JOURNAL OF THE ROYAL SOCIETY OF WESTERN AUSTRALIA INCORPORATED VOLUME 46 (1963) PART 1 PUBLISHED 1st. MAY. 1963 REGISTERED AT THE G.P.O., PERTH FOR TRANSMISSION BY POST AS A PERIODICAL THE ROYAL SOCIETY OP WESTERN AUSTRALIA INCORPORATED COUNCIL 1962-1963 President Past President Vice-Presidents Joint Hon. Secretaries Hon. Treasurer Hon. Librarian Hon. Editor W. D. L. Ride, M.A., D.Phil. J. E. Glover, B.Sc., Ph.D. C. F. H. Jenkins, M.A. W. R. Wallace, Dip.For. Margaret E. Redman, B.Sc. C. V. Malcolm, B.Sc. (Agric.). R. P. J. Pillow, Dip. P.T.C. Ariadna Neumann, B.A. J. E. Glover, B.Sc., Ph.D. N. H. Brittan, B.Sc., Ph.D. R. W. George, B.Sc., Ph.D. J. G- Kay, B.Sc. R. J. Little. D. Merrilees, B.Sc. R. T. Prider, B.Sc., Ph.D., M.Aust.I.M.M., F.G.S. R. D. Royce, B.Sc. (Agric.). L. W. Samuel, B.Sc., Ph.D., F.R.I.C., F.R.A.C.I. Journal of the Royal Society of Western Australia Vol. 46 Part 1 1. — Description of a New Freshwater Fish of the Family Theraponidae from Western Australia By G. F. Mees* Manuscript received — 20th Novemher, 1962 A new species of freshwater Therapon from Millstream Station. West Pilbara. Western Aus- tralia. is described, and notes are given on some other freshwater fishes occurring in the same region, the fish fauna of which was hitherto very imperfectly known. Moreover Mesopristes jenkinsi from the Ord River is shown to be synonymous with TJierapon alligatoris, and the range of the latter is extended to include the Kimberley Division of Western Australia. When Whitley (1947» published his survey of the fresh-water fauna of Australia, he could list only three species of fishes from the mid-west and north-west of Western Australia (his Greyian Fluvifaunula) , and correctly stated that ichthyologically the area could; . . be consid- ered unexplored . . The three species listed by Whitley were Therapon unicolor, Hypseleotris compressus, and the eyeless Eleotrid Milyerinya veritas which is restricted to subterranean waters near the tip of the North West Cape Peninsula and therefore cannot be regarded as character- istic of the area as a whole.t Shipway <1950, 1953) has since recorded Eleotris aurea from the Murchison River and a Melanotaeniid which he identified as Melanotaenia nigrans, from the Bul- lawarrina River, West Pilbara, but the situation has remained essentially unaltered. When therefore, in July and August 1958, staff members of the Western Australian Museum visited West Pilbara for collecting purposes special attention was devoted to the fish fauna of the region. A popular account of the visit was given by Ride (1959), and there is a map of the area in Mees (1961). Fishes were col- lected by Miss K. C. A. Vollprecht (now' Mrs. Thies) and the author. Eight species were ob- tained and I am convinced that, with the ex- ception of an eel, pre.sumably Anguilla bicolor, which was reported by locals but not seen by us. they give a fairly complete picture of the fish fauna of the Fortescue River and its tributaries in the area where we worked (about 70 miles inland). The fishes obtained belong to the fol- lowing families: Ariidae (one species), Ploto- * Western Australian Museum. Perth. Western Australia. V CraterocephaLus cuneiceps from the Murchison River, though not mentioned by Whitley, may also be included: it occurs not only in the upper branches of the river, but was found in its lower course by Shipway (1950): see also Whitley (1955). sidae (one species), Melanotaeniidae (one species), Dorosomidae (one species), Gobiidae (one species), and Theraponidae (three species). A complete report on this collection had been planned, but the systematics of several groups of Australian freshwater fishes are in such dis- order that they can only be solved by major revisions. Fortunately two families, the Melano- taeniidae and the Dorosomidae are now under review by Mr. I. S. R. Munro (C.S.I.R.O. Fish- eries and Oceanography, Cronulla), who has re- ceived all our material of these groups on loan. Until he has completed his work it is of little use to comment on members of these groups. While the main purpose of this paper is the description of a new Therapon, it seems never- theless useful to give some information on the other species collected. Though catfishes have been known to occur in the Fortescue River since its discovery, have been eaten by explorers (F. T. Gregory in Gregory and Gregory 1884. p. 62-63), and have been men- tioned by later authors like Ride (1959), and Ride and Serventy (1961), their specific identity had never been ascertained. I found these fishes to be Arius australis, a species that in Western Aus- tralia had previously been recorded from Noon- kanbah, Kimberley Division (Rendahl 1921). though Munro (1957) did not include Western Australia in its range. This is apparently the common catfish of the State: the Western Aus- tralian Museum has recent material from Langey Crossing, Pitzroy River, west Kimberley Divi- sion, collected on 24.VI.1960 by F. W. Monck, P 5090, and from the King Edward River, Kal- umburu, north Kimberley Division, collected on 27.VI.1960 by A. M. Douglas and G. F. Mees. P 5091. The goby is Glossogobius giuris. The Plotosid and the Dorosomid are as yet uniden- tified. Of the Theraponidae one is the widely distributed Therapon unicolor which occurs over the whole northern part of Australia, in Western Australia at least as far south as the Murchison River. Watson (1958), followed by Ride and Serventy (1961), recorded Therapon unicolor from the Greenough River south of Geraldton. but at the suggestion of Messrs. J. O. Knight and B. J. Parkes, who could find only T. cauda- vittatus in the Greenough River, I examined 1 Watson’s specimens, which were donated to the Western Australian Museum, and found them referable to the latter species. The second Therapon is T. vercoides and the Melanotaeniid is doubtless identical with the one that in literature has been called Melano- taenia niffrans (Whitley 1948, Shipway 1953. Ride and Serventy 1961) or M. australis (Whit- ley I960). It may be recalled that de Castelnau (1875) described Neoatherina australis and Therapon fasciatus from Swan River. While the record of Neoatherina ( Melanotaenia) has generally been accepted as erroneous, the Swan River, meaning south-western Australia, is at present still included in the range of Therapon percoides (of which T. fasciatus has been con- sidered a synonym) on the basis of de Castel- nau’s specimen (Ogilby and McCulloch 1916. Rendahl 1922, Whitley 1948, Nichols 1949, Whit- ley I960). I even believe that de Castelnau’s re- cord is the sole basis for the inclusion of the species in the fauna of Western Australia. While it is true that the name Swan River, or Swan River Colony, used to be applied to an ill-defined but large area of which Perth was the centre, it wculd hardly have included any part of the ranges of Melanotaenia and Therapon percoides, as I do not believe that either of these species occurs far south of the Tropic of Capricorn. In the introduction to his paper, de Castelnau states that the collector of his material from Western Australia, the Reverend G. J. Bostock. who lived at Fremantle, also provided specimens from the Dampier Archipelago, and it is likely that the types of Neoatherina australis and Therapon fasciatus were actually obtained somewhere on the mainland opposite the Dam- pier Archipelago. The Kimberley Division, as was only to be expected, can be included in the range of T. percoides, as in June 1960, I found it in several tributaries of the King Edward River near Kalumburu. The third Therapon apparently represents an undescribed species that may be characterized as follows: Therapon aheneus species nova (Pig. 1) Description. D (XI)-XII-(XIII). (8^)-9L A 111.8^, P 13-14 (ii.lO.i or ii.lO.ii), VI.5, C 19-20 ( ignoring some very short undivided rays on each side: usually ii.15.ii), gillrakers on outer branchial arch 6-8+1 + 11-12, branchiostegals 6, scales in lateral line 40-44, scales under lateral line 38-44 in a longitudinal series. Differs from all other species of the genus by the combina- tion of fin formula, gillrakers, and scale num- bers. A very normal representative of the genus; body moderately deep. Head 2.7 to 3.0 in stan- dard length, about equal to depth of body or slightly less; profile of snout and forehead slightly concave, becoming convex behind eye: eye large, 3.8 to 4.0 in head; snout of about same length as eye diameter or a little longer, maxil- lary reaching to below anterior border of eye;* nostrils well separated, anterior one just above upper lip, with an elevated rim that might pass for a very short tube; posterior one in front of middle of eye, in the anterior part cf a nar- row longitudinal groove with an elevated rim: mouth moderate: each jaw has an outer series of fairly small teeth, followed by a narrow band of very fine teeth: no teeth on palatine or tongue; no canines: posterior border of preoper- culum free, finely serrated; operculum with a flat spine that does not normally protrude be- yond the border cf the soft operculum; branchi- ostegals six: suprascapular bone exposed. Lateral line slightly arched, following the pro- file of the back, becoming straight on the caudal peduncle, and continued to the basis of the tail. Body scaled, with the exception of the upper surface of snout and head, and chin; scales largest on the sides on the middle of the body. Dorsal fin normally with twelve strong spines, increasing in length from the 1st to the 5th: the 2nd twice the length of the 1st; the 5th * In specimens fixed with open mouth, these proportions are distorted. Fig. 1. — Outline drawing of Therapon aheneus, type, nat. size. 2 longest. 1.6 to 2 0 times eye diameter; from there onwards each following spine slightly shorter than the preceding one, until the 11th which is of the same length as. or slightly shorter than the 2nd. and slightly shorter than the 12th; spiny dorsal much higher than soft dorsal. Of 22 specimens. 18 have D XII. 9L two DXIII. one D XIII. Sh, one D XI. 104- Anal fin with three strong spines and eight rays (all but the first divided), the last of which is double. Soft part rounded. Of the spines, the second is the strongest, more than twice the length of the first; it is not or only a little longer than the third, but much stronger. Pectorals rounded, fairly short, about 1.6 in head, with thirteen developed rays, of which nine or ten are divided. Ventral fins with one strong spine and five divided rays; their origin almost exactly oppo- site origin of D. Caudal fin slightly emarginaie, not quite symmetrical, the upper lobe slightly larger than the lower lobe, with 19 or 20 rays, of which 15 are divided. Size. As the largest specimen obtained mea- sured only 101 mm in standard length, and as no larger individuals were seen, the conclusion seems justified that T. aheneus is a small species. Colours. In life, colour bronzy blackish-brown to dark brown, sides slightly tinged pinkish, small specimens indistinctly banded vertically. P slightly yellowish. Type. A specimen of 90 mm standard length, collected on 18. VII. 1958 by K. C. A. Vollprecht and G. F. Mees at Millstream Pool. W.A.M. regd. no. P 5350. Further material. Eleven specimens, varying from 51 to 101mm standard length, same data as type. P 5351 (two specimens of this lot have been donated to the Leiden Museum); five specimens, standard length 68-95 mm. 25. VII. 1958. same locality. P 5373; three specimens, standard length 59-70 mm. 18. VII. 1958. Mill- stream bath house, P 5358; two specimens, standard length 47^ and 51 mm. 20. VII. 1958, Fortescue River at Millstream, P 5423; one speci- men, -standard length 50 mm. 21. VII. 1958. Fortescue River at Millstream. P 5424. Distribution. Therapon aheneus was plentiful in Millstream Pool, near Millstream Station homestead, in the small creek connecting the pool with the Fortescue River, and in the Fort- escue River at Millstream, which at the time of our visit consisted of a chain of large, deep, pools. We, and several other collectors who have recently visited the district and donated freshwater fishes to the Western Australian Museum, failed to find the species anywhere else, so that it is likely that it has a very limited distribution. Related species. Working from keys and descriptions it appeared that T. aheneus is clos- est to T. argenteus, a species that has not yet been recorded from Western Australia and of which the Western Australian Museum had no material. Thanks to the courtesy of the Chief Inspector of Fisheries of Queensland, two large specimens were received for comparison, one of which we w'ere allowed to retain for our collec- tion. These specimens differ by a slightly differ- ent fin formula, D XII. lOL A III. 8L smaller scales (about 52 below lateral line), number of sillrakers, x 1 ^ 15. relatively deeper body; different position of nostrils. T. argenteus grows also to a much larger size than any specimen of T. aheneus I have seen: Ogilby & McCulloch (1916) examined specimens of up to 262 mm in length, and Weber & de Beaufort (1931) mention a length of 275 mm; it is also different in colour. The affinity to T. argenteus places T. aheneus in the group of species separated by Whitley (1943) under the generic name Mesopristes. While I agree with Whitley that a subdivision of the large genus Therapon might be useful, I prefer for the moment to keep all species under the one name. Whitley (I960) placed in the genus Mesopristes two other nominal species which therefore, as fairly close relatives of T. aheneus. require discussion. These two species are Therapon alligatoris Rendahl (1922), described from the South Alli- gator River and the McKinley River, Northern Territory, and Mesopristes jenkinsi Whitley (1945) from Ivanhoe Station. Ord River, West- ern Australia. In the original description of M. jenkinsi. no reference was made to T. alli- gatoris. The type localities of these two nominal species are in the same general geographic region, and only about 250 miles apart, and comparison between the description of T. alli- gatoris and the unique type of M. jenkinsi (which is in the Western Australian Museum, regd. no. P 2763). shows but very few differenccvs. The type of M. jenkinsi has a standard length of 113 mm. DXIII.il. A III.8. gillrakers 9 1 + 19 (of which the last two or three rudimentary). T. alligatoris differs only in hav- ing D XII. 12-13, gillrakers 9 + 17. Whitley (1945) described the colour of M. jenkinsi as fairly uniform dark slate grey: this was when the specimen was still fresh (the register shows that it was received on 30 October 1944); at present it is light brown in colour, and agree.s with Rendahl’s colour description of T. alliga- toris: “Colour in alcohol light brownish, paler on the ventral parts. Back with a slight tinge of purplish. Pectorals yellowish, the other fins with dark membranes, on the posterior border of the soft dorsal and anal a black basal blotch”. The only differences are that the darkening of the fin membranes has only just set in; there is already a darkish blotch at the end of the dorsal fin, but the anal fin is as yet without one. Hitherto both the species T. alligatoris and M. jenkinsi seem to have been known from their respective types cnly; three specimens of T. alligatoris. and one of M. jenkinsi. Besides the type of M. jenkinsi. the Western Australian Museum has six specimens, all collected by A. M. Douglas and G. F. Mees. Two specimens, standard length 182, 187 mm, 26.VI.1960, Kal- umburu. P5384; two specimens, standard length 151, 159 mm, June 1960, Kalumburu, P. 5385; two specimens, standard length 249. 268 mm, July 1960, Beverley Springs, P4386. The fin formulae of these specimens, in the sequence in which they are listed above, are: DXII.12i, AIII.9i: D XIII. 12i, A III.9i; D XII. 121. A 111.8^: 3 D XII.llL A III. 7^; D XII.llL A 111.8^; D XI. 12-2, A III.8L At the moment of writing (Oc- tober. 1962), over two years after their capture, these specimens are still blackish in colour. The large variation in fin-ray counts of the additional material, D XI-XIII.ll ‘ -12^-, A 111.7-^- 9-L shows that the slight differences in fin- formula between T. alligatoris and M, jenkinsi are well within the limits of individual variation. Therefore I have no hesitation in concluding that Mesopristes jenkinsi Whitley is a synonym of Therapon alligatoris Rendahl. T. aheneus differs from T. alligatoris in hav- ing fewer dorsal rays, larger scales, fewer gill- rakers, and in its different colour. I have, besides the paper by Ogilby & McCul- loch (1916), used Fowler’s (1931) revision, but none of the additional species listed by him seems to be close to T. aheneus. References Castelnau, P. de (1875). — Researches on the fishes of Australia. Viet. Off. Rec. Philad. Exit. Fowler, H. W. (1931). — Contributions to the biology of the Philippine Archipelago and adjacent regions. The fishes of the families Pseudo- chromidae. Lobotidae. Pempheridae, Pria- canthidae. Lutjanidae. Pomadasyidae, and Teraponidae, collected by the United States Bureau of Fisheries steamer “Albatross”, chiefly in Philippine seas and adjacent waters. Bull. U.S. Nat. Mus. 100 (11). Gregory. A. C. and Gregory, F. T. (1884) .—“Journals of Australian Explorations” (Government Printer: Brisbane). Mees, G. F. (1961). — An annotated catalogue of a col- lection of bird-skins from West Pilbara, Western Australia. J. Roy. Soc. W. Aust. 44; 97-143. Munro, I. S. R. (1957). — Handbook of Australian fishes, pp. 37-40. Fisheries Newsletter. Nichols, J. T. (1949). — Results of the Archbold Expedi- tions. No. 62. Fresh-water fishes from Cape York, Australia. Amer. Mus. Novit. 1433. Ogilby, J. D. and McCulloch, A. R. (1916). — A revision of the Australian Therapons with notes on some Papuan species. Mem. Qd. Mus. 5: 99-126. Rendahl. H. (1921). — Results of Dr. E. Mjdbergs Swedish Scientific Expedition to Australia 1910-13. XXVIII. Fische. K. Svenska VetenskAkad. Handl. 61 (9). — (1922). — A contribution to the ichthyology of north-west Australia. Nyt Mag. Naturv 60- 163-197. Ride, W. D. L. (1959). — A museum expedition to the Hamersley Range. Aust. Mus. Mag. 13: 94-98. Ride. W. D. L. and Serventy, D. L. (1961).— The fauna of Western Australia, in Little (editor): “Official Year Book of Western Australia. 1960, N.S. 2”: 59-70. Shipway, B. (1950). — Notes on the aquatic natural history of the lower Murchison River. W. Aust. Nat. 2: 73-77. (1953). — Additional records of fishes occur- ring in the fresh waters of Western Australia. W. Aust. Nat. 3: 173-177. Watson, J. A. L. (1958). — The occurrence of northern fish and dragonflies in the Greenough River. W. Aust. Nat. 6: 184. Weber, M. and de Beaufort, L. F. (1931). — “The Fishes of the Indo-Australian Archipelago” VI (Brill: Leiden). Whitley, G. P. (1943). — Ichthyological notes and illus- trations (part 2). Aust. Zool. 10: 167-187. (1945). — New sharks and fishes from Western Australia (part 2). Aust. Zool. 11: 1-42. (1947). — The fluvifaunulae of Australia with particular reference to freshwater fishes in Western Australia. W. Aust. Nat. 1: 49-53. (1948).- A list of the fishes of Western Aus- tralia. W. Aust. Fish. Dept., Fish. Bull. 2. (1955). — Freshwater Atherines from Western Australia (Pisces: Atherinidae). W. Aust. Nat. 5: 25-31. (1960). — “Native Freshwater Fishes of Aus- tralia” (Jacaranda Press: Brisbane). 4 2. — Copper Poisoning in Sheep in Western Australia By A. B. Beck* and H. W. Bennettst Manuscript received — 16th October, 1962 Investigations were made to ascertain the cause of high-copper status of sheep in the Wiluna area of Western Australia. It was con- cluded that this was due mainly to the ingestion of plants naturally high in copper. An occurrence of copper poisoning in sheep in the Toodyay area was also investigated. Here the copper content of pastures generally was not high. Although the factors responsible for the high copper status of sheep were not determined, histological evidence suggested that this was associated with a hepatotoxic principle of plant or fungal origin. Introduction During investigations relating to the deficiency of copper in stock in Western Australia and in the course of routine laboratory diagnosis, we have encountered numerous cases of copper poisoning in sheep. In some instances the cause has been due to an over-generous use of copper-containing licks or fertilizers, often to both. In other cases the poisoning has been associated with lupinosis when the damaged liver accumulates excessive amounts of copper, particularly if copper supplements have been fed. Investigations into this disease and the significance of the storage of heavy metals in the liver are being continued by the Depart- ment of Agriculture at the present time. In eastern Australia, copper poisoning is commonly associated with plants containing pyrrolizidine alkaloids, particularly heliotrope (Heliotr opium europaeuvi) and “Pattersons Curse” (Echium plantagineum i (Bull 1961; St. George-Grambauer and Rac 1962). The first plant is rare in southern Western Australia and no cases of copper poisoning due to either plant have been reported. The purpose of the present paper is to record the results of detailed investigations of the cause of the high copper status of sheep at Wiluna in the North Eastern Goldfields pas- toral region, and at Toodyay. The High Copper Status of Sheep in the Wiluna Area In 1951 one of us (H.W.B.) visited the Eastern Goldfields area of Western Australia to enquire into possible causes of sheep losses in the 1950-51 drought. Some pastoraiists had considered that the losses were unduly severe and could not be attributed solely to the effects of drought. On some stations it was reported that sheep had * Formerly Division of Animal Health, C.S.I.R.O., present address. Division of Plant Industry. C.S.I.R.O., c/o Department of Agriculture. South Perth. Western Australia. i Formerly Department of Agriculture of Western Aus- tralia, present address, 44 Louise Street, Nedlands, Western Australia. lost their appetite for top feed which should have been adequate for their requirements. As a result of these observations it was decided to investigate the remote possibility that cobalt deficiency was responsible for this reported anorexia. During the analysis for cobalt con- tent it was noted that the livers contained large amounts of copper. Shortly afterwards chronic copper poisoning was diagnosed at “Albion Downs” Station. The present investi- gation was then carried out to ascertain the reasons for the high copper status of sheep at Wiluna and the area of country affected. Because of the remoteness of the region and of the large areas of the properties concerned, the scope of the investigation was restricted to the determination of copper in the livers of sheep from nine station properties, to the analysis of herbage from three properties where sheep showed high copper status and to the deter- mination of copper in well waters on one pro- perty. Histological examination was made on many of the liver samples to check for the possible effects of hepatotoxic plants. Materials and Methods At “Albion Downs" Station, every species likely to be eaten by sheep was collected in the 1953 sampling. The sampling was much less comprehensive at later dates and on other stations. Soil contamination was avoided in the col- lection of samples but in some short and semi- prostrate species it was not possible to avoid this entirely. In these cases an iron determina- tion was done to obtain some indication of soil contamination. As levels of molybdenum, manganese and in- organic sulphate are known to influence cop- per metabolism in sheep, determinations of these constituents were made on many of the herbage samples. Copper, molybdenum and inorganic sulphate were determined as described previously (Beck 1962), manganese by the periodate method (Willard and Greathouse 1917) and iron by the thioglycollic acid method (Mayer and Bradshaw 1951), Except for one sheep which died from copper- poisoning, all liver samples were from healthy sheep killed for rations. The livers were pre- served in copper-free alcohol or alcohol-formalin mixture. The analysis of all samples is reported on the dry-matter basis. No correction was made for the fat content of livers, but no obvious fat was noted in any of the samples. 5 Fig. 1. Results An examination of well waters was made on “Albion Downs” Station in 1953 but the copper levels were less than 0.04 mg per litre and would not be responsible for the high levels of copper in liver. The analytical data for liver samples and for herbage material are set out in Tables I and II respectively. The livers of normal Western Australian sheep contain 50 to 400 p.p.m. copper and as sheep with levels over 1000 p.p.m. are liable to develop copper poisoning under condi- tions of stress, the percentage of such samples is indicated in Table I. The geographical dis- tribution of properties is shown in the accom- panying map. With Goodenia Mueckeana and Eremophila leucophylla there was little difference in copper content at the different times of sampling, but in all other species the levels in the 1953 samples were appreciably higher. There was no obvious difference in other constituents of species at different sampling times or on the different properties. A number of livers from ration sheep have been examined histologically for evidence of ingestion of hepatotoxic plants. Some of the high-copper samples showed small areas of megalocytosis but there was no significant liver damage in any of the specimens. Prior to the outbreak of copper poisoning in 1952 it had been noted that the affected sheep had been grazing heavily on Goodenia eremo- phila. This species did not reappear in appre- ciable amounts until 1959 when a sample was forwarded to the C.S.I.R.O. Chemical Research Laboratories, Melbourne, for alkaloid determina- tion. The analysis showed a very low alkaloid content (0.015 per cent, tertiary base, on as- sumed molecular weight of 300; N oxides were absent) which would be unlikely to cause any liver damage. Discussion Prom the data obtained it would seem that the occurrence of high copper status in merino sheep is restricted to an area within 100 miles of Wiluna. It seems probable that the development of the high liver copper levels was primarily due to the ingestion of herbage of high copper con- tent. Data from the agricultural areas of West- ern Australia (Beck 1941, 1962) indicate that in those regions, levels of copper in pastures 6 TABLE I Copper Content of Wiluna Sheep Livers Values as p.p.m. Cu on dry liver Uver (‘11 I'ropcrty Date .\o. ol Samjiles Mean ami Ranjie Percentage of values above 1 OOO ]).p.iii. Details 1 “ Albion Downs " Sept.-Xov., 1952 7 1010 300-1700 43 One slieep died Cu poisoning •Ian. -April, 1953 •filly, 1953 <; O 1410 380- 1990 SOU 780 940 83 0 Ageil w(*tbers Sept., 1953 3 ItiOO 1190-2100 100 Aged ewes Sept., 1953 Alls;. -Sept., 1954 7 s .520 440- OOO 810 31)0-1390 0 37 5’oimg sli(H‘i) 12 30 inonths Sept., 1955 12 soo 390-1940 33 Mixed age (‘ues and wetliers Sept. -Oct.. 1951) (i 1250 300 -21 10 83 Aged wetliers ’■ Vfclcrri(( ” ... May-Oct., 1959 () 1)20 220- 1440 17 Aged ew(*s and wetliers .Ian., 1954 9 1820 790-2730 (17 Wethers mixed ages ■■ Limia (ilcn ' l-'eb.-Auii., 1955 s 780 210 1900 25 5\'ethei's 3 (5 years Mareh-April, 1954 5 I •>•)(» 520-1850 (10 Ag('d sliee]) ‘ Lake \\ ay " Xov., 1951-Feb.. 1952 . , Ort.-Xov.. 1952 3 () 870 310-17.50 720 350-2000 33 17 •lan.-Man-h. 1953 3 800 450-!380 33 Wetliers 3 4 years ■ \^ aniiJinoo ’’ .fnly-Auji., 1951) () 450 180-740 0 Wethers 3 7 years ‘ Uranite Peaks” Au^f.-Xov.. 1952 (> 380 150-820 0 Wethers 3-5 years ■ Cunyu ” An SO, Sampled Sam])les p.p.m. p.p.m. p.I>.m. p.p.m. 0 . n Gotxleuiu M .Dilv. 195:1 a e 4 19-9 <0- 1 900 0-15 (2) 17 1--20-9 790-1100 0- 14-0-lfi Pel). -Aim., 195.5 a 1) 4 19-1 <0-1 7(50 190 (2) 0-11 1:}-S-2I -7 570-1000 150-2:}() 0-08 0-18 Oct.. 1952 a 1 15-7 .luiv, 195;l a :} 11 - S <0- 1 4:}() lO-O-l:}-:} <0-l-0- ll> :}40-500 Sept., 1954 a 1 ()-S <01 490 0:}C Aug.. 1955 a h •> s-;} .'(1-1 1100 0 1(5 8-0-S-7 810-14U0 0-08 -0-25 Oct.. 1959 a 1 7-9 0 -n :}70 0-14 Up! ir/i /■//-< ii in Dll ren port i 1 Julv. 1952 a 1 20-;} ( Kvcrlastiiiii) Sept.. 1954 a 1 15-:} 95-100 0- 51-0 -74 Diitixnix iflani'ifoliiis .(iilv, 195:} a 1 l:}-() <0-1 Trir/iiniinn ohorntinii Julv. 195:} a 1 1 :} • 5 0-28 100 (Cntton Imsli) .\im., 1955 a h 2 5 -a 0-72 57 1(55 (1) 0-1(5 5 • 5 -5 • S 0 21-1 -l:} 5()-(55 0-1,5-0-17 7’. I’.aiUiitiiiti Julv. 195;} ... r 1 10-1 0 • 8() Miin'orppfioliix (iiierinne Julv. 195:} a 2 12-7 0-28 (I) :)0() (1) 1 10 (liiltv Imtton) 10-8-14’() 0-7:}-l-47 Aim. 1955 a ii 2 9-4 ()• 19 450 445 (1) 0-:}2 7 • 5- 1 1 • 4 0 08-0-:}0 200-4)90 0-25-0-40 tJrpiHOphilii h‘urop!ii/!l(i ■Julv. 195:} a 4 12-:} 0-15 (;}) COO 0-21 (Povt*rty or Ilaniud l)iish) 9-8-i:}-9 0-l:}-017 400-740 ()• 15-0-27 Aim.. 1955 .. a 1. ;} 12-8 0-22 450 1(50 (2) 0-26 10-:}- 14-5 0-l:}-0-29 190-870 .150-170 0-22-0-:}() EroiUiiiii rijf/Noniin . . Julv. 195:} a V () 11-1 0-17 (5) l:}0 (1) 0-;}4 (2) 9-:}-l4-a <0- 1-0-;}7 0-24 0-44 Kemieib/a sp. Julv. 195;} . a 1 lo-;} ()•:}:} 0-14 SU/ii ('ornufKta Julv. 195:} a 1 9-8 0-24 Danthunid bipurtHa (Wan- July, 195:} . . a 4 9-0 0-15 (;}) 290 (2) ()-:}l (2) darru* grass) 7-7 lU-4 0-10-0-25 250-:}:}0 0-:30-0-:}2 Iliielinijia loxophnUa Julv. 1952 ... a e 9-0 275 (1) 0-12 (•) -9-11-0 0 • 08-0 • 1 r> Ualorrhaijix sj». Aim., 1955 a 1 8-8 - (1 ■ 1 2400 207 0-:}7 lias-sia sj>j). Julv, 195:} a (• 5 8-«> 0-42 (2) :}•:} (4) 7-:}-9-8 0-27-0-58 0-8-C-0 Coiloiwrarjnis vutinifoUnx July. 195:} .... a 8-G 0-()2 (1) :}(>0 (1) 0-79 (1) (Xative jioplar) 7 •9-9-:} Jiriirhi/xenia ( 'l/innhprsii July. 195:} a 2 0-:} (dowers onlv) (>•0-1) -7 Aravin i/enistoidrs Salt lUishes July, 195;} a 1 :}•;} 0-10 Korkin pi/rawi(!ata Julv, 195:} a 1 8-2 5-0 Atrijile.r palndoxa Julv, 195:} a 1 4-9 1-0 Jii/rium aaxtraJr ... July. 195:} ,. a 1 9-:} 2-0 * “ a ” hidicatfs '• AU)ion l)(,iwns.” “ li ” “ Yeelirrie ” and “ o ” Lt)rna Oleii.” 8 TABLE III Copper. Molybdenum and Inorganic Sulphate Levels of Toodyay Pastures Values expressed as mean and range on dry matter basis 1 No. of Details ('ll .Mo SO, SniTjples p.I>.in. p.]).ni. O 0 Oct., H»r).> .. .) Caf'eweed 14-1 0 ■ (if) 14-2 14- () 0d0-()-70 0-28-0-:l2 1 Siih. eIov(*r , . 14- ti 0-40 0-24 1 >tixp(l .«ur. clovfr and cajiewcrd 9-7 0 •:'>{) 0-22 iPofi ... fi (trasses ... 0- 7 I :17 0-19 4 • :}-S • H ()(>2--2-2(? 0- 12-0-24 4 Capeweed ()-7 1 -87 0- 19 1 0-2 -d 0- 17-0-22 Mixed pa.'^ture 9-7 1 10 0-17 1C2-10-2 0-0.5 i -2r> - 1 Capeweed 14-4 0 • 49 0-19 1 Plan(nr/o crf(H‘a SC2 0-49 0-28 ! .Mixed pasture 1l() 0-dO 0-18 Any., iy;')7 . 4 Capeweed 115 0(l:l 0-28 Ol- 14-U 0-48 -0-94 0-17O-49 1 (Jrasses . S-4 0(51 0-17 2 SuVt. e!ov«‘r 120 0-H) 0-12 10-2-i:Ct> 0- 14 0- 18 0-07 0-18 Pasture Investigations The pastures were of the annual Mediter- ranean type common to the 20-25 inch rainfall belt. The main species were Trifolium subter- raneum, capeweed ( Cryptostemma calendula) , wild geranium (Erodium botrys) and annual grasses f Bromus spp. Vulpia myuros and Wim- mera rye grass, Lolium spp.). Surveys were car- ried cut by the former Government Botanist (Mr. C. A. Gardner) in August, 1955, and Sep- tember, 1956, but the only unusual species found was Plantago cretica. An examination of dry paddocks in March, 1957, showed no unusual plants and none considered likely to cause toxic effects. Echium plantagineum (“Patter- sons curse”) and lupins (L. varius and L. angustifolius) , known to cause liver damage, were completely absent from the property. Analysis of pasture samples is set out in Table III. Levels of molybdenum and inorganic sul- phate were determined as these are known to affect copper metabolism. Two samples (Aug- ust. 1956) were analysed for manganese content but normal levels were found (72 and 85 p.p m. cn dry matter). As the histology described in the follow^lng section had suggested the action of a hepato- toxic alkaloid, determinations for alkaloid con- tent were made on P. cretica at the C.S.I.R.O. Chemical Research Laboratories, Melbourne. As the manager of the property had stated that cases of poisoning only occurred in years of rank capeweed growth, this species was also examined, even though there was no suggestion that it caused trouble elsewhere. Both species showed very low alkaloid content (P. cretica, 0.018 per cent, tertiary base, quaternary and weak bases and N oxides absent; capeweed, 0.014 per cent, tertiary base and 0.004 per cent. N oxide) . Animal Studies Chemical. — The liver of the sheep dying of copper poisoning in August, 1955, showed 900 p.p.m. copper (dry basis); a similar sheep in July, 1957, showed 600 p.p.m. These values are rather lower than those usually found in cases of haemolytic jaundice due to copper poisoning. In June, 1956, liver samples for chemical and histological examination were collected at the abattoirs from 30 sheep from the property. The sheep had been without food for at least 18 hours and consequently the livers contained more fat than usual. There had been no losses from haemolytic jaundice during yarding. Some of the livers were macroscopically abnormal; two were small, five were yellowish and one rather fibrous. Chemical analysis showed the following results which are expressed as the means with range of values in parenthesis: the values for fat are on the dry material and values for iron and copper on the dry, fat-free material: — copper 1500 p.p.m. (490-3120), iron 990 p.p.m. (280-2410), fat 23 p.c. (15-40). Seventy-three per cent, of the livers contained over 1000 p.p.m. copper. Livers from normal Western Australian sheep contain 50-400 p.p.m. copper, 200-800 p.p.m. iron and less than 10 per cent. fat. Histology. — Sections were available from the two moribund animals and the thirty abattoirs animals mentioned above. In the liver from the first affected animal (August, 1955) the interstitial tissue in the portal tracts was slightly increased and abnor- mally cellular. A slight fine diffuse fibrosis was present. Exces-s bile pigment was present in the ducts and canaliculi. There was some new bile duct formation. Ceroid and protein in- clusion globules were plentiful, but megalocy- tosis and central necrosis were absent. The kidney tubules were laden with haemoglobin casts and degradation products. The second liver (July. 1957) showed marked portal tract fibrosis infiltrated with lymphocytes and poly- morphs. These reactive cells were also signifi- cantly increased in numbers, both diffusely and focally throughout the liver, presumably as a reaction to necrosis of liver cells. There was a great variation in nuclear size. The reticulo- endothelial cells were increased in numbers, swollen and packed with degenerating red blood cells and bilirubin. There was some small bile duct proliferation in the portal tracts. The livers of many of the abattoirs sheep showed cellular reaction in the portal tracts 9 with fibrosis and bile duct damage. The Kupffer cells frequently showed yellow-brown granules. Discussion Copper levels of pasture were at times mod- erately high but it is not considered that these alone could have caused a dangerous accumu- lation of copper in sheep livers. Molybdenum and inorganic sulphate levels were normal. The histological data on livers were limited but suggested that two processes were involved. The first consisted of fibrotic and bile duct changes probably leading to some degree of excretory obstruction. This had probably been acting for some time before deaths occurred, and could have been due to ingestion of a plant containing a hepatotoxic alkaloid or to recovered facial eczema due to fungal toxicity. The absence of megalocytosis indicated a dif- fernt type of poisoning from that due to Helio- tropiuni (Bull 1961). The second change in the livers was due to an acute haemolytic process which caused the actual deaths. Although the clinical findings indicated that this was due to copper poisoning, the histological picture gave some suggestion that it may have been due to difficulty in the excretion of bilirubin normally produced. The relatively low liver copper liver values for the two sheep which died also gave some support to the idea that copper toxicity was not the primary cause of death. It is not possible to give a satisfactory ex- planation for the massive accumulation of copper in the thirty abattoirs sheep but it was probably consequent on the liver damage as in heliotrope and lupin poisoning. Acknowledgments This investigation was carried out cooperat- ively by the Division of Animal Health, C.S.I.R.O. and the Western Australian Department of Agriculture. The authors wish to express their indebted- ness to the station owners of the Wiluna region who co-operated in this investigation and in particular to Mr. and Mrs. R. S. Howard, of “Albion Downs” Station. Acknowledgment is also made to Dr. J. R. Price of the C.S.I.R.O. Chemical Research Laboratories, Melbourne, for arranging the determination of alkaloids in herbage: to the Botany Branch of the Western Australian Department of Agriculture for identification of species; to Dr. L. B. Bull of the Division of Animal Health. C.S.I.R.O., and Dr. M. R. Gardiner of the Department of Agri- culture of Western Australia for discussions on the histology of the Toodyay samples, and to Messrs. C. N. Macliver. A. Negrin and V. McLinden for analytical assistance. References Anon. ( 1956).— Toxaemic jaundice in sheep. Final report of Investigation Committee. Aust. Vet J ^2: 229-236. Anon. (1957-58). — Tenth Annual Report of C.S.I.R.O. ( Aust.) : 51. Barker. S. (1961). — Copper, molybdenum and inorganic sulphate levels in Rottnest plants. J. Roy. Soc. W. Aust. 44: 49-52. Beck. A. B. (1941). — A survey of the copper content of Western Australian pastures. J. Dep. Agric. W. Aust. (2) 18: 285-300. Beck, A. B. (1962). — The levels of copper, molybdenum and inorganic sulphate in some Western Australian pastures. Aust. J. Exp. Agric. Aiiivi. Hush. 2: 40-45. Bull, L. B. (1961). — Liver diseases in livestock from intake of hepatotoxic substances. Aust. Vet. J. 37 ; 126-130. Dick, A. T. (1954). — Studies on the assimilation and storage of copper in crossbred sheep. Aiist. J. Agric. Sci., 5: 511-544. Gubler. C. J., Cartwright. G. E., Taylor. D. S., Eichwald. E. J. and Wintrobe, M. M. (1953). — Chronic manganese and copper poisoning in rats and its possible relation to hepatolenticular de- generation in man. Fed. Proc. 12: Abstract 1366. Mayer, A. and Bradshaw, G. (1951). — Photometric de- termination of small amounts of Iron in magnesium and magnesium alloys by thio- glycollic acid. Analyst 76: 715-723. Spais, A, (1956). — A contribution to the study of enzootic ataxia of lambs in Greece. University of Thessaloniki: 121 pp. (Table 8). St. George-Grambauer, T. D. and Rac, R. (1962). Hepa- togenous chronic copper poisoning in sheep in S. Australia due to the consumption of Echium plantagineu7n (L). Aust. Vet. J. 38- 288-293. Willard, H. H. and Greathouse, L. H. (1917). — The colori- metric determination of managanese by oxidation with periodate. J. Amer. Chem. Soc. 39: 2366-2377. 10 3. — Two New Western Australian Cockroaches By K. Princis* Manuscript received — 21st August, 1962 I received recently from Mr. Athol M. Douglas, of the Western Australian Museum, a small lot of cockroaches for identification. These cock- roaches had been collected partly in caves partly in mines and although but two species were represented they proved to be new to science. For the opportunity to study them I wish to express my gratitude to Mr. A. M. Douglas. Shawella douglasi, sp. nov. (holotype). Western Australia. Jurien Bay. Limestone Caves <30° 17'S, 115°E), IX. 1958 ( associated with the droppings of small cave dwelling bats Eptesicus pumilus), A. M. Douglas leg. (W. Aust. Mus.); 3 larvae, the same data (W. Aust. Mus. and Lund Mus.). 6 . Light brown with rather weakly chitinized integument. Eyes somewhat reduced (they are rather narrow with acute apices instead of rounded). Maxillary palpi and antennae rather long. Pronotum flat. Tegmina slightly overlapping in the middle of dorsum and reach- ing with their rounded apices to the 4th tergite; subcosta provided with rami anteriores in its distal part. Wings vestigial, reaching with their apices to the 2nd tergite. The 1st tergite with a well developed glandular area. Supra-anal plate (Fig. 1) quadrangular, with rounded latero-caudal angles; its caudal margin rather heavily spined and mesally slightly emarginate, Hypandrium (Fig. 1) considerably exceeding the supra-anal plate, markedly asymmetrically de- veloped; the right style is somewhat larger than the left one and is situated about on the median line, while the left style is strictly confined to the left side of the plate; the upper surface of both the styles is heavily spined. Legs rather long, heavily spined. Lower anterior margin of front femora armed after type At and bearing 3 heavy distal spines. Pulvilli lacking. Tarsal claws symmetrical, unspecialized. Small arolia present. Length of body 17.5 mm; length of pronotum 4.5 mm; width of pronotum 5 mm; length of tegmina 9 mm. The female of S. douglasi is still unknown. The new species differs from S. couloniana mak- ing up about 40 per cent, of the rock. It is massive and porphyritic, with hard porphyro- blasts in a softer, serpentinous, fine-grained groundmass. Small flakes of chlorite are dis- seminated throughout the rock. In thin section the texture is pseudoporphy- ritic. Large pyroxene porphyroblasts * have in- clusions of olivine, some of which are euhedral. No crystal faces are developed on the pyroxene. The following minerals are present: Antigorite (approx. 45%), very fine-grained fibrous or laminated mineral, occurring as pseudomorphs after olivine giving a characteristic mesh tex- ture of “grains” 0.2 mm in diameter. There is a strong tendency for it to be oriented with X and Z lying in the same directions throughout the whole slide. Fine inclusions of magnetite dust occur everywhere within it and particularly along the contacts of “grains” pseudomorphous after olivine. Alteration is to clinochlore by re- action with spinel. Bronzite (approx. 33%). rather irregular grains, from 1 to 5 mm in diameter, with abundant inclusions of anhedral. sieve-textured olivine averaging 0.5 mm in * The term porphyroblast is used here to describe prominent crystals which have grown in a solid or crystalline medium. 15 ciiameter, and also of rounded grains of spinel and magnetite 0.1 mm in diameter. Olivine (ap- prox. 16%^. occurs mainly as sieve-textured in- clusions in bronzite 0.5 mm in diameter, and as relics 0.1 mm in diameter in mesh-textured antigorite marginal to bronzite. Some of the grains included in the bronzite show develop- ment of the 010, 001 and 021 faces, but such euhedral grains, though significant, are excep- tional. Alteration is to antigorite and magnetite dust, or to hornblende. Hornblende (approx. 3% t . rare prisms up to 0.5 mm long or pseudomorphs after olivine. Spinel (approx. 2%K dark yel- lowish green 5 GY 5/4t rounded grains (0.1 mm in diameter), included in bronzite. Alteration is to magnetite and isotropic, fine-grained chlorite, by reaction with antigorite. No spinel remains except where it is armoured with an- other mineral against antigorite: reaction rims are not well developed in this specimen. Clvio- chlore (trace), books up to 0.05 mm by 0.3 mm associated with magnetite. These associated laminae cut across the mesh texture and exert their form strongly. The relationship of the olivine and bronzite in this specimen is most significant. The tex- tures that may be produced from bronzite with abundant poikilitically included olivine will be discussed later. Hornblendic bronzitite. — Specimen 39739 was collected from 084268 where the outcrop con- sists of scattered boulders. The rock appears to consist of normal bronzite peridotite (with the typical rough weathered surface) in which are ellipsoidal blebs of material about 9 in. by 4 in. by 6 in. which weather to a smooth sur- face. These blebs are outlined by a weathered groove, about ^ in. deep and in hand specimen they are uniform grey and of even hardness. Numerous grains of amphibole averaging about 1 mm long compose about 50% of the blebs. In thin section the texture is poikilitic. Large masses of bronzite enclose numerous hornblende prisms, which in ouantity exceed the host crys- tal. Bronzite (approx. 4:0%), forms oikocrysts up to 5 mm in diameter, which are cut by small veins of antigorite and chrysotile. Alteration is to chrysotile. Olivine (trace), occurs only as rare relics in mesh-textured antigorite inter- granular with respect to bronzite. Hornblende (approx. 52%), occurs as euhedral. subhedral or anhedral prisms included in bronzite and is commonly traversed by chrysotile veins. This specimen is notable in several respects: — (a) The absence of spinel and clinochlore both of which are found almost universally throughout the ultrabasic body. Possibly both of these minerals were in the rock, but they have now been altered and their components used to form hornblende, 'b) The absence of olivine included in the bronzite. Conditions appear to have been particularly favourable for the formation of hornblende, and included olivine would have been converted to hornblende, (c) The predominance of hornblende inclusions and the general texture. There is a textural . Colo\irs of minerals given are from the colour stand- ards of - the Roek-Colour Chart prepared by the National Research Council, Washington. D.C similarity in thin-section between this rock and specimen 39725, in which the bronzite porphyroblasts have abundant inclusions of sieve-textured and euhedral olivine. Serpentinite. — Specimen 39751 was collected from 040120, which is 50 feet south-east of Black Rocks. The outcrop consists of scattered boulders, with smooth weathered surfaces. In hand specimen the rock is greyish green, uni- formly soft and has an uneven fracture. In thin section the specimen has a fine- grained feathery texture with some recognizable mesh-texture Areas up to 3 mm across contain reticulate, or knitted-textured masses of anti- gorite. Antigorite (approx. 90%), occurs as feathery groundmass. in reticulate areas, and as anhedral “grains” up to 0.8 mm in diameter. Much is recrystallized and oriented so that when the nicols are crossed and the gypsum plate in- serted. if the stage is rotated the field appears alternately dominantly orange and blue. Mag- netite (approx. 3%), is present as fine inclusions up to 0.2 mm in antigorite. and as ragged grains up to 1 mm in diameter associated with chlorite. Chlorite occurs as laminae frequently associated with magnetite grains or laminae. The charac- teristic occurrence is in books 0.3 mm by 0.1 mm. Two varieties were recognized, clinochlore (spprox. 3% ) and pennine (approx. 4%). The variation between these two varieties is prob- ably continuous. Clinochlore formed by the reaction of antigorite with aluminous spinel, and positive and negative Pennine resulted from a reaction between relatively aluminous clinoch- lore and antigorite. This rock is a highly serpentinized bronzite peridotite in which the original pseudoporphyri- tic texture is preserved. The reticulate masses of antigorite, 3 mm in diameter, represent earlier porphyroblasts of bronzite. Serpentinite. — Specimen 39734 was collected at 048129, that is, on the immediate north of Black Rocks from an outcrop of scattered boulders. The hand specimen i.s a uniform, fine- grained dark green rock with numerous minute silvery chlorite flakes. Under the microscope the rock appears to consist dominantly of anti- gorite with the characteristic mesh texture, in which are several partly serpentinized horn- blende prisms and also a bronzite porphyroblast altered to structureless serpentine. The chlorite flakes are always associated with magnetite laminae. This rock bears a striking resemblance to specimen 15425 (collected from a dyke just south of Toodyay) described by Prider (1944, p. 128) which was shown from analysis to be a normal serpentinite. Veins having their origin withm the ultrabasic body. — A variety of veins have their origin vdthin the ultrabasic rocks themselves. They are easily distinguished in hand specimen as they are normally subject to differential weather- ing with respect to the enclosing rocks. In thin section however, they have less distinct boun- daries, being composed of altered or recrystal- lized minerals on a planar or distorted planar surface. The chrysotile veins are of particular i.nterest for it is veins of this character which 16 Fig. 3.— Polished specimen (x showing bronzite crystals (pale grey) in a serpentinous groundmass (dark grey). The left half of the specimen consists of normal bronzite peridotite, in the centre-right is the pyroxene- rich vein, to the right of this is the zone of intense serpentlnization, and on the extreme right there is normal bronzite peridotite. A magnetite-clinochlore vein passes from the upper left of the specimen to the lower right of the pyroxene-rich vein, but does not stand out strongly in the photograph. form the commercial deposits of chrysotile asbestos elsewhere. Further, chrysotile veins alone have produced noticeable dilation. Pyroxene vein. — The only outcrop of this type of vein may be seen on the northern vertical surface of the cutcrop referred to as Black Rocks. The vein has a relief of about 1". is about 2" wide and crops out for a length of some 30 feet. Specimen 39727 is 13" by 9" by 6", and one face of it has been polished to reveal the relationship of the vein to the enclosing bron:ite peridotite (see Figure 3). The rock and the pyroxene vein are both traversed by magnetite-clinochlore veins. Inter-relations of the minerals in the vein are essentially the same as in the normal bron- zite peridotite. Large bronzite porphyroblasts comprise about 66% of the vein: olivine in- cluded in the pyroxene and adjacent to the pyroxene has been partly preserved from ser- pentinization. On one side there is a zone of intense serpentlnization. The only feature fundamentally distinguishing the vein from the enclosing rocks is the high proportion of bronz- ite. Magnetite-clinochlore veins. — These are pre- sent in specimen 39727 and are relatively abun- dant throughout the rest of the ultrabasic body. As seen in thin section they consist cf strings of roughly equidimensional composite grains (averaging 0.2 mm in diameter), which are com- posed of alternating lamellae of clinochlore and magnetite. They may transgress bronzite crystals in which case the bronzite immediately adjacent to the vein is altered to clinochloi-e. Magnetite-antigorite veins. — These veins are characteristically weathered to a depth of up to 4 mm deeper than the surrounding locks, giving an entrenched appearance. Under the micro- scope they appear as numerous strings and groups of magnetite granules and occasional composite magnetite-clinochlore grains in a serpentinous material. Examination under high power magnification shows that some of the magnstite consists of dust (0.001 mm in diameter) surrounding antigorite “grains", which are approximately 0.02 mm in diameter. The antigorite is oriented with X normal to the plane of the vein. The veins themselves are traversed by gamma chrysotile veins. Eremite peridotite breccia. — Specimen 39754 was collected at 057290. The zone of brecciation is not more than a few feet wide, and in it the bronzite peridotite is broken into angular frag- ments averaging 40 cm in diameter. The groundmass is composed of antigorite and mag- netite with numerous fine striae parallel to the margins of the fragments. The antigorite is either in large crystals or else in parallel-oriented small crystals since areas up to 4 mm in diameter show essentially the same optical orientation. This brecciation appears to be a late-stage effect due to adjustments within the ultrabasic :mas 3 after its consolidation. Chrysotile veins. — These are typical cross-fibre chrysotile veins and occur in two varieties. The oldest is characteristic of the more highly ser- pentinized rocks and forms swarms of intric- ately distorted thin lenses averaging 0.02 mm in thickness. This variety appears to be more recent than any antigorite-bearing veins or antigorite in the groundmass of the bronzite peridotite. These irregular thin veins are in turn cut by young regular chryseti'e veins which are of uniform sheet-like form from 0.1 mm to 1.0 mm in thickness. These later veins exhibit only miner distortion. Fig. 4.— Zoned chrysotile vein cutting serpentinite. Crocsed nicols. x 15. 17 Alter'ed rocks associated with the pegmatite dykes. — The ultrabasic mass has been intruded by several small pegmatite dykes and addition of material has caused considerable local alteration. At 110010 on the south bank of the creek, some tens of feet from a pegmatite dyke, bronz- ite has been altered and replaced by a fine- grained felted mass of hornblende <25%), clino- chlore (80S ). talc (35%), anthophyllite (10%) and magnetite (trace). The original ground- mass, presumably antigorite. has been mainly replaced by hornblende, which is medium- gi'ained and comprises about 90 'v; of the present groundmass. Thus the pseudoporphyritic tex- ture is preserved even though the mineralogy has been almost entirely changed. Xenoliths of ultrabasic material included in the pegmatite dykes have been very extensively alter- ed, and all former textures destroyed. The aniphi- boles. however, are still those of the ultrabasic body. A typical xenolith is 8 in. in diameter and has a core composed entirely of pure coarse- grained hornblende: clinochlore, magnetite, spinel and the serpentine minerals are conspicu- ous by their absence. Surrounding this core is a rock in which hornblende is still dominant but chlorite, vermiculite or chlorite-vermiculite mix- tures comprise about 30% of the rock. Talc or anthophyllite may be present. The outer shell of the xenolith is either chlorite or vermiculite schist one or two inches thick and is very friable. In one specimen examined alteration had proceeded to a stage at which hornblende had been corroded and replaced by oligoclase. The resulting rock is composed 60% of hornblende, 30% of oligoclase * Allis — An;i7), and 1% of quartz. Talc bodies are found in the vicinity of the pegmatite dykes. These have a core of talc of variable thickness and are separated from the bronzite peridotite by a sheath of anthophyllite needles of the order of 4 cm long, oriented with their c axes normal to the contact. Anthophyl- lite may also occur in the core as feathery masses but is always heavily corroded by talc, which is either pseudomorphous after antho- phyllite 01 ' in irregular masses. Interleaved with the anthophyllite needles are rare flakes of vermiculite and Pennine (cf. specimen 39765). One of the talc bodies had a core consisting cf about 80% of talc with corroded inclusions of chlorite (diabantite) up to 1 mm in diameter, anthophyllite and hornblende being present in trace quantities. Anthophyllite and talc are the products of progressive siliciflcation; chlorite has resulted from alteration in an environment rich in water but deficient in silica, but subsequent addition of silica has led to steatitization of the chlorite. Alteration associated loith the quartz dolerite dyke.— A quartz dolerite dyke about 150 feet wide has been intruded across the south-western part of the area studied. Near this dyke bronzite has been extensively altered to talc, hornblende and anthophyllite. This alteration stai'ted along cracks and cleavages and worked outwards converting adjacent bronzite to amphi- bole. Antigorite is present in the normal ouantities but occurs as a feathery groundmass with associated Pennine and magnetite. Dis- tinct mesh texture is not developed. At 036014 the bronzite peridotite has been completely altered to a very fine-grained mass of intricately intergrown clinochlore and talc. No amphiboles are present at this locality. Xenoliths in the ultrabasic body. — In certain areas there are patches of mixed float of bronzite peridotite and amphibole schist. These amphi- bole schists are thought to be xenoliths, as their general appearance is different from either the country rocks or the ultrabasic rocks. Three .specimens which are essentially quartz-tremo- lite schists have been examined in detail. The texture varies from schistose to granoblastic with poor platy orientation and there may be a poor or a strong linear orientation. The colour is pale green but may be reddish due to oxidation. Tremolite (9%, -100%. ' occurs as subhedral prisms with lamellar twinning parallel to 001. 1.630 — 1.648, apatite and monazite are common inclusions, alteration is to talc. Quartz (trace — 90%) forms equant to elongate grains with inclusions of apatite, monazite and more rarely rutile. The tremolite is similar to the amphiboles of the ultrabasic body and quite unlike the horn- blendes of the basic granulites in the country recks. As all the outcrop is in the form of float it has not been possible to correlate the linea- tions with any structure elsewhere. These rocks have a mineralogical affinity with the ultrabasic rocks but have different textures, the accessory minerals apatite, monazite and rutile, and occur only within the boundaries of the ultrabasic body. Mineralogy Olivine varies from forsterite to chrysolite. Originally this mineral probably constituted over 95%r- of the “magma”, but now. due to sili- eifleation and serpentinization, it is uncommon to find rocks with more than 5% of olivine. d ranges from 1.663 to 1.674 indicating a varia- tion in composition from Fa.7 to Fa. 12 (Polder- vaart 1950 >. Spinel is a rare inclusion. Where the inter-relations can be seen, olivine is idio- morphic towards pyroxene, though it is usually embayed until only sieve-textured relics remain. Bronzite, hornblende and chry.sotile have evi- dently all formed from olivine under appropriate conditions, but usually chrysotile appears to have recrystallized to antigorite. The only pyroxene encountered is bronzite. No 010 sections of clino-pyroxene have been seen, nor have any sections perpendicular to the optic axes given figures similar to those of clinoyproxene. Nevertheless clinopyroxene pre- sent in small amounts would be difficult to detect unless the grains were favourably oriented. Bronzite occurs throughout the ultrabasic body except where it has been serpentinized or inetasomatized to produce hornblende or talc and anthophyllite. Even where complete ser- pentinization has occurred, structureless ►ser- pentine pseudomorphs, or knitted-texture anti- gorite pseudomorphs testify to the earlier presence of pyroxene. Similarly where the rock has been metasomatized, talc, or talc- anthophyllite pseudomorphs after pyroxene 18 clearly indicate its original presence. These pseuaomorpns are equivalent in size to the fresh pyroxenes and are also present in the same pro- portions with respect to the groundmass, as the pyroxenes are in the less altered rocks. Measurement of 2V and 7 has shown that these pyroxenes are aluminous (see Table I) and do not conform with the composition — optical property diagram of Poldervaart (1950). As the percentage of ferrosilite is between 11 and 25, and the Mg:Pe ratio is between 9 and 2.33, all these pyroxenes are bronzite. This nomen- clature is in accordance with Poldervaart ( 1950 ) even though some of these pyroxenes contain less than 70% of enstatite. In thick sections the bronzite is pleochroic with X colourless. Y pale pink and Z pale green. The bronzite porphyroblasts are distinctive not only in thin-section and hand specimen but also in their persistence in shape and size in replaced and altered parts of the ultrabasic body. Their origin and relation to other minerals is of interest. There are three im- portant points in the inter-relations of olivine and bronzite. First, the bronzite is never euhedra! and never show.s any crystal faces; second, the bronzite usually includes sieve- textured or embayed olivine which was prob- ably once euhedral (cf. specimen 39725); third, the margins of the bronzite porphyroblasts are commonly rich in olivine inclusions, in fact some may be so well coated with olivine that the serpentinous groundmass of the rock is only rarely in contact with the pyroxene. The bron- zite porphyroblasts are the result of introduc- tion of silica into an aggregate of olivine crystals. Thus, the porphyroblasts contain relic granules of olivine cemented together by bronzite formed by the silicification process; the persistence of olivine depended upon the availability of silica and the armouring effect of the already-formed bronzite. Svinel is found in all of the less-altered bron- zite peridotites. Grains of ragged magnetite associated with clinochlore in the serpentinites testify to the earlier presence of spinel. The stages — spinel, spinel with magnetite and fine- grained chlorite reaction rims, spinel with mag- netite-clinochlore reaction rims, and finally mag- netite grains with clinochlore— clearly illustrate the progressive “serpentinization" of these grains. The normal amphibole, a near-tremo- lite, is apparently undersaturated with respect TABLE I Estimation of the comvositioii of vyroxenes f from the o^Hlcal data diagram of Winchell and Winchell 1951. p. 406>. •iVy y .\i. AI.O, II W si.n 7 S' 1 r>70 70'\, 1 1 IO"o :U>72n •s7'' 1 • «)7*» (•■s"o 14"„ ls“o :{5)72S >1 1 r*7<> 70‘'o 11“.. I0''„ 7-S 1 ■ ()7<> 7->ti 11“.. I0“o :iS)72.'. .s4 1 <(70 71% 1 (“n 1 :{97:i(5 si 1 • «>7<> 71“., 1 1“.i 1 :{‘»7:(7 ■S.S 1 70% UC',. 1 1">, .s;} 1 ()7:( 71''.. 1 (>“.. :M)7gl !)0 1 -<)7I (!'.'% I-')",. H»7 1 f> 7 s to A1'>0 :h. Regional metamorphism has led to the absorption of magnetite and spinel by horn- blende. Although the previous existence of spinel in these rocks cannot be proved, it was clearly present in the rest of the ultrabasic body to the extent of 1 or 2 per cent. The refractive index varies from 1.778 to 1.786. As the grains are generally less than 0.2 mm in diameter, no measurements of specific gravity have been attempted. The variety is probably ceylonite or pleonaste. The colour is close to moderate yellowish green 10 GY 4/4. with vari- ation in hue and absorption towards light olive 10 Y 5/4. Regional metamorphism has caused the par- tial recrystallization of the ultrabasic rocks to produce hornblende. There is no primary horn- blende. Near the contacts with the country rocks it is abundant, but towards the centre of the mass it is rarer, being of the order of 10% or less. Hornblendites are developed adjacent to the pegmatite at 109010. The optical properties of the hornblende vary over a well-defined range, viz: — ", 1.640-1.655 (mean 1.645); 2Vx 75"-90' (mean 85'^'); Z c 17’ -21° (mean 18 ); d 0.017-0.022 (mean 0.019). These properties vary independently and .so no generalizations can be made regarding the trends or changes in the composition of the hornblende. The composition is evidently near to tremolite with Mg:Fe -- 85:15. In hand speci- men the hornblende is dark green, but in thin section it is almost colourless, except in some rocks rich in hornblende. The strongest colour.s recorded were X very pale yellowish green dO GY 8/2), Y pale yellow-green (10 GY 7/2) and Z pale green (5 G 7/2) with absorption X. Dolerite intrusion (liydrothermal activity ami ( a nietasomatism) Late veining . . . . talc, anthophyllite and clinochlore ; also rodiiigite.H ciirysotile veins Dolerite intrusion was clearly ))ost regional metamorphism persistent chrysotile probably post regioiifl! metamorpliism 23 contact of the dolerite. Alternatively, the' rodingites could have originated in the inter- vening ultrabasic material between the dolei'itc and the gneisses. Garnet rodmgite. — Specimen 39773 was found at 005108 in mixed float consisting of dolerite. pegmatite and weathered bronzite peridotite. No other rocks of this type have been found. The specimen is about 12" by 9" by 5": at one cud it is a uniform dark greenish rock of uneven liardness and from this it grades through a blotchy green, pink and white sugary rock to a dense, white, fine-grained “cherty" rock at the other end- The dark greenish rock is composed o± sub- I'ound gariiets in a chloritic groundmass. The garnet is in discrete colourless grains up to 3 mm in diameter, and in granular masses: it is weakly birefringent and makes up about 50 h of the rock. Flakes of chlorite (pycnochlorite' 'approx. 40%) 0.5 mm by 0.1 mm fill most of the interstices between the grains of garnet. Irregular grains of diopside ^approx. 10%) are distributed throughout the chloritic ground- mass. Chromite is present as rare euhedra oi- irregular translucent brown grains up to 0.1 mm in diameter. The blotchy green, pink and white sugary i-ock is similar but contains only about 20 of chlorite, the pink colour being due to ii'on stains. The “cherty” rock is a fine-grained mass of nearly pure garnet ( grossularite ) . This mass of garnet is traversed by yellow veins of diopside up to 1 mm wide. Also in the cherty rock are rare dark brown grains of allaii- ite. Deep red iron staining is common in radi- ating cracks surrounding the allanite grains. The garnet adjacent to the allanite is noticeably more birefringent than that elsewhere Prehnite rodingite. — Specimen 39774 was col- lected from a stone heap at 005106. It is dense- white. "cherty”, and of uniform hardness. Prehnite with its inclusions constitutes more than 90';^ of the rock and occurs in three ways: first as subhedral crystals, averaging 0.7 mm by 0.4 mm with curved fractures and abundant minute inclusions of high refractive index ; second as clearer equant grains with coarse sutured contacts, which are up to 0.3 mm in diameter, and are arranged in more or less continuous strings; third as apparently form- less masses with mottled interference colours Examination of this material in suitable im- mersion oils revealed fine grains, averaging 0.05 mm in diameter which have sutured contacts with a material of much higher refractive in- dex. This materia] has n 1.70 approx, and is probably zoisite. The remainder of the rock is made up of subhedral diopside prisms averaging about 0.4 mm in length. Clinozoisite-diopside rodingite. — Several bould- ers of the clinozoisite-diopside rodingite where found at the contact of the ultrabasic racks with the country rocks at 015086. Specimen 39775 from this heap has a yellowish white ground- mass enclosing sub-lenticular grey patches up to 3 mm in diameter. The groundmass consists of subhedral prisms of diopside up to 3 mm long. which are altered to clinozoisite and epidote around the margins. The grey patches are crystals of clinozoisite, which are up to 3 mm long, have simple twins on 001 and on 100 and are strongly zoned. Associated with the clino- zoisite are veins composed of fine equant granules of epidote in a chloritic groundmass. The veins are about 1 mm wide and comprise 5S of the rock. Diopside-prehnite rodingite. — Specimen 39776 was collected from the creek bed 100 feet west and 2,300 feet north of the south-west corner of the area. It is mottled grey with black spots and is traversed by numerous sub-parallel veins, up to 5 mm wide, of transversely oriented actinclite. Examination of the thin section with the naked eye and comparison with thin sections of bronzite peridotite shows a remarkable simi- larity in texture. Bronzite porphyroblasts have been replaced by clear prehnite grains with zoisite inclusions and the antigoritic groundmass has been replaced by murky prehnite with abun- dant fine inclusions and anhedral prisms of diopside. The diopside is commonly altered to amphibole or chlorite. Unfortunately, none of the original minerals of the bronzite peridotite remain. The texture however is unmistakeable. This particular rodingite is an altered bronzite peridotite. Petrogenesis The rodingite suite conforms to the chemical and mineralogical reouirements of the rocks de- scribed by Marshall (1911. pp. 31-35), except in the presence of zoisite. The frequent occurrence of zoisite in rocks of this tyoe was recognized by Grange (1927, p. 162). In 1927 Grange reviewed the previous work rn redingites. and concluded that they were formed by the garnetization of gabbros. Grange considered that prehnite and grossularite were secondary after feldspar, but direct evidence that garnet formed after feldspar was not available: serpentinization of diallage was suggested as a ccurce of lime and silica for these reactions. Similar recks have been described by various authors including Miles (1951) and Baker (1957). The association of rodingites with ultrabasic rocks requires that ultrabasic rocks play an es- sential role in their formation. The present author considers that garnetization of gabbros is inadequate tc explain the origin of the rodingites described here. The rodingites from Nunyle show no relic gab- broic textures, but specimen 39776 has a texture derived from the bronzite peridotite, and speci- nien 39773 shows the association cf garnet rodin- gite with a chloritic rock. These recks are the result of alteration of ultrabasic rocks, and not of gabbros. Furthermore the serpentinization of calcic pyroxene cannot be invoked as a source of lime, since lime-rich minerals are not present in the bronzite peridotite. The probable mode of formation of these rodingites therefore appears to be lime and alu- mina metasomatism of bronzite peridotites. Con- sidering serpentine as the original mineral, the 24 constituents of the rodingite suite could be de- rived as follows: — 1. 2H20.3Mg0.2Si02 (serpentine! -I- 3CaO + 4SiO-i ^ 3 (Ca0.Mg0.2SiO2> (diopside> + 2H2O. 2. 2H-20.3Mg0.2Si0-2 (serpentine) + 3CaO + Al^Oa + Si02 SCaO.Al^O.jSSiO^ (grossu- larite) + 2 H 2 O + 3MgO. 3. 2(2H20.3Mg0.2Si02> (serpentine) + 6CaO + 5Si02 + 3 AI 2 O 3 = 3(2Ca0.Al203.3Si02. H 2 O) (prehnite) + H 2 O + 6MgO. 4. 4(Ca0.Mg0.2Si02) (diopside) -h 3Al20.-i + HjO “ 4Ca0.3Al20;}.6Si02.H20 (clinozoisite) + 2 S 1 O 2 + 4MgO. The reactions involve addition of lime, silica, alumina and water, and the removal of water, magnesia and silica. The removed oxides could be represented by the actinolite veins in speci- men 39776. The equations 1, 2 and 4 appear to be petro- graphically sound. Equation 3 is not supported by the petrography and prehnite is more probably derived by means of a two-stage reaction with either intermediate diopside or grossularite. The formation of prehnite by the decomposition of grossularite was recognized by Marshall (1911, p. 34), though it is probable that the prehnite. which is pseudomorphous after bron- zite, was formed in some other manner. The source of the metasomatic fluids was certainly neither the ultrabasic rocks nor the pegmatites but probably the quartz dolerite dykes which contain abundant alumina and lime. However, there are two difficulties con- nected with such a source; first, hydrothermal action clearly connected with the dolerite has given rise to talc and anthophyllite; second, the later stages of dolerite intrusion were sodic rather than calcic. Nevertheless, the present author considers that this rodingite suite is the product of fluids derived from the dolerite dyke acting on the bronzite peridotite. The Country Rocks IniroAuction The country rocks in the area studied con- sist of gneisses, quartzites and basic granulites. There are not granite batholiths in the imme- diate vicinity, though in places the gneisses are intruded by narrow dioritic dykes, which are a few feet wide. Gneisses make up most of the rocks in the district and vary from almost massive to strongly foliated and lineated. Half a mile to the east and west of the ultrabasic body the trends in the gneisses are remarkably uniform and the dip is steep to the east. To the south of the area the trends are disturbed by an east-west trending fault or shear zone along which the northern block appears to have moved east. Outcrops of gneiss are fair on the hill slopes and excellent in the creek beds. The quartzites occupy a north-south strip of country which varies from 300 feet wide to 600 feet depending on the dip and other struc- tural features. They are generally coarse- grained and in the north-west corner of the area, euhedral quartz crystals, up to 6" in length, are common. Foliation and lineation are well-developed in the quartzite except where they are obscured by the very coarse grain-size. No cross-bedding, ripple marks or other primary sedimentary structures have been observed in the area. The quartzite probably had rather impure bands which gave rise to garnetiferous quartzite; this rock has only been seen on talus slopes and its stratigraphic loca- tion within the quartzite is unknown. Lenses or bands of basic granulites occur within the quartzite. Benches and belts of red soil trav- ersing the quartzite indicate that these are common, but proof in the form of solid out- crops is seldom available. Outcrops of quartzite are good, constituting the highest features in the area. The basic granulites appear as lenses and boudins in the gneiss and quartzite, and are generally of the order of 10 to 20 feet wide. Outcrops are only fair except in the creek beds, whez’e boudin structures are well-developed and original beds may be traced as strings of discrete boudins. It is probable that the basic granu- lites occur everywhere as boudins and lenses, and no continuous bands remain, but the gen- eral paucity of outcrop of this rock type makes this uncertain. Foliation is not well-developed but lineation is evident on close examination, in hand specimen. The basic granulites have been studied in detail as they are the best metamorphic facies indicators in the area. Metamorphic Facies Recognition of the grade of metamorphism is generally dependant on the presence of a diag- nostic mineral assemblage and at Nunyle such assemblages are rare, so that no conclusion has been reached regarding the exact meta- morphic gi'ade of the country rocks at the time the bronzite peridotite was intruded. In the absence of diagnostic assemblages, a study of all the assemblages in the country rocks has indicated a definite trend in the metamorphism of the area. Seven mineral assemblages occur in the coun- try rocks: 1 . Hypersthene-andesine-hornblende. 2. Diopside-andesine-hornblende. 3. Microcline-oligoclase-biotite-quartz. 4. Quartz-muscovite. 5. Oligoclase-hornblende-biotite. 6. Chlorite-muscovite-epidote-actinolite. 7. Saussurite-hornblende-quartz. Of these, the first four evidently represent the metamorphic facies prior to the intrusion of the bronzite peridotite. Granoblastic textures, with little or no sign of corrosion occur in rocks con- taining these assemblages. Assemblages 3 and 4 have no diagnostic value, but 1 and 2 indicate high amphibolite or granulite facies. Abundant greenish yellow hornblende may be due to P- or (OH)- in the rock inhibiting the develop- ment of more than one type of pyroxene and so masking the exact facies assemblage. Assemblage 5 has resulted from down-grading of 1 and 2. The hornblende is green contain- ing small blebs of quartz and is characteristic- ally coarser in grain-size than the greenish yellow hornblende in assemblages 1 and 2. 25 Assemblage 6 is of the lowest metamorphic grade and occurs only in a shear zone, which passes through 118063. The temperature had fallen before the shearing, so that the rocks within the zone have recorded the intensified retrogressive metamorphism. Assemblage 7 has resulted from saussuritiza- tion adjacent to the dolerite dyke at 012233. A similar assemblage of hornblende-saussurite- chlorite is partly developed at 194063. The highest metamorphic grade was attained some time before the intrusion of the bronzite peridotite, when all the basic rocks contained pyroxene. The temperature fell prior to the bronzite peridotite intrusion so that at the time of the intrusion, where there was differential movement and possibly the introduction of OH-. pyroxene-andesine-hornblende recrystallized to oligoclase-hornblende (which may represent low amphibolite facies). Shearing took place at still lower temperatures producing low grade rocks in I'estricted loci. Minor subsequent alteration has been caused throughout all the country rocks by fluids which accompanied the dolerite intrusion. Petrogenesis All of these rocks have been completely re- crystallized, and. with the exception of bedding, no primary structures are preserved. The bulk composition has probably not altered greatly, except in the addition of alkalies to the gneiss. The gneisses and quartzite are lineated, bedded and appear to be conformable. The origin of the basic granulites has been clearly demonstrated by Prider (1944, p. 121 ), who described and analysed rocks similar to those at Nunyle. Originally these were tholei- itic flows or sills which recrystallized to form pyroxene-andesine-hornblende granulites, some of which were later altered to oligoclase-horn- blende granulites. During this alteration there was no significant change in the bulk compo- sition. At Nunyle the alteration to oligoclase-horn- blende granulites has only occurred in critical localities where there have been structural ad- justments with attendant rock flowage. Summary and Economic Considerations The rocks at Nunyle consist of gneisses, quartzites and interbedded pyroxene-andesine- hornblende granulites. similar to those of the Jimperding “Series” described elsewhere. They have been developed by metamorphism accom- panied by regional folding, of the original sand- stones, sandy claystones and tholeiitic sills or flows. After the main period of folding the bronzite peridotite was injected. During the intrusion, gneiss and basic granulites were mobilized in certain critical localities, with the result that some of the pyroxene-andesine-hornblende granulites were converted to oligoclase-horn- blende granulites. At the same time there was further folding of the quartzite. Intrusion of pegmatite dykes followed the con- solidation of the bronzite peridotite. Emana- tions from these produced talc, anthophyllite- and clinochlore in the ultrabasic mass. The rocks were subsequently intruded by dolerite dykes, which caused epidotization in the country rocks and minor hydrothermal altera- tion in the ultrabasic mass, and were probably also responsible for the formation of the rodingites. The high Mg/Fe ratio and lack of associated gabbro place the Nunyle bronzite peridotite in the class of alpine type serpentinites. Alpine type serpentinites rarely carry metallic mineral- ization but may be favourable loci for chrysotile asbestos deposits. As only moderate thermal metamorphism will cause alteration of chryso- tile to antigorite Archaean deposits are not likely to have survived. It would be interesting to know if the late chrysotile veins have formed as a result of continued serpentinization of high temperature silicates or by solution of antigorite and redeposition of chrysotile. In the past by far the greatest production of chrysotile has come from rocks of post-Archaean age but there are significant producing mines situated within the shield areas. The ultrabasic mass at Nunyle is small and contains a very low percentage of chrysotile. Other serpentin- ites of similar age will certainly be found in the district as geological mapping proceeds and it is possible that one of these may prove to be of economic interest. Acknowledgments This work was undertaken at the University of Western Australia as part of the require- ments for the Honours degree in geology and the author wishes to thank Professor R. T. Prider and Dr. A. P. Wilson for their super- vision and assistance. Other members of the Department of Geology. University of Western Australia gave valuable advice and encouragement. In particular the author would like to thank Mr. G. Playford for assistance with the surveying, and for his criti- cism, encouragement and advice; Mr. J. G. Kay for his criticism of the petrology; and Mr, I. M. Threadgold for useful discussions on the origin of serpentinites and rodingites. Special thanks are due to Mr. F. L. Billing for prepar- ing the photographs. The author deeply appreciates the kind hos- pitality of Mr. and Mrs. J. S. McKay of “The Range”. Toodyay. References Bilker. G. (1957). Nickeliferous serpcntiiiite and rodiim- ite from Anderson’s Creek Area, near Beaconsfleld. Northern Tasmania. C.S.I.R.O. (Mineragraphic Investigations) Rep. No. 682. Barshad, I. (1948). — Vermiculite and its relation to bio- tite as revealed by base exchange reactions, X-ray analyses, differential thermal curves and water content. Ainer. MUi. 33: 655-678. Bowen. N. L. and Tuttle, O. F. (1949).— The system MgO — SlOa — H 2 O. Bull. Geol. Soc. Amer. 60: 439-460. Clarke. E. de C, (1930). — The Pre-Cambrian succes.sion in some parts of Western Australia. Rcv- Aust. Ass. Advance. Sex. 20: 155-192. Dana, E. S. (1954). — “A Textbook of Mineralogy." 4th Ed. (Wiley: New York.) 26 Geological Map of NUNYLE 960 Western Austral ia LEGEND GNEISS QUARTZITE BASIC GRANULITE ULTRABASIC ROCK XENOLITHS IN THE ULTRABASIC BODY OOLERITE PEGMATITE EPIDOTE IMPREGNATIONS Al 111 vlum EPIOOTE-FILLEO SHEARS ESTABLISHED GEOLOGICAL BOUNDARY APPROXIMATE GEOLOGICAL BOUNDARY INFERRED GEOLOGICAL BOUNDARY DIP AND STRIKE OF FOLIATION IN THE COUNTRY ROCKS DIP AND STRIKE OF PLATY FLOW STRUCTURE DIP AND STRIKE OF JOINT SURFACE TREND AND PLUNGE OF LINEATION CONTOUR LINE (feetobovc seo level) Qua rt zitic Quortzite '90 BLACK ^OCKS 2 0 . ■/y' ToluS uartzite Talus Alluvium Alluvium CP.£ /9se Francis, G. H. (1956). — The serpentinite mass in Glen Urquhart, Inverness-shire, Scotland. Avier. J. Sci. 254: 201-226. Grange, L. I. (1927). — On the “rodingite” of Nelson. Trans. N.Z. Inst. 58: 160-166. Hess, H. H. (1938). A primary peridotite magma. Amer. J. Sci. 35: 321-344. Hey, M. H. (1954). — A new review of the chlorites. Miner. Mag. 30: 277-292. Johnstone, M. H. (1952). — The geology of the Hamersley Siding Area. J. Roy. Soc. W. Aust. 36: 45-75. Marshall, P. in Bell, J. M., Clarke, E. de C. and Marshall. P. (1911). — The geology of the Dun Mountain Subdivision, Nelson. Bull. Geol. Surv. N.Z. 12: 31-35. McWhae. J. R. H. (1948). — The geology and physiography of the Lawnswood area. J. Roy. Soc. W. Aust. 32: 49-74. Miles, K. R. (1951). — Garnetized gabbros from the Eula- minna district, Mt. Margaret Goldfield. Bull. Geol. Surv. W. Aust. 103. Nagy, B. (1953). — The textural pattern of the serpen- tines. Econ. Geol. 48: 591-597. Nagy. B. and Faust, G. T. (1956). — Serpentines: natural mixtures of chrysotile and antigorite. Amer. Min. 41: 817-838. National Research Council, Washington, D.C. (1948). — Reck Colour Chart. Poldervaart, A. (1950). — Correlation of physical properties and chemical composition in the plagioclase. olivine, and ortho-pyroxene series. Amer. Min. 35: 1067-1079. Prider, R. T. (1944). — The geology and petrology of part of the Toodyay district. Western Australia. J. Roy. Soc. W. Aust. 28: 83-137. Rabbitt, J. C. (1948). — A new study of the anthophyllite series. Amer. Min. 33: 263-323. Tertsch, H. (1921). — Studien am Westrande des Dunkel- steiner Granulitmassivs. Miner, petrogr. Mitt. 35: 177-214. Turner, F. J. and Verhoogen. J. (1951). — “Igneous and Metamorphic Petrology.” (McGraw-Hill: New York.) Wilkinson, J. F. G. (1953). — Some aspects of the alpine- type serpentinites of Queensland. Geol. Mag. 90.- 305-321. Winchell. A. N. and Winchell. H. (1951). — “Elements of Optical Mineralogy.” Part II. 4th Ed. (Wiley: New York.) Yoder, H. S., Jr. (1952).— The MgO — Al20;t — Si02 — H 2 O system and the related metamorphic facies. Amer. J. Sci. Bowen Vol.: 569-627. 27 5. — Homing Behaviour in the Quokka, Setonix brachyurus (Quoy and Gaimard) ( Marsupialia ) By W. C. Packer* Manuscript received — 16th October, 1962 Quokkas collected in the settlement area of Rottnest Island exhibit homing behaviour fol- lowing translocation over distances of up to 1.75 miles, the maximum distance tested. Within familiar territory they are able to orient them- selves rapidly and behave appropriately in rela- tion to the physical character of the habitat but behave inappropriately in unknown areas. It is not clear, however, whether homing involves true orientation and navigation or a random- search technique. The area within the settle- ment in which any individual animal was re- captured was extremely variable in size but in most cases probably did not represent the com- plete home range. Introduction Dunnet (1962) and other investigators (un- published data ) suggest that the quokka ^Setonix ’brachyurus) is generally rather re- stricted in its movements, although this may vary considerably from population to population. Dunnet found that the animals have relatively restricted individual ranges on the eastern end of Rottnest Island but have been recaptured over distances of up to 2,000 yards from the original site of marking. He suggests, however, that those on the western end of the island are more restricted in their range (no more than 300 yards on the basis of limited data). Following from these observations, the ques- tion arose as to whether or not the quokka possessed the ability to return to its home area if translocated from it. Associated with this is the question of the size of the individual ranges. Methods and Materials Rottnest Island. 12 miles west of Perth, Western Australia, supports a large population of quokkas. There is a tourist resort at the eastern end of the island, and with the excep- tion of one series (see below) all of the animals were taken from this settlement area. Although not typical quokka habitat in any sense, this area was chosen because the animals there are accustomed to people and can be approached closely. This allowed one to identify marked animals without the need for physically handling them subsequent to marking. Each animal was marked in two ways. They were fitted with plastic collars which bore in- dividually recognizable symbols of reflective tape (see Ealey and Dunnet 1956) which were easily distinguishable with the naked eye at distances of 10 to 20 feet. A numbered operculum tag (see Dunnet 1956) covered with red reflection tape, was fastened to the pinna of one ear. This greatly reduced the time spent in locating Zoology Department. University of Western Australia, Nedlands, Western Australia. marked animals since this marker was easily visible in the beam of a head-torch up to 100 feet and stood out clearly when the animal was facing directly toward or away from the observer at which times the collar was obscured. In January and February, 1961 (mid-summer) four series of animals were collected in the southern end of the settlement and marked. The release points for these series. Bungalow No. 1 at the northern most end of the settle- ment, Herschell Lake, Lake Bagdad and Para- keet Swamp, are shown in Figure 1 and numbered, respectively 1 to 4. At the same time two additional series were marked but not trans- located. One of these came from the southern end of the settlement and served as the control and the other was captured at the fresh-water soak on the north shore of Herschell Lake. The settlement area was searched for marked animals about three times a week for six weeks and then at irregular intervals until the middle of May. All recaptures were visual sightings only, and all were recorded in order to obtain a measure of the individual ranges. During January and February, environmental conditions are quite severe for the quokka. Over most of the island, fresh water is available only from soaks along the edges of the lakes. The lakes themselves are highly saline with a chlorinity of approximately 70 parts per thous- and in Bagdad and Herschell (Hodgkin 1959). Fig. 1. — Map of eastern half of Rottnest Island. Western Australia, showing release points (1-4) of translocated quokkas. 28 Most of the vegetation is dry also. The settle- ment area at this time is an “oasis” where leaking taps and friendly tourists keep the animals supplied with water and food. As a result, the animals here are never as poor physiologically as their “non-civilized” cousins (unpublished data). In the event that the settlement was an exceptionally attractive area during the summer, so producing atypical homing results, a .small series of animals was translocated from the settlement to the Lake Bagdad site in August, 1961 'mid-winter). At this time of the year fresh water is widely available, the vegetation is green and tourists are few. The control series consisted of animals marked the previous sum- mer and seen in the settlement on the two evenings when the translocated series was collected. Because of other commitments on the mainland, the settlement was searched for marked animals only at irregular intervals over the next two months. Results The number of animals marked in each series and the number and per cent, recaptured one or more times in the settlement area are shown in Table I. The per cent, of summer animals returning to the settlement decreased as distance of translocation increased, but there is no signi- ficant difference between the three closest points on the basis of a chi-square test. A significant decrease in the number making the return is found, however, in the group released at Para- keet Swamp when compared to the other sites istatii-i- from captma* site Nmnher marked Ueea]itiire 1 or more times in settlement \ri *'• “utJinu'r n-d;} mile 14 12 Hf) • 7 ll.TM-lioll r. () t>7 rtiile 2;{ to S2 • i'y I>. Uaudad 1 • 2:') mile 10 7«-0 I'arakt'f! Suanip 1 ■ 7a mill* 25 10 40 0 roiifrols .42 40 04 *K Herschell controls 20 1 .5-0 Winter L. Ha^rdail 1 •2;-> mile 21 10 47-(> Controls Ht 00-0 series was ever seen in the settlement. Prior to this recapture this individual was observed in an intermediate location between Herschell Lake and the settlement but was never recap- tured subsequently. Of the other 19 animals in the series. 12 were never recaptured except along the soak at which they had been collected originally; two were recaptured at Herschell and on the golf course w'hich in part separates the lake from the settlement; and two were re- captured only on the golf course. Three were never recaptured. The release points for the translocated series were checked on the two nights following release and at irregular intervals after that. No marked animals were seen except on the first night. Traverses by road over the north-east corner of the island were made as often as possible, but no marked animals were seen except at the rubbish dump ( point D on Pig. 1 ) which at that time supported a large population of quokkas. These included one from Herschell, two from Bagdad and four from Parakeet. These animals were not seen on every occasion, but they were never recaptured subsequently in the settlement or elsewhere. There was no significant difference between the sexes in the translocated series as to whether they homed or not 'x" 0.85. 1 d.f.), and there is no evidence to suggest that age was a factor either, although no animals less than 18-20 months old were tested. The number of animals returning from Bagdad during the winter was significantly less than in the summer (x^ 3.95, with 1 d.f.), but whether this is because of more favourable environmental conditions during the winter is not clear. Except for the few summer animals that apparently established new ranges around the rubbish dump, none of the translocated animals who failed to return to the settlement were seen. It is not known v;hsther they set up new ranges elsewhere or if they even survived. A total of 419 recaptures of marked animals was made, comprising 176 of controls and 243 of translocated animals. The frequency of re- capture ranged between 1 and 14 with a medium of 5.4. Although the area in which any one animal was recaptured probably does not repre- sent its home range (see Discussion), neverthe- less, certain measurements can be made utilizing the locations of recapture, i.e., area and proximity to marking site of subsequent recap- ture. Area: — The site of original capture and of all recaptures for each animal was plotted, and the area enclosed within a polygon, formed by con- necting the most peripheral of these points, was calculated. The size of the area thus obtained was extremely variable and ranged from approxi- mately 3.900 to 178,000 square feet with a median of 26,300 square feet. The values were not ncrmally distributed about the mean (c.41,700 square feet) but were skewed toward the low- side. The relationship between area and number of recaptures was also extremely variable and ranged from 10 recaptures within an area of 5,400 square feet to three recaptures within 151,500 square feet. The minimum number of 29 Figure 3 shows the frequencies of recapture at various distances for the first and for the closest recapture. The distribution of the dis- tances of first recapture 'Fig. 3 A and C) are somewhat different from each other, while those for the closest point . DISTANCE IN ?EET DISTANCE IN FEET Fig. 2. — Distance of all recaptures from the site of ' marking of (A) summer translocated animals and (B) ^ ^ non-translocated summer animals; grouped for various < distance classes. i On 3 further comparison is worthy of note. The area in which the winter controls were recaptured overlapped the area in which these same animals were found during the summer in eight of the nine cases; and in the one which did not overlap, the two areas were separated only by about 100 feet. This would suggest that there is no change in location, at least in this portion of the home range, between summer and winter. The winter areas were recaptures required to establish the size of the range could not be determined from the data. Proximity of recapture: — No true measure can be made of the precision of the return because of a lack of knowledge of the nature of the home range and of how to interpret the position of the original capture site. Nevertheless, it is interesting to note how close subsequent recap- tures were made to the point of marking. Figure 2 shows the frequency distribution of the distance between the site of original mark- ing and all subsequent recaptures for the observations, animals from the southern part of the settlement could range as far as the two closest release points. All that can be said here on this point is that no control animal from the settlement was ever seen at either of these two points, and only one animal marked at Herschell Lake was ever seen in the settlement. It is unfortunate that this latter individual was never recaptured again. Although homing in this species may not in- volve navigation, a series of unplanned observa- tions in the course of this woi’k clearly suggests that the quokka is able to orient itself rapidly within its home area and to respond appro- priately. The fresh water soak at which animals were collected for the non-translocated series at Herschell runs for about 200 yards along the edge of the lake. During the summer the level of the lake is reduced, and the shore is 20 to 50 feet wide. Behind this rises a wall of Pleisto- cene dune limestone which shows a bench and notch configuration resulting from fluctuation of sea -level during geologically recent times: this is figured by Teichert (1950) for one of the other lakes. The animals move freely over this cliff on pathways. The animals in this non-translocated series were collected, placed in bags, taken to the settlement, marked and then returned to the shore of the lake and released. Upon release most of them appeared to be in an excited state and rapidly moved off up or down the beach or up the cliff-face. On the other hand, some of the translocated animals released at Hers- chell, although handled in the same way. responded quite differently. When released on the shore, they too moved off rapidly; but if they happened to be heading toward the water, they just kept going, regardless. They could be followed in the water m the beam of a head- torch for at least 100 feet. This behaviour was never observed in the control series: and the quokka has never been observed to enter the salt lakes except very rarely when chased by man. although they frequently wade in one or two inches of water around the edges of the fresh-water swamps during the winter. The morning following release, three dead animals mot included in Table I> were found washed up on the shore. Arising from these observations, an attempt was made to see if the quokka could and would orient within a restricted area. For this pur- pose an octagonal enclosure 150 feet in diameter was built. The terrain was such that the only visible landmark was the lighthouse which is situated on the highest point in the centre of the island. Single animals were released by remote control in the centre of the area and were captured in traps situated at the eight angles. All tests were run on clear, moonless nights and involved animals captured in differ- ent directions from the enclosure and at various distances up to two miles. As can be seen in Figure 4a, the animals were captured at random around the enclosure in relation to the home direction. In a series of preliminary experi- ments, the animals responded negatively in relation to the observer, but after modification of the apparatus this no longer occurred (Figure 4b). Although the number of animals tested was relatively small, there is nothing to suggest that under the conditions of the experiment addi- tional animals would have changed the results. Animals taken from one-quarter mile away were no better in their orientation that those from the settlement two miles away. This is some- what unexpected, since the near-by individuals were from an area from which animals are known to move to the site of the enclosure. Those animals might be expected to be familiar with the area and. therefore, to return to the home area by the most direct route. While these results do not support an hypothesis in- volving true orientation and navigation as a factor in homing, it is probably too early to say that the mechanism involves only random- search. Different techniques will be required to differentiate the two. HOME OBSERVER Pig. 4.— Position of recapture within an enclosure in relation to (a) the home area and (h) the observer. 31 An interpretation of the individual areas within the settlement in which recaptures were made is difficult to make. The method used in determining the size of these areas was essentially the minimum area method used by several authors (reviewed by Brown, 1962), although in the present work the points repre- sent visual sightings, not trap captures; and. therefore, the matter of boundary strips does not arise. The area certainly does not correspond to the home range, at least as that term is applied by Burt (1943). Many of the smaller areas, par- ticularly those in which ten or more recaptures were made, are open, grassy plots providing no diurnal shelter. In these cases the animals probably spend the day either in the near-by scrub or else under a building from which they emerge at night to occupy a small range in which they feed and possibly mate. On the basis of the present data, it is suggested that this is the general situation for this species, although the size of the area will vary with the individual. Excursions of various lengths with various frequencies may occur, perhaps in- fluenced at the present site by the activities of tourists. A discussion of the relationship be- tween excursions and the home range concept is beyond the scope of this paper. Acknowledgments The author is grateful to Mr. D. Sullivan, Managing Secretary, Rottnest Island Board, for the valuable assistance provided by him and his staff. The work was financed by a research grant from the University of Western Australia, and by a grant to Professor H. Waring from the C.S.I.R.O. for marsupial research. References Brown. L. E. (1962). — Home range in small mammal communities pp. 131-179. In "Survey of Biological Progress". Vol. I (Ed. B. Glass) (Academic Press: N.Y. and London.) Burt. W. H. (1943). — Territoriality and home range concepts as applied to mammals. J. Mammal 24: 346-352. Dunnet. G. M. (1956). — A population study of the quokka, Setonix brachyurus Quoy and Gaimard (Marsupialia) . I. Techniques for trapping and marking. C.S.I.R.O. Wildl. Res 1(2) : 73-78. (1962). — A population study of the quokka. Setonix brachyurus (Quoy and Gaimard) (Marsupialia). II. Habitat, movements, breeding, and growth. C.S.I.R.O. Wildl. Res. 7(1): 13-32. Ealey, E. H. M. and Dunnet, G. M. (1956). — Plastic collars with patterns of reflective tape for marking nocturnal mammals. C.S.I.R.O. Wildl. Res. 1(1) : 59-62. Hodgkin. E. P. (1959). — The salt lakes of Rottnest Island. J. Roy. Soc. W. Aust. 42: 84-85. Teichert, C. (1950). — Late Quaternary changes of sea- level at Rottnest Island. Western Australia. Proc. Roy. Soc. Viet. 59: 63-79. 32 INSTRUCTIONS TO AUTHORS Papers may be submitted to the Society in accordance with Rules and Regulations 38 to 41 inclusive (see below). They should be addressed to The Honorary Secretary, Royal Society of Western Australia, Western Australian Museum, Perth. Authors are solely responsible for the factual accuracy and for any opinion expressed in their papers. They are particularly requested to verify references. Alterations to MSS. submitted to the printer will be allowed only under excep- tional circumstances, and no changes will be permitted after galley-proof stage. In the preparation of MSS. authors are required to follow the C.S.I.R.O., Guide to Authors (C.S.I.R.O., Melbourne, 1953), except that papers longer than 10,000 words (30 foolscap pages of pica type, with 6 inch lines 40 to the page) will not normally be accepted. Authors may be required to meet half the cost of preparation of the blocks of diagrams and illustrations. Authors shall receive a total of 30 reprints free of charge. Further reprints may be ordered at cost, provided that such orders are submitted with the MS. RULES AND REGULATIONS 38. Every paper intended to be read before the Society or to be published in the Society’s Journal must be sent to the Secretaries at least seven days before the date of the next ensuing Council meeting, to be laid before the Council. It will be the duty of the Council to decide whether such contribution shall be accepted, and if so, whether it shall be read in full, in abstract, or taken as read. All papers accepted for publication must be read or otherwise communicated at an ordinary meeting prior to publication. 39. A Publications Committee, appointed by the Council, shall recommend to the Council whether a paper presented to the Society shall be published in the Society’s Journal. The Publications Committee may obtain an opinion from any person it may select on the suitability of any paper for publication. 40. Publication in the Society’s Journal shall only be available to (a) Ordinary Members, (h) Honorary Members, (c) Non-members resident outside Western Australia, who must communicate the paper through an Ordinary or Honorary Member. No paper shall be accepted from a Non-member resident in Western Australia. 41. The original copy of every paper accepted for publication by the Society, with its illustrations, shall become the property of the Society, unless stipulation is made to the contrary, and authors shall not be at liberty to publish their communicated papers elsewhere prior to their appearance in the publications of the Society unless permission for so doing is given by the Society, or unless the Society fails to publish the paper in the Journal of the year in which it is read or otherwise communicated, or of the succeeding year. Journal of the Royal Society of Western Australia, Inc. Volume 46 1963 Part 1 Contents 1. — Description of a New Freshwater Fish of the Family Theraponidae from Western Australia. By G. F. Mees. 2. — Copper Poisoning in Sheep in Western Australia. By A. B. Beck and H. W. Bennetts. 3. — Two New Western Australian Cockroaches. By K. Princis. 4. — Bronzite Peridotite and Associated Metamorphic Rocks at Nunyle, Western Australia. By C. R. Elkington. 5. — Homing Behaviour in the Quokka, Setonix hrachyurus (Quoy and Gaimard) (Marsupialia). By W. C. Packer. Editor; J. E. Glover Assistant Editors: G. F. Mees, R. D. Royce Annual Subscription : Forty Shillings The Royal Society of Western Australia, Inc., Western Australian Museum „ Perth 66922/1/63—535 ALEX. B. DAVIES, Government Printer, Western Australia JOURNAL OF THE ROYAL SOCIETY OF WESTERN AUSTRALIA INCORPORATED VOLUME 46 (1963) PART 2 PUBLISHED 1ST AUGUST, 1963 REGISTERED AT THE G.P.O., PERTH FOR TRANSMISSION BY POST AS A PERIODICAL THE ROYAL SOCIETY OP WESTERN AUSTRALIA INCORPORATED COUNCIL 1962-1963 President Past President Vice-Presidents Joint Hon. Secretaries Hon. Treasurer Hon. Librarian Hon. Editor W. D. L. Ride, M.A., D.Phil. J. E. Glover, B.Sc, Ph.D. C. F. H. Jenkins, M.A. W. R. Wallace, Dip.For. Margaret E. Redman, B.Sc. C. V. Malcolm, B.Sc. (Agric.). R. P. J. Pillow, Dip. P.T.C. Ariadna Neumann, B.A. J. E. Glover, B.Sc., Ph.D. N. H. Brittan, B.Sc., Ph.D. R. W. George, B.Sc., Ph.D. J. G. Kay, B.Sc. R. J. Little. D. Merrilees, B.Sc. R. T. Prider, B.Sc., Ph.D., M.Aust.I.M.M., F.G.S. R. D. Royce, B.Sc. (Agric.). L. W. Samuel. B.Sc., Ph.D., F.R.I.C., F.R.A.C.I Journal of the Royal Society of Western Australia Vol. 46 Part 2 6. — Studies in the Diagenesis of Some Western Australian Sedimentary Rocks Presidential Address, 1962 By J. E. Glover, B.Sc., Ph.D.* Delivered — IStli July, 1962. Textures arising from diagenesis in some Western Australian sedimentary rocks are des- cribed, and particular emphasis is placed on determination of the order in which the dia- genetic minerals have been developed. Rocks studied in detail include specimens from the Clark Sandstone, Septimus Limestone, Cockatoo Sandstone, Birdrong Formation. Arrowsmith Sandstone, Enokurra Sandstone and Mokadine Formation, and many others are described briefly. As recommended by Gilbert and Turner (1949) the universal stage has been used to clarify re- lationships between adjacent authigenic minerals, and to obtain optical and morphological data. It has enabled fairly detailed investigation of some optical properties of K-feldspars in the Clark Sandstone, Mokadine Formation and Arrow- smith Sandstone. The stage is particularly useful for the study of textures in which relatively hard and soft minerals are adjacent, for example quartz and calcite or feldspar and barite. Such mineral pairs are commonly separated by faces of one of the minerals, and these faces, which can be recognized and determined rather easily with the stage, are often otherwise overlooked. An attempt has been made to summarize the main causes of diagenesis, and to categorize re- sultant textures according to their origin. Some textures do not conclusively demonstrate the order of formation of the authigenic minerals, but many do, and the latter are almost as useful petrologlcally in sedimentary rocks as standard or well-known textures are in igneous and meta- morphic rocks. The diagenetic textures are divided into (a) enlargement textures, which in- clude simple enlargement textures, indentation textures and enclosure textures, (b) pressure- solution textures, (c) micro-drusy textures, which include simple and composite micro-drusy tex- tures, (d) reorganization textures and, (e) re- placement textures. Reference is made to much of the literature on sedimentary petrology in Western Australia. Table of Contents Page Introduction 33 Use of the Term Diagenesis 34 Aims of the Investigation 34 The Universal Stage in Sedimentary Petrology .... 34 General Advantages .... .... .... .... 34 Detection of Crystal Faces .... .... .... 35 Examination of Replacement Textures .... 35 Diagenetic ^Minerals and Textures .... 35 Feldspathic and Arkosic Sandstones 35 Clark Sandstone .... .... .... .... 35 * Department of Geology, University of Western Aus- tralia, Nedlands, Western Australia. Page Mokadine Formation .... .... .... 36 Arrowsmith Sandstone .... .... .... 37 Enokurra Sandstone .... .... .... .... 38 Quartz Sandstones .... .... .... .... 38 Calcareous and Dolomitic Rocks .... .... 39 Clastic Limestones .... .... .... .... 40 Dolomitic Quartz Sandstone .... .... 40 Septimus Limestone .... .... .... .... 40 Miscellaneous Silicified Limestones .... .... 41 De-silicated Rooks .... .... .... .... 41 Miscellaneous Diagenetic Minerals and Textures 42 Glauconite .... .... .... .... .... 42 Phosphate Minerals .... .... .... .... 42 Gearksutite .... .... .... .... ... 42 Tourmaline .... .... .... .... .... 42 Concretions .... .... .... .... .... 43 Kaolinite .... .... .... .... .... 43 Allophanoids .... .... .... .... 43 Outstanding Causes of Diagenesis .... .... .... 43 Reactions Near the Sediment-water Interface .... 43 C'ompaction .... .... .... .... . .. .... 43 Intrastratal Solution and Precipitation . .. 44 Solution .... .... .... .... .... 44 Precipitation .... .... .... .... ... 46 The Carbonate-silica Relationship .... .... 48 The Dolomite Problem .... .... .... .... 50 Common Diagenetie Textures and their Interpretation 52 Enlargement Textures .... .... .... .... 52 Simple Enlargement Textures .... .... 52 Indentation Textures .... .... .... 52 Enclosure Textures .... .... .... .... 52 Pressure-solution Textures .... .... .... 52 Pore-filling or Micro-drusy Textures .... .... 54 Simple Micro-drusy Textures .... .... 54 Composite Micro-drusy Textures .... ... 54 Reorganisation Textures .... ... 54 Replacement Textures .... .... .... . .. 54 Conclusions .... .... .... .... .... .... 54 Acknowledgments .... .... ... .... .... 54 References .... .... .... .... .... .... 54 Introduction The study of sedimentary rocks in thin sec- tion has received increasing attention during the last fifteen or twenty years. Contrary to earlier practice, general petrographic texts now tend to allot sedimentary rocks as much space as igneous and metamorphic rocks, and several 33 new texts devoted entirely to the petrology of sedimentary rocks have appeared. This is not to say that sedimentary petrology has reached the position of igneous and metamorphic pet- rology, in which perhaps most advances in our understanding of petrogenesis come from attempts to duplicate in the laboratory con- ditions under which the rocks and their con- stituent minerals form. Thus, although increas- ing attention is being paid to synthesis of minerals common in sediments, and to their response to various physico-chemical conditions, there is still scope for much petrogenetic inference from microscopic studies of the naturally occurring sedimentary rocks.'-' One of the most rewarding avenues for such studies is in the investigation of the minerals and textures arising from diagenesis. A sur- prisingly complex diagenetic sequence may be unravelled by detailed examination of a rock whose petrogenesis is at first apparently simple. Use of the Term Diagenesis Diagenesis has been variously defined. Ac- cording to Pettijohn (1957. p. 648) the term refers primarily to reactions which take place within a sediment between one mineral and another, or between one or several minerals and interstitial or supernatant fluids. Most authors agree that diagenesis is a low-temperature phenomenon that grades with increase in tem- perature, into metamorphism. In this paper diagenesis will be divided into early and late diagenesis, though no attempt will be made to define their limits precisely. Siever’s discussion of early diagenesis seems the most satisfactory. Siever (1962, p. 140) con- siders early diagenesis to include that stage in the history of a sediment during which it is buried only up to a few tens of feet. It is bac- terially active, is not greatly compacted, has high porosity and water content, may be mod- erately permeable, and is subject to the upward passage of waters from compacting sediments below. Changes attributed to late diagenesis may per- sist until weathering of the exposed rock: in fact the processes of diagenesis and weathering, though distinguished by many authors, must be gradational. There is much petrographic evidence for the reality of the distinction between early and late diagenesis, but two examples will suffice here. Calcareous concretions that preserve fishes (see Weeks 1957) are obviously of early origin. On the other hand Topkaya (1950) describes Jurassic conglomerates from Arvel in Switzer- land which are free of authigenic silicates despite the fact that they have been partly de- rived from Triassic formations now exception- ally rich in them.. This is evidence of late diage- nesis. for the authigenic minerals could have formed only after the erosion of the Triassic rocks. • It is strange that sedimentary petrology should have lagged behind igneous and metamorphic petrology. The first rock to be examined in thin section with a microscope and polarized light was a sediment fSorby 1851). and Sorby later published widely on sedimentary petrology, maintaining an interest for many years in such contemporary sedimentological problems as dolomitizatlon and the formation ot aragonite (Sorby 1904). One of the main problems has been to decide wlu'u diagencsis becomes metamorphism. It has been suggested by Packham and Crook (1960, p. 404) that retention of the original clastic fabric of the rock would indicate that it had been subjected only to diagenesis, whereas extensive modification, involving substitution of hornfelsic or schistose fabrics, w'ould demon- strate metamorphism. There is no difficulty in distinguishing diagenesis from metamorphism in most of the rocks discussed below. Some of them, as indicated later, have been intruded by dolerite. Apart from that, none has been metamorphosed in the sense of having been heated by intrusions, strongly folded or gener- ally sheared; nor are any adjacent to later igneous bodies that might have metasomatized them. Further details about their age. location and tectonic setting can be obtained from Mc- Whae et al. (1958). Aims of the Investigation The purpose of this paper is to show how texture may demonstrate the order in which diagenetic processes have taken place, and how it can reveal some of the causes of those pro- cesses. This has been done before, notably by Gilbert (1949) and Gilbert and Turner (1949), and comprehensive reviews of the literature on diagenesis have been given by Pettijohn (1957) and Carozzi (I960). Although the techniques used here are similar to those of Gilbert and Turner, there are several reasons, stated im- mediately below, why presentation of the re- sults of this study is warranted. Firstly, refer- ence is made to all known papers dealing with diagenesis in Western Australian rocks. Secondly petrographic descriptions of some of the sedimentary rocks treated here have not previously appeared in the rather sparse West- ern Australian literature. Thirdly, some of the textures are different from those described else- where. Finally, an attempt is made to group the various diagenetic textures with similar origins into categories, so that their petrologic interpretation is thereby facilitated. The Universal Stage in Sedimentary Petrology General Advantages Gilbert and Turner point out that the uni- versal stage is particularly valuable in the microscopic investigation of sedimentary rocks, and state that it is essential in a modern sedi- mentary laboratory. They use the stage to de- termine the optical properties of the minerals, and the crystalline faces developed, and to ob- serve the three-dimensional interrelationships of the minerals. Faces are identified tentatively by noting their relationship to optical directions within the crystal, and their identity is con- firmed by measuring angles between the faces and comparing them with published data, not- ably those in Dana (1899). Gilbert and Turner note, as was found in the present investiga- tion. that appropriate tilt cn the stage reveals crystalline faces quite unsuspected following routine microscopic examination of authigenic minerals. They recommend the Federow pro- cedure in which all measured directions and 34 planes are plcttid on a stereographic or equal- area projection net. Details of procedures are clearly set forth by them, and will not be re- peated here. Detection of Crystal Faces Where the boundary between two authigcnic minerals is planar, the plane is commonly a face cf one of the minerals. The fact that a boundary is nlanar can only be seen if the plane is essentially parallel to the microscope axis; if it is oblique the boundary does not appear sharp and is represented by a narrow band in which the two minerals, perhaps with very different optical properties, overlap. Moreover, unless both minerals have a similar resistance to abra- sion. irregularities due to grinding can cause an inclined planar boundary to appear uneven. This effect, which is common with such mineral pairs as quartz and barite, quartz and calcite, and quartz and dolomite, is illustrated in Fig. 1. and Plate I, Figs. 1 and 2. The only reliable way to establish the planar nature of boun- daries between such minerals and hence to dis- cover crystal faces is by tilting until the plane is parallel to the microscope axis. Examination of Replacement Textures Where carbonate cement completely encloses quartz grains it commonly penetrates their margins and partly replaces the quartz. On the other hand, the margins of well-rounded grains that have undergone no replacement may show apparent irregularities for reasons indi- cated in the previous paragraph and these ir- regularities may suggest marginal replacement. Appropriate tilt, revealing smooth, rounded margins devoid of re-entrants, prevents mis- interpretation. Diagenetic Minerals and Textures Textures resulting from diagenesis are con- sidered below, the constituent minerals are de- scribed, and the order of their formation is dis- cussed. Cei’tain diagenetic textures are com- monly associated with particular lithologies, and some textures are virtually restricted to one lithologic type. For this reason, the nature of the textures is illustrated by describing the mineral relationships commonly encountered in some of the main types of sedimentary I'ock. Discussion of the ultimate cause of diagenetic growth and solution of minerals is however left to a later section of this paper. Fig. 1.— Diagrammatic sketch of unmounted thin sec- tion of quartz-barite rock. Quartz ( stippled i and barite (cleaved) meet in a plane. However, grinding has lowered the level of the softer barite with respect to quartz, and has plucked out cleavages, forming rough valleys. The quartz-barite boundary, viewed from above, appears highly uneven. The planar nature of the contact will be evident only if the section is tilted so that the plane is parallel, or almost parallel, to the microscope axis. Feldspathic aiid Arkosic Sa7idst:.nes Clark Sandstone. — The Clark Sandstone (Traves 1955) of the Carlton Basin, in north- eastern Western Australia, is a highly fossilifer- cus, reddish or green, glauconitic, Cambrian sandstone. It is faulted but not strongly folded, and it is not intruded by later igneous rocks. The preservation of abundant brachiopods and trilobites, and much unaltered glauconite, shows that the remarkable outgrowths on quartz and feldspar observed in this study must be ascribed to diagenesis rather than metamorphism. The specimen examined or microcline (ex- cept for one unusual grain with a composite core of adularia and microcline). Cleavage can commonly be seen passing without break or change of direction, from both types of core to the authigenic rim, but any twinning in the core is sharply truncated at its boundary with the rim (Plate I. Figs. 5 and 6). The 001, 010 and 110 cleavages and faces were identified by measuring the angles between their normals and X, Y and Z. and comparing them with Niki- tin’s data (see Gilbert and Turner 1949, Table IV). These measurements, which were made with the universal stage, also served to con- firm the identity of the core, for they enable distinction between untwinned microcline and adularia. The authigenic rims have remarkable optical properties, but due to their narrowness, numer- ous inclusions and natchy extinction, precise • Numbers refer to the General Collection of the De- partment of Geology. University of Western Aus- tralia. C0571— (2) 35 measurements are possible only on selected grains. Patchy extinction is not accompanied by obvious bending- or distortion of cleavage. The angles between the principal optical direc- Uons and the normal to 010 and 001 correspond approximately with those quoted by Nikitin for orthoclase. Ihis correspondence can clearly not be everywhere precise in rims with patchy ex- tinction, for the patchiness is caused by varia- tions of up to 6° in the principal optical direc- tions. Thus Z commonly departs by several de- grees from the normal to 010 and the rims are not everywhere monoclinic. Moreover, there is generally a variation of 2V within the one rim. The maximum variation observed within one rim is from 11° to 45.5°, and the overall range is from IT to 64° (see Table D. This range of 2V in K-feldspar would thus include the min- erals, as sometimes defined, adularia (2Va 50° - 70°T orthoclase (2Va 25° - 50' ). and sanidine t2Va 0° - 25°. optic plane i 010) I see Chaisson (1950> for the optical classifica- tion used here ! . No grains whose optical plane is parallel to 010 (i.e. “high” sanidine) were observed. All of the authigenic feldspar described above came from one hand specimen 2 in. x 2 in. x 1 in., and the considerable range in 2V between different grains, and within individual grains, is puzzling. Variations in the optical properties of pctassic feldspars are said to depend upon (1) compositional differences, (2) submicro- scopic intergrowths due to unmixing, and (3> degree of Si/Al order (see Hewlett 1959). Neither the composition of intrastratal solu- tions at any given time, which influences the composition of the feldspar precipitated, nor the cooling history of the feldspar, w^hich affects the degree of unmixing, and of Si/Al order, can have varied much. This suggests the influ- ence of ether, highly localized factors. Differ- ences between the clastic cores, which are seed crystals, could cause precipitation of slightly different material on each. Perhaps slight com- positional differences within one core could cause or even accentuate, differences in material precipitated in different parts of the one rim. The shape and volume of the intergranular pores may have affected the rate of passage of solu- tions in a significant way. This, however, is speculation: no reason can confidently be ad- vanced. at present, for the variations in 2V be- tween different authigenic rims, and within in- dividual rims. No explanation is offered, either, for the remarkably low values of some readings (see Table I) . Mokadine Formaticn . — The Mokadine Forma- tion, of probable Proterozoic age, crops out on the eastern margin of the Perth Basin near Moora. It is a sequence of arkose, feldspathic sandstone, siltstone, and claystone that has been intruded by dolerite dykes. Authigenic growths cn detrital quartz grains are mentioned by Logan and Chase (1961) in their brief descrip- tion of the petrography of the formation. The specimen described in this investigation (No. 36909, collected from just above the type sec- tion of the underlying Dalaroo Siltstone) is a brown, strongly lithified, moderately well-sorted, fine-to-medium-grained arkose and it does not •appear to have been altered by dolerite intru- sions, the nearest exposed dyke being .l-mile dis- tant. It has the following composition (by volume): quartz 65%, K-feldspar 25%, opaque iron-ore grains, largely changed to haematite, 8%, and interstitial haematite and limonite, 2%. The quartz includes a few fine-grained quartzite grains. The K-feldspar includes microcline and adularia (2Va down to 58°), and ranges from clear fresh grains to grains that have been com- pletely converted to grey and yellow-brown clay minerals. The interstitial iron oxide is present mainly as a film that outlines the clastic grains, which range from sub-angular to very well rounded, TABLE I Variations in the optic axial angles of authi- genic feldspar in the Clark Sandstone. Values greater than 25° are reproducible to within 7°. and values less than 25° are reproducible to ivithin 2°. Clastu* Curf' 'iVa of aiitliiireiiie rim Micr(K-liuf‘ 17 Microrline ('IVa ^ ■SO'') 47 , 37.i- Adularia (2\7,r) 43.r. 4U Microcline 2.or Almost all quartz and many K-feldspar grains have been authigenically enlarged. Secondary quartz is abundant, forming about 5% of the rock, and secondary K-feldspar makes up per- haps 0.5%. The following points are signifi- cant: 1. A few clastic quartz grains penetrate each other to give minor and poorly developed niicro- stylolitic intergrowths. (Quartz grains commonly penetrate feldspar grains by as much as 0.025 mm without development of micro- stylolitic boundaries, and the feldspar has apparently undergone almost all the solution in such pairs (see Plate II, Fig. D. Much less frequently a fairly sharp-cornered feldspar grain slightly penetrates the flattish surface of a quartz grain, showing that initial shape and perhaps crystallographic orientation influence the final relationship of the two minerals. Feld- spar grains are also penetrated by each other. No clastic grains have been penetrated by authigenic outgrowths, so compaction preceded authigenesis, which was a late diagenetic feature. The volume of secondary material is far greater than that available from local solution by clastic grains at their point of con- tact. and most of it must have come fi’om elsewhere. 2. Boundaries between quartz outgrowths are either irregular, or roughly planar, but the planes generally do not seem to correspond to common rational faces of either grain. Out- growths of quartz and feldspar meet in boun- daries that occasionally correspond to the 001 face of the feldspar, but are more often approxi- mate planes that are not rational faces of either grain. These relationships may mean that times 36 of growth of the minerals overlapped (see Gil- bert and Turner 1949. p. 13). 3. Each feldspar growth is kaolinized to about the same degree as its clastic core. This shows that most kaolinization occurred after formation of the authigenic rims, so that the clastic feldspars were practically unaltered when deposited. Kaolinization has apparently lowered the optic axial angle of some grains, for some readings on clastic microcline cores are as low as 72°, about 10° lower than those normally quoted for the mineral. As the rims are also altered, their range in 2Va from 68° to 77 gives no reliable indication of their nature. 4. Many authigenic rims on microcline grains have uneven extinction, apparently a vague con- tinuation of the crosshatching (albite and peri- cline twins) of the core. It has been suggested that crosshatching indicates inversion from an earlier, high-temperature monoclinic form < Laves 1950), and the fact that authigenic (low- temperature) microcline described by Baskin 0 956) does not show crosshatching, seems to accord with this view. Baskin’s grains show no clastic cores and have a type of fourling twin, and their unusual morphology may well result from cold formation. On the other hand the poorly developed extension of crosshatching from clastic cores to rims in microcline of the Mokadine Formation does not necessarily mean that they grew at high temperature. These rims seem to have their structure determined by that of the core rather than by temperature of for- mation, for they are crystallographically con- tinuous with the ad.1acent part of the core. The same effect is observed with strained quartz grains, whose rims show wavy extinction con- tinuous with that of the adjacent quartz, and with plagioclase. the twinning of w'hich extends from the core into the rim (see Plate II, Fig 4). The vague extension of crosshatching noticed here therefore probably has no significance in terms of temperature. However, if the type of crosshatching just described is not an indication of high temperature formation, it is strange that it is not more common. It is not known whether authigenic quartz and feldspar replaced an earlier mineral cement, or wdiether they grew in voids between the grains. Logan and Chase (1961) found specks of undigested calcite in the authigenic quartz cement of arkoses in the Mokadine Formation examined by them, but no calcite could be detected in this rock. The only incorporated materiels are the iron oxide coatings outlining the clastic grains. Arrowsmiih Sandst07ie . — The Arrowsmith Sandstone is an unfossiliferous, lower Palaeo- zoic or Proterozoic formation exposed in the north-eastern part of the Perth Basin near Arrino. All specimens so far examined by the author are well-sorted, medium-to coarse- grained arkoses, and many of them contain almost as much lithic material as feldspar, and therefore they nearly grade into lithic sand- stones. The lithic fragments include sericite- quartz schist, composite quartz-feldspar grains, and myrmekite. but the most abundant are vol- canic fragments. A previous account of the petrography (Glover 19605) drew attention to authigenic quartz and chlorite, but re-examina- tion with the universal stage has revealed a complex diagenetic sequence. Specimen Pf2* is an arkose with the following composition (by volume): quartz 40%, feldspar 27%, volcanic fragments 13%, other lithic frag- ments 13%, authigenic cement 6%, other minerals 1%. The grains are angular to well rounded, and there is slight pressure-solution on some boundaries of adjacent grains, causing penetration of one by the other, generally with- out discernible micro-stylolitic intergrowth. Almost all clastic grains are covered by a thin brown film that generally appears as a very faint stain on the grain surface. This film is only 0.002 mm thick in cross section, and is composed of one, or in places, two or three layers or flakes of a brown mineral with fairly high relief. The mineral is pleochroic from red-brown to yellow, with absorption greatest when the flakes are elongated parallel to the polarizer, but it is too thin for other optical pro- perties to be measured or even clearly observed. The mineral seems most closely allied to brown mica, and it may represent the transformation product of a thin argillaceous coating on the sand grains. It has disappeared locally where pressure-solution has caused slight interpenetra- tion of clastic grains. Other authigenic minerals beside the brown, mica-like mineral, are quartz, pale green almost isotropic chlorite, and feldspar. The feldspar is untwinned, with relief less than quartz, and usually has patchy extinction. 2Va ranges be- tw'een 43“ and 55 ' with an average of 50 ' (10 grains) and typical measurements of the angles between X, Y and Z and the pole of the main cleavage are as follows (the corres- ponding angles between 001 and X, Y and Z for orthoclase, as given by Nikitin, are quoted in brackets) : X .\ 1 cleavage 81.5° (85°). Y A 1 cleavage 9° (5°). Z A 1 cleavage 88.5° (90°). The feldspar is close to the orthoclase-adularia boundary, according to the classification of Chaisson (1950). It is uncertain whether the differences between the optical measurements given above and those of orthoclase taken from Nikitin are due to experimental errors arising from the smallness of the grains and their un- even extinction, or whether there is a slight departure from monoclinic optics. The authigenic minerals listed occupy the spaces between clastic grains, and examples of the resultant diagenetic textures are illustrated below. Plate II. Fig. 2, shows one of the simplest arrangements. The minerals lining the void are a brown, mica-like mineral on the surfaces of the clastic grains, chlorite which has a serrated fringe toward the centre of the void as though it had grown inward, and feldspar in the centre. This texture would not arise from replacement of an earlier cement, but is consistent with pre- cipitation in an empty pore space. As with similar drusy textures, which form by precipita- tion from fluids in cavities, the first-formed mineral must be the outermost, and the last- * Specimen number assigned by West Australian Petrol- eum Pty., Ltd. 37 formed the innermost. In the texture illustrated m Plate II. Pig. 2, the brown, mica-like mineral formed first, the chlorite second, and the feld- spar last. Some of the intergranular micro- drusy textures are more complex, and there have been minor variations in the order of formation of the minerals: nevertheless, the basic pattern is the same, and the brown mica- like mineral is always the first, with chlorite either second or third. In some places there is a similar texture in which quartz is substituted for feldspar. In the texture illustrated in Plate II. Fig. 3. quartz forms both before chlorite and after it. The most complex texture is shown in Fig. 6b. where quartz forms before chlorite and both quartz and feldspar form after it. The last-formed quartz has perfectly developed prism faces and. as the mineral was growing into a void, it must have formed before the feldspar. The order of development in this texture was therefore brown mica-like mineral, (2) quartz, (3) chlorite, (4) quartz. (5» feldspar. Elsewhere in the Arrowsmith Sandstone ■similar textures, but with development of slightly different minerals, are encountered. The composition of the chlorite changes, and its colour and birefringence vary perceptibly. The texture shown in Plate II, Fig. 4, is notable, for it shows how albite twinning of plagloclase I determined by the curves set forth by Turner (1947) as Am] persists into outgrowths. The original clastic grain is outlined by the brown micaceous mineral. Any reasonable conjecture regarding the origin of the diagenetic textures in the Arrow- smith Sandstone must seek to account for the presence together of up to four minerals in the spaces between the clastic grains, and should explain their sequence and arrangement. The following hypothesis appears to satisfy these conditions. A well-sorted sand, with most of the mud winnowed out to deeper water, is assumed to have compacted to a porous sand- stone. Slight pressure-solution causing minor penetration of clastic grains accompanied this compaction, or was brought about by later earth movement. The thin film adhering to the grains was dissolved at their points of contact: it is not clear when this mineral, probably originally argillaceous, was changed to the brown mica- like mineral that now coats the grains. In any case, in one form or another, it must have pre- ceded the diagenetic chlorite, quartz and feld- spar, which were precipitated from percolating v.'aters. These waters, which varied in com- position both in time and place, were magnesian, potassic, Eodic and siliceous, and all the solutes would have been available within the formation itself, as it contains volcanic fragments, micro- cline. orthoclase, sodic plagioclase and quartz. Precipitation could have taken place at any time after compaction of the sediment in the late Proterozoic or early Palaeozoic. It is con- ceivable that equilibrium between the fluid and surrounding rock was established several f.mes, only to be disturbed by flowage of the intersti- tial fluids caused by earth movements. If so, processes leading to attainment of the final texture might have persisted even for millions ot years. It might indeed be speculated whether future techniques will allow absolute dating of authigenic minerals in such textures. The dates obtained could well apply to tectonic events, such as folding or uplift, tnat brought about the disequilibrium mentioned above. Enokurra Sandstone . — The Enokurra Sand- stone is an unfossiliferous Proterozoic or lower Palaeozoic formation exposed in the north- eastern part of the Perth Basin, near Yanda- nooka. It consists of fine- to very coarse- grained sandstone which is locally conglomera- tic. and shows well-developed cross-bedding. It has been described previously by McWhae et al. (1958). Glover (1960b) and Bastian (1961). The specimen investigated here (No. 38725. from 3 miles northeast of Yandanooka) is a coarse-grained arkose with the following ap- proximate composition (by volume) : quartz 67%. feldspar (microcline and oligoclase) 15%, lithic fragments (mainly gneiss) 14%, other minerals (muscovite, garnet, opaque grains, clay-sized material. (?) limonite, haematite) 4%. The clastic grains are outlined by a patchy coating of haematite. Cementation has been effected partly by muscovite and patches of sericitic to clay-sized material, and partly by secondary quartz. The secondary quartz fills many of the spaces between the clastic grains, and is usually in optical continuity with one or more of the visible quartz grains. In places, outgrowths from several grains meet in more or less planar surfaces, but none have so far been identified as rational crystal faces. Also present in these areas of authigenic quartz is a fringe of colourless to very pale-green sericite (2Va 38% adhering to the brown, iron-stained clastic grain surfaces. The fringe, which is made up of minute flakes at right angles to the grain surfaces, is commonly difficult to observe, even under crossed nicols, for the interference colours of the minute flakes are hard to dis- tinguish from those of the surrounding authi- genic quartz (see Plate II. Fig 5). Tlie texture here is basically the same as that observed in the Arrowsmith Sandstone, and could have arisen only by precipitation in voids between grains. A few of the clastic grains penetrate each other, and some of the contacts are micro-stylolitic. Even if quartz dissolved at the points of contact of these grains were de- posited nearby in voids, in the manner envisaged by Waldschmidt (1941), there would have been insufficient available from that source to account for all the secondary quartz in the rock. Moreover, some other source for the sericite would be necessary. Precipitation from potassic and siliceous solutions percolating through the porous rocks was the probable cause of this diagenetic texture. Quartz Sandstones One of the most common types of diagenesis is the enlai'gement of quartz grains, especially in v/ell-sorted quartz sandstones such as some from the Birdrong Formation. Cockatoo Sand- stone and Wogatti Sandstone. Diagenetic enlargement of quartz is also common in felds- pathic sandstones from the Clark Sandstone, Mokadine Formation, Arrowsmith Sandstone, and Enokurra Sandstone, discussed earlier in this paper. Petrologic descriptions of sand- stones in the Billeranga Beds, Wenmillia For- mation and Moora Group that contain abundant secondary quartz have been given by Arriens and Lalor (1959). Ranford and Shaw (I960) and Logan and Chase (1961) respectively. More petrographic work will undoubtedly confirm the widespread nature of authigenic quartz in unmetamorphosed Western Australian sand- stones: for example recent petrographic descrip- tions of sediments from the Canning Basin by Johnson and Dallwitz iin Veevers and Wells 1961) mention it in the Noonkanbah Formation, Grant Formation, Godfrey Beds and Kidson Beds. Solution of quartz is also an important process in sedimentary rocks, but in sandstones it is most usually found where there is an abun- dant calcareous, ferruginous or argillaceous matrix or cement. The dissolved quartz is almost always replaced by the cementing material. However, solution processes and their resultant textures are considered elsewhere (p. 44 et seq.). The sandstone illustrated in Fig. 2 shows the type of texture that results from enlargement of quartz grains in a well-sorted quartz sand- stone, where the growth does not fill the pore spaces between grains. This sandstone is from the Cretaceous Birdrong Formation in the west- ern part of the Carnarvon Basin. The original rounded shapes of many grains have been out- lined, apparently by a coating of the same argillaceous material that now partly occupies the pores. Some of the grains are in direct contact with each other, but there seems to have been no pressure-solution, and the secondary quartz must come from another source. Bound- aides between secondary quartz outgrowths in adjacent grains are planar, and all measured planes are prisms or rhombohedra of one or the other of the grains. Relationships between the grains are worth studying, for the light they may throw on the process of crystallization in a porous sandstone. For example, in Fig. 2 the contacts between grain A and grains D and B are a rhombohedron and a prism respectively of grain A, indicating that growth on these faces had been completed before the quartz of grains D and B grew out to meet them. Similarly, grain D finished growing before grain C along their common boundary, and grain C in turn finished before grain B. Thus the sequence, beginning with the first to complete its growth, is A, D, C, and B. The position in this sequence of grain E is unknown, except that it finished crystalliza- tion before adjacent parts of grain C. The .sequence inferred above does not imply that individual grains accomplished all their growth one after the other, during separate and dis- tinct intervals of time. Their growth almost certainly overlapped, as the rock was sufficiently porous to allow fluids to percolate widely. An example of quartz sandstone in which silicification has almost completelv filled spaces between the grains is specimen 48624 from the Devonian Cockatoo Sandstone of the Bonaparte Gulf Basin. The formation is made up mainly cf cross-bedded sandstones, and soecimen 48624 comes from Cockatoo Springs, about 65 miles Fig. 2. — Authigenic enlargement of quartz in sandstone from the Birdrong Formation (Rough Range Bore No. 1. core 7, 3.633-3.636 ft). The stippled areas are partly filled with clay-sized material. Quartz crystal faces are indicated by p (prism) and r (rhombohedron). The sketch is slightly idealized to show faces clearly. Width of field 0.6 mm. east-south-east of Wyndham. It is made up of rounded quartz grains with a haematite coating (85%), authigenic cement (11%) and other minerals including microcline and adularia with authigenic outgrowths, quartzite fragments, green tourmaline, muscovite, and kaolinized composite grains. Numerous planar boundaries can be identified as common faces of quartz, but other boundaries are not strictly planar; of the latter some almost correspond to rational faces, but many cannot be identified. Some ap- parently irregular or curved faces appear to be made up of many very small planes. This tex- ture is apparently that which would have devel- oped in sandstones of the Birdrong Formation, had silification proceeded to the point where most outgrowths interfered with each other. The tendency for the development of some rational crystallographic faces, despite mutual interfer- ence, has also been observed in sandstones studied elsewhere (Basumallick 1962). Most authigenic outgrowths on the rare K- feldspars have developed faces, among which 001, 010, 110 and 130 have been identified. Some of the feldspar outgrowths abut the clas- tic cores of the quartz grains, and have been moulded by them. This is the same as the tex- ture illustrated in Fig. 5e. and is proof that the authigenic feldspar preceded the authigenic quartz in this rock. Calcareous and Dolomitic Rocks Many clastic limestones are mineralogically and texturally simple. On the other hand, num- erous calcareous and dolomitic rocks, ranging from calcareous or dolomitic quartz sandstones to limestone, contain complex suites of dia- genetic minerals, of which the most usual are pyrite, glauconite, quartz, opal, chalcedony, albite, orthoclase, microcline, calcite and dolo- mite. The order of formation of these minerals may often be established by their textural re- 39 lationships. Some textures, and thus some cor- responding sequences of formation, appear to be more usual than others. The reason for the common presence of authigenic siliceous min- erals in carbonate rocks has been debated for a long time, but satisfactory explanations are only now being advanced. Textures from West- ern Australian calcareous and dolomitic rocks will now be considered. Clastic Limestones. — Our understanding of the genesis of clastic limestones has increased greatly in the last ten years, particularly with the research into the formation of Recent lime- stones on the Bahama Banks by Illing <1954), and later w^orkers. The classification of Polk *1959) incorporates much of the new informa- tion on the significance of textures in these rocks. The only published petrologic work on Western Australian calcarenites (Glover 1955) deals in introductory fashion with Devonian rocks from the Lennard Shelf, Canning Basin, but was done before the results of Illing’s work were known. Devonian clastic limestones of the Canning Basin, because of their abundance, strong outcrop, and lack of metamorphism, would be especially suitable for the type of de- tailed quantitative petrographic investigation carried out by Stauffer (1962). Moreover, they are associated with well-exposed reefs and in places provide excellent examples of dolomitiza- tion for study. Two clastic limestones are illustrated here to show textures arising from partial cementation of a calcarenite from the Pleistocene Coastal Limestone at Shark Bay (Plate III, Fig. D and complete cementation of an oolite from the Devonian of the Lennard Shelf (Plate III, Pig. 2). Both textures are formed by precipitation of calcite from solutions in voids, and can readily be distinguished from textures due to re- crystallization of a fine matrix (Pig. 6e). Dolomitic Quartz Sandstone. — Specimen 43799 is from the Palaeozoic sequence in the Bona- parte Gulf Basin and is a fine-grained, silici- fied, dolomitic quartz sandstone containing quartz ( 88% ) , dolomite <10% ) , pyrite ( 2% ) . and traces of other minerals (zircon, tourma- line. black iron-ore, and microcline). The Fig. 3.— Indentation texture in dolomitic quartz sand- stone (No. 43799). Dolomite is partly surrounded by authigenic quartz. The clastic core of the quartz grain is clearly shown. This texture does not reveal whether the dolomite or quartz grew first. Crossed nicols. Width of field 0.25 mm. Fig. 4. — Indentation texture in dolomitic quartz sand- stone (No. 43799) showing special case where dolomite IS partly surrounded by authigenic quartz, but has been moulded onto the clastic core. Dolomite grew before the authigenic quartz. Crossed nicols. Width of field 0.25 mm. clastic quartz grains are fairly well sorted and generally well rounded, and the sandstone is undoubtedly a mature, multicycle sediment. There are three authigenic minerals: pyrite, dolomite and quartz. Pyrite is present mainly as irregularly shaped grains with scattered cubic and octahedral forms, and as rare, minute spherical grains. Small pyrite grains are in- cluded in some dolomite rhombs and secondary quartz, and pyritic films partly outline the cores of some of the secondarily enlarged quartz grains. Pyrite therefore preceded both minerals. Many small dolomite rhombs are completely enclosed by secondary quartz, but it is notable that dolomite never penetrates the detrital quartz cores. Where enclosed dolomite impinges against detrital quartz, the shape of the contact is governed entirely by the shape of the original quartz grain (see Figs 4, 5e and 5f). The re- lationships conclusively show that dolomite grew before authigenic quartz. Therefore the order of formation of the diagenetic minerals is: pyrite first, dolomite second and quartz third. A remarkable feature of some dolomite boundaries is that they are parallel to the margins of clastic grains, but are separated from them by a thin film of authigenic quartz. This shows that some dolomite grains have been slightly replaced by the last-formed mineral, quartz. Set-)timus Limestone.— Textmes, that are fairly widespread in Western Australian Palaeozoic limestones are found in a calcareous sandy dolomitic rock from the Carboniferous Septimus Limestone in the Bonaparte Gulf Basin. The rock studied (No. 36879. from the foot of Mt. Septimus on the northern side) is slightly weathered and is made up mainly of sparry calcite (6%). quartz (40%), and dolomite (53%). The dolomite has a striking appearance as it contains large, brown, opaque limonitfc zones which represent the light-grey zones of clay-sized impurities found in unweathered rocks of the formation (Plate III, Fig. 3>. Brachiopod and crinoid fragments are abun- dant elsewhere in the formation but are sparse in this specimen. Many of the quartz grains were initially well rounded but are now angular from enlargement. The environment of deposi- tion seems to have been neritic, and there are three diagenetic changes of importance; namely formation of the zoned dolomite, formation of the sparry calcite, and growth of the quartz. Zoned dolomite crystals have been recorded elsewhere by numerous workers, including Gil- bert im Williams. Turner and Gilbert 1954, pp. 350-351), Pettijohn (1957, pp. 422-3) and Taft (1961). The last-named author describes dolo- mite, probably forming at present near the sediment-water interface in Florida, which con- tains dark cores apparently of organic matter. It does not therefore seem necessary to postu- late two periods of growth to account for the dark and clear zones in dolomite of the Sep- timus Limestone, for by analogy with the Florida dolomite the mineral is most likely to have formed in a marl or impui'e calcilutite matrix, the impurities of which were arranged to form the dark zones. The patches of sparry calcite certainly represent a recrystallized mat- rix for they are too large to have been pre- cipitated in pores between clastic crystals, and furthermore, they do not have the micro-drusy texture characteristic of calcite precipitated in pores. Non-calcareous material must therefore have been expelled during recrystallization. There is other evidence to support the view that the dark zones of the dolomite are impurities derived from a matrix which has since recrys- tallized and expelled non-calcareous material. Dolomite crystals that replace clay-sized mat- rices rarely have a preferred orientation, whereas dolomite that replaces calcite crystals, such as crinoid fragments, generally adopts the same crystallographic orientation as the replaced calcite crystal. In this rock numbers of dif- ferently oriented, zoned dolomite crystals are found in the one sparry calcite crystal, indicat- ing replacement of a fine matrix before its recrystallization (see Plate III, Fig. 3). Dolomite formed before the quartz out- growths. for although it is commonly partly surrounded by secondary quartz, it never pene- trates clastic quartz grains when it abuts them. However, the relationship between the dolomite and quartz is more complex than indicated, for the secondary quaitz has partly replaced some dolomite, leaving irregular boundaries between them, but has not attacked other dolomite, so that their boundaries are planar. All planar boundaries between secondary quartz and dolo- mite are dolomite faces. On the other hand, all planes developed between quartz and calcite are quartz faces. This suggests, but does not prove, that the secondary quartz grew after ihe dolomite but before transformation of the matrix to sparry calcite. It is worth noting that where adjacent calcite and dolomite are replaced by the same quartz outgrowth, the quartz commonly penetrates more deeply into the calcite than into the clear marginal parts of the dolomite crystal. This. and the fact that some dolomite is enclosed but not replaced by quartz, indicates that it was less prone to replacement by quartz than the calcitic matrix. The diagenetic history of the rock, as tenta- tively deduced from its textures, can now be summarized. The original sediment is believed to have been a shelly and sandy impure calcare- ous mud (or marl.) The first mineral to form was dolomite, which partly replaced the matrix, perhaps before significant consolidation, and arranged included non-calcareous material in zones. Then the quartz grains were enlarged, the secondary quartz partly surrounding dolomite crystals and partly replacing them, but evi- dently replacing the adjacent calcitic matrix m.ore easily. Finally, perhaps because of the pressure of compaction, the remaining matrix was converted to coarse crystalline calcite, and impurities were dissolved and squeezed out. Miscellaneous Silicified Limestones . — A detailed description of a partly silicified limestone, the Sakmarian Woolaga Limestone Member of the Holmwood Shale, in the Irwin River Basin, has been given by G. Playford (1958, pp. 53-62). Playford describes a diagenetic sequence that includes intrastratal solution (producing vugs), reorganization of calcite, and finally, partial replacement by chalcedony. The best known sedimentary sequence in Western Australia whose lithology is ascribed to silicification of limestone is that of the Coomberdale Chert, a formation over 3,300 feet thick within the Moora Group. The chert has been described by Logan and Chase (1961) as a novaculite made up of microcrystalline quartz, with only rare chalcedony. They cite consider- able evidence to support their theory of re- placement, including the observed transition of dolomitic limestone to chert, the presence of siliceous fossils and ooliths believed to have been originally calcareous, relict carbonate inclu- sions in the chert, siliceous rhombs apparently after dolomite, and quartz overgrowths on detrital quartz in some clastic members. Some of the original limestone is silicified only along joints, showing that the process occurred after llthification. Many American and European limestones and dolomites yield, on solution, residues of fine- grained, euhedral authigenic quartz and feldspar, generally without recognizable clastic cores. Mere cataloguing of the literature on these minerals, which were the first to be recognized as authigenic. would unduly lengthen this paper, but most early references are in Boswell (1933>. Authigenic quartz, feldspar and tourmaline have been recorded from dolomite in the Gap Creek Formation (Glover 1955, p. 3) but generally there has been little attempt at extraction of such residues from other Western Australian limestones. De-silicated Rocks The replacement of carbonate minerals by various forms of silica is not a one-way process, and it is probably at least as usual for car- bonate to replace silica. Penetration of rounded quartz grains by the calcareous matrix of sand- stones is an unmistakeable and rather common indication of this type of replacement (Plate III, Fig. 4) and it is not unusual for the corroding calcite to form deep re-entrants or even to split the quartz grain in two. In Western Aus- tralia, clear examples of these replacement textures are found in parts of the Pleistocene Ccastal Limestone I see Glover in Ride et al. (1962, p. 231; and Hodgson hi Hodgson et al. <1962)1. Hodgson has also described rocks from the Cheyne Bay area in which quartz grains are corroded by opal, limonite and goethite, and he suggests that these minerals replaced a calcilutite matrix that was initially responsible for the corrosion. Although the formation of iron oxide cements may generally be a function of weathering rather than diagenesis, if it causes quartz to dissolve and be available for later precipitation perhaps at considerable depth, it clearly has diagenetic significance. Many fer- ruginous sandstones in Western Australia con- tain corroded quartz grains, as for example those described by McLellan (in Brien and McLellan 1962) from the Cockleshell Gully sand- stone, but there has not been enough work to justify generalization about the role of the ferruginous cement. Miscellaneous Diagenetic Minerals and Textures Glauconite . — The mineral glauconite, a product of early marine diagenesis, which has been mentioned in the discussion of the Clark Sand- stone, has been recorded from the following Western Australian units: Clark Sandstone (Cambrian); Emanuel Formation. Pander Greensand (Ordovician): Blina Shale (Triassio: Langey Siltstone, Callawa Formation (Jurassic); Birdrong Formation, Muderong Shale, Windalia Radiolarite, Alinga Formation, Toolonga Cal- cilutite, Peepingee Greensand, Molecap Green- sand. Gingin Chalk, Poison Hill Greensand, South Perth Formation. Madura Shale (Cretace- ous); Boongerooda Greensand, Wadera Cal- carenite (Palaeocene) ; Jubilee Calcarenite, Giralia Calcarenite. Osborne Formation, Plan- tagenet Beds (Eocene). Glauconite is present in other formations, and the known number of occurrences will probably increase with further drilling. Veevers and Wells (1961 p. 146) point out that glauconite is found in practically all subsurface occurrences of Mesozoic finer grained rocks in the Canning Basin, and its absence in outcrops is almost certainly due to removal by weathering. Hodgson *1962) has discussed the origin of the glauconite in the Plantagenet Beds at Cheyne Bay and concludes that much of it has been derived from mica. Phosphatic Minerals . — Phosphatic nodules have been recorded from several Western Australian units, namely Bogadi Greywacke (Permian); Colalura Sandstone. Bringo Shale (Jurass'c); Molecap Greensand, Poison Hill Greensand (Cretaceous). The phosphatic nodules in the Colalura Sandstone and Bringo Shale have been described by P. E. Playford (1959) who considers that they grew before burial in the sediment, possibly by growth from coprolitic cores. Fossil wood replaced by cellophane occurs in the Colalura Sandstone. The best known phosphate deposits are near Dandaragan in the Molecap Greensand, wliere calcium phosphate (collophane) nodules are also associated with phosphatized wood. Other phosphatic minerals present in places in the matrix, and apparently derived from collophane, are dufrenite. vivianite, beraunite, wavellite and minylite (Matheson 1948). Matheson be- lieved that the phosphatic material was initially dissolved from organic remains and deposited as collophane nodules during sedimentation and for some time after, before consolidation of the sediments. A pale to dark-brown mineral that replaces shell fragments in the Pleistocene Coastal Limestone was tentatively identified as colb- phane by Hodgson {hi Hodgson et al. 1962). and its identification has been con- firmed by chemical tests (Hodgson, personal communication). The mineral is very widely but sparsely distributed in the formation, and has been seen by the present author from Shark Bay and Rottnest Island, generally com- prising less than one per cent, of the rock. It is possible that during the Pleistocene there were numerous widely scattered and short-lived pockets of very high organic activity, in which the shells became phosphatized. The eustatic fluctuations of the Pleistocene would have depleted populations of these localized environ- ments only to allow them to become strongly established elsewhere. The phosphatized shells would be incorporated in the repeatedly re- worked terrigenous and organogenic detritus, and thus attain their present wide but sparse distribution. Gearksutite . — The rare mineral gearksutite. a hydrous fluoride of aluminium and calcium, has been described from a phosphatic layer near the base of the Cretaceous Poison Hill Greensand, about two miles east of Gingin. by Simpson (1920). He ascribes its formation to growth in situ, from interaction between solu- tions of the nearby minerals fluorapatite (present as nodules) and gibbsite (a constituent of over- lying laterite). The reaction postulated by Simpson seems to be on the borderline of late diagenesis and weathering. Jarosite . — The secondary mineral jarosite has been noted by Prider (1943, p. 39) in the Pre- cambrian Cardup Shale and by Clarke et al. (1951 ) in Permian formations of the Irwin River area. The latter authors believed that euxenic conditions led to formation of pyrite or mar- casite, from which the jarosite was later developed. This view’ accords with the origin suggested for jarosite in Mexico by Rough (1941). and in Yorkshire by Hartley (1957). As with gearksutite, its formation is due to either late diagenesis or weathering, depending on arbitrary definition. Tourmaline . — -Many ouartz sandstones con- tain rai'e but widely distributed grains of clastic tourmaline showing authigenic outgrowths. Clarke et al. (1954, p. 20) have described authi- genic elbaite on rounded schorlite grains from quartzites of the Precambrian Mt. Barren Group but it should be noted that these rocks have been metamorphozed. Prider (1943, p. 41) has described tourmaline euhedra from the Pre- cambrian Cardup Shale, but again there is doubt 42 about their origin, and they may be meta- somatic. Concretions . — Concretions, mainly calcite and limonite, are common in many Western Aus- tralian sediments, but they have generally not been the subject of detailed studies. Calcareous and sideritic concretions from the Learmonth Formation, Carnarvon Basin, have been de- scribed, and evidence for their formation before much compaction of the rock has been cited (Glover 1960a). Barite concretions are known from the Wogatti Sandstone and Gearle Silt- stone of the Carnarvon Basin. The varied con- cretions in sedimentary rocks of the Irwin River area have been suggested as a subject of re- search by Clarke et al. (1951. p. 47). Kaolinite. — Crystalline kaolinite. apparently of authigenic origin, is found in sandy claystone of the Wagina Sandstone, 13 miles east-north- east of Mingenew. The specimen examined with the universal stage consists of a clay mineral (80%), angular quartz grains (18%), opaque material (2%) and muscovite and bio- tite (<1%). The clay mineral, which was shown to be kaolinite by differential thermal analysis, is present in two forms. Most of the clay consists of a brown paste, but scattered through it are pale-brown, anhedral to sub- hedral crystals about 0.3 mm long. These crystals have uneven extinction, and their main cleavage is generally parallel to the bedding, but some are bent and have a vermicular form. A few grains, when cut parallel to their cleavage, show pseudo-hexagonal habit due to lines of minute inclusions arranged at 60“ (Plate III, Fig. 5). 2Va ranges from 7® to 40° (20 read- ings), and most values are between 15° and 30°. a little lower than those generally quoted for the mineral. Kaolinite is too soft to survive transport as large crystals, and these must have grown in the sedimentary rock. Post-compactional growth due to percolation of solutions can be ruled out, as the rock is impervious. The grains therefore probably grew before compaction, or during compaction as fluids were squeezed out. Allophanoids . — A white, isotropic, tuberose mineral, probably an allophanoid, has been re- corded by Prider (1948, pp. 112-113) as an authigenic constituent in the finer grades of the ferruginous sandstones of probable late Meso- zoic age at Ridge Hill. Outstanding Causes of Diagenesis Reactions Near the Sediinent-water Interface It is generally accepted that much glauconite and collophane have formed by reactions on or just below the sediment-water interface, and the evidence has been summarized in Carozzi (1960) and other recent texts. This origin accords well with the appearance of the glau- conite and collophane seen in this investigation, and with their relationship to other minerals. Pyrite is another mineral that is commonly of early diagenetic origin, and it is found in many rocks containing glauconite and collophane. It has long been suspected that much of the pyrite of black organic shales has been precipitated by sulphate-reducing bacteria, and the process has been described recently for sediments on the McMurdo Sound region of Antarctica by Barg- hoorn and Nichols (1961). Love (1958) believed that residues left after solution in nitric acid of framboidal pyrite from the Lower Car- boniferous Oil Shale of Scotland, were the remains of such bactei’ia. The present writer obtained similar residues from the Jurassic Learmonth Formation of Western Australia, but described them as remains of organic fragments including spores, which were etched and frayed (but not much compressed) during pyritization. The spores and other organic fragments were thought to have provided micro-environments of putrefaction in which the pyrite grew by bac- terial action (Glover 1960a). Pyrite, as minute spherical and subspherical grains and as crystals (commonly cubes and octahedra) often occupies the cells of wood fragments, and has evidently formed before compaction of the wood. It is found in some calcareous and sideritic concretions that also contain virtually uncompressed fossils and must themselves have formed early. It is frequently included in dolomite rhombs and in quartz out- growths, clear evidence that it formed before them. In ancient sediments, therefore, pyrite is often demonstrably one of the first products of diagenesis, and its common association with organic detritus accords with bacterial origin in a reducing environment. Ricour (I960) has drawn attention to the possibility of bacterial origin for dolomite, and that mineral is certainly the first diagenetic product to have formed after pyrite in the Western Australian sediments examined, and it could therefore have grown in reducing muds. There is no doubt that dolomite has formed before the sparry calcite matrix of some lime- stone (see the description of Septimus Lime- stone), but its oi'igin warrants fuller comment, and will be considered later. Limestones commonly contain authigenic quartz and feldspar euhedra (that is. com- pletely authigenic grains as distinct from the secondary outgrowths on clastic cores found in many sandstones), and many early workers attributed them to growth in the original cal- careous mud. There are generally no significant clues to indicate whether growth was early or late. However, there are rare records of authi- genic feldspars in shales and siltstones (Gruner and Thiel 1937) and it is hard to see how these feldspars could have formed after lithification. for the rocks would presumably have been impervious to solutions that might have pre- cipitated them. Much of the literature on diagenesis, particu- larly early diagenesis, assumes that it is promoted or assisted by the presence of organic matter. Sujkowski (1958, p. 2694) emphasized this view in his discussion of the agents of diagenesis and stated “Water is the main agent of diagenesis and organic matter is an auxiliary.” Compaction Many mineralogic changes in fine-grained sediments such as claystones and shales, and in sediments with an abundant clay-sized fraction, such as greywackes, seem to have been caused at least partly by compaction. These changes includes the formation of micas, chlorites and 43 coarse kaolinite crystals in claystones and shales, and the growth of chlorites and other flakey minerals that penetrate sand grains (with corresponding solution of quartz) in greywackes. It is unlikely, in view of the impermeability of these rocks, that solutions could have circulated in them after lithification. Lutites compact most rapidly during and shortly after deposition •Jones 1944) and although there may be slow compaction during deepening burial, the abun- dant liquids squeezed through the compacting rocks in the early stages probably assist in the changes listed above. The formation of sparry calcite from calci- lutite. apparently caused by compaction in some limestones, is commonly accompanied by expul- sion of carbonaceous matter, clay minerals and other finely divided impurities. This can be demonstrated in dolomitic limestones where dolornite crystals have replaced some of the matrix and are grey from zones of included impurities, whereas later recrystallization of the surrounding matrix has left it clear and sparry. Other changes generally ascribed to compac- tion include development of macro- and micro- stylolites in limestones (Stockdale 1922. Dun- nington 1954). Waldschmidt (1941) suggested that pressure between quartz grains of sandstones at their points of contact leads to solution, develop- ment of micro-stylolitic contacts and precipita- tion of the quartz nearby on grain boundaries where pressure is less. This is probably re- sponsible for some redistribution of quartz in sandstones, but as Pettijohn (1957, p. 657 » points out, many micro-stylolitic contacts in sand- stones are between the enlarged grains rather than adjacent detrital areas. Pettijohn earlier <1949, p. 481) attributed all well-developed sutures between enlarged grains to orogenic de- formation after authigenic growth, and all examples so far observed by the present writer have been in deformed rocks. Purely physical effects of compaction are seen ill the bending of micas and chlorites around sand grains, and in the flattening of some fossil remains. Intrastratal SolutiGU and Precipitation. The main causes of diagenesis in arenites are solution and precipitation by fluids percolating between the grains. The chief fluid by virtue of its ubiquity, is undoubtedly water, but i.he possible diagenetic effects of oil have provoke! little attention. Kulbicki and Millot (I960) consider the role of oil in some early Palaeozoic Saharan sandstones, and conclude that it has preserved kaolinite from the effects of salt waters, which elsewhere in the sequence hnve altered it to illite. Solution . — Textural evidence of solution is afforded by the raggedness of many heavy minerals, particularly ferromagnesian silicates, and this is generally best observed in grain mounts of heavy minerals concentrated from disaggregated sedimentary rocks. The sus- ceptibility to solution of many heavy minerals has long been recognized, notably by Bramlette (1929, 1941), Edelman and Doeglas (1932), and Smithson (1941). Pettijohn (1957) has summed up much of the literature on the subject. How- ever, not only the heavy minerals are affected. Textures resulting from the solution of quartz and feldspar grains, and of lithic fragments, are readily evident in thin sections of many "Western Australian sandstones. The grains show ragged boundaries, re-entrants and even skeletal textures, and are mostly replaced by carbonate or ferruginous cement. Many examples of corrosion of sand grains by diagenesis have been reported in the recent literature, and in fact it is now appreciated that surface textures such as frosting and pitting may be due to shallow corrosion as well as to abrasion^. A significant feature of this corrosion is that it provides an apparent source for much of the material that elsewhere forms the authigenic outgrowths on quartz, feldspar, and other minerals. strangely enough, frosting is said to be produced both by etching (Walker 1957) and by deposition of secondary silica (Humphries 1961). PLATE I Fig. 1.— Barite-cemented quartz sandstone from the Jurassic Wogatti Formation (Rough Range No. 1 Bore, core 9, 3.738-3.752 ft), as seen with the universal stage. The three quartz grains (light grey, no cleavage) appear to have irregular boundaries with the barite (dark grey, cleavage). Plane polarized hght. Widtn of field 0.4 mm. Fig. 2. Same fleld as in Fig. 1 after tilt of the stage through 25'. An authigenic quartz outgrowth with at least three faces is evident, and other faces can be detected with different tilt. Plane polarized light. Fig. 3. Glauconite and quartz in the Clark Sandstone (No. 48629). Authigenic quartz outgrowths tend to be moulded by the glauconite pellet, which is crossed by a small crack or micro-fault. Glauconite to the right of the crack is securely wedged between two quartz grains, but below the crack it has been displaced about 0 02 mm. apparently by growth of the quartz grain in the lower left of the held. Plane polarized light Width of field 0.6 mm. Pig. 4.— Glauconite and K-feldspar in the Clark Sandstone (No. 48629). The core of the feldspar which cannot be sharply distinguished from the authigenic rim in the photograph is microcline The rim has’ patchy extinction with 2V negative, about 64. and is adularia. 001 cleavage (parallel to the well-developed faces) and no cleavage pass from core to rim. The authigenic feldspar is moulded against the glauconite and its growth has apparently caused the glauconite to yield along a small fault. Plane polarized light Width of field 0 6 mm. Fig. 5.— Microline with an authigenic rim in the Clark Sandstone (No. 48629). The rim has slightly uneven extinction, is crowded with haematite inclusions, and has optics close to sanidine. In the rim 2V ranges in different places, f’-om 41° to 18^ (approx.). 001 and 010 cleavages are present in the core, and the 001 cleavage can be seen passing from the core to the rim. Plane polarized light. Width of field 0.35 mm. Fig. 6. — Same feldspar grain as in Pig. 5 under crossed nicols. Note how the wedge-like albite twins of the microcline core stop at the margins of the authigenic feldspar. Other grains are glauconite (right) feldspar (upper left) and quartz. 44 PLATE I 45 Preclpdation . — Precipitation from intrastratal solutions is a major factor in the diagenesis of arenites, and accounts for most outgrowths on quartz and feldspar. Pressure-solution of grains at their contacts, and precipitation nearby on grain boundaries where pressure is less, is probably responsible for some redistribu- ticn of quarts. However, pressure-solution is only a minor feature of the texture of most West Australian sandstones examined, and the process seems inadequate to account for the abundant secondary quartz in many of them. Moreover, some silicified sandstones show no pressure-solution effects and it can only be con- cluded that the quartz grains have acted as seed crystals and have been enlarged by percolatin'', siliceous solutions. The same process is re- sponsible for enlargement of feldspar grains. Naturally, sand grains do not act as seed crystals if they are not themselves crystals, cut are fragments of opaline or chalcedonic chert, schist, slate, volcanic glass or fine-grained volcanic rock, or if they are pellets and ooliths composed of clay-sized calcareous material. Shell fragments (except echinoderms, which break into coarsely crystalline debris) do not act as seeds. Therefore in lithic sandstones and cal- carenites many pores between clastic grains can not be filled with minerals that are optically continuous with the clastic fragments. These pores instead are commonly filled with pris- matic or fibrous minerals with their length toward the centre of the cavity, or else they are lined with several concentric mineral bands. Both of these textures could probably arise only by precipitation from circulatiiig solutions. Finally, a significant point of nega- tive evidence favouring precipitation from cir- culating solutions is the general absence of en- larged grains in greywackes and sandy lutites. where post-compactional circulation of fluids is practically impossible. These rocks often con- tain corroded quartz and feldspar grains, and the dissolved material was probably removed by the solutions squeezed out during compac- tion. to be made available for precipitation else- where. One of the problems connected with the pre- cipitated material is whether or not it normally travels a long way before deposition. There is evidence that it often does not, for it is charac- teristic of many authigenic minerals that they are most abundant in those sediments which are able to supply their component elements. Thus it was pointed out by Gilbert Hn Williams, Turner and Gilbert 1954) that quartz cement is abundant only in quartz-rich sandstones, authi- genic feldspars are found chiefly where there is also detrital feldspar, and carbonate cements, although found in all types of sandstones, are invariably formed wherever the original sand contained primary carbonate. He also points out that laumontite and chlorite are generally found in sandstones containing volcanic frag- ments and are much less abundant in others, and lists several Tertiary Californian sand- stones with volcanic detritus and chlorite cement. Montmorillonoid. a mineral that re- sembles chlorite in containing magnesium, has been described coating grains in late Tertiary lithic (volcanic) sandstones in California by Lerbeckmo (1957), and authigenic chlorite is described in similar sandstones from Mazakh- stan by Chernikov (1980). Australian chloritic sandstones with volcanic detritus include the Arrowsmith Sandstone (pp. 37-38) and numer- ous Permian and Triassic sandstones from the Bowen Basin, Queensland. A local exception to the generalization seems to be the Poole Sand- stone of the Canning Basin, Western Australia, which is a feldspathic sandstone with chlorite shells around the grains, but apparently no vol- canic material (Johnson and Dallwitz in Veevers and Wells 1961, p. 267). As Gilbert also noted. PLATE II Fig. 1.— Slightly kaolinized arkose from the Mokadine Formation (No. 36909). All grains except for opaques and the three clear qtiartz grains at bottom left are K-feldspar. Authigenic quartz and feldspar outgrowths have practically eliminated pore space. None of the boundaries of authigenic material in this field corresponds precisely with a common crystal face. A quartz grain has slightly penetrated the long feldspar grain, and the clear microcline (centre, right) has slightly penetrated the kaolinized adularia below it. The penetrations were probably caused by compaction before authigenesis. Plane polarized light. Width of field 0.6 mm. Fig. 2.— Micro-drusy texture in the Arrowsmith Sandstone {No. Pf2). Clastic grains are K-feldspar (left,, with cleavage) and quartz. The cavity between them has been filled by three minerals, which are, from the outside: d) a thin film of brown mica-like mineral, (2) a pale green chlorite layer with minute serraticns directed toward the centre of the cavity, and (3) an infilling of orthoclase (2Va — 43°). A little chlorite, probably projecting from a surface just grazed by the section, can be seen apparently within the orthoclase, Width of field 0.3 mm. Plane polarized light. Fig. 3.— Micro-drusy texture in the Arrowsmith Sandstone (No. Pf2), The cavity on the right has been filled with three minerals which are, from the outside: (1) a thin film of brown mica-like mineral. (2) a pale green chlorite with minute serrations directed toward the centre of the cavity, and (3) an infilling of adularia (2Va = 55°). The sequence is different in the cavity on the left and is. from the outside: (1) the mica-like mineral. (2) quartz (visible only on the lower right of the cavity). (3) chlorite, and (4) quartz, optically con- tinuous with (2). Dark clastic grains are iron-stained volcanic fragments, and others are quartz and feldspar. Plane polarized light. Width of field 0.6 mm. Fig. 4. — Authigenic enlargement of plagioclase (An.-,) in the Arrowsmith Sandstone (No. Pf61). Twinning continuous without interruption into the rim (lower right). The grey mineral is calcite: clastic fragments include quartz. K-feldspar and opaques. Crossed nicols. Width of field 0.9 mm. Fig. 5.— Micro-drusy texture in Enokurra Sandstone (No. 38725). A fringe of sericite whose minute flakes point toward the centre of the cavity, adheres to the iron-stained surfaces of the clastic grains. This was the first authigenic mineral to form. The rest of the cavity is filled with quartz, most of it optically continuous with the clastic grain in extinction flower left). Crossed nicols. Width of field 0.3 mm. Fig. 6 —Authigenic quartz in the Cockatoo Sandstone (No. 48624). Most of the pore space has been filled. Some boundaries correspond to common crystal faces, but many do not. and some are curved. The clastic grains are outlined by a film of haematite. Plane polarized light. Width of field 1 mm. 46 V PLATE II. 47 part of the tendency for particular cements to concentrate in certain sandstones is due to the presence of sand gi’ains which act as crystal nuclei on which cement of the same composition is precipitated. This holds with quartz and feldspar, but not with chlorite, so the presence of the latter in sandstones with volcanic material is generally convincing evidence for its local derivation. It must clearly be under- stood that the preceding discussion applies to precipitated chlorite, and not to that which results from reorganization of the finely divided matrix in rocks such as greywackes and shales. The replacement of carbonate in many lime- stones by quartz, chalcedony, opal and feldspar seems contrary to the general principle of intra- sti'atal (as opposed to interstratal) origin for the solutions. There can be no doubt that many such solutions have originated elsewhere in more siliceous strata, but the replacement process is sufficiently complex to warrant brief separate treatment below. The Carbonate-silica Relationship. — Silicifica- tion of carbonates has to be considered with its reverse process, the equally common car- bonation of silica and silicates in calcareous sandstones and other rocks. The considerable literature on the chert problem, particularly as it applies to the distinction between primary and secondary chert, is well summarized in Pettijohn (1957), and will not be reviewed here. Some of the views about the causes of replace- ment. where it can be demonstrated, will, how- ever, be mentioned. Much of the discussion about replacement has hinged on the chemistry of calcite and amor- phous silica, particularly on their solubility under different pH conditions. It has been shown that at 25 C the solubility of amorphous silica is practically independent of pH except where it exceeds pH9, when the silica becomes increasingly soluble with increase in alkalinity (see Krauskopf 1959). On the other hand, calcite becomes less soluble with increase in pH. Recent papers to use these data include that of Rouge et al. (1959) who explained a decrease in quartz and silicates of certain limestones as they became more strongily dolomitized by assuming that the quartz and silicates dissolved because of the increased pH that caused dolo- mitization. Examples of chert replacing car- bonate and vice versa in the same rocks have been cited by Walker (1962) and ascribed either to variations in pH where the interstitial waters are highly alkaline (pH > 9) or to variations in temperature attendant on deep burial. The prospect of sufficiently high pH values for these reactions in the natural environment has also been accepted by Dapples (1962, p. 917), but not by Siever (1962) who has discounted the signifi- cance of pH. and has put forward the idea that clays act as semi-permeable membranes causing solution of carbonate and precipitation of silica. He has also appealed to several other mechanisms. Generalizations about the distribution of authigenic quartz, chalcedony and opal are possible, though much is still unknown about their origin. Quartz forms very commonly as outgrowths on clastic quartz nuclei in many rocks including calcareous sandstones and sandy limestones, and these outgrowths are generally fairly free of inclusions. Quartz also grows as minute, widely distributed euhedra without obvious clastic cores in some limestones, but the euhedra are often crowded with carbonate in- clusions.'" Authigenic feldspar assumes both forms, but is far less common. Fine-grained to micro-crystalline quartz forms many cherts, some of which, like the Coomberdale Chert, are considered to be epigenetic. Chalcedony also replaces carbonate widely, but opal generally seems to be Tertiary or younger, and it tends mainly to replace clay-sized carbonate, though it is not restricted to that lithology. Most pre- Tertiary opal has probably crystallized to micro- crystalline and fine-grained quartz. Selective replacement by different forms of silica often leaves precise details of the original texture as in Plate III. Fig. 6. It has been assumed by Millot et aL (1959) that percolating waters with low concentrations of dissolved silica cause pre- cipitation of quartz, whereas solutions with high concentrations, and often many impurities, result in opal and chalcedony. However, the * Although quartz and feldspar euhedra in limestones were the first generally recognized products of authigenesis they have yielded practically no textural evidence of their time of formation, variously suggested as very early to late in the history of the limestone. PLATE III Fig. 1.- Micro-drusy texture in the Coastal Limestone (No. 50267). Calcite fibres partly cement a coarse- grained calcarenite. Plane polarized light. Width of field 0.8 mm. Fig. 2— Micro '■’nny texture in an oolite from Devonian limestones of the Lennard Shelf (No 50269) Sparry calcite completely fills the voids between individual ooliths. Plane polarized light. Width of field 1.75 mm. Pig. 3.— Dolomite and sparry calcite in the Septimus Limestone (No. 36879). There are a few quartz <^rains in the top right corner. All calcite is part of the one crystal (see cleavage). Slight weathering has oxidized Iron and emphasizes the zone of impurities in each dolomite rhomb. The dolomite crystals show no preferred orientation, and apparently grew in an impure calcareous mud which later crystallized to sparry calcite expelling impurities. Plane polarized light. Width of field 1 mm. Fig. 4.— Replacement in Coastal Limestone (No. 8656). Rounded quartz grains are partly replaced by a finely divided clay-calcite matrix. Plane polarized light. Width of field 3.6 mm. Fig. 5.— Authigenic kaolinite showing pseudo-hexagonal structure due to lines of inclusions in sandy •claystcne of the Waglna Sandstone (No. 36927). The matrix is finely divided kaolinite. and there’ are a few quartz grains. Plane polarized light. Width of field 0.6 mm. F'<^ 6.— Siiific-+ion in the Toolonga Calcilutlte (No 50768. from (>-mile north-west of varin°a North Homestead). This silicification is not widespread, but is very selective. Foraminlfera are replaced by ctialcedonic f’lartz. and the groundma^s is onal with calcareous inclusion . Glauconite (not shown here) is not replaced. Plane polarized light. Width of field 1 mm. 48 i tendency observed by Millot (1960, p. 143) for chalcedony to predominate in replacement of carbonate later suggested to him that some in- trinsic feature of the calcite itself might in- fluence the type of mineralization. It will be evident from the above summary that much is yet to be determined about carbonate-silica relationships in sedimentary rocks, and that one of the best foundations for advancement is still in the amassing and recording of petrographic data. The Dolomite Problem Dolomite is one of the enigmas of sedimentary petrology. It is common in Precambrian and Palaeozoic limestones, less common in Mesozoic limestones, unusual in Tertiary and Quaternary limestones, and apparently forms at present under only exceptional circumstances. Examples of its present day formation are described by Alderman (1959), Miller (1961) and Taft (196D, comprehensive discussions of the origin of the mineral are given by Pairbridge (1957), Petti- john (1957) and Carozzi (1960), and a recent description of its laboratory synthesis is pre- sented by Siegel (1961). The texture of many dolomitic rocks points to formation of the dolomite in partly compacted muds below the sediment-water interface, where it is unaffected by wave or current action. The dolomitic grains are euhedral, and commonly enclose pyrite, showing that they foim after that early diagenetic mineral. These features are consistent with the general absence of dolo- mite at present in shallow, unconsolidated mar- ine muds. On the other hand the texture of most dolomitic rocks indicates that the dolomite had finished growing before their final consoli- dation. For example, dolomite is found in some shales and the impermeability of those rocks seems to preclude precipitation of the mineral .from percolating solutions after their final compaction. Furthermore, the textures of many limestones show that dolomite formed before micro-stylolites and before crystallization of sparry calcite from calcilutite, both of which are probably results of intense compaction. It is also probable that the porosity of many dolomitic rocks indicates formation of dolomite before solution of interstitial calcite, for the dolomite seems to have served as the framework from which the calcite was leached. Finally, in this respect, it should be noted that much dolomite has clearly preceded authigenic quartz in some sandstones, although in others the reverse re- lationship has been established. The above textural features, taken together, would apply very well to dolomite that formed as compaction began, taut before those diagenetic features that commonly accompany the last stages of con- solidation. Some dolomite, however, probably forms as a product of late diagenesis in consolidated rocks. Other dolomite undoubtedly foims very early in evaporites. Recently Sabins (1962) has recog- nised rounded “primary” dolomite in Cretaceous sandstones and states that it formed before burial of the sediment, and hence was subject to abrasion and sorting. Primary dolomite is said by Sabins to be restricted to marine sand- stones, and he lists 42 formations containing it. Fig. 5 £nlargeme7it Textures Simple Enlargement Textures (a) Calcarenite showing sparry calcite in crystallographic continuity with crinoid debris. Stippled fragments are calcareous pellets. (b) Quartz sandstone with quartz outgrowths crystallographically continuous with clastic cores. Note faces. Lightly stippled areas represent pores. (c) Quartz sandstone with pores completely occupied with secondary quartz. Some outgrowths bounded by plane surfaces, some not. No sutured boundaries. Indentation Textures (d) Dolomitic sandy marl in which dolomite is partly surrounded by, or has partly penetrated, quartz outgrowths. Texture does not reveal whether quariz or dolomite grew first. (e) Dolomitic sandy marl, same as (d), except one of the dolomite grains is moulded onto a clastic quartz core. Dolomite therefore preceded secondary quartz. Enclosure texture (f) Dolomitic sandstone with dolomite completely enclosed by quartz. Dolomite therefore formed first, and order is confirmed by moulded dolomite (upper centre). Note how a moulded dolomite has retreated marginally (left centre) due to slight solution during silicification. Pressure -^solution Textures (g) Quartz sandstone with clastic grains showing sutured boundaries due to compaction, deformation or both. Secondary quartz(s) fills voids, may have come partly or completely from quartz dis- solved along sutured contacts. (h) Calcarenite with micro-stylolite due to compaction. Micro-stylolite outlined by iron-stained argillaceous matter and small quartz grains, both insoluble in the particular conditions of its formation here. Sparse distribution of quartz in rock suggests compaction equivalent to field of view. (i) Quartzite with sutured boundaries between outgrowths due to deformation after diagenesls. As much a metamorphic as a diagenetic texture. Micro-drusy Textures Simple Micro-drusy Textures (j) Calcarenite partly cemented with fibrous calcite. Fibres are elongated normal to grain boundaries. (k) Calcarenite completely cemented with sparry calcite. Long axes of calcite crystals are normal to grain boundaries. (l) Lithic (volcanic) sandstone cemented by fibrous chlorite. Texture basically the same as in (j) and (k). 50 Fig. 5. 51 To sum up, dolomite almost certainly forms in different ways, and may grow early (by direct precipitation in marine sandstones during sedi- mentation, or by evaporation) or late (replacing other diagenetic minerals in porous rocks). How'ever. textures in ancient rocks suggest that most dolomite in limestones forms during initial compaction, but before final consolidation. Perhaps the composition of Precambrian and Palaeozoic atmospheres helped to account for its abundance during those eras, as Ronov ( 1959 ) believes. A hypothesis which explains the requir- ed concentration of magnesia from sea water is that of Adams and Rhodes (1960) who postulate alteration of limestones by magnesia-rich solu- tions seeping from evaporite lagoons. Further information from the drilling of areas where reefs are now growing seems desirable, though it must be agreed with Ladd et al. (1953) that the drilling of atolls has so far yielded puzzling results. Common Diagenetic Textures and their Interpretation. Diagenesis, like magmatic crystallization and metamorphism, causes many textures, and, as with those processes, certain basic textures appear again and again. These textures fall in the following categoi’ies : enlargement textures, pressure-solution textures, pore-filling or micro- drusy textures, reorganization textures, and replacement textures. Most of them have been encountered in the rocks described earlier, and they are now represented diagrammatically in Figs. 5 and 6. A discussion of their main features and significance follows below. Enlargement Textures All enlargement textures, are the result of precipitation of minerals on crystal nuclei of the same or similar composition and the best examples are found in porous arenites, where solutions are able to circulate freely. However, some outgrowths have replaced earlier minerals which partly filled the cavities, and these out- growths may contain clues in the fcrm of minute inclusions. Simple Enlargejnent Texfiires.— Tourmaline and other heavy minerals grow either small projections, or faces, but crystalline calcitic debris, quartz and feldspar usually grow large, simple and well-developed faces, except where prevented by mutual interference (see Figs. 5a 5b, 5c). The commonly developed faces are, in calcite, rhombohedi a : in quartz, prisms and rhombohedra: and in feldspar, basal and side pinacoids and prisms (110 and 130). Faces which are less common in feldspar, but are not rare, include the pyramid (111) and dome (101). Indentation Textures.— Authigenic outgrowths on one mineral may partly surround, or be partly penetrated, by another authigenic mineral to give an indentation texture. This texture, illustrated in Fig, 5d, does not reveal which of the two minerals grew first. However, in the special case where one of the authigenic min- erals has grown so that it is moulded against the clastic core of the other, the moulded mineral has been the first to grow (Fig. 5e). Enclosure Textures.— These are a development of the indentation texture in which the authi- genic rim of one mineral grows sufficiently to include a second authigenic mineral. The in- cluded mineral forms first (Fig. 5f). Pressure-solution Textures The secondary quartz of some sandstones has been attributed to pressure-solution effects. Where quartz grains impinge on each other under intense pressure, they partly dissolve at points of contact, and interlock with micro- stylolitic boundaries. The dissolved quartz is supposed to precipitate in adjacent pores, where pressure is less. The common association of secondary quartz and micro-stylolitic contacts a* u Micro-drusy Textures (continued) Composite Mlcro-drusy Textures The order of formation was ( 1 ) micaceous mineral (or its precursor) {2\ chlorite (Based on a diagenetic sequence in Arrowsmith Sandstone)^ )- ( ) chlorite. (3) feldspar. Reorganization Textures Id) orYst^. 1 ^. Fragility of the crystals Is proof of in situ formation. ml- ™ »rre1SS-ent"?e\fn- “ (e) Fontainebleau sandstone, in which calcite has reorganized to form large crystals. Replacement Textures ?alcUe.“‘'’ replaced by quartz euhedra with calcareous inclusions. The long fragment is The dolomite ' '* repV\"smt?d'1?rp|tchS®of tr£^^ impurity , "and Tsotted pyme causTd^rreplacement of— sandstones may be fj) A complex hut fairly common texture, in which dolomite has partly renlaced the matrix of a Thlre l^s beeS^latir?I?r^^^^ inclusions form a dark zone in each dolomite rhomb, mere nas oeen later recrystallization of the matrix to sparry calcite. with expulsion of impurities. 52 a b c d e seems to accord with this origin, but there is often too much quartz to be explained by the micro-stylolites. Appeal must then be made to intrastratal solutions for at least part of the quartz. In some sandstones (better called quartzites), micro-stylolitic boundaries are con- fined to adjacent outgrowths, a texture prob- ably due to imposition of great load, or faulting or folding, after diagenetic enlargement. These textures are therefore practically metamorphic. Micro-stylolites that traverse the rock, cutting across grains, are another form of pressure- solution common in limestones, but not absent from quartzites. Textures arising from pres- sure-solution are illustrated in Pigs. 5g, 5h, 5i. Pore- filling or Micro-drusy Textures All micro-drusy textures are believed to form by precipitation in porous rocks, generally arenites. They differ from enlargement tex- tures in that clastic grains do not act as nuclei on which the precipitated material grows in crystallographic continuity. This is because the clastic fragments are non-crystalline (e.g. vol- canic glass, some chert), or are aggregates of minute crystals (e.g. schist, shale, fine-grained volcanic rock, ooliths. pellets, etc.) rather than single crystals, or because they are single crys- tals, chemically very different from the precipi- tated material (e.g. quartz grains and chlorite cement). The authigenic minerals may form a fringe on which the serrations are normal to giain sui’faces and point into the cavity, or they may form a series of elongate crystals 01 fibres normal to grain surfaces and directed into the cavity. This texture is practically proof of growth in an empty pore cavity, another difference from enlargement textures, some of which are due to replacement of earlier pore- filling material. Simple Micro-drusy Texfures.— Simple tex- tures. involving one authigenic mineral are Illustrated in Figs. 5j, 5k, 51. Composite Micro-drusy Texfwres.— Composite textures may involve several authigenic minerals, of which the outermost are the first to form' and the innermost the last. See Figs. 6a, 6b. Reorganization Textures Reorganization textures are like replacement textuves (discussed below) except that in the former the substituted material is chemically the same as, or similar to, the original material. The 1 eoi ganization is generally to a coarser crystalline form such as the change from cal- cilutite to sparry calcite, or from the clay-sized matrix of greywacke to aggregates of chlorite, mica, and other flaky minerals. Other changes are from finely divided clay minerals to coarse vermicular crystals, and from aragonite to cal- cite in fossil material, although the latter does not involve coarsening. Most reorganization textures seem to have been the result of com- paction. Some of them are illustrated in Figs 6c. 6d. 6e. Replacement Textures Replacement textures are very common dia- genetic features of sedimentary rocks, and some are illustrated in Figs. 6f-j. They are ap- parently the result of almost simultaneous solution and precipitation, and in those lutites which are practically impervious to the passage of solutions, the process probably took place during compaction and the expulsion of con- tained fluids. As already noted, some en- largement textures can also be classified as replacement textures. Conclusions There is no reason to suppose that the pro- cesses leading to diagenesis are less complex than those leading to solidification of magmas, or to metamorphism. Nevertheless, they are becoming increasingly well understood, and consequently the examination of diagenetic textures should lead to petrologic inference as reliable and useful as that resulting from examination of igneous and metamorphic tex- tures. The scheme set forth above is an attempt to group diagenetic textures so that they form a useful and simple guide to the origin of the sedimentary rocks in which they occur. Ackn 3 wledgments Some of the rock specimens examined during this investigation were kindly supplied by West Australian Petroleum Pty. Ltd. Helpful discus- sions were had with Dr. P. j. Coleman. Dr. P. E Playford and Professor R. T. Prider. References c.. ana itnoaes, M. u. ( 1960 ) .— Dolomitization by seepage refluxion. Bull. Amer Ass Petrol. Geol. 44: 1912-1920. Alderman. A. R. (1959). — Aspects of carbonate sedimenta- tion. J. Geol. Soc. Aust. 6: 1-10. Arriens, P. A., and Lalor, J. H. (1959). — The geology of the Billeranga Hills, Western Australia. ■a V, )^ep., Univ. W. Aust. Barghoorn, E. S„ and Nichols, R. L. ( 1961). -Sulphate- reducing bacteria and pyritic sediments in Antarctica. Science 134; 190. Baskin. Y. (1956).— A study of authigenic feldspars ./ Geol. 64; 132-155. Bastian. L. V. (1961).— Geology of the Green Brook Area. Mlngenew District. Western Australia. Thesis. Geol. Dep., Univ. W. Aust. Basumallick, S. (1962). — Crystal-optic study of secondary overgrowth in quartz. Amer. Min 47: 1473-1478. Boswell. P. G. H. (1933). — “On the Mineralogy of Sedi- mentary Rocks.” (Murby: London.) Bramlette, M. N. (1929). — Natural etching of detrital garnet. Amer. Min. 14: 336-337. Bramlette. M. N. (1941). — The stability of minerals in sandstone. J. Sediment. Petrol. 11; 32-36. Brien. J. W.. and McLellan. G. A. (1962).— Geology of part of the Hill River area. Western Aus- tralia. Thesis. Geol. Dep., Univ. W. Aust. Carozzi. A. V. (1960). — “Microscopic Sedimentary Petro- graphy.” (Wiley: New York.) Chaisson, U. (1950). — The optics of triclinic adularia. J. Geol. 58: 537-547. Chernikov, O. A. (1960). — Certain authigenic neoforma- tions in the Lower Carboniferous sandy- silt rocks of the Ashaninsk Trough (Northwest Kazakhstan). Dokl. Acad. Sci. USSR. 135: (4); 958-960. (English Translation. Amer. Geol. Inst, for 1961; 1274-1277.) Clarke. E. de C.. Prendergast. K. L., Teichert. C.. and Fairbridge, R. W. (1951) — Permian succes- sion and structure in the northern part of the Irwin Basin. Western Australia, J. Roy. Soc. W. Aust. 35; 31-84. 54 Clarke, E. de C., Phillipps, H. T. and Pridei. R. T. (1954).— The Pre-Cambrian Geology of part of the south coast of Western Australia. J. Roy. Soc. W. Aust. 38: 1-64. Dana, E. S. (1899). — “The System of Mineralogy. Des- criptive Mineralogy.” 6th Ed. (Wiley; New York. ) Dapples. E. C. (1962). — Stages of diagenesis in the de- velopment of sandstones. Bull. GeoL Soc. Amer. 73: 913-934. Dunningtcn, H. V. (1954). — Stylolite development post- dates rock induration. J. Sedwient. Petrol. 24: 27-49. Edelman, C. H., and Doeglas, D. J. ( 1932).— Reliktstruk- turen detritischer Pyroxene und Amphi- bole. Miner, petrog. Mitt. 42: 482-490. Pairbridge. R. W. (1957). — The dolomite question. Spec. Publ. Soc. Econ. Paleont. Mineral. No. 5: 125-178. Polk, R. L. (1959). — Practical petrographic classification of limestones. Bull. Amer. Ass. Petrol. Geol. 43: 1-38. Gilbert. C. M. (1949). — Cementation of some California Tertiary reservoir sands. J. Geol. 57: 1-17. Gilbert. C. M., and Turner, P. J. (1949). — Use of the universal stage in sedimentary petro- graphy. Amer. J. Sci. 247: 1-26. Glover. J. E. (1955). — Petrology and petrography of limestones from the Fitzroy Basin, Western Australia. Rep. Bur. Miner. Resour. Aust- in. Glover. J. E. (1960a). — Concretionary sandstones from the Learmonth Formation. Carnarvon Basin. Western Australia. J. Roy. Soc. W. Aust. 43: 68-73. Glover, J. E. (1960b). — A contribution to the petrology of the Yandanooka Group. J. Roy. Soc. W. Aust. 43: 97-103. Gruner, J. W.. and Thiel. G. A. (1937).— The occurrence of fine grained authigenic feldspar in shales and silts. Amer. Min. 22: 842-846. Hartley. J. (1957). — Jarosite from Carboniferous shales in Yorkshire. Trans. Leeds Geol. Ass. 7: 19-23. Hewlett, C. G. (1959). — Optical properties of potassic feldspars. Bull. Geol. Soc. Amer. 70: 511- 538. Hodgson. E. A. (1962). — Origin of glauconite in some sandstones of the Plantagenet Beds, Cheyne Bay, Western Australia. J. Roy. Soc. W. Aust. 45: 115-116. Hodgson. E. A.. Quilty, P. J. G.. and Rutledge, D. I. (1962). — The geology of the south coast in the vicinity of Cheyne Bay, Western Aus- tralia. Thesis. Geol. Dep.. Univ. W. Aust. Humphries, D. W. (1961). — The Upper Cretaceous white sandstone of Loch Aline. Argyll. Scotland. Ptoc. Yorks. Geol. Soc. 33: 47-76. Tiling, L. V. (1954). — Bahaman calcareous sands. Bull. Amer. Ass. Petrol. Geol. 38: 1-95. Jones, O. T. (1944). — The compaction of muddy sedi- ments. Quart. J. Geol. Soc. Loud. 100.- 137-160. Krauskopf. K. B. (1959). — The geochemistry of silica in sedimentary environments. Spec. Publ Econ. Paleont. Mineral. No. 7: 4-19. Kulbicki. G.. and Millot, G. (1960). — L’evolution de la fraction argileuse des gres petroliers cambro-ordoviciens du Sahara Central. Bull. Carte Geol. Als. Lorr. 13 (4): 147-156. Ladd. H. S.. Ingerson, E.. Townsend, R. C.. Russell, M., and Stephenson, H. K. (1953). — Drilling on Eniwetok Atoll. Marshall Islands. Bull. Amer. Ass. Petrol. Geol. 37 (10): 2257-2280. Laves, F. (1950). — The lattice and twinning of micro- cline and other potash feldspars. J. Geol. 58: 548-571. Lerbekmo, J. F. (1957). — Authigenic montmorillonoid cement in andesitic sandstones of central California. J. Sediment. Petrol. 27: 298- 305. Logan, B. W,. and Chase, R. L. (1961). — The stratigraphy of the Moora Group, Western Australia. J. Roy. Soc. W. Aust. 44: 14-31. Love. L. G. (1958). — Micro-organisms and the presence of svngenetic pyrite. Quart. J. Geol. Soc. Lond. 113: 429-440. Matheson. R. S. (1948). — The Dandaragan phosphate deposits. Min. Resour. Bull. Dep. Min. W. Aust. 4. MeWhae, J. R. H., Playford, P. E.. Lindner, A. W. Glenis- ter, B. F. and Balme, B. F. (1958). The stratigraphy of Western Australia. J. Geol. Soc. Aust. 4 (2): 1-161. Miller, D. N. Jr. (1961). — Early diagenetic dolomite as- sociated with salt extraction process, jnagua, Bahamas J. Sedunent. Petrol. 31: 473-476. Millot, G,, Radier. H.. Muller-Feuga, R.. Defossez, M., and Wey, R. (1959). — Sur la geochimie de la silice et les silicifications sahariennes. Bull. Carte Geol. Als. Lorr. 12 (2): 3-14. Millot, G. (1960). — Silice, silex, silifications et croissance des cristaux. Bull. Carte Geol. Als. Lorr. 13 (4): 129-146. Packham, G. H.. and Crook. K. A. W. (I960).— The prin- ciple of diagenetic facies and some of its implications. J. Geol. 68: 392-407. Petti]ohn, F. J. (1949). — “Sedimentary Rocks.” (Harper: New York.) Pettijohn, F. J. (1957).— “Sedimentary Rocks.” 2nd Ed. (Harper: New York.) Playford, G. (1958). — Permian stratigraphy of the WooJ- aga Creek Area. Mingenew District. West- ern Australia. Thesis. Geol. Dep., Univ. W. Aust. Playford, P. E. (1959). — Jurassic stratigraphy of the Geraldton district. Western Australia. J. Roy. Soc. W. Aust. 4? 101-124. Pough, F. H. (1941). — The formation of jarosite on pyrite ornaments. Amer. Min. 26: 562-564. Prider, R. T. (1943). — The contact between the granitic rocks and the Cardup Series at Armadale. J. Roy. Soc. W. Aust. 27: 27-55. Prider. R. T. (1948).— The geology of the Darling Scarp at Ridge Hill. J. Roy. Soc. W. Aust. 32: 105-129. Ranford, L. C., and Shaw. S. E. (I960).— Geology of part of the Wooderarrung River Mullewa area Western Australia. Thesis. Geol. Dep.. Univ. W. Aust. Ricour, J. (I960). — De la genese de certaines dolomies. C. R. Acad. Sci. Paris 251 (17): 1798-1800 Ride. W. D. L., Mees, G. F., Douglas, A. M.. Royce, R. D.. Tyndale-Biscoe, C. H. (1962).— The results of an expedition to Bernier and Dorre Islands, Shark Bay, Western Australia in July, 1959. Fauna Bull. Fish. Dep. W. Aust. Honov. A. B. (1959). — On the post-Precambrian geo- chemical history of the atmosphere and hydrosphere. Geochemistry (Geokhimiya ) 1959 (5)’ 493-506 Rouge, P.-E., Charpal, O., and Montadert. L. (1959).— Relations reciproques entre dolomitization et fraction silica tee des roches carbonatees: influence de la dolomitization sur la frac- tion silicatee. C. R. Acad. Sci. Paris 248 (3): 435-436. Sabins. Floyd. F. Jr. (1962).— Grains of detrital, seernd- ary. and primary dolomite from Cretaceous strata of the Western Interior. Bull. Geol. Soc. Amer. 73: 1183-1196. Siegel, F. R. (1961). — Factors influencing the precipita- tion of dolomitic carbonates. Bull. State Geol. Surv. Kans. 152 (5): 127-158. Siever, R. S. (1962). — Silica solubility, 0°-200° C., and the diagenesis of siliceous sediments. J. Geol. 70: 127-150. Simpson. E. S. (1920). — On gearksutite at Gingin, West- ern Australia. Min. Mag. 19: 23-39. Smithson. F. (1941). — The alteration of detrital minerals in the Mesozoic rocks of Yorkshire. Geol Mag. 78: 97-112. Serby. H. C. (1851). — On the microscopical structure of the calcareous grit of the Yorkshire coast. Quart. J. Geol. Soc. Lond. 7: 1-6. Sorby, H. C. (1904). — Notes on the coral rock of Funafuti in “The Atoll of Funafuti.” pp. 390-391 (Roy. Soc.: London.) Stauffer. K. W. (1962). — Quantitative petrographic study nf Paleozoic carbonate rocks. Caballo Mountains, New Mexico. J. Sediment. Petrol. 32: 357-396. Stcckdale. P. B. (1922). — Stylolites: their natiire and orgin. Ind. Univ. Stud. 19: 1-97. Sujkowski, Zb. L. (1958). — Diagenesis. Bull. Amer. Ass. Petrol. Geol. 42; 2692-2717. Taft, W. H. (1961). — Authigenic dolomite in modern cprbonate sediments along the southern coast of Florida. Science 134: 561-562. Topkaya. M. (1950). — Recherches sur les silicates authi- genes dans les roches sedimentaires. Bull. Lab. Geol. Min., Mus. Geol. Univ. Lausanne 97. Traves, D. M. (1955).— The geology of the Ord-Victoria region, Northern Aixstralia. Bull. Bur. Miner. Resour. Aust. 27. 55 Turner, P. J. (1947). — Determination of plagioclase with the four-axis universal stage. Amer. Min. 32: 389-410. Veevers. J. J.. and Wells, A. T. (1961).— The geology of the Canning Basin, Western Australia. Bull. Bur. Miner. Resour. Aust. 60. Waldschmidt, W. A. (1941). — Cementing materials in sandstones and their probable influence on migration and accumulation of oil and gas. Bull. Amer. Ass. Petrol. Geol. 25: 1839-1879. Walker, T. R. (1957). — Frosting of quartz grains by carbonate replacement. Bull. Geol Soc Amer. 68: 267-268. Walker, T. R. (1962). — Reversible nature of chert- carbonate replacement in sedimentary rocks. Bull. Geol. Soc. Amer. 73: 237-242. Weeks, L. G. (1957). — Origin of carbonate concretions in shales, Magdalena Valley, Colombia. Bull Geol. Soc. Amer. 68: 95-102. Williams, H.. Turner, F. J., and Gilbert, C. M. (1954) _ “Petrography.” (Freeman: San Francisco.) 56 7. — Round Australite Core from Graball, Western Australia By George Baker* Manuscrivt received — 20th November, 1962. A large round australite core found at Graball, Western Australia weighs 168 grams and is 57 mm in diameter and 34.5 mm thick. It is the fourth largest known australite and the second largest known round core. A sculpture pattern of flow swirls and pits on the posterior surface is apparently of primary origin, produced in an extra-terrestrial environ- ment. Meandrlne grooves, crater-like depressions, and minute pits on the anterior surface, are due largely to differential solution-etching in soils. They are limited to a thermally stressed surface left after considerable ablation of the primary form. Ablation was by frictional heating during aerodynamically stable transit at ultrasupersonic velocity through the earth’s atmosphere. A well- developed flaked equatorial zone was evidently initiated circumferentially around the edge of the specimen during the end phases of high- speed infall, and subsequently became etched by terrestrial weathering. Introduction A large round australite core, discovered in 1952 on the roadside near Graball telephone ex- change, approximately seven miles north-west of Mt. Walker, near Narembeen, 155 miles east of Perth, Western Australia, has recently been brought to notice through the courtesy of Mr. W. H. Cleverly of the School of Mines, Kalgoor- lie, and Mr. W. Meacock of the Mt. Walker Gov- ernment School. Western Australia. The specimen, weighing 168 grams, was kindly loaned by the owner, Mr. W. Berry, of Mt. Walker, Western Australia, for examination. Only three other australites are known to be heavier than this specimen, each of them weighing over 200 grams. Three casts of the specimen, prepared in “Artificial Stone” at the Geology Department. University of Melbourne, are lodged: one in the University of Melbourne geological collection, one in the tektite collection of the National Museum of Victoria, and one in the author’s private collection of tektites. Size of Specimen The round australite core has a diameter of nearly 57 mm and a depth ( thickness* of 34.5 mm. It weighs 168.28 grams, a value ex- ceeded by only three other australites. The largest is a fractured oval-shaped form weigh- ing 238 grams from Warralakin, Western Aus- tralia (Baker 1962a); the next is a large round australite core weighing 218 grams from Lake Yealering, Western Australia, described by Fen- ner (1955) as a lens form; the third largest is a boat-shaped form weighing 208.9 grams from * C.S.I.R.O. Mineragraphic Investigations, c/o Geology Department, University of Melbourne. Parkville, N.2, Victoria. Karoonda, South Australia (Fenner 1955, Plate VII Nos. 3 and 4). The Graball specimen is thus the second largest known round core type of australite; it shows comparable sculpture patterns to those on the larger Lake Yealering round australite core figured by Fenner (1955, Plate VII, Nos 1 and 2). The specific gravity of the Graball specimen, determined in distilled water (T 19.3°C.) on an air-damped chemical balance, is 2.434. From the specific gravity — silica relationships of tek- tites, this value indicates a silica content of approximately 72 per cent. Structure and Sculpture The specimen is reasonably well preserved, and shows the following features: Posterior surface . — The posterior surface (Plate I, A), which remained directed back along the flight path during ultrasupersonic atmospheric flight earthwards, shows no features assignable to the effects of aerodynamic heating and abla- tion or aerodynamic sculpturing, hence its sculp- ture pattern is considered to have been de- veloped in an extra-terrestrial milieu. It has been but slightly modified by terrestrial erosion since landing on the earth’s surface a few thousand years ago. An artificially chipped area around the region of the rear pole on the posterior surface, is approximately 10 mm in diameter. It reveals the characteristic vitreous lustre of freshly broken tektite glass (Plate I, A). The sub- conchoidal fracture surface shows secondary ripple fracture lines and carries five “knobs” up to 1.5 mm in diameter which are unusual structural features on the broken surfaces of fractured australites generally. The rest of the posterior surface has prin- cipally a primary sculpture pattern of flow swurls showing schlieren, in places arranged in complex fold-like structures. Round to elon- gated pits averaging 1 mm in diameter, occur towards the edges of the posterior surface (Plate I, A). After removing soil constituents from the pits, their walls revealed a lacquer- like lustre produced by a process of natural solution-etching which has also brought out some of the fine schlieren on the pit walls. The general surface of the tektite at the level of the pit openings, however, shows a rather duller, almost sub-vitreous lustre. Flaked equatorial zone . — The flaked equatorial zone (Plate II, A and B) occurs circumferen- tially in the mid-regions of the specimen as viewed in side aspect. It is approximately 10 mm broad. The rim separating it from the 57 £2 X5 P £2 bC a ? O £i (Q a; c P . w — 53 2 p -p !/l o a T3 G cJ 0) s ■d G d S a 5 d « h. (0 OJD 0) o b ao ^ a 0) G-^ -G -p 1 ° w y p <^d y s d « “ y p j3 flj d .G pi .2 8 P O d +j Pt G y d 1 d y ffl CO 58 posterior surface (uppermost in Plate II > is sharply defined, but the anterior surface (lowermost in Plate II * and the flaked equa- torial zone merge more gradually into one another. Vertical, oblique, and less commonly mean- drine short grooves 2 mm long and 1 mm wide, and longer straight grooves up to 10 mm long and 1 mm wide, cut across the flaked equatorial zone, and mainly trend parallel with the breadth. Like the pits on the posterior surface, these grooves occasionally reveal fine schlieren cross- ing their walls at various angles. Anterior surface . — The anterior surface (Plate I. B), which was fully exposed to aerodynamical forces during high-speed flight through the earth’s atmosphere, remained directed forward down the flight path as long as the specimen was maintained in aerodynamic stability at the high entry velocities. Meandrine (some “vermicular”), straight, and curved grooves on the anterior surface vary in length and average about 1 mm in width (Plate I, B). Sometimes they are S-shaped in pian (left-hand side, Plate I. B), occasionally J- shaped. In cross sectional aspect they ere largely U-shaped. The walls of some plainly reveal a few small etch pits; more frequent are fine flow lines that cross the grooves at right angles, or obliquely, or sometimes trend parallel with their length. Occasional grooves are at a slightly lower level than others, and neigh- bouring grooves sometimes cross or merge with one another in places. A few depressions 3 mm across, 0.5 mm deep, and circular in outline (lower central portion, Plate I. B) are crater-like compared with smaller pits (right, central portion, Plate I, B). They sometimes carry smaller pits and occasional fine flow lines that cross their walls in random direc- tions. The asymmetrical silhouette of the anterior surface observed from one side aspect of Uic specimen (Plate II, B), evidently has arisen from weathering. This gives a flattening effect which is faintly discernible in plan aspect ( top portion of Plate I, B). Curvature and Volume The radii of curvature of the posterior (Rii) and of the anterior (Rk) surfaces were deter- mined from a silhouette tracing equivalent to a cross section containing the polar axis of the specimen (i.e. equivalent to an enlarged sil- houette trace of Plate II, A). The values so obtained are: Rit 4.0 cms (40.25 mm). Rf 3.7 cms (36.75 mm). The posterior surface is evenly curved in all directions radially outwards from the rear pole of the specimen. Its arc of curvature corresponds with that of a constructed circle with radius 4.0 cms. but only about 28 per cent, of the original curved surface remains after ablation, assuming the parent form was a sphere. The arc of curvature of the anterior surface was not even in all directions outwards from the front pole (cf. Plate II, B), so the radius of curvature was determined from a silhouette tracing of the most regular curvature shown in one side aspect (cf. Plate II. A). The area of the front surface (27rrh). where h is the height of the anterior cap above the rim of the specimen, was calculated as 58.13 cm- assuming that the area of the front surface is equivalent to the area of the cap of a smooth sphere. This represents the frontal area projected in the flight direction during the closing stages of the aerodynamic heating process. The height of the anterior cap is 2.5 cms, and the height of the posterior cap is 0.95 cms; together these constitute the depth or thickness (3.45 cms) of the australite. The volume of the specimen is 69.1 cm='. The volume of a sphere of australite glass wdth R 4.0 cms and of specific gravity the same as that (2.434) determined for the residual australite shape (large round core), was cal- culated as 268.1 cm^ This is not quite 21 times the volume of the largest round core so far recorded from the Port Campbell district some 1,500 miles distant in Victoria (Baker 1962b). Loss by Ablation If the parent form of the large round aus- tralite core from Graball, Western Australia was a sphere, or a spheroid close to a sphere then the volume of tektite glass lost (a) during entry at high speed, by ablation, and (b) during the few thousand years that the specimen has lain exposed to weathering on the earth’s sur- face, amounts to 199 cm'*, this being the differ- ence between the parent sphere volume and the volume of the residual australite romrd core; it represents a loss of approximately three quarters of the primary form. Thus, as a con- sequence of aerodynamic heating with attendant ablation and fusion stripping, followed by ter- restrial erosion, only 25.8 per cent of the original form remains. Allowing for the small chip lost by artificial fracturing (see central portion of Plate I. A), this value can be rounded-off at 26 per cent. Although it is impracticable to assign definite proportions to either of these processes causing loss of australite glass since initial entry into the atmosphere, largely because of the unknown loss by erosion, it is expected that the combined effects of ablation and fusion stripping and pos- sibly some stress exfoliation in the subsonic region of the flight path, significantly dominated loss by subsequent terrestrial weathering. Assum- ing. as a generality, that erosional loss in the Australian strewnfield amounted to at least ten per cent, of the tektite glass that reached the earth’s surface, it is apparent that over 70 per cent, of the Graball primary form was consumed by aerodynamic ablation. This is in accord with the calculated ablation losses of other primary spheres from which large round cores were produced (Baker 1962b). As measured graphically, and calculated from 2(RB) — D where Rn is the radius of curvature of the posterior surface, and D the depth of the specimen measured from the back to the front pole, the depth of ablation from the front pole of the reconstructed primary sphei*e to the front pole of the remnant australite core, is 45.5 mm. 59 PLATE II Bound australite core from Graball. Western Australia. ^ aspect showing flaked equatorial zone separating posterior (uppermost) from anterior (lowermost) surfaces. ® side aspect in a position normal to that of (A) above, showing flattened arc of curvature of anterior surface in lower left portion of the photograph. Most grooves crossing the flaked equatorial zone trend parallel with the line of flight which was in a direction from top to bottom of the photograph. Scale is in centimetres and millimetres. Photographs by Alfred A. Baker. 60 This is 56.9 per cent, of the original sphere diameter, and thus comparable with the average percentage value (56.5 per cent.) for ten round australite cores from Port Campbell, Victoria (Baker 1962b). Among other things, this is proportional to the aerodynamic heating (fide Dr. Dean R. Chapman, N.A.S.A., U.S.A.). Compared with the depth of ablation (7.5 mm) of a hollow australite of comparable diameter containing an eccentrically disposed internal cavity (Baker 1961), the depth of ablation along the polar axis of the solid australite core from Graball has been six times as great. The indica- tion is that the hollow form was therefore a much more effective dissipator of the frictional heat input during aerodynamically stable transit through the earth’s atmosphere. By comparison with the average for perfectly developed, excellently preserved australite but- tons having well-formed circumferential flanges (Baker 1962b), and with the value for the hollow australite which is also perfectly developed and very well preserved (Baker 1961), the aerody- namic heating (determined from A.H. — k/\ Rk, where Rf is the radius of curvature of the anterior surface) for the large, well-developed, reasonably well preserved, round australite core of solid tektite glass from Graball. Western Australia, is 0.520 (see Table I). TABLE I 1 Australltf Shape Typ«- J.orality Aerodynamic Heating JVrfpiit llatigMl hiittniis I’ort Canipbcll. 0-921)* \'irtoria Larfj*' roiiiifl con* (solid y:lass) i (irahall. Western Australia 0-.’>2() Larife round hollow f<»rm Horsham, Victoria 0-187 * This is an average value for 23 perfect australite buttons, the range for which was 0.86 to 1.09. The aerodynamic heating value for the hollow form is lowest, and equal to only one fifth that for the perfect australite buttons. The value for the large solid core is intermediate and its aerodynamic heat input was a little over half that for perfect buttons, approximately two and three quarter times greater than the input for the hollow round australite. Relative to the perfect flanged buttons, which are smaller, the solid round core from Graball has a larger mass per unit frontal area, hence, during high speed earthward flight, more of the aerodynamic heat received was dissipated, largely by radiation. The amount radiated, however, was considerably less than for the hollow round australite from Horsham. Conclusions The sculpture of the posterior surface of the large, round austrahte core from Graball, Western Australia is evidently primary and of extra-terrestrial origin; unlike most other large australites from the more westerly parts of the Australian strewnfield, it has been subjected to less severe weathering by terrestrial agents. The sculpture patterns of the anterior surface and the flaked equatorial zone are secondary. The curvature of the anterior surface was largely determined by aerodynamic sculpturing during frictional heating processes arising from high speed, aerodynamically stable transit through the earth’s atmosphere. Evidently all the fused glass from this secondary period of melting was shed almost immediately during flight. There is (i> no evidence on large cores such as this, that ring-wave formations were developed to give flow ridges and flow troughs as on buttons of smaller size (Baker 1962b, Plates XII to XIV), (ii) no evidence of flange glass remaining frozen-in circumferentially as on the flanged buttons, and (iii) no evidence that flange glass was present on landing and was subsequently removed by weathering. Accepting the parent form as having been a sphere of australite glass of 8.0 cms diameter, then more extensive loss of all the melt glass and vapourised glass during aerodynamic fusion has occurred to produce this large core than in forming flanged button-shaped australites from originally smaller spheres. The depth of ablation in the stagnation point (front polar) regions was proportionally much greater than for flanged buttons where all melt glass was not shed, because some was re-frozen in equatorial regions of the ablated smaller spheres to form circumferential flanges. Depths of ablation in forming perfect flanged australite buttons range from 5.5 mm to 16.5 mm, and average 9.5 mm. This is 47.3 per cent, of the diameters of the originally smaller spheres which ranged from 12.7 mm to 27.1 mm and averaged 20.1 mm in Port Campbell specimens (Baker 1962b). Depth of ablation (45.5 mm) of the sphere from which the large round core from Graball was produced, represents 56.9 per cent, of the ori- ginal sphere diameter, i.e. approximately 10 per cent. more. Since not all of the melt glass was shed in forming the australite buttons, and a relatively substantial amount went into the building of the circumferential flanges, there was an in- crease in the frontal area. This increase was equal to the area of the anterior surface of the newly generated circumferential flange. Con- sequently, flange formation led to reduced aero- dynamic heating ^fide Dr. Dean R. Chapman ( as an outcome of increased amounts of drag arising from increased frontal area. Such a process did not occur in the large core types of australites. so that the aerodynamic heat input was maintained at a steady rate as long as high enough velocities prevailed. This being so, greater aerodynamic heating leading to melting and ablation was experienced than in the phases of button generation when circumferential flanges were formed. The present sculpture pattern of grooves, craters and pits on the anterior surface, and probably also on the flaked equatorial zone, is fundamentally a result of natural solution- etching developed after landing on earth surface. It is not easy to assess how far a process of exfoliation on cooling may have affected the anterior surface during the final phases of flight, i.e. between the end of the aerodynamic heat- ing stage and the time of impact with the earth. 61 the specimen was collected from a region where soil deflation processes operated to a significant degree in recent geological times, and australites are usually discovered with their anterior sur- faces uppermost, a position found by experi- mentation to be their stable position of rest on the earth’s surface. The most dominant cause of weathering was evidently chemical, with soil solutions acting partly differentially to bring about some direc- tionalized and some random natural solution- etching. The meandrine character of certain of the grooves might be indicative of confined biochemical reactions associated with plant rootlets lying in contact with the top of the specimen as it lay embedded in soil. Such re- actions are expected to be more prevalent on the uppermost (anterior) than on the lower- most (posterior) surfaces of australites buried in soils, even though, as deflation progressed in some areas, the posterior surface remained longer in contact with soils on exposure. This may have been interconnected with, or possibly completely superseded by, subsequent spallation from diurnal temperature changes while resting on the earth’s surface. After the australite had passed from (1) the phase of flight where ultrasupersonic and then lesser supersonic speeds prevailed, through (2) the transition zone of transonic speeds, and into (3) the subsonic region where much lower speeds supervened for the final few seconds of earth- ward trajectory, processes of fusion and ablation created by frictional heating had ceased to operate. The last-formed thin skin (1 mm and under) of the forwardly directed surface that was heated and softened at the end of this stage of flight, then cooled rapidly in the lower- most region of the atmosphere. It was not soft on impact with the earth, as deduced from the available evidence, but compared with the underlying glass, it would have been in a relatively highly stressed condition. Such a crust would tend to exfoliate in thin layers from the anterior surface, either before, and/or after landing. This would result in exposure of a sub-surface of the anterior surface to which the secondary process of aerodynamic heating had not penetrated. Gradual chemical attack of this surface by terrestrial agents, mainly etchants in moist soils, further modified the anterior surface to produce its present sculpture nattern. If the nature of the exfoliation deter- mined the initial trends of solution-etching, it is no longer evident because of the degree to which etching has advanced. The indications on the Graball round aus- tralite core are that mechanical agents causing modification by abrasion played no major role in the process of terrestrial erosion, unless the somewhat flatter appearance of the anterior surface as seen in one aspect (Plate II, B, bottom left-hand side) is a faceted area arising from abrasion by wind-driven sand and soil. There is no proof that such faceting occurred, although Acknowledgments The author is grateful to Mr. Alfred A. Baker of the Geology Department, University of Mel- bourne, for photographing the specimen, and to Mrs. A. Alexieff for preparing “Artificial Stone” casts of the australite. References Baker. G. (1961). — A perfectly developed hollow austra- lite. Amer. J. Sci. 259: 791-800. Baker, G. (1962o). — The largest known australite and three smaller specimens from Warralakin, Western Australia. J. Roy. Soc. W. Aust. 45: 12-17. Baker. G. (1962b). — Volumenbeziehungen von wohler- haltenen Australit-Knopfen, -Linsen und -Kernen zu ihren primaren Formen. Chem. d. Erde 21 (3/4): 269-320. Fenner. C. (1955). — Australites Part VI. Some notes on unusually large australites. Trans. Roy. Soc. S. Aust. 78: 88-91. I 62 8. — A Review of the Gekkonid Lizard Genus Heteronota Gray, with a Description of a New Species from Western Australia By Arnold G. Kluge* Manuscript received — 21st August, 1962. A review of the nominal species of the gek- konid lizard genus Hetercnota reveals that binoei has been until present the only recog- nizable member in the genus. A new species of Heteronota is described from the North-West Natural Region of Western Australia. This new species appears to be restricted to subterranean cavities. Introduction Collections of herpetological material from the more isolated regions of Australia have greatly increased within recent years and doubt- less will provide the basis for the description of many new species. It must be emphasized, how- ever, in view of the large number of names already applied to various parts of the popula- tions and the almost complete lack of know- ledge of relationships, that a generic revision or review of the nominal species should accompany the recognition of these novelties. As an example of the case in point, the extremely variable dorsal body scalation and colour pat- tern found in the gekkonid lizard genus Hetero- nota has led to considerable confusion, with the description of a large number of species. The following generic diagnosis and review of the nominal species of Heteronota is a necessary preliminary to the description of a new species in the genus. Diagnosis of the Genus Heteronota Heteronota can be distinguished from all other genera of the Gekkonidae by the following combination of characters: terrestrial species with long slender digits; distal phalangeal elements slightly angulate: subdigital lamellae large, rectangular and swollen; a pair of sub- apical plates; two rows of scales covering sides of digits; subcaudals greatly enlarged trans- versely: dorsal body scales heterogeneous, con- sisting of large trihedral tubercles in regular or irregular longitudinal rows and separated by small smooth or keeled conical granules; ventral body scales large, smooth and imbricate; a short angular series of preanal pores in males; cloacal sacs in males and females; a single pair of cloacal bones in males; mental and post- mentals large; primary postmentals in contact behind mental: rostral and first supralataial border nostril; pupil with emarginations on both anterior and posterior margins. * Department of Biology, University of Southern Cali- fornia, Los Angeles 7, California. Post-graduate Fulbrlght Scholar, 1961-2, Department of Zoology, University of Western Australia. Nominal Species of Heteronota The present review of nominal species in- cludes all described forms known to be based on examples of Heteronota (Group A) and also those which have been erroneously referred to the genus (Group B). Each name is presented in its original form and followed by a citation of the original description and type locality. The present status of the species, where it differs from the original, follows the citation with a reference to the author (s) who made the initial change. If no reference is given the present author assumes the responsibility for the synonymy. The generic name which I re- gard as being applicable to the genus is placed in square brackets when it differs from that of the author who made the initial change. Fol- lowing both Groups A and B there is a discus- sion elucidating some of the synonymy. Group A Heteronota binoei Gray 1845, Cat. Lizards Brit. Mus., p. 174, type locality: Houtman’s Abrollos. Western Australia. Type species of Heter- onota by elimination. Eublepharis derbianus Gray 1845, Cat. Lizards Brit. Mus., p. 274, type locality: Port Essington, North- ern Territory = Heteronota binoei Gray fide Gunther 1867, Ann. Mag. Nat. Hist., ser. 3, vol. 20, p. 50 and Gray 1867, The Lizards of Australia and New Zealand, London, p. 6. Hoplodactylus ( Pentadactylus ) australis Steindachner 1867, Reise der Novara (Reptilien), p. 18, type locality: New South Wales - Heter- onota binoei Gray fide Gunther 1867, Ann. Mag. Nat. Hist., ser. 3, vol. 20. p. 50 and Gray 1867, The Lizards of Australia and New Zealand. London, p. 6. Phyllodactylus anomalus Peters 1867, Mber, Akad. Wiss. Berlin, p. 14. type locality: Rockhamoton. Queensland -- Heteronota derbiana fGray) fide Boulenger 1885. Cat. Lizards Brit. Mus., vol. 1, p. 75. Neither Gunther (1867) nor Gray (1867) gave any evidence for synonymizing Eublepharis derbianus with Heteronota binoei and a number of later workers continued to recognize both species (Boulenger 1885. Oudemans 1894 and Zietz 1920). Lucas and Frost (1896), Procter (1923), Kinghorn (1924), and Loveridge (1934) studied large series of H. binoei and supposed E. derbianus and have confirmed Gunther’s and Gray’s original action I believe beyond any reasonable doubt. Group B Heteronata kendallii Gray 1845, Cat. Lizards Brit. Mus.. p. 174, type locality: Borneo = Gonatodes \Cnemaspis] kendalli (Gray) fide Boulenger 1885, Cat. Lizards Brit. Mus., vol. 1, p. 63. 63 Heteronota pelagica Girard 1857, Proc. Acad. Nat. Sci. Philad.. p. 197, type locality: Feejee and Navigator Islands — Gymnodactylus [Cyrtodactylus] pelagicus (Girard) fide Boulenger 1885, Cat. Lizards Brit. Mus.. vol. 1. p. 40. Gyrnnodactylus (Heteronota) arfakianus Meyer 1874. Mber. Akad. Wiss. Berlin, p. 129. type locality: New Guinea = Gymnodactylus [Cyrtodactylus\ pelagicus (Girard) fide Boulenger 1885, Cat. Lizards Brit. Mus., vol. 1. p. 40. Heteronota jasciata Macleay 1877, Proc. Linn. Soc. N.S.W.. vol. 2. pt. 1. p. 100, type locality: Hall Sound. New Guinea — Gymnodactylus [Cyrtodactylus] heteronotus Boulenger 1885, Cat. Lizards Brit. Mus.. vol. 1. p. 41. Hetoronota marmorata Macleay 1877, Proc. Linn. Soc. N.S.W.. vol. 2, pt. 1, p. 100, type locality: Pitzroy Island and Endeavour River, Queensland = Gymnodactylus [Cyrto- dactylus} cheverti Boulenger 1885, Cat. Lizards Brit. Mus., vol. 1, p. 41. Heteronota ehoracensis Macleay 1877. Proc. Linn. Soc. N.S.W. vol. 2, pt. 1. p. 101. type locality: Cape York, Queensland = Cyrtodactylus pelagicus (Girard). Heteronota walshi Kinghorn 1931, Rec. Aust. Mils., vol. 18, no. 5. p. 268, type locality; Boggabri, New South Wales ~ Pliyllurus walshi ( Kinghorn ) . Heteronota binoei remained as the only species by Gray’s original definition of that genus, when Boulenger (1885) referred kendalli to the genus Gonatodes. Heteronota binoei is therefore con- sidered the type species by elimination (see International Code of Zoological Nomenclature, 1961, Rec. 69B, 3). Boulenger (1885) referred Heteronota fasciata and H. marmorata to the genus Gymnodactylus mow in part Cyrtodactylus, see Underwood 1954) apparently solely on the basis of Macleay’s original descriptions. Boulenger was forced to provide new specific names (heteronotus and cheverti) for both species as Macleay’s names Pig. 1 . — The distribution of Heteronota spelea. were preoccupied in Gymnodactylus. Loveridge (1934) has since synonymized G. heteronotus and G. cheverti with C. pelagicus. Loveridge (1934) referred Heteronota ebora- ce7isis to the synonomy of H. binoei, however, its affinities appear to lie within the genus Cyrtodactylus. Specimens of Cyrtodactylus pela- gicus from the Cape York Peninsula, Queens- land. agree in all respects with the original description of H. eboracensis. The holotype of Heteronota walshi has not been examined, but, judging from the original description alone, the species is clearly referable to the genus Phyllurus. From this review of nominal species, Heteronota binoei is considered to be the only recognizable species in the genus. While study- ing the large series of H. binoei in the collec- tions of the Western Australian Museum (W.A.M.) and the Department of Zoology of the University of Western Australia, eight speci- mens from several localities (see Pig. 1) in the North-West Natural Region of Western Aus- tralia (Clarke 1926) have proved to be extremely different and are here described as a new species. All of the recently collected material of this species, where accurate locality and habitat in- formation are available, indicates that it is restricted to mines and natural subterranean cavities and it is therefore described as: Heteronota spelea, sp. nov. Holotype: W.A.M. R12638: collected in “Pro- phecy West” mine at Bamboo Creek. Mar- ble Bar District, Western Australia, by A. M. Douglas and W. D. L. Ride on October 12 or 13, 1957. Paratypes: W.A.M. R12639-40; also collected in the “Prophecy West” mine and an unnamed adit at Bamboo Creek by A. M. Douglas and W. D. L. Ride on October 12 or 13. 1957. Diagnosis: Heteronota spelea differs from H. binoei in possessing regular longitudinal rows of very small trihedral tubercles on the dorsum of the body (see Table D and four distinct brown bands on the body and nine to ten on the tail (Fig. 2). In H. binoei the tubercles are larger and more randomly scattered and the colour pattern is extremely variable. Description of holotype; Head somewhat flattened: snout long; rostral rectangular, twice as broad as deep; dorsomedian rostral crease one-half height of rostral; nostril moderately large, directed posterolaterally, surrounded by rostral, first supralabial, one postnasal and two supranasals; anterior supranasal greatly enlarg- ed, meeting counterpart on midline: scales pos- terior to supranasals greatly enlarged; loreal region strongly concave; 11/12 (right and left sides respectively) scales between postnasal and anterior margin of orbit; dorsal surface of snout slightly concave; supralabials 7/8 (from rostral to immediately below vertical pupil) ; fourteen scales between centrolateral margins of orbit (excluding supraciliaries and supraocular gran- ules); external ear opening a small obscure slit at level of angle of jaw; mental triangular, much broader than long; primary postmentals meet on midline, greatly enlarged, almost twice as 64 TABLE I The range of variation of certain meristic characters of Heteronota spelea and H. binoei from the zone of sympatry. indicates the number of specimens examined. Kju'leii (S*) binoei Ml. Kd<;ar .Marble bar (2(,l*) (lO-^) NmnbPi- of suj)ra- flto 9 (7-0) 5 to 0 (5-4) 5 to 0 (5-5) Niniihor of scales between post- nasals and ])re- oeiilar granules 10 to l:l (11 -2) Stoll (9-0) 8 to 11 (9-8) Niimb('r of keeled scales in ]>rini- ary }>araverte- bra! row l)e- tween axilla an -F Si/iinfhopi^is rmsKiruudaUi ((.iould) -F Fatnilv Peramelidae Maerofi.s hufotis (Reid) + PenimelpH hmmu'mi'illel (^uoy and (iainiard Family .Macrojiodidae liptfunqia lemeuri (Qiiov' and (iaimard) ... lieltorKjia pfuicilhiln Oray .. . -F MarrojniH sj). + (.)rdcr Hodentia J*'ainily Muridae \otomyn mitchelU (Otxilby) .. + + P.'tpinlomi/K rau'Vniniie Troutditon JjpqqaiUmt hprmnrmxhurqenHia (Waite) Jjpporilhoi coiulifor (Sturt) .... LppnriUux apiealU ((loullt)(4()) 27-0 (5-1) 02-8 (2-2) U)-2 (1 2) 4-8 (0-7) 1. ocvbhnialiH 17 (14) 40-7 (:L5) :l8-2 (2-2) (H *0 (4-5) 29-5 (:i-9) Of) -4 (5-7) 114U-2) 4-0 ((t-7> 1. IcvIh 7 (()) (1-0) :9) (> (0-9) 52-H (4-9) 21-2 (1-.5) 02-7 (2-0) 10-7 (1 -4) 4-8 (()•«) rerfehralis 10 (0) :i9-5 (2 '9) 24-0 (2-S) .'iiM (4 5) 19-2 (2 -.5) (50 (5 (4-4) n- 2 (l- 2 ) 5-2 (O-O) liieciKsnuux U (10) 4r>-l (2-4) 38-2 (1 4) 49-1 (2-0) 1:M (2-2) 07-8 (2 -2) 12-5 (0-9) 5-2 (0-8) 89 Measurements The following measurements were made on all specimens: total length, i.e. head plus trunk 'including neck) plus tail (measured from the vent, not the post-pygal constriction as was apparently done by De Vis when measuring the type of levis) hind leg. and width of head and tail (all to the nearest 0.5 mm* ; and the length of ear aperture and the horizontal diameter of visible part of eye (both to the nearest 0.1 mm). These data were expressed as ratios; means and standard deviations for each taxon are set out in Tables I and II. The eight taxa of Nephrurus inhabiting Western Australia are listed in the tables in a generaly north-south sequence, revealing geo- graphical trends in the ratios which transcend species limits. The ratios, ear aperture to trunk and width of head to its length, decrease as one goes south; whereas diameter of eye and length of hind leg increase. Clinal variation is especially marked in the ratio, length of ear aperture to diameter of eye: this ratio yields a good separation of the northern and southern races of wheeleri, as illustrated in Figure 2. Ideally ratios should remain constant through- out life, a condition that is not always fulfilled in the present genus, where the growth of appen- dages in adults tends to slow down with respect to the trunk. For example, the ratio hind leg to trunk in levis decreases slightly after a trunk length of 55 mm is attained (see Figure 3). The proportions within an appendage, however, seem to be unaffected by age, e.g. width to length of head and tail, and length of ear aperture to diameter of eye. TABLE II Mea7i ratio: width to length of head and tail, and length of ear aperture to horizontal diameter of eye (with standard deviation in brackets ). Width to Width to Ear Lfiigth of Length of Aperture lo Head Tail ]•:>•(* diaper 1 - 05 0-42 0-75 II'. rinffnii 0-96 (0 06) 0-4:i (0 06) 0’62 (0-U)) If. u'heelpfi 0-95 (0-04) 0-:B6 (0-06) 0-50 (0-07) 1. pilharennU 0-96 (0-04) 0-44 (0-09) 0-46 (0-09) 1. (ircidentaHii 0-93 (0-05) 0-44 (0-07) 0-41 (O-OS) /. Irris- 0-92 (0-05) 0-40 (O-Oli) 0 -45 (0 07) I'l'rfphriiH'i 0-S7 (0-05) 0-31 (0-04) 0-47 (0-06) Iiiprt.'ishinix 0-85 (0-04) 0-27 (005) 0-42 (0-08) Discussion To a large extent the various species of Nephrurus are still allopatric, so that over much of the State any one region is occupied by a single form of the genus (see map, Fig. 1). In the Kimberleys asper alone has been found; in the De Grey drainage, only levis pilbarensis: in the Fortescue, wheeleri cinctus: in the East Murchison, nominate wheeleri: and in the Carnarvon Basin, levis occidentalis. It is only in the Great Victoria Desert that two forms (viz. levis levis and laevissimus) are widely sympatric. Elsewhere sympatry has been established at Mundiwindi and Warburton Range, between vertebralis and respectively wheeleri cinctus and levis levis. The status of vertebralis has only been re- cently settled. Glauert (1961) treated it as a colour variant (his No. 2) of levis. At first the present writer too was inclined to regard it as no more than a well-marked race of levis. The available specimens at that time all came from the country immediately east and south-east of the range of levis occidentalis. As nominate levis occurred further to the east, vertebralis appeared to be not only allopatric to occidentalis and nominate levis but also geographically inter- mediate between them. But only in minor ways was vertebralis morphologically intermediate between these forms, and the writer began to doubt the propriety of including it in levis. These doubts inci’eased when a specimen of vertebralis was collected at Yuin, which is only 30 miles east of the straight line between Mul- lewa and Nai’ryer, at both of which a levis occidentalis had been collected, neither speci- men showing any intergradation with vertebralis. The problem was finally solved last year, when Mr. Mark de Graaf collected a specimen each of vertebralis and levis near the Warburton Range Mission. This absolute sympatry neces- sitated the promotion of vertebralis to a full species. Occidentalis and the closely related pilbarensis are themselves very distinct from nominate levis, from which they are geographically separated by a large area of heavy, frequently stony, soils which are dominated by mulga and occupied by wheeleri. However, levis could well have a continuous distribution in the far interior of the state. Although Well 15 on the Canning Stock Route (a pilbarensis locality) is 300 miles WNW of Warburton Range (the nearest locality of nominate levis), the inter- vening desert is almost certainly inhabited by some form of the species. References Glauert, L. (1961). — “A Handbook of the Lizards of Western Australia.” (W.A. Naturalists’ Club: Perth.) Kinghorn, J. R. (1924). — Reptiles and Batrachians from South and south-west Australia. Rec. Aust. Mus. 14: 163-183. Loveridge. A. (1934). — Australian reptiles in the Mus- eum of Comparative Zoology, Cambridge, Massachusetts. Bull. Mus. Comp. Zool. 77 ( 6 ). Lucas, A. H. S., and Frost, C. (1896). — Reptilia in ‘‘Re- port on the Work of the Horn Scientific Expedition to Central Australia.” (Dulau: London. Melville, Mullen, and Slade: Mel- bourne.) 90 13. — The Angler-fish Ceratias holboelli from Western Australian Waters By B. K. Bowen* Manuscript received — 11th June, 1963 The first recorded adult specimen of the deep- sea angler-fish Ceratias holhoelli from Australian water was taken from a whale’s stomach at Albany in 1957. The specimen is figured and described. On August 31st, 1957, a male sperm whale 38.5 ft in length, was caught at 35° 34' S., 117° 36' E. (38 miles south-west of Albany). When the stomach of the whale was opened on the flensing deck at Albany, it was found to contain a specimen of the deep-sea angler-fish Ceratias holboelli Kroyer (Fig. 1). The specimen was a fully metamorphosed female without an attached male. Its standard length was 48 cm. This specimen of C. holboelli is the twenty- second adult female recorded throughout the world and the first to be taken in Australian waters. Bertelsen (1951) records all specimens known up to 1951 and more modern records are by Krefft (1954) and Van Utrecht (1957). The only previous record of this species from Aus- * Fisheries Department, Perth. Western Australia, tralia is that of a female larva taken on Febru- ary 25th, 1929, at Dana Station No. 3665, loca- tion 29° 37' 5" S., 156° 46' E., off the east coast of Australia. The specimen was badly lacerated on the right side in the region of the exhalent aperture and in addition all fin rays were broken and lacked their distal ends. The specimen v^as otherwise in good condition. The eyes had completely regressed but could be exposed by an incision. Bertelsen (1951) has suggested that there are two sub-species of C. holboelli, C.h. holboelli from the northern hemisphere and C.h. tentacu- latus from the southern hemisphere. C.h. holboelli possesses a single escal filament, while C.h. tentaculatus possesses two filaments which may be further branched. Unfortunately our specimen (Western Australian Museum No. P4266) has a damaged tip to the escal bulb and dees not provide further information in this regard. Pig. 1. — Female deep-sea angler-fish Ceratias holboelli tiken from the stomach of a sperm whale at Albany. (Photograph: Government Printer.) 91 The measurements shown in Table I were taken in conformity with those taken by Clarke (1950). TABLE I Measurements (in cm) of Ceratias holboelli (W.A.M. No. P4266) Standard length 48.0 Length of illicum ... 8.8 Height of caruncle, stalk included 1.6 Snout to base of tentacle 9.5 Snout to dorsal fin 35.0 Snout to anal fin 37.3 Snout to base of pectoral fin 17.6 Base of tentacle on back to dorsal fin .... 5.6 Posterior caruncle to dorsal fin 4.4 Length of lower jaw 7.6 Greatest depth of body 17.2 Depth of caudal peduncle 4.2 Base of tentacle on head to base of tentacle on back 20.0 Dorsal fin to cadual fin 8.1 Anal fin to caudal fin 7.1 Base of anal fin 2.6 All of the measurements with the exception of the height of the caruncles, fit the allometric growth graphs drawn by Clarke (1950). The caruncles measure 1.6 cm in height, whereas in the specimens considered by Clarke (1950) not one of the measurements exceeds 0.75 cm. Clarke has shown that the maximum height of the caruncles is reached when the fish is about 12 cm standard length, after which they regress in absolute size. He suggests that the onset of this enantiometric (i.e. absolute negative) growth is associated with a crisis in the development of the fish, which is probably the attainment of sexual maturity. The caruncles of the new specimen, which is a mature fish, also differ in Fig. 2.— The caruncles showing a division into a head and stalk region. Diameter of field 4.3 cms. (Del. M. Walsh). Fig. 3. — Caruncle length plotted against a standard length on a double logarithmic scale. (The numbers of the points are those allocated to specimens by Clarke 1950. The new specimen is designated N.) shape from other mature fish studied. They are divided into a head and stalk region similar to the juvenile rather than the regressed club- like form of the mature fish (Fig. 2). If the measurement for the new specimen is incorporated in Clarke’s (1950) allometric growth graph for the “length of caruncle”, (Fig. 3), it will be seen that it lies to the right of a projection of the line derived from the plotting of the caruncle heights of immature fish and above the line obtained for mature fish. Thus the point can not be definitely associated with either line. Also, it is impossible to state whether or not the absolute size of the caruncles decreased after metamorphosis. However, the fact that the caruncles still resemble the juvenile form points to the probability that growth has been progressive throughout the life of the fish. It is apparent, therefore, that the growth of the caruncles of the new specimen has differed from those studied by Clarke (1950) and confuses his hypothesis that they reach maximum size at sexual maturity. It could be argued that the females mature over a length range and that on maturity the caruncles cease to grow in length but gradually thicken to the club-like form. This would explain the differences in caruncle lengths and shapes. However, for such an explanation to hold true the range of maturity would have to be approximately 6 to 19 cm. This seems unlikely and, therefore, the growth of the caruncles of Ceratias holboelli remains an inter- esting problem. References Bertelsen, E. (1951).— The ceratioid fishes. Dana Rep. 39. Clarke, R. (1950).— The bathypelagic angler fish Cera- tias holboelli. "Discovery” Rep. 26: 1-33. Krefft. G. (1954).— Ein Tiefseeangler aus islandichen Gewassern. Fischereweilt 6: 78-79. Van Utrecht, W. L. (1957). — Een diepzeevis uit de maag van een potvis. Visserji-Nieuws 5: 72-73. 92 14. — The Callionymidae of Western Australia (Pisces) By G. P. Mees * Manuscript received — 19th March, 1963 In Western Australian waters the Calliony- midae are represented by two genera and nine species: Dactylopus with one species and Cal- lionymus with eight species. A key for their identification is given and their distribution and synonymy are discussed. Introduction Thanks to McCulloch’s fl923, 1926) beautifully illustrated revisions, the Callionymidae of Aus- tralia are systematically an easily accessible and comparatively well-known group. However, sev- eral of the species described by McCulloch were known from single specimens only, and their known distribution was limited to the type locali- ties. Prom Western Australia. McCulloch '1929- 1930) knew four species: Dactylopus dactylopus, Callionymus calauropoinus, C. calcaratus, and C. apricus. In subsequent years Whitley (1944. 1945) added C. goodladi (a new species), and C. papilio to the State list, bringing the total number to six (Whitley 1948b). Very recently I was able to record another two species (Mees 1959). Even this increased number of eight species does not complete the state list, for amongst material recently received by the West- ern Australian Museum are several specimens of yet another species. The purpose of this paper is to give a key to the species of Callionymidae known from Western Australia, as well as additional informa- tion, mainly on their distribution. It has further been possible to synonymise two names which had been given as a consequence of insufficient knowledge of sexual dimorphism. It is possible that for some of the species listed here older names will be found to be avail- able when Australian material is compared with species described from the East Indies and the Philippines. Major revisional work, however, goes beyond the scope of this small contribution. The bulk of the material this study is based on was provided by Mr. R. J. McKay of the Fisheries Department, and by Messrs. W. and W. Poole, owners of the “Bluefin”. Mr. G. Mack, director of the Queensland Museum, Brisbane, sent on loan a cotype of Callionymus grossi Ogilby and specimens of C. limiceps Ogilby. Mr. I. S. R. Munro, of the C.S.I.R.O. Laboratory ot Fisheries and Oceanography, Cronulla. N.S.W., sent on loan his whole collection of Calliony- midae which includes many specimens from Western Australia. Mr. G. Palmer allowed me. during a visit to the British Museum (Natural History) in May 1962, to examine the specimen of C. grossi from the Monte Bello Islands recently recorded by him. Mr. T. D. Scott, * Western Australian Museum, Perth, Western Australia. South Australian Museum, provided some speci- mens from South Australia, amongst w'hich Callionymus papilio. a species of which no material from Western Australia was available. Mr. G. P. Whitley gave information on material of C. calcaratus in the Australian Museum. Sydney. In the list of material examined, numbers pre- ceded by a P refer to specimens in the collection of the Western Australian Museum, numbers preceded by A and C are in the C.S.I.R.O. collec- tion, Cronulla. Key to the Genera known from Western Australia la. Spine and first ray of ventrals separated from the rest of the fin .... ... Dactylopus b. Ventrals without detached rays Callionymus Genus Dactylopus Gill * Dactylopus Gill. Proc. Acad. Nat. Sci. Phllad.. 1859, p. 130 — type by monotypy. D[actylopus\ Bennetti Gill — Callionymus dactylopus Valenciennes. Vulsus Gunther, Cat. Fish. Brit. Mus. Ill, 1861. p. 151 — nomen novum for Dactylopus Gill, allegedly preoccupied. Characterised by the detached spine and first ray of the ventral fin. Though this character is very useful and convenient for identifying the species, it seems hardly of generic value; how- ever, the preopercular spine is also different from that of any other species I have seen, and as the genus has now been generally accepted for a century, I prefer to maintain it. Only one species known. Dactylopus dactylopus (Valenciennes) Callionymus dactylopus Valenciennes, in Cuvier & Valenciennes. Hist. Nat. Poiss. XII. 1837. p. 232 — no locality. Dactylopus Bennetti Gill, Proc. Acad. Nat. Sci. Philad.. 1859, p. 130 — nomen novum for Callionymus dactylopus Valenciennes, proposed to avoid tautonomy. Dactylopus dactylopus; Ogilby. Ann. Qd Mus. 9. 1908. p. 38 (Moreton Bay, Queensland); Ogilby, Proc. Roy. Soc. Qd 23, 1910, p. 46 (Stradbroke Island and Wynnum in South-Eastern Queensland, and Moreton Bay. Queens- land); McCulloch, Zool. Res. Endeavour III, 1915, p. 149. pi. XXVIII (Shark Bay. Western Australia, and off Hervey Bay, Queensland); McCulloch & Whitley, Mem. Qd Mus. 8, 1925. p. 173 (Queensland: Moreton Bay. Stradbroke Island, Wynnum. off Hervey Bay); McCullQCh, Mem. Aust. Mus. 5. 1929. p. 337 (Queensland, Western Australia); Whitley, W. Aust. Fish. Dept.. Fish. Bull. 2, 1948, p. 27 (Western Australia); de Beaufort & Chapman, Fish. Indo-Aust. Arch IX. 1951. p. 80 (Western Australia. Queensland). Diagnostic characters. D IV-8L A Tg, P ii.15.ii or ii.15.iii, or ii.16.ii, sometimes i.l5.ii or i.l6.iii, V I. 1-4, C ii.7.ii or ii.7.iii; soft rays of dorsal fin divided except the first one which may be * Synonymy throughout this paper is confined to Australian records. 93 (a) (b) (c) Fig. 1.— Right preopercular spines, (a) Dactylopus dactylopus (P 5263, standard length 129 mm), (b) Callionvmus calauTopomus (P 4563, s. 1. 180 mm), (c) C. calcaratus (P 5392, s. 1. 154 mm), (d) C. goodladi (P 5440 s 1 118 mm) 5400, s. 1. 128 mm), (f) C. limiceps (P 5407, s. 1. 131 mm), (g) C. papilio (F 3076, s. 1. 62 mm), (h) C. phasis {from literature), (i) C. rameus (P 5431, s. 1. 129 mm). The spines of C. calauropomus and C. phasis which are curved upwards, are shown in lateral view, the others in dorsal view. 2i x natural size either simple or divided: origin of D 1 in advance of a line connecting the gill openings; preoper- cular spine with on each side three to five hooks (Fig. la); usually first and second rays of pectoral undivided, sometimes only the first. Distribution. A widely distributed species, known from the Philippines, the East Indies, Queensland and Western Australia. In Western Australia known from Exmouth Gulf, Shark Bay, and the Perth area, from Scarborough to Point Peron. The species apparently reproduces at City Beach: in the aquarium of Mr. V. A. Dawson of Como I have seen several live specimens of less than 5 cm length, caught on a sandy bottom in shallow water at that locality. Material examined, 43 specimens, varying in standard length from 64 to 150 mm. Exmouth Gulf (C 2319, C 2342, C 2363, C 2364, C 2381, C 2382, C 2383, P 4378, P 4952, P 5102 (3 sp.), P 5352, P 5367 (2 sp.) P 5368, P 5422), Shark Bay (P 4368, P 4375 (2 sp.), P 5451, P 5454), north of Peron Plats, Shark Bay (P 5268), between Kck’s Island and Point Quobba, Shark Bay

. South Head. Mary River. Q. - A 1142. A 1143. A 1144. A 1145 >. Tangalocma Point. Moreton Bay. Q. A 1030’. ofi Point Lookout. Stradbroke Island. Q. 'A 1035. A 1036 ^ Tween Heads. N.5.W. -C 2060 h Wallis Lake. NS.W. -A 861'. Princess Royal Harbour. N.S.W. 'A 856. A 857 . Twofold Bay. N.S.W. (C 2846 k Callionymus soodladi ‘ Whitley * CalliurzcrizK^s goodlad’. VTiiitlev. Alls*. Zool. 10. 1?44. p. 270 — Chevue Be-acli. Albany cis'ric*. Wes’ern Aus- tralia. Ca;.':urtc.^i;AL5 coodJac:. W Aust. Fisn. Dept. Fish. Bun. 2."l54£. p. 27 : Western Austraua'. Diagnostic characters. D A Ti. P i:14.: or ii.lS.i or ii.l5ui or ::.15.iii or ii.l6i. V 1.5. C ii.T.ii or il.7.hi: only last ray of dorsal fin divided; precpercular spine straight, spear-like, with about 8-10 small antrorse teeth on the inside, and with near its base a large antrorse spine on the outside • Fig. Id > ; snout large, broad and prominent, twice length of eye; origin of first dorsal nearly on a hne with gill openings; first and second pectoral rays undivided. Distribution. As far as hitherto known con- fined to Western Australia, where recorded from the south coast; Cheyne Beach and near Michaelmas Island. King George Sotind. and from the west coast: Cockbum Sound. Shark Bay ‘entrance to South Passage, and without precise locality * and Exmcuth Gulf. This distri- bution Is unusual inasmuch as it includes bctn tropical and temperate waters. Material examined, 27 specimens var3hng in standard lensrth from 81 to 160 mm. Exmouih Gulf 'P 4377. P53o5'. Shark Bay P2541. P2542. P2543. P 4360. P 4369 4 sp.‘. P 5438. P 5439. P 5440 P 5441. P 5442 * . entrance to South Passage. Shark Bay ^P 4966 k Cockbum Sound ‘ A 1359. A 1361. A 1362 k 'Cockbum Sound yro'H'. Ogilby Proc Roy. Soc. Qb 23. *9*1 p. 43 — Mor'-yor. Ba.y Callionymiu . nartctui McCUi*oczi. Res. Ende&Tcur V. I'M p :^<7. p: li;— 12 TTs::es scu*n- east from Cap^ Capr.oorr. Qu^.oslar-cJ Csyr-.—.coro:. ? crc^.*-:: McC'niccb. Hec Aus: M-is *4. 1523. r H Caue M:r^:ou McCunccb dc Wb:::ey. Mem Qc M'us :. 152S. ? 173 Mcre-tou Bay. Cape ilcretcu McCuilocn, B::! R.es rmcesTcur V 1926. ?. 155 ir. '^ey ■ no Iccaliry' Cay.io'zy'n.iii c'05r.. McCullccn Au5i Mus Mem. 5. 1929. p. 333 Queensland ; Scnmrz. Bml ^ S Mus f-i-3 Mcnoe EeUc •Lini Marsnai:. Iin-rnvGloeica.: Sz~.es p. 5 Miretcn Bay cn Pams Pom: CoKm -oc ^ TO - i O'O*'— 7".-: Mees TV A'ls: ? 15r5 p . 5 SnsJTi Bay CalJiumco:’ .f W-iTley Mamne Fisnes It. 1562 •70.= oro 2 ^96^. p . ■='.-2 HiFt. ' 13 1562 Cc'.'.icr.v'^. •— i SZ' BZZ5'Z3.TLe L. 1951. CG-tion^’r; CZ Fisn. d-=?t . t ; Diagnostic characters. D A 85 in one specimen. 7 ; k ? n.l3.iii or ii.l4 ii or ii.l5-i. V i^. C ii.7.:i:: only last ray o: dorsal hn ciyndeG; precpercular spine straight, spear-like. serrated cn The inside with about 10-18 small antrorse teeth, and with on the outside near its basis an antrorse spine Fig. le ; origin of first dorsal on a line with gill openings, first and second rays of pectoral undi'.uced. interesting features of this species are that both sexes have the rays of the first dorsal fin elongated; and the sexual dimorphism in snout length see disc’usslcn Distribution. Known from Queensland and Western Australia. In Queensland recorded from Moreton Bay typs locality; also Marsna.1 1951 . and from Cape Capricorn type Icxiality of ria6-utii-s . Marshall's statement that tne type cf nasTAti^-s was trawled cr Cape Moreton. South Queensland, m'ust be a slip. In Western Australia the species is known from Shark Bay. Exmcuth Gulf and the Mcnte Bello Islands Disc’jssion One of the mam diagnostic features given by McCullcch 1926 in his key to and cescription of C. oosuius. which was oasec on a single individual- is the long snout. Mar- shall 1951 had six specimens and as he makes no comment to the contrary, it can oe asrumed that they all have a snout tnuch longer tnan the eye. Two specimens from :rhark Bay. re- ported upon by me Mees 1959 showed tne same feature. Later, however, specimens were received which agreed m every detail wttn C. nasuti..s 'cut for the fact that they had shorter snouts and a snaFer anal papilla. In the long- snouted specimens the anal papilla is large and long more than half an eye dmmeter m the short -snouted specimens it is very small . Though the result cf an attempt at sexing was incon- clusive. I feel confident that sncut-length in this species is a sex-aal character, the long- snouted fishes 'oeing males, the short-snouted this, when I had the .e the mdr.udual cf y Palmer 19:2 . -his proved I nac come as tar opportunity to exai C. gr-c-557 record^ TO be identical iard^ as females cf nGStitu-i. and mdeec. in McCuFoch's 1926. p. 195 key. the only charac- ter aiven to cisiing'rish C. ^cr^rns from C. IS tite length cf the sneut In order to make muite certain that C c-cs?: and C are female and male cf the same sp-ectes - cc :e c: tne term* t'uFv confirmed :c‘ twee expecta- Materml examinee specimens varying in standard lenstn :nr=i Tf to 14f rant Mcnte 3eUo 3r-t M-ns Nat. Hist. IJtl-S .6 es n-GU-n GsLf ? case 3 sp ? 5-5- C 2378. C 2379 C 2737 C -758 C 2735 Zsnncntn Gnld 57 or Shark Bay (P 5493), Shark Bay (P 4376 (2 sp.). P 5394, P 5395, P 5396. P 5397. P 5400. P 5447, P 5107). Also examined a specimen from Bulwer, Moreton Bay, Queensland (Queensl. Mus. I 1579: cotype of species), and two speci- mens from Tangalooma Point. Moreton Bay. Queensland (C 2092, C 2093). Callionymus limiceps Ogilby Callionymus limiceps Ogilby. Ann. Qd Mus. 9. 1908, p. 35 — Moreton Bay, Queensland. Callionymus limiceps vur. sublaevis McCulloch. Biol. Res. Endeavour V. 1926. p. 204 — 7-10 miles north-west of Hummocky Island, near Cape Capricorn. Queensland. 14-16 fathoms, and 13 miles south-east from Cape Cap- ricorn. Queensland. 13 fathoms. Callionymus limiceps: McCulloch. Rec. Aust. Mus. 14. 1923. p. 9 (between Hervey Bay and Port Denison, Queensland, at various depths between 13 and 26 fathoms): McCulloch & Whitley, Mem. Qd Mus. 8, 1925, p. 173 (Moreton Bay: between Hervey Bay and Port Denison); McCulloch, Biol. Res. Endeavour V. 1926, p. 203 (various localities off the coast of southern Queens- land): Whitley. Rec. Aust. Mus. 17, 1929 (27 June), p. 115. figs. 3 and 4 (over Sow and Pigs Reef, Port Jack- son); McCulloch. Mem. Aust. Mus. 5. 1929 (28 Nov.), p. 340 (Queensland); Schultz, Bull. U.S. Nat. Mus. 202, 2, 1960, p. 4C3 (no locality). Callionymus limiceps var. sublaevis; McCulloch, Mem. Aust. Mus. 5. 1929, p. 340 (Queensland). Velesionymus limiceps: Whitley, in McCulloch. Fish. N.S.W., 3rd ed., 1934, suppl. no. 418a (Port Jackson). Diagnostic characters. D IV-95, A 9k, P i.l4.iii or i.lS.i or i.l5.ii or i.l5.iii or i.l6.i or i.l6.ii or i.l7.i, V 1.5, C ii.7.ii or ii.7.iii; only last ray of dorsal fin divided: preopercular spine with a tip bent inwards, with one antrorse hook on the inside, and with an antrorse spine near its base on the outside (Fig. If); cnly first pectoral ray undivided: origin of first dorsal well behind gill openings. Males have all four rays of first dorsal elongated: there is only a small amount of black on the first dorsal fin; females have a short first dorsal, which is largely black. Dlsiribution. Queensland, New South Wales and Western Australia. In Queensland the specie^ has been recorded from a number of localities between Hervey Bay and Port Denison; for a full list of localities I refer to McCulloch (1926). For New South Wales there is only Whitley’s (1929) record of a single specimen caught at Port: Jackson. In Western Australia known from Shark Bay, Exmouth Gulf and the Dampier Archipelago. Discussion. From the Queensland Muspum T received on loan a pair of C. limiceps, collected ana identified oy Ogiloy himself; the material frcm Western Australia differs from these only in having the bony upper surface of the head almost smooth with two slightly elevated radiat- ing centres. This material therefore fully agrees with McCulloch’s variety sublaevis as figured by Whitley (1929). As there are no other dif- ferences. I 6^ not think that limiceps and sub- laevis are different species, and they cannot be recognised as subspecies either, as they co-occur along the Queensland coast. Material examined, 17 specimens varying in standard length from 53 to 131 mm. Dampier Archipelago (A 1461), Exmouth Gulf (P 5353, P 5354. P 5425. C 2380), Shark Bay (P 4370 (2 sp.), P 5401, P 5402. P 5403. P 5404, P 5405, P 5406. P 5407, P 5408, P 5421 (2 spJ). Also examined two specimens from Moreton Bay, Queensland (Queensl. Mus. I 487), and one from Tangalooma Point. Moreton Bay. Queensland (A 868). Callionymus papilio Gunther Callionymus papilio Gunther. Ann. Mag. Nat. Hist. (3) 14. 1864. p. 197— Melbourne. Callio7iymus ocellijer Castelnau, Proc. Zool. Accl. Soc Vlct. 2. 1873, p. 49 — Cape Schanck. Callionymus lateralis Macleay. Proc. Linn. Soc. N.S.W. 5. 1881. p. 628 — Port Jackson. Callionymus macleayi Ogilby. Cat. Fish. N.S.W.. 1886. p. 37 — nomen novum for Callionymus lateralis Macleay. nec Callionymus lateralis Richardson. Callionymus Papilio: Macleay. Proc. Linn. Soc. N.S.W. 5. 1881. p. 627 (Melbourne); Macleay, Descr. Cat. Aust. Fish I, 1881, p. 262 (Melbourne). Calli07iymus lateralis: Macleay. Descr. Cat. Aust. Fish. I. 1881. D. 263 (Port Jackson); Johnston. Pap. Roy. Soc. Tasm. (1890). 1891. p. 33 (Tasmania). Callionymus latealis: Tenison-Woods. Fish and Fish- eries N.S.W.. 1883, p. 19 (New South Wales). Callionymus papilio: Lucas. Proc. Roy. Zool. Soc. Viet. N.S. 2, 1890, p. 29 (Hobson’s Bay); Waite, Mem. N.S.W. Nat. Cl. 2. 1904. p. 51 (no locality New South Wales): McCulloch. Aust. Zool. 2. 1922. p. 103 (no locality - New South Wales); Lord, Pap. Roy. Soc. Tasm., (1922). 1923. p. 69 (no locality _ Tasmania); McCulloch. Rec. Aust. Mus. 14. 1923, p. 13 (New South Wales, from Port Jackson southward to Victoria and Tasmania): Lord & H. H. Scott, Synops. Vertebr. Anim. Tasm., 1924, p. 12. 78 (Tasmania); McCulloch. Fish. N.S.W., 2nd ed.. 1927. p. 77 (no locality New South Wales); Lord in Glblin. Lewis & Lord (editors), Handb. Tasm., 1927. p. 87 (Tas- mania, especially in the north): McCulloch. Mem. Aust. Mus. 5, 1929, p. 338 (Victoria. New South Wales, Tas- mania); McCulloch. Fish. N.S.W., 3rd ed.. 1934, p. 77 (no localitv New South Wales); E. O. G. Scott. Pap. Roy. Soc. Tasm. 87. 1953, p. 157 (Tasmania); T. D. Scott. Fish. S. Aust.. 1962, p. 168 (South Australia. Victoria New South Wales and Tasmania). Foetorepus papilio; Whitley, Aust. Zool. 6, 1931. p. 323 (Victoria): Whitley, Aust. Zool. 11. 1945, p. 42 (Cottesloe); Whitley. W. Aust. Fish. Dept.. Fish. Bull. 2, 1948. p. 27 (Western Australia). Diagnostic characters. D IV-lk, A Qk, P 18 cr 19. V 1.5, C ii.7.iii: first dorsal ray simple, others divided at the tips; preopercular spine as in C. calauropomus pointing outwards and upwards, with, besides the decurved tip, one single hook, curved forwards, on its inside; no antrorse hook at its bf)se (Fig. Ig). A email species. r*. 2 ^apilio can easily be distinguished from other Australian species of the genus by the short dorsal and anal fins, though I note that McCulloch (1923) mentions a specimen that has eight rays in its dorsal fin instead of the usual seven. Distribution. New South Wales. Tasmania. Ifictoria. on? record for South Australia (Cape Elizabeth, Yorke Peninsula), and two specimens recorded from Western Australia (Cottesloe). The Cottesloe soecimens should be in the col- lection of the Western Australian Museum, but I have been unable to find them. Material examined. No material from West- ern Australia has been available. The only specimen examined was one of 62 mm standard length from Cape Elizabeth, Yorke Peninsula. S.A. (S. Aust. Mus. F3076).=" Callionymus phasis Gunther Callionymus phasis Gunther, Voy. Challenger. Zool. I. 1880, p. 23, pi. XV Fig. C— Twofold Bay; 120 fathoms (?). Callionymus apricus McCulloch, Biol. Res. Endeavour V. 1926, p. 209. pi. liv. Fig. 2— Great Australian Bight, south from Eucla, 350-450 fathoms. Callionymus phasis: Macleay, Proc. Linn. Soc. NS.W. 9, 1884, p. 35 (Twofold Bay); Waite, Mem. N.S.W. Nat. Cl. 2, 1904, p. 51 (no locality _ New South Wales): Mc- Culloch, Aust. Zool. 2, 1922. p. 103 (no locality - New South Wales); McCulloch. Rec. Aust. Mus. 14. 1923, p. 9 *Amongst our unidentified collections, I have since found a single individual of this suecies from Salmon Bay, Rottnest Island, collected 25.11. 1955. reg. no. P5691. standard length 52 mm. 98 (Gippsland Coast. Victoria, 80 fathoms); McCulloch. Biol Res. Endeavour V. 1926. p. 212 (Gippsland coast. Victoria. 80 fathoms: South of Cape Everard. Victoria. 200 fathoms): McCulloch, Fish. N.S.W., 2nd ed., 1927. p. 77 (no locality -- New South Wales); McCulloch, Aust. Mus Mem. 5, 1929. p. 338 (New South Wales. Victoria): McCulloch, Fish. N.S.W.. 3rd ed.. 1934, p. 77 (no locality New South Wales); Norman. Fishes. B.A.N.Z.A.R.E., Rep. (B) I, 1937, p. 56 (off Tasmania. 42°40'S., 148°27'30"E.. 122m.). Callionymus apricus; Waite. Rec. S. Aust. Mus. 3. 1927, p. 231 (the Australian Bight in 350 to 450 fathoms); McCulloch. Aust. Mus. Mem. 5, 1929. p. 339 (Border of South and Western Australia). Yerutius apricus: Whitley. Rec. Aust. Mus. 18. 1931. p. 115 (no locality); Whitley, W. Aust. Fish. Dept.. Fish. Bull. 2, 1948. p. 27 (Western Australia (south coast)); T. D. Scott. Fish. S. Aust., 1962, p. 170 (Western Aus- tralia and South Australia in the Bight). Diagnostic characters. D IV- 82 ' or A 7 - 2 -: all (iorsal rays divided but apparently simple in small specimens: preopercular spine curved up- wards at its distal extremity to form a hook of the same size as two others on its upper margin: no antrorse spine at the base below (Fig. lh>. Eyes large, rising high above the profile of the head; interobital very narrow. Anterior margin of D 1 only a little behind gill openings. Distribution. Temperate seas of Australia, where known from Twofold Bay, New South Wales: the coast of Victoria, the east coast of Tasmania, and the Great Australian Bight south from Eucla. Apparently a deep water species that has been taken at depths of from 80 to 350-450 fathoms. Recently recorded from Japan (Ochiai, Araga and Nakajima 1955 ». Discussion. I have not examined material 3 f this species, which as far as I am aware, in Western Australia is known from the type specimen of C, av^'icus only. The particulars given are after McCulloch (1926), who noted the close resemblance of C. apricus to C. phasis Gunther (1880) and, though not actually saying so, more or less suggested that the two might be indentical. Nearly forty years have passed since, and no additional material that might confirm the validity of C. apricus has turned up. As the type locality of C. apricus is in a region where C. phasis, already known from the coasts of New South Wales, Victoria and Tasmania, would be expected to occur, and as both were found in deep water, it seems justified to synonymize the former. McCulloch (1926) men- tioned only colour differences to distinguish between the two, but there may well be sexual dimorphism in this character. All specimens recorded until recently, in- cluding the types of C. phasis and C. apricus, were reported to have divided dorsal rays: their size range was from 48 to 123 mm standard length. In the specimen of 44 mm standard length recently recorded from Japan, the dorsal rays are described as simple. Callionymus rameus McCulloch Callionymus, Calliurichthys, rameus McCulloch. Biol. Res. Endeavour V, 1926, p. 201. pi. liii — Cape Capricorn. Queensland. Callionymus rameus: McCulloch, Aust. Mus. Mem. 5, 1929, p. 339 (Queensland); Schultz, Bull. U.S. Nat. Mus. 202. 2. 1960. p. 4C3 (no locality). Orhonymus rameus; Whitley. Aust. Zool. 11. 1947. p. 150 (no locality). Callionymus (Calliurichthys) rameus: Mees, W. Aust. Fish. Dept., Fish. Bull. 9. 1959. p. 9 (Shark Bay); Palmer. Ann. Mag. Nat. Hist. U3) 4. 1962. p. 548 (Monte Bello Islands). Diagnostic characters. D IV- 8 i. A 72 , P i.lS.ii or i.lS.iii or i.l 6 .ii or i.l7.i or 17.iii or i.l 8 .i, V 1.5. C ii.7.ii to iii.7.iii: all dorsal rays divided or the first ray simple: preopercular spine with about five small teeth on the inside and with an antrorse hook on its base (Fig. li); origin of D 1 on a line with the gill openings, only first ray of P simple, once even the first ray divided: snout short, slightly shorter than eye. Ap- parently no sexual dimorphism. Distribution. Queensland: Cape Capricorn, also 25 miles south-east from Double Island Point. 33 fathoms, and 4-20 miles north-east of Gloucester Head. 19-35 fathoms (McCulloch 1926). Western Australia: Shark Bay (without exact location) and 40 miles South of Carnarvon (Western Australian Museum); off the Monte Bello Islands (Palmer 1962). Material examined. 12 specimens varying in standard length frcm 80 to 153 mm. Various localities in Shark Bay (P 4361. P 5106. P 5267. P 5269, P 5366, P 5413 (2 spJ. P 5414. P 5415. P 5416. P 5431. C 565). References Gunther, A. (1861). — ‘'Catalogue of the Acanthopteryglan Fishes in the Collection of the British Museum" III (London.) (1880). — Report on the shore fishes procured during the vovage of H.M.S. Challenger in the years 1873-1876. Rep. Sci. Res. Challenger. Zool. I. McCulloch. A. R. ( 1923).— Notes on fishes from Australia and Lord Howe Island. Rec. Aust. Mus. 14: (1926). — Report on some fishes obtained by the F.I.S. "Endeavour" on the coasts of Queensland. New South Wales. Victoria. Tas- mania. South and south-western Australia. Biol. Res. Endeavour V : 155-216. ( 1929-1930).— A check -list of the fishes re- corded from Australia. Aust. Mus. Mem. 5. Marshall. T. C. {1951) .—Ichthyological Notes (Brisbane). No. 1. Mees, G. F. (1959).— Additions to the fish fauna of West- ern Australia — 1. W. At(Sf. Fish. Dept.. Fish. Bull. 9 : 1-11. Ochiai. A.. Araga. Ch.. & Nakajima. M. (1955).— A re- vision of the dragonets referable to the genus Callionymus found in the waters of Japan. Publ. Seto Mar. Biol. Lab. 5 : 95-132. Cgilby. J. D. (1910). — On new or insufficiently described fishes. Proc. Roy. Soc. Qd. 23 : 1-55.* Palmer G (1962).— New records of fishes from the Monte Bello Islands. Western Australia. Aim. Mag. Nat. Hist. (13) 4. (1961) : 545-551. Schultz. L. P. et al. ( I960).— Fishes of the Marshall and vol. 2. Stead D G (1906). — "Fishes of Australia :a Popular and Systematic Guide to the Study of the Wealth within our W^aters" (William Brooke Sz Co. : Sydney.) Whitlev. G. P. (1929).- Studies in ichthyology. No. 3. Rec. Aust. Mus. 17 : 101-143. (1931). — New names for Australian fishes. Aust. Zool. 6 : 310-334. (1944).— New sharks and fishes from Western Australia. Aust. Zool. 10; 252-273. (1945). — New sharks and fishes from Western Australia. Part 2. Aust. Zool. 11 ; 1-42. (1948a).— New sharks and fishes from Western Australia. Part 4. Aust. Zool. U : 259-276. (1948b). — A list of the fishes of Western Aus- tralia. W. Aust. Fish Dept.. Fish. Bull. 2. =^The volume in which Ogilby’s article appeared bears the date 1911. but the Royal Society of Queensland pre-issued author’s reprints and according to Mc- Culloch (1929-1930). the paper was published m November. 1910. 99 INSTRUCTIONS TO AUTHORS Papers may be submitted to the Society in accordance with Rules and Regulations 38 to 41 inclusive (see below). They should be addressed to The Honorary Secretary, Royal Society of Western Australia, Western Australian Museum, Perth. Authors are solely responsible for the factual accuracy and for any opinion expressed in their papers. They are particularly requested to verify references. Alterations to MSS. submitted to the printer will be allowed only under excep- tional circumstances, and no changes will be permitted after galley-proof stage. In the preparation of MSS. authors are required to follow the C.S.I.R.O., Guide to Authors (C.S.I.R.O., Melbourne, 1953), except that papers longer than 10,000 words (30 foolscap pages of pica type, with 6 inch lines 40 to the page) will not normally be accepted. Authors may be required to meet half the cost of preparation of the blocks of diagrams and illustrations. Authors shall receive a total of 30 reprints free of charge. Further reprints may be ordered at cost, provided that such orders are submitted with the MS. RULES AND REGULATIONS 38. Every paper intended to be read before the Society or to be published in the Society’s Journal must be sent to the Secretaries at least seven days before the date of the next ensuing Council meeting, to be laid before the Council. It will be the duty of the Council to decide whether such contribution shall be accepted, and if so, whether it shall be read in full, in abstract, or taken as read. All papers accepted for publication must be read or otherwise communicated at an ordinary meeting prior to publication. 39. A Publications Committee, appointed by the Council, shall recommend to the Council whether a paper presented to the Society shall be published in the Society’s Journal. The Publications Committee may obtain an opinion from any person it may select on the suitability of any paper for publication. 40. Publication in the Society’s Journal shall only be available to (a) Ordinary Members, (b) Honorary Members, (c) Non-members resident outside Western Australia, who must communicate the paper through an Ordinary or Honorary Member. No paper shall be accepted from a Non-member resident in Western Australia. 41. The original copy of every paper accepted for publication by the Society, with its illustrations, shall become the property of the Society, unless stipulation is made to the contrary, and authors shall not be at liberty to publish their communicated papers elsewhere prior to their appearance in the publications of the Society unless permission for so doing is given by the Society, or unless the Society fails to publish the paper in the Journal of the year in which it is read or otherwise communicated, or of the succeeding year. Journal of the Royal Society of Western Australia, Inc. Volume 46 1953 Part 3 Contents 9.— The Plantagenet Beds at Hummocks Beach, Bremer Bay, Western Aus- tralia. By J. G. Kay, J. E. Glover, and Rex T. Prider. Memorials: Enid Isabel Ahum, Ludwig Glauert. 10. ~Vertebrate Remains from the Nullarbor Caves, Western Australia. By Ernest L. Lundelius, Jr. 11. — Some New Western Australian Sawflies of the Euryinae and Phylacteo- phaginae (Hymenoptera, Pergidae). By Robert B. Benson. 12. — The Gekkonid Genus Nephrurus in Western Australia, including a New Species and Three New Subspecies. By G. M. Storr. 13. — The Angler-fish Ceratias holboelli from Western Australian Waters. By B. K. Bowen. 14. — The Callionymidae of Western Australia (Pisces). By G. F. Mees. Editor: J. E. Glover Assistant Editors: R. W. George, R. D. Royce Annual Subscription: Forty Shillings The Royal Society of Western Australia, Inc., Western Australian Museum Perth 72638/7/63—570 ALEX. B. DAVIES, GovernmenI Printer, Western Australia JOURNAL OF THE ROYAL SOCIETY OF WESTERN AUSTRALIA INCORPORATED VOLUME 46 (1963) PART 4 PUBLISHED 18TH DECEMBER, 1963 REGISTERED AT THE G.P.O., PERTH FOR TRANSMISSION BY POST AS A PERIODICAL THE ROYAL SOCIETY OP WESTERN AUSTRALIA INCORPORATED COUNCIL 1963-1964 President Past President Vice-Presidents Joint Hon. Secretaries Hon. Treasurer Hon. Librarian Hon. Editor C. F. H. Jenkins, M.A. W. D. L. Ride, M.A., D.Phil. W. R. Wallace, Dip.For. J. H. Lord, B.Sc. Margaret E. Redman, B.Sc. J. G. Kay, B.Sc. R. D. Royce, B.Sc. (Agric.). Ariadna Neumann, B.A. J. E. Glover, B.Sc., Ph.D. A. S. George, B.A. A. B. Hatch, B.Sc., Dip.For. L. E. Koch, M.Sc. R. J. Little. D. Merrilees, B.Sc. W. C. Packer, Ph.D. P. E. Playford, B.Sc., Ph.D. L. W. Samuel, B.Sc., Ph.D., F.R.A.C.I., F.R.I.C. Journal of the Royal Society of Western Australia Vol. 4(> Parti 4 15.— Some Pelecypods from the Cretaceous Gingin Chalk, Western Australia, together with Descriptions of the Principal Chalk Exposures By F. R. Peldtmann* Manuscript received — 18th September, 1962 In this paper a number of pelecypods (lamelli- branchs), mostly new. from the Gingin Chalk are described in detail and named, and descrip- tions are also given of the principal chalk exposures. The pelecypods include species of Perna. Anomia. Spondylus, Plicatula. Ostrea. Gryphaea, Pycnodonta and Exogrya. In the descriptions of the species occasional reference is made to specimens from the Toolonga Cal- cilutite of the Lower Murchison area. Introduction Pioneer geological and palaeontological work in the Gingin area was carried out by Mr. Ludwig Glauert, firstly as a member of the Geo- logical Survey staff, and subsequently as Curator of the Western Australian Museum. Mr. Glauert was responsible for several papers and made the first geological map of the area. With the exception of the Pectens (Feldtmann 1951) no new pelecypods from the Gingin Chalk have been described since the publication of Etheridge’s paper on the Cretaceous fossils of the Gingin Chalk (Etheridge 1913). Glauert’s list of Upper Cretaceous fossils (Glauert 1926) included, in addition to those described by Etheridge, species of Pecten, Chlamys and Avius- slum, but no descriptions were given. Etheridge described and named Pycyiodonta ginginensis, Chlamys ellipticus and Mytilus piri- formis, and in addition to mentioning various species of Inoceramus, which he compared to species from Queensland and elsewhere, he also mentioned briefly and figured, but did not name, three species of oysters under the titles of ‘‘Ostrea sp. ar “Ostrea sp. hr and ‘‘Ostrea oi Pycnodonta (Junr.).” These three last are de- scribed and named in this paper as well as species of Perna, Anomia. Spondylus, Plicatula, Ostrea, Gryphaea, Pycnodonta and Exogyra. In connection with distribution of the various species, brief descriptions have been given of the more important exposures of the Gingin Chalk. In 1944. a very large number of fossils was collected by the late Professor E. de C. Clarke and Dr. Curt Teichert from the Toolonga Cal- cilutite at various localities on the northern side of the lower Murchison River ( Clarke and Teichert 1948), and more recently, a number of * Honorary Research Associate. Department of Geology University of Western AustrnliH, Nedlnncls. Western' Australia. specimens was collected by Dr. B. F. Glenister and Mr. B. E. Balme from Thirindine Point in the same area. In the description of the Gingin species occasional reference is made to specimens from these collections. Location and General Geology The small farming town of Gingin is situated 50 miles by rail north of Perth, on the Midland Company’s line to Geraldton. It lies on both sides of the permanently flowing Gingin Brook, which, north-east of the town, flows in a south- south-westerly direction, but, after taking an abrupt U-shaped bend to the east immediately east-north-east of the town, changes to a west- south-westerly course through and west of the town. The town lies near the western edge of remnants of a former plateau of Mesozoic rocks situated between the main north -striking Darl- ing Fault, which separates it from the Precam- brian meta-sediments to the east, rather more than 8 miles east of Gingin railway station, and a second fault striking about north-north-west, which apparently diverges from the Darling Fault a little south of Bullsbrook some 24 miles to the south-south-east. Shot-holes put down for the West Australian Petroleum Proprietary Co. Ltd., and examined by Mr. S. Warne (unpub- lished data) indicate that this second fault is approximately U miles west of Poison Hill, 4 miles north-north-west of Gingin railway station. In the immediate neighbourhood of Gingin, the plateau has been largely eroded by Gingin Brook and its tributaries, but from about a mile north- west of Poison Hill, northward, the western escarpment of the plateau is remarkably regular with a general trend of N.24°W., and this is very probably the approximate strike of the fault. This would put it about a mile west of the escarpment north of Poison Hill and approxi- mately the same distance west-south-west of One Tree Hill and Gingin railway station. From 3 miles south of Gingin it is probably very close to the Midland railway line. The higher points on the plateau are now mostly between 700 feet and 780 feet above sea-level. West of the fault, the surface is much lower, mostly between 180 feet and 270 feet above sea-level, and the rocks are largely obscured by the sands of the coastal plain. 73061 - (U The Mesozoic succession, and the approximate thicknesses of the formations are as follows; — Upper Cretaceous. — Poison Hill Greensand— possibly 170 feet to 250 feet. Gingin Chalk — nil to 70 feet. Molecap Greensand — 20 feet or less to more than 90 feet? Unconformity. Lower Cretaceous. — Strathalbyn Beds — between 170 feet and 180 feet exposed in the Gingin area. Strathalbyn Beds The Strathalbyn Beds were formerly regarded as Upper Jurassic from the identification by Walkom (1944) of plant remains found in ferruginous sandstone near the top of the formation in Cheriton Gully, 22 chains east of McIntyre Gully. In addition to the species identified by Walkom, specimens of an Otoza- vutes, associated with Taeniopteris, were found more recently at approximately the same hori- zon about 15 chains west of McIntyre Gully. The beds are now regarded by Mr. B. E. Balme (personal communication) as of Lower Cre- taceous (Aptian) age, from material obtained from some of the shot-holes and from shallow holes below the floor of Molecap quarry. The longest exposure of the beds is in McIntyre Gully, in the Strathalbyn property, where they consist mainly of fine and coarse ferruginous sandstones, party bleached and finely micaceous in places. In McIntyre Gully the uppermost beds are pale grey to pale brown- ish yellow, and appear to be more shaly in character. At the northernmost exposure in the gully, the Strathalbyn Beds are separated from the overlying Molecap Greensand by a pale yellowish lateritic layer, four inches thick, containing fragments of fossil wood. This lateritic layer, found elsewhere at the top of the beds in places, doubtless represents a former land surface, much eroded in places before the Upper Cretaceous marine sediments were laid down. Less weathered examples of the beds occur in two small watercourses near the foot of the southern slope of Moorgup Hill, about li miles south-east of the railway station. Here the beds, immediately below their contact v/ith the Molecap Greensand, consist of bands, from 4 inches to 8 inches thick, of fine-grained, pale grey arenaceous shales containing a few coarse quartz grains and with inch-wide flakes of muscovite on bedding planes, alternating with bands of coarser red sandy grit. Small angular fragments of black carbonaceous material were noticed in one 6-inch band of more sandy shale. The beds here have a very slight easterly dip. The beds also occupy a low ridge partly enclosed by the U-shaped bend of Gingin Brook north-east of the town. Here, pale grey, fine- grained, apparently argillaceous sandstone with much fine white mica is exposed in a few shal- low potholes on the backbone of the ridge. In Molecap quarry, auger-holes exposed yellowish brown clay, very similar in appearance to the top layers of the Strathalbyn Beds in McIntyre Gully, below the greensand at 7 feet 8 inches below the main floor of the quarry. Water was encountered at the junction of the two formations. Molecap Greensand The Molecap Greensand (Peldtmann 1934' a glauconitic sandstone, is a fairly fine-grained, dark greyish green rock, usually of homogeneous appearance and without visible bedding. At the type locality, a thickness of about 28 feet is exposed in the quarry, and auger-holes show it to have a total thickness here of 36 feet. The last 2 feet 6 inches encountered in the auger- holes above the Strathalbyn Beds was oxidised to a ferruginous brown sand. A thin band with phosphatic nodules was said to have been found ot the base. In an auger-hole put down for the company quarrying the greensand, the glau- conite being used as a water-softener. At the top of the greensand, a band, from about 3 inches to 30 inches thick, of dark reddish brown ferruginous material, with phosphatic nodules up to about 8 inches in diameter as well as rare bone fragments and other fossils, separates it from the overlying chalk. At One Tree Hill, nearly I j miles north-north- west of Molecap, the thickness of greensand appears to be about the same. The phosphatic band is absent here. At McIntyre Gully, the observed thickness between two points was only 21 feet 7 inches and as the junction between the Strathalbyn Beds and the greensand showed a slight southerly dip the thickness is probably even less farther north, and in a shot-hole three-quarters of a mile north-east it was only 20 feet. It appears, however, to thicken fairly rapidly to the west. At one point in McIntyre Gully, the lowest foot of the greensand has been altered to a yellowish brown colour and contains small irregular phosphatic nodules. The uppermost five feet of the green- sand in the gully becomes gradually paler in colour as it approaches the base of the chalk and appears to pass into it without any defined break. Although the Molecap Greensand is usually of even textui’e and fine grains, occasional coarser facies occur. In small watercourse rather more than half a mile north-west of One Tree Hill, greensand with fairly numerous grains of quartz and orthoclase up to about 3 mm in length was seen. Not many fossils are found in the Molecap Greensand. The writer found two small saurian limb bones, one about 7 inches long, as well as smaller fragments, in McIntyre Gully at about five feet below the base of the chalk and specimens of Sptrulaea gregaria, small belem- nites, and two species of Chlamys w-'-^re found by Dr. R. W. Fairbridge at about 10 I>et below the base of the chalk (Feldtmann 1901, p. 24). A few rare bones occur near the top of the formation at Molecap and bone fragments were also seen, near the top. in a watercourse immediately north of the Mooliabeenie road, a short distance east of Musk’s Chalk. The writer also found specimens of fossil wood in the greensand in the more westerly small gully near the base of the southern slope of Moorgup 102 GEOLOGICAL SKETCH MAP OF GINGIN POISON HiLL '/ / / M Sti'othafbyi north G ' N G l N L) P ' I LEGEND 'h'ilL Poison Hil' Go^jnsond C i> I k Mol^cop Crjensood G t NG I N SliaO'olt"/n Beds f^poliaitt iti'f ocoiogicol boi'ndories Fossi L lV riH>5 f hoik Ool< ro| M O L E C A p r -A ^MOORGi^^P F, R. Feldtmonn. 1963 Fig. 1 103 At Poison Hill, where the chalk is absent, the Molecap Greensand appears to pass into the Poison Hill Greensand without a break. The heavy mineral content of the Molecap Greensand, mainly magnetite, was estimated at about one per cent, by Carroll (1941). Gincjin Chalk The Gingin Chalk forms a discontinuous layer of varying thickness between the Molecap and the Poison Hill Greensands. It consists mainly of carbonate of lime with varying proportions of grains of glauconite, quartz and some ortho- clase. The heavy minei’al content was describea in detail by Carroll (1939) who estimated it at 0.07 per cent. Detailed descriptions of the prin- cipal chalk exposures are given on subsequent pages. Poison Hill Greensand The best exposures of the Poison Hill Green- sand are at the type locality and immediate vicinity and at a prominent landslip, about 4 -mile south-south-east. Good exposures of the base of the formation, above the chalk, are visible in two small gullies south-east of the landslip and about half-way between Poison Hill and Ginginup, and also at Southern’s Chalk=^ about i-mile north-west of Ginginup. The slightly weathered rock was exposed on the southern slope of Mooi'gup, about a mile south- east of the railw'ay station, by a landslip, about 40 years ago, but has since been partly covered up by further slipping. Like the Molecap Green- sand. the Poison Hill Greensand is a fairly fine- grained dark greenish rock consisting mainly of glauconite and quartz with some orthoclase; a little clay is present in places. According to Carroll (1941. p. 85), the heavy mineral content is very small, less than in the Molecap Green- sand, The unweathered rock is usually of more or less homogeneous appearance, but an expo- sure of much-weathered oxidised rock on the south-eastern face of the Ginginup escarpment shows it too consists of beds, three or four inches thick, of both fine and coarse sandstones, some even approaching a grit. Small pea-shaped nodules of the rare min- eral gearksutite (CaP2AlF0H(P,0H)H:.0) were found by Simpson (1920, p. 27) in a rather clayey bed, about 12 inches thick, with abundant glauconite and small phosphatic nodules. The bed was exposed in a small shaft on the north- eastern slope of Moorgup, about 2 chains south of the Mooliabeenie road and nearly opposite Musk’s Chalk. Simpson also mentioned other somewhat clayey bands carrying small phos- phatic nodules farther up the slope. So far as I know the only organic remains found in the Poison Hill Greensand are copro- lites. which, according to Simpson (1937, p. 38, Fig. 3) are abundant in certain bands, associated with bands carrying the phosphatic mineral dufrenite (Pe 2 (OH);POi), at the landslip south- south-east of Poison Hill. * Note that the terms Southern's ChalK, Musk’s Chalk, Hosking’s Chalk and North Chalk are not formation names. They refer to exposxires of the Gingin Chalk, and may be located on the map (Fig. 1 ) from the information la the text. On much of the high ground between Gin- ginup and Poison Hill, the greensand is capped by dense ferruginous laterite up to about two feet thick in places. No Upper Cretaceous rocks have been found south of Lennard’s Brook, about two miles south- south-east of the railway station. Eclipse Hill, south of the brook and about 3i miles south- east of the station, presents a very different appearance from Moorgup, north of the brook, on the southern face of which the Upper Cret- aceous succession shows up very noticeably, whereas on Eclipse Hill, which is slightly higher, there are no rock outcrops and the hill is entirely covered by white or grey sand very different from the dark reddish brown sand derived from the greensands. It is probable that the straight easterly portion of Lennard’s Brook follows a line of transverse faulting. Pleistocene Drift An unexpected discovery early in May. 1949, was that of a fairly extensive deposit of Pleis- tocene drift in the upper portion of McIntyre Gully. The best exposure is on the west side of the gully. 120 feet south of where joined by the second small eastern tributary, and 590 feet south of the head of the gully. Here, portion of a lower jawbone with four large teeth, identified by Mr. Glauert of the Western Australian Museum as belonging to the extinct giant mar- supial Nototheriuni initchelli, was found lying on the floor of the drift by a student, D. A. Wood- man. Other bone fragments were also found lying on the drift-floor, which is 2 feet 6 inches above the present floor of the gully. The base of the drift, which here is lying on greensand, is marked by an inch-thick layer of red ferruginous material. Above it, the drift consists of com- pacted greensand containing nodules and some larger boulders of chalk to 1 foot 9 inches above the base, followed by a 3-inch layer of grit, then soil to about 3 feet above the base, followed by soil with layers of grit to about 6 feet 6 inches above the base. The top of the drift is about 16 feet above the present stream-bed. On the w^est bank of the gully, the drift appears to tail out a few feet south of the exposure, but it ex- tends northward for about 160 feet. Further search, a fortnight after the first discovery, re- vealed a second exposure of the drift about 135 feet north of the first and a few feet north of the junction with the small eastern tributary. Here, the drift is resting on the chalk and consists of material derived therefrom. Bones found resting on the floor of the drift at this exposure in- cluded a limb bone, scapula, vertebra, and two terminal toe-bones. The top of the drift here is also about 16 feet above the present stream bed. The post-Pleistocene portion of the gully is considerably narrower than that of Pleistocene times, indicating rejuvenation. The drift appears to be present in places on the eastern side of the gully, but its boundaries are ill-defined and there are no exposures. The Chalk Exposures In the Gingin area the Gingin Chalk forms a more or less tabular layer of varying thickness deposited when and where conditions, such as depth of water, were favourable between the 104 lower or Moiecap Greensand and the upper or Poison Hill Greensand. Where fully exposed the chalk grades into the greensands below and above, the boundaries between them being some- what indefinite and difficult to determine with accuracy. At most of the exposures however, the section is incomplete, either the lowest or upper- most layers of the chalk being absent and the junctions between the chalk and the greensand abrupt. The northern limit of the Gingin Chalk is about 3h miles north-north-west of the Gingin railway station and half a mile south of Poison Hill, the most prominent point on the escarp- ment. At Poison Hill, where it is absent, the Moiecap Greensand appears to pass without a break into the Poison Hill Greensand. The combined formations pi'obably range from Coniacian to Campanian or even Maestrichtian in age (McWhae et al. 1958, pp. 116, 117, Table 1), the Gingin Chalk itself being Santonian. The Gingin Chalk reaches its maximum thickness of about 70 feet in the vicinity of McIntyre Gully, a little more than two miles north of the rail- way station. To the south, it appears to be thin- ning again at the southernmost exposure on the southern slope of Moorgup Hill, not quite I 2 miles south-east of the station, and near the southern limit of the Upper Cretaceous rocks. The chalk consists mainly of varying pro- portions of carbonate of lime, quartz grains and glauconite, and according to Carroll ( 1939, p. 228) grains of orthoclase are also present. Clay is present at some horizons, particularly at McIntyre Gully. In a typical specimen of purer chalk from Moiecap quarry. Carroll esti- mated the carbonate of lime at 87.3% and the glauconite at about 9%; the heavy mineral con- tent was estimated at about 0.07%: this last (Carroll 1939, pp. 228-230* consisted mainly of minerals of metamorphic origin, without doubt derived from the Chittering Valley belt of meta- sediments and gneisses east of the Darling Fault. Prom south to north, the principal exposures of the Gingin Chalk are at Moiecap Hill, Musk’s Chalk, One Tree Hill, McIntyre Gully, Southern’s Chalk, the Springs Gullies. Hosking’s Chalk, and North Chalk (refer Fig. 1). Of these, the most important is McIntyre Gully, just over two miles north of the railway station, as only there, is the formation complete, either the lower of upper layers or both, being missing from the other exposures. A detailed survey of the gully and its tributaries by the writer in 1939. showed the total thickness of the formation to be ap- proximately 67 feet 4 inches, but the position of the base of the Gingin Chalk, at its junction with the lower or Moiecap Greensand, is deter- minable only with difficulty under the most favourable conditions, and is more or less arbi- trary, as the last five feet of the greensand becomes progressively paler and less highly glauconitic, with small lenticles of chalk, and appears to pass into the formation without any defined break. As determined by the writer, the base of the Gingin Chalk is situated near the base of the second small eastern tributary, at its junction with the main gully at about 470 feet south of the head of the gully. Here it is approximately 2 feet 6 inches above the stream bed of the main gully. It is also exposed on a cliff face about 200 feet farther south, where the gradual change from a fairly dark grey, mottled, highly glauconitic rock, with a few lenticles of chalk and a few fragments of small pelecypods, to a paler rock with less glauconite and more numerous lenticles of chalk can be seen. Above the base, the Gingin Chalk may be described as follows: — 0 ft-4 ft 2 in.: mottled grey glauconitic chalk, probably somewhat clayey. 4 ft 2 in. -about 10 ft: more even-textured pale grey chalk. 10 ft-19 ft 3 in.: fine-grained pale yellowish to white, somewhat sandy, fairly typical chalk, slightly glauconitic. 19 ft 3 in. -20 ft 3 in.: very clayey fine- grained grey chalk. 20 ft 3 in. -23 ft: fairly typical, nearly white chalk. 23 ft-about 33 ft: mottled grey, probably clayey, marly chalk with more glauco- nite. 33 ft-about 42 ft: less distinctly mottled, grey, marly, somewhat glauconitic chalk. 42 ft-48 ft: grey glauconitic chalk without mottling. 48 ft-62 ft: increasingly glauconitic and darker greenish grey rock. 62 ft-67 ft 4 in.: rock practically indistin- guishable from greensand in appear- ance. The last two sections were exposed in the first small eastern tributary, which enters the main gully at about 307 feet south of the head. The position of the top of the formation was determined on the noticeable change of soil and a marked change of slope, rather than on the appearance of the rock itself. A definite outcrop of the Poison Hill Greensand was exposed a few inches above the assumed top of the Gingin Chalk. Regai’ding the fossil distx’ibution, plates of Ui7itacri7ius are fairly common between the base of the chalk and 13 ft 3 ins. above, particularly between 4 feet and 10 feet. At approximately 13 ft 3 in. Umtacrinus gives place to Marsupites. An overlap of the two stalkless crinoids has not been established in this locality, but at Moiecap there is a definite overlap of about 8 inches. Between 8 ft 9 ins. and 10 feet in the main gully, a layer of the large quadrate Inocerariius com- pared by Etheridge (1913, p. 21) to I. maxivius Lumholtz forms the lip of a small waterfall. This species appears to be confined to the Vmtacrmus zone, as do the small coral CoelosTuilia gmginen- sis, Clila7nys subtilis, and the large pachydiscoid ammonite compared to the English Parapuzosia by Spath (1926, p. 54). Tubulosthmi pyramidale is commonest in this zone, but its vei’tical range extends to 17 feet or 18 feet above the base of the formation. Some of the commoner fossils, such as Cidaris spines, Spirulaea ( Tubulostium? ) gregaria, the brachiopods Bouchardiella cretacea, Ki7ige7ia 7nese77ibri7ius , and l7iopi7iatarcula acan~ tliodes, Plicatula glauerti and the oysters Ostrea etheridgei, PycTiodonta gmgmensis and Exogyra va7'iabilis have a much wider range, extending from the base to at least 44 feet or 45 feet above. 73061 — ( 2 ) 105 The Marsupites zone extends from 13 ft 3 ins. to 20 ft 3 ins. above the base. A few plates of the smooth form are found in the lowest portion of this zone, plates with the typical M. testudi- narius sculpture being more numerous in the middle and upper portions. This zone is richest in fossils, characteristic forms being Peronella globosa, various brachiopods including, in addi- tion to the commoner forms, species of Tere- bratulina and Magnithyris. Pectens, including Syncyclonema siibreticulata and Chlamys gin- gi?iensis occur in the upper half of this zone, where also Pycyiodonta ginginensis attains its greatest size, and plates of the cirripedes Calan- tica gvigmensis and Scalpelhmi glauerti are not unccmmcn. The characteristic species of l7io- cerarnus is a large ovate nearly smooth form, of which a nearly perfect specimen 27 inches in length was found by Dr. Curt Teichert between 17 ft 9 ins. and 18 ft 3 ins. above the base. Equally large specimens seem to have been par- ticularly common at this horizon, although, judg- ing by the thickness of the fragments found, the .species evidently extended to about 35 feet above the base. The band of grey clayey chalk betweei:i 19 ft 3 ins. and 20 ft 3 ins. is particularly rich in small smooth or faintly plicate rhynchonellids, some- what resembling ’'Rhynchonella” limbatcL Sow. and probably representing more than one species. A new Terebratulina, as well as other brachiopods. are fairly common at and just below this horizon. A few small sponges, in- cluding conical forms, a rare squat, mushroom- shaped species, and one resembling Peronella globosa Eth. hi., but more elongate in shape, were found immediately above the clayey band. A narrow zone between 21 feet and 23 feet above the base is characterized by numerous specimens of Ostrea philbeyi, which is almost wholly re- stricted to this zone, although a single small right valve was found close to the top of the Marsupites zone. The species is particularly common in the second small eastern tributary. Specimens of a small finely ribbed rhynchonellid resembling some species of Burmirhynchia occur between 23 feet and 28 feet above the base and a few' unusually large specimens of Inopinatarcula W'ere found at about 28 feet. From about 29 feet to 44 feet the fossils are mostly restricted to the commoner forms, but specimens of a small Terebratulina were found between 32 feet and 39 feet, and a few of the sm.all pecten Pseudainussium candidus were found in the second eastern tributary at about 35 feet above the base. No macrofossils have, so far as I know, been found between about 45 feet and 64 feet above the base. Between 64 feet and the top of the formation, however, fragments of a thin-shelled Inoceramus, prob- ably of a species different from those occurring at lower horizons, were found. At Molecap Hill, half a mile south-east of the railway station, the thickness of chalk and chalk soil is 14 feet 7 ins., about 11 feet of chalk being exposed in the quan-y, but of this only the lower 9 feet contain recognizable fossils. Distribution of the fossils in general, and of the two stalkless crinoids in particular, indicates that, compared with McIntyre Gully, approxi- mately the low^est 11 feet of chalk is missing. probably due to contemporaneous erosion, and as the 28 ft 6 ins. of Molecap Greensand exposed in the quarry— bores show the total thickness of gi-eensand to be 36 feet — is homogenous in appearance, possibly as compared wdth McIntyre Gully, a small thickness is missing from the top of this formation also. The junction be- tween the chalk and the greensand is abrupt and is marked by a feiTUginous band from 3 inches to 30 inches thick, containing numerous phosphatic nodules. Fossils found in this band include many small shark teeth, portion of the lower jawbone, as well as a few small teeth, pTobably of a small Mosasaurus, several centra of Ichthyosaur vertebrae, a vertebra of one of the Salmonidae. and fragments of fossil w'ood containing numerous Teredo casts. The Gingin Chalk at Molecap Hill has proved rich in all the commoner fossils and single speci- mens of species not found elsewhere have been found here. These include a single valve of Crania sp., a small rhynchonellid unlike those of McIntyre Gully, and a small conical coral. The lowest foot of the formation is more highly glauconitic and of slightly coarser texture than that above and contains numerous coarse quartz grains and a few small phosphatic nod- ules. The remainder consists of typical white or pale yellowish rather sandy rock from which most of the glauconite has disappeared. There is a definite overlap of 8 or 9 inches of the two stalkless crinoids, Vintacrinus extending from the base of the formation to at least 3 ft 2 ins. above, where a smooth plate of Marsupites was found at approximately 2 ft 6 in. The two forms of Marsupites also overlap, the smooth form ex- tending to about 5 feet above the base, whereas the sculptured forms range from about 4 feet to the top of the less weathered portion of the exposure, some large plates occurring between 7 feet and 8 feet. Of the rarer fossils, the large pachydiscoid ammonite appears to be confined to the lowest foot of the formation, whereas the other ammonites Eubaculites and Glyptoxoceas have been found in the Marsupites zone. Perna and Spondylus appear to be confined to the Uintacrinus zone and the lower portion of the Marsupites, as do also Mytilus piriformis and a Dentalium. Of the pectens, Chlamys subtilis is confined to the Vintacrinus zone, and C. glngineyisis to the lower portion of the Marsupites zone. Syncyclonema ranges from about 3 feet to 8 feet above the base. Strangely enough, whereas in McIntyre Gully Pseuda- mussiuju candidus was found only at about 15 feet above the top of the Marsupites zone, at Molecap it is common between 3 feet and 6 feet above the base of the formation. Rare frag- ments of a Holaster-li^e echinoid occur in both zones. Although fragments of Inoceramus are very common, some occurring in thin layers, recognizable specimens are rare. Most of the fragments appear to belong to the large smooth form characteristic of the Marsupites zone. A few poorly preserved specimens resembling 1 cripsi have also been found. Very rare needle-like teeth of one of the Salmonidae. resembling those of the genus Apateodus, have been found in the lower portion of the Marsu- piles zone, as w^ell as rare posterior plates of an amphineuran. At Musk’s Cha'k. about 130 chains east of the railway station and a short distance north of the Mooliabeenie road, about 6 feet of chalk was exposed in a cut, apparently just above the top of the Molecap Greensand. About Ih feet was also exposed in a second small cut. the top of which was about 12 feet above the base of the first. When last visited by the writer, both cuts were largely obscured by soil and vegetation. The rock consists of fine- grained greyish glauconitic chalk with very few quartz grains and appears to be mainly, if not wholly, in the Uintacrinus zone, as, so far as I know, Marsupites has not been found here, whereas plates of Uintacrmus are common in the lower cut. A well-preserved specimen of Spondyhis ginginensis was found in highly glauconitic chalk at the base of the low^er cut about 40 years ago by the late Dr. E. S. Simpson, as well as porticn of a gastropod resembling a Pleurotomaria in brown weathered greensand, probably from just below the chalk. Only Cidaris spines and fragments of Inoceramus were seen in the higher cut. At One Tree Hill, just over a mile north- north-west of the railway station, the thickness of chalk and chalk soil is about 18 feet. The junction between the chalk and the Molecap Greensand is exposed behind an old lime-kiln on the southern slope below the quarry. Here, it is abrupt and w^ell defined, suggesting that seme of the lower chalk is absent, though perhaps not as much as at Molecap. The Mole- cap phcsphatic layer is absent here. The junc- tion ^s also expesed near the base of a small cut on the south-eastern slope of the hill. Here the junction is very irregular, suggesting the presence of local currents. Fossils found in tlii> cut include Porosphaera glohularis and casts cf other small sponges, Coelosmilia gin- ginensis. rare plates cf Holaster, very numerous plates of JJintocriiius. Kingena mesevihrinus. Ostrea etheridgei, Pycnodonta ginginensis. and large shark teeth, as well as a few ceprolites. The main quarry is 8 feet to 9 feet deep, the top be'ng about feur feet below the highest point of the hill. It was somewhat deeper at the s:uth-west corner, now partly filled in, w'here plates of Uintacrinus are common in the darker grey, more glauconitic chalk. Many fossils w^ere obtained during the excavation of the quarry. Glauert <1910. p. 117) stated that “the large Lamellibranchs are found in the upper portion of the main bed. and seem rare or entirely absent in the low’er strata, wdiore dw^arfed Corals, Brachiopods, Lamellibranchs and Gastropods, as well as numerous Serpulae and Echinoderm spines represent the animal life of the day ’’ Good specimens of ammonites are said to have been found near the .south-western corner, as well as a very few well- preserved echinoids resembling Holaster and Hemiaster. Marsupites has been recorded from the quarry, but appears to be very .scarce. The upper portion of the quarry is in fine- grained white chalk, which differs from that of the other exposures in the occurrence, in places, of fairly numerous tabular, lenticular layers of indurated and probably silicified chalk, usually thin, but up to 6 or 7 inches thick in places. Southern’s Chalk is situated about 2^ miles north-north-west of the railway station, about 50 chains north-west of the head of McIntyre Gully, and immediately north of the northern boundary fence of the Strathalbyn property. It is in a small gully which runs north from the fence to join a more mature west-running gully 194 feet farther north. The junction of the chalk with the Poison Hill Greensand is exposed a few* feet north of the fence. The chalk, of which a thickness of about 12 feet is exposed, differs from that of the other exposures, consisting of a very fine-grained even-textured fairly dark grey to putty-grey rock, with much glauconite in fine grains. The relative position of Southern’s Chalk is not easy to determine, as neither Marsupites nor Uintacrinus has been found here. Although actually situated at a lower level than Hosking's Chalk, 32 chains farther north, it appears to re- present a higher horizon, which may have reached its present position by slipping. The only fossils of index value found by the writer were a single small left valve of Ostrea philbeyi. associated with Pycnodonta ginginensis, from about four feet below^ the top of the chalk, and two faintly plicated small rhynchonellids from 6 or 7 feet below the top. These suggest that the exposure extends from about the middle of the Marsupites zone to about 7 feet above that zone, and that, compared with McIntyre Gully, about 45 feet is missing from the top of the chalk. Other fossils found here include a fairly large rudistid from about 6 feet below the top: some large Pycnodonta and Kingena from between 5 feet and 7 feet: also a few specimens of Boucliar- diella and Ostrea etheridgei, and the posterior plates of an amphineuran. A large capillate Magnithyris, associated with unusually large specimens of Inopinatarcula acantJiodes was found a little lower dowm. The Springs Gullies are situated on the lower slope of the escarpment between 60 and 70 chains north-west of the head of McIntyre Gully and about 25 chains north-north-east of Southern’s Chalk. They include three fairly deep narrow* gullies w*hich ow^e their origin to springs emerg- ing at a few feet above the top of the chalk. They 1 un westward to join a rather more mature gully from farther south, which, in turn, runs north-westerly to join a still older gully' from tlie north-east. The junction of the Ginsin Chalk w*ith the Poison Hill Greensand is well exposed near the heads of the tw^o more northerly of the three small gullies, which also give the best exposures of the chalk. The chalk here is coarse-grained and highly glauconitic, especially near the top. Specimens ot the small, finely ribbed rhynchonellid found between 23 fest and 28 feet above the base of Ihe chalk in McIntyre Gully are common between the top of the chalk and 6 feet below, associated with many relatively large specimens of Bcuchardiella and a few tiny Terebratulina. Spirulaea. Kingena and Inopinatarcula were found at about 6 feet in the middle gully. Fossils found below' 6 feet in the more northerly gully include Serpula. Spirulaea, a small conical sponge, Kingena. Inopinatarcula, Magnithyris and numerous Plicatula. A single small valve of Ostrea philbeyi, associated with a fairly PLATE I Perna coolyenensis X. — The holotype (48932); cast of right valve and ligament pits of left i^lve left valve, further enlarged to show pattern of shell fragment. X 2 X 11; 2. — The same (48932), '3.— Paratype (48933): cast of left valve. X li: 4. X i; (All from Molecap.) Cast of small right valve, X 1^; 5. — Cast of small left valve, Anomia fragilis 6.-Exterlor; 6a.-Interlor of holotype (48936), a left valve X 3; 7 -Exterior; 7a -Interior ^8937). a verv large right valve, X 2; 8.— Exterior; 8a.— Interior of a smaller right valve, X 3, 9.— ^terior of a ver> ^ small right valve (48938), with well-developed umbo. X 3. (All from McIntyre Gully.) Anomia prideri 10— Exterior; 10a.— Interior of holotype. a left valve (48939), X 2; U.— Exterior; lla.-Interlor of a right valve, the exterior covered by a colony of bryozoans (48940), X 2. (Both from McIntyre Gully.) 108 large Pycnodonta was found at about 8 feet below the top, and also portion of a new Chlaviys, A fragment of the carapace of a new crustacean was also found. Tiny specimens of Magriithyris and Terebratulina were found just below 9 feet. The horizons at which the index fossils were found suggest that approximately 40 feet is missing from the top of the chalk. Hosking’s Chalk is situated at the top of the western end of a long spur which runs west from the escarpment north of the Springs Gullies. It is 32 chains north-west of Southern's Chalk and one mile north-west of the head of McIntyre Gully. Only about six feet of chalk is exposed here, about half being in the Uintacrinus zone and half in the Marsupites zone. The exposure has proved to be fairly rich in the commoner fossils found at this horizon. Rarer forms found here include a few small smooth rhynchonellids, many Magni- thyris, plates of both Calantica ginginensis and Scalpellum glauerti, small Lamna teeth, and one probably of one of the Salmonidae. Good sr 17 Length '20- -2 19-1 1 20 IS") 10? It; U? Thickness 10 ■1 9- I Hinge of left valve of holotype, 16 mm. Description. — Shell of moderate size, the test thin and fragile; nearly equivalve. inequilateral, height greater than length, both valves moder- ately inflated, the right valve very slightly more so; beak sharp, apical angle about 65°; hinge-line long and straight, multiv'.n- cular, postero-ventral margin rounded, anterior margin slightly concave to nearly straight, posterior margin slightly convex and forming a very obtuse angle with the hinge-line. Surface of test with low, somewhat irregular but clearly defined growth-lines, about 2 to 24 mm apart at their widest on the holotype. Length of hinge nearly two-thirds the height of the shell. The interior of the hinge of the left valve of the holotype, 16 mm in length, is exposed to show six U-shaped ligament pits. Remarks. — The Gingin specimens show some variation in shape. The holotype and paratype show a slight resemblance to the casts from the Cambridge Greensand flgured by Woods (1905, p. 94), the holotype particularly to figure 19C, which, according to Woods, was referred to P. lanceolata Geinitz by Seeley, and to figure 19D, P. semielliptica Seeley and the paratype to flgure 19F, referred to P. subspathu- lata Reuss by Seeley. The ventral portion of the Gingin spee'es is, however, relatively wider than that of P. lanceolata and the anterior and posterior margins are straighter than those of P. semielliptica and P. subspathiilata and the hinge-line appears to be longer than those of the English specimens. The third large cast from which the beak and postero-ventral por- tion are missing was apparently relatively narrower than the others and more nearly resembles a specimen of P. raulineana d’Orb.. figured by Woods (1906, Plate XII, Fig. 9), as does the small right valve, of which, however, the ventral portion is less produced. The small right valve is the only specimen of which the horizon was recorded. This was the lower portion of the Marsupites zone, at flve feet above the base of the chalk. The other specimens most probably came fi’om about the same horizon. Superfamily ANOMIACEA Family ANOMIIDAE Gray Genus ANOMIA Linne, 1758 t Registered number. Geology Department Collection University of Western Australia. 109 Anomia fragilis, sp. nov. Plate 1. Figs. 6-9 Eighteen specimens, mostly small and some poorly preserved, were available for examination. These included ten from McIntyre Gully and eight from Molecap. Dimensions are given in Table II. TABLE II Dimensions of Anoviia fragilis. sp. nov. Iloln- tvix* Para- type 4S9:17 Para* tvpje 4S9:18 Shell and rant 1.. li. K. K. 1- K. valve valve valve valve valve \ alve miu. mm. mm. mm. mm. mm. lu-*) 19-7 9 ■ 9 UM 12-.^) 7-r> 10 1 :^2 0 lOO 1 0 • r> 10-8 7-4 Bfscripticn: — Shell rather small, thin to moderately thick, disc sub-circular, ovate, to subtrigonal in shape, relative proportions of height and length variable, usually slightly incQuilateral: nearly equivalvs, slightly con- vex. usually most mflated near the dorsal m'^i'Ern. Disc surrounded by a thin flange, apparently cf varying width, taut very imperfect in the specimens examined. Umbo small, at cr near the margin. Hinge very nar- row, edentulous. Holotype decorated with a number of well-defined, rather irregular over- lapping concentric lamellae cf varying width and traces of indistinct and rather irregular radial striae near the ventral margin. Interior of right valve shows nearly circular depressed muscle area, immediately below the hinge, about 6-2- mm in height on the holotype, with a large well-defined elongate elliptical adductor scar, situated immediately posterior to the median line and extending practically to the posterior edge of the muscular depression. The edge of the disc is marked by well-defined ridge, apparently strongest on the postero-dorsal portion. The pallial line is simple and fairly well defined. Interior of the disc generally smooth, rarely showing faint concentric ridging near the ventral margin. Left valve shows depressed nearly circular muscle area about equal in height to one-third the height of the shell, but with a U-shaped extension from its posterior end, extending ventrally to about half-way down the shell. Major byssal scar, elliptical in shape, high up under the umbo and immediately anterior to the median line: nearly separated into two halves by a narrow, curving ridge which enters it from its posterior side. Adductor scar sub- trigcnal in shape, situated immediately below the major byssal scar; minor byssal scar not so well defined, apparently more nearly elliptical in shape and situated at a distance from the adductor scar a trifle greater than the width of either and dorsally of a line at right angles to a line running through the other two scars. Remarks. — The largest right valve somewhat re- sembles in shape a specimen of A. pseudoradiata d’Orb. figured by Woods (1899, Plate VI. Fig. 2) but lacks the fine radial riblets of that species. It also resembles fairly closely A. subtrigonalis Meek and Hayden from the Fort Pierre group of the North America Cretaceous (Meek 1876, p. 22, Plate 16, Fig. 4a). Other specimens show a resemblance in shape to specimens of A. ponticulana Stephenson (1952, Plate 20, Figs. 1-4), but the umbo of that species is farther from the margin and the hinge is stronger. At McIntyre Gully the range of the species is from the base cf the chalk to 18 feet above. At Molecap, it has been found between one foot and seven feet above the base. Most of the specimens from Molecap are particularly fragile, especially the left valves, and it is almost impossible to obtain one intact. Anomia prideri, sp. nov. Plate 1. Figs. 10, 11 Two specimens from McIntyre Gully differ from those of the preceding species in their marked obliquity and in the greater depth of the hinge. The larger specimen (the holotype), a left valve, is from about 14 feet above the base of the chalk; the location of the smaller specimen, a right valve, was not recorded. Dimensions are given in Table III. TABLE in Dimensions of Anomia prideri, sp. nov. Rulutvpe I’aratvpe 48939 48940 1,. valve 1 K. valve mm. mm. Height 14-0 12-5 Length 15-8 , t I 11-0 1 Description. — Shell fairly small and moderately thin, obliquely ovate to piriform, inequilateral, the anterior half being somewhat the larger. Proportions of height to length variable, both valves slightly convex, the right valve rather more so. Umbo very small, marginal, directed slightly anteriorly. Hinge edentulous fairly thick and deep, particularly posterior to the umbo. Exterior surface of holotype nearly smooth, that of the paratype is almost wholly obscured by a colony of tiny bryozoans. Pallial line simple and well defined. Interior of right valve smooth, the muscle attachment area only very slightly depressed, with the large elliptical adductor scar immedi- ately posterior to the median line and extend- ing from about 2.5 to 5 mm below the umbo. The interior surface of the left valve is much eroded. The anterior margin of the specimen shows the remnants of a fairly wide flange; 6 or 7 tiny crenulaticns are present immediately below the hinge on the ridge separating the body of the valve from the flange. The muscu- lar area, circular in shape, extends to about half the height of the valve below the beak; the position and shape of the minor byssal and adductor scars are difficult to determine; the major byssal scar is slightly larger than that of A. fragilis; it is situated a little below the hinge immediately posterior to the median line. 110 with the adductor scar at about half way down the muscle area below it and slightly anterior to it; the minor byssal scar is very faint; it appears to be larger than the adductor scar, circular in shape and situated nearly opposite the middle of the major byssal scar below the posterior portion of the hinge. Remarks. — In its obliquity and shape, A. prideri resembles a species of Anomia from the Crackers of Atherfield. figured by Woods (1899, p. 28, Plate V, Figs. 4, 5) but it lacks the radial ribs of the English species. It also resembles somewhat some specimens of A. psaviatheis Bayan from the Auverian and Bartonian of Aquitaine (Cossman 1922, p. 215, Plate XV, Figs. 23-25) but the shape cf the muscular area appears to be different. Superfamily PECTINACEA Family . SPONDYLIDAE Genus SPONDYLUS Linne. 1758 Spondylus ginginensis, sp. nov. Plate II. Figs. 1, 2 Represented by four more or less imperfect specimens of united valves of varying size, three from Molecap and one from Musk’s Chalk, as well as a small imperfect right valve showing the interior and wholly attached to a fragment of Incoceramus from the North Chalk. Dimen- sions are given in Table IV. TABLE IV Dimensions of Spondylus gingmensis. sp. nov. Uolotvpc 48941 .Molorap l*aratvrM‘ 48942 Molerap -Mnlccap 48948 Chalk I.. Villv«‘ 1 _l K. valv»' 1-. valv<‘ IE, 1 ,. valve j IE. 1 valve 1.. IS. valve valvi- min. mm. mm. mm. 1 mm. mm. mm. mm. IJeiuhf 4.5 • (5 51 5?| 87 40-.S?! 80 • 2 80 • 4 29-0 82 5 liPIlKtll 40 : 44 II? 8.5 • 2 .87 ()?: 28 0 V 28-2 20-0 Thirk- iu*ss ■6 2s ■3 14 ■ II 20-7 Description. — Shell fairly large, ovate, oblique, height greater than length, the left valve usually highly inflated the right valve less so, but higher and perhaps slightly wider. Attached by the right valve, usually by the umtaonal part only, this part usually much produced, some speci- mens showing a marked gap between the um- bones of the two valves. Exterior surfaces of both valves ornamented with 70 to 80 fine evenly-spaced radial threads, and rather indis- tinct concentric growth lamellae; fine concen- tric threading, more noticeable on the right valves, is also present. A number of irregularly spaced small spines are present on the dorsal half of the left valve of the paratype. Left valve usually highly convex, the beak terminal and sharp, and curved approximately at a right angle to the commissure. Postero- dorsal margin of body of shell, where joined by ear. concave, the anterior and ventral margins convex. Ears descending slightly from the umbo; posterior ear large, extending on the holotype, from the beak to 24 mm below: dorsal margin straight, distal margin concave: anterior ear smaller, extending to 17 mm below the beak, dorsal and distal margins straight, meeting at a very obtuse angle. On the paratype the width of the posterior ear is about 8 mm. Both ears show faint concentric threading continued up- wards from the body of the valve. The left valve of the holotype is ornamented with about 80 radial threads. On the other specimens the number is about 70. Right valve less convex than the left, but considerably higher and apparently slightly longer. The umbo is produced well above that cf the left valve. The area of attachment is very variable in size; on tho holotype the at- tachmsnt is to a smooth surface of Inoceramus, on the paratype the umbonal half of the at- tached portion shews fine radial ribbing, the ether half well marked concentric lamellae: the other two specimens do not show any area of attachment. Ornament similar to that of the left valve, except that spines appear to be absent. Remarks. — The third specimen cf which the dimensions are given is much flatter than the others, its thickness being only half the dimen- sion of the length of the shell, wnereas in the other specimens the proportion is three-quarters. It may represent a different species, but in some European species the proportions of thickness to length appear to be variable. In the Musk’s Chalk specimen, the point of greatest width is situated higher than in the other specimens. S. gmginensis appears to resemble most nearly the European species S. gibbosus d'Orbigny, particularly the specimens from the Cambridge Greensand figured by Woods (1901. Plate XX. Figs. 5-11), but the right valves of the Gingin specimens are relatively flatter, except in the Musk’s Chalk specimen, and the ears of the left valves appear to be larger. Woods (1901, p. 118) states that the right valve of S. gibbosus is variable, flattened when attached by its entire surface, more convex when attached by a part only. The regularly spaced stronger ribs found on some specimens of S. gibbosus (Woods 1901, Fig. 5) are absent from the Gingin specimens. The holotype of S. gingineiisis was found in a block of chalk that had fallen from the back of the Molecap quarry. The paratype was from 2 ft 6 ins. above the base of the Gingin Chalk, at the junction of the Uintacriiius and Mar- supites zones. Judging from the dark brown colour of the test, the Musk’s Chalk specimen must have come from near the base of the formation at that locality. The North Chalk specimen was from the junction of the Uinta- crinus and Marsupites zones. The range of the species would appear to be the Uintacrinus zone and lower half of the Marsupites zone. Four imperfect right valves, wholly attached to fragments of Inoceramus, similar to the North Chalk specimen, were found by Dr. Glenister and Mr. Balme in the Toolonga Calcilutite of the Murchison River valley at Thirindine Point and a similar fairly large specimen was ob- tained by Clarke and Teichert from Pillarawa Hill farther north, as well as two small ones from Meanarra Hill south of the Murchison River. Ill PLATE II Spondylus ginginensis Right valve of holotype (48941), from Molecap, nat. size: la.— Left valve of same (48941), nat size; Posterior profile (48941), nat. size: 2.— Paratype. right valve (48942), nat. size: 2a. — Paratype. left valve (48942). nat. size. Plicatula glauerti 3 Right valve of holotype (48944); 3a.— Left valve; 3b.— Posterior profile. Hosking’s Chalk, X 3; 4. — ExteriOT of paratype. a right valve (48947); 4a.— Interior, showing teeth. Molecap, X 3; 5.— Interior of right valve, with teeth Middle Springs Gully (48946), X 3; 6. — Exterior of imbricated right valve (48948); 6a. — Interior showing anastomosing ribbing. McIntyre Gully. X 3; 7. — Exterior of a left valve (48945); 7a. — Interior showing ribbing, Hosking’s Chalk, X 3. 112 Genus PLICATULA Lamarck, 1801 Piicatula glauerti, sp. nov. Plate II, Figs. 3-7 About 150 specimens, mostly right valves, from various exposures, were available for examination. Left valves are comparatively rare and I have only seen three specimens of united valves. The hinge with the charac- teristic Piicatula teeth was preserved in only four right valves. The umbo was absent from nearly all the specimens. Dimensions are given in Table V. TABLE V Dimensions of Piicatula glauerti. sp. nov. 1 ]*aru- I'nra- ' Ilolotvpe t V]»0 tVTH' Imhricalod 48945 1 48944 48947 4894(1 . 48948 Hos- Mnle- 1 kiiigs ' eap i iloskiuKs Mole- Spi'inus Meinlyre Chalk , Chalk ' 1 9-1 1 (M) 1 90 lO-O 80 7 • 5 (1-0 l.cufftli 8-8 8-7 1 ()• 1 B-5 9;i 8-1 ()-7 (>•‘1 The thickness of the united valves of the holo- type is 3.9 mm. Description. — Shell small, obliquely ovate to broadly piriform, the degree of obliquity vary- ing; height and length nearly equal, but height usually the greater: inequivalve, the right valve moderately to strongly inflated except where wholly attached, the left valve rather less so. Attached by the right valve, but individual specimens range from unattached to wholly attached: in many, the area of attachment appears as a small subcircular truncation of the umbonal area; this is present in about 40 of the specimens examined; margins rounded. Body of shell thin near the umbo, thickened at the margins, especially the ventral margin, the width of the thickened portion varying considerably. Pallial line simple, well-defined on better preserved specimens. Right valves range from those in which the whole exterior surface is covered with from about 20 to 40 closely spaced small radial rib- lets or plications to those from which radial ornament is absent; the riblets may be confined to the ventral half of the shell. They usually become obsolete near the anterior and posterior margins. Very rarely a single rib bifurcates near the ventral margin. The ribbed forms appear to predominate, although ribless speci- mens are fairly common. Concentric ornament of a number of closely spaced slightly irregular growth rings, of which from three to five may be more prominently imbricating. The interior surface appears to be slightly eroded as I could find no trace of an adductor scar on any of the specimens, the innermost layer of the shell apparently consisting of more soluble material. The interior decoration visible consists of 16 to 20 rather irregular bifurcating riblets. The hinge shows the typical divergent teeth, of which the anterior is slightly the longer. The left valves are usually more nearly circular in outline, and are less inflated than the right valves, and very rarely may be even slightly concave; radial ribs are absent, the valves showing only concentric ornament, similar to that of the right valve. Where the hinge is present there is usually a nearly circular hole immediately below. The interior decoration consists of 15 or 16 anasto- mosing riblets. On some small specimens the interior ribbing is very indistinct, the ornament consisting of four or five concentric threads. Remarks. — Piicatula glauerti is one of the commonest Gingin fossils and has a wide vertical range extending, at McIntyre Gully, from the base of the chalk to about 43 feet above. It is particularly common at Molecap aird Hoskings Chalk and has been found at the Springs Gullies and the North Chalk, but. so far. I have not found it at One Tree Hill or Southern’s Chalk. In general appearance and in type of ribs the Gingin species most nearly resembles the European Cretaceous species figured by Woods. P. minuta Seeley (16-20 ribs) and P. barroisi Peron (13-25 ribs) (Woods 1901. p. 138, Plate XXV, Figs. 22-25 and pp. 141-143, Plate XXVI. Figs. 12-18 respectively), particularly the former species. Woods .says that it seems probable that P. minuta is only the young form of P. gurgites (Woods, 1901, Plate XXV, Figs. 13-21) but that species shows little resemblance to the Gingin shell. P. glauerti is, however, consider- ably larger than P. minuta and its ribs are usually more numerous. P. barroisi also is smaller than the Gingin species, its ribs are usually fewer, and bifurcation of the ribs is common. The Gingin species also resembles fairly closely specimens of the supposed genus Diplo- schiza, founded by Conrad on imperfect speci- mens of a single species D. cretacea Conrad and revived by Stephenson (1934, pp. 273-280, Plate 38). The Gingin shells are rather more oblique than those of Conrad’s species and plicated specimens resembling those of Stephenson’s variety D. cretacea striata (Stephenson 1934, Plate 38. Figs. 15-17) are more common; also the ribless form of the Gingin species appears to be rather more rugose than Stephenson’s neotypes (Stephenson 1934, Figs. 3-9). Stephen- son’s specimens were from the southern ex- tension of the Pecan Chalk Member of the Taylor Marl of Texas. The genus Diploschiza was founded on speci- mens of which the hinge was imperfect and from which the innermost layer of the shell was missing. According to the description of the genus given by Shimer and Schrock (1944. p. 407) teeth are absent, and the inner surface of both valves is lined to near the margin with fine sharp irregularly spaced radiating ridges. Stephenson (1934, p. 276) stated that the hinge is edentulous, but also mentioned that “an occasional right valve supported by the fragment of extraneous shell to which it is attached exhibits a faint suggestion of a pair of small short crural ridges diverging inwardly” (the Piicatula teeth). He suggested that the absence of muscle scars on the inner surfaces 113 PLATE III Ostrea philbeyi 1 —Exterior- la Interior of holotype: a left valve (48950). nat. size; 2. — Interior of a left valve: the largest soeclmen seen (48957) nat size; 3. — Exterior of a slightly alate left valve (48958), nat. size; 4. — Interior of a riiht valve attached to a larger left valve (48951), nat. size; 5.— Interior of a small right valve (48952), nat. (All from Mclntvre Gully.) 6.— Exterior: 6a.— Interior of a small left valve (48954). Nungajay Springs area, 8^ miles N. of the mouth of Murchison River. Nat. size. Ostrea macintyrei n Pvfprinr- 7fl —Interior of holotype. a left valve (48955), x 2; 8.— Exterior; 8a.— Interior of a small right u^xterior. /a. xntc (48956), X 3. (Both from McIntyre Gully.) 114 “may be due to the failure of preservation of a thin inner layer of the shell.” This is a regular feature of the Gingin species and Eudes-Des- longchamps (1858, p. 2) has mentioned that some of the Plicatulas from the neighbourhood of Caen were greatly attenuated in the cardinal region, and did not show in the interior any trace of teeth, groove, ligamental cavity, or muscular impression, the intex'ior surface show- ing only growth striae, whereas others, from rocks of a different character, with thick test in the cardinal region showed the muscle im- pression, the teeth and groove. He attributed the different states of preservation of the two groups to their occurrence in rocks of different character. Without doubt the shells of Conrad’s species are really Plicatula from which the teeth and innermost layers of the shell have been eroded. Therefore, as a generic name Diploschiza is invalid. The wholly attached valves of P. glauerti are most commonly found on nearly flat surfaces such as fragments of Inoceramus, valves of Ostrea philbeyi and right valves of Pycnodonta ginginensis, but I have found them also on O. etheridgei, on a larger Plicatula and even on the small brachiopod Bout char diella cretacea. Superfamily OSTREACEA GolaTuss Family OSTREIDAE Lamarck Genus . OSTREA Linne, 1758 Ostrea philbeyi,* sp. nov. Plate III, Pigs. 1-6. Ostrea sp. a. R. Etheridge. Junr., Geol. Surv. Bull. No. 55. p. 17, Plate IV. Figs. 8. 9, 1913. W. Aust. Etheridge’s brief mention of this species is: “A single example of a very thin -shelled flat valve, of common form and with no particular characters. The type is, however, new to our Cretaceous recks, and therefore of interest, but resembles an Oyster met with in the oolitic beds of the Greenough River.” His exce’lent drawings are of a fairly well-preserved speci- men a trifle more oblique than the holotype. A fairly large number of specimens frem McIntyre Gully, nearly all left valves and mostly somewhat eroded, was available for examination, as well as single small left valves from Southern’s Chalk and the northernmost of the Springs Gullies. Right valves are com- paratively rare and only four were available. Dimensions are given in Table VI. All specimens from McIntyre Gully except 48954 which is from Nungajay, Murchison River area. Description. — Shell fairly large, moderately thick, ovate, slightly oblique, height slightly greater than length, nearly equivalve but the right valve may be slightly smaller than the corresponding left valve, inequilateral, rarely alate. Left valve slightly inflated, right valve nearly flat. Umbo straight, or slightly curved, fairly sharp when well-preserved, left valve occasionally showing small area of attachment. Ligamental groove broadly triangular; liga- • After the late Mr. W. R. Philbey, who collected the specimens described by Mr. Etheridge. 115 TABLE VI Dimensions of Ostrea philbeyi, sp. nov. Holo- type 48950 Para- type 48957 Alate 48958 : 48951 L. valve with small R. valve at- lachedi 48954 L. valv<‘ valve I,. valve L. valve L. valve Right valvcR Height Length mm. 37-8 32 • 8 mm. 65 62 mm. .52-5 50-5 inm. 50 44 mm. 46-7 37 mm. 30-5 24-5 mm 38 1- 32-5 27 36? 30-6 24 -5 mental area large, with well-marked trans- verse striae. Exterior surface fairly rough, with closely spaced rather irregular growth larnellae. Interior surface smooth except for the well- defined adductor scar. Depressed area for accommodating the body of the animal well- defined, extending below the umbo for about two-thirds of the height of the shell. On the Ulterior of the valve the anterior submargin, immediately below the ligamental area, shows a line often to eighteen short transverse, rarely bifurcating crenulations extending usually to approximately opposite the top. rarely the middle, of the adductor scar. On the posterior submargin is a .shorter, wider area of slightly waved crenulations. Adductor scar fairly large, nearly elliptical in outline, the centre situated from slightly less than one-third to about two- fifths the height of the shell below the umbo. Right valve usually flatter and thinner than the left. So far as could be judged from the poorly preserved specimens the exterior surface IS smeother than that of the left valve. The crenulations on the interior submargins appear to be similar to those of the left valve. The position of the adductor scar varies consider- ably. On the tw^o larger specimens and also cn a small specimen (48953) from the Mar- supites zone it is situated at about the same height as those of the left valves, but on the .smallest specimen figured it is about half-way down the valve. In Clarke and Teichert’s large and compre- hensive collection from the Murchison River area the only representatives of O. philbeyi are feur small specimens from the Nungajay Springs area on the westward-facing escarp- ment about 8^ miles north of the mouth of the Murchison River. Two of these specimens are similar in shape to the McIntyre Gully specimens, but the other two. and particularly the one figured (Plate HI. Figs. 6. 6a) are narrower and more obliquely piriform in shape; the anterior margin is curved inwards, the shell is much thicker, particularly the interior marginal portion enclosing the pallial line: and the exterior surface more rugose. These features suggest their growth in waters more disturbed by local currents than those in the McIntyre Gully area. Remarks. — Etheridge (1910) stated that O. philbeyi resembled “an oyster met with*'in the Oolitic beds of the Greenough River,” prob- ably from the Newmarracarra Limestone, east of Geraldton. Etheridge’s figure shows the interior of a right valve obliquely subtrigoaal in outline (Etheridge 1910. Plate IX, Fig. 2). The specimen does not show the large triangular ligamental area characteristic of O. philbeyi and the adductor scar is larger, apparently less deeply cut. more centrally situated and more nearly circular in outline. The Gingin species somewhat resembles the European 6. leymerii Leymerie (ex Deshayes) (Woods 1912. pp- 355-358, text-figs. 139. 140 > but the latter is a much larger and apparently thicker shell. In shape and size O. philleyi most near-y resembles the widely spread O. acutirostra Nillson, especially to the rather pooily preserved specimens from the Carclita teaumonU beds of Baluchistan figured by Coss- man and Pissaro (1927. Plate I. Figs 1-5). Possibly ow'.ng to erosion, these do not show the ligamental groove and transversely striated aiea characteristic of O. philbeyi. The shells eppear to be thicker, the adductor scar of the Ipft valve is situated much lower than m the Gingin shell, being below the transverse median line and where the specimens are slightly alate, the alation is anterior instead of posteiioi. The slightly curved beak shown by some of the specimens frcm the Arrialoor beds figuied by Stoliczka (1871. Plate XLV, Figs. 3 and 3a) is absent from the Gingin species. Stoliczka s specimens show the ligamental groove and area. As far as I know. Ostrea philbeyi has not been found below the uppermost foot of the Marsnvites zone, in which one small light valve was found, and at McIntyre Gully is piactically nstiicted to a zone between about 21 feet and 23 feet above the base of the chalk. Although found in the main gully it is commonest in the second small eastern tributary. Fiorn its restricted occurrence, the species appears to be a good horizon marker. Ostrea macintyrei, sp. nov. Plate III. Figs. 7. 8 n^trea sp or Pycnodonta sp. (JuveiUle foi-ini: R. gthendge. Junr., -Geol., Surv^ Aust- Bull. No. 55. p. 19. p. 29. Plate Til. Figs. 10, 11. 191J. Etheridge'S figures show a small right valve of which the original dimensions were apparently about 14 mm in height by 13 mm m length, unfortunately, the ventral portion is imperfect, and it is impossible to say whether the specimen was originally ovate or subtrigonal in shape Etheridge evidently regarded his specimen as a young form of some larger shell, but it was more probably an adult shell as the only othei specimens that could be assigned to the same species are even smaller. A single small left valve *48955), height 12.3 mm. length 14.2 mm. from McIntyre Gully at about 17 or 18 feet above the base of the chalk, doubtless belongs to the same species as Etheridge’s specimen, as does a sinall right valve (48956). height 9.8 mm. length 10 mm. from about 19 feet above the base. Description. — Shell small, fairly thin, both valves slightly convex, obliquely subtrigonal to ovate: length usually slightly greater than height, nearly equivalve, inequilateral, the posterior portion produced; umbo small, shaip, except where truncated by area of attachment; attached by the left valve; posterior margin straight and making an obstuse angle with the ventral margin, anterior margin rounded and passing almost insensibly into the ventral margin: exterior surface rugose with strongly marked concentric ridges and rather faint growth lamellae. Interior surface smooth, but shallowly grooved in harmony with exterior concentric ridges. The left valve shows a small but prominent area of attachment immediately behind the beak. The exterior shows two fairly prominent rather irregular concentric ridges as well as fine, rather indistinct growth lamellae. The hinge is small and shows a rather narrowly triangular ligamental groove, and shallow triangular area on which transverse striae could not be de- tected. On the posterior submargin, immedi- ately below the hinge, is a short, broad area showing three fairly strong dental crenulations. nearly parallel to the median line. A slightly longer line of crenulation is present on the much narrower anterior submargin. The in- terior of the valve is shallowly grooved in harmony with the exterior ridges. The adduc- tor scar is fairly large, rather shallow, elliptical in outline, and is situated at a distance below the hinge nearly equal to its major diameter. The pallial line is well defined and remote. Right valve similar in shape to the left. Its exterior also shows two low broad rounded con- centric ridges separated by a narrow groove; growth lamellae very faint. The rather narrow ligamental area is very similar to that of Etheridge’s specimen and does not show any defined groove. Interior of valve with faint grooves corresponding to the exterior ridges. Adductor scar rather small and very shallow, semi-elliptical in shape and situated at about one-third of the distance from the median line to the postero-ventral angle. Remarks. — This species is very rare. The only two examples of which the locality is known are from the uppermost portion of the Mar- supites zone. So far as I can determine it does not ressmble any of the European Upper Cretaceous Ostreidae. Ostrea etheridgei, sp nov. Plate IV, Figs. 1-4 Ostrea sp. b. R. Etheridge, Junr. Geol. Surv. W. Aust Bull. No. 55. p. 17. Plate II. Figs. 19-21, 1913. Etheridge stated : — ‘T have associated to- gether a few very small and delicate valves, perhaps only the young of some larger form. They are rudely deltoid in outline, with high pointed umbos, and very short area. The adductor impressions, on the other hand, are large for such small shells. The two largest have lost much of the pointed umbo feature, and the valves have broadened out.” Etheridge’s figures are of two small, rather squat specimens, not very characteristic in shape. Right valves of this species are among the commonest Gingin pelecypods. but left valves are rare and I have only seen four small speci- mens. of which only one is reasonably well preserved. Dimensions are given in Table VII. 116 PLATE IV OStVQCL CthSTid'ffGt' l.-Exterior; la.-Interlor of holotype, a right valve (48959), X 2; 2.-Exterior; right valve (48960), X 2; 3.— Exterior of a large piriform right valve (48961). X 2, 4.— Exterior, 4a.— Interior of a ^ small left valve (48962), X 3. All trom McIntyre Gully except (48961). Gryphaea teicherti (23025) 5— Exterior of cast of a left valve, nat. size; 5a.— Posterior profile of same cast, nat. size; 5b. -Front view of cast, nat. size. Locality uncertain. Gryphaea viinuta (48964) fi Tflst of left valve of holotype, X 4; 6a.— Cast of right valve of holotyp^ showing trace of test, X 4; 6.— Cast of left vaive oi holotype. X 4. From McIntyre Gully. Pycnodonta ginginensis. Eth. fil. 7.— Left valve; 7a.— Bight valve of a well-preserved specimen from Molecap (48965), X 1^. 117 TABLE VII Dimensions of Ostrea etheridgei. sp. nov. Kialif valvfv^ Left valves Holii- ; Para- lvp(‘ 489.M) 4SIMU) .M<- Mc- Intyre Irityre (iJiilly (Jiiily l.arg- est Spec. 48901 Mole- , cap ' i Para- type , 48902 Mc- Intyre Oiilly Alate 1 , .Mole- 48900 Aorth p..,, Chalk i 1 nun. mm. mrn. mm. mm. 1 1 mm. ' mrn. Height 1:0 2 lO-O 8-r> 11-8 1 10 -7 8-9 Lengtli 8-2 9-7 1! :{.5 O-,") , 12-0 1 lO-O 0-9 Description. — Shell small, fairly thin, shape very variable, usually considerably higher than long, obliquely piriform or ovate, the degree of obliquity varying greatly, inequivalve, usually very inequilateral, rarely nearly equilateral; beak usually directed posteriorly, rarely upright: right valve slightly convex, left valve more so: ligamental area varying in depth. Posterior margin of right valve usually strongly concave below the beak for about half the height of the shell, passing insensibly into the convex ventral margin, but it may be nearly straight or even slightly convex; anterior margin convex. The exterior surface slightly rugose to nearly smooth with numerous closely spaced growth lamellae. Interior surface smooth except for the unusually large, shallow, more or less elliptical or piriform adductor scar of which the centre is situated from a little more than three-fifths to nearly three-quarters of the distance from the beak to the ventral margin. Ligamental groove apparently absent; ligamental area varies somewhat in size, and usually shows faint transverse striae. Marginal crenulations are absent. Left valve very similar in shape to right but considerably more convex. The paratype (Plate IV, Figs. 4. 4a ) appears to show a small flattened area of attachment immediately below the beak. The ligamental area is rela- tively large. The adductor scar is very faint, but appears to be elliptical in shape, and situated immediately below the hinge-line and slightly nearer the median line than the posterior margin. Remarks. — At McIntyre Gully, right valves of this species are common between the base of the chalk and 20 feet above. Above that horizon they appear to be rare, but have been found as high as 42 feet above the base. They ai'e also common at Molecap and other ex- posures in the Uintacrinas and Marsupites zones. Etheridge (1913, p. 17) compared the outline of this species with that of a specimen described by Meek (1876. p. 18. Plate XI, Pigs. 4a, b) as O. patina var. C. but the latter is a much larger, coarser shell. There is also seme resemblance to Gryphaea ar7'ialoorensis Stoliezka <1871, p. 464, Plate XLV. Figs. 13, 14) but the left valves of that species appear to be more slender and more oblique than those of O. etheridgei. On the whole the Gingin species appears to resemble most nearly the specimens of O. incurva Nilson figured by Woods (1912, Plate IX). Like those of O. etheridgei individual speci- mens of O. incurva vary greatly in shape, and except that the Gingin specimens do not show the radial riblets seen on some of Woods’ speci- mens, especially his Pigs. 12, 14 and 16. the re- semblance is very close. The beak of O. incurva, however, appears on the average perhaps to be more sharply curved posteriorly and the adductor scar of the right valve is situated much higher and is more elliptical in shape than that of O. etheridgei. Genus GRYPHAEA Lamarck. 1801 Gryphaea teicherti, sp. nov. Plate IV, Figs. 5, 5a. 5b This species is represented by a single cast of the left valve of a typical Gryphaea from which the lower portion of the anterior part is missing (23025). Unfortunately, neither the exposure nor the horizon from which it came was recorded- but the partly brown-stained somewhat glauconitic chalk of which it is com- posed closely resembles the lowest foot of the Molecap Chalk. There is only an irregular mass of chalk where the right valve should be. Ap- proximate original dimensions of the specimen are: height 38.5 mm, length probably about 30 mm, thickness 21 mm. Shell inequivalve. Left valve fairly large, higher than long, strongly inflated, inequilateral, slightly oblique, exterior contour in a fairly even curve, slightly more marked at the umbo. Exterior of cast smooth. Posterior half noticeably alate, wuth a shallow depression between the alate portion and the body of the shell; alation extending nearly to the ventral margin. Anterior half of specimen imperfect but apparently without alation. Valve apparently narrowing towards the ventral mar- gin. Umbo prominent, directed approximately at right angles to commissure, rather short, about 4 mm in length and about 6.5 mm thick at the base, distal end rounded, slightly incurved. Araong the European Cretaceous species fig- ured by Woods. G. teicherti appears to resemble most nearly G. vesiculosa Saw (Woods 1912. Plate LV, Figs. 10-14; Plate LVI. Figs. la. lb» in general shape, but the beak of the Gingin species is much thicker, blunter and more pro- minent than those of Woods’ specimens. In profile, G. teicherti is very similar in shape and convexity to a specimen of G. vesicularis Lam. figured by Stoliezka (1871, Plate XLIII, Figs. 1. la) but here also the beak of the Gingin vspeci- men is larger and more rounded. Gryphaea msnuva, sp. nov. Plate IV, Figs. 6, 6a, 6b This species is represented by a ,sin:^'le well- rreserved cast of united valves from McIntyre Gully (48964), so small that were it not for the number of well-defined growth rings, one would regard it a young specimen of a larger species. Dimensions.—Height 4.8 mm: length 4 mm: thickness 2.3 mm. The height of the right valve was probably 3.7 mm. Description. — Shell very small, outline ovate approaching trigonal, higher than long, nearly equilateral, very inequivalve; left valve convex, right valve flat or slightly concave. 118 Left valve nearly semi-circular in profile, its continuity broken by several well-defined growth lamellae. Anterior margin strongly convex near the umbo and the ventral margin, nearly straight between: posterior margin rather more eveirly convex; ventral margin straight in the middle, strongly convex at junction with anterior and posterior margins. Beak small, prominent, fairly sharp, slightly incurved. Exterior surface with apparently five growth ridges of which the middle three are more prominent, and are fairly high and rounded. Right valve broadly ovate in outline; it shows faint traces cf the test. The exterior surface appears to be smooth and no growth lines could be distinguished. I have been unable to find descriptions of any ether species resembling G. minuta at all closely. Subgenus PYCNODONTE Fischer de Waldheim 1835 Pycnodonta ginginensis Eth. fil., 1913 Plate IV. Fig. 7; Plate V, Figs. 1-3 Pyc7iodonia gingmensis R. Etheridge junr. Geol. Surv. W. Aust. Bull. 55. pp. 17-19, Plate III. Pigs. 6-9. Plate IV. Figs. 3-7. 1913. Pycnodenta ginginensis v/as described in detail and figured by Etheridge but he made the mis- take (Etheridge 1913, p. 18) of regarding the lower inflated valve as the right valve instead of the left. Although, as stated by him, the abductor impressions are sub-central, they are definitely posterior to the median line. Apart from some species of Inoceramus, P. ginginensis is probably the commonest pelecypod in the G ingin area and it appears to be equally common in the Murchison River area, but the shells, though thick, are brittle and really well- preserved specimens are rare, few being suffici- ently well-preserved for accurate measurement. With the exception of a specimen (Plate V, Figs. 1 la) from Toclonga Hill near the Murchison River, I have not seen any as complete as those shown by Etheridge’s Plate IV, Figs. 5-7. Dimen- sions are given in Table VIII. TABLE VIII Dimensions of Pycnodonta ginginensis Left valvi'.-i Kight valves Mr- iiit yrc (Jully 4S9t)5 MnU- rat> Mole- cap 4S9(i() Too- lolliiJl Hill 48!)«)7 1 : Moh!- eap 4.s‘M)8 .Me- Intvre (iully Height, l.fUfftil mm. 42-U? 58-1) luin. 30-7 32 • 0 I ' mm. 30 0 31 -r> mm. 3(5 0 37-5 , mm. (52-5 (H»-U mm. 5S ? 72-5 The two right valves are the largest I have seen. Only fragments of the corresponding left valve of the McIntyre Gully specimen were re- covered and this was evidently even larger than the right valve. These fragments show a very large tabular area of attachment which makes an obtuse angle of about 105° with the sides of the valve. Description.— Shell large, thick, longer than high, oblique, inequilateral, very inequivalve, the left valve highly inflated, globose, and larger than the right, the right valve flat or concave; attached by the left valve. Left valve variable in shape, alate. the pos- terior alaticn usually the larger and more strongly lobate, occasionally separated from the body of the valve by a deep narrow groove. An- terior alation not always noticeable and rarely lobate. Exterior surface smooth except for the widely spaced, somewhat irregular margins of thin growth laminae. In some large specimens, these are very noticeable and imbricating. Ethe- ridge’s Plate IV, Fig. 7, shows the umbo as fairly long and strongly incurved with the point close to the body of the valve; almost invariably how- ever, the beak is truncated by a flat or concave area of attachment which may be small or may occupy a large proportion of the surface of the valve (Plate V, Fig. 3a). Owing to truncation of the umbo the ligamental area and groove are rarely present; the groove is fairly wide and shallow; the rather small area shows faint trans- verse striae. Interior of the valve smooth: ad- ductor scar fairly large, shallow, semi-circular in shape and situated in the dorsal half of the valve, immediately posterior to the median line. Right valve broadly and obliquely elliptical in outline; usually slightly concave, but some valves are very slightly convex with reverted edges (Plate V, Figs. 2, 2a); in others, the umbonal half is convex, probably corresponding to the attached area of the left valve, the remainder concave and making a considerable angle with the convex portion, a well-marked groove separ- ating the two portions. The umbo is rarely very noticeable being, as a rule, raised only slightly above the general surface, but in some concave valves the umbonal portion takes the shape of a low dome or boss, about equal in length to half the length of the valve (Plate IV. Pig. 7a). Exterior surface usually smooth, with faint growth rings; rarely, fairly regularly spaced faint radial striae are also visible. Cardinal margin long and straight: the ligamental area varies in size, but is usually wide. In the large Molecap valve (Plate V, Figs. 2, 2a). it is about 30 mm in width, with a wide groove, but it is relatively small in the even larger McIntyre Gully specimen. Only two specimens, from Thir- indine Point in the Murchison River area, are sufficiently well-preserved to show the trans- verse dental crenulations on each side of the ligamental area figured by Etheridge (1913, Plate IV, Fig. 3). Adductor scar large, semi-circular to nearly circular in shape, deeply incised in aged shells, and situated immediately posterior to the median line and usually just above a transverse median line. Remarks. — At McIntyre Gully, the vertical range of P. ginginensis is from just above the base of the chalk to about 42 feet above. The largest specimens are usually found in the upper half of the Marsupites zone, the largest found by the writer being from about 18 feet above the base of the chalk. At Molecap, the largest speci- mens occur near the middle of the Marsupites zone. The general resemblance of P. ginginensis to Pycnodonta vesicularis Lam. was noted by Ethe- xddge (1913, p. 19), and the resemblance of speci- mens both from Gingin and from the Toolonga Chalk of the Murchison River area to the speci- mens of P. vesicularis figured by Woods (1912. Figs. 143-182) is so close that doubt arises as to 119 PLATE V Pycnodonta gxnginensis Eth. fil. 1. — Exterior; la.— Interior of hypotype, a left valve from Toolonga Hill, Murchison River area (48966), X li; 2. — Exterior; 2a. — Interior of an aged right valve from Molecap (48967), X 1; 3. — Interior of a very large right valve from McIntyre Gully; 3a.— The same, with fragments of left valve superimposed (48968), X §. Pycnodonta strathalbynensis 4— Exterior of left valve of holotype (48969). X 2; 4a.— Exterior of right valve of holotype (48969), X 2; 4b.— Posterior profile of holotype (48969). X 2. From McIntyre Gully. 120 whether they do not actually belong to the one species. The radial striae characteristic of the right valves of P. vesiciilaris were visible on only two right valves from Molecap, the exterior sur- faces of other right valves both from Gingin and from the Murchison River area being too eroded to show whether this feature was origin- ally present. In Etheridge’s profile (Etheridge 1913, Plate I, Fig. 7) the beak is more incurved than that of Woods’ Fig. 178 (Woods 1912, p. 371) but in some small specimens from Thirindine Point in the Murchison River area, the curva- ture is very similar or even less than that of Woods' figure. I have not seen any right valves of which the umbo is as well developed as that of Woods’ Fig. 150 (p. 367) but on the whole, the great similarity between that of P. gingineusis and P. vesicularis suggests that they belong to the same species and that any differences are those of individuals. Pycnodonta strathalbynensis, sp. nov. Plate V. Figs. 4, 4a, 4b, Plate VI. Fig. 1 A single fairly well-preserved small shell showing both valves, from McIntyre Gully at about 37 feet above the base of the chalk (48969), differs greatly in its proportions from typicai specimens of P. gingineusis and appears to re- present a different species. The horizon from which it comes is nearly 20 feet above that at which the largest specimens of P. gingineusis occur. Dimensions are given in Table IX. TABLE IX Dimensions of Pycnodonta strathalhyensis, sp. nov. I.cft villvo Hi'jlif vnlM- The thickness of the united valves is 8.6 mm. Description.— Shell small, fairly thin, higher than long, inequilateral, obliquely ovate, very inequivalve, the left valve inflated and larger than the right, the right valve convex in the um- bonal area, the remainder concave: attached by the left valve. Left valve grypheate, fairly strongly inflated, obliquely ovate; the lateral margins somewhat compressed to form a rounded ridge extending from the umbo to the ventral margin: slightly alate posteriorly, the alate portion separated from the remainder of the valve by a shallow furrow: anterior portion of the valve slightly concave for a short distance immediately below the umbo, the remainder convex: posterior mar- gin very slightly concave for about a quarter of the height of the valve, becoming strongly con- vex in the alate portion; the anterior-ventral and postero-ventral margins meet in a fairly sharp arch. Exteiior surface of valve slightly rugose with fairly distinct growth rings, shaped similarly to the shape of the valve. Umbo fairly high and grypheate, but truncated in the holo- type by a small, slightly concave ai'ea of attach- ment. Ligamental area high and fairly wide with a narrow groove directed posteriorly and set obliquely with the top immediately below the posterior edge of the truncated umbo, the anterior edge of the triangular ligamental area being much longer than the posterior. Hinge- line straight and not very long. Interior of valve smooth, the adductor scar small, shallow', nearly circular in shape and situated a little below^ the base of the umbonal ai*ea, posteriorly to the median line. Right valve obliquely ovate, the umbonal por- tion convex, the remainder concave, a shallow' groove separating the two portions in the anterior half of the valve. Anterior margin fairly evenly convex: posterior margin slightly concave immediately below the hinge-line, thence convex and bulging outwards almost to an angle opposite the alate portion of the left valve, thence nearly straight to form a rounded arch with the anterior margin at the, ventral end of the valve. Umbo marginal, sharply pointed and set obliquely with its point directed tow*ards the posterior end of the hinge; the hinge area long, flattened and bent slightly forward in the middle, the ligamental area occupying the flattened portion. Exterior of valve smooth except for a few' faint narrow grow'th rings round the umbo. Interior surface smooth except for a few low^ growth ridges in the middle third, adductor scar large and ovate in shape, situated immediately posterior to the median line and occupying a large part of the ventral half of the concave portion w'hich corresponds to the convex umbonal portion of the exterior. The concave area extends ventrally for nearly half the hei^'ht cf the valve, being deepest immedi- ately below the hinge. Remarks. — The principal difference between this species and P. gingtnensis is in the propor- tions of height to length. In both valves of P. gmginensis the length is almost invariably greater, sometimes much greater, than the height, whereas in P. strathalbynensis the height of the left valve is much greater, that of the right valve very slightly greater, than the length. P. strathalbynensis shows a very close resemb- lance to Gryphaea vesiculosa Sow. from the Upper Greensand of Warminster. England (Woods 1912, pp. 374, 375; Plate LV, Pigs. 10-14. Plate LVI, Fig. 1). Indeed Woods’ description of that species might very well fit the Gingin shell. So far as can be judged from the single specimen, P. strathalbynensis is rather more oblique than G. vesiculosa, the ligamental area of its left valve being particularly so and a median line from the apex of the ligamental groove to the middle of the ventral margin would show a much greater curve than those of the specimens of G. vesiculosa figured by Woods, which show the ligamental area of the left valve to be directly below the umbo, whereas in the Gingin specimen it is situated posteriorly to the umbo, which is evidently grypheate, but owing to truncation of the holotype its exact shape is unknown. Genus EXOGYRA Say, 1820 Exogyra variabilis, sp. nov. Plate VI. Figs. 2-10 Specimens of Exogyra are fairly common in the Gingin area, especially at McIntyre Gully, where their vertical range is from the base of the chalk to possibly 40 feet above. They are also fairly common at Molecap. PLATE VI Pycnodonta strathalbynensis 1. — Interior of right valve of holotype (48969), X 2. Exogyra variabilis o_Fxterior- 2a —Interior; 2b.— Posterior profile of holotype (48970), a left valve from McIntyre GuUy. X 2; 3 * Exterior’ 3a. — Interior of left valve, variation A (48971), from McIntyre Gully, X 2; 4. — Exterior; 4a. Interior* 4b Posterior profile of left valve, variation B (48972), from Molecap, X 2; 5. Interior of iGft valve, vflrifltioA c’ (48973) from McIntyre Gully, X 2; 6.— Exterior; 6a.— Interior of left valve between variations B Ind 6 H8974) from McIntyre Gully. X 2; 7.-Exterior; 7a.-Interlor of left valve of variation D (48975) from Mclntvre Gully X 2* 8.— Exterior; 8a.— Interior of a typical right valve (48976), from Molecap, X 2; 9.— Exterior; 9 a— Interior of another right valve (48977), from Molecap, X 2; 10.— Exterior of a more nearly circular right valve (48978), locality not stated. X 2. 122 Individual left valves vary so greatly in shape that at first sight one is inclined to regard them as representing more than one species. A care- ful analysis of 34 fairly well-preserved speci- mens however, showed every gradation between the extreme forms and all would appear to be merely variations of the one species. The different variations do not appear to be charac- teristic of any particular horizon and two very different forms may come from approximately the same horizon. The principal variations are as follows: — The type. A neai’ly elliptical and nearly equilateral form, non-alate or with only a trace of posterior alation. Approximately half the specimens are of this type, from which the other forms probably originated, and a well- preserved specimen from McIntyre Gully has been chosen as the holotype of the species. A. A high trigonal non-alate form, widest near the ventral margin, with high beak which is twisted, rather than curved, posteriorly. B. A broad, very inequilateral form with broad and deep rounded posterior alation only. C. A bi'oadly ovate, more nearly equilateral form with nearly equal rounded posterior and anterior alation. D. A very inequilateral more oblique form with high, very elongate, and sharply pointed posterior alation. Two small valves without posterior alation show narrow anterior alation. Only six specimens show definite areas of attachment, the position depending on the form of the valve. One large imperfect specimen is almost wholly attached. The right valves of the species are operculi- form and are all very similar to each other whatever the form of the corresponding left valve. Dimensions are given in Table X. TABLE X Dimensions of Exogyra variahilis, sp. nov. UiL^lU \alvi‘< llt.In- 1 (yp‘‘ Mc- 1 nt\iv (hilly \’ari;int 4S97 1 ^]r- 1 iityrt* Uiiily Variaiil U 48072 Mole- eap Vaiijuil (' 1897;? Mr- l nt\rc (iully Variam 1) 4897r> Mo Intyn’ ( hilly .Ml I'niin MnleCfl]) mm. mm. mm. mm. mm. mm. Height 1-2 7 12-.') 9-4 9-8 lO-.s 9--I 1 1 •() 7-> l.englli h-7 7-5 7-9 9 • ■) i;?-7 9'1) 80 7-2 Description. — Shell small, thin, ovate, height greater than length, usually slightly oblique and inequilateral, very inequivalve, the left valve highly inflated, non-alate to alate. and larger than the right which is operculiform and usually concave. The beaks of both valves are curved posteriorly, that of the left valve being moder- ately large, that of the right valve very small. The type (Plate VI. Figs, 2, 2a, 2b).— Left valve highly inflated, fairly evenly convex in profile, ovate, with the widest portion opposite the middle of the valve, higher than long, the length equal to approximately three-quarters the height. nearly equilateral, non-alate; margins fairly evenly convex, the anterior and posterior merging into the ventral. Umbonal portion large and shaped like a hood above the remainder of the valve, with the hinge-line situated at about one-third the height of the valve below the top of the umbo. The hinge- line forming an even concave curve with the highest point anterior to the median line. The beak itself is long and broad, but not very prominent: it extends from the top of the valve almost to the hinge-line, and is curved pos- teriorly. Except for faint growth ridges the exterior surface is fairly smooth, except near the posterior margin. Interior surface smooth. Adductor scar of moderate size, elliptical and situated high up in the posterior half of the valve, being nearly hidden by the umbonal hood. Right valve operculiform, concave, obliquely ovate to nearly circular in outline: height greater than length, inequilateral; anterior margin with an even convex curve and merging into the ventral; posterior margin nearly straight to strcngly convex in the nearly circular specimens. Beak exceedingly small, curved posteriorly. Exterior surface decorated usually with about 7 to 9 fairly evenly spaced strongly imbricating growth-rings, each ending in an arch directed ventrally. A line joining the points of the arches would show a fairly marked curve, convex anteriorly. Interior of valve smooth ex- cept foi the fairly large, shallow ovate adductor scar situated about half-way between the median line and the posterior margin and with its centre on or slightly above a median transverse line. On the average the height of a right valve is about three-fifths that of the corresponding left valve. Variation A (Plate VI, Figs. 3. 3a). — Left valve moderately inflated, height much greater than length, sub-trigcnal in outline, approaching a scalene triangle with the most acute angle at the umbo, an interior margin being the longest; anterior margin of specimen imperfect but ap- parently nearly straight; posterior margin slightly concave and considerably shorter than the anterior; ventral margin fairly evenly con- vex but imperfect in the specimen figured. The valve widest at a little more than three-quarters the height of the valve below the umbo. The umbo twisted rather than curved posteriorly; top of umbo of the specimen flattened by a small concave area of attachment. Hinge-line fairly sharply aragular. Exterior surface some- what rugose, with narrowly-spaced growth- ridges on the umbonal half; interior surface smooth. The specimen is from the Ostrea phil- heyi zone of McIntyre Gully at about 22 feet above the base of the chalk. Variation B (Plate VI, Figs. 4. 4a. 4b). — With posterior alation only. Left valve strongly in- flated, height somewhat greater than length, slightly oblique, veiy inequilateral: markedly alate posteriorly, the alation both wide and deep, with the widest portion about opposite the middle of the valve. The posterior portion of the valve is concave between the umbo and the alation, but the remainder of the valve is strongly convex; the margins are all convexly curved; the hinge-line showing a fairly even 123 curve and merging into the anterior and poster- ior margins which, in turn, merge into the ven- tral. The umbonal portion of the valve extends in depth to nearly two-fifths of the height of the valve, the beak itself is large, rather thick and shows a marked posterior curve. Exterior surface of the body of the valve slightly rugose, with rather faint growth-lamellae, the alate portion strongly rugose: interior of valve smooth: adductor scar fairly large, ovate, hidden by the umbonal hood. The specimen figured is from Molecap. Variation C (Plate VI. Fig. 5'.— The bi-alate form. Left valve fairly strongly inflated, height only slightly greater than length, nearly equi- lateral in outline. Hooded umbonal portion wide and fairly deep; the beak large, curved poster- iorly: hinge-line of specimen somewhat irregular, thickest and with a convex curve at the top of the posterior wing; margins below the hinge-line fairly evenly rounded. Posterior alation usually larger than the anterior, but in one small speci- men, the two are approximately equal; the ala- tions are usually moderately wide and deep and merge into the ventral margin. The widest por- tion of the valve is at about three-fifths of the height of the valve below the top of the umbo. The position of the area of attachment of this form and of variation B is usually at the base of the umbo. The specimen figured is from McIntyre Gully, but from what height above the base of the chalk was not recorded. Variation D (Plate VI, Figs. 7, 7a).— The high- winged form. Left valve fairly strongly inflated, very inequilateral. Excluding the wing, the valve is obliquely ovate, the length about equal to three-quarters the height: with the wing, the length is much greater in well-developed specimens, but in some the wing is quite narrow: in the specimen figured, the length of the wing is equal to more than half the height of the valve. The wing is not sharply marked off from the body of the valve: the widest portion is in continuation with the hinge-line and about op- posite the base of the umbo from which it is separated by a fairly wide furrow; the doisal margin is nearly straight and ends in a sharp point from which the posterior margin descends, usually in a straight line, to the ventral maigin, anterior margin of the valve convexly curved, the convexity being greater in the ventral half. Hinge-line slightly curved. Umbo usually not so deep, proportionally, as in the other forms and the beak itself is relatively slightly smaller. Exterior of valve fairly smooth except for rather faint growth-rings, the alate portion being rather more rugose. Interior of valve smooth: the ad- ductor scar situated very high up and hidden by the umbo. The area of attachment is in the furrow between the base of the umbo and the wing. ... ^ 1 . 1 . ^ This form apparently merges into the broad mid-wing form of variation B. The specimen ficrured is from the Ostrea vhilheyi zone of jj"^tyre Gully at about 22 feet above the base of chalk. Remarks.— Exogyra variabilis shows some re- semblance to the European E. cu^iiculuto. Sow. ‘(Woods 1912. pp. 375-379, plate LVI, Figs. 2-16), the more alate specimens figured by Woods in particular resembling less markedly developed specimens of Variation D. Woods’ Fig. 6 is somewhat like the non-alate form of E. vari- abilis but is narrower and more sharply ovate. The right valves of the two species are very similar. Apart from the direction of curvature of the beak, E. imriabilis resembles fairly closely shells of the sub-genus Gryphostrea Conrad, in which, however, the beak is always curved anteriorly: the non-alate form much resembling a speci- men of G. inscripta d’Arch. figured by Cossman (1922. Plate XIII, Pigs. 8, 9. 21) and specimens of Variation C are not unlike G. borissaci Doncieux which Cossman (1922. p. 211, Plate XIII. Figs. 28, 29) regarded as an alate variety of G. inscripta. Authors differ as to the relationships of the subgenus Gryphostrea which appears to have taken the place of Exogyra in Eocene times. It first appeared in Upper Cretaceous times as G. vomer Morton, considered to be identical with the Eocene G. eversa Deshayes. chosen as the type species by Conrad. Stoliczka (1871) placed Gryphostrea under Gryphea, but Mayer (1875) and some later authors regarded G. eversa as an Exogyra (see Cossman 1922, p. 210). Woods who recognised only Ostrea and Exogyra as genera, considered (1912, pp. 378. 379) that Gryphostrea canaliculata was probably related to Pycnodonta vesicularis. Gardner (1916, p. 579) stated; "Gyphostrea suggests Exogyra in the gyrate umbones of the left valve. The beak of the right valve of the former, however, is crthogyrate or at the most slightly inclined, and this, together with the uiflation of the beak of the left valve, allies it more closely with Gryphaea than with Exogyra.” Gardner’s lemarks could be applied equally well to hxogyra of the E. variabilis group. Cossman (1922) placed Gryphostrea together whth Pycnodonta under the genus Liostrea Douville. of which Ostrea sublainellosa is the type species. Shimer and Schrock (1944) gave generic rank to Gryphaea, but placed Gryphostrea under Ostrea. The only important difference between Exogyra and Gryphostrea is in the direction of curvature of the beak, which in Exogyra is curved posteriorly in Gryphostrea anteriorly. Other differences are of no greater than specific value, and the difference in the direction of curvature may be only one of a few degrees; as, for example, between the holotype of Exogyra variabilis (Plate VI. Pig. 2a> and the specimen of Gryphostrea inscripta figured by Cossman (1922, Plate XIII. Fig. 8), and it seems most probable that Gryphostrea was originally derived from species of Exogyra and should be regarded as a subgenus of that genus. Acknowledgments I am greatly indebted to the Research Grants Committee of the University of Western Aus- tialia for the funds necessary to carry out my researches. I must express my deep gratitude to Professor R. T. Prider for his encouragement and for permitting me to make full use of the facilities afforded by the Geology Department of the University. I am indebted to Dr. B. F. Glenister for generous help and advice during the early stages of the work; to Dr. J. E. Glover for much helpful advice; to Miss Natalie Sugden and later to Mr. S. Blotnicki for help in obtain- ing the necessary literature; to Mr. F. L. Billing and Mr. K. Bauer for the many excellent photographs which accompany this paper; and particularly to Mr. F. L. Billing for much generous help in many ways; I am indebted to Mrs. Jean White for the typescript. References Carroll. Dorothy (1939).— A note on the mineralogy of the Gingin Chalk. Proc. Geol. Ass. Lond. L Pt. 2: 227-234. Carroll, Dorothy { 1941) .—Heavy residues from some Upper Cretaceous sediments at Gingin. J. Sediment. Petrol. 11: 85-91. Clarke, E. de C.. and Teichert C. (1948).— Cretaceous stratigraphy of lower Murchison River area. Western Australia. J. Roy. Soc. W. Aust. 32: 19-47. Cossman, Maurice (1922).— Synopsis illustre des Mollus- ques de Pliocene et de roligocene en Aquitaine. Mem. Soc. Geol. Fr. 55, Tome 24. Ease. 1-2. Cossman. Maurice, and Pissaro, G., in Vredenburg, E. (1927)— The Mollusca of the Ranikot series. Palaeont. Indica 10 Mem. No. 2. Dali, W. H. (1898).— Contributions to the Tertiary fauna of Florida. Trans. Wagner Inst. Sci. Philad. 3 (4). Davies, A. Morley (1935).— “Tertiary Faunas.” Vol. 1 ( Murby : London . ) Etheridge, R. Junr. (1910).— Oolitic fossils of the Greenough River District. Western Aus- tralia. Bull. Geol. Surv. W. Aust. 36. Etheridge, R. Junr. (1913).— The Cretaceous fossils of the Gingin Chalk. Bull. Geol. Surv. W. Aust. 55. Eudes-Deslongchamps, J. A. (1858). — Essai sur les Plicatules Possiles des Terrains du Cal- vados. Mem. Soc. Linn. Normandie 11 (1). Feldtmann, F. R. (1934).— The glauconite deposits at Gingin. South-West Division. Annu. Progr. Rep. Geol. Surv. W. Aust. for 1933: 6 - 8 . Feldtmann. P. R. (1951).— Pectens of the Gingin Chalk. J. Roy. Soc. W. Aust. 35: 9-39. Gardner, Julia A. ( 1916) .—Mollusca in Systematic palaeontology, Upper Cretaceous. Mary- land Geol. Surv.: 371-733. Glauert. Ludwig (1910).— The geological age and organic remains of the Gingin Chalk. Bull. Geol. S^lrv. W. Aust. 36. Glauert, Ludwig ( 1925) .—Further notes on the Gingin Chalk. J. Roy. Soc. W. Aust. 12: 5-12. Glauert. Ludwig (1926). — A list of Western Australian fossils. Bull. Geol. Surv. W. Aust. 88: 36-72. MeWhae, J. R. H.. Playford, P. E.. Lindner. A. W., Glenister, B. P., and Balme, B. E. (1958). — The stratigraphy of Western Australia. J. Geol. Soc. Aust. 4 (2). 1-161. Meek, F. B. (1876). — A report on the invertebrate Cretaceous and Tertiary fossils of the Upper Missouri Country. Rep. U.S. Geol. Surv. (of the Territories) 9. Shimer. H. W., and Schrock. R. R. (1944).— “Index Fossils of North America”. (Wiley: New York.) Simpson, E. S. (1920). — On gearksutite at Gingin, West- ern Australia. Miner. Mag. 19.- 23-39. Simpson, E. S. (1937). — Contributions to the mineralogy of Western Australia. J. Roy. Soc. W. Aust. 23: 27-29. Spath. L. F. (1926). — Note on two ammonites from the Gingin Chalk. J. Roy. Soc. W. Aust. 12: 53-55. Stenzel. H. B. (1947). — Nomenclatural synopsis of supraspecific groups of the family Ostreidae. J. Paleont. 21: 165-185. Stephenson, L. W. (1934). — The genus Diploschiza from the Upper Cretaceous of Alabama and Texas. J. Paleont. 8: 273-280. Stephenson. L. W. (1952). — Larger invertebrate fossils of the Woodbine Formation (Cenomanian) of Texas. Prof. Pap. U.S. Geol. Surv. 242. Stoliczka, Ferdinand (1871). — Cretaceous Fauna of Southern India. Palaeont. Indica, Ser. 6. Vol. 3. Teichert, Curt (1947). — Stratigraphy of Western Aus- tralia. J. Roy Soc. N.S.W. 80: 81-142. Walkom. A. B. (1944). — Fossil plants from Gingin, West- ern Australia. J. Roy. Soc. W. Aust. 28: 201-204. Withers, T. H. (1924). — The occurrence of the Crinoid Uintacrinus in Australia. J. Roy. Soc. W. Aust. 11: 15-18. Withers, T. H. (1926). — The Crinoid Marsupites and a New Cirripede from the Upper Cretaceous of Western Australia. J. Roy. Soc. W. Aust. 12: 97-104. Woods, Henry (1899). — A monograph of the Cretaceous Lamellibranchia of England. Vol. 1, pt. 1, Palaeontogr. Soc. (Monogr.) 53: 1-72. Woods, Henry (1901). — A monograph of the Cretaceous Lamellibranchia of England. Vol. 1, pt. 3, Palaeontogr. Soc. (Monogr.) 55: 113-144. Woods, Henry (1905). — A monograph of the Cretaceous Lamellibranchia of England. Vol. 2, pt. 2, Palaeontogr. Soc. (Monogr.) 59: 56-96. Woods. Henry (1906).— A monograph of the Cretaceous Lamellibranchia of England. Vol. 2, pt. 3, Palaeontogr. Soc. (Monogr.) 60: 97-132. Woods, Henry (1912). — A monograph of the Cretaceous Lamellibranchia of England. Vol. 2, pt. 9, Palaeontogr. Soc. (Monogr.) 66: 341-473. 125 16. — The Effect of the Suppression by the International Commission on Zoological Nomenclature of Zimmermann 1777 upon the Stability of the Generic Name Macropus Shaw 1790 By W. D. L, Ride* Manuscript received — June, 1963 As the result of a ruling of the International Commission on Zoological Nomenclature. Yerhoa gigantea Zimmermann (the type species of Macropus Shaw 1790} has no status in nomen- clature. Macropus giganteus Shaw 1790 must thus be regarded as new and as the type species of Macropus but it Is a junior secondary homonym of Jaculus giganteus Erxleben 1777 which is in turn an objective synonym of Mus conguru Statius Muller. It is proposed that the situation be stabilised by the selection of the type of Mus canguru as the lectotype of Macropus giganteus. Introduction The type species (by monotypy* of Macrovus is Macropus giganteus Shaw 1790, which is generally regarded as a junior objective synonym of Yerhoa gigantea Zimmermann 1777. This latter name is entirely based upon material collected by Captain Cook’s party at the Endeavour River, Queensland. To date no author working with marsupials has doubted the validity of Zimmermann’s name Yerhoa gigantea as applied to Captain Cook’s Kangaroo and all (including myself and co-workers) have accepted it (see Thomas 1888, Cabrera 1919, Iredale & Troughton 1934. 1937, Raven 1939, Tate 1948, Morrison-Scott & Sawyer 1950, Calaby, Mack & Ride 1962) although there is some disagreement as to the animal species represented by it. However, I now find that the International Commission on Zoological Nomenclature ruled in 1950 that Zimmermann 1777 (Specimen Zoologiae Geographicae) is not available for zoological nomenclature (Bull. Zool. Nomencl. 4: 547 1 ; thus Yerhoa gigantea Zimmermann 1777 carries no more status than a vernacular name. It is now necessary to ensure that: (a) the generic name Macropus (of which Yerhoa gigantea Zimmermann 1777 was believed to be the type species by monotypy) is stable in its present usage, and (b) the objective synonymy of Mus canguru Statius Mlillert and the various usages of the specific name giganteus as applied to Macropouidae are retained. All of the authors since 1777 (Erxleben, Vol. 1, p. 409) who have used canguru have regarded these as synonyms. * Western Australian Museum. Perth, Western Aust- ralia. t- This author’s name is often given as Muller, or as P.L.S. Muller. Holthuis & Jrwige (1958), in a foot- note. show that the family name is Statius Muller— a name today well known in Holland. To Stabilize Macropus As a result of the action of the International Commission, the description of Macropus Shaw 1790 now contains no reference to any valid species name other than Macropus giganteus Shaw 1790. This name can now be regarded as a new name although it is a junior secondary homonym of Jaculus giganteus Erxleben 1777. No type specimen is known to exist for Macrcpus giganteus Shaw and only one speci- men is known to be in existence today which had been seen by Shaw. This is a spirit-pre- served juvenile in the collection of the British Museum (Nat. Hist.), No. 145b of a manuscript catalogue by Gray; it weighs 2 lb. 4 oz. and is noted to be “the one described by Dr. Shaw”. It is not known whether this note refers to Shaw's 1790 description, or to his later work of 1800, so the specimen cannot be assigned to the type series of M. giganteus and is there- fore unsuitable for selection as a lectotype. Since the Grey Kangaroos possibly merit treat- ment at a subspecific level, it is desirable that the types of any names of Grey Kangaroos should have adequate locality data and this specimen has none. I therefore reject it for the purposes of neotype designation as well. Shaw’s 1790 description of M. giganteus is based upon material from various sources among which are the three specimens collected by Cook’s party at the Endeavour River. Thus any of these (and in particular the holotype of Mus canguru Statius Muller ) is available for selec- tion as the lectotype of M. giganteus Shaw. Subsequent to Shaw’s 1790 description, in 1800 Shaw himself replaced the name giganteus with the replacement-name major formally stat- ing at the same time that the new name major was synonymous with giganteus Shaw and also with Didelphis gigantea of Gmelin and Schreber. This last name is simply the employment of Jaculus giganteus Erxleben 1777 by these authors in combination with the different generic name Didelphis. Thus, it is clear that Shaw regarded both his major and giganteus as being equal to the earlier Jaculus giganteus Erxleben 1777 which is itself no more than a replacemen)-, name for the earlier Mus canguru Statius Muller 1776. Macropus major, Jaculus giganteus and Mus canguru thus all possess the same type specimen (International Code Article 72(d)). Since Shaw (1800, p. 505) formally equated his giganteus with all of these, I hereby propose the holotype of Mus canguru Statius Muller as its lectotype. 126 The holotype of Mus canguru now no longer exists and Calaby, Mack & Ride (1962) have proposed an undoubted specimen of a Grey Kangaroo as its neotype- — Queensland Museum specimen No. J 10749 male, skin and skull, col- lected at Kings Plains, 20 miles south of the Endeavour River, November 24. 1960. by D. P. Vernon and S. Breeden and as figured in Calaby. Mack & Ride 1962. Plates 5, 6, 7, 8, An applica- tion for recognition of this neotype is at present before the International Commission on Zoo- logical Nomenclature. By adopting this procedure the generic name Macropus is unequivocally fixed to the Grey Kangaroos. The alternative to the procedure which I out- line here would be to have proposed a neotype from Botany Bay for Macropus giganteus Shaw because Shaw’s description is based in part upon material from that locality described by Pen- nant. But, since M. giganteus Shaw is a junior homonym of Jaculus giganteus Erxleben (and of Schreber which was in the early 19th century in use as its senior homonym; see Waterhouse 1846, p. 62), it would then require as a sub- stitute the first available replacement name for the Grey Kangaroo. This is possibly Dipus tridactylus Perry 1811 or Kangurus lahiatus Desmarest 1817. iMacropus major Shaw is an objective synonym of Mus canguru Statius Muller and is not available as a replacement name for M, giganteus Shaw). I believe that this would cause greater upset in the literature than the course which I follow. The Synonymy of the Names cangaru and giganteus Calaby, Mack & Ride have proposed that the Queensland Museum specimen mentioned above should be recognized as the neotype of M. can- guru Statius Muller and the lectotype of M. giganteus Zimmermann. This reference to Zimmermann is now no longer necessary. Erxleben (1777, p. 409 > proposed Jaculus giganteus as a replacement name for Mus can- guru Statius Muller and the name giganteus is thereby an objective synonym of canguru with- out the need for selection of a lectotype since it automatically possesses the same type speci- men as the name which it is proposed to replace 'Code Art 72).* The Controversy over Captain Cook’s Kangaroo While this controversy is only of indirect concern here, some comment should be made on Iredale & Troughton's (1962, p. 183) state- ment that our arguments (Calaby, Mack & Ride 1962) are based upon a specimen of doubtful authenticity. Iredale & Troughton do not qualify this remark but, because it is made from their venerable position in Australian mam- malogy, it casts very real doubt upon our published conclusions. Briefly, the factual basis of our argument is that Cook’s party was known to have collected • The only other usage of gigantea for Macropodldae in 1777 Is Schreber. Saugethlere, 3, p. 552. Pages 455 onwards were published after Erxleben (Sher- boru 1891, p. 588. footnote). three animals and it is possible to find pub- lished records of three specimens which, it is reasonable to assume, were brought home to Britain by the party. One is a robustus (identified from a contemporary drawing by Morrison-Scott &; Sawyer, whose specific identi- fication was confirmed by ourselves), the other two are Grey Kangaroos (identified by J. E. Gray 1843, by Owen 1853, and by Flower 1884). One of these three specimens must be the holo- type of canguru. From these three we elimin- ated the juvenile spirit-specimen identified by Gray as a Grey Kangaroo because the holotype was known to have been eaten. The robustus was also eliminated because our investigations showed that it would clearly have been too large to have agreed with the known weight of the holotype: we were thus left with only a specimen in the Hunterian collection which from Richard Owen’s description we then demonstrated is at a stage of dentition (based on both dental pro- gression and eruption ) consistent with the weight specified by Statius Muller for the holo- type in his original description. The specimen listed in the Hunterian cata- logues has unambiguous data. i.e.. it was presented to Hunter by Banks and came from the Endeavour River. It was identified as a Grey Kangaroo by Owen and later by Flower. Its exact dental-age was specified by both authors who were eminent in this field. I suspect that the reason for Iredale & Troughton’s statement is that the Royal College of Surgeons’ lantern slide of this skull (pub- lished as a plate in Morrison-Scott & Sawyer 1950) probably has a mis-matched mandible. (The skull has been destroyed and this is the only known illustration of it). In this photo- graph, the .skull is clearly numbered on the maxilla with its catalogue number but. whereas the catalogue states that it has a right mandi- bular ramus I? only], the illustration is of the left side of a skull with a left mandibular ramus. This ramus is unnumbered. The photograph cannot have been reversed because the catalogue number is the correct way round. To this extent the illustration of the skull — as a picture of a "complete skull" (i.e. cranium and mandible) could be said to be unreliable. I drew Trough- ton’s attention to this some years ago. It is important to realize that the arguments in Calaby, Mack & Ride are not based, in any way. upon this illustration but upon’ Owen’s published description of 100 years earlier. They are also based upon Owen’s and Flower’s identi- fication of the specimen with the Grey Kangaroo and their statements that it was a Banksian specimen from the Endeavour River. Unfortu- nately. since our argument as to the choice of the holotype from these specimens depended, in part, upon the probability that certain weights and dental ages can be correlated, we believed that action by the International Commission on Zoological Nomenclature was warranted in order to establish nomenclatural stability through placing the nomenclatural problem beyond the upsetting effect of a biologically sterile argu- ment which has gone on since 1927. We thus asked the Commission to accept an undoubted Grey Kangaroo as the neotype of canguru. 127 In fairness to Morrison-Scott & Sawyer, whose conclusions are strongly criticized by Iredale &: Troughton. it must be pointed out that Iredale & Troughton neglect to bring out certain facts which argue against their own conclusion that Captain Cook’s Kangaroo is a Whiptail Wallaby. These are, firstly that the outcome of their controversy with Morrison-Scott & Sawyer over the interpretation of the ambiguous descriptions by Solan der, in Latin, of the incisors is still an argument against the species being the Whip- tail. They conclude that the third upper incisor of Captain Ccok’s Kangaroo has a smaller anterior lobe. In fact, the anterior lobe of the third incisor of the Whiptail is not smaller than the posterior lobe, being slightly larger (to markedly greater) than it. Of a series of 12 skulls in the British Museum measured by me the index (Anterior lobe I'V/Total length of I'b x 100 has a mean of 56.7. with a range from 64.4 to 50.7. Secondly, Iredale & Troughton (1937, p. 68) said that the general coloration of the Cook^own Whiptails agrees with the nondescript colour accorded to Cook’s species in the early accounts- This is not so. The Whiptail Wallaby of the Endeavour River is prominently and vividly marked and quite distinct from the Grey Wallaroo and the Grey Kangaroo of the same area. While it is conceivable that a person unfamiliar with kangaroos and wallabies could group Grey Wallaroos and Grey Kangaroos together in the same sample, he could not avoid remarking on the brilliant facial and hip pat- terns of the Whiptail. The large series collected by the Queensland Museum in the Cooktown district in order to clarify this problem, illus- trates this well. Finally, I am unable to understand the state- ment (Iredale & Troughton 1962, p. 177 > that “the name giganteus as applied to the Great Grey Kangaroo is superseded by major as type of Shaw’s genus Macropus''. The facts are otherwise and are summarized thus: (a) Until 1950 (Bull. Zool. Nomencl. 4: 547), the type species of Macropus by monotypy was Yerboa gigantea Zimmermann 1777. 'See Shaw 1790, Vol. 1. text to Plate 33). (b) Today the type species is Macropus gigan- teus Shaw also by monotypy. (c) Macrcpus major was first described in synonymy with Macropus giganteus Shaw, Didelphis gigantea Gmelin, and Schreber. Accordingly, it is either a replacement name for these, i.e. a junior objective synonym of the most senior of them (Code Art. 72(d)), or, being originally described in synonymy with other more senior names, has no status in nomenclature (Code Art. IDd) ). It is certainly not the type species of Macropus, and unless the Commission uses its plenary powers in accordance with the recommendations of Calaby. Mack &: Ride, its use in any form at all is invalid. References The only works cited here are those not listed in Calaby, Mack & Ride 1962. Cabrera, A. (1919). — “Genera Mammalium. Monotre- mata, Marsupialia". (Museo Nacional de ciencias naturales: Madrid.) Calaby, J. H., G. Mack, & W. D. L. Ride. (1962). — The application of the generic name Macropus Shaw 1790 and of other names commonly referred to the Grey Kangaroo. Mem. Qd Mus. 14: 25-31. Erxleben, J. C. P. (1777). — “Systema Regni Animalis . . 1: Mammalia (Weygand: Leipzig.) Holthuis, L. B.. & G. C. A. Junge (1958). — The specific names in Tunstall’s Ornithologia Britan- nica 1771. Ardea 46: 167-170 [footnote p. 1 / 0 ;. Iredale. T., & E. le G. Troughton. (1962). — The actual identity of Captain Cook’s Kangaroo. Proc. Linn. Soc. N.S.W. 87: 177-184. Schreber. J. C. D. von (1777). — “Die Saugethiere . . 3: (Leipzig.) Shaw, G. (1790). — Naturalist’s Miscellany. 1. (Nodder & Co.: London.) Sherborn, C. D. (1891). — On the dates of the parts, plates and text of Schreber’s “Sauge- thiere”. Proc. Zool. Soc. Lond. 587-592. Tate, G. H. H. (1948). — Studies on the anatomy and phylogeny of the Macropodidae (Marsupi- alia). Bull. Amer. Mus. Nat. Hist. 91: 237- 351. Thomas, O. (1888). — “Catalogue of the Marsupialia and Monotremata in the Collection of the British Museum.” (British Museum (Nat. Hist.): London.) Waterhouse, G. R. (1846). — “Natural history of the Mammalia. 1. Marsupiata”. (Bailliere: London.) 128 17. — Eight New Plants From Western Australia By C. A. Gardner* and A. S. George* Manuscript received — 16th July, 1963 Descriptions and illustrations are given of one new genus {Neogoodenia) and eight new species of plants from Western Australia, viz., Grevillea variifolia, Grevillea prostrata, Stackhousia um- bellata, Verticordia staminosa, Calytrix superba, Eremaea rosea, Leschenaultia subcymosa, and Neogoodenia minutifiora. Introduction The following eight species have been collected during the past three years. We are indebted to Mr. C. Chapman for the specimens of Caly- trix superha, and to Mr. W. H. Butler for the specimens of Verticordia staminosa. The English descriptions do not parallel the Latin, but include additional information on certain characters. The holotype and some syn- type specimens will be housed in the Western Australian Herbarium. Syntypes of all except Neogoodenia minutifiora will be distributed to Kew Herbarium and the National Herbarium of Victoria. Only a single specimen of Neogoo- denia has been collected to date. It is, however, an excellent one, bearing full flowering and fruiting material. PROTEACEAE Grevillea variifolia, sp. nov. Fig. 1 Frutex diffusus. 30-70cm altus, ramis pubes- centibus. Folia oblanceolata vel anguste- cuneata, 15-40 mm longa, 3.15 mm lata, 3-7- acute-lobata. sericea vel supra demum glabra, marginibus recurvis. Flores rubrae. racemis 20-65 mm longis densifloris terminalibus; pedi- cellis 2. 5-3.5 mm longis sericeo-pubescentibus divarticatis. Perianthium 7-8 mm longum, sub limbo revolutum, ad basin dilatatum, extus parce pubescens. intus supra medium dense hir- sutum. Ovarium glabrum stipite 3 mm longo; stylus 25-27 mm longus, glaber, exsertus, disco stigmatico laterale, glandula hypogyna laterale, toro fere recto. Folliculus obovoideus, 10-11 mm longus, 5 mm latus. Semina angusta, 6-8 mm longa, anguste alata, marginibus revolutis. A spreading shrub 30-70 cm tall. Branches silky-pubescent, becoming glabrous with age. Leaves oblanceolate or narrow-cuneate, 15-40 mm long. 3-15 mm wide, shortly petiolate, with 3-7 pungent lobes, occasionaly entire, both sur- faces covered wuth fine appressed hairs, the up- per surface becoming glabrous, margins recurved. Flowers red in dense terminal racemes of 20-65 mm long; rachis silky-pubescent, the lower part without flowers; pedicels 2.5-3.S mm long, divari- cate or somewhat reflexed. Perianth 7-8 mm long, revolute under the limb, broadened in lower half, with scattered appressed hairs out- ♦ Western Australian Herbarium. Perth. side, short spreading hairs along margins of segments and inner surface densely hirsute above the middle. Ovary glabrous on a stipes of 3 mm; style 25-27 mm long, slender, glabrous, exserted, stigmatic disc lateral, hypogynous gland lateral, torus slightly oblique or straight. Follicle obovoid, 10-11 mm long, 5 mm wide, smooth, the style persistent. Seeds narrow, orange, 6-8 mm long, margins narrowly winged and revolute. Holotype and Syntypes; Cape Range, near No. 3 Well, in red sand over lime- stone, A. S. George 2477. June 2, 1961. Grevidea variifolia falls in the Section Leio- gyne. It is closest to G. thelemanniana, Hueg., and G. stenomera, F. Muell.; but the leaves are never pinnatisect or divided into linear lobes and are grey-green in colour. The species is noteworthy as being the only repre- sentative of the Section known to occur outside the South-West Province in Western Australia. The following collections have also been examined: Cape Range (Charles Knife Rd.), A. S. George 1340, August 30. 1960. — A shrub of 1.3 m (no fls.); leaves up to 50 mm long and 20 mm wide. Vlaming Head (Lighthouse Hill), A. S. George 1369, August 31, I960.— A shrub to 70 cm, leaves silver-green, fls. red; style 20 mm long. 1 mile S. of Vlaming Head lighthouse, on W. side of Cape Range, A. S. George 2577, June 3, 1961. — A spreading shrub of 30-70 cm. fls. red; perianth 9-10 mm long, style 28-30 mm. 79 miles S. of Learmonth, in red sand with spinifex and low shrubs, A. S. George 2402, June 2. 1961. — A shrub of 70 cm, fls. red; leaves with 3, sometimes 5 divaricate lobes, perianth 6-7 mm long, style 19-23 mm. Learmonth rd., 22 miles N. of Warroora turn- off, in red sand, A. S. George 3286, February 22, 1962. — Lea-f lobes very variable, sometimes divaricate; leaves on young branches lanceolate, entire. Grevillea prostrata, sp. nov. Fig. 3, K-P Frutex parvus prostratus. Rami hirsuti, demum glabri. Folia 2-4.5 cm longa, petiolata. pectinata lobis linearibus oppositis marginibus revolutis. hirsuta vel supra demum glabra. Flores in racemis densifloris in pedunculos terminales et laterales. Pedunculi hirsuti- pubescentes et parce glandulosi. Bracteae spathulatae, 1.5 mm longae, deciduae. Pedicelli 7-8 mm longi, glabri. Perianthium angustatum. 3-4 mm Icngum, sub limbo revolutum, omnino glabrum. Stylus curvus, glaber; ovario minute glanduloso breviter stipitato; toro recto, glan- dula hypogyna laterale; disco stigmatico laterale. Folliculus juvsnis obovoideus, fere sessilis. 129 Fig. 1 Grevillea variifolia, sp. nov. A, B, C, D. — Leaves, natural size; E.— Opening flower. 5 X natural size; F.- Open flower, 3 X natural size; Q. — Two perianth segments showing inner surface, 4 X natural size; H. — Ovary and torus, 5 X natural size; I. — Fruit, 2 X natural size; J — Seed, 4 X natural size (A-J from A. S. George 2477); K. — Leaves from A. S. George 3286, natural size; L. — Leaves from A. S. George 2577, natural size; M — Leaves from A. S. George 2402, natural size. 130 Pig. 2 Stackhousia um'bellata, sp. nov. A. — Umbel of flowers, 4 X natural size: B. — Apex of corolla, enlarged: C. — Attachment of stamlnal filaments and free basal lobes of corolla to calyx tube, enlarged; D. — Fruit, 12 X natural size; E. — Single coccus, 8 X natural size; F. — Style and ovary, 8 X natural size: G. — Anther, much enlarged; H. — Diagram of portion of a branch to show habit. 0.5 X natural size. (A-C. F-H from A. S. George 2585; D. E from A. S. George 1380.) 131 A prostrate shrub. Stems hirsute, becoming glabrous, not much branched. Leaves 2-4.5 cm long, petiolate, pectinate, lobes linear, opposite, with revolute margins, hirsute, becoming glab- rous above. Flowers in short, dense racemes on terminal and lateral branchlets. Rachis hirsute- pubescent, with glandular hairs also. Pedicels 7-8 mm long, glabrous, subtended by deciduous, spathulate bracts 1.5 mm long. Perianth 3-4 mm long, narrow, revolute under the limb, glabrous inside and out, the segments separating when open. Style curved, glabrous: ovary minutely glandular, shortly stipitate: torus straight, hypgynous gland lateral: stigmatic disc lateral. Young fruit obovoid, almost sessile, sparsely glandular. Holotype and Syntypes: On sandplain out- side Pallarup Rocks, S.E. of Lake King, A. S. George 1652, October 14, 1960. “FIs. cream on reddish uedicels". The species falls in the Section Lissostylis, Series Occidentales. Although the ovary is minutely glandular, it is never densely hirsute as in the Sections Hebegyne and Eriostylis. The affinities are w'ith G. crithmifolia, R.Br.. the dif- ferences being the completely prostrate, less crowded habit, the pinnate leaves, smaller floral bracts, the perianth smaller, glabrous inside and out, the shorter ovary stipes, and the smooth fruit. STACKHOUSIACEAE Stackhousia umbellata, sp. nov. Fig. 2 Herba perennis, diffusa, glabra. Rami sulcati. Folia sQuamata. Flores umbellatae terminales, bracteatae, pedicellis 1-1.5 mm longis. Calyx 4.5 mm longus, 10-nervosus, lobis liberis rhomboideo- oblanceolatis, minute denticulatis, quam tutaum longioribus. Corolla quam calycem longiora bis, lutea, lobis late lanceolatis acutls patentibus, Stamina inclusa, filamentibus inaequalibus, an- theris angustis. Ovarium 3-lobatum. Stylus brevis 3-lobatus. Cocci 3. ovoidei. 2.5 mm longi. reticulati, minute papillosi. A diffuse, glabrous perennial herb. Stems slender, sulcate. Leaves reduced to small, scat- tered scales. Flowers in terminal umbels of up to 14 flowers, each subtended by 2 very small bracts: pedicels 1-1.5 mm long. Calyx 4.5 mm long, 10-nerved in lower half, lobes free, rhom- boid-lanceolate, minutely denticulate, acute, longer than tube. Corolla twice as long as calyx, yellow, scented, lobes broadly lanceolate, acute, spreading. Stamens included, filaments un- equal, anthers narrow opening in longitudinal slits. Ovary 3-lobed. Style short, 3-lobed. Cocci 3 'often only two developing in fruit), ovoid, 2.5 mm long, reticulate, minutely papillose. Holotype and Syntypes: Cape Range, near No. 2 Well, in red sand over limestone. A. S. George 2585, June 4. 1961. Stackhousia umbellata differs from all other species in its umbellate inflorescence, and from all except S. scoparia, Benth. and S. dielsii. Pampanini, in its diffuse, leafless habit. In floral structure it is closest to S. viviinea, Sm.. but the flowers are much larger, the corolla lobes broader, and the cocci less rugose. The following collections have also been examined: Cape Range 'Charles Knife Rd.) in and around spinifex clumps, A. S. George 1336 August 30. 1960: Vlarning Head (Lighthouse Hill) in red sand over limestone, A. S. George 1380. August 31, 1960. MYRTACEAE Verticordia staminosa, sp. nov. Fig. 3, A-J Frutex ramulosus diffusus. Rami setosi, demum glabri. Folia ad apices ramorum con- ferta, 7-14 mm longa, linearia-teretia, in basi- bus pulvinifoi’inis setosis subpersistentibus in- sertia. Flores lutei, in pedicellos 4-6 mm longos glandulosos axillares. Bracteoli magni, 5-6 mm longi, ovati, scariosi, subpersistentes. Calycis tubus turbinatus. 1.5-2 mm longus, 2-2.5 mm latus, 10-costatus, glaber. Lobi calycis orbicu- lares, longe 5-7-pectinato-lobatis, 5 mm longi. Petala ovata, longe 5-7-subulato-lobatis. 5 mm ionga. Stamina longe exserta, 9-10 mm longa, in tubum longum connata, filamentibus in dimi- dio superiore planis, liberis; staminodia subu- lata in exteriore tubo inserta: antherae basifixae. 2-porosae, glandulis dorsalis prominentibus. Stylus staminiis aequalis. stigma parva, pulvini- forma. Ovarium 2-ovulatum. A spreading, much-branched shrub. Branches setose, at length glabrous. Leaves crowded to- wards the ends of the branches, linear-terete. 7-14 mm long, on short, thick, setose bases which remain on the branch for some time after the leaves have fallen but at length are decidu- ous. Flowers yellow, on slender glandular pedi- cels in the upper axils. Bracteoles large, 5-6 mm long, ovate, scarious, red-brown, persistent for some time but deciduous with the flowers. Calyx-tube turbinate, 1.5-2 mm long, 2-2.5 mm wide, 10-ribbed, glabrous; lobes orbicular, deeply divided into 5-7 pectinate-ciliate lobes, the whole 5 mm long. Petals ovate, divided into 5-7 subulate lobes, 5 mm long. Stamens much exceeding the petals, united for about h of their length in a tube, the free portion of the filaments flat: staminodes subulate, inserted on the out- side of the tube; anthers basifixed, 2-porose, the dorsal connective gland prominent. Style as long as the stamens (9-10 mm), stigma small, cushion-shaped. Ovary with 2 ovules. Fig. 3 — opposite Vertico7',dia stammosa, sp. nov. A.— Bracteoles and pedicel. 3i X natural size; B.— Calyx tube, 6 X natural size; C.- -Ovules, enlarged; D. — Calyx lobe, 6 X natural size; E.— Petal, 6 X natural size; F.— Staminal tube, showing staminodes, 6 X natural size; G.— Anthers, much enlarged; H.— Leaf, 3 X natural size, and leaf base, enlarged; I.— Portion of stem, showing setae, enlarged; J— Style end. much enlarged. (All from W. H. Butler Wongan Hills 12.vi.l961.) Grevillea prostrata, sp. nov. K.— Leaf, natural size; L— Young fruit. 2 X natural size; M. Bud. 4 X natural size; N— Style, ovary, and torus, 8 X natural size; O.— Bract, 14 X natural size; P.— Flower with pedicel, 10 X natural size. (All from A. S. George 1652.) 132 A. S Gforg^ Fig. 3 133 Holotype and Syntypes: Wongan Hills. W. H. Butler, June 12. 1961. Verticordia staminosa falls in Bentham's Sec- tion Ic of Euverticordia, containing V. grandi- fiora. Endl.. V. acerosa, Lindl., etc., but it dif- fers in the prominently exserted stamens with long tube and external staminodes, the setae on the branches, the crowded, linear leaves, and the long, glandular pedicels. CaSytrix superba, sp. nov. Fig. 4, A-K Frutex ramis erectis. Folia erecta vel appressa, alterna, oblongo-linearia. crassa, concava. 4-8 mm longa, breviter petiolata. Flores ramos terminantes. Bracteoli 10-12 mm longi, fere ad basin liberi. marginibus scariosis, apicibus acutis recurvis. Tubus calycis gracilis, 11-14 mm longus, 10-sulcatus, supra ovario solidus, lobis ad basin orbiculatis in setas minuter scataridas productis, 11-15 mm longis. Petala rosea, quam lobis calycis Icngiora, late elliptica, acuta. Stamina ca. 30. filamentibus 4-8 mm longis purpureis luteisque, in medio crassis. Anthera oblonga, glandula parva. Stylus gracilis. 5-6 mm longus. Ovuli 2, recti. A slender shrub with erect branches. Leaves erect or appressed, scattered, oblong-linear, thick, concave. 4-8 mm long, shortly petiolate. Floral leaves similar but with white scarious mai’gins. Flowers terminal. Bracteoles 10-12 mm long, free almost to base, margins broad, scarious, apices acute, recurved. Calyx tube slender, 11-14 mm long, 10-ribbed. solid above the ovary: lobes orbicular at base with slender awns becoming finely scabrous towards the apices, 11-15 mm long. Petals longer than awns, bright pink, broadly elliptical, acute, rather deciduous, the stamens remaining after the petals have fallen. Stamens about 30, fila- ments 4-3 mm long, purple and yellow, swollen in the middle. Anthers versatile, oblong, open- ing in longitudinal slits, connective gland small. Style slender, 5-6 mm long. Ovules 2. erect, attached basally on a short lateral placenta. Holotype and Syntypes: Eneabba. C. Chap- man, late December, 1961. The flowers of Calytrix superba are much larger than in any other species. It is also dis- tmguished by the scarious margins of the biacteoles, the swollen, bi-coloured staminal filaments and the extremely small connective of the anther. Eremaea rosea, sp. nov. Fig. 4, L-V Frutex ramosus erecto-diffusus, 30-50 cm altus. Rami juvenes setosi, mox pubescentes, demum glatari. Folia 6-9 mm longa, alterna. bTeviter petiolata. lineare-lanceolata, concava. carinata, 1-3-nervosa. glabra nisi marginibus setosis. Flores ramos terminantes, sessiles, solitarii vel raro geminis. Bracteae plures. ovatae. obtusae. imbricatae, breviter pubescentes. marginibus ciliatis. Calyx ab bracteis fere occultus, 5-8 mm longus, pubescens, lobis marginibus scariosis ciliatis quam tubum brevioribus. Petala rosea, quam sepala longiora, spathulata, patentia. Stamina in phalangibus petalis oppositis conjuncta, filamentis roseis, antheris luteis. Stylus glaber, staminibus aequalis. Ovarium 3-loculatum, apice convexo dense hirsute. Fructus sessilis, urceolatus. valvis inclusis, lobis calycis deciduis. A spreading shrub with several stems 30-50 cm high. Branches setose when young, later pubescent and finally glabrous. Leaves 6-9 mm long, scattered, shortly petiolate. linear- lanceolate. obtuse, concave, keeled, 1-3-nerved, glabrous except for setae along the margins which become glabrous with age. Flowep terminal sessile, solitary or rarely two within the one series of bracts. Bracts numerous, ovate, obtuse, imbricate, shortly pubescent with ciliate margins, somewhat scarious and striate. Calyx almost concealed by the bracts, 5-8 mm long, pubescent, lobes shorter than tube with scarious ciliate margins. Petals pink, longer than the calyx-lobes, spathulate, spreading. Stamens united in bundles of 12-15 opposite the petals, filaments pink, anthers yellow. Style glabrous, as long as the stamens. Ovary 3 -locular, apex convex and densely hirsute. Fruit sessfie. smooth, urceolate, 6-7 mm long, 7-8 mm wide when young, 10 mm long. 10-11 mm wide when mature, valves included, calyx-lobes deciduous. Holotype and Syntypes: Maida Vale, in sand. A. S. George, 4161, September 10, 1962. Eremaea rosea differs from E. acutifoUa, F. MuelL, in the flat, obtuse leaves, the pubescent calyx with lobes deciduous in fruit, the deep pink flowers and the ovary convex on top. It differs from E. paucifiora, (EndlJ Druce in the broader, flat leaves, the bracts almost completely covering the calyx, the stamens in 5 definite bundles, the deep pink colour, the ovary convex on top and the fruit larger with a wider orifice. GOODENIACEAE Leschenaultia subcymosa, sp. nov. Fig. 5 Herba perennis usque ad 30 cm alta. ramis erectis floridis et ramis diffusis foliolatis. Rami striati. Folia linearia, usque ad 10 mm longa, tj’iquetra. recta. Flores sessiles paniculati ramis cymosis. Bracteae lineares. Calycis lobi lineares, 1.5-2 mm longi, acuti. Corolla alba. Fig. 4 — opposite Calytrix superba, sp. nov. A. — Flowering branch, natural size: B. — Corolla, X natural size; C. — Calyx lobe, 3^ X natural size; D. — Calyx tube, 3 X natural size: E. — Ovary in vertical section, enlarged: F. - Bracteoles, 3 X natural size; G. — Floral leaf, 3^ X natural size; H. — Leaves, 3^ X natural size; I. — Style end, enlarged; J. — Stamen, 10 X natural size; K. — Anthers, enlarged. (All from C. Chapman. Eneabba, late December 1961.) Eremaea rosea, sp. nov. L. — Bracts surrounding flower. 3 X natural size; M. — Flower with bracts removed. 3 X natural size; N. — Staminal bundle. 3^ X natural size; O. — Anthers, much enlarged. P. — Vertical section of ovary, with style, enlarged; Q. — Young fruit, 2 X natural size; R. — Mature fruit, 2 X natural size; S. — Leaf, 2^ X natural size; T. — Flowering branch. 2 X natural size; U. — Fertile seed, 4 X natural size; V. — Sterile seeds, 4 X natural size. (All from A. S. George 4161.) 134 A. 5. Georgl - ^ s Fig. 4 135 Leschenaultia sybcymosa, sp. nov. A. — Diagram of flowering stem, less than natural size; B. — Portion of stem, with T.S. of a leaf, enlarged; C. — Flower, 7 X natural size; D. Base of corolla spread open, enlarged; E. — Indusium from above, much enlarged; P. — Indusium from below, much enlarged; G. — Indusium from front, much enlarged; H. — Indusium from side, much enlarged; I. — Anthers, 10 X natural size; J. — Fruit 4 X natural size; K. — Seeds before falling, 9 X natural size; L. — Seeds. 14 X natural size; M. — Embryo. 14 X natural size. (All from A. S. George 2433.) 136 Neogoodenia minutifiora, gen. et sp. nov. A— Branch and floral scape. ^ X natural size; B.— Leaf, 2 X natural size; C.— Calyx with bract, 9 X natural size; D.— Corolla, 6 X natural size; E.— Corolla spread open. 11 X natural size; F.— Stamens. 18 X natural size; G.— Style and indusium, 18 X natural size; H.— Fruit. 10 X natural size; I.— Seed, 10 X natural size; J.— Embryo, 12 X natural size. (All from A. S. George 910.) 137 15-17 mm longa. lobis subaequalibus alatis patentibus. tubo intus pubescente, ad basin minute glanduloso. Stamina plus minusve libera, quam stylo breviora. Stylus glaber, 10-11 mm longus, indusio in dorso pubescente, labiis ciliatis. Ovarium lineare cylindricum 12-15 mm longum. Ovuli usque ad 15 per loculum. Capsula 25-30 mm longa, erostrata. Semina cylindrica, 1.5-2 mm longa, testa dura horizontaliter tuber culata. A perennial herb to 30 cm high, with a short, thick basal stem and numerous erect or spread- ing flowering stems and shorter, spreading leafy ones. Stems striate. Leaves linear, to 10 mm long, triquetrous, erect. Flowers sessile, in panicles with cymose branches, bracts linear. Calyx-lobes linear, 1.5-2 mm long, acute. Corolla white, 15-17 mm long, lobes subequal, winged, spreading, tube pubescent within, minutely glandular towards the base. Stamens more or less free in the open flower, shorter than the style. Style glabrous, 10-11 mm long, indusium pubescent on the back, lips ciliate. Ovary linear, cylindrical, 12-15 mm long. Ovules up to 15 per locule. Capsule 25-30 mm long, not beaked. Seeds cylindrical, 1.5-2 mm long, testa hard with large horizontal tubercles. Old fruit persistent for some time. Holotype and Syntypes: 56 miles S. of Lear- month, in red sand with spinifex and low shrubs, A. S. George 2433, June 2, 1961. Other collections: Learmonth rd„ 22 miles N. of Warroora turnoff, in red sand. A. S. George 3288, February 22, 1962. Learmonth rd., 18 miles S. of Bullara turnoff, in red sand, A. S. George 3293, February 22, 1962. The species is closest to L. stenosepala, Pritzel, differing in the habit, the shorter calyx- lobes, the white corolla, and the longer ovary and fruit. Genus Neogoodenia, gen. nov. Calycis tubus ovario adnatus, lobis liberis. Corolla minima glabra bilabiata, lobis subae- qualibus exalatis trinervis. Stamina libera. Stylus glaber, indusium labiis ciliatis. Ovarium uniloculatum. Ovulus 1, basiflxus. Fructus multo compressus, indehiscens. Semen mag- num, anguste alatum. Embryo in endospermo copioso I’ectum. Herba prostrata annua. Folia alterna plana. Flores racemosi. Type species: N. minutifiora, sp. nov. Neogoodenia minutifiora, sp. nov. Fig. 6 Herba annua prostrata, glabra calycum nodo- rumque exceptione. Rami 20-40 cm longi, plures. racemis terminantes. Folia alterna, cordato-rhomboidea, plana, obtuse 5-7-lobata, longe petiolata, capillis albis in axilibus. Race- mae aphyllae, multiflorae. Flores breviter pedi- cellatae bracteis linearitaus bracteolatis nullis. Calyx ad ovarium adnatus, compressus, ovalis, 6-nervosus, pubescens vel breviter hispidus, lobae breves. Corolla minima, 1 mm longa, glabra, bilabiata, mox decidua: lobae subae- quales, exalatae, trinervae. Stamina libera, 0.4 mm longa. Stylus 0.5 mm longus, glaber; indusium glabrum, labiis ciliatis. Ovarium 1- loculatum, 1-ovulatum. Fructus compressus, ovalis-obovatus, pubescens, 3.5-4 mm longus, 2.5- 3 mm latus, pericarpium tenue. Semen ovalis-obovatum, marginatum sed etiam anguste alatum. Embryo 2 mm longum in endospermo copioso rectum. A prostrate annual, glabrous apart from the flowers and leaf nodes. Stems 20-40 cm long, slender with one or more branches from each node, all ending in racemes. Leaves alternate, cordate-rhomboid, flat, obtusely 5-7-lobed, on long petioles, with a tuft of white hairs in the axils. Racemes leafless, many-flowered, elongat- ing as flowering progresses. Deciduous white hairs around young buds. Flowers each sub- tended by a linear bract, bracteoles none. Pedi- cels short. Calyx about 1.5 mm long, adnate to ovary, compressed, ovate, each face 3-nerved, pubescent or shortly hispid, lobes linear, short. Corolla very small, about 1 mm long, 5-lobed almost to base, bilabiate, lobes subequal, with- out wings, 3-nerved. Stamens free, 0.4 mm long, filaments flattened. Style 0.5 mm long, glabrous; indusium glabrous, lips ciliate; stigma strap- shaped, protruding in older flowers. Ovary 1- celled, 1-ovulate. Fruit much enlarged as flower matures and falls, much compressed, oval. 3.5- 4 mm long, 2.5-3 mm wide, pubescent; calyx- lobes persistent: seed rim apparent through thin pericarp. Seed basifixed, with a thickened rim. narrowly winged, similar to fruit in shape and size. Embryo straight, erect, embedded in endo- sperm. Holotype: 10 miles south of Mt. Magnet, in red loam, A. S. George 910, August 20, 1960. This extraordinai-y plant falls in the section including Scaevola, Dampiera, Verreauxia and Diaspasis. It differs from them all in its annual habit, the minute wingless corolla and the flat thin-walled fruit. Other differences from indi- vidual genera will be apparent from the descrip- tion. The epithet “Neo-” in the generic name has two shades of meaning. It refers to the unusual characteristics, observed in this plant, which have not been seen in the family Goode- niaceae before: and in a lesser sense it indi- cates that it is the most recently discovered genus in this family. 138 INSTRUCTIONS TO AUTHORS Papers may be submitted to the Society in accordance with Rules and Regulations 38 to 41 inclusive (see below). They should be addressed to The Honorary Secretary, Royal Society of Western Australia, Western Australian Museum, Perth. Authors are solely responsible for the factual accuracy and for any opinion expressed in their papers. They are particularly requested to verify references. Alterations to MSS. submitted to the printer will be allowed only under excep- tional circumstances, and no changes will be permitted after galley-proof stage. In the preparation of MSS. authors are required to follow the C.S.I.R.O., Guide to Authors (C.S.I.R.O., Melbourne, 1953), except that papers longer than 10,000 words (30 foolscap pages of pica type, with 6 inch lines 40 to the page) will not normally be accepted. Authors may be required to meet half the cost of preparation of the blocks of diagrams and illustrations. Authors shall receive a total of 30 reprints free of charge. Further reprints may be ordered at cost, provided that such orders are submitted with the MS. RULES AND REGULATIONS 38. Every paper intended to be read before the Society or to be published in the Society’s Journal must be sent to the Secretaries at least seven days before the date of the next ensuing Council meeting, to be laid before the Council. It will be the duty of the Council to decide whether such contribution shall be accepted, and if so, whether it shall be read in full, in abstract, or taken as read. All papers accepted for publication must be read or otherwise communicated at an ordinary meeting prior to publication. 39. A Publications Committee, appointed by the Council, shall recommend to the Council whether a paper presented to the Society shall be published in the Society’s Journal. The Publications Committee may obtain an opinion from any person it may select on the suitability of any paper for publication. 40. Publication in the Society’s Journal shall only be available to (a) Ordinary Members, (b) Honorary Members, (c) Non-members resident outside Western Australia, who must communicate the paper through an Ordinary or Honorary Member. No paper shall be accepted from a Non-member resident in Western Australia. 41. The original copy of every paper accepted for publication by the Society, with its illustrations, shall become the property of the Society, unless stipulation is made to the contrary, and authors shall not be at liberty to publish their communicated papers elsewhere prior to their appearance in the publications of the Society unless permission for so doing is given by the Society, or unless the Society fails to publish the paper in the Journal of the year in which it is read or otherwise communicated, or of the succeeding year. Journal of the Royal Society of Western Australia, Inc. Volume 46 Part 4 Contents 15. — Some Pelecypods from the Cretaceous Gingin Chalk, Western Australia, together with Descriptions of the Principal Chalk Exposures. By F. R. Feldtmann. 16. — The Effect of the Suppression by the International Commission on Zoological Nomenclature of Zimmermann 1777 upon the Stability of the Generic Name Macropus Shaw 1790. By W. D. L. Ride. 17. Eight New Plants from Western Australia. By C. A. Gardner and A. S. George. Editor: J. E. Glover Assistant Editors: R. W. George, R. D. Royce Annual Subscription: Forty Shillings The Royal Society of Western Australia, Inc., Western Australian Museum. Perth 73061/7/63—570 ALEX. B. DAVIES, Government Printer, Western Australia