, in 1918, from gall-bladder (ten specimens from one host); Varanus varius, Townsville, Queensland, collected by Dr, G, A. M, Heydon in 1927, from gall-bladder (two specimens from one host). The description is based on 24 stained and mounted specimens from Hemiergis peronii, with details from living specimens, Notes on the trematode from Trachydosaurus rugosus are given in the discussion. 2 ts IS TAINS ANN AS Fins. 1-6 Paradistomum ecrucifer, Drawings were made with the aid of a camera lucida, Scales in millimetres, Figs. 1 and 6 to same scale. Evws not drawn exactly to scale. Figs. 1-3. 5, 6, from Memiergiy pereniiz 1, 2, 6, adults, showing variation in form, and in extent of uterus, (2. flattened); 3. cirrus sae, fram specimen shown in fig, 2, enlarged; 5, youny adult, excretory system from living specimen. Fig. 4, type specimen redrawn (fronr Delma frasert). ¢, cirrus; eb, excretory bladder; mg, Mehlis’ gland; 0, ovary; bie, prostiie gland cells; rs, receptacuhim seminis; sv, sernimal vesicle: vd, vitelline cluct. 62 lL. MADELINE ANGEL any FATRICIA M. MAWSON The numbers of trematodes in the infected H. peronii ranged from I-15, and generally included juveniles as well as adults. One gall-bladder contained 65 specimens, but all were very young, Nearly all specimens were found in the gall- bladder, but a few were in the liver, and four were taken from the intestine. OF these last, three were in two hosts whose galJ-bladders were uninfected, and the fourth was in a lizard whose gall-bladder contained twelve trematodes. In one lizard only, the trematodes were found not only in the gall-bladder (thirteen). but alse in ducts which appear to be pancreatic, running through the clongated panereas from the gall-bladder to the intestine (fifteen). Specimens were fixed, under a coverslip with only slight pressure, in formol acetic aleohel, stained in Van Cleave’s combination haematoxylin stain, and mounted in Depex. DESCRIPTION Shape: Elongated ov leaf-like when living, fixed specimens with bluntly rounded posterior and often somewhat elongated anterior end. Colour pale pink, with caeca yellow. Rody 0748-2890 mm long by 0:306-1 -496 mm wide. Cuticle not spined, Oral sucker rounded or slightly oval, 223-494 » (average 342 4) Jong by 129- 493 yp, (330 p) wide. Acetabulum rather inconspicuous, rounded, 170-353 4 (235 «) long hy 141-306 » (239 ») wide; anterior border near anterior third of body, Ratio of width of oral to width of ventral sucker 1:0:7 to 1:0:9. Pre- pharynx short; pharynx 50 by 50 «—1L0 by 100 », dorsal to oral sucker or partly posterior to it; oesephagus short; caeca wide, sinuous, extending nearly to pos- terior end of body. Testes symmetrical, at sides of acctabulum, rounded in living, may he irregulae in Gxed specimens, equal or sub-equal, 71-165 ~ (110 4) long by 5Y- 176 p. (106 ») wide. Cirrus sac 92-184 p» long by 45-100 » wide, between suckers, at an angle to left of mid-line, Seminal vesicle internal, much coiled. Cirrus inconspicuous, slightly coiled, surrounded by diffuse prostatic cells, Cenital pore median, near posterior border of oral sucker. Ovary post-ncetabular, sinistral, irregular, §2-223 » long by 71-188 ,» wide. Receptaculum seminis mostly dorsal ta ovary, may be posterior or to cither side of it, 47-141 » (96 ») by 59-212 » (107 »). Mehlis’ gland irregular, generally posterior to ovary and slightly to right. Laurer’s canal not seen. Uterus varying in extent from condition shown by Nicoll, with relutively few eggs (Fig. 6), to one in which it fills all of hind-body and an area in front of acctabulum; passing to right of cirrus sac and opening by museulur nctraterm at genital pore. Eygs variable, larecst 40 » by 22 p. Vitellaria extracaecal, restricted to middle of body, reaching level of posterior third of acetabulum anteriorly; 270-764 » (476 «) in extent; Jobules variable, some moderately large. Transverse yolk ducts widening at junction in mid-line, but forming no obvious yolk reservoir. Lxeretory bladder elongate, main arms leaving anterior end, Exerctory pore terminal. DISCUSSION The muture trematodes found in Hemiergis peéronii in this study show a considerable variation in size, ranging from 0-748 by 0-374 mm to 2-890 hy 1258 mm, and in appearance. The trematodes found in the pancreatic ducts were greatly elangated, while the worms from the gall-bladder tended to be TELMINTHS FROM SOME LIZARDS 63 foliate, As immature specimens are often present with the adults, some of the differeices among egg-beuring adults ire probably due to differences in age. For example, the testes and ovary may be as large in small worms as in much larger ones (Figs, 2, 6) giving a disproportionate emphasis to these organs in the small specimens, The specimen shown in Fig, 1 shows the acetabulum very near the oral sucker; & specimen of similar size from the same host showed a similar arrangement of eggs, but the body was more elongated and the acetabulum was in the more Gypical position, Even among specimens of the same size there is great variation in the number of eggs, and this leads to differences of form andl general appearance, In the original description of Paradistomum crucifer, Nicoll stated that the uterus, especially in the less mature specimens, had a charavteristic cruciform course, but that in more mature specimens this arrangement was, to some extent, obliterated. Among the sinall trematodes of my collection. some (Fig. 6) have a uterus showing the crucitorm course, while in others (Fig. 2) the eges form a inore or less solid mass in the hind-body and sometimes anteriorly to the acctabulum, so that most of the organs are obscnred, Even among the larger specimets there is alsoa good deal of variation in the number of eggs. By courtesy of the Director of the School of Public Health and Tropical Medicine, Sydney, one of us (L.M.A.) has been able to examine the type of P. crucifer. Some of our smaller specimens are very similar to it in appearance, aml one of them is almost identical with it, not only in overall size, but in the sive and arrangement of the organs. The trematodes from Hemdergis peronii must be regarded as Paradistomum ertcifer. Although the range of measurements given by Nicoll shows that his speci- mens were umformly smaller than the trematodes found in H. peronit, Nicoll's description was based only on “a few" worms. In addition, all measurements given by Nicoll are consistently smaller than those now made on the type (Table 2), It is possihle that this is due to some fattening of the specimen over the years; it 1s also possible that Nicoll had made a mistake in his scale. MacCallum (1921) found about twenty trematodes in the gall-bladder of a stump-tailed lizard, Trachydoseurus rugesus, which died in the New York Zoo, He stated: “there were three different sizes among them, which, with the decided differences in form, would almost make it necessary to divide them into three species, but as they are in many particulars alike, and for brevity’s sake, we shall describe them as one species”. This he named Paragonimus trachysaurt, Later in the same paper MacCallum stated that he had found two different trematodes in the lizard, one being P. trachysauri; the second he described as Cephalogonimus frachysouri. Malin (1939) regarded the two as identical, and thus synonymous with Paradis!omum trachysauri, and this has been accepted by later workers. Although the largest specinens frotn H. peronii are smaller than the measure- ments given by MacCallum for Paradistomum trachysauri, the measurements of ovarics and testes are comparable, the ratios of the suckers appear similar, and the specimens resemble the one figured hy MacCallum. Differences in fixation could gecount for some differenves in body dimensions, as also could the relative sives of the hosts*. There seems no reason ta regard the specimens from Hemieryis perenié as distinct from those deserihed by MacCallum trom Trachydosaurus gugosus, so that Paradistomum trachysauri must become synonymous with I’. crucifer, (It should be noted, however, that P. crucifer as now interpreted * It is of interest to note that Delma frasert (the type host of Paradistomum orucifey), a yeapodid lizard, and Temicrais peronil are much smaller than Trachydosaurus ttigosus. 1 ave Tut Uisseted D. freseri, ut if can be expected that its gall-bladder is very lithe Larger than that at If neronit, which is much smaller than that ef T. rugasts, EL anp PATRICIA M. MAWSON ANG - 4, L. MADELINE ‘Kem Auw ur pauayyey uoog oavy o7 deedde you op tsueureds qaidg *(uoIssnasip vax) gevosad sebsarniazy Waa] osoyA Wey, pousy}EP ssay sueundadg ‘suauiloeds QE JO JSeTeWs pus ysesreyl ayy eon ¢ ‘BUY UBY} JOpBOIg BAL LUNTUGEYoUR vYI QpIYAX ‘pHoId UBtZ iatuo, sei JayONs [BIO aly) ‘sraspans qseypeus yp ueumvede ayy ut ‘aeyq7 yovy oy Aq Joy paqunodae st FesuRt OM} By} JO qued waMo] ayy ut Loundosastp qooindde ayy, y “da] [Rue SB UTYTLIG’eyao8 |YY SAVOYE ANS sy ‘YOU oy} se ofA] SB ST TUNTNG BoE O1]_, FEY} 9X91 OY] UL SolEIS UMTBQoBW YSNOUIY ¢ "pastel Baa wwosIeT 5 UaATS St saaqots Jo YZpIM ory ‘Apnjs quagord ayy Ur opeut eyuetMeuNsTETI Uy 7 ‘SUOJOITLE UT Que YAM ‘sBFe roy qdeoxa ‘sar AUIl]PIVU UT are syueMadusealL [Ly | 208 X OF wz ¥ TE rte X GE 2 X OF 0% X OF 0z ¥ OF 28% X OF Je X BS-08 Saag ve — £9e-0- 881 -0-1TL0°0 = = 680-0 ¢.RO}SAT SB OZIS Yyprsa— — = €EL-0- €ee-0-E80-0 O8E-0 e-0 ¥60-0 AEBS FOG ¥,, yasuey— AIBsg = 7 FRI-U- §£1-0-650-0 = wre | F60-0-280-0 — Yyypr - TY LEO £91 -O-TLO- 1 006-0 ST-0 | BOL O-TL0-0 F0-0 Yp suo] — 82788 [, — TFI+0 LOT-O0-€£0°0 ) O0T-0-TES0-0 =, — es FOO XUATEU (48) 81-0: EESs-01 ROVOS OT) GOL OT rT & PLO k E8-OT OLED maens GGEO-OLE( | SOE-O-1FE-O | GOGO -STT-0 | e908 -0-TFL70 8c8h-O = Sta SLO pana eqaay GGF-O-OLE-0 | GIF-O-TFE-O | SRE-O-LEL-O | +8éF-O-GET-0 0GE-0 = ¥6e-0 81-0 reyons [CI G6G-T-880-T | 816-0 F8S-0 | S16-O-H8e-0 | 96F-1-90E-0 04:0 00°S FL -0 F-0-98 0 WoPTM s9TS-2-BPT-G | ceOT B-OEL-S | obhLL-€-880°L | 068-E-BbL:0 O0+% 00-¢ | FECT 8-0-£:0 qyauary ¥ & “G te € Lb é! I way @,, (PoulluExe Tequun pT Teh] 1Z61 Aq waar Spnqs quoseagq | Apngs quaserg | Apnys quesorg | Apnys quosorg | ‘umyppegoeyt ‘umpegoeyy = Apnys lueserg] FIG] “TOOUN RUOUIAINS BOP 8I6L LEG R9GT/E LAG St/a/0¢ STs/0g = = paiwsypoa 9y8q “PLO “PLO CN Ws ‘ehprug MMOISPE[Y IL ‘ATTLASTIAAO ABLINPL “a Yoveq B]PPIIY | oog yOA AvaN | OO yIOR Many pursuant) YON AVpeaory saprouutoy snzobnit qrnosadl snsobnt snsobnie pnb? sry snunep 4 | snunvsophiyanay,| sibsauazy | enanpsopfiyonuy,| snunvsophkyovd,7, | ataspef DWC WoT wafiansa safranaa dafionsa safronda PAN DSAYIOL] wn pafpond (waturaeds ad) uinwo siping | wnuumsipniwg | wnwapsipolug | wriwoipuddg |snuaucbeppyda,| snuiuobveng wafionda puUcaghmngy SU poqiasac] (TOaTN) saftonsa wemuowrpyeDgy & ATA HELMINTHS FROM SOME LIZARDS 6n shows considerable variation in size and general appearance. Variation has already been reported for another species of Parailistomum by Dollfus (1922. 328, footnote), who referred to the great morphological variations he had found in P. miutabile (Molin). This is the first Australian record of 4 trematode from « lizard since MavCallum’s report of P. trachysauri. In 1932, Harvey Johnston reported that “a mimber” of 7, rugosus had been searched for parasites from time to time, but thut no trematodes had been found, After P. crucifer was found in H. peronii in the present study, as many T. rugosus as possible were examined. Jt was not until this paper was completed that P. crucifer was found in this host, in two lizards from the Murray Bridge district. (Of the forty-five T. rugasus dissected only nine were from the study area north of Adelaide. The remainder came from a number at different localities.) In the first lizard there were 34 trematodes in the gull- bladder smd 12 in the liver. In the second, there were 56 living and a number of disintegrating worms (which were mere collections of eggs) in the gall-bladuer, and 13 living worms in the liver, The lizards had been in captivity for almost two months when they were dissected, All the trematodes were mature, but great ditterences in size were found. The specimens were fixed in formol acctic alcohol under a coverslip. Virtually no pressure was required to fix them Rat, so that the measurements are not truly comparable with those of the specimens from A. peronii (which were fixed with slight pressure) but would probably have been greater (certainly in the width of the body) had the worms heen fixed with the same pressure. Measurements of the largest and the smallest specimens fram the first infected lizard aré given in Table 2. The following observations were made on the living trematodes fram T. mugosus. The caeca appear hright yellow (due to bile), and contain many crystals, as recorded by MacCallum. These are tetragonal in shape, and vary in size from fine slender crystals to forms up to 223 by 35 by 35 w. (Crystals were not present in P. erucifer from H. peronii; this is presumably due to a difference Hi composition of the hile of the two lizards.) The part of the hody which is not coloured by bile or obscured by eggs is pale pink in colour, Twenty mature eggs tukem at random from the liver washings measured 34-39 « (37 «) hy 21-24 p» (22 4). P. crucifer fram both hosts was very sensitive to changes in the medium in which it was kept. In 0-65% suline it very quickly hecame swollen and <¢lied. It vould be kept alive in bile at 6°C for several days. LIFE HISTORY The only species of Parudistomum for which investigations on the life-history have been recorded is P. mutabile (Molin). Timon-David and Timon-David (1967) infected Helicella arenosa experimentally and obtained brevicercous xiphidiocercariae, (HH. arenasa was not the normal host since it did not occur on the islands on which Paradistomum mutabile was commonly found in livards). The cyst stage was not found, but the authors thought it probable that a second intermediate host is necessary, and suggested an isopod or an insect. In the present study no information could be gathered on the life-history of P. erteifer, The only land snails found in the study area are Avstrosuccinee australis Ferrusac, Omegapilla australis Anvas, Australbinula margaretae Cox and Paralaoma stabilis Iredale, The last three of these ave very small snails and quite diffieult te find in the litter in which they occur, (P. stabilis was not. in fact, found 66 J,, MADELINE ANGEL ann PATRICIA M. MAWSON in the open at all, but was recovered from the stomachs of lizards, in which it occurred quite often). It was therefore not possible to conduct any trematode life-history studies with these molluscs, Attempts were made to infect Austro- succinica autralis (collected from another area) but it proved impossible to keep the snails alive Jong enough to obtain any results, It will be seen from Table 1 that whereas fiftcen of seventy-one Hemiergis peronii from Middle Beach were infected with Paradistomum crucifer, none of ninely-nme of these lizards from Port Gawler harboured the parasite. We can suggest no reason for this. If, as seems likely, a second intermediate host is necessary, it is to be sought among the animals listed by Smyth (1968) as found in the stomachs of H. peronii. Of these, the most common are weevils, free-living mites (very small species), ants, cockroaches, moths and snails. Microphallus sp, Host and Locality. Trachydosaurus rizgosus, Middle Beach. One specimen, in upper small intestine. Trematoda, ? sp. Host and Locality, Rhodona bougainvillii, Port Gawler, One specimen, in Intestine. CESTODA The authors, neither of whom is a cestodologist, are greatly indebted to Dr, John Hickman, of the Zoology Department, University of Tasmania, for identification of Buerietta sp, and verification of Oochoristica spp. Further identi- fication of the material will be made by Dr. Hickman. Oochoristica trachysauri (MacCallum) (Fig. 7) Tuenia trachysuuri MacCallum, 1921, 229. Oochovistica trachysauri (MacCallum), Johnston, 1932, 65. Oochoristica australiensis Spassky, 1951, 547. Host and Locality. Trachydosaurus rugosus, Middle Beach, South Australia, This species was fully described by Johnston (1932). Spassky (1951) con- sidered Johnston’s specimens different from those of MacCallum and proposed for therm a new species, O. australiensis. The material now examined, all from one host, shows similarity to all earlier collections—some specimens with an obvious rostellum, some with rounded anterior end; the mature segments vary (sometimes abruptly, fig. 7) from more or less square to elongate. It is considered that all helong to the same species. Oochoristica sp, Hosts and Locality, Rhodona bougainvillii, Hemiergis peronii, Phyllodactylus marmoratus, Port Gawler. More than one species may be present. Baerietta sp. Hosts and Locality, Rhodona bougainvillii, Hemiergis peronii, Port Gawler. The cestodes from these hosts arc similar and probably belong to the same specics. HELMINTHS FROM SOME LIZARDS 67 \ a../ 2 lo (3 Soa 300 Fig. T. Ovochoristica trechysauri, part of strobila showing variation in shape of segments. Figs, 8-10, Thelancros kartana, 8, anterior end; 9, vontral yiew of male tail; 10, female tail, Figs. 11-14, T. trachysauri, 11, anterior end; 12 and 13, yentral and lateral views of male tail: 14, tail of female. Figs, 9, 12, 13, to scale beside 13, figs. 8, 10, 11 and 14 to seale beside 10. NEMATODA The nematodes taken from the lizards at Port Gawler show an interesting host distribution. None were found in Ablepharus greyii or Rhodona bougain- villii. Oxyurids were found in Hemiergis peronii, Trachydosaurus rugosus and Phyllodactylus marmoratus, and in almost all cases each of these hosts carried only its own species. The exceptions were two P. marmoratus, in which were Thelandros kartana, typically present in H, peronii, Although H. bougainvillit is apparently free of nematodes in this area, it carries the same cestode species as does #7, peronii and these species are in turn different from that from T. rugosus. This difference in parasites may be duc to a high degree of host specificity among the oxyurids, and perhaps of resistance among the lizards, or it may be explained in part by the different niches occupied by the lizards. Dr, Smyth has informed us that R. bougainwillii and H, peronti live mainly under certain bushes, A. greyii and P. marmoratus mainly under wood, stones, etc., and T. rugosus (a much larger lizard) may be found in either habitat. Most of the lizards are diurnal and insectivorous; T. rugosus is diurnal and predominantly vegetarian; P. marmoratus is nocturnal and insectivorous. Oxyurid eggs might be expected to be quite common over the surface in the area, especially those of Thelandros trachysauri which occurs in hundreds in each host animal, 68 L. MADELINE ANGEL sxp PATRICIA M. MAWSON Thelandros kartana Johnston and Mawson (Figs. 8-10) Thelandtag kartana Jolinston and Mawson, 1941, 145; from Iemiergis peronii, Kangaroo Slana, Hosts and Localities. Hemiergis peronii, Port Gawler, Middle Beach. Phyllodactylus marmoratus, Port Gawler. Thelandros kartana was taken from 22 specimens of I. peronii from Middle Beach and 27 from Port Gawler. Female worms with eggs were taken from two P. marmoratus, No more than four worms were present in any one lizard, and usually only one or two. They occurred almost always in the short caecum at the junction of small and large intestine. The new specimens have been com- pared with paratypes of T. kartana and found to agree closely. The original description can now be augmented by Figures 8-10 and measurements in Table 3. The measurements of the specimens from P, marmorattus are within the range of those from H. peronii, Figs. 15-21, Skrjahinodon smythi, 15 and 16, anterior end of specimens in relaxed and con- tracted states respectively; 17, posterior end of male; 18, ventral view of cloacal region, male; 19, tail of female; 20, part of spike of female tail showing ‘spines’; 21, egg, Figs. 17 and 18 to same scale, HELMINTTIS FROM SOME LIZARDS 42 Thelandros teachysauri Johnston and Mawson (Vigs. 11-14) T Relandyye 5 frachysairt Jobeston and Mawson, 1947. 24; fron Traehytlosanris rupovun, Jselehnue, Host and Localities. Trachydosaurus rugosus, Port Gawler, Middle Beach. Thelandros trachysauri has been found in nearly all T, rugosus dissected in this Department—more than thirty-five from various places. It is present in very large numbers in the middle and posterior parts of the large intestine. The col- lections [rom the Port Gawler area agree with the original description, which ean now he amplified by measurements of more specimens (Table §) and by Figures 11-14, The lateral alae, present in both sexes, are not very wide; in the mule they extend from the posterior oesophageal region to near the cloaca; in the female they are restricted. commencing just posterior to the oesophagus and reaching to, not past, the vulva. The nerve ring is 190-210 4 from the head in females, rather more anterior than described earlier. The eggs are oval in shape, contain a bent larva, and measure L0G « by 59-60 p. Skrjabinodon smythi n.sp. (Figs, 15-21) Type Host and Locality. Phyllodactylus nurmoratus, Port Gawler. Other Jovalities: Middle Beach, Chowilla, Loxton, Lock, all in South Australia, Type 6 and allotype @ will be deposited in the South Australian Museum. This species appears to be common in P. marmeratus throughout southern Australia, although another species (unpublished) apparently takes its place in northern parts of the state, tt has been taken from fourtcen of seventeen host animals examined, In most cases there are about 6-8 worms in cach host, but in some there are more, the greatest number being sixty-seven of which 35 were males, 7 females with eggs, and 22 females without eggs, All ovcur in a mass in the small caecum at the origin of the large intestine. Where few worms are present, the gravid lemales are stuffed with eggs, but where there are many worms, the eggs are few, There is an apparent variation in the position of the vulva and excretory pore in the females. These appear to be oesophageal structures in some collec- tions and well behind the oesophagus in others. This however is largely dependent on the degrec of contraction of the body. It has been noted that those worns in which the excretory pore (and vulva) are oesophageal are stiff and barreltike in appearance, with strongly marked ringed or ruched cuticle, whereas in flaccid specimens with smooth cuticle these two pores are further back, This is shown hy the measurements, in Table 8, for $. smyth! as, although the total length measurements show a great variation, the length of the ocsophagus and the [gil spike do not. In this Table, measurements are also given of some very flaccid specimens from Chowilla. Lateral alae are present in both sexes, from about the midlength of the oesophageal region to the level of the anus, The amphids are large. mere promi- nent, and slightly further back, than the four large cephalic papillae. The three lips are bilobed in the female and single in the male. At the anterior end of the oesophagus of the male are three small teeth, not present in the female. The position of the nerve ring is not clear in mast specimens, The excretory pare is a transverse slit with cuticularised lips and it lies posterior ta the wcsuphagus dn relaxed specimens, It is more posterior in the male than in the female, The male is without caudal alae; the male tail spike has a few very minute spines. There is no spicule, but cuticularised projection of the cloacal wall is 70 L.. MADELINE ANGEL ann PATRICIA M. MAWSON TABLE 3 Measurements of Thelandros kartina, T. trachysaurié and Skrjubinadon wmylhi- All nvusurements ara in jo unless otherwise indicated. The tail of the male S. vnylht fram Part Gawler is broken. Hpuciex T. kartana T. trneh yaar. | Si smytht Locality ' Port Gawler | Middle Bouwh | Port Gawler | Port Gawler Chowilla. ae of wonlraatior (eontracted) (rotaxedl) (flaccid) Males: myth (111) 163-0 | 1628-5 1-40-1-78 2-4-2-5 1) 75-23-10 vosophapus 320-480 =| 9 5A0-6H0) 220-300 | 250-275 200-31) antr. end-cxer, pore | 50-1000 Oo0-OtN) SAU-8 70) 600 650 400-490 tail spike 50-74) \ 100-120 ahU-410 4110420) 450. FAO apicule 65-75 100 150 — — Yemales: | | length (rn) 2-1-6-5 3°3-3-5 31-4 Fi 4-0-7°-0 | 6-9 Bed beso phage 400-L200 820 980 ANO-AfiN) 470-500 440-600 antr. end-exer, pore 520-1400 00-1000 200 40 5HU-GH) 550-700 -yvulva 3-0-4°8 1-2: 350-45) 650-TH0 GON-7580 tuil 100-150 340-450 1390-1500 1350. 1800 15-1600 taal spike —_ 1900-1050 800 -1050 HO0=LO00. present, The preanal, adanal, and postcloacal papillac are almost evenly spaced on the ventral surface, Dorsulaterally to each adanal papilla is a papilja-like extension into, but not lifting, the cuticle, and terminating at a tiny pore, These: are probably the phasmids. The tail of the female bears about 7-9 irregularly spaced projections which are more digitiform than spinous (Fig. 20). The vulva is a transverse slit, without thickened lips. The eggs are 150-165 » long, with a plug at cach end and slightly flattened on one side, The most mature eggs contain a larva 120 p long. The genus Skrjabinodon was proposed by Inglis (1968, 179) for some species which had been placed in Parathelandras Baylis but which differed from the type species of this genus, and from other species which he described at that time. The species attributed to Parathelundros are all from Australian frogs, and those to Skrjubinedon are from lizards, mostly from places other than Australia, but one, §. oedurae (Johnston and Mawson) from an Australian lizard. The present specimens agree generally with other species of Skrjabinodon and with the gencric diagnosis proposed by Inglis, except in the two characters which he marks as doubtful, namely the absence of onchia at the anterior end of the oesophagus {present in the male of $. smythi) and the lip shape, which does not uppear to be bilobed in the male of S. smythi. The species is differentiated from S. anolis (Chitwood), the only other species of the genus in-which the spicule is absent and the female tail bears large “harbs”, by the shape of the barbs, the larger size of the eggs, the longer oeso- pligus,. the more anterior position of the anus in the female, and by the rather different spacing of the male caudal papillae. Skrjabinelazia sp. Host and Localities; Phyllodactylus marmoratus, Port Gawler, Middle Beach, The material available consists of several female worms from the intestine of five geckos. No males have heen found. The worms are large, up to 18°3 mm in length, and the cuticle at each end is markedly inflated. The anterior end bears four large papillae and two amphids, There are no lips, The mouth, more or Jess circular, Ieads into a short buccal HELMINTITS FROM SOME LIZARDS Tl cavity from the walls of which project 2 ring of tiny teeth, like an internal leaf erown. The oesophagus, 1:2 mm long, widens towards its posterior end, but is not obviously divided into muscular and glandular parts, The nerve ring is at about one third its length from the head, and the excretory pore at the same level. Cervical papillae were not secn, The posterior end narrows suddenly 300 p behind the anus, and the bady proper ends in a short spike about 110 j long. The inflated cuticle extends behind this for about 50 p. The vulva lies shortly behind the nerve ring, about 490 » from the head, The eggs are large, sub-spherical, 90 » in diamcter, and contain a coiled embryo, The egy shell is thin and apparently not rigid, as some change their shape under pressure. These worms are very similar in appearance to those of the genera Skyjabinelazia Sypliakova und Salobrella Freitas. The species of these genera differ, as far as the female is concerned, in the absence or presence, respectively. of lips, so the present specimens are identified as Skrjabinelazia sp. In the absence af inales, no further identification is attempted. No species of Skrjabinelazia, or af Salabrella, has so far been recorded from Australian hosts, ACANTHOCEPHALA Host and Locality: Wemiergis peronii, Port Gawler (2) and Middle Beach (3), Five acanthocephalan cysts were taken from the mesenteries of these hosts. They have been examined by Dr. S. J. Edmonds, who has kindly given the following information. All the cysts appear to belong to the same genus, In only one is: the introvert extended enough to permit an estimate of its measurements, The. ellipsoidal cyst is approximately 800 » long by 300 » wide, The introvert, which is cylindrical and bears numerous hooks, is about four-sevenths extended, and is estimated 45 1100 » long and 260 » wide, The number of hooks is hard to estimate, possibly about 80 rings cach of 26 hooks. The only acanthocephalan so far recorded from Australian reptiles is Sphaerechinorhynchus rotundocapitatus (Johnston, 1912), from the black snake Psewdechis porphyriacus Shaw, The cysts from H. peronii da not belong to this species, REFERENCES Curiwoon, B. G., 1934, Reports on the collections ohtained by the first Johnson-Smithsyonian sleep-sea expedition to the Puerta-Ricun Deep. Two new nematodes, Smithsonian Mise, Coll. (Publ, 3243) 91 pp. Duturvus, R, P.. 1922. Variations dans: la forme dw corps, la position et Ia forme des tostionles chez Dicracoelinm lanceolatum (Rudolphi), Bull, Sue. zool, France, 47, pp. 312-342. Faenmas, J. F. Teixeira de, 1941. Sobre um intercssante nematodes parasita de reptil (Spiru- roidea) Mem. Inst. Osw, Crnz., 28, po, 239-245. Incers, W. G... 1968, Nematorles parasitic in Western Anstralian trogs. Bull, Bro Minw mut Hist. (Zool.), 16, pp. 13-183. Jemaxston,, T. H., 1912. Notes on some Entozoa. Proc. Roy. Soc,, Queensland, 26, pp. 6+91_ JouNston, T, 1,, 1932, The parasites of the “stumpy-tail" lizard, Trachysaurus rugosus, ‘Trans, R. Soc. S, Aust., 56, pp. 62-70; Jounstow, T. H. and Mawson, P. M., 1941. Some nematodes from Kangaroo Island, South Australia. Records S. Aust, Museum 7, pp. 145-148. Jouxsron, T. H. and Mawson, P. M., 1947. Some nematodes from Australian lizards. ‘I'rans. RK, Soc. $. Aust, 71, pp, 22-27- MacCattum, CG. A., 1921. Studies ia helmintholosy. Pe 1 Trematodes. Tt. 2 Cestodés, Pt & Nematodes, Zonpathologica 1, pp. 137-284, 72 L. MADELINE ANGE! ann PATRICIA M, MAWSON Matan, J. R., 1939. Some helminths of South African lizards. Onderstepoort Jour. Vet. Sci. and Anim. Ind., 12, pp. 21-74. Nicott, W., 1914. The trematode parasites of North Queensland. 1. Parasitol, 6, pp. 333-350. Smytu, M., 1968. The distribution and natural history of a skink, Hemiergis peronii (i Fitzinger ) Trans. R. Soc, S. Aust. 92, 51-58. Svassxy, A. A., 1951, Anoplocephalate cestades of domestic and wild animals. (In Russian), Osnovy Tsestodologii, 1, 735 pp. SypuiAxova, O. M., 1930. Sur un nouveau nématode des reptiles: Skrjabinelazia taurica, n.g., nsp. Annis. Parasit. hum. comp., 8, pp. 615-615. Timon-Davip, J. and Timon-Davin, P., 1967. Contribution 4 la connaissance de la biologie des Dicrocoeliidae (Trematoda, Digenea). Recherches expérimentales sur le cycle vital de Paradistomum mutabile (Molin) parasite de la vésicule biliaire de Lacerta muralis (Laurenti). Annls. Parasit, hum, comp., 42, pp. 187-206. Travassos, L., 1919. Contribuicao para a sistematica dos Dicrocoeliinae Jooss, 1899. Archivos da Escola Superior de Agric. e Med. vet., Nictheroy, Brazil, 3, pp. 7-24. Travassos, L., 1944. Revisio da Familia Dicrocoeliidae Ohdner, 1910. Monog. Inst. Osw. Cruz,, 2, 357 pp. TWO SPECIES OF SACCOGLOSSUS (ENTEROPNEUSTA) FROM SOUTH AUSTRALIA BY I. M. THOMAS* Summary A new species of the Enteropneusta, Saccoglossus aulakoeis is described. It possesses a deep, dorsal, longitudinal groove on its proboscis. It is compared with three previously described members of the genus, which have similar grooves. One of these species, Saccoglossus otagoensis (Benham) has been found in the same vicinity. This is a new record for Australia. The two species however, occupy different habitats, S. aulakoeis being found in coarse sand and shellgrit in amongst the roots of Zostera while S. otagoensis, in this locality, has been found only under stones at or below the level of low water spring tides. TWO SPECIES OF SACCOGLOSSUS (ENTEROPNEUSTA) FROM SOUTH AUSTRALIA by I. M. Troxtas* [Read 8 August 1968] A new speeies of the Enteropneusta, Saccoglossus aulakoeis is deseribed. LU possesses a deep, dorsal, longitudinal groave on its probosvis. It is campared with three previously described members of the genus which have similar grooves. One of these species, Saccoglossus otagoensis (Benhani) has been found in the same vicinity. This is a new record tor Australia, The two species however, occupy different habitats, $, aulakocis being found in coarse sand and shellgrit in amongst. the routs of Zostere while 8. ofagoensis, in this locality, has been found only under stones ut or below the leyel of low water spring tides. During an investigation to determine the distribution of Saccoglossus apan- tesis. (Thomas, 1956) in South Australia, two other members of the genus have been found. One of these is a new species and the other a new record for Aus- tralia.. Material of both species was fixed in Bouin’s-in-seawater and scctioned at 10 4. Sections were stained either with Ehrlich’s haematoxylin and eosin or with Mallory’s triple stain. The latter was very eflective for showing the basement membrane and the skeletal parts derived from it. Proboscis skeletons were extracted by maceration of unfixed material in 4% borax for several days. After carelul cleaning by brushing, they were stained in aniline blue and mounted in cedarwood oil. Saccoglossus aulakoeis n.sp. The trivial name is descriptive of the deep, median grooye on the dorsal side of the proboscis (avAaé = a groove or furrow), Specimens have been found at Port Willunga, South Australia (lat. 35° 16° 50” S, leng. 138° 27’ 20” E) in a shallow tidal pool about half an acre (0-2 hec- tares) in extent about a quarter of a mile (400 m) south of the rnins of the Port Willunga jetty. The pool is covered to a depth of five or six feet (about 1-6 m) at high spring tides and about half its floor is exposed at low spring tides. Zostera tasmanica G, V. Martens grows sparsely in small patches at about the level of normal spring lows and in amongst the roots of some of these patches, the enteropneust has been found, Il is by no means common, The soil over the under- lying rock is not more than an inch or two-deep and consists of sand and shell grit with little or no mud, The animal has not been found in sand without Zostera and many patches of Zostera do not contain it. One specimen has been found in amongst the roots of Cymodocea antarctica (Labill.) which grows abundantly in slightly deeper water in the pool. The pool is partly protected by a low, rocky elevation (covered at mid-tide) on its scawards side, but it is open to the sea at all states of the tide. At Brighton, (lat. 35° 02’ S, long, 138° 31’ E), about thirty miles north of Port Willunga, after a heavy storm, some masses of Zoster roots were washed * Zoology Departnient, Uniyersity of Adelaide, Trans. Roy. Soc. S.A. (1968), Vol. 92. T4 T. M. TITOMAS ashore on a sandy beach, In amongst the roots were found four fragments of enleropneusts similur in size to the Port Willunga specimens, One of these frag- meyts included a portion of a pruboscix with a deep dorsal groove, Sectioning of this material has confirmed that it is $8. aulakoeis. The species agrees with the diagnosis of the genus Saccoglossus Schimke- witsch, 1892 (syn, Dolighoglossus Spengel, 1893) iu the possession of the follow- ing characters; (a) proboscis very long, (b) collar about as long as broad, (¢) lateral genital folds present but dorsal gonads absent, (d) gill pores small lait dis- tinct, (e) perihacmal spaces present, (£) posterior oesophageal pores present, anterior ones absent, EXTERNAL FEATURES S. aulakoeis (Fig, 1) is a species of small to moderate size, Full grown, intact specimens are diffeult to obtain but a specimen of reasonable size which las heen relaxed by the method recommended by Ledingham und Wells (1942) (isotonic magnesinm chloride). had the following dimensions, Proboscis 16 mm; collar 2+5 mm; branchial region 5 mm; oesophageal region 2 im; hepatic region 13 mm; intestinal region 24 mm: genital region (which overlaps from the branchial to the hepatic region) 10 mm. The gonads were not folly mature. The proboscis is orange-red and its base, usually hidden by the collar, slightly darker, On this the preoral ciliary organ (Fig. 1) shows up as a U-shaped yellow line. The anterior three-fifths of the collar is similar in shade to the proboscis, The remaining two-fifths bears two broad, slightly elevated bands, paley and yellower and separated by a narrow groove, the posterior wall of which shows wp as a still paler band. This latter band corresponds to zone four in the histological divisions of the collar epidermis (Fig. 6). Dorsally and laterally, the branchial region is slightly browner than the pro- boscis and is beset with irregular flecks of yellow-orange. These are groups of eland cells (epidermal glandular eminences, Figs, 1, 7, 8 and 10), The ventral musculature of this region is again similar in shade to the proboscis. but with narrow and slightly paler transverse striations. (Fig, 1), The oesophageal region is similar in colour to the dorsal branchial region hut the flecking, dorsally and laterally, is denser. In the hepatic region, the body wall is somewhat translucent and the surface flecks on dhe epidermis are smaller and sparser. Tn livmg soeci- mens, the deep brown colour ef the lateral saccules of the liver region of the alimentary canal show through clearly. The intestinal region, too, is transhicent and the sandy gut contents show through, The flecks on the epidermis are still evident but ure more widely spaced and are smaller than they are on the hepatic repion. In preserved specimens, the proboscis tapers slightly throughout its length but in ving aud fully relaxed material, it has the form of « very elongate pear. heing appreciably thicker at its base. It has a deep mid-dorsal groove which, at its deepest, is a third to a half of the diameter of the organ and it extends almust to the anterior extremity (Figs. 1 and 4). Posteriorly, in the vicinity of the proboscis complex, it is shallower so that on the basal face of the proboscis, its depth is about one-fifth of the diameter. The preoral ciliary organ has a pattern similar to that in other members of the genus in which it has been described (Brambell and Cole, 1939; Brambell and Goodhart, 1941; Thomas, 1956). It is U-shaped and lies an the base of the proboseis closer to the stalk than to the outer border of the hese. The dorsal tins of its arms, however, are detected inwards and herein it dif- fers slightly from the organ as it is pictured by Brambell and Cole (1939b) in 5, suber (syn. 8. cambrensiv Burdon-Joncs and Patil, 1960) in which the arms are ON a a Fig, 1. Saccoglossus aulakoeis, anterior end, lateral view. Fig. 2, Oesophageal region, dorsal view. Fig. 3. Transverse section, dorsal proboseis yroove. Fig. 4. Transverse section, proboseis. Fig. 5. Proboscis skeleton, yentral view, Fig. 6. Zones of collar epidermis, b., body of proboscis skeleton; b.I., basement layer: ¢., collar; ce, 1 to 5, zones of collar epidermis numbered from anterior end; d.pr.. dorsal groove. of pro-~ hoseis; e.gLe,, epidermal glandular eminence; g.p,, gill pore; g.r, genital ridge; h., horn of proboscis skeleton; i:c.m.pr., inner circular muscle layer af proboscis; k,. keel: Lb)., lateral blood vessel; lm.pr., longitudinal muscles of probosvis; m.d.bl., mid-dorsal blood vessel; n.J., nerve layer, O.c,m.pr,, outer circular muscle layer of proboscis: o.p., sesaphayeal pore; p.c.or., preoral ciliary organ; pr.proboscis; pric., proboscis coclom; pr, sk,, proboscis skeleton; s., sole, formed hy ventral longitudinal muscles; th.n.l., dorsal thickening of nerve layer: v.gle., vacuole of epidermal gland cell, v6 1. M, THOMAS straighter, There is a single, slit-like proboscis pore on the stalk, close to its junc- tium with the collar and slightly to the Jeft of the mid-te. The collar is longer than its diameter in the proportion nf about 1/0 to 0-77, It is also longer dorsally than it is veatrally im the proportion of about 1:0 to 0-95. This is dne partly toa slight retraction of the ventral part of the anterior Hange under the mouth and partly to the dorsal part of the posterior lange extending farther over the branchial region than does the baistrale part (Fig. 1). The pos- terior two-fifths, approximately, of the collar bears two broad, slightly elevated bands of about equal width and separated hy the narrow groove already mentioned, Anteriorly, the branchial region is somewhat quadrangular in transverse sec- tion as most of the main musculature is veutral in position, The ventral surface of the body thns toms a broad sole on which the animal creeps (Fig, 1). Farther back in the branchial region, the sole has a shallow median ventral groove { Figs, 7 and 8). There are from twelve to twenty-five pairs of small gill pores set in shallow dorse-lateral depressions. The anterior two or three pairs are coyered by the pos- terior flange of the collar. The Just five or six diminish in size and get successively closer to the mid-dersal line (Fig, 2). In small decinatte) specimens the first gonads are seen wbout half-way alony the branchial region but in filly mature specimens, they hegin closely behind the collar, Depending on the maturity of the specimen, the gonads form more or less conspicuous genital ridges (Fig, 7). In immature specimens (Fig. 8) the ridges are relatively inconspicuous, These extend to the beginning of the hepatic region in mature specimens and end rather abruptly. In immature specimens they do not extend as far back, In the branchial region, the gonads are rather lateral in position but behind the gill pores they hhecume more dorso-lateral oe 2). About 1 mm behind the last gill pore lie the ocsophageal pores (Fig. 2). These are arranged in two rows one on each side of the mid-dorsal line. The: pores number from two to seven or eight pairs but the more anterior ones are not patent. The specimen shown in Fig. 10 hud two patent (1 and 2) and three rucdti- mentary (3, 4.and 5). The disposition of the rows is variable. Seven specimens were examined in this regard, In three of these the rows were almost parallel to each other and ta the mid-dorsal line, in three they were slightly divergent. the more posterior being the farthest apart and in one they diverged at an angle of about 30°. This specimen is shown in Fig. 2. The larger and patent pores are the most posterior. In the anterior oesophageal region the sole (Figs. 7 and 8) is as broad as it is in the posterior branchial region but begins to narrow towards the posterior cnd of the oesophagus. The hepatic region can he recognized hy the more or less regular lateral dilatations of the alimentary canal which can he seen through the somewhat transparent body wall, The sole here narrows further and becomes less in width than the dorsal part of the body though its lateral bulges aré still apparent, In the intestinal region the hody tapers slightly to the terminal anus and the ventral musculature diminishes in amount so that the sole disappears sind the hody is rounded in transverse section. INTERNAL ANATOMY Praboscis The glandular and ciliated epidermis of the praboseis extends to the bottom of the dorsal groove and beneath it Hes the nerve layer (Fig. 3 and 4), Over most af the proboscis, this is about 12 to 14 « thick, At the bottom of the proboscis groave there is. a marked ridge in the nerve layer so that at this puint it is 30 to 32 TWO SPECIES OF SACCOCLOSSUS 77 Fig. 7. Succoylossus aulukodis, transverse section of nearly oature fermale in first region of oesophugns. Vig. §, As Fig. 7 but of immature female. Fig. 9. Longitudinal sagittal section of proboscis conyplex. Fix. 10. Longitudinal section showing oesophageal pores. hic,, buccal cavity; c.. collar: c.co.. collar cord; c.cl., collar coelom; ¢.s, central sinus (blood sinus); d.bly.. dorsal bload vessel; dium. dorsal longitadinal muscle; dn.c. dorsal nerve cord; e.gle., epidermal glandular emi- nence; gl,. glomerulus; hev., heart vesicle (pericardium ): ic.m.pr., inner layer of vircular muscles of proboscis: |m.pr., longitudinal muscles of proboscis; Lin.p.c.. longitudinal muscles of peri- huemal cavity; Lw.d.gr., vertical section through wall of dorsal groove; m., mouth; n.ljpr., nerve Jnyer of proboscis; o.c.mpr,, outer layer of circular muscles of proboscis; oes, 1, first region of oesophagus: o.p. 1 and. 2, first and second oesophageal pores (patent); o.p. 3, 4, and 5, third, fourth and fifth acsephageal pores (rudimentary ); o.p.sk., supporting skeleton of cesophageal pores; ov., ovary; pr., proboscis; pr.c., praboscis cotlom; pr, ca., proboseis canal (endsav); p.sk.. proboscis skeleton: sid., distal portion of buceal diverticulum; st.p., proximal portion of buccal diverticulum; tr.c., trunk coelom: y.bl., ventral blood vessel; v.lam., ventral longitudinal mus- cles: v.n.e., ventral nérve cord. Ts LM. THOMAS zw thick (Fig, 3). There fs also i general thickening nt the nerve layer at the bise of the proboscis. particnlarly under the preoral ciliary organi where it may be three times its normal thickness (Fig. 9) The outer layer of circular iuscles immediutely under the basement membrane is a half to two-thirds of the thickuess of the nerve layer. The longitudinal muscles form nine or ten concentric layers separated by loose connective tissue. These are compressed and barely distin cuishable immediatcly under the dorsal grouve (Fig, 4) and they are reduced to five or six in number posteriorly in the vicinity of the proboscis complex (Fig. 9), An inner, thin Jayer of circular muscles limes the coelomic cavity, The latter extends almost to the tip of the proboscis. Anteriorly it is narrow, being only about one tenth of the diameter of the prohoseis, but it expands considerably pos- terlorly to accommodate the proboscis complex. Two blood vessels are present immediately outside fhe outer circular musele layer on the erests of the ridges formed as the result of the presence of the dorsal wroboseis groove (Fig. 4). A small, subneural vessel (median dorsal vessel) has heen seen in some specimens in the mid-dorsal line but a median ventral blood vessel lias not heen observed in specimens so far examined. The buecal diverticulum (stomochord) (Fig. 9) has a wide, ventral pocket the posterior wall of which is indented by the blunt, anterior tip of the proboscis skeleton. The diverticulum is bent slightly backwards at its tip under the end of the skeleton. The lumen of the pocket is wide and broadly in contact with the main lumen of the buccal diverticulum within the neck of the proboscis but it is not confluent with the huceal cavity. It is occluded at the level! of the peint of migin of the horns of the proboscis skeleton. Anterior to the ventral packet, the buccal diverticulum has no continuous lumen, but only a series of uncon- nected cavities which: diminish in size anterforly, This anterior part of the buecal diverticulum differs from the form usual in the genus in that it is sharply dividect into two regions of more or less equal length, The proximal part is thiek and conical, while the distal part is very thin being only about one sixth of the diameter of the proximal part at its widest, The distal part is attached antero- ventrally and curves dorsally to end near the anterior extremity of the glomerulus. Its cells are much smaller and less vacuolate than those of the proximal part but thy separate portions of its lumen are apparent almost to its anterior end. The glomerulus (Fig. 9) surrounds the distal part of the bueeal diverticulum bul al the level of the proximal part it is almost entirely lateral and ventral. [+ ceases posteriorly at the level of the ventral pocket of the huceal diverticulum. The curdiac vesicle (pericardium) and ventral sinus (Fig. 9) call for no speciul voninent, The dorsal mesentery of the probuseis extends forwards nearly to the level of the constriction of the buccal diverticulim, The ventral musentery is shor- ter, extending forwards only. to the level uf the ventral pocket of the buccal diver- Hivulum, The Teft coelomic pouch so formed, communicates as is usual in the genus, throngh the proboscis canal (endsac) with the prolioseis pore which is dor- solateral on the left side of the proboscis stalk, The proboscis skeleton (Figs. 5 and 9) is slender, terminating anteriorly in a rounded tip which partly penetrates the posterior wall of the ventral pocket of the huceal diverticulum. The ventral keel is well formed and posteriorly it bifureates to become continuous with the homs. These extend aliout halfway along the collar and embrace about half the circumference of the buccal cavity. Collar The five transverse, epidermal zones of the collar distinguished by Spengel (1893) are present (Fig. 6). The first, which overlaps the anterior Range of (he collar, contains cells which have large vacunles distally. This zone meres into the TWO SPECIES OF SACCOGLOSSUS 7 second. This forms an epithelium, which, at its thickest, is more than three times as thick as the first zone, and occupies nearly three-fifths of the length of the colliv. Its cells contain large numbers of small basophil granules in the imyer three-quarters Of (heir lengths. The third zone and the fourth and fifth zones com- bined, form two elevated bands which are readily visible externally. The third and fourth zones are little more than half as thick as the second, The third is histalogi- cally similar to the second except in that the basophil granules are concentrated in the inner half to two-thirds of its cells. The fourth zone is by far the narrmayest and forms part of the posterior wall of the groove. Its granules are more densely packed and are fairly evenly distribnted throughout the length of its cells. In the fifth zone the granules are more sparsely and evenly distributed and towards its posterior end there are peripheral vacuales which are characteristic of the epithe- lium of the brinchial region with which this zone merges. The nerve cord of the collar is solid throughout its length, The dorsal anvseu- lery is Incomplete and ends anteriorly a little behind the point where the pro- koscis skeletan divides to form the two horns, The ventral mesentery is mare variable in extent, In some specimens it extends as far forwards as the level of the posterior tips of the hors. ie. about halfway wong the collar while in others it js apparent only near the posterior end. Trienk The gill pores imerease in size over the first two or three and decrease in size uver the last four or six. The posterior gill pores are very small and may lack gill youches. Behind these there may be two or more rudimentary gills. The tongucs project farther into the pharyny than do the septa. The gills extend a little more than half-way around the circumference of the pharynx. In the oesophagus there are three regions. In the first, the epithelium is similar to that of the ventral part of the pharynx but its walls are thicker and have a higher proportion of gland cells, Its lateral walls are irregularly sacculated (Fig, 10). In the second region the walls are much thicker and the lumen correspond- ingly narrower, At its untero-dorsal end there are deep grooves in the dorsolateral walls into which the oesophageal pores open. Of these, two lo five are patent and two to four do not open to the surface but are represented hy outpushings of the dorsolateral grooves. There may be cormesponding indentations of the epidermis wbove them, All ure supported by skeletal eloments which usually fuse to form an almost continuous plate which is perforated in the positions of the patent and the nou-patent pores. These clements are, like the other skeletal structures of the animal, thickenings of the basement laver which underlies the nerve layer over the whole of the body. In the third region of the oesophagus the walls are thinner and the lumen wider, Laterally there is an irregular sequence of shallow pouches. This region merges into the hepatic region which differs from it maimly in that the lateral pouches are Jarger and more regularly arranged. In the intestinal region, the alimentary canal is simple. Its wall is thin and the lumen wide. There are hawever, two ventral thickenings separited by a narrow, median, longitudinal groove which extend nearly to the posterior end. COMPARISON WITH OTHER SPECIES Three other species af Saccoglossus have been described as possessing dee dorsal grooves an the proboscis. These are 8. mereschkowskil (Nic, Wagner) 188) 5 S. ofayvoensis (Benham) 1809, and S, sudcates (Spengel) 1893. While the first two have been adequately described (van der Horst. 1939), the third was described from three anterior ends only which became dried up so that detailed examination was impossible (Spengel, 1893), ait 1. M. THOMAS The main differences between these three species and S. aulakoels are listed in Table 1. The relevant data on S. mereschkowskii, 5. otagoensis and S, sulcatus, have heen taken, in the main, from van der Horst, 1930 and 1939. §, aulakoeis is established on its possession of the following combination of characters. 1. A deep dorsal proboscis groove. 4. The ventral musculature forms a promiticut “sole” in the posterior bran- chial and ocsophageal regions, 3. The collar is slightly longer than broad. 4, The fourth epidermal zone of the collar epidermis is very narrow. 5. There are two to five pairs of patent oesophageal pores preceded by twa to four rudimentary ones. 6. The longitudinal muscles of the proboscis ure arranged in nine or ten concentric layers. Epidermal gland cells extend to the bottom of the proboscis groove. 8. The glomerulus covers the anterior end of the buccal diverticuhim. 4, The buccal diverticulum has a very narrow distal portion which is not conspicuously bent and in which the lumen is incomplete. 10. The proboscis skeleton is bluntly rounded anteriorly, 1l. There are no cavities in the dorsal nerve cord of the collar, DISCUSSION 8. aulakoeis has been found only in the restricted regions indicated in the introduction. Even here it is not common. It is rarc to find as many as two or three specimens in a spadeful of soil, It is interesting however, that frequently specimens varying in size from 1-5 cm to 5:5 cm may he found at the same time. This suggests that the breeding season for the species is an extended one or that there is a considerable variation in growth rate, However, animals with mature or maturing gonads are generally seen in late winter and early spring (Fig, 7), while in the summer months the gonads are invariably small (Fig. 8). Mature specimens do not coil markedly in the post-lranchial and oesophageal regions as do those of S. apantesis (Thomas, 1956), This is, no doubt, to be associated with the lessee degree of development of the genital ridges, In all specimens examined, with the exception of those fixed without adequute narcosis, the collar is slightly longer than it is broad, The definition of the genus states that the collar is “about as Jong as it is broad”. The proportion ot the length to breadth of 1-0 to 0:77 is considered to fall within the limits of the definition but the extent of the ratio is noteworthy. Non-patent oesophageal pores have heen described for $, ruber (syn. 8. cam- brensis Brambell and Cole, 1939) and they also occur in the present species, If will be necessary to determine their presence or absence in several other forms before reasons for their existence can be discussed, ‘There is a possible association between the dorsal thickening of the uerve layer and the dorsal groove of the proboscis. Both S. ofagoensis and S. aulakoeis have deep dorsal grooves and a thickening of their nerve layers. Similar thicken- ings of the nerve lavers have been described in S. ruber and in $. horsti (Brambell and Goodhart, 1941) and in these too there is a dorsal groove in the proboscis TWO SPECIES OF SACCOGLOSSUS $1 though it is net nearly as deep as it is in the first two species named. In S. apani- tesis however, the dorsa! groove is only slightly developed and the thickening of the nerve layer is also slight. While the description of this species was being prepared, some specimens which clearly belonged to the same genus were found under stones on Snapper Point, This hes about half a mile (about 800 m) south of the pool in which S. aulakoeis had been found. These specimens were at first thought to be the same species as they also had u deep dorsal groove on the proboscis, Closer examination and Sater, a study of longitudinal and transverse sections showed that this was not the case but that they were Saccoglossus ctagoensis (Benham). Balanoglossus ofagoensis, Benham, 1899 Dolichoglossus otagoensis, (Benham) v:n der Horst, 1930 Succoglossus otagoensis, (Benham) Van der Lorst, 1939 (p. 399) This is the first record of the species outside New Zealand, Three fairly com- plete specimens and two fragments were found under stones in about one foot of water at low tide on the northern (more sheltered) side of Snapper Point Reef (Jat. 35° 17° S, long. 138° 26’ 30” E), This is an extensive wave-cut platform of almost horizontally bedded sandy limestone of Pliocene age. It ix relatively hard in parts but on its northern border and in some other regions, it is softer and somewhat friable. Thus, along its northern margin, a secondary, narrow: reef flat has been formed about two feet below the general level of the main rect. This is covered by about au foot of water st normal spring lows. One Specimen was found under a stone in a permanent rock peal about six inches deep on the main reef surface. It is a crawling rather than a burrowing species, as is 5S. aulakocis and it agrees closely with the description of 8. etagoensis as given by Benham ( 1899). and van der Horst (1930 and 1939). The points of undeniable resemblance are asterisked in Table 1. In regard to the remaining points listed, the collar is rather shorter than it is broad in the proportion of abont 1-0 to 0-70. This may be due to contraction in fixation. The specimens from Snapper Point are relatively immature so that the irregularity in the lateral genital bulges is not very apparent. In regard to the presence of gland cells in the proboscis groove, van der Horst (1930) states that they are absent from the base of the groove and his Fig. 2 (p. 137) shows them to be present in the lateral walls, This is the case in the Snapper Point speci- mens too. However in his diagnosis of the species (1939) he says “Keine Driisenzellen in der Epidermis in der dorsalen Furche’, which implies that they are absent from the groove altogether. This seems to be an oversight as they are present in the walls of the groove (though not at its base) in specimens in the author's possession which were collected at Portobello. New Zealand. which is close to the type locality of the species. Van der Horst (1930 p. 139 and 1939 p, —61 ! describes the buceal diverticu- lum as haying two marked flexures in front of the ventral diverticulum. This is figured in his 1930 description (Fig. 4, p. 139), The Snapper Point specimens do not have these marked flexures. It is telt that these may well be fixation artifacts in van der Horst’s specimens, In New Zealand, the species is found in amongst coralline algxe at Wellington and amongst the holdfasts of Macrocystis at Portobello. This shows it to be a crawling rather than a burrowing species and indeed, Benham in his original account of the specics writes of it as crawling on ua stem of seaweed, 1, M. THOMAS pepunor Ayunyq aoz -ajays sfasoqoid jo dr} Joraywy ¢ “Kraarp yeoonq Jo dy puosaq | Ajroliayue = spuayxa snyniauta[g | sonxay payrwist ynoyyM nq ApIOUe}UR LOL -Inu AOA WNpUIAearp Peony yayood [ey -UaA OU} JoLia}ue snonuyuos you WH[NOWaAD PRONG JO WeluNT AYAYD YINOUL OJUT Nodo JOU seop woynoaniaarp ywoong jo AyAvy SBULI OLIUaOUOS {)T 10 § WL sapsnut [eUTpPNySuoT 9AQOIS slosoqoid [esiop jo gseq ur jquasaid spars ypeuuepidy | gjeutay pue opeul yyoq | ut qvpnBar Apey sprog TeMuaD peor wey] dasuop Apysys sBoD | proaq sv suo] sv ynoge AvTJOD | juiod dreys “auoj yya uo} -ajays stosoquad yo dy stopeyay, wnNITyIIAIp Jeaong Jo joy Ul payun jou SnMsawo9p3s ayy JO saaTey OML, APIOMA}UR MOLTB fsarnxay jowayue way} peIyUaA poyrelw quan winqMaLAaAtp yeoong snonuiyu09 Un MonAeATp [Been jo SULIT AYACI TQNOU OUT uedo Avot lunMoHaArp [e90ngG JO AWARD, SSULL OLNQUSAIOD P JO © UL saposnul [BUIpUpPSWOT, aAooIs stosogoid [esuop yo aseq Ul spur[s [euLspide ON ayewoay Ul saspnq ey] -NSaIM FO WIOF UL SPlOF TeWUOS) proiq sv Buoy se qnoqr seyjOD qurod divys “suo, yyy U0} -g[ays stasoqoid jo diy Joey umypnoyiaarp Teoonq jo [Je Apuau sano snpniewlo[s qysreys APPAQU[aT WinynoAwaAip Rong snonurywo09 WINnpWarHeATp yeaonq fo wsuM’y slosoqgold SBULL S1}UB009 OL OF L Ur epost Teurpnysu0T A[ao apeuiay uy Surmeadde 34y -ngal ssay Jo 3100r Sp[OF pequsy pvorq se Suaoy sv ynoqe AOD saingoa y [DULay : slaoynynn *s Ssnjoopns °S SISUBOTDIO 'S MYSMOYYISAAIUL *§ a “sla0yD]ND *g pu snynopns. *g “HYSMOYYIsasauL -g ‘s|sUaORDJO snssojzog9vg Jo samyvay aysouderp Jo wosriedurop T HTaViL 83 TWO SPECIES OF SACCOGLOSSUS ‘sualujoads puvjeaz MeN pue UeyeYsny NOG UsaAjaq AjETLUTS yo sjaiod sayeorpuy , Bypeasny wos AK IOVFO ‘NY Ys FIND qsoura 4g AJVAG QUO UL BAO Ad aroUr IO g soxod AIVUaTUIT pI jo sired % 0} g pue sarod wad -vydosao yuayed jo sued ¢ 0} suoyoeas ¢ YYWAX susdeydosag Ss yo sued og 04 Z] saspii [v4912] stonordsuos WUOF Sapsuu = RUIPNySuey yeruaA piod aarauU [sIOp UL SaVIAvD ON ajo[dmoour yng yasaid Solayuaselt [LIJUVA PUR TesI0g sauoz [Butapida 3Aly SSoUYIIY] [LUIOU Jo syutseprd| AYIABI tqnour Fey yNoqe aoviquia pur Ipyjoa Buope Ava-Jey joqe pua) “x9 UO}IfeXs stIasoqoid Fo su10}F{ uvdef | SITES Jo sured [] 0} OT FOOOULA "19 pug PuRlfeoZ Mon AIBAO OUO UT BAO adir 9 uel} azo TUOPT9S gy sarod yeaseydosso jo amd aug, suolpes g YIM snsrydosag Sys JO sed cy 07 OTs saspi [Blaze] WIoF Jou Op Yyuns jo Sappsnur [RUIPNysuoc], [eyooA, p40. vAtou JPSIOp UT SaT}TAvr), yuasqv SaLlo}Uasaul [BQUaA PUL [BSIOC sauoz Teuoapida dear], , ssauyaty] [eUuOU jo stuaprd'yz, yoo et) pus Jopuny yeal pure yeyuozoy udzJoOAS stosoqoid jo SUIOY ,. JO Rag OY} O} LIssNY UIsIION Uu01}D90°T saiod [vaseydosao jo sued y ynoqy suotpoas § Ym suseydosag S]T4 jo sed gg Jnoqy yun, pi02 OAraU [esIOp UT saryTatry aqyayduioaur ynq yuosaid SASPUSsAU [RIYUDA puUe Tesi0dy (war co‘9) IGT AIA stuwopidy Ingjog Slaoyojnp *¢ snyp2 ns “S HSUZOBDIJO “S inysmoyyosasau ‘5 _-e_,———— Eee 84 J, M, THOMAS REFERENCES Bennam, W. B. (1899). Balanoglossus otugoensis n.sp. Quart. J. Micr. Sci. 42; 497-504. BramMBE.L., F. W. R. and Core, H. A, (1939a). The preoral ciliary organ of the Enteropneusta; its occurrence, structure and possible phylogenetic significance. Proc. Zool. Soc. London. B., 109, 181-193. BRAMBELL, F, W. R. and Coreg, H. A. (1939b). Saccoglossus cumbrensis sp.n., an Enterapneust ovcurring in Wales. Proc. Zool. Soc. London. B., 109, 211-236. Braneert., F. W. R. and Gocpmart, C. B. (1941). Saccoglossus horsti sp.n., an Enteropneust cecurring in the Solent. J. Mar. Biol. Ass. U.K. 25, 283-301. Bunpex-Jones, C. and Patiz, A. M. (1960), A revision of the genus Saccoglossus (Entero- pneusta) in British waters. Proc. Zool. Soc. Lond., 134, 635-645. Horst, C. J. van der (1930). Observations on some Enterapneusta. (Papers from Dr. Th. Mortensen’s Pacific Expedition 1914-16). Vidensk. Medd. naturh. Foren. Kobenhavn. 87, 135-200. Hors, C. J. van der (1939). Hemichordata, Bronns Klassen u, Ordnungen des Tierreichs Bd. 4, Abt. 4, Buch 2, Tiel 2. LepincuaM, Isaper. C. and Wexxs, G. P. (1942). Narcotic for marine invertebrates. Nature, London, 150, 121-22. SpENcEL, J. W. (1893), Die Enteropneusten des Golfes von Neapel. Fauna u, Flora des Golfes von Neapel. Monogr. 18. Tuomas, I. M. (1956). Saccoglossus apantesis a new species of Enteropneust from South Australia, Trans, Roy. Soc. S. Aust. 79, 167-176. LUNETTES IN SOUTHERN SOUTH AUSTRALIA BY ELIZABETH M. CAMPBELL* Summary Lunettes are low crescentic ridges, which commonly occur on the eastern side of ephemeral lakes. They are of two types-those composed predominantly of sand, and those composed of sand, silt and clay. The mineral composition of lunettes varies, though lunettes rich in quartz or in gypsum are common. The morphology, composition and distribution of lunettes in southern South Australia are examined in the light of suggested theories of origin. The deflation hypothesis of Stephens and Crocker and many other workers can account for most, though not all, of the field evidence. Lunettes develop as a result of wind action, but wave transport is an important factor in concentrating debris on the Ice sides of lakes. The available evidence suggests that although some lJunettes are apparently still forming, they are essentially relict features dating from the recent past. LUNETTES IN SOUTHERN SOUTH AUSTRALIA by ExizapetH M. Campseii® [Read 12 September 1965] SUMMARY Tameltes: are low éréscentic ridges which ecmmonly occur on the castern side of ephemeral Jakes. They are of two types—those composed predosuinantly of sand, and those composed of sand, silt and clay, The mineral womposition of Juncttes varies, though lunettus rich im quartz or in gypsum are common, The morphology, composition and distribution of Juncttes in southern South Australia are examined in the light of suggested theories of origin. The deflation hypothesis of Stephens and Cravker and many other workers can account for most. thongh not all, of the Held evidence. Lamettes develop as a result of wine action, but wave trausport is an important factor in concentratng debris on the lee sides of lakes. The available evidence suggests that although some Iuncttes are apparently still forming. they are essentially relict features datiny from the recent past. INTRODUCTION The purpose: of this paper is to describe the shape, cornposition and internal structures of lunettes in southern South Australia, and to discuss their possible age and origin. The term lunette as here used includes not only the crescentic dunes located on the lee side of lake and swamp depressions and composed variously of silt-clay, clay-loam and sand, but also the so-called gypsum dunes of Jack (1921) and others. | Features of similar morphology and situation have been described from several parts of the world, including North Africa (Tricart, 1954a; Boulaine, 1954) and North America (Coffey, 1909, Huffman and Price, 1949), They have also been recorded from several parts of the Australian continent. In general, they are best and most commonly developed in the semi-arid regions of Victoria, South Australia and Westem Anstralia, thongh they occur also in humid zones, for example Tasmania, and in truly arid regions as, for example, the north of Sonth Australia, DESCRIPTION Morphology and Size The dune on the eastern side of Lake Bumbunga, Mid North, 6-4 km long, 21 m high, and 1:2 km wide, is an example of the larger lunettes known in South Australia, At the other end of the scale are the small rises about 0+S km long, 4-5-6-0 m high and 45 m wide, typical of the flat, swampy interdune areas of the Southeast. There is no consistent and direct relationship between the size of each Iunette and the extent of its associated lake, Lake Greenly, Eyre Peninsula, which is about 23 km? in area, is bordered by a Junette which is only just dis- tinguishable. Lake Baird, Eyre Peninsula, on the ather band, which covers only 1-3 km? is bordered by a lunette 2 km long, 21 m high and 0-8 kin wide (Fig. 2), Nor is there any consistent distributional pattern; in a given district, some lakes have associated limettes, athers do not. ~ Geography Department, University of Adelaide. 1 For the locution of all South Australian place naiues mentioned, sec Fig. 1, Trans. Roy. Soc. S.A. (1968), Vol. 92, 86 ELIZABETH M. CAMPBELL Yongala *Hutlant Lageon Bordertown Moone Flab Ss * Sere wy odulla eXyanculta wHites Lagoon eSnowlawn *Loke Gumbunga én mark i Barri | I *ADELAIDE #Cookes Plains Lake to mots oe eCumm'nt Greenly® li \ ®Port bineoln Mt ing Yorketawn Siske fowler hearst White aqean Enlargemest A “Sarndurce 4 Kongoreo Island PReth + Burden iow { Enlargamen! 6B Frances ; i ” Narorsa:tel loke Cadn . Cockatao Lake Lo] 4 loo mj as f “soll tale { ° Penola I ° igo kilometres Mount "Lochaber 5 um bey ; = Swamp Unnamed ve Hynom Swarnp . ae ae et Hy 6 am pant Nayacoerle Unnamed « Smiles lose d — week Rest ys — | a lakilomet es —— Roads ———! ='™Stale & Skilemerres \ Bel lagoon Bowndory ny i Baebes Fig, 1, Southern South Australia—Location, Enlargement A. Enlargement B. Lunettes are crescent-shaped in plan and extend between one-quarter and one-half the way round the perimeter of the lake (Figs, 3, 4, Pl. 1).* They are most commonly found either on the eastern or southeastern side. The smooth form of the lunette on its western or lake side contrasts with the irregular or sinuous shape of the eastern slope of many lunettes. The occurrence ® For reasons not yet clear, an wonamed lake on the Jaterite plateau of Kangaroo Island is completely surrounded by a dune, whilst several depressions south-east of Renmark are bordered on their western side by gypsum dunes. In the south-west of Western Australia between Kindinin and Corrigin where summer easterly winds are significant linettes occur on the western side of depressions (C. R. I'widale—personal communication). See opposite page Fig. 2. Contour plan of Lake Baird lunetic. Plane table survey. Five feet = 3-3 metres, 500 yards = 500 metres, LAKE BAIRD LUNETTE CONTOUR INTERVAL to FEET ARBITRARY CATUM seater tin yarns} o igh - os Height in ¢ a o6 j o * o 8 » o sa Vertical Exaggeration 3-75 8&8 ELIZABETH M, CAMPBELL LOCHABER SWAMP LUNETTE CONTOUR PLAN Contour interval S feet Arbitrary datum Lochaber Swamp Scale in feet ro) Q 500 19000 isoo 7 ‘ct = st Artificial Drain oO PROFILE A-B Vertical exaggeration 7:1] me an 8 a0 4 40 = 3 = 20 ae 20 > 0 A x 500 1000 «=1SO00 feet B 17-18°5-65,E.C+ DC. Fig. 3. Contour plan of Lochaber Swamp lunette. Plane table survey. Five fect = 1-7 metres, 1500 feet = 500 metres. LUNETTES IN SOUTHERN SOUTH AUSTRALIA LEGEND Arbitrary datum Contour interval 10 feet ——— Road Quarry 2400 feet Fig. 4. Contour plan of Cooke Plains lunette. Redrawn from King 1949a. Ten feet = 3.3 metres, 2400 feet = 800 metres. 89 90 ELIZABETH M. CAMPBELL of consolidated rock outerops at the Jake edge, as for example at Lake Greenly, introduces complications and further irregularities in the plan, Many lakes are bordered not by a single Junette, but by two or more luncttes arranged concentrically, or nearly so, around the eastern shore. At White Lagoon, Kangaroo Island (PI. 2), for example, there are no fewer than three distinct and separate dune ridges bordering the lake on its eastern side. The pair of lunettes which borders Lake Fowler, Yorke Peninsula, and those close to the eastern shore of Hiles Lagoon, Mid North, are of similar composition, but at White Lagoon the outer or easterly dune is sandy, the westerly ridge immediately bordering the lake is composed predominantly of silt and clay, and the one between these is : —____ fe er eer ee —I, — 500 7000 See an vee 2 ——— ee ae 2nd 490 HORIZONTAL DISTANCE Fig. 5. Transverse profiles across selected lunettes from west to east (left to right). A—Moona Flat lunette, B—Lake Cadnite lunette, C—Lochaber Swamp Imette, D—Cockatoo Lake lunette, E—Hutton’s Lagoon Junette, F—Hiles Lagoon lunette, G--Bool Lagoon lunette. LUNETTES IN SOUTHERN SOUTET AUSTRALIA a composed of sand, silt and clay. More complex patterns of lake and multiple lunettes are displayed at Bool Layoon and Salt Lake near Naracoorte, Southeast (Pl. 3). Here complexes of smal! lakes each with their associated lunette or lunettes, occur within larger lake basins on the caste margins of which are found individual luncttes. Most Junettes in southern South Australia are asymmettical in transverse section, the steeper slope being the western or lake slope (Fig. 5). The gradient of the western slate of Bool Lagoon linette (Pl. 4) averuges 1 in 3+3, whilst the eastern slope is 1 in 18-3, Some few, however, are roughly symmetrical and several examples have been noted in which the castern slope is steeper than the western. This is so along part of the Lake Baird hmette (Fig. 2), part of the middle lunette on White Lagoon and along most of the inner lunette at the same site. The western slopes are tor the most part smooth and rectilinear and, though varying from one hinctte to another, are essentially uniform on any given feature, The inclination of the eastern slope, however, varies considerabl , even on the same Iunette. In longitudinal section, the crests of lunettes typically rise gradnally to a high point near the centre of the feature (Fig, 3) but many crests are undulating (Pigs. 2, 4). On the innermost of the White Lagoon linettes, there are no [ewer than four distinct crest lines, all of them undulating, the high points of one crest invariably located opposite depressions in the adjacent ridge (Fig. 6), Fig G Diagram showing the multiple crestline of the linette mnmediately adjacent tw White Lagoon. viewer! from the est. Composition Samples of lunctte material were collected on the crest where it reached its maximum development, and from the lake hed adjacent to this position, Borings were made to below the level of soil development. On the basis of particle size, lunettes are of two distinct types—those com- posed of almost pure sand and those composed of sand, silt and clay (Fig. 7), The predominantly gypscous and predominantly quartzose sandy Junettes are clearly differentiated. The sorting value developed by Trask (Krambein and Sloss, 1963 p. 101) is a measure of the uniformity of the sample and is derived from the formula Sy = Q:/Qs where §, is the sorting value and Q; and Qs, represent the grain size values corresponding to the 25% and 75% marks respectively. of the sample. The sorting value of the sandy Iunettes is 15 or less, that is. they are well sorted, whilst the fine-grained lnnettes are poorly sorted, with a value ranging from 5-6 to 7:6, Field determinations of samples collected at various depths and from various positions across the linette indicate that the mechanical composition is generally uniform, although there are Iocal and minor irregularities, 92 ELIZABETH M. CAMPBELL 100 1 I, H Per { ——$ _—__—__—— a = 4 © Fy . e Lunettes. | e 1, Cocke Plains Gypteoyve) | a 1 r. Zlake Fawler Sandy Quariose \ A 3, Loke Cadnite Lunerics Sandy . . 4, abe Board } Lumens 7 £ 5, White boagoon 2 b Lochaber Swenp £50 = > FHynem Swamp | a 8 Cockulos Luhe = \ “ o o r a s = Fineegriined Lunettes 25 _ -|— eC —_ | ° cas Al 4 n 3 E 3 u 3 = 1G 4 i} 4s 4 05 2) 005, “007 Pesticle Size (millimeters } Fig. 7. Cumulative percentage curves of the size distribution of Imette samples, The mineralogy of samples from six Junettes and their associated Jake beds was analysed by Australian Mineral Development Laboratories (A.M.D.L.), Adelaide. The < 0-002 mm fraction (Table 1) of the lunette samples is com- posed predominantly of “amorphous” material, which was not positively identified but which shows physical properties similar to illite and smectite, or “amorphous” material and illite. Quartz and kaolin ave present in all cases, and illite-chlorite, illite, calcite and chlorite in some. The > 0-002 mm fraction was separated into a light and a heavy [raction (S.C. 2-96), The light minerals (Table 2) include quartz which is abundant in all and dominant in most, silt- and clay-sized particles too small to be identified by the methods used, and, in some samples, minor amounts of calcite, layer silicates, organic remains and opal. The heavy minerals (‘Table 3) vary from 0-007%-0-31% of the total sample, They are numerous and vary from onc lunette to the other. Hematite, magnetite, other opaqnes, tourmaline and zircon are the most common.- Gypseous lumettes were not included in the selection analysed by A.M.D.L. Apart from varying amounts of gypsum which is mainly the “seed” variety, calcite, halite and insoluble materials, including quartz, opaques and many other minerals are present in some samples (Table 4), There is generally a close relationship between the mechanical and mineralo- gical composition of samples taken from the hinette and from the associated lake bed, though minor variations occur, The mineral composition (‘Tuble 4) is LUNETTES IN SOUTHERN SOUTH AUSTRALIA 93 TABLE 1 Semiquantitative analysis of clay minerals in samples from Iunettes and associated lake beds. Minerals Present. Weight | Amor- | Illite- Sample % of | phous | chlorite | Quartz | Kaolin | Tllite | Caleite | Chlorite total | Material oe joes. open or bs: _ a Moons Flat lunette < 10 bed 50% § == Subdominant 20-50%, A = Accessory 10-20% T = Trace <10%, F = Faint trace just detectable. TABLE 2 Analysis of the light fraction mineralogy (8.G. <2+96 and diameter ~-(-002 mm) of samples from lunettes and associated lake beds. | Mincrals present a3 percentage of light fraction Sample Weight Silt Layer Organic oof | Quartz + Opal sili- Calcite | Remains total clay eatcs. Moona Flat linctte > 90 100 bed >90 of 5 Lake Cadnite lunette >90 100 bed &3 75 20 5 bed (surface) > 90 100 <2 Lochaber Swamp lhinette 68 BO 65 5 bed so | 85 15 Coekatoo Lake lunette 63 | 75 as) 20 <5 bed 37 30 | 50 20 Hutton’s Lagoon lunetto 92 24 70 5 bed 78 40 5h | 5 Bool Lagoon lunctte 72 20 i 50) 25 5 bed 24 | 15 i 5 60 10 10 | J} ILIZABETIT M. CAMPBELL 4 ‘Aye pug ApuyypezsAro tood jo esneoeg LOLyBoylyUapt eautsod UON = 4&4 (uoowdy Aavey Jo % OT >) Tueserg = q (uoorry AABOY Jo % 07-07) UouuIE)) = 9 (uoyariy AARoy Jo %OF<) yaRpunqy = | 20-0 LOO +0 s0-0 60-0 S00 01-0 c+) £00 Té-0 el: 90°+0 9T-0 OT+0 l | 1 5 —|=-|a)/—|/—|/ala a)—p—l~ia,alvljal|o}al— ad|—|d/ai-|~|a a d | vid|vlaj— -|— 1 A} di—~);/o};a}) wy) a|— 1 a Ley) poh i ] od d ¥y)|¥ | — SS ee d. | | : d d Viw|d - a d d d|—~!l—;,—) ad ad ral d Vv e) ad pa é dA d ~—~iji—~|d d;ja7}]d/|—id d | V d|—i-~]o — | — || d ad fae d d j—|]7/]) a oo! da2//]9 Vi—i—|d ; d ral i => i a - od d | V dj}-|da --j}—]a@]/—:—!|a]a ala }—j/—lai—j}o}oj}o;—|— i d d a ) d |] a a lB Peet Pon et ee ca a eal] = '—|a)/ajy—la}a —|—!—|—-la@)al vw} oa}¥}—|— 2 | ole wpl>la slolwl/>lz|/elsilels Be le/e F/El2 S/S FF 2/2/2128) 2) 2) F e;S |e /F/F/E|F BLES) S|) a's /8)/8 | 8/8] 8 SE |S | Pl ele Sala Ee) PR SR] E & | & = es h > se) a | = Fy a o mf =e a|/F)e Ble] a! = Fi eile]? 3 5 2 od a 5 | z § o |S g e | | | z S| et vle 4 folie | | + | \ x quesad T S]RaoUTyAy O80 | paq TO-0 | ayJoUN] UGOse'T [oog 90-0 | paq O1-0 ayyauN] Woose'yT s uoyNy]T FI-0 | peq FI-O | 2}, GUN] BYVTY] VOY VO) 90-0 | paq LO-0 | ayjaun, duremg aaqeyooT Th-O | (oanyans) poq ér-a peq 90-0 | Spwuny] syUpeEy eye] 91-0) | peq OT-0 | a]jJOUNy ZBL BUCO]Y FV a | a og ig ame | ajduaeg 33S =| | 32, “spaq eYE] poyeLoosse pu soz, 9UN] WoIy saydures ur {"wIW Z0-0 < dayeWeIp put 9g:g < OS) syesaUTO AAvaT, Jo sis(jeue Arey yUrnhruag £ ATAVL LUNETTES IN SOUTHERN SOUTH AUSTRALIA 85 TABLE 4 Mineral comnposition of two pypscous lunettes and their associated lake beds. Minerals present “ree Gypsum pp Sodinm | Ferric | Insolu- Total E bonate Chloride! Oxide bles Cooke Plains lunctte (average of 16 samples) (a) 83°88 2°80 na. na. 11:21 ' 97°89 Cooke Plains bed (trial hole , $12:2'107-3'2”) (a) 47-77 | 21-70] ma | na. 18-43 - 87-90 Lake Fowler lunette (average of 33 samples) (b) 92-9 | Mel Q+7 (hel 3:9 98-7 Lake Fowler bed (sample C4:0-294”) («) 75-70 1-11 1:95 0-34 18-3 97-4 Source: (a) King (19494 p. 145) (b) King (1949b p. 66) (¢) Hiern (1957 p. 40) n.@, not available. TABLE 5 Percentages by weight of sand, silt and clay. (The silt and clay fraction was not separated. in some samples: sill and clay—less than 0-066 mm. sieve size.) | Sample i Sand Silt & Silt Clay Clay Cooke Plains lunctte 98:4 16 Lake Fowler lunette 98-7 1:3 Lake Baird lunette ! 99-4 0:6 White Lagoon lunette 93-2 6-8 Moona, Flat lunette : 96-1 3:9 bed 96-4 3:6 Unnamed lake 2 lunette 97-0 3:0 bed , 97-2 2-8 Unnamed lake 1 lunette 94-2 5:8 bed 92°5 7-5 Lake Cadnite lunette 95+2 4-8 bed T4e4 25-6 bed-surface 2” 98°38 1-7 Hynam Swamp lunette | 37-5 B7-d 25:0 Lochaber Swamp lunette 45-0 32-1 21-9 bed 85-0 4-0 1150 Cockatoo Lake lunette 33-0 37-0 40-0 bed 79 92-1 Hutton’s Lagoon lunctte 34-2 65-8 bed 37-7 §2-3 Tliles Lagoon lunette 29-7 70-3 bed 41-1 58-9 96 TABLE 6 percentage of the total sample. s from lunettes and lake beds each fraction expressed us 1 Percentage size distribution (by weight) of sample size in millimetres: range of particle Yas Pipette Anal ELIZABETH M, CAMPBELL ‘ a -.7 SMONRM SOKO Her MAMI HMMA gs AM SOPRANO HOA I GAARAS w id A AMARA SSE ia ro FO 0-G0070 na amor 8000-F00-0 esos = Lm Moon 10-0-800-0 conn See | Aaas TEO:0-9T0-0 mrewnn — = j IDK 20 06 990-0-T80-0 RRS or -= oH st oHas Sigecnmarcrhemedtdscon 680°0-990-0 | Sofa poe nennateetoeoweas (SAE ANHONARAINSO HO MSEetTAN FEL-O'G80-0 | Soa ma nartaonsemangerncedts aN = Ae “4 Hee IEE SH err tte Me KoINge SLI- Fel - 0 HN DOB OSSU SHQHSNHOR HINS NOAA Naas 4 MOMHROMMOADHIMOMANMAANOAMSA G8S-ORLLO | Km omoanr aes traporanseren NANA HAN oe “= Zor OSSCHe AE ONE StHSS Come 2 (294+) SCHOBHODHNOBNNR HN SON Hee oo tel ed eS ene ee) Mo HSeGAAMS Se soOSNIHesE NOg-O-SSh +0 HBEAASS AMHR oSSOMAONSONANS a a SCAGCSCH RTI SANAMotoNSsoaN ETL-0-008 0 SSASSSSOAMAONS Sessoangeon ars - Sa = P nae : ian) Ho SI DS) ole 1 EM TIL 0 gists esnanss Ne ams 71 f00°-L= | sis x = eS z 3 22t_2 4% i ec 7 s He ra eal es t, Seo SEITE FE Beyond stZoR £ESFLEvsFs * seeeb Sesser. 5 1” 3s a = eeg bum wn SSETH Fe EF ca FRatn are gee i] C 3 2 = ; os 5 oO a aff ag Pay =o 5 =Sae5Ro g s32 2 & es Hee ome + gv a2 3 @ «& 4 #8382 83 2 42 fa 4 4 2 2 Seiecr tz ts Re og mm Ww ‘Roo sat a 3 a g c Pa load eHAge &€ € LV a¢ ro 2 4 woor ed 8 € sa Ff BS oH#M ES fF EF Y Bos ‘s zw = Seats = -— 3 Bo 6 § = (CHRBR DP PA 5 Oo f = LUNETTES IN SOUTHERN SOUTH AUSTRALIA 97 TABLE 7 Quartile measures (in mms) and sorting af samples from hinettes ancl associated lake beds, \ Sample Q, | Mg Q: | Sa Cooke Plains hinette 0-85 0-61 0:35 | Ls Lake Fowler lanette 0-68 0-61 0-37 143 Lake Baird Inneatte O26 ! 0-2 0-16 1:3 White Lagoon Iunette 2] | 0-16 O12 1-3 Monona Flat lunette 0-18 : O15 : 0-12 1-2 bed 0-18 15 OIL 1-3 Sandy Lake. hinette 0-23 | 0-19 (14 1-3 beet 0-25 0:20 0-16 1-4 Northwest. ake lanetic O32 423 O- 16 1-4 bed H Oap 0:20 - 14 I-5 Linke Cadnite hinctte | 0-38 0-26 0-18 bed bed 0-26 | 0-16 0-06 2-4 bed—surtace 2" +25 0-18 O14 1-3 Wynn Swamp limette 0-086 0+ 0084 00018 7-6 Lochaber Swamp linette 0+096 O15 0- 0026 t-1 bed Q-17 0-098 0-44 1-0 Cockatoo Lake linette Q-Q31 | ()- 0966 (he (01 56 bed | na. H T.w, ra, Lith. Huttou’s Lagoon lunette 0: 107 | TB, na. ied. bel (125. i iA. Lh. ni. Hilegy Lagoon lunetty ()- LL nA na, 11.8. bored 0-105 | Tha. nea. | 1,8. Sy O1/ Qs mem > notoayailable, the only information available by which to compare the gypseous sandy lunettes with their associated Jake beds. The same minerals are present in both cases and the percentages vary only slightly. The percentages by weight of sand, silt and clay (Table 5), the size distribution (Table 6, Fig. §), the quartile measures and degrees of sorting (Table 7), and the mineralogy (Tables 2, 3) all indicate that the composition of the quartzose sandy lunettes is very similar to that of the associated lake bed. The size distribution of samples from the fine-grained luncttes varies slightly from that of the lake bed samples, especially in the small size grades (Table 6, Fig. 8), but the mineralogy of the two is similar (Tables 1, 2, 3). The sample from Lochabcr Swamp lunette, Southeast, shows a lower percentage of light minerals than might be expected from the percentages observed in the same lake bed (Table 2). Uowever, the composition of the Jake bed may have been altered by the man-made drain entering the swamp (Fig.3). Other dissimilarities, for example, at Cockatoo Lake, can be accounted for by the difficulties associated with the analysis of the samples. Structures Internal structures have not been observed in fine-grained Innettes, bul there are several exposures in yypseous and quartzose sandy Junettes, The most exten- sive of these are in Cooke Plains and Lake Cadnite lunettes, Southeast. Cooke Plains linette is composed almost entirely of seed gypsum of sand- grain size, In the quarry on the southern end of the lunette, near the swamp edge, the dip of the beds is about 2° towards the Juke; near the centre of the lunette the din is in the same direction but varies over short distances from 20° ta almost flat (Pl. 5); and on the side of the Iunette away from the swamp, the dip is consistently abeut 30° away from the depression, 98 100 ELIZABETH M. CAMPBELIL LAKE CADNITE 75 oy 2 SIZES) 25 - CERTAIN ry {LARGER TRAN 3 Ss 75 30. PERCENTAGE a5 CUMULATIVE COCKaToOo PARTICLE 100 7s 40 05 PARTICLE “01 ‘O05 -O01 DIAMETER (MILLIMETERS) i i | HUTTONS LAGOON HILES LAGOON (MILLIMETERS) DIAMETER Fig. 8 Cumulative per centage curves of the size distribution of lunette and lake bed samples. The solid line represents the Iunette sample, and the dashed line the sample from the lake bed. In the case of Lake Cadnite, the dotted line re- presents the sample from the surface of the lake bed. LUNE TTES IN SOUTHERN SOUTH AUSTRALIA bY The structures of Lake Cadnite lunette, composed jlmost entirely of quartz sund, are only visible in isolated places where the quarry face has been exposed to weathering. Typical measurements near the centre of the lunette are 2° to 12° lo Uhe east (PL. 6), but whatever the value near the centre of the feature, the dips in all cases Hatten to the east. A series of superimposed soil profiles is exposed in a Quarry face to the cast of the crest for a distance of a few metres, Each «oil consists of a black humus-rich layer which grades to light brown with depth and overlies buft-coloured sand. This rests with sharp demareation on the black honzon of the next lower soil, Vegetation Cover The surface of nearly all the Iunettes in southern South Australia is protected by a dense cover of grasses, Trees are rare, hut red gums about 6 m high are lound on Lake Cadnite lunette, and the Cooke Plains lunette is covered hy dense Mallee scrub, The grass cover has apparently been sufficient to protect the lunettes from crosion; the only occurrence where modifications to the original Junette form are evident are those in which wave action and slumping have resulted in a cliff on the lake cdge, or where overgrazing by shecp and rabbits, or overcropping, have resulted i erosion of the surface. ORIGIN Precious Investivations The two principal hypotheses advanced in explanation of Innettes are those of Hills (1939, 1940) on the one hand, and Stephens and Crocker (1946) on the other, Lt should be noted, however, that although Stephens and Crocker first gave prominence to the deflation theory, similar ideas are either implicit or explicit in the writings of Coffey (1909), Grayson and Mahoney (1910), Wyherzh (1918), ane ITarris (1939), Hills suggested that spray droplets whipped from the surface of the lake by wind and dew formed preferentially on the dawnwind side of the lake cause the deposition and retention of atmospheric dust on the lee shore of the lake, The presence of the water in the lake is thus an essential feature of this mechanism. Hills did not preclude the possibility of the addition of minor amounts of material by deflation from the dry lake bed but considered this of minor siznificance. Stephens and Crocker and many later workers {Huffman and Price, 1949; Gautier, 1953; Boulaine, 1954; Tricart, 1954a, b, 1955; Jennings, 1955) considered, huw- ever, that this latter process was of prime importance. They considered that wind scouring of the unconsolidated sediments exposed in the dry Jake bed provides material which is trapped hy vegetation geawing on the lee shore of the lakes. Once the lunette has developed, it propagates itself by deflecting the wind over it whilst the material in transport is deposited on the lunette. The essential contrast between these two hypotheses is that Hills’ requires water in the lake while the deflation hypothesis required a desiccated lake hed. Thus it is that these lwo hypotheses have been construed by different authors as implying the association of the hinette cither with wel ( glacial) or dry (inter- glacial) phases of the Quaternary. In view of the marked seasonality of climate and lake conditions in southern South Australia at present, it seoms immecessary lo invoke such definite and long term changes of climate. But the age or ape range of lunettes is certainly significant, for, if the features developed under 2 Various other hypotheses have heen put fonvard io explain ximilag landfornis in other parts © the world (Cooke, 1934, 1940, and Raisz, 1934: Grant, 1945: Melton and Schriever. 133; Frouty, 1952), hut none seems ta have any relevance to the situation in southern South Australin, 101) ELIZABETH M. GCAMPBRU), conditions widely different from those which obtain at present, modern climatic parameters have little or no relevance to considerations of their evolution. Thus, before any assessment of the various lines of evidence, and a discussion of the possible genesis of these Teulures, it is necessary to examine (heir likely age or age range. Age of Lunettes Soil profile development on lunettes varies considerably but most have leer sufficiently stable to allow the development of mature soils on the surface, Sisni- larly, a layer of “flour” gypsum, the weathering product of the “seed” varicty, occurs at the surface of the gypseous lunettes. This is an indication that weather: ing has latterly outpaced accumulation, suggesting, that lumette formation has ceased or has ut least slowed down. However, it is not necessarily evidence af antiquity; as soils can develop in a relatively short time, as evidenced by horizon development on sediments no more thar 100 years old at Beefaeres, an Adelaide suburb (C. R. Twidale, personal communication ), There are isolated indications that linetles are forming today; the bed of an unnamed depression 13 km. northwest of Naracoorte is bare, but flat-topped remnants standing 4 metre above the bed are vegetated. This is taken to mean that fhe mesa slopes and the bottom of the unsalted depression have heen recently removed, apparently by deflation; movement of material across the lunette on ITutton’s Lagoon has heen observed: and there are reports of luneites forming today in Western Australia (Bettenay, 1962), Victoria (Baldwin, Burvill and Freedman, 1939) and Texas (Caftey, 1909; Huflman and Prive, 1949) under conditions similar to those in southern South Australia. In any case, the evidence points to the luncttes heing geologically youthful rither than to their antiquity. From a variety of evidence throughout southern Australia, Stephens and Crocker (1946) concluded {hat “it is apparent that the halk of the Tnettes belong to the lute Pleistocene to Recent: Period”. King (419492) concluded that the Cooke Plains gypsum dime postdated the emergence of fend after the wid-Recent high sea-level when conditions permitted the evapo- ration of water and the precipitation of salts, In the Riverine Plains of south- eastern Australia, Pely (1866) noted that lanettes predate the most recent period uf stream incision into the plain but postdate the lust phase of deposition by the prior streams which occurred in the Pleistocene (Pols, 1964), Bowler and Ifarford (1966) in the same area, worked out a geomorphic sequence in which they retog- nized three phases of lunette formation: the first oceurred about 7,000 B,P., indicated by a date from lacustrine silts in the hed of the lake on which the hinette formed: the second occurred following shrinking of the lake: and the thirel alter a period of prior stream activity, deposits from which were dated by the carbon 14 method at 4.000 B.P. Bowler and Harford thus showed that linette formation in this area began at least 7,000 years ago, and continued at least until 4,000 years ago, that is, through the middle Recent." Whilst lunettes in southern Australia may vary in age, it is likely that their age range is of the same general order, and it can be assumed that the lumettes in southern South Australia developed during the Recent. They are, therefore, a relatively youthfnl feature of the landscape. Sinew this time, the climate of sonttiert: Australia is thought to heve changed only slightly se Uiat conditions fayunrable to linettc formation are not very different Gout the range of those ‘Carbon 14 determination of material at the use anel at the snrlice ofa linete offer Te next satislactory answer to the range af time clupine which formation Wok place, Datahl fuiterial has so Fur heen found only vear the surface of lunetles in South Austrlia aud poage donominitions lave heen carried out. LUNETIES TN SOUTITRRN SOUTLIE AUSTRALIA td expericnced at present. An examination of present climatic parameters is there- fare relevant in a sludy of the origin of luncttes. Discussion of Orisin The fine-grained hinettes in southern Soufh Australia are composed of 76-90% mulerial less thay O-L mim diameter (Fig. 7), ie. it can be transported in stes- pension, All the hinettes deserfhed by Hills (1939, 1940) ia north-western Victoria wre dominuntly fine-grained. Most of these lunettes could have been formed by the deposition of atmospheric dust on the lee side of the lake, Deflation is not, however, excluded, Most of the particles in the sandy lunettes and sore in the fine-grained luncttes are too large to be transported jn suspension, and thervfore cannot be explained by the mechanism suggested by ills, Other lines of evidenve fram scuthern South Australia throw further doubt on this theory. No measurements for the precipitation of dust over rural areas. are available, but the amount precipitated in wetter months, as required by the atmospheric dust hypothesis, must he extremely small (R. Culver—personal communication ). ly view of the youthfulmess of lunettes, it is considered that there has heen in- sufficient time tor the large amounts of material required to make up a lunette to have heen deposited from material suspended in the atmosphere. Variation in composition of linettes In close proximity and in an area of uniform soil types is incompatible with the atmospheric dust hwpothesis. Cackutna Lake and Lochaber Swamp, no more than 10 km apart, are associated with lunettes of quite different composition (Tables 1, 2, 3), and the composition ot the three hunettes associated with White Lagoon, as stated previously, varies considerably, The position of the hinettes on the castern side of lake depressions indlcates formetiun by westerly winds, and, in general, winds in southern South Australia hlow froin this seetor (Fig. 9), In most areas where luncttes are numerous, these winds are the least dust-bearing as they blow for only short distances across land (Stephens and Crocker, 1946), Twe morphological characteristics of lmettes which arc atypical of dunes, their smooth and regular contours as opposed to # hummocky surface, and their asymmetry with the steeper slapes on the windward rather than the leo side, were interpreted by Hills (1940) as evidence against the deflation hypothesis of Innette formation. There are, however, several factors in the lunette environment different from that of dunes in general, which can account for these anomalies, The surface of Junettes is covered and protected hy vegetation, thus preventing major re- shaping of the material, and the development of a hummocky surface. A slow and steady (although not necessarily continuous) rate of accumulation which js in- tlicated by the presence of low angle dips and the absence of marked diseon- formities in bedding.” would allow such a vegetation cover to develop and be maintained. This vegetahion cover, together with the facts that hmeltes are stationary, the rate of supply of material is limited, and most deposition takes pluce near the Jake shore, can account for the steeper windward slope ot luncttes. Tt is suggested that « large and continuous supply of material destroys the vegetation cover and permits the development of a hummocky surface and of steeper lee slopes. In fact, sume lynettes, for example Cooke Plains lunette ( Fig. 4) and some of those in northern South Australia (C. R. Twidale, persunyl communication) do liave a hummocky surface due to wind and water erosion, The 30° dips of the beds in the Cook: Plains limette on the side away fron the luke presumably represent true foreset bedding, hut the low angle beds in Cooke Plains and Lake Ciucluite lonettes, although nat characteristic af transverse dunes, are consistent with an aeolian (rigin, Servicéton G9OOhrs | i Noracoorte OFOChLS Coimas elle Calms Serviceton |5Q00nrs Yoreetown GFOGhes Noo cales jeetieded th SNOWFOWN OF DD hia Yange a OF Obes Colmes Vergata 3G hrs Cols KYAMECUTTA 1500hrs SNOWTOWN fad brs Cabins Al, PARRDANA OF00 hry Cains PORT LINCOLN O8906 hrs BERRI OFOO hrs Cayms ula. Calms sills t PAINDANA [S500 his Calms PORT LINCOLN ‘SDQOh-2 Calms Calnis MOUNT GAMBIER @830 his MOUNT GAMBIER 1490 bys Fig, 9, Wind roses—all winds, The mouthly percentage frequency distribution of wind directions from ¢ight points af the compass. The sides of the ovtigons face tawards the cardingl and semicardinal points—north is at the top. Projecting from cach side are twelve colunms representing the twelve months of the year and the lengths of the columns are propor- tionul to the percentage frequency of winds from the given direction working round clockwise from January to December. The solid black part of the column represents the winds of speeds 13 m.p.h. (21 kilometres per hour) aid greater, and the open part of the colunin represents the winds of speeds less than 13 xi-p.h. The scale is shown in the ventre of each wind rose. The percentage frequency of calms in the twelve months is shown to the right of the wind rose, working from left to right from January tu Devember. The distribution of winds at 0900 hours is shown for each station and at 1500 hours for some stations, The wind roses for Mount Gambier show the distribution of winds from sixteen points of the compass. for 0830 and 1430 hours. ‘The solid black part of the columns represents winds of speeds 15 uup.h. (24 kilometres per hour} and greater, and the open part winds of speeds Jess than 15 m,p.h. 14 ELIZABETIL M. CAMPBELL and some, for example parts of Lake Baird lunette (Fig. 2) are asymmetrical with j steeper lee side, ‘These unusual morphological characteristics of Janettes represent conditions similar to those of normal dune formation, and can be expected in the hmette environment as a result of the season! and Jonger tern yatiations in climalic conditions such as are experienced at preseul. Similarly, normal variations in the climate of today can explain the multiple lanettes which developed at various stages in the recession of lake shorelines." There is little evidence in southern South Australia to support the atrospherie dust theory and much which contradicts it. On the other hand, there im ample evidence in support of the deflation hypothesis of lunette formation, As stated previously, the particle size distribution, the mineralogy, the quarble mneasures and the sorting values of each Innette sample analysed are similar to those of the associated lake bed sample. Several components charac- teristic of, though nat confined to lakes, have been found to luncttes; Hingston aud Bettenay (1960) reported alunite in lunettes, and gypsum exystals make up « large proportion of many, such as those at Cooke Plains and Take Fowler specimens of Coxiella sp. have been found in Lochaber Swamp hinctte and Jhuraceae oospores im Hiles Lagoon lunctte, The close similarity between Ute composilon of the lunette and of its assuciated luke bed suggests that the two are eaisally related, ie, that the material inthe hnictte was derived from the lake hed. as required by the deflation hypothesis, Although Hills discounted the likelihood of the formation of agyrcyates: af particles in the lake beds of north-western Victoria, ageregales have been noted ihuring analysis of samples from southern South Anstralia, and have bee seen in transport across dry lake beds in Texas (Coffey, 1909). Although the process of ageregate formation is incomplelely mmderstood, aggregates are considered to explain the large proportion of particles in fine-grained Iunettes (76-90% ) which are ton small to be transported by defation as individual particles, hut whieh, when bound together, behave as sand grains, The size cornposition of the sandy Jnmettes is also consistent with the deflation hypothesis. If the Jake depressions and their associated lnnettes ure not causally related, the lakes must be explained otherwise than by deflation, In southern Sonth Aus- tralia eximples can be cited where wave action, solution. blocking of surface drainage of Huviatile action have contributed at least in part to the origin of the lake, Lut these proecsses cannot salistuctorily explain all depressions. Owing to tectonic disturbances and the semi-arid climate of the area, condifiars.are Lavon able for the creation of basins of interna] drainage. The periodic aceumulation of water-transported debris in the centres of these basins under conditions of strong evaporation makes possible the removal of this material by wind, There are fumerous examples in southern South Australia where depressions have been formed by this continuing process of deflation. The origin of the lakes is thus consistent with the deflation theory of limette fornntion. ‘i is gonerally accepted (J, T. Hutton—personal communication) that ery conditions favour the generation of electric charges on particles, and these charges murkedly assist in the deflation of material. In southern South Australia the following air temperature and relative humidity relationships usually apply (B. Mason—personal communication ); 6 Variations in campositian of fanettes associated with the one Jake result front variations in Che material in the lake bed whieh bas heen transported by strewoss of varied rates of flow or (iim varivd sources, LUNETTES IN SOUTHERN SOUTIT AUSTRALIA 105 Air Temperature— FP. Relative Humidity—% < 60 > 50 60-80 35-55 > 80 <5 The number of days on which the air temperature reaches these categories varies throughout the arca, bul statistics for Mount Gambier and Yongala represent vilues near the two extremes: No. of Days Temp. reached Air Temperature— PF’. Mt. Gambier Yongalu < 60 120 LOL 60-50 204 159 > 60 41 105 For i considerable part of the year, dry conditions assist in the process uf deflation, However, the deflation hypothesis of lunctte formation, though on hroad terms consistent with the observed evidence, presents some difficulties. First, lunettes are common on the edge of salt encrusted depressions from whieh deflation of the underlying material is impossible. However, the sult erust may be a more recent development. Second, although the climate where luncttes are found varies from arid to subhumid, they are best developed in semi-arid arcas, and not in truly air regions where deflation is most pronounced. Third, an examination of the strong winds (greater than 13 miles per hour— about 21 km, per hour)? for selected stations in southern South Australia shows that. in the wet season (May to October, based on the uumber of rainy days per inonth) westerly and northwesterly winds are predominant, whilst in the dry season (November to April) the stroug winds most commonly blow from a southerly to westerly direction (Fig, 10). Lunettes occur most commonly on the eastern or south-eastern sides of lake depressions. That is, they facc the wet season winds, those which blaw when the lakes are occupied by water, when deflation from the lake bed is impossible. They are not primarily associated with the dry season witids which could achieve deflation from the drv lake bed. These difficulties could he overcome by postulating a change in climate, for exiunple, a change in temperature-humidity relationships resulting in different evaporation rates, a decrease in rainfall during the wet season, or a change in wind direction. However, there is no evidence that such drastic chunpes. ol ave more commonly required, 13 m.p,b. is also the lower limit of one of the classes on the Beaufort scale. Wet KYANCUTTA ae; ‘ Aa wT | Wet PORT LINCOLW Nn | | N | ‘\ aL. aes ———— \ a — th | ai Ca ied ‘Ss a ra ] AQ fe “a a a Wer s . ‘ ee Dry i Wet 5 Mes “ tee 3% ay PARNDANA SERVICETON MOUNT GAMBIER BERRI Fig, 10, Wind roses—strong winds in the wet and dry halves of the year. These wind roses show the distribution of strong winds from the cardinal and semicardinal points of the compass in the wet and dry halves of the year, differentiated on the basis of numbers of rainy days per month, and averaged over the selected stations, The monthly percentage frequency of the strong winds (13 m.p.h. (21 kilometres per hour) and greater) at 1500 hours for November to April and for May to October are expressed as a percentage of strong winds for the year. LUNETTES IN SOUTHERN SGUTH AUSTRALIA 107 Price and Kornicker (1961) nated the importance of waves aid currents in transporting shell fragments on to the lower slupes of clay dunes similar to Irmettes. in Texas. Bowler (1964) suggested that medium io coarse sand is con- centrated in beaches before being blown by the wind on to the lunctte, allhough finer particles, le considered, could not be concentrated in this way. However, although silt- and clay-sized particles are generally carried in suspension in water, some may be conventrated in a beach by one or more of the following methods, Claw particles may adhere to sand grains or themselves form aggregates. The salinity of the water in the lake will determine whether the clay particles wre in (vue colloidal suspension, or are capable of transport by wives in the form ot agerceutes, As J, T. Hutton pointed out (personal conwiunication) wave acting sufficient to cause transport would probably also be capable of breaking up the aguregates; but many remained unseparated after mixing during the analysis of samples. Silt and clay in susperision may be filtered out by the sand and debris ina beach or hy vegetation growing on the edge of a lake, It is suggested, there force, that wave transpnrt of sand, and possibly of silt- ancl clay-sizecl particles, is important in lunette formation. As Stephens and Crocker (1946) pointed ont, the vegetation growing on the vlge af a lake provides the rongh type of surface required ta cause deposition of wind-Lransperted material, The line of vegetable matter, floated on the surface of the Jake and deposited as the water level recedes. has a similur result, as was recognized lyy Woods (1862, pp, 28-29). The effect of waves and the presence of vegetation on the edge of the lake “an overcome each of the difficulties encountered by the deflation hypothesis, First, in the case of those lunettes occurring on the edge of the salt-encrusted lakes, the major part of deflation ocenrs from a beach built hy wave action in fhe wet scason, and not from the luke bed. Second, the reason for lunettes being best developed in semi-arid rather than arid regions, is that in arid areas there is insufficient vegetation on the lake edge to Uap the material which has been scourcd from the lake hed and to prevent later degradation, and the lakes in arid areas do not contain water lony cnough for significant wave transport* to occur," The third diffienlty concerns the distribution of the Junette in relation to the lake. The position of the Junette on the eastern or south-eastern side of the lake oppasite the wet season winds is a result of wind generated waves. A beach is built on the castern and south-eastern side of the lake, and the material in the beach is able to he transported by the wind. In the dry season, once the water level has fallen and with the winds blowing from the west and south-west, deflation from the lake hed takes place. In this way, « cambination of wave and wind action results in the formation of the lunette on the eastern side, and where waye action and wind transport in winter are relatively more important on the south-eastern side. ‘hus, the anulogy of hinettes with primary coastal foredunes is readily apparent. ACKNOWLEDGEMENTS The research. reported in this paper was made possible by a grant from the Uprversitv of Adelaide. A mineral analysis of samples from lunettes and their associated lake beds was carried ont by Australian Mincral Development Labora- 5 1n suh-humid vegians. the raintall is sufficient to allow the development of co-erdinaled drainage systeros rather than dry busing of internul druinage,.and the evaporition mite is lower. Therefore, deflition is limited, ‘In several districts in sootherm South Australia, for example near Snowtown, sume likes have associated lometles. whilst others do not. Such factors as the amonnt aod nature of material curried Inte the lake, and the length of time the lake is oceupied by water (enabling wave transport) or is dry (enabling deflation) must be considered in explaining this distribution. Ws ELIZABHTH M, CAMPBELL torics, Officers of the Bureau of Meteorology, Adelaide, kindly supplied informa- tion [rom which the wind roses were drawn. The thanks of the writer are due to Dr C, RK. Twidale, Department of Geography, University fo Adelaide, for helpful advice on many aspects of the research, and for a critical reading of the text, and to Mn. J. T. Hutton, of the Division of Soils, C,S.1.R,0,, Adelaide, and Mr. B. Mason of the Department of Geography, University of Adelaide. REFERENCES BacNoLy, K. A., 1941. The physics of blown sand and desert dimes. London, 263 p, Barpwin, J. G., Rurvine, GH, and Preepsray, J.B, 1999, A soil survey of part of ihe Kerang Irrigution District, Victoria.. C.S2.K.0. Bull. 125. Betrenay. EB. 1962. The salt lake systems and their assauiated acolian features iu the semi ariel regions of Western Australia. Jour. Sail Seience 13; 10-17. Bouratve, J., 1954. La sebkha de Bien Ziane eh sa “Wanette” ow boorrelet. Aen. de Geon, Dynamique 3: 102-123, 2 Bownen, J. M.. 1964. Environmental significances of lunettes, Arid Zone Newsletter, p. 29. Bow.en, J. M. and Hanronp, L. B., 1966, Quaternary tectonics and the evolution of the Riverine Plain near Echuca. Victoria. Jour. Geol, Soe, Aust. 13: 339-354. Caspers. £. M.. 1967. The morphology and eenesis of lumuttes in southern South Avstralia. M.A, Thesis, University of Adelaide, (Unpublished,) Currin, W. 5., 1945. Dynamics of wind erosion, Sui! Sct. 60: 30520, 397-411, 475-80. @1- (67-77. 237-63, 331-40, Ouerit,, W.S., 1958, Soil conditions that influence wind erosion. US. Dept. Agric, Tech. Bull. 1155. ae W. S., 1959. Wind eralibility of farm Relds, Jour, Sail ¢- Water Consers. 14; S14. 219, Curein, YW. S. anc Mixer, KR. A., 1939. Comparative study of soil drifting in the field and in a wind tunnel. Sci. Agr. 19: 249-57. : CGorrry, G. N., 1909. Clay dunes. Jour. Geol. 17: 754-5. Conpbon, KR. W,, and STAXNARR, M, F., (957, Erosion in western New South Wales, Bt U1, Forms of wind erosion. Jour, Soil Cons, N.S.W, 13; 17-26, Cooke, ©, W,. 1934, Discussion of the origin of the sanposed meteorite sears pf South Garolina Jour, Geol. 42: 88-96. Cooxr, C. W., 1940. Filliptieal bays in South Carolina and the shipe of eddies. Jour. Ceal. A8: 205-LL. Gauner, M., 19353. Les chotts. machines evaporiutoires complexes, pp. 317-325, in Cellogues Tilernalianaux du Centre National de la Recherche Setentifque, 35. Guana, C., 1945. A dsiological exphiuetion of the Carolina bays, Set Monthly, Gl; 443-50, CGnavson. H, J.. and Manony. D. J.. 1910. The geology of the Camperdown iui Mount Klephant districts. Geol. Sur. Viet. Memairs 9. pared J.. 1939. Physiography of the Echuca District. Proc, Roy. Sec. Viet, SL ONS): a! {). Tlmas, M. No 1987. Lake Towle gypsinn deposit. §. Aust. Dept. Mines Mining Ree, 108: O58 41. Ifinas, B.S. 1939. The physiography of Northwestern Victoria Lroe. Roy. Soe. Viet. 51: (N.5,); 297-323, . Hiuws, £, S.. 1940, The Janette. «mew landform of aeolian origin, Aust. Geoerapher’3; 15-21, Ifacsion, F. J. aud Brerreway, E., 1960, A laboratory exammatian af the soils of Uv Moerecin aret, Western Anstraliz, C-S1.B.Q. Div, Soils, Dio, Rep, 7/60, Hurvscan. G. Go tmd Prick. W. A, 1849. Clay dune formation near Corpas Christi, Texns. Jour, Sed. Petrel. 19; 118-127. Tack, i, ei 1921. The sult und gypsuin resources of Sonth Austealia. Geol, Suri. S$. Anst. Bull. 8. Jeanincs, J. N,. 1955, Le conplese de sebkhas. Rew. de Géom, Dynamique, G: 49-72. King, D., 19492, Cooke Plains gypsum deposit, $, Aust, Dept. Mines, Mining Hev. 91; 141-5, King, 1., 1949b, Lake Fowler gypsum deposit. S. Aust, Dept, Mines, Mining Rev, 92: G07. Kuumpein, W. C,, amd Sross, L, L., 1963, Stratigeaphy and Sedimentation, Quad vd. San Francisco 660 yp. Meirox, F, A. and Seamenvnn, W., 1939. The Carolina “Bays’—are they wnetecovitie scurs? Jour, Geol, 41; 52-tits, Pais, 5., 1964, Quaternary sedimentation hy prior streams orn the Riverine Plains, southwest of Onitfith, New South Wales, J. Proe, Hoy. sae. NoS.W. 97; LOT-15. LUNETTES IN SOUTHERN SOUTH AUSTRALIA 109 Pevs, S., 1966. Late Quaternary chronology of the Riverine Plain of southeastern Australia, Jour. Geol. Soc. Aust. 18: 27-40. Pewk, T. L., 1951. An observation on wind blown silt. Jour. Geol. 59: 399-401. Price, W. A., and Kornicxer, L. S., 1961, Marine and lagoonal deposits in clay dunes, Gulf Coast, Texas. Jour. Sed. Petrol. 31; 245-55, Prouty, W. F., 1952. Carolira bays and their origin. Geol. Soc. Amer, Bull. 63: 167-224. Rasz, E., 1934. Rounded lakes and lagoons of the Coastal Plains of Massachusetts. Jour, Geol. 42: 839-48. SrepHEns, C. G., and Crocker, R. L., 1946. Composition and genesis of lunettes., Trans. Roy. Soc, S, Aust. 70: 302-12. Tricart, J., 19542, Une forme de relief climatique: Les sebkhas. Rev. de Géom. Dynamique 5: 97-101. Tricart, J., 1954b, Influence des sols salés sur la déflation éolienne en Basse Mauritanie et dans le delta du Sénégal. Rev, de Géom. Dynamique 5: 124-132. Tricart, J., 1955, Aspects sédimentologiques du delta du Sénégal. Geol. Rundschau. 43: 384-97. Woops, J. E. Tentson, 1862. Geological observations in South Australia, London. 404 p. Wysercu, W., 1918. The limestone resources of the Union. Vol, 1. Geol. Surv. S. Afric. Memoir 11, p. 74 Piate J {LIZABETH M. CAMPBELL PLATE 1 Hitton’s Lagoon linette—aerial view. The smoothly rounded eastern shore is typical of the lakes in southern South Australia, The dashed line indicates the position of the crest of the lunette. (Reproduced by courtesy of South Australian Department of Lands. ) EvimaBetru M, CAMPBELL Phare 2 White Lagoon multiple lunette—ae view. White Lagoon is an example of the more or less concentric arrangement of a multiple lunette bordering ore lake. The three ridges are marked by dished lines. (Reproduced by courtesy of South Australian Department of Lands. ) PLATE 3 Exurzazetru M, CAMPBELL PLATE 3 Salt Lake multiple lnnette—aerial view. A complex of small lakes, each with its associated lunette or limettes. occurs within a larger lake basin on the edge of which is a large lunette. (a) (b) (c) (d) (Reproduced by courtesy of South Austraiian Department of Lands. ) PLATE 4 (See next page ) Western slope of Bool Lagoon lunette. The smooth and rectilinear western slope of this lunette is typical. Bedding in Cooke Plains lunette near the crest. The dip of the beds is variable over short distances. but nowhere exceeds 20) degrees. View looking north-east. Bedding in Lake Cadnite lunette. The characteristic dip of the beds in this lunette near the crest is 2 to 12 degrees to the east. The example shown here indicates a rare cross- bedded sequence. Shoreline of Lake Greenly, Gneissic material outcrops on the south-eastern shoreline of Lake Greenly. A shore platform and a beach with rounded pebbles are evidence of the effectiveness of waves in transporting debris even in this shallow lake. The white material near the top of the photograph is “Hour” gypsum on the surface of the lunette. (Photo- graph by C. R, Twidale. ) PLari 4 CAMPBELL M. ELizABETH ORIGIN OF WAVE ROCK, HYDEN, WESTERN AUSTRALIA BY C. R. TWIDALE* Summary Wave Rock is a long high overhanging natural wall located on the northern side of Hyden Rock, in Western Australia's Wheat Belt. Though it displays some apparently unusual features, it has much in common with similar forms described from Eyre Peninsula and other areas both within Australia and overseas. The available evidence suggests that Wave Rock also evolved in similar fashion: by strong scarp foot weathering, and subsequent erosion of the weathered debris. This and other hypotheses advanced in explanation of Wave Rock are discussed in light of the field evidence. ORIGIN OF WAVE ROCK, HYDEN, WESTERN AUSTRALIA C. R. Twipate* [Read October 10 1968] SUMMARY Wuve Ruel isu long high overhanging natural wall located on the northern side of Myden Rock, in Western Australia’s Wheat Belt. Though it displays some upparertly unusual features, it has much in common with similar fornis deseribed from Eyre Peninsula and other areas both within Australia aud bverseas. The available evidence suggests that Wave Rock also evolved in similar fashion: by strong scarp foot weathering, and subsequent erosion of the weathered debris, ‘Chis and other hypotheses advanced in explanation of Wave Rock are discussed in ight of the ficld evidence, INTRODUCTION Various opinions haye been offered concerning the origin of Wave Rock, a well-known and spectacular feature located near Hyden, W.A. In Land of the Southern Cress (1956), for example, a popular picture book designed to attract immigrants and tourists, it is attributed to the work of “wind and rain” (p, 89). In Vincent Serventy’s Landforms of Australia (1967, p, 25) it, and similar sharply sigmoidal or flared forms at Ayers Rock, are tentatively attributed to erosion hy wind-driven sand. An alternative explanation has heen offered by O'Riley (1967) who, citing professional opinion, suggests that Wave Rock is due fundamentally to selective erosion by running water. It is suggested that the upper, and earlier exposed, surface of Hyden Rock, on which Wave Rock occurs, has been indurated with silica to 4 greater extent than the more recently revealed flanks, Water coursing over the bare rock surface therefore crodes the weaker lower slopes more than the toughened upper surface; the lower slopes are “undercut by the contin- ued flow of water”; and the stone wall erected around the flattish upper surface of the rock is intended to “arrest erosion”. Comparable, though less spectacular, flared forms on Eyre Peninsula (S.A, ) and elsewhere (Twidale, 1962; 1964; 1965, pp. 145-147, 347-350) are evidently due to strong localised subsurface scarp foot weathering and subsequent erosion. Because this interpretation is at variance with the opinions cited above concerning the origin of similar forms, it may be both useful and interesting to describe the process of scarp foot weathering and its characteristic results; and to match the several hypotheses with the field evidence at Flyden Rock, SCARP FOOT WEATHERING AND EROSION Flared forms developed on granitic rocks have been examined on Eyre Penin- sula, in the Kosciusko region of southern N.S.W., in the western Mt. Lofty Ranges (S.A.), and in the Sierra Nevada of California. They occur on sandstone outcrops in the Flinders Ranges, and at Ayers Rock, as well as on the conglomerate of the © Department of Geography, University of Adelaide. Trans. Roy. Soc. 5.A. (1968), Vol. 92. L1G ©, Ro TWIDALE Olgas massil, Norther Territory, Although Hares are well developed on the north tn west facing slopes, and although exceptions have been noted, they are best and most commonly developed on the southern and eastern, ie. the shady, aspects af the residual hills or inselbergs on whicl they occur. Double and multiple Hares nr Convavities are in places displaved. The flarcd slopes are not everywhere found in horizontally disposed zones of steepening or overbang. At several sites they are inclined or even irregularly distributed, though in all cases the disposition of the steepened zone varies in evident sympathy with that of the hill-plain or rock-soil, junction. Flares are very well developed on the points of spurs. Incipient flared slopes are developed beneath the land surface, beneath weathered country rock which is in situ. At the margin of the inselberg, the suv- face of the fresh rock, continued subsurface as the weathering front (Mabbutt, 1561) or limit af weathering, characteristically phmges steeply for a few fret heneath the land surface before gradually fattening out. Snch concavities in the fresh bedrock surface, and formed beneath the natural land surface, have: been observed in dry reservoirs located on the inselberg margins, and have also heen detected by augering und excavation, Finally, small flared slopes oceur in clefts within the inselhergs, around boulders and on gentle slopes, in association with accumulations of soil or weath- erect debris, Although several possible explanations of flared slopes have been entertained (Twidale, loc. vit. 1962; op. cit. pp. 347-350), all save one are found wanting when the field evidence is idtched against the deducible consequences of the hypotheses. The interpretation which has survived rigorons testing and which acuournts for the field evidence is that the flares evolve in two distinct stages involving first, strony lovalised scarp foot weathering, and, secon], marked) difter ential erosion of the scarp foot zone and exposure of the erstwhile weathering front (Fig. 1). HYDEN ROCK Ceneral Setting Hyden Rock lies about 3 km, east of Hyden township, which is, in turn, 296 km, E.S-E. of Perth (Fig. 2). The Rock is about 2 km. long on an east-west axis and on a N.N.W-S.S.E. line a litde more than 0:8 km. wide (see Fig. 3), It is composed of granite, most of which is strongly porphyritic, though pegmatilic, even and medium grained, aid fine gramed, phases occur in patches and veins, No horizontal dat-lying zones of petrological differentiation have been detected in the granite. The granite is jointed but most of the joints are tight and widely spaced. Avenues of weathering within the Rock are few, and its greater resistance to weitherng aud crosion, as evidenced by ils remaining as an upstanding hill, is probably due to its being built of especially massive joint blocks, This is demrons- trably the case in places on Eyre Peninsula and elsewhere, Hyden Rock is a granitic inselherg which rises abruptly to a maximum eleva- tion of some 55 m, above the surrounding plains, whieh, on the evidence of lores wd dam excavations, are also underlain by granite, Tere, however, the rock is deeply weathered: Berliat (1965) reports over 20 m, of weuthered grunite wbuut 400m), from the margin of a residual located 63 km. east of Hyden and well over AQ nm. of disintegrated rock elsewhere beneath the plains. The plains are not Aat, for there is a pronounced slope down to the north; the lew vaguely defined watercourses How toward a complex of ephemeral Jakes ORIGIN OF WAVE ROCK, ITYDEN 7 | | F pr --2=-"F71', basistaet during dry season ! F Fig. 1. Evolution of flared slope by scarp foot weathering and sub- sequent erosion, (a) Strong weathering by water (derived from run-off) at the lower margin of the residual hill. Numbered lines represent stages in advanue of weathering front. (b) Lowering of base level, erosion of weathered debris, and exposure of weathering front as a flared slope, with further weathering beneath the new plain surface, (c) Develop- ment of double flare by repetition of the twa-stage process. c , -}- located immediately to the north of the Rock. The plain to the south of the Rock is about 25 m. higher than the plain on its northern side, The inselberg consists of three distinct hills, the central dome-shaped hill being both slightly higher and more extensive than its neighbours to cast and west, All three hills are delineated by prominent vertical or near vertical joints, the more important of which run E.N.E.-W.S.W., N.W.-S.W., and N.W.-S.E. Many clefts within the hills follow along such joints. Flat lying or gently dipping joints are also important, for, although discontinuous, they subdivide the hill into mas- sive slabs or sheets. The weathering and erosion of several of the latter gives the Rock its stepped appearance in broad view ( Fig. 4). Flared Slopes Flared and basally steepened slopes are well developed on all three hills, not only around the margins but in clefts (Pl. 1) and amphitheatres at higher levels also. The flared and steepened slopes are well formed on the long southern wall of the Rock, for instance, adjacent to the 13th fairway of Hyden Golf Club (A on Fig. 3) and they are well developed along much of the northern foot wall. In clefts and depressions the south-facing wall is commonly, though not in all cases, steeper than that opposite, 118 Cc. R. TWIDALE PERTH @Hyden Sr & ——______ limit of granite outcrop major joint watercourse, loke ming kts road 0 Tmile —— 0 Vim. Fig. 2 and Fig. 3, Location map and map of Hyden Rock (drawn from W.A. Lands Dept. air photographs). Sites A-D are referred to in the text. N. sheet >, S. boulders formed by 25 disintegration of sheet ee SS gm. \ incipient flare —-"S, 3 Bet 25 Fig. 4. Diagrammatic N.-S. section, not to scale, of Hyden Rock, showing stepped form, sheet structure, flared slopes, probable present weathering fronts, and the disparity in elevation between the plain north and south of the Rock. ORIGIN OF WAVE ROCK, HYDEN 119 But what is unquestionably the most majestic and impressive Aared and over- hanging slope is Wave Nock (PI. IT) which is 4 steepened and overhanging hasul slope 10-12 m. high located on the north side of the central hill, Some fared slopes on Pildappa Hill, Eyre Peninsula, are as high or higher than Wave Rock, some at Ucontitchie Hill (also on Eyre Peninsula) overhang toa greater extent or are more complex, but as a high continuous overhanging wall, Wave Rock indubi- tably stands alone. At the base of the flared slopes on the northern side is a continuous, virtually level platform up ta 10 m. wide (Pl. IL) and located at, or a litle above, the present plain level. Similar platforms have been observed marginal to the insel- bergs of northwestern Eyre Peninsula, hut nowhere are they as extensive and level as at Hyden. Another notable feature is the angular junction between flared slope and platform. Double concavities or flares are fairly common, as. for instance, on Wave Rock itself (PJ. IL), Some flared zones on Hyden Rock are inclined with respect to the horizontal, as, for mstance, on the northern slope (PI. IIT), in the enclave hetween the central and eastern hills. Finally, it is very typical that the weathering front—the surface between the fresh and the weathered bedrock— plunges steeply beneath the weathered granite which, with a veneer of colluvial debris, underlies the plain (Pi. IIL). Excavations reveal that this steeply inclined slope continues beneath the plain surface for rather more than one metre, a which depth it shows ne sign of levelling out. Thus, in many respects the flared slopes at Hyden Rock, and particularly Wave Rock, share characteristics with similar forms in other regions. The great contrast lies in Wave Rock itself, which is on the northern, and not the southern side of the inselberg, and which is located not on the point of a spur, but in a broad, prohably joint-controlled, crbayment (B on Fig. 3). These apparent anomalies are, however, susceptible of ready explanation in terms of the subsurface water weathering hypothesis. Being in a broad cmbayment, to which drains nim-off from a wide area of the central hill, the plains marginal to the inselberg in that area undouhtedly reocive a more than average quantity of water, so that especially pronounced and deep scarp foot weathering is probably developed there. Furthermore, inmedi- ately to the north of the Rock and virtually at the same clevation as the plain on the northern side is a complex depression—probably an old drainage line—which receives water from a wide area and which in winter is full of water. Ground- waters from this drainage line may penetrate to the Rock itself. The comparative degree of erosional exposure on the northern and sonthern aspects of Hyden Rack should also be considered. As already mentioned, the northem plain, clase to the drainage depression, is about 25 m. lower than the plain to the south of the Rock (Fig. 4): indeed, the whole plain surrounding the inselberg slopes down to the north, Any lowering of the plain by strcams must have been initiated from the old drainage line ‘ Fig. 3) aid. therefore would have affected the northern slope of Hyden Rock before the southern. Thus, it may be argued that while greater erosion and lowering of plain level has already exposed the deep weathering front on the northern face, similar lowering of the surfaer has not yet extended to the southern side. where an incipient flared slope may be present beneath the surface (Fig. 4). Thus, local circumstances may have combined to cause unusually deep and pronounced subsurface weathering of the granite on the northern side of Hyden Rovk. The same northern slope may have been exposed by erosion to a greater extent than has the southern. These are the probable reasons for the wnhusual degree of steepening of the northern slopes, and for Wave Rock in particular. 120 CB. TWIDALE The other suggested mechanisms tor the development of Waye Rock either cannot explain the field evidence, or are inconsistent with it. Por instance, aeolian sand blasting cannot account for the occurrence of incipient Hures beneath the sresont land surface, below weathered granite in site, and for the fares at high Fevels an the Rock (sand blasting is most effective within 1 m. or so of ground level). The areal distribution of the fares is also at odds with the éxpectable consequences ot sand blasting. Luncttes duc to deflation of exposed lake deposits are located on the western side of lakes in the Hyden-Corrigin area, indicating a prevalence of strong winds from the east, and in these terms the “undercut” slopes slintilel be best developed on the eastern aspects of the hills, No such consistent and preferred orientation has been detected. Furthermore, the very fact of the inselberys being there would disturb the air flow and divert winds (and sand} around the residuals; the abundant vegetation growing aronnil the hills ar the better watered scarp foot zone also serves to protect the lower rock walls against avalian action. The suggestion concerned with running water (O'Riley, loe. cit.) faces oom- parable difficulties: (1) Induration of the snrtace of Hyden Rovk has certainly acewrred, partly as a result of climatic conditions, partly resulting from the work of the lichens which grew in profusion on the rock surface. Iron oxides have been concentrated! at and newr the surface, though much of this indurated layer appears to have heen eroded, only small Hat-topped nobs remaining in a few places. But the induration appears equally distributed over most of the surface, It is certainly aol preferen- tially developed at upper levels of the Rock surface, so that selective crosion cannot be invoked. urthermore, though the upper area must have been first exposed, there is a strong suggestion that far from being tougher than the lower slopes, the upper part of the rock has suffered cousiderable weathering, the upper sheets of granite having disintegrated iuto groups of blocks or ronnded boulders. (2) Some form of induration has undoubtedly affected many boulders, and preferential weathering has given rise to cayernous forms or tortoiseshell rocks. But in these forms the hardened crust forms thin projecting lips or visors, which da nat oceur in the flared slopes. At one site on the northern slope of eastern hill (C on Fig, 3) there is a concentration of cavernous and alveolar weathering which may be due to petrological factors, or may be a reflection of local concen- trated weathering at or just below a former plain level (the site is some 6 m. above the present plain). In this area, quite deep hollows are developed beneath thin external crusts, (3) Although rimning water, hecanse of its downslope increase in velocity and volume, may appear capable of producing slopes of irereasing downslope inclingtion (see for instance Fenneman, 1908; Cotton, 1952), particularly in an embayment such as those at Wave Rock and at D (Fig, 3) on the southern side of Tlyden Rock, where flow is centripetal, there are sound hydrological reasons for suggesting that it dacs not do so (Twidale, loc, cit., 1962). Moreover, ounce a cortain critical inclination is atlained, fast Howing water becomes detached fram the rock surface to form a jct or waterfall. Only the slowest flowing thin films of water adhere to the steep rock face by surface tension—vyet these are supposed to aucomplish considerable erosion, and even pronounced undercutting. At Wave Rock in June 1968, some thin films of water were creeping down the overhanging slope, but there was a continuous drip of water from the upper lip to the foot wall. Also, on the upper part of Ilydun Rock, pronounced overhanging Mares are develoned on spurs of very limited catchment, and.in other places on boulders of similarly limited potential run-off, ORIGIN OF WAVE ROCK, HYDEN 121 (4) Selective erosion by running water cannot explain multiple flares, the development of fares in clefts and amphitheatres (nor, indeed, the amphithcatres themselves), the inclination of the flared zone, and their apparent occurrence suly- surface marginal to the Hyden residual. The searp foot weathering hypothesis can readily account for these and other relevant details of the flared forms, and it is suggested that it is this process which is responsible for the development of Waye Rock and other similar features at Tyden Rock, as well as on other granite residuals in the southwest of Western Australia. Angular Footwall The pronounced angularity of the junction between the fared slope and the basal platform, us displayed at Wave Rock, and other similar features at Elyden Rock and other inselbergs in the region, develops as a result of subaerial weather- ing, Ramwater and seepage waters trickle down the steepened lower slopes and cause the platform immediately adjacent to the hillslope to be wetted. The rnck here disintegrates, and is subsequently washed out, leaving a miniature depres- sium, one flake (1-2 ems) deep between the main platform level and the base of the fared slope, Such depressions, observed at the base of Wave Rock and on Gorge Rock, near Corrigin, W.A., carry water after rain, causing further weather- ing of the granite with which it comes into contact, inclnding that exposed at the searp foot, As the shallow depression is extended laterally by such weathering, the basal slope is undermined and steepened, and an angular junction is deyeluped between hillslope and platform (Fig. 3). Amphitheatres Another feature well displayed at Hyden Rock, but not so far described from the Eastern States. is the amphitheatre, Joint clefts with flared hounding slopes occur on Hyden Rock (Pl, T) and, for instance, on several of the inselhergs of northwestern Eyre Peninsula. These two areas have similar clirmates (both with hot summers, coo! winters, 35-45 cm. average annual rainfall and a marked dry season), but at Hyden Rock the slopes of the clefts have suffered marked reces- sion as a result of soil moisture weathering and as a result have been widened to form large amphitheatres (Fig. 6). It is not known whether this more advanced stage of development reflects a time factor or the less resistant character of the porphyritic granite exposed at Wyden Rock, but such developments clearly contri- brite ta ihe breakdown of the sheets of granite and to the formation of an overall stepped morphology on the inselbergs. CONCLUSION Though Wave Rock displays apparently unusual features, it hag much in common with similar forms described from Eyre Peninsula and elsewhere. The available evidence suggests that Wave Rock evolved, as did the other flared slopes, by strong scarp foot weathering followed by erosional exposure. This interpreta- tion accounts for the field evidence more readily than any of the other hypotheses so far advanced. ACKNOWLEDCEMENTS The writer wishes to thank Professor M_]. Webb, Department of Geogniphy, University of Western Australia, and the Director, Geological Survey of Western Australia, for facilitating in vations ways the investigation of Ilyden Rock and several other inselbergs in the region. Some travel expenses were met from a Research Grant from the University of Adelaide, _ w to C. R. TWIDALE Fig. 5. Development of an angular junction between flared slope and platform by local subaerial water and pool weathering. (a) Water trickling down flared slope soaks the bedrock at the inner edge of the basal platforin and. weathers it. (tb) The weathered debris is washed away, a shallow pool of water accumu- lates in the hollow so formed, and further weathering takes place. (e) The base of the shallow pool flare is steepened and made angular as a result \ of weathering by pool water, and consequent b - 7 lateral extension of the pool depression. grenite debris \ *_weathered granite Fig. 6. Sequential development of an amphitheatre from a joint-controlled cleft by soil mois- ture, weathering, undermining of bounding slopes, and progressive widening. Several such phases of weathering, with intervening spells of debris evacuation, are evidently represented in these clefts and depressions with complex sides. ‘The phases of weathering and erosion in the clefts, like those to which the flared stoned are attributed, are probably controlled by climatic changes or fluctuations or even seasonal and storm effects. a-d represents a sequence, represen- tatives of which have been observed on Hyden Rock. d1 shows the result of baselevel lowering in a narrow cleft (cf. Pl, 1). The depth of fill must decrease as the area of the depression floor increases, ORIGIN OF WAVE ROCK, HYDEN 123 REFERENCES Anon., 1956 Land of the Southern Cross. Australian Publicity Council, Melbourne, Australia; BERLIAT, Xe, 1965 Exploratory drilling for water in the sand-plain country 40 miles east of Hyden. pp. 8-9 in Annual Report (1964) Geol. Surv. W. Austr., Perth, 72 p. Corton, C, A., 1952. The erosional grading of convex and concave slopes. Geog. Jour., 118, pp. 197-204. FENNEMAN, N. F., 1908. Some features of erosion by unconcentrated wash. Jour. Geol., 16, pp. 746-754. Massutt, J. A., 1961. “Basal surface” or “weathering front”. Proc. Geol. Assoc. London, 72, pp. 857-358. O’Riey, [., 1967. Wave that turned to stone. Walkabout, 33, (6), pp. 12-15. ServeNTY, V., 1967. Landforms of Australia. Angus & Robertson, Sydney, 132 p. Twipace, C. R., 1962. Steepened margins of inselbergs from north-western Eyre Peninsula, South Australia. Zeitschr. f. Geomorph. (N.S.), 6, pp. 51-69. Twimate, C. R., 1964. Contribution to the general theory of domed inselbergs. Conclusions derived from observations in South Australia, Trans. and Papers Inst. Brit. Geogr., 34, pp. 91-1138. Twwateg, C. R., 1968. Geomorphology, with special reference to Australia. Nelson, Melbourne, A406 p. . BR. Twipate 1 4 PLates 1 to 3 (CSTEPIAL “WD Ofoyd) — “yloursnc ute yoo SAUAA JO ysPA "UN QOS qwos YyooYy uepAFE JO 9p!s you ayy uo sumypd quaorlpe Tyvauaq judy AuLatpTAM jo asunjd pure ‘umopyepd yoo, dimos ‘auoz poasy surdoys (ApEn WD ood) ‘sayddns raya sy sawp uvapapy toryar wory HOALaSal af} OYUL PO-uNe FAALTP OF papuazue sl OAbA\ BY PAOYT [[RA MOP at, “MOET eseq ptoaq ayy Fupwoys UALUARQU at] JO pula Usayska aly Wor, YoY aarA, (OPBPIV,L YO God) “oY Uopayp jo Ayttttoayea UsaIsva AY} MAU Yapo payporMo.-julol Bur sadojs pamwpy THT 98d uU ‘I Vd Vy at OBITUARY: THOMAS DRAPER CAMPBELL 1894-1967 Summary OBITUARY THoMAs Drarer Cantppen. (1894-1967) Born at Millicent in the South-east of South Australia Campbell received his early education at Prince Alfred College and undertook tertiary study ut the University of Adelaide where he graduated in dental surgery in 192! continuing these studies ta receive a doctorate in 1923, His doctoral thesis was published by the University in 1925 under the title “Dentition ae rain of the Australian Aboriginal” and has been widely acclaimed as a pioneer work in this field. Soou after graduation Campbell entered the Dental Department of the Royal Adelaide Nospital as the first Dental House Surgeon, and in 1926 was appointed Superintendent. He wus elected Dean of the Faculty of Dentistry in 1938, and appointed to the full-time University position of Director of Dental Studies in 1949. He became the first Professor of Dentistry in 1954 antl upon retirenent four years later was made Professor Emeritus, During his professional career Prof, Campbell received high recaguition for his work, In 1948 he was made F,D.S.B,C.S, (London); in 1959, F.D.S.R.C.S. (Edinburgh): iv 1952 he wus. clected an Honorary Member of the Qdontological Section of the Royal Society af Medicine, London, and in 1966 « Fellow af the Australian College af Dental Sargeurs, At all limes Professor Canipbell combined his dental studies with a keen interest in anthropology. In December 1926 he was associated with the formation of the Board for Anthropologic:l Research, University of Adelaide, and has played a leading role in its research activities. As organiser and leader of many expeditions ta Central Australia in the 1920’s and 30's he was responsible for the collection of a large amount of important data and material on many aspects of the traditional life of the aboriginal, Storie technology fascinated him and he spent countless hours searching for and making a detailed study of aboriginal stone implements. He contributed important papers to, the literature on this subject. Well known. as an aecom- plished amateur film producer, Professor Canmphell made a series af colour-sound movics on ge bf aboriginal technology for the Board. He was chairman of the Board at the time of tis death; He became a member of this Society in 1922 and spent ten years on the Council, holding the office of President in 1934-35. Following appointment to the Board of the Public Library. Musenm and Art Gallery in 1932, he became a foundation member of the Museu Board in 1939, a position which he held until his death. In addition to being a Fellow of the Australian Research Council, Foundation Member of the Australian Institute of Aboriginal Studies and a Lite Member of the Anthropological Socicty of South Australia, Professor Campbell was an panty member of the Anthropology Staff of the South Australian Musewn from 1923 until us death, Te was well known as a keen debater and will be remembered for lis participation in many lively discussions at Royal. Society meetings. A member of this Society with an out- standing international reputation, he died at his home at Tusmare on 8th December, 1967, Robert Edwards, BALANCE SHEETS Summary 126 NN 66089" TS S9GI ‘auUN{ WINE aoumpg_ Sutsopp 99 SLIT OSET salupung 99°9 agoucinsuy O0O°OLT surpurg GL'¥G surunig SL LE6 saat My 00°06 asvysogq squauhing jonpaqd $ $ SLNAWAYd £9°e08's ors6e's = —— 1c'91 3 * a ” ” a yorajuy yue_g 6TS.TEe i 7 suorjousuery, JO ales a P "208% ‘AR[ JST aounjppg suyuadg SLdI HORN we eee ee SS SS eee SS s96t “ANAL 108 GHGNA YVAA AHL OA SLNAWAVd GNV SIdIGOSY AO LNAWALVLS INNOSOY AUVUArT NN —— creies S96T ‘UML Og aaun—ng BursozD GL°FS0'S GFFSL soupung 69'SoT aounimsuy FLSP ALQUOES GF 601 sasieyo) aoraras ole edayes yseg Aneag 88-961 5 aseysog 68181 sjsor) Bururayg SLCIl sysor) Burysipqng squawmAng jonpag $ $ SINGAWAVd FS OLS'S BC C9T Ss ST'CSL ata Saatays FO gus) auiooul Arpung Leo9 (qsau9yU] puogl puw yurg oOo 00L"e _yarrg yusuraiA0s) 9B FEST } saondrosqngs 96° LOL § LOGT “sane Sl aounjpg Buruado $ S Ld SOA em S961 “ANAL 08 GAANA UVAA AHL YOU SLNAWAVd CANV SidIdOed JO LNAWALYLS ENQODOY ‘IVWENG) CALVHOdHOONI VITVULSNVY HINOS JO ALYIOOS TVAOWU AHL et HUYUNODY parsuryqg 3 “OO AOUVAS ‘SNAARLS ‘ANTUN "S96T “rquiaydag jo Wg ‘eYeNY yng ‘aprepapy “89GT ‘unf qIOg I syuaUySsAUT JO s[NpayosS ay} poynea osye FAcy OA, “poytugns AJaWog ayy Jo syoog ayy Aq HAMOYS Sk puL sn Oo) Woah suoyeuEldsa oyy pur conPugozur mma jo 3894 9Y} 0} Aurproooe gogl ‘AaUn[ WOE papua Eat oy} 30j AyaIog ay} Jo suo ousuTsy yseo aq} pinsar op dh uanip Apadoid ase pung WOUAOpU PUP Arerqry ‘Tessuag ayy Joy syuawAeg pus sydiacay JO SUIUSTEIS paypeye ayy ‘uorurda ino ur “paynbes vary aa suojeueldxa pue ueneuiezut ay} qe paulejqo aary puy ggg ‘eunf moog POpUS ans OH? 30} (AALVYORYOONT) VWIIVELSAVY HLAOS JO ALBLIOS TVAOU AHL Jo Syunosoy pue syoog’ ayy pouruers savy aay yeqy Hoda. aA, Iuoday suouday eee OOLFE FITS. B96T “sun WOE yos Aynby jo ane yeyeyY sazoyy SUEES ESS 8O°Lg¢ i " u ; f “yor ye ysey 00°O0E ‘ 7 ° " 7 2POIS pequosuy LL°SPL1T . es'Ist‘h £8S—'spigQ-—ssurpjoyy Pooayyjos 00°'UU8 ; j * UOF—Spalg—saulpjoH Pooayjos e6BSU'L is i 1H 7 : _ OOL—qMoasusT O0G°E90°T t " ~ 7 ooS— TIT 0O'OSET “ , . . OSe—stooy, OO°L9E°T 7% GQOE—SouTY, 4ppaaAy pue pyerezy OO'ESS 1 hy - ‘ 00F—WS'*D OOELS TT : “ P0S—WataaD spreapy OOLETT | . = 0OF—TO'V OOOUEST ~ : we ™ ~ O06—SYHompoo yoos Aunby 00°006°0T 7 00°000°S d To%eL—SUONVAA ONT 00°00. 3 : ! e i ~ 28—"T a oo'000'S ' "f i > bYL—W Oa 00';000'S at : 4 LO WWD oO'O00'F . ; %9—eaoueury Yyvasuay sqysodaq] party $ $ (809)38) S961 ‘ANAL Woe LY SINANLSAANI JO ATINGTHOS Bo Less WY i. 89GT “Bunt inVE aounjog Sursoj+ ST'199'9 LL°€L0°0 ‘i : , . | : 00°000°S : ~4 ' ‘SQ9T SUNTAN “PTO'N'A 000°C sa9cg'S [ora , 00°€90'T a , ~ . M SPIO ‘TOT gor erPRoe : Y Sa1ey§ “d'H'dA O0E Je a"g GA'SSO'T “ ‘ “IPIQ) aOUFNTT YHomsuayT OO, 00°000°1 " a ‘ saamquaqaq “Oy wondurapay 00'D08 fi ‘a ii ‘spaig SAUIpjO]Y poomyos OF fOLS * Ff ‘i 7 ° ir ysarszUT yuUgy SeIsvT " r ‘SPAQ SBUIP[OHT pooayyog Ego Ea’ eer't vy 7 Peajwoay SpuaprArq: put qsazazuy —8}au}saAuT JO aseymg sidiaooy ppy $ $ LU¥ES ub ‘ ; LOGIT “AlN 3ST aounjng Fuivado spuaufog jonpad : s § SLNAWAYE , SLdIYOae cx. $$ eee 8961 “ANN! Woe GHANA UVAA AHL WO SLNAWAVA ANY SIdISO0uu AO LNAWSALYLS GNO J HOuwyasay OLULLNALOG GNy INGNAOGNG CaALVYOdUOONI VIIVULSAVY HLOAOS AO ALAIOOS ‘IVAOY AHL REPORT ON ACTIVITIES OF THE COUNCIL, 1967-68 Summary REPORT ON ACTIVITIES OF THE COUNCIL, 1967-68 Meetings The usual eight ordinary meetings were held in the Society's rooms during the past year at which the attendance averaged 32, One special meeting was held on December 7th to ratify the new Rules and By-laws, A total of 12 papers were read devoted to the following disciplines:— Zoology 8, Botany 3, Geology 1, Lectures were given at each meeting and two exhibits presented, Membership Twenty-two new members were elected during the year and nine resigna- tions were received. The membership now stands at 266, the highest number in the history of the Society. The Council records with great regret the death of Fellows; Emeritus Pro- fessor T. D. Campbell, a past president of the Society; Sir Tom Bart Smith and Mr, J. K, Powrie. Rules and By-laws During the year the new set of Rules and By-laws approved at the December 1967 meetings Was issued to all members. One amendment has since been passed at the meeting of September 12th, 1968. Rule VI (2) now reads;— The nomination furm shall be lodged with the secretary and shall be sub- mitted to the Council at its next meeting, Upon approval by the Council the nomination shall be submitted to the next meeting of the Society and an election shall be held at the next meeting of the Society thereafter, Sub-Committees of Council The business of Council has been facilitated by the work of four sub- committees ;— Library Committee—management of the Library Publications Committee—production and maintenance of standards in the Transactions and any other publications of the Society Awards Committee—nominations to the Council of names of members pro- posed for awards by the Society and other bodies Research and Endowment Fund Committee—manayement of investment funds. Library The library has operated efficiently and profitably during the year, the Library Account having a credit balance of $1630 as at June 30th 1968. The Council wishes to express their grateful thanks to the Assistants, Mrs. Dunlop and Mrs. Dougal for the services they have rendered to the Society both in the library and in assisting the Treasurer. Rechecking of the bookstock has resulted in 200 new entries being forwarded to the C.S.1.R.0, Index, Fifty volumes have been bound and another 160 prepared for binding. The exchange list has been completely reviewed and adjustments made 129 where necessary. Some 340 journals are now received on exchange from 46 countries, 10 new exchanges were negotiated during the year. In addition 37 continuing subscriptions are now in operation, Altogether 245 volumes were borrowed during the year, mainly on inter- library loan within the state, interstate and with New Zealand. Author and subject indexes for volumes 45 to 91 have been prepared for publication. Publications Vol. 91 of the Society's Transactions was published in December 1967, It contained 204 pages compared with 191 pages in Vol. 90. The cost of printing seems to rise steadily each year but it has been possible so far to include all papers received before the June Council meeting which is normally taken as the final date for acceptance of material for the current volume. Research and Endowment Fund As a result of moves which were initiated two years ago, to improve the management of the fund, its finances are now on a reasonably sound basis. Accordingly this year it was found possible to make available some of the income, in the form of grants, to aid scientific research. Two grants for the calendar year 1968 were made, Dr, C, R. Twidale was awarded a grant to study the geomorphology of the Arcoona Plateau, and Mr. I. M. Thomas and Mr. S. Shepherd on behalf of a group of workers will receive a sum of money to finance an expedition to Pearson Islands to study the flora and fauna of the islands and adjacent waters. Verco Medal The Council awarded the Verco Medal for the year 1968 to Mr. R. C. Sprigg. OFFICERS FOR 1967-68 Summary 130 ROYAL SOCIETY OF SOUTH AUSTRALIA INCORPORATED Patron: IUS EXCELLENCY LIEUTENANT GENERAL SIR EDRIC M. BASTYAN, K.C.M.G., K.B.E., C.B. COUNCIL FOR 1967-68 President: K. R. MILES, D.Sc., F.G.S. Vice-Presidents: H. B. S. WOMERSLEY, D.Sc. F. J. MITCHELL Treasurer: Secretary: S, A. SEIEPHERD, B.A., LL.B. W. K. HARRIS, B.Sc. Editor: Assistant Editor: 1. M. THOMAS, M.Sc., M.1.Biol. J. K. TAYLOR, B.A., M.Sc., B.Se.Agr, Programme Secretary: Librarian: Db. E, SYMON, B.Ag.Se. N. H. LUDROOK, M.A., Ph.D., D.LC., F.G.S, Members of Council: J. A, PRESCOTT, C.B.E., D.Se., M, J. TYLER F.RS., FAA, FRAC. J. T. HUTTON, B.Sc., A.S.A.S.M. K. E. LEE, D.Sc. C. B. WELL, M.Ag.Se. H. E. WOPFNER, Ph.D. Auditors: Messrs. MILNE, STEVENS, SEARCY & CO. LIST OF MEMBERS Summary 131 ROYAL SOCIETY OF SOUTH AUSTRALIA LIST OF NEW MEMBERS Apam, J. R., B.E., A.M.Aust..M.M., Dept. of Mines, Box 38 P.O., Rundle Street, Adelaide, S.A, Ancove, P. C., R.D.A., M.A.LA.S., Dept. of Agriculture, Gawler Place, Adelaide, S.A. von Bexrens, D., Naracoorte High School, Naracoorte, S.A. Bursserr, A. H., M.Sc., A.M.AustJ.M.M., F.G,§., Dept. of Mines, Box 88 P.O,, Rundle Strect, Adelaide, $.A, Bonsai Miss S., BA. (Mus.), Dip. S.T., Bedford Park Teachers College, Bedford Park, S.A, Cannick, R., Ph.D., Mawson Inst. for Antarctic Research, University of Adelaide. Davy, R., B.Sc. (Hons.) Ph.D., L.RB.LC., F.G.S., C/o AM.D.EL., Conyngham Street, Frewville, S.A. Dyxr, N, W.. A.C.A.A., Mem. Inst. Inst. & ‘Coni.. Aust,, Seaton, S.A. Ernst, L. K., 50 Airdrie Avenue, Findon, S.A, Famsurn, W., B.Sc. (Hons.) A.M.I.M.M., F.G.S., Dept. of Mines, Box 88 P.O., Rundle Street, Adelaide, S.A. TlENnnincsen, M.. B.E., F.S.A.S.M., A.M.LE., Aust,, A.M.Aust.LM,M., Dept. of Mines, Box 38 P.O., Rundle Street, Adelaide, S.A, Hiren, M. N., B,Sc., Dept. of Mines, Box 38 P.O., Rundle Street, Adelaide, S.A. Huronson, A, R., 218 Shepherds Hill Road, Bellevue Heights, S.A, Jenxins, R. J. F., B.Sc. (Hons.), Dept. of Geology, University of Adelaide. Kennepy, Miss G. R., B.Sc., Ph.D., Flinders University, Sturt Road, Bedford Park, S.A. McGowan, B., B.Sc., (Hons.), Ph.D., Dept. of Mines, Box 38 P.O., Rundle Street, Adelaide, S.A. Mutter, P. G., B.Sc., (Hons,), A.M.Aust.LM.M., Dept, of Mines, Box 38 P.O., Rundle Street, Adelaide, S.A. Mooxcrort, E,, B.Sc., Dept. of Mines, Box 38 P.O., Rundle Street, Adelaide, S.A. Moutns, M, S., F.R,E.S,, 14 Chisholm Street, Greenwich, N.S.W. 2065. Otsen, A. M., M.Sc,, Dept. of Fisheries and Fauna Conservation, Adelaide, S.A, Wu.son, R. B., B.Sc., (Hons.), 22 Chetwynd Street, West Beach, S.A. Wo aston, Miss E, M., Ph.D., Botany Dept., University of Adelaide. LIST OF LECTURES AND EXHIBITS, 1967-68 AND AWARDS OF THE SIR JOSEPH VERCO MEDAL Summary LIST OF LECTURES GIVEN AT MEETINGS DURING THE YEAR 1967-68 Sept., 1967 Dr, H. B.S. Womenstry: “Aspects of Coral Reef Biology”. Oct., 1967 Mr. C, A. Martin: “Moomba, a South Australian Gas Field”, Noy., 1967 Mr. T. R. N, Lorman; “The work of the National Parks Commission of South Australia’. April, 1968 Mr. D. BR, Cunme: “The Industrial future of South Australia”. May, 1968 Prof. C. M. Donacn: “Rural development in the Northern Territory”, June, 1968 Prof. R. W. R. Ruriann: “A structural view of Continental Drilt”. July, 1968 Mr. R. C, Spiuce: “The Role uf Research in Industry”. Aug., 1968 Prot |. BR. M. Ravox; “Waves and tides of the South Australian oast”. EXHiBITs Dr. B. G. Forses: Some recent maps issued by the Geological Survey of South Australia. Mr. M. J. Tyuen: Relationships between musculature and vocal sac structure m frogs. AWARD OF THE SIR JOSEPH VERCO MEDAL 1929 Prov, WALTER Howain, F.G.S, 1980. Jonw Mc. Brack, A-L.S. (931 Pro, Sm Doucias Mawson, O.B.E., D.Sc., B.E., F.R.S. 1933 Pror, J. Burros Cre.anp, M.D, 1935. Pror, T, Hanvey Jownston, M.A., D.Se. 19388 Pror. J. A, Prescott, D.Sc., F.A.C.. 1943 Hinnenr Womemnsey, AIS. FR,E.S. 1944 Prov, J, G. Woon, D,Sc., Ph.D. 1945 Craw T, Mamaan, M.A. BE. D.Se. F.GS. 19486 Herienr M. Hare, O.BL. 1953 L. Krrma Wann, LS,0., B.A,, B.E.. D Se. 19356 N. B. Tsnvaze, B.Sc. 1957 ~-C, S. Prrer, D.Se. 1959 GC. G. SterHens, D.Sc. 1960 H. IL. FirnAysox 1961 KR. L. Spretrr, Ph,D- 1962 IT, G. ANpnewanrHa, M.Ag.Sa.. D.de., TAA, 1963 N. H. Lupproox, M.A., Ph.D. D.G., F.G.S. 1963 RB. ¥, Sourneorr, D.Se., M.D,, B-S.. D.T.M. & I. 1966 Pror. A. R. ALpeRMaw, D.Sc., Ph.D., BGS, 1967 IL, D. Pryor, M.Sce., Dip.For. 1968 BR. C. Spnicc, M.Sc. For stimulating and versatile contributions to the geology of South Austrilia in the fields of sedimentation, stratigraphy, structural geology, and regional napping. Significant in these are the discovery and description. of the Ediacara medusoid fauna in the Pound Quartzite. formerly considered to be Early Cumbrian but now regarded as of Precambrian ae; initiation of a systematic. study of the Adelaide System and sediinentation in the Adelaide Geosyncline; description of coastline migrations and geochronology of the Quaternary of the South-East Province and of submarine canyons off the Sonth Australian coasts; direetion of renewed interest in structures in the northeast of South Australia, thus encouraging petroleum exploration in the area in which gasfields were subsequently diseavered. Mr. Sprivg is the anthor of some 62 scientific papers, 12 of which have heen published in the Transactions. CONTENTS D. R. Smytu and C. M. Pumporr: A field study of the Rabbit Bandicoot, Macrotis lagotis, Marsupialia, from Central Western Australia - R. M. E. Wexsourn and R. T. Lance: An analysis of vegetation on stranded coastal dunes between Robe and Naracoorte, South Australia R. Scuoppe: Further Taxonomic notes on the species of Millotia Cassini CASteraCeae:) > — yk EN Petey i hl ee te Mea oe G. Beni: A new species of Ptilotus from South Australia - - - J. A. Harris: Age structure, growth rate and spawning cycle of a popula- tion of Yellow-eye Mullet Aldrichetta forsteri (Cuv. and Val.) from the Coorong Lagoon, South Australia - - - - - - M. Smyru: The distribution and life history of the Skink, Hemiergis peronii (Fitzinger) - : ¢ : : > é - ‘ L, M. Ancex and P. M. Mawson: Helminths from some lizards mostly from SeuthAnetraliss te ced tee Sak a tee.) ben hie I. M. Tuomas: Two species of Saccoglossus (Enteropneusta) from South Australia - er 3 ee eee eek E . E. M. Campsetu: Lunettes in Southern South Australia - = aya C. R. Twmate: Origin of Wave Rock, Hyden, Western Australia - - Obituary: Thomas Draper Campbell IN, am Ney be eee General Account, Library Account - - A272 ‘ : J Research Fund - - - - : F 2 : Z 2 Report on Activities of the Council 1967-68 — - = - = F a Officers for 1967-68 - - - - 2 : i = Se a, ES, List of Members - - > = - = = m ~ - List of Lecturers and Exhibits, 1967-68 - - - E - s < Awards of the Sir Joseph Verco Medal - - - - - - .- 37 51 115