LOA i 166 WELLIN ETON p p Fy TRANSACTIONS AND PROCEEDINGS OF ree a ae WADE [Wits Etcur Patzs, AND Tuirty Ficukes tn THE Tuxt.] \ EDITED BY PROFESSOR WALTER HOWCHIN, F.G.S. Assistep By ARTHUR M, LEA, F.ES. [Each Author is responsible for the soundness of the opinions given, and for the accuracy of the statements made in his paper.] PRICE: FOURTEEN SHILLINGS. Adelaide: : ; ; PUBLISHED BY THE SOCIETY, f ROYAL SOCIETY ROOMS, NORTH TERRACE, ADELAIDE, DECEMBER 24, 1931. Printep By GILLINGHAM & Co. Limirep, 106 anp 108, Curriz STREET ADELAIDE, SoUTH AUSTRALIA, : aoe: « Parcels for transmission to the Royal Society of South Australia from the United States of America can be forwarded through the Smithsonian Institution, Washington, D.C. TRANSACTIONS AND PROCEEDINGS OF THE ROYAL SOCIETY OF SOUTH AUSTRALIA (INCORPORATED) VOL. LV. [With Eicur PLates, AND Tuirty Figures tn THE TEXxT.] EDITED BY PROFESSOR WALTER HOWCBHIN, F.G.S. Assistep ny ARTHUR M. LEA, F.E.S. [Each Author is responsible for the soundness of the opinions given, and for the accuracy of the statements made in his paper.] PRICE: FOURTEEN SHILLINGS. i Adelaide: PUBLISHED BY THE SOCIETY, ROYAL SOCIETY ROOMS, NORTH TERRACE, ADELAIDE, DECEMBER @4, 1931. PrintEp BY GILLINGHAM & Co. Limiten, 106 anp 108, Currre STREET, ADELAIDE, SouTH AUSTRALIA. Parceis for transmission to the Royal Society of South Australia from the United States of America can be forwarded through the Smithsonian Institution, Washington, D.C. _—_ ROYAL SOCIETY OF SOUTH AUSTRALIA (INCORPORATED). Patron: HIS EXCELLENCY BRIG.-GENERAL SIR A. G. A. HORE-RUTHVEN, V.C.,, K.C.M.G., C.B., D.S.O. OFFICERS FOR 1931-32. President: PROFESSOR T. HARVEY JOHNSTON, M.A., D.Sc. Vice-Presidents: PROFESSOR J. A. PRESCOTT, M.Sc. A.LC. J. M. BLACK. Hon. Editor: PROFESSOR WALTER HOWCHIN, F.G.S. Hon. Treasurer: Hon. Secretary: B. S. ROACH. RALPH W. SEGNIT, M.A., B.Sc. Members of Council: J. F. BAILEY. ARTHUR M., LEA, F.E.S. C. T. MADIGAN, M.A., B.Sc. SIR JOSEPH C. VERCO, M.D., F.R.C.S. T. D. CAMPBELL, D.D.Sc. HERBERT M. HALE. Hon. Auditors: W. C. HACKETT. O. GLASTONBURY. CONTENTS. Cuewines, Dr. C.: Delineation of the Pre-Cambrian Plateau in Central and North Australia, with Notes on the Impingent Sedimentary Formations. With Map ADELAIDE University FIELD ANTHROPOLOGY, CENTRAL AUSTRALIA— Fry, Dr. H. K.: No. 8.—A Table showing the Class Relations of the Aranda Fry, Dr. H. K.: No. 9—On the Class System, Kinship Terminology, and Marriage Regulations of the Australian Native Tribes Best, E. W.: The Anatomy of an Australian Leech, Helobdella bancrofti .. Trinpae, N. B.: Geological Notes on the Iliaura Country North-east of the MacDonnell Range, Central Australia .. Lea, A. M.: Notes on some Miscellaneous Coleoptera, with Descriptions of New Species. Part viii. Prescott, Pror. J. A.: Atmospheric Saturation Deficit in Australia .. Fintayson, H. H.: On Mammals from the Dawson Valley, Queensland. Part I. Plates i-iii, .. Murray, B. J.: A Study of the Vegetation of the Lake Torrens Plateau, South Aus- tralia. Communicated by J. G. Wood, M.Sc. Plate iv. Howcnin, Pror. W.: The Dead Rivers of South Australia. Part I. The Western Group. Plate v. Buack, J. M.: Additions to the Flora of South Australia. No, 29. Plate vi. Rocrers, Dr. R. §.: Pollination of Caladenia deformis, R. Br. .. Woops, N. H.: Pelecypoda from the Abattoirs Bore, including twelve New Species. Plates vii. and viii. .. CLELAND, Pror. J. B.: Australian Fungi. Notes and Descriptions. No.8 .. Fintayson, H. H.: Notes on some South and Central Australian Mammals. Part 2 ALDERMAN, A. R.; Petrographic Notes on some Basic Rocks from the Mount Barker and Woodside Districts ABSTRACT OF PROCEEDINGS ANNuAL REportT Ostruary NOoTIcE Sir Josepa VERco MEDAL BALANCE SHEETS ENDOWMENT AND SCIENTIFIC RESEARCH FuNpD Donations To LIBRARY List or FELLows Past AND PRESENT OFFICERS OF THE SOCIETY SUGGESTIONS FOR THE GUIDANCE OF AUTHORS Rutes, as AMENDED, 1931 By-Laws INDEX Page 1 12 20 a) 32 39 65 67 91 113 136 143 147 152 161 163 168 184 185 186 187 189 190 196 199 200 201 204 208 2 “GP ano? Well fone) } 1g) 3 i Conal¥3 Well/ape) i Wop bbe Flot Well ¢ ae ° 4 5H ~ hos "1920, + ey ee . a he 1 + “Green Swomp™ om Dairmid Hill Mt Davidson 1 tgarie WH Silerrenu WH = Mt Lothario ae al eBore| Hole 8 WMoevonsan? 22, doitle Swamp Well eGreen Swomp Well Pd x 2 eHit or Miss\Well oy Soo. \ ; co se Folge Wet! ' f3 ‘ Native Well “Tomson RH Allent RES iz Boombrorn W.H. ad (Pert) ef 3¢ Bennett Hills Hills =H 4¢Shepherd Knoll Hi Taylor Crossing Well 3 Wyckham Ca 145) «Jardaiyarda to one another. These are represented by the al’ and al” groups (thick and thin line, respectively) mentioned above. If the Aranda table be compared with Table II., in which the class names of a typical four-class tribe, the Kariera, are tabulated by a similar representation of marriage and descent, it will be seen that the eight-class table represents a quadruplication of a four-class diagram, by a duplication of class names (ipmunnas) and a change from a two-generation to a four-generation cycle. The complexity of the linkage lines of the eight-class table is due to the inter- change of the female ipmunnas Al’ and Al”, Bl’ and B1”, A2’ and A2”, and B2’ and B2”, in the third and fourth generations as compared with the first and second generations. If the class names of 4 two-division tribe such as the Warrunjerri be charted on the same plan as above, it will be found that the pattern of the four-class diagram is reproduced. So far tribes of paternal descent only have been considered. In the case of tribes of two moieties or four classes, such as the Urabunna and Kamilaroi, respectively, in which maternal descent is the custom, the four-class diagram will be found to hold good if the al, Al, etc., terms be transposed so that the male vertical lines then show the alternation of moiety in succeeding generations. No instance is known of a fully-fledged eight-class system with maternal descent, but the table would work out equally well for maternal descent. There is the interesting example of the Dieri, with two moieties only but with an eight-class organisation. Howitt’s genealogical table of the Dieri‘?) will be found to follow the pattern of the Aranda table if the al, Al terms be transposed. 14 With regard to the form of the four-class table, the following more simple linkage would appear to give a satisfactory picture of the marriage and descent relationships :-— a A ; B i ; re 4 5 =e a B b A a B | | | | _ a A b B a A But it will be noticed that the only marriages provided for in this arrange- ment are those of a man’s marriage with his father’s sister’s daughter, and a woman’s marriage with her mother’s brother’s son, The converse marriages with mother’s brother’s daughter and [father’s sister’s son are not represented. To provide for these the long links of the four-class diagram are necessary, and it will be found that this diagram in a new way represents a three dimensional form. Imagine a square vertical column seen from the front of one face, and let each vertical edge represent one of the vertical lines of the diagram. The Banaka BURUNG marriage links will run across the front face, the BANAKA Burung across the back face. The marriage links Palyeri KARIMERA, and PALYERI Karimera, will run from before backwards on the lateral faces of the column. Marriage links of alternate generations are, therefore, diagrammatically on planes at right angles to one another. This can be shown in another way. If each line of the diagram be written as if seen from above, the form of the three rows will be as follows :— Al bl B2 b2 Al bl al Bl a2 A2 al Bl First Row Second Row Third Row where sides aA, bB represent brother and sister relationship, and sidcs aB, bA that of husband and wife. The geometrical form of the eight-class table can now be reviewed and interpreted in the light of the information provided by this simpler table, The point of view is from above in the following representation of the suc- cessive rows (generations) of the more complex table. Al’ bl’ Al” bl” B2’ b2’ B2” b2” al’ Bl’ al” Bl” a2’ A2’ a2” A2” First Row Second Row Al” bl’ Al’ bl” B2” b2’ B2’ b2” al’ Bl” al” Bl’ a2’ A2” az” A2’ Third Row Fourth Row 15 If these diagrams are compared with the four class, it will be found that exactly the same process is taking place, only the marriage links of the second and third rows join corresponding units of two squares. In the first row marriage lines connect units of front and back rows of individual squares. In the second row the lines, instead of running back along the sides of individual squares, join the “opposite number” of the other square, a2’ links with B2” instead of B82’, b2’ with A2” instead of A2’, and so on. In the third row the transverse connec- tions run between squares round the ring of the diagram, and in the fourth row the linkage is front to back in individual squares, as in the second figure of the four-class diagram. The ring structure of the diagram is readily reconstructed for the four-class table; for the Aranda table a form comprising the outer and inner surfaces of a hollow ring would appear to be necessary. Turning now to the practical value of the table. Firstly, there are two interesting features in regard to nomenclature. In tribes with an eight-class organisation and a four-class nomenclature, the named classes are almost invari- ably those represented by the al’, b1’, a2’, b2’ positions in the table. The Mara‘ and Anula tribes are exceptions. Here the four named classes, when charted, occupy the top line of the diagram in the positions al’, bl’, al”, b1”. Otherwise the tabulation works out quite normally. Spencer states that the Mara represent an instance of direct as opposed to an ordinary indirect paternal form of descent. If I read the table rightly, the Mara represent an unusual form of nomenclature not of descent. Some eight-class tribes, such as the Warramunga“) (male descent) state that their classes are grouped in pairs, and the paired classes bear a mutual relation of mother’s mother to one another. If the class names are tabulated according to the Aranda diagram, it will be found that these paired names fall into relative posi- tions, such as that of PURULA and NGALA. A most important service is rendered by the table in the study of relationships. The vertical lines of direct male and female ascent have been mentioned before, the male showing four groups of subclasses of constant moiety and two- generation cycle, the female two groups of alternating moiety in a four-generation cycle. For diagonal lines of symmetry the ring nature of the diagram demands a repetition of the plane table in all directions. Firstly, take any female name of “A” moiety. One generation below and one female place to the right will be found brother’s daughter. Lay a ruler on the diagonal line set by these two names, and the diagonal will be found to traverse a series of female names representing the relationships: father’s father’s sister, father’s sister, self, brother’s daughter, brother’s son’s daughter. This series represents a complete circuit of the ring, and therefore is capable of extension ad infinitum. From the woman’s brother’s point of view the same series of relations is expressed by the same line, except that the names are altered to father’s father’s sister, father’s sister, sister, daughter, and son’s daughter. The relationship of wife is one generation directly above that of daughter, and one female place right of sister in the same generation. An exactly similar series will be found for “B” moiety males and females, only in this case the diagonal extends from above downwards to the leit. ‘ +Since this paper was written, I find that Brown has made this same deduction, “Oceania,” vol. i., 1930, p. 40. 16 VIVON Bean VIYANM BLInwy viInund ening VIAINVNVd BYURUET Ad ald alV al® AG Td AV le I : 1 ' Se ekee =e 7 | : ' , ‘ ' — ' t | ' 1 1 i VIVONVE euetiiqn VUVAVA ereyed Vuvitvd BIPIUE VNVILIAN eiesueg ev raat ca uc? utV e at ce ‘ 1 . i rs es H i rs 1 | 1 f | 1 1 1 1 : : bow nme ome ee ‘ee = ee oe ' i 1 : ; j i 7 t ' 1 : 7 t VTonawNnd ees N VINVNVd BLinuy VIVON pning VIYOANS equeued Ad ald AV al® ld A alV Ae ‘ f , | re Cli i is \ . ; : ; 4 1 i laiatalialllenietaiel | et ee mamnmer 1 H ci 1 r Py i ' ; . VUVLIVd eueliqa VNVILIGW Bseyeg VLVONVG BIVUIe Sy VUVAVSL epesurg weV uc rast wee wv cd ed we i ; i I [rteeeesesees | ! ! | _ I 1 ; | oe eee se ee i ' ; ' 1 ‘ ; 1 ! 1 ‘ et i 1 t a VIVON ues Ni VION eNU yy vinund ening VUNVNVd equeued ald al al V ul® Ad Ad AV Ae ‘VONVAY FHL JO SNOILVIRY SSVI) AHL ONIMOHS— J ATEV 17 ONNaANG suning VUIVNVG eyeued ONAN suning VAVNVG eyeued Ta TY IV 12 1e3 | Tq IV Te 1 4 ' i] 1 \ H t i Idd ATVd BIOUIC YT VUAWIaVa Madjeg TYaATVd ELUTE] VaAAWNIdvya wade ZV cq ca ce ZV eq cd ce eee eee eel a i ‘ 1 {Sa Cen peed ! ' | © : t H 7 ONNANG suning VAVNVd eyeueg ONNaANG Suning VAVNVd eyeueg Id 14 TV ye Ta TY TV ; 1? I f ry . 1 : i I i] omcaeomemny eleiatietetalatatatale : i 1 ! ‘ ! i H : ' TYAATVd EAVES VaAWNIiva Mead[eg IWHATVd PAUTIE VUAWIAVI Headed ev <4 ca ce ZV 24 2d ce . 1 1 4 a a , ij t ’ ONNANE Suning VAVNVG eyeurd ONNUNG suning VAVNVG eqeueg Ta 14 IV Te Ta 14 IV re ‘SdIHSNOILVTAY SSVID ANNOY ONIMOHS—T] aIavy 18 | See a & VIYON STSSN ud wbF i Set WRIVd eBUubeg TON WNYC LIGN . wed SS a ~S BlB4 Led ace 3 fj sewuuuAeu & wa etna vu d At A’ ; 8 a he ww ced mre eee — ; Va waaurey YNVCLIGN eaesueg gq FAS ed y4 Wwrt sqesueg “ qd x oa" 2 1 A . § ‘ #5 : _~S ~ : Se. d BUBB J lets. 19 The relationship series mother’s mother’s brother, mother’s brother, brother, son, and daughter’s son, from a female standpoint, will be found to lie on a diagonal swinging down and right for subclasses Al’, Al”, B2’ B2”, down and left for subclasses A2’, A2”, Bl’, B1”. Husband relationship keeps step one generation above that of son, and one male place right or left of brother. This line indicates the series mother’s mother’s brother, mother’s brother, self, sister’s son, sister’s daughter’s son, from the male point of view. On these diagonal lines of relationship, it will be found that the female names recur in pairs of constant moiety, the males in two groups .of alternating moiety. This is the converse of the vertical lines of relationship. In Table III, where these lines are indicated, it will be seen that there are four lines of male and four of female connections, the same number as are repre- sented as vertical lines in Table I. If any female oblique line series is compared with any male oblique line series, it will be found that the class names of the respective male and female of the same generation bear to one another a relation- ship of husband and wife, brother and sister, “ipmunna” to husband or wife, ‘“ipmunna” to brother or sister, in a cycle of successive generations. This again is the exact counterpart of the relation of male and female expressed by the vertical lines. In other words, the pattern of the oblique system of lines is that of Table I., charted for maternal instead of paternal descent, which, as we have stated before, involves no change of pattern of the linkage system. Consequently, the table can be looked upon as a lattice system, hinged on class names of one hori- zontal row. If the lattice is built with the paternal descent pattern in rectangular form, swinging the vertical lines to a diamond pattern will give the maternal descent form in vertical series; conversely, if built for a maternal descent pattern in the rectangular form, swinging it to a diamond lattice will give the paternal descent arrangement in vertical series. The reciprocating nature of the oblique and. vertical relationship series of Table I. with the alternation of paternal and maternal descent is clear. An expert practical craftsman devoid of theoretical knowledge, can adapt his materials to a new use with the correct technique. In the same way the Urabunna, a two-moiety tribe with maternal descent, and the Aranda, a tribe with an eight- class organisation and paternal descent, are able to make satisfactory arrange- ments for inter-marriages®), This is an interesting example of successful prac- tical application without, necessarily, any appreciation of the theoretical principles involved. The lattice for the four-class system is similar to the eight-class table, except that in the four-class type there are no “ipmunnas,” so the relationships of the class names in the same row in both vertical and oblique systems are those of sister and wife alternately, and four vertical and four oblique lines with a two- generation cycle comprise the relationship series. REFERENCES, (1) Spencer and GiLten, “Native Tribes of Central Australia,” p. 71. (2) SPENCER and GILLEN, “Northern Tribes of Central Australia,” p. 118, SPENCER, “Native Tribes of the Northern Territory,” p. 60. (3) Howitt, “Native Tribes of South-East Australia,” p. 159, (4) SPENCER and Gitten, “Northern Tribes,” p. 104. (5) SPENCER and GILLEN, “Native Tribes of Central Australia,” p. 69. ADELAIDE UNIVERSITY FIELD ANTHROPOLOGY, CENTRAL AUSTRALIA NO. 8 - ON THE CLASS SYSTEM, KINSHIP TERMINOLOGY, AND MARRIAGE REGULATION OF THE AUSTRALIAN NATIVE TRIBES BY H. K. FRY, B. Sc., M.B.B.S (ADEL.), ETC Summary In a recent paper I have described tables of Australian class systems.‘'’ These were a development from one of several attempts to represent compactly the class organisation of the Aranda. Class names were charted in vertical lines of male and female descent, and horizontal lines were drawn to indicate marriage unions. The class names of a four-class system were arranged, then, on a similar plan, and the implications of the eight-class table began to appear. The class names of a two- division tribe were found next to fall in the same pattern as the four-class table. An attempt to find a more simple pattern for the two-class tribal system gave the clue to the “double cross-cousin marriage” nature of the form of the table which had been found previously. This led to the full interpretation of the form of the eight-class table. Diagonal lines of “wife” relationships were apparent in the Aranda table as soon as it was drawn up. Other lines of relationship series were found by charting simple relationships on the table in the manner of plotting graphs on squared paper. The tables were drawn up, therefore, without any reference whatever to native relationship terminology, but show patterns of relationships. 20 ADELAIDE UNIVERSITY FIELD ANTHROPOLOGY, CENTRAL AUSTRALIA. No, 9—ON THE CLASS SYSTEM, KINSHIP TERMINOLOGY, AND MARRIAGE REGULATION OF THE AUSTRALIAN NATIVE TRIBES. By H. K. Fry, B.Sc., M.B.B.S., D.P.H.,, DipLAnth. [Read July 9, 1931.] In a recent paper I have described tables of Australian class systems.) These were a development from one of several attempts to represent compactly the class organisation of the Aranda. Class names were charted in vertical lines of male and female descent, and horizontal lines were drawn to indicate marriage unions. The class names of a four-class system were arranged, then, on a similar plan, and the implications of the eight-class table began to appear. The class names of a two-division tribe were found next to fall in the same pattern as the four-class table. An attempt to find a more simple pattern for the two-class tribal system gave the clue to the “double cross-cousin marriage” nature of the form of the table which had been found previously. This led to the full interpretation of the form of the eight-class table. Diagonal lines of “wife” relationships were apparent in the Aranda table as soon as it was drawn up. Other lines of relationship series were found by charting simple relationships on the table in the manner of plotting graphs on squared paper. The tables were drawn up, therefore, without any reference whatever to native relationship terminology, but show patterns of relationships. Radcliffe Brown has made an intensive study of Australian social organisa- tion, especially from the point of view of native kinship terminology. He has drawn up tables of Kariera type and Aranda type kinship terminology.@? In these, kinship terms are correlated with an arrangement of classes which are represented by a notation which is different to the general notation used in my tables. Brown adopts capital letters for males, small type for females, I happened on the opposite convention. Brown’s notation corresponds with mine as follows :— In the Kariera tables, A=al, B=bl, D=a2, C=b2; in the Aranda tables, Al=al’, A2=al”, Bl =bl’, B2 = bl”, D2 = a2’, D1 = a2”, Cl = b2’, C2 = b2”. The similarities between the two sets of tables are interesting. The Kariera tables are very much alike. In the Aranda tables the same males appear in the vertical lines, which are in different order. If the order of the lines in the kinship table is altered from P R S Q to P R Q,5S, so making the left half of the table symmetrical with the right half, it will be found that the correspondence of the males in the two Aranda tables is as close as that shown in the two Kariera tables. The four vertical series of “wives” in the kinship table are represented in four diagonal lines in the class table, and conversely the four vertical lines of direct female descent in the class table appear in four diagonal lines in the kinship table, in the P R QO S rearrangement. The relationships are expressed in the kinship table as translations of the actual native terms, which Brown supplies elsewhere.“ “) Several alternative translations are possible for many of the native terms, as will be seen by tracing 41 out the linkages of either the kinship or the class table. Either table is equally effective. This demonstration is of importance because it shows that virtually identical results can be obtained by working out the implications of either native kinship terminology or class division. Brown has taught“) that the regulation of native social life in general, and of marriage in particular, is the result of native kinship terminology alone. These views have been widely accepted. But this acceptance has been more passive than active, owing to the extreme difficulty of comprehending the complexity of the terms, involving multiple variants, with which the theory in question has been supported. With the aid of the class tables it is possible to visualise the sig- nificance of involved relationship terms, A close study of the data, which have been used to support the theory that social organisation and marriage are dependent only upon the laws of relationship terminology, will show that an explanation is provided equally well by the use of class considerations alone. The same will be found to be true in regard to the “characteristic features” of Type I. and Type II. marriage law. Further, the marriage permitted by the Variety (a) of Type II. marriage law will be found to be, not a variety of, but a direct contradiction of, the marriage law Type II. In typical Australian tribes, all the tribal members are representatives of some one class, and are all relatives. Class tables are genealogical tables of male and female representatives of all the tribal classes, and show relationships genealogically. The virtual identity, with these, of the tables of kinship terminology is due simply to the fact that the tribal representatives in the latter are arranged in accordance with the genealogical interpretation of their respective relationship terms, and are built up in this way into a genealogical table. It will be seen, therefore, that Radcliffe Brown in his tables, and in his writ- ings generally on Australian social organisation, is dealing, not with mere native kinship terminology, but with the genealogical significance of native kinship terminology, which is identical with the genealogical significance of the native class system. The theory that Australian native social organisation and marriage are deter- mined by kinship terminology, and not by class system, therefore depends upon a subtle distinction which is actually non-existent. The question as to what has been the determining factor underlying the development of the more complex class systems is a matter for speculation. There is the explanation from the premises of kinship terminology that an antipathy to marriages of near consanguinity has demanded marriages between individuals of local groups representing comparatively distant degrees of consanguinity, and that a systematisation of such arrangements has resulted in a dichotomy of class nomenclature This can not be disproved in the absence of historical evidence. On the other hand, with the key of the existing class systems, it is a simple matter to draw up organisations for sixteen and thirty two-class systems, the latter with two main alternatives and several minor alternatives. It would be a very difficult matter to work out such arrangements from the basis of relationship terminology. An argument based on class arrangements could explain the more complex organi- sations by dichotomy or agglutination of pre-existing units. Whether class con- siderations of themselves have been a primary factor in such development or not is again only speculation, but they have the virtue of enabling things as they are to be expressed graphically and simply. 22 _. Similarly the conditions of native marriage may be summarised simply as follows :— 1. The tribal class system determines that a man may marry any woman of his own generation and of a certain class or subclass, named or un- named, and each such woman is called “wife.” 2. “Local rules” in different tribes may: (i.) Prohibit marriage with certain “wives.” (ii.) Permit marriage with women of the same class as “wives” but belonging to a second ascending or descending generation. tii.) Permit “prohibited” marriages with women of a different class Py bs a rs & to that of “wives” under exceptional circumstances. The attempt to formulate these conditions into “fundamental laws of relation- ship terminology” has resulted in an increased complexity of reasoning, and fallacy in deduction. There is also a constant tendency for arguments based on native kinship terminology to drift into arguments involving relationships of con- sanguinity only. The class nomenclature system provides relatively simple and safe lines of argument in the investigation of the Australian social organisation, and a useful check on deductions drawn from the more complex terms of kinship. .ADDENDUM. An illustration of the above is provided by Brown’s description of the Yaralde kinship and marriage system, in “Oceania,” vol. i. part iv., 1931, p. 452, in terms which imply a latent organisation into at least thirty-two subclasses in a tribe without any class nomenclature. This, of course, is not impossible, but certainly raises the question whether an error has not crept in somewhere. REFERENCES. (1) H. K. Fry, Trans. Roy. Soc. 5S. Austr., vol. lv., 1931, p. 12. (2) A. R. Brown, Oceania, vol. i., 1930, pp. 49 and 50. (3) Ibid, p. 323, (4) A. R. Brown, Journ. of the Royal Anth. Inst. of G. B. and L, vol, xliit., 1913, pp. 153 and 154. (5) Ibid, p. 190, et seg. THE ANATOMY OF AN AUSTRALIAN LEECH, HELOBDELLA BANCROFTI BY EFFIE W. BEST (NEE DELAND), M.SC Summary The material studied consisted of three specimens collected by Dr. T. L. Bancroft, and presented to Professor T. Harvey Johnston, by whose kindness I was enabled to examine it. Two individuals were in the form of whole mounts, and the third has been prepared as a series of transverse sections. The leeches were obtained from a turtle, Emydura krefftii, in the Burnett River, Queensland, but no information is available as to their colour and markings in life. The crops of all three specimens were distended with blood. 23 THE ANATOMY OF AN AUSTRALIAN LEECH, HELOBDELLA BANCROFTI. By Errig W. Best (née DeLanp), M.Sc. [Read May 14, 1931.] The material studied consisted of three specimens collected by Dr. T. L. Bancroft, and presented to Professor T. Harvey Johnston, by whose kindness I was enabled to examine it. Two individuals were in the form of whole mounts, and the third has been prepared as a series of transverse sections. The leeches were obtained from a turtle, Emydura krefftii, in the Burnett River, Queensland, but no information is available as to their colour and markings in life. The crops of all three specimens were distended with blood. This species of Helobdella is small, measuring only 6°8 mm: in length and 2°6 mm, in greatest width, and is greatly flattened dorso-ventrally. The leaf-like general form is shown in fig, 1, The mouth is sub-terminal, lying on the second annulus, and the anterior sucker is inconspicuous. The posterior sucker is circular, with a diameter of 1 mm. and is distinctly marked off from the body. The anus opens in the centre of its disc. A pair of eyes is conspicuous at the anterior end of the body, but no segmental sense organs could be recognised. The genital apertures are situated immediately in advance of ganglion 12, and are separated by a single annulus. Annulation was not obvious in the preparations as mounted, so that internal structures are referred for their position to the nerve ganglion rather than to the superficial marks of segmentation unless the contrary is definitely stated. The general form of the body wall shows a certain amount of variation in different regions. This is due to the variable proportions of muscular, glandular, and other elements present, rather than to any alteration in the structures com- posing it. A typical section is shown in fig. 7. The cells of the epidermis are very irregular in shape, approaching a columnar form only at the extremities of the body and in the neighbourhood of the genital apertures. A very large number of epidermal cells are highly granular and modified as unicellular glands. Some of these may be sunk two or three times the depth of the epidermis below the surface, in which case their secretion is poured out through a narrow duct-like prolongation. The epidermis rests on a layer of fibrous connective tissue, the cells of which have particularly dceply-staining, compact, spindle-shaped nuclei. There is no sign of the definite arrangement of longitudinal, circular, and dorso- ventral muscle layers which characterises the more highly organised leeches. Most of the muscle fibres, except those which run in the incomplete septa dividing the somites, are longitudinal. ‘They form irregular masses beneath the epidermis and are almost lacking at the margins of the body (fig. 5). Lying amongst the muscle fibres and. scattered in the connective tissue are a number of very irregular, large, pigment cells containing a highly refractive, granular substance. Minute capillary vessels of the coelomic system form an intricate network among the superficial layers of the body wall, and the larger collecting sinuses with which these ultimately communicate, lie in the deeper layers. The main coelomic’ spaces are four in number, dorsal, ventral, and lateral. The largest is the ventral sinus (fig. 4, 5, 8), in which jie the nerve ‘cord, ventral blood vessel, and the various ducts of the reproductive systems, as well as portions of the alimentary canal, and into which the ciliate funnels of the nephridia open. For the greater part of the body length this space assumes almost the dimensions 24 25 and relationships of an ordinary coelomic cavity, At both extremities of the body the ventral sinus narrows considerably and is lost among the network of fine cavities connecting it with the dorsal and lateral sinuses. The dorsal sinus is much smaller and would be indistinguishable from the larger of the subcutaneous sinuses if it were not for the presence of the thick-walled dorsal vessel within it. It is connected at both ends with the ventral and lateral sinuses by the network of capillaries already mentioned. The lateral sinuses run very close to the margin of the worm, and the body wall above them is composed of the epidermis and a loose parenchyma, but no muscle fibres. These sintises are dilated somewhat in each segment and give off metamerically-arranged branches (the transverse coelomic spaces) to the ventral sinus, They also receive numerous branches from the subcutaneous system of spaces whose arrangement has already been described in the account of the body wall. Except in the large size of the ventral sinus this arrangement of sinuses closely follows that described by Bourne in Glosst- phonia (ulepsine). All these coelomic spaces are lined by an epithelium of large, squamous cells whose nuclei project into the lumen of the sinus. The coelomic corpuscles are small, usually rounded and often binucleate (fig. 7). Closely connected with the sinus system are the so-called blood vessels. There are two main trunks, dorsal and ventral, lying in the corresponding sinuses and connected at both ends by a series of capillaries indistinguishable from those of the sinus system. ‘The dorsal vessel is strongly muscular and lies immediately above the alimentary canal (fig. 5, 6,8). The ventral vessel is larger, thin-walled, and lies just above, and usually close to, the nerve cord. The alimentary canal has the same general arrangement as is present in other members of the genus. There are a retractile proboscis, an oesophagus, a crop with seven pairs of diverticula, and an intestine. The proboscis sheath is lined by a layer of squamous cells continuous with those of the epidermis and, like them, often granular, showing their glandular function (fig. 6). Outside this is a layer of loose muscular tissue whose fibres are more or less circular, and in which the thick-walled dorsal blood vessel is present. This, in turn, is surrounded by the ventral coelomic sinus. The lumen of the proboscis is trifid in transverse section, and is lined by an irregular columnar epithelium. The muscles of the proboscis consist of a very thin sheet of circular fibres and a number of radial fibres with which are interspersed groups of granular secreting cells. In a whole preparation these radial fibres and the gland cells are seen to be quite regularly arranged, giving the peculiar appearance shown in fig. 15. The epithelium surrounding these structures is squamous, like that of the sheath. A very fine ctiticle was observed in places, lining the lumen of the proboscis and of the sheath, and covering the former. Towards the base of the organ the structure of its wall becomes rather looser, and the gland cells stain more deeply with eosin. At the base of the proboscis a thin-walled portion of the alimentary canal receives the ducts of the ocsophageal glands, and may he termed the anterior region of the ocsophagus. This forms loose coils when the proboscis is retracted, but is probably drawn taut by its extension. The posterior region of the oesophagus is lined by a columnar epithelium, similar to that of the anterior part and surrounded by large glandular cells with peculiar and very obvious nuclei (fig. 8). The stomach or crop follows upon the oesophagus in somite 12 and with its diverticula occupies the greater part of the body from somite 11 to the base of the posterior sucker. The form of these structures may be seen in fig. 1. The stomach passes into the intestine at somite 19. The latter is lined by a glandular, columnar epithelium and bears four segementally arranged caeca whose epithelium is of a similar type. At somite 24, the intestine opens by a sphincter into the anterior swollen portion of the thin-walled hind gut. The narrow rectum opens by the anus in the centre of the posterior sucker. " B 26 “i SSS ee Fn = : Cyl terns ge 3 é se Aig, : Ercan gr noe Be Bs y f Ngee ce OY Aeraninans f See he my f 1 = ued Ny, Ty LSS ot NM Ny ad 27 t ‘The oesophageal glands are a pair of conspicuous, compact, triangular organs -tying in somites 8 and 9. The cells composing them are large, with densely granular protoplasm which stains deeply with eosin. The nuclei are rather small and are situated against the outer boundary of the organ. The cells are pyriform and their tapered ends unite to form the duct of the gland, so that the constituent cells retain their individual connection with the oesophagus in spite of the compact nature of the gland. There are six pairs of testes, situated between the crop diverticula in somites 13/14 to 18/19. From each testis a thin-walled vas efferens unites with the vas deferens of its own side. This duct passes forward within the ventral sinus and median to the ovary where this is present. Each vas efferens runs side by side with the vas deferens for some distance, eventually joining it at about the level of the preceding testis. In somites 10 and 11, the vas deferens of each side is thrown into coils within the space median to the anterior cacca of the stomach. This coiled ejaculatory duct passes anteriorly into a large club-shaped vesicula seminalis on each side of somite 10. These organs extend posteriorly to the boundary of the annulus containing the male aperture, and are 1 mm, in length and ‘04 mm. in their greatest diameter. The wall of the vesicula is composed of an outer layer of circular muscle continuous with that of the ejaculatory duct and an inner zone of very large cells, clear towards their outer extremities and very granular towards the lumen of the organ. At the narrow anterior end these cells merge into the clear, tall columnar epithelium of the ejaculatory region of the vas. The nuclei of the secreting cells are small and close to the muscular coat. The lumen of the vesicula is irregular, the tapered distal portion of each secreting cell projecting into it in the form of a minute papilla (fig. 8). This arrangement is doubtless connected with the formation of spermatophores within the organ. The two vesiculae seminales unite near the midline, below the ventral nerve cord, to form a short muscular common duct opening at the male aperture. Below this duct is a small blindly-ending depression lined with columnar epidermal cells similar to those covering the immediate neighbourhood of the genital apertures, and opening at the male pore. In the sectioned specimen male activity was apparently nearly past. The testicular sacs contained only a few scattered sperm morulae and a number of large cells with a reticular or highly vacuolate proto- plasm. On the other hand, the portions of the male ducts contained within the ventral sinus, both vas deferens and vasa efferentia, were swollen with masses of sperms embedded in some sort of prostate secretion, The histological form of these various cells is shown in fig. 9a, b, c. The ovisacs apparently vary considerably in size according to the sexual con- dition of the individual. In oné specimen examined as a whole mount they extended very little behind the ganglion of somite 12, whereas in the material sectioned they reached somite 16 posteriorly and showed. in addition. an anterior caecum which extended into somite 10. The structure of the wall of the ovisac varies little in its entire length. Both within and without there is a squamous epithelium which may be thrown into small folds. The main thickness of the tube consists of a layer of very small muscle fibres in a connective tissue matrix. In the neighbourhood of the genital pores these and the fibres of the common male duct approach those of the body wall in size and become indistinguishable from them. Nephridia are absent from the part of the body anterior to the genital aper- tures and from the posterior sucker. The nephridial funnel is connected with the ventral sinus and the nephridiopores open ventrally in the median third of the body. The coils of the nephridia extend to the margin of the body just inwardly from the lateral sinus. The ciliate funnel (fg. 10) is long and narrow and lined with very long cilia. Its extremity is bifid, each lobe being further partly sub- 28 29 divided into two, and each of these four divisions bears a nucleus. Another nucleus is present at the base of the structure. The funnel projects a little into the dilatation following upon it, and here two smaller nuclei are present. This dilatation takes the form of a large sac whose walls are formed of a single layer of cells surrounding a fibrous capsule. Its lumen is filled with coelomic corpuscles. The tubules of the nephridium extend outward to the margin of the body and then return to open by the nephridiopore just ventral to the ciliate funnel and a little behind it. Typical cells of this nephridial tissue, with their intracellular ducts, are shown in figs. 11,12. The ciliate funnels lie just in advance of the ganglion in the somite which contains them. In all these respects the nephridia resemble those of other members of the same family fairly closely. The central nervous system has the general character described in detail by Ilemmingway for Placobdella pediculata, The highly organised eyes lie well beneath the surface on the third annulus. Each consists of a cup-shaped mass of pigment of the same type as that contained in the ordinary pigment cells of the body wall, which are in this region very numerous (figs. 17, 18). The cup is filled with clear cells of the usual type, but no axial fibres were observed. The front of the cup is filled by a mass of cells reaching the surface of the body at one point and extending slightly beyond the limits of the pigment layer. The nuclei of these cells are seen in section to be situated round the periphery of the mass which is enclosed in a distinct capsule. The protoplasm of these rod-like cells is finely granular. They are homologous with the “tactile cells” described by Whitman, but appear to be more specialised as an optical medium than any studied by him, judging from Miss Merrill’s summary of his work. In the present species these cells form a distinct, clear cornea-like structure, filling the space between the cells of the optic cup and the surface. The nuclei of these “tactile cells” differ from those of similar cells of the marginal sense organs in the more open nature of their chromatin network. Segmental sense organs or sensillae of the usual type were not observed, but the anterior margin of the body is bordered by marginal sense organs. These take the form of groups of rod-like sensory or tactile cells whose protoplasm is densely packed with fine granules and whose nuclei are spindle-shaped. Capillary vessels occur among the sensory cells. Attached to one of the specimens were many solitary, peritrichous ciliates, collected along the edges.of any depression on the surface, Their form is shown in figs. 13, 14. The new species of leech has all the characters of the genus Helobdella R. Blanchard, 1896 (Glossiphoniidae), and the name H. bancrofti is proposed for it in recognition of assistance rendered by Dr. T. L. Bancroft. Its most obvious specific characters are the absence of a dorsal scute, the very compact nature of the oesophageal glands, and the small size of the intestinal caeca. This appears to be the first record of a member of this genus from Australia, though Goddard (1908-9) described several species of the related genus Glossi- phoma. The type slide is being deposited in the South Australian Museum. 30 co oS \ \ qh ts \\ NY LIST OF FIGURES. 31 Fig. 1: Entire animal. Fig. 10: Section of nephridial funnel (semi- » 2: Detatls of region of genital aperture. diagrammatic). ,» 3: LS. Oesophageal gland. » 11: Lateral loop of nephridium. » 4: Portion of T.S., showing sex ducts » 12: Cell from median loop of nephri- within the ventral sinus. dium. » 5: T.S. in region of intestine. » 13: Epizoie ciliates. » ©: T.S. proboscis and. sheath. » 14: Group of similar ciliates. » 7: T.S. body wall. , 15: Anterior end of leech. » 8: Portion of T.S. passing through male » 16: Marginal sense organ. aperture. , 17: Oblique section through eye. » 9a:T.S. testis; 9b: sperm mass from vas , 18: Eyes viewed from above. deferens; 9c: detail of contents of testis. All the transverse sections are somewhat oblique. EXPLANATION OF LETTERING. a.g.m., anterior ganglionic mass; a.oe., anterior region of oesophagus; b., vertebrate blood in crop; b.c.c., binucleate coelomic corpuscle; b.v., blood vessel; ca., capsule. sur- rounding corneal cells; caec., caecum of crop; cap., capillary; c.c., clear cells; c.c.t., corneal cells, “tactile”; c.f, ciliate funnel; c.m., circular muscle; coe., coelomic space; coe.c., coelomic corpuscle; c.t., connective tissue; c.t.c., connective tissue corpuscle; d., duct; d.m.a., depression below male aperture; d.s., dorsal coelomic sinus; d.v., dorsal vessel; e, eye; ej.d, ejaculatory duct; ep., epithelium; epd., epidermis; g.c. gland cell; i, intestine; ic., intestinal caecum; ic.d., intra-cellular duct; Im., longitudinal muscle; ls., lateral coelomic sinus; m., mouth; m.f., muscle fibre; mi.n., micronucleus; m.1., mus- cular layer; m.n., meganucleus; m.s.o., marginal sense organ; n., nucleus; n.c., nerve cord; n.d., nephridial dilatation; o.d., oviduct; oe.d., duct of oesophageal gland; oe.g., oesophageal gland; o.n., optic nerve; ov., ovisac; ova., mass of developing ova in ovisac; p., proboscis; pb.s., proboscis sheath; p.c., pigment cell; pig., pigment cup; p.oe., posterior region of oesophagus; p.s., prostate secretion; p.t.c., prolongation of corneal cells (“tactile”) towards surface (cut obliquely); r.m., radial muscle; s.m., sperm morula; s.n., sperm nucleus; t.c., tactile cell; tv.c., transverse coelomic space; v.c., vacuolate cell; v.d., vas deferens; y.e,, vas efferens; ves., vesicula seminalis; v.s., ventral coelomic sinus; v.v., ventral vessel. BIBLIOGRAPHY. 1884—Bournp, A. G.: ‘Contributions to the Anatomy of the Hirudinea.” O.J.M.S. 24, 1884, 419-506. 1908-9—Gopparp, E. J.: “Contribution to Our Knowledge of Australian Hirudinea.” Part I, P.L.S. N.S.W., 33 (2), 320-342; Part IL, Id. 33 (4), 854-866; Part IIT., Id. 34 (3), 467-486. 1927—-Harpinc, W A., and Moorg, J. P.: ‘‘Hirudinea, The Fauna of British India.” 1898—Lampert, A. M.: “The Structure of an Australian Land Leech.” P.R.S. Victoria, 10, 1898, 211-235. 1894—MerriLL, —: “Preliminary Note on the Eye of the Leech.” Zool. Anz., 17, 1894. 1912—-NacutTries, H. F., Hemminoway, E. E., and Moore, J. P.: Leeches of Minnesota.” “The GEOLOGICAL NOTES ON THE ILIAURA COUNTRY NORTH-EAST OF THE MACDONNELL RANGE, CENTRAL AUSTRALIA BY NORMAN B. TINDALE, SOUTH AUSTRALIA MUSEUM Summary During the combined Adelaide University and Museum Anthropological Expedition to MacDonald Downs, August-September, 1930, some notes were made on the geology of the country forming the headwaters of the Mubunji (Bundey) Creek and its tributaries, the Abmoara (Fraser), Alpara, and Irukaru Creeks. 32 GEOLOGICAL NOTES ON THE ILIAURA COUNTRY NORTH-EAST OF THE MACDONNELL RANGE, CENTRAL AUSTRALIA. By Norman B, Trnpare, South Australian Museum. [Read June 11, 1931.] During the combined Adelaide University and Museum Anthropological Expedition to MacDonald Downs, August-September, 1930, some notes were made on the geology of the country forming the headwaters of the Mubunji (Bundey) Creek and its tributaries, the Abmoara (Fraser), Alpara, and Irukaru Creeks. The outward journey from Alice Springs was made by means of motor trucks, and followed a new track on the northern side of the MacDonnell Range, zeta Bird and Turner Wells, South Point, Tlarndanga (Kerr’s Station), Tjeruka (Peaked Hill), and down the Abmoara Creek to Lilatara (135° 9’ east long. x 22° 25” south lat.), at the junction of Abmoara and Alpara Creeks, where the head station of MacDonald Downs (owned by Mr. C. O. Chalmers) is situated. The outward journey did not afford many opportunities for detailed observations, but on the return trip, which was made by a different route wa Mlarndanga, the Upper Mubunji, Table Hill, Hart’s Range, Arltunga, and thence by the main track through Undoolya to the Alice, several short detours were made to points of interest in the area herein discussed. During our three weeks’ sojourn, the localitics examined included the vicinity of Lilatara; Undala or Bundey Gap, six miles north; Arapia, eight miles north: India Range, three miles south; Ataparapara (Mount Ultim), twelve miles east by south; Mopunja Range, eight miles south-west of Mount Ultim; also Table Hill and its vicinity. The Alpara Creek rises near Mount Swan and follows first a north-easterly and then a northern course to Lilatara (lila, creek; tava, two), where it joins the Abmoara, On the present official maps it is wrongly shown as a tributary of the Apewunga (Plenty) River, which has its source near the Hart Range, and flows eastward and then somewhat morc southward past the Jervois Range. Irukaru Creek has its sources in the gently undulating plateau country south- west of Mount Ultim and, after skirting the western flanks of the latter, enters the Mubunji (Bundey) at Undala. The Alarinjela (Marshall River) has one of its sources on the southérn side of Mount Ultim, and flows south-eastward past Atnoala Springs, to join the Apewunja River. The low plateau between the Alarinjela and Irukaru is, there- fore, on the north-south divide of the MacDonnell Range, which has hitherto been placed much too far to the north-west. Mistake Creek rises on the eastern flank of Mount Ultim and flows east, and then north, to the Sandover. Apparently no previous description of the geology of the area has been published. The country further east was examined by Brown (1), who approached it by a southern route via Arltunga and the lower reaches of the Plenty River. The MacDonnell Range Pre-Cambrian complex underlies the area at no great depth. Gneisses, schists, with both acid and basic igneous rocks, are present. At Mopunja Range (fig. 2) these rocks form an extensive peneplain emerg- ing from a mantle of sediments. At this place three small outcrops of basic igneous rocks have been extensively worked by aborigines for the manufacture of stone axes. 33 uotypadxy oyy fo aqnod . OF San J@Mol1y Oo arn Snes meee en | . S S%1lWwo TUBING 3 eal LNIod AWG Fr ss re Aa rr *) ( fa a i ee coal VW, y ade vpoquvuigenne, 39 ’ Raat VA td MND 53 4 % VaI Wy Wired hee SANDAL DID] T Bae °s R FNS, of: 0 se Mery i “ ‘ i wmnuowy se. NG | mardiroun, SLHOI3H “ WY \ WILTN * Pe) VURdIVENTS | unyndy] oon Fig. 1. Sketch Map of the Hiaura Country, North-east of MacDonnell Range, Central Australia. 34 Several vertical dykes of coarse granite-porphyry run east and west across the country on the Pre-Cambrian plateau south of Peaked Hill. This plateau extends eastward from Ilarndanga (Kerr’s Station), for about twelve miles, to within a mile of Atjerukarukuda Rock, a characteristic outcrop of weathered granite rising a hundred feet from the alluvial plain. Further south the Old Mepenga Barge sid Papp : 7 Eepabiceurtate SE : Pop ma ater Hels. EE) schist and qos a baste Iymeus, Fig. 2. Sketch Section at Mopunja Range (c, 2 miles). Rocks are concealed by limestone and chalcedony beds which form low table- topped hills. Still further south they reappear in the area dissected by the head- waters of the Apewunja River. Lying directly on the Pre-Cambrian gneisses and schists at Mopunja Range is an extensive sedimentary series, consisting in the main of shales and quartzites (fig. 2}. These beds dip at an angle of about 25° to the east-north-east, presenting jAmberlon Bangs _Uilsters of Macdomla Downs Fig. 3. Sketch Section, Abmoara Creek to Amberakana Range (3 mile). on their western aspect a steep quartzite scarp and a slope estimated to be seven hundred feet in height. Further to the south the strike of these beds sweeps to the east in an even cutve, while to the north-west the strike turns gradually to the west, so that at Lilatara similar beds dip a little east of north. At India Range, which runs almost due east and west, sections of beds which appear to be somewhat higher up in the Mopunja Range series are met with (fig. 4). The beds dip at about 20° to the north-north-cast. They consist of Abmoaya or Frases Armberikana Barge Mabunji Bundey “Monrange Benge ree) 1 + $ a ,Tndia Range b Lilatare r ban 1 preyia ne * Macdwnald. Downs ‘tation. » {WM Goorsy a Sketch Section, India Range to Bundey Gap (10 miles). alternating quartzites (sometimes very fine grained, but usually coarse and fels- pathic) and altered shales containing abundant large flakes of detrital mica. West of the gap (three-quarters of a mile wide) through which the Alpara Creek passes, 35 the Mopunja Beds are overlain by horizontally-bedded limestones capped with chaleedony. The unconformity is well revealed in a series of small side valleys. Three miles north of this gap, across an alluvial plain strewn with chalce- dony gibbers derived from the breaking down of the limestone plateau, a further section of the Mopunja Beds is revealed at Lilatara (figs. 3,4). The well on the bank of the Abmoara Creek (in which good water was struck in a coarse fels- pathic grit at 90 fect) revealed some of the features underlying this plain :— Feet. Red soil .. oe ae se iy .. to 10 Horizontal Cemented grit with white chalcedony pebbles... ., 18 Beds. Limestone - hg 2 me ey «620 Feet. Beds Purple and yellow shales Py + .. to 64 Dipping Coarse red siliceous sandstone .. wep a tat a fF c. 20° to Shale ie is any nh, on an 3) 78 North. Felspathic grit » 90 Purple shales similar to those present in the well outcrop at the base of Uldaritja Hill, several hundred yards to the north-west, and are conformable with alternating thin quartzites and shales, with some fine conglomerates. These beds may be enumerated here, in descending order (with approximate thicknesses), as being typical of the Mopunja series as a whole :-— ; Coarse felspathic quartzite .. 7 7 = .. 10 Coarse quartz grit .. i ie ” he on . 45 White shale .. 5 is ms ae re! _ .. 10 Coarse felspathic quartzite a. Ke $4 ahs cm 2 Purple shales rae oh sg st a ei ne .. 30 Coarse felspathic quartzite .. ahs i fy. Me ae Tine quartzite a <* te oe ~ Me .. 10 Coarse felspathic quartzite .. af 3 af ite o. 55 Shales with a thin quartzite and some argillaceous sandstone .. 90 Coarse felspathic quartzite _ ts ka . .. 20 Coarse iron-stained grit or fine conglomerates with unabraded fragments of crystalline constituents ag Bch 73 20 Purple shales The shales everywhere show signs of disturbance, and the felspathic quartzites are permeated by polished slickenslide faces. The aborigines naively account for the latter by saying that some mythical ancestors milled grass seeds upon the rocks. Looking northward from Mopunja, and from Lilatara, across an alluvial plain about five miles wide, ithere can be seen a series of almost horizontally-bedded sediments (of different character from those already described) which extend for some twenty miles across the northern horizon, These beds consist of bright red sandstones and fine quartzites, together with a few argillaceous sandstone horizons. They dip at a low angle (about 3-5°) to the north-north-east, and by their character indicate that they are of littoral origin. Current-bedded and ripple- marked sandstones are well developed, together with numerous fossil beds. In character and content they stand in marked contrast to the beds of the Mopunja series, and, therefore, it is suggested that an unconiormity may be discovered between them. Between Undala and Mount Ultim these sandstones form a plateau sloping gently to the north; it stands four to cight hundred feet above the plain. The south-western margin of this plateau forms a steep scarp and slope, only slightly cut into by a series of small valleys at regular intervals along its margin. Upon 36 the top of the plateau there is a series of smaller flat-topped residuals (culminat- ing in Mount Ultim itself) which have resisted weathering. North of Undala this plateau dips gradually towards the vast Sandover Plain, Fig. 5. Sketch Section, from Alarinjela Creek to Mount Ultim (2 miles). The first observations on these beds were made at Arapia and at Undala, where fossils, in the form of “worm-tracks” and ? Orthoceras casts were dis- covered, but, as the same horizons were afterwards identified and examined in detail at Mount Ultim, the latter occurrence will be described in preference. kon, Mabini o Bundey Creek Moran Arapio Fig. 6. Sketch Section, from Undala to Arapia (c. 2 miles). The Mount Ultim occurrence consists of a thickness of about 800 feet of sediments (fig. 5) which, in descending order, consist of :— Feet Current-bedded Sandstone “i £i rs .. approx. 100 Massive red sandstone... 80 White quartzite with an Orthovera as hotizot ati its base bs 20 Red sandstone with Raphistoma ons be, _ 7 100 Current-bedded_ gritty quartzite ei aL ee 30 Argillaceous sandstone with “worm- tracks” and ripple- marks .. es te A A a s - 50 Fine Quartzite f. ei a ty As 4 v 10 Red sandstone with “worm-tracks” and ripple-marks .. __,, 60 Quartzite .. - Per: c.5 White argillaceous satidstarte with fagsits Ws aS: a 100 Quartzite Ne i % ee _ of 3 10 Argillaceous sandstone... és ik ee a a Sy 120 Highly crystalline quartzite : i i thy 20 Grey argillaceous sandstone with foedhe me ere: 100 Shale a f if, os a A - 7 The current-bedded quartzite forming the summit of Mount Ultim is weathered into numerous shallow caverns and leaning rock-shelters, which have been made use of by aborigines. The mode of weathering is so characteristic a 37 feature of this bed wherever it outcrops that the natives have a rational explana- tion for the occurrence, which takes the form of a legend explaining how a mythical being commenced shelter-making operations many miles to the west and proceeding south-eastward, excavated shelters in turn at Arapia, Mount Ultim, Mistake Creek, etc. At Undala (fig. 6) a bed which was identified with “fine quartzite” of the Mount Ultim series forms a smooth flat surface, upon which the natives have cut a few simple figures. Above and below it there are, as at Mount Ultim, red y Mt OLTIN : Ath _ Cc , Fig. 7. Block diagram of the vicinity of Lilatara. sandstone beds with abundant ‘“worm-tracks.” Many of the boulders also show fine ripple marks. Beds lower in the series do not outcrop. North of Arapia the current-bedded sandstone, corresponding to the beds forming the summit of Mount Ultim, are capped by silicified sandstone. Further to the north the beds appear to dip at a low angle towards the alluvial sediments of the Sandover. Limestone beds, capped by chalcedony, are stated to occur further to the north, but were not examined, A representative series of fossils from these beds has been lodged in the Palaeontological Collection of the S.A. Museum, and it is hoped that their identi- fication will be soon carried: out. 38 Several references have been made to the occurrences of limestone beds capped by chaicedony. At India Hill they lie unconformably on the Mopunja series (fig. 4), while according to Mr. C. O. Chalmers similar beds to these occur north of Arapia. At the first-named place the beds in descending order were observed to be :— Feet Chalcedony .. a 4 7 ws bs - .. 6 Limestone... bd a 2S ae tf ve . 6-20 Kaolin roe - ne ns ok bs a .. 10 Red ferrugineous grit i 7 vt oe ie a ? In the whole of the south-eastern part of the area under discussion and around Table Hill and South Point this formation has given tise to a partly dissected plateau, traces of which can be seen in the form of table-topped hills up to one hundred feet in height. The breaking up of this extensive formation has also given rise to the larger or smaller chalcedony gibbers which everywhere strew the alluvial plains and, as coarse grit or gravel, helps to choke the wide sandy beds of the creeks. SUMMARY. Representatives of what appear to be three, possibly four, geological series (fig. 7) have been noticed :— A. Pre-Cambrian gneiss, schist, with granitic and basic intrusions. B. Mopunja Range series of coarse felspathic quartzites and purple shales resting upon a peneplaned Pre-Cambrian pavement. These beds have a general dip of about 20°-25° to the north-east; the strike vary- ing from north to east and then back to north as one travels from east to west. C. Almost horizontal red sandstones, fine-grained quartzites, and some argillaceous beds, littoral in character and containing numerous fossils, which suggest a tentative “Ordovician” age for the beds. D. Younger limestone series of thin horizontal beds capped by chalcedony, found resting unconformably on sediments of the Mopunja Range type. In the accompanying block diagram (fig. 7) the relationships of the four series are set out in a.simplified manner. ACKNOWLEDGEMENTS. We are indebted to Dr. C. Chewings, who suggested that these notes should be placed on record, and to the members of the Expedition, especially to Dr. H. K. Fry, who was particularly interested in the geology, and to Prof. T. Harvey Johnston, who located the “Orthoceras” horizon at Mount Ultim. Mr. C. O. Chalmers provided native guides and horses, and thus enabled us to visit several otherwise inaccessible spots. Finally, we are indebted to an Mliaura native, Akoambaka by name, who on our first arrival in the district explained the principal details of the physiography of his country by means of a relief map, voluntarily constructed in the sand of the creek bed. REFERENCE, 1. Brown, H. Y¥.L.: “Reports on Arltunga Goldfield . . . and Explora- tions North-east of Hart’s Range, in South Australia.” Parlia- mentary Papers, No. 1,353, Adelaide, 1897, pp. 5-8, and map No. 2. NOTES ON SOME MISCELLANEOUS COLEOPTERA, WITH DESCRIPTIONS OF NEW SPECIES. PART VIII BY ARTHUR M. LEA, F.E.S. Summary The transverse arrangement of the prothoracic granules, often so exaggerated that the prothorax appears traversed by fine carinae, is a very distinctive feature of many species of the genus, and easily recognisable, although abrasion is sometimes necessary to see it. In the 1926 key, a special section, "G," was given for fourteen of them. Other species not previously referred to "G," but with transverse arrangement, are: M. canalicornis, n. sp., contortus, n. sp,, excavatus Lea, ferrugineus Lea, incisipes, n. sp., medianus, n. sp., melancholicus, n. sp., octagonalis Oke, and valgus Pasc. 39 NOTES ON SOME MISCELLANEOUS COLEOPTERA, WITH DESCRIPTIONS OF NEW SPECIES. PART VIII. By Artuur M. Lea, F.E.S. (Contribution from the South Australian Museum.) [Read July 9, 1931.] Family CURCULIONIDAE. MANDALOTUS. ‘Since the key given in the Records of the South Australian Museum, on March 31, 1926, species of this genus have been dealt with as follows :— 1927. Lea, Proc. Linn. Soc., N.S.W., pp. 356-357. 1929. L.c., pp. 528-533. 1931. Oke, Proc. Roy. Soc., Vic., pp. 181-190, The species there dealt with, and the new ones described in the following pages, may be associated with the key in the following positions :— B. ae ere pentagonoderes Lea G,qq... .. .. medians, n. sp. C, ddd, 5a Be leat Oke, and parenthet-G,t. .. .. .. canalicornis, n. sp cus, 1. sp. G,w.. .. 6. «corrugicollis Lea C,h. .. .. «. sternocerus Lea G,v. 2... 2.4. melancholicus, n. sp. Cn... .. dolens Lea H. -. +e +e 6goudiet, n. sp. Dis... 1. «. tstgnis, n. sp. I(or NN,v) .. femoralis Lea DD. e. .. .. rufipes Lea Jum... 6... minusculus Oke DD, eee. .. .. egenus Oke Jipp. 1. .. «. gramnicollis, n. sp. DD, nnn, .. .. fimbriatus Lea K (or YY) .. willosipes Lea DD, r. «. .. explanicollis Oke NN (or NNN) oculivorts, 1. sp. F,dd. .. .. .. tuberipennis Lea NNN. .. +. modicus, n. sp. F,l. ow. 4.) 6.) armicoxis Lea 0. se ee ae 6Geanthocnemis Lea, and F,m(orNN,ww) octagonalis Oke bryophilus Oke G, q. .. contortus, n. sp. and W. 1. oes 4.) 6Cinereus, 1. Sp. tnctsipes, 1. Sp. This leaves two species (excluding synonyms and others transferred to Timareta) for which no positions have yet been suggested, as their types are females. M. imponderosus Lea. A minute species (1°5 mm.), from Queensland. M. latus Lea. A wide, tuberculate, densely clothed species, from Tasmania. The transverse arrangement of the prothoracic granules, often so exaggerated that the prothorax appears traversed by fine carinae, is a very distinctive feature of many species of the genus, and easily recognisable, although abrasion is some- times necessary to see it. In the 1926 key, a special section, “G,” was given for fourteen of them. Other species not previously referred to “G,” but with trans- verse arrangement, are: ‘4. canalicornis, n. sp., contortus, n. sp., excavatus Lea, ferruguneus Lea, incisipes, n. sp., medianus, n. sp., melancholicus, n. sp.y octagonalis Oke, and valgus Pasc. On fairly numerous species the middle coxae are armed, although to see the armature clearly it is sometimes necessary to twist the leg, or to view it from several angles, and a small amount of grease or dirt may easily obscure it. 40 On several species there is a shining ridge, but not a dentiform process, The species so armed, owing to the exigencies of tabulation, were not all associated together in the key. The following are also so armed:—M. contortus, n. sp., ferrugineus Lea, dentipes Lea, medcoxalis Lea, medianus, Lea, oclagonalis Oke, oculivorus, n. sp., oryomus Lea, (more a ridge than a tooth), and valgus Pasc. I do not think the femora in any of the species could fairly be regarded as dentate, the apical incurvature in several species, from certain points of view, appears sudden, but the part before it is rather the abrupt termination of a swell- ing than a distinct tooth, and clothing may also cause deceptive resemblance to dentition, Although in figure 3 (especially on Cand H) Mr. Oke has shown quite strong tecth, he nowhere mentions femoral dentition in his descriptions. The tibiae are very distinctive on the males of many species, but it is usually necessary to examine them from several points of view, or even to detach them from the body (in the case of species with the inner side distinctive), to see their structure clearly; clothing and dried mud are also apt to disguise their features, so in the sketches given no clothing was shown. In sending specimens of M. goudici, Mr, Goudie called my attention to the fact that each of its claw-joints was apparently terminated by a single claw; this at first appears to be the case, but on close examination it may be seen that there are really two claws, very closely applied together; and they are very similar to those of the other species (H of the key) with the scape very thick (except in M. nodicollis, on which the claws are normal), vis —M. ammophilus, crassicornis, herbivorus and pondericornis. This character was previously overlooked, except that for amnophilus it was noted: “Claws subsoldered together at base.” On M. howensis, with a heavy scape, although less heavy than on the species of H, the claws are also approximate. On M. acutangulus, on which the scape is stouter than on most species, other than those of H, the claws are normal. Mandalotus parentheticus, n. sp. 4. Dark brown, antennae and tarsi paler, some parts obscurely darker. Densely clothed with dull brown and grey scales, becoming almost uniformly pale grey on under parts; in addition with stout and usually curved setae, on the elytra confined to a regular row on each interstice. Rostrum with median carina normally concealed. Antennae moderately long. Prothorax moderately transverse, sides strongly rounded, derm concealed, Flytra subcordate, shoulders rounded, base wider than prothorax, interstices even except for feeble alternate elevation; punctures large, but appearing much smaller through clothing, Intercoxal process of mesosternum small, but obtusely conical, Metasternum very short. Basal segments of abdomen flattened in middle. Legs moderately long, front coxae touching, T.ength, 3°5-3°8 mm. 9. Differs in being slightly more robust, intercoxal process of mesosternum unarmed, abdomen more convex, and legs slightly shorter. Australia (Dr. W. Horn). By the upper surface practically indistinguishable from M. blackmorei, but the mesosternum armed in the male, and both sexes distinct from those of that species by the front coxae in contact. The intercoxal process of the prosternum is but little produced, and is obtusely pointed, but in the key the species could only be placed in C, ddd, and associated with M. vacilans, in which the process is also rather feeble; but on that species the front coxae are distinctly, although not widely, separated. The clothing is also different, although not much reliance is to be placed on this. Both specimens have the derm brownish or castaneous. as may be seen where slight abrasions have occurred ; the only parts that are apparently black are on the head. he scales on the prothorax are mostly pale, with five distinct dark lines from base to apex, the median straight, the others evenly 41 curved (less distinct on the female than on the male) ; on the elytra the paler scales are in the minority, and are irregularly distributed. The derm of the pronotum is entirely concealed, but feeble granules are indicated. There are no striking features on the legs, the right front tibia of the male has a minute denticle near the base, but it is not present on the left one. Mandalotus insignis, n. sp. Figs. 1, 14, 17, 18. g. Black, parts of antennae and tarsi obscurely reddish. Densely clothed with scales and setae. Rostrum with median carina indicated throughout. Antennae moderately long. Prothorax slightly transverse, sides slightly increasing in width from base to apical third, and then rapidly narrowed to apex; with large, normally con- cealed granules. Elytra rough, base narrower than widest part of prothorax and unevenly arcuate; with rows of large punctures, distinct on sides, but almost or t 2 3 ‘ 4 , 18 tt EXPLANATION OF FIGURES. t ; 4 ip uN 1" , : 1, front tibia of AMandalotus insignis Lea; 2, of M. contortus Lea, type; 3-5, of M. con- tortus {rom Barrington Tops; 6, of M. melancholicus Lea; 7, of M. incisipes Lea; 8, of M. medianus Lea; 9, of M. acanthecnemis Lea; 10-11, of M. pentagonalis Lea; 12-13, of M. tibialis Lea; 14, middle tibia of M. insignis Lea; 15, of M. glaber Blackb.; 16, of M. canali- cornis Lea; 17-18, hind tibiae of M. insignis Tea; 19-20, of M. medianus Lea; 21, of M. decipicns Lea; 22-23, of M. glaber Blackb.; 24, antenna of M. canalicornis Lea; 25, inter- coxal process of mesosternum of M. niger Lea; 26, variety of same. All without clothing. on 13 s 2% AL 3 quite concealed elsewhere ; alternate interstices irregularly elevated, the third and Afth tuberculate. Basal segment of abdomen with a strong subconical tubercle on each side of middle, slightly nearer base than apex. Front coxae widely separated, but not quite as widely as middle ones; front and middle tibiae notched near apex, hind ones strongly bisinuate on lower surface, the apex incurved and bidentate. Length, 6°5-8°0 mm. 9. Differs in being wider in proportion, elytral tubercles less conspicuous, basal segment of abdomen more convex and non-tuberculate, and tibiae not notched. New South Wales: Bombala. Types, in Australian Museum. 42 Remarkably distinct by the bituberculate abdomen and tibiae of the male. In the key could be associated with M. glaber and decipiens, two polished black species, with very different tibiae, The clothing is so dense that the derm is everywhere concealed, and the type is rather dirty, To the naked eye it appears of a muddy-brown, but on close examination numerous small golden scales may be seen; the setae are numerous, and all the tibiae are fringed with long hairs. The third interstice on each elytron has a fairly large, round tubercle, crowning the apical slope, the fifth has a swelling at the basal third, and then curves outwards, and has three tubercles, one before the one on the third, and two beyond it, there is also a small posthumeral tubercle, invisible from directly above. The pronotum of the female appears to have four feebly elevated tubercles: two in the middle, and two at the base; on the male the two basal ones are very feebly indicated, but not the two median ones. Mandalotus contortus, n. sp. Figs. 2-5. 8. Black, some parts paler, parts of antennae and tarsi obscurely reddish. Densely squamose and setose. Rostrum with median carina obscured but traceable, Antennae moderately long. Prothorax slightly wider than long, angles rounded off, but sides subparaliel in middle; granules conspicuously transversely arranged or subcarinate. Elytra at base narrower than widest part of prothorax, but quite as wide across the post- humeral swellings; with rows of large punctures, partly or entirely concealed by clothing; suture on apical slope, and parts of odd interstices elevated. Meta- sternui and two basal segments of abdomen widely and shallowly concaye. Front coxae widely separated, middle ones each with a conspicuous tooth; front tibiae dilated and suddenly deflected at apex, with an obtuse notch near outer apex, the apex itself acute, middle tibiae strongly arched near apex and acutely pointed, hind tibiae rather strongly curved. Length, 7-9 mm. @. Differs in being more robust, elytra less strongly narrowed behind the posthumeral swellings, two basal segments of abdomen gently convex, middle coxae unarmed, front tibiae less suddenly deflected at apex, the other tibiae shorter, and all with shorter clothing. New South Wales: Ebor (C. F. Deuquet), Barrington Tops (H. J, Carter). A remarkable species. ‘The prothoracic granules transversely arranged, dentate middle coxae, and front tibiae notched near apex, associate it with M. dentipes, from which it differs in being much larger, and elytra rougher. The front tibiae are much wider near apex, the external notch, although distinct, is rather shallow (it is less defined on the type than on the Barrington Tops specimen), and the tip is actually pointed (it is necessary, however, to examine the tibiae from several directions to see these particulars). The general outlines are much as those of M. niger, but the legs are very different. Secn from behind, the base of the elytra appears strongly trisinuate, but from directly above it appears almost evenly arched, with the shoulders clasping the base of the pro- thorax; the third interstice is distinctly elevated near the base, and again beyond the middle, the elevation abruptly terminated at the summit of the apical slope, so as to appear subtuberculate, From one direction the tooth on each middle coxa is seen to be flat, and wider than long, from another it appears as an acute spine. The hind tibiae are shining internally, with transverse granules or short ridges, denoting an approach to the numerous transverse ridges of M. niger. The clothing of the type is of a rather light brown, becoming. paler on the under surface; on the upper surface there are many pale setae, in the majority on the pronotum, in the minority on the elytra, on the legs they are about evenly divided, The specimen from Barrington Tops has the clothing obscured by dried mud, 43 and the hairs on the lower part of thé front tibiae are compacted, so as to appear to be fascictlate near the apex, its front tibiae are longer and more complicated at the apex (figs. 3-5) than on the type (fig. 2), but it was not made the type on account of its poor condition. Mandalotus melancholicus, n. sp. Fig. 6. 4. Black, parts of antennae and tarsi reddish. Densely clothed with sooty or sooty-brown scales, interspersed with sloping or curved setae; on the elytra almost confined to a single row on each interstice; tibial fringes rather long. Rostrum with median carina glabrous throughout. Antennae rather long. Prothorax slightly wider than long, sides rounded and widest slightly in advance of the middle, median line well defined; with flattened granules transversely arranged, or altered to short transverse or oblique ridges. Elytra slightly narrower than widest part of prothorax, base trisinuate, posthumeral prominences feeble and scarcely visible from above; with rows of large punctures, appearing much smaller through clothing, alternate interstices slightly raised. Metasternum and basal segment of abdomen rather shallowly depressed. Front coxae decidedly but not very widely separated, the middle ones almost twice as widely; front tibiae multigranulate internally, somewhat dilated towards base, and then suddenly narrowed to base itself, apex acutely pointed. Length, 5-5-6°5 mm. New South Wales: Armidale (C. F. Deuquet). Two specimens. The transverse or oblique arrangement of the prothoracie granules is not as pronounced as on the species referred to G, in the key, but regarding it as correctly placed there, it could hardly be associated with M. abdominalis (a much smaller and otherwise different species), as the basal segment of abdomen is punctate and clothed; passing that species it could only be associated with M. crawfordi, also much smaller and otherwise different. Regarding it as belong- ing to GG, it differs from M. foveatus, in the much less depth of the depression common to the metasternum and abdomen, that species also has quite rounded prothoracic granules; passing it, it could be placed with M. albonotaius, which is a smaller species, with very different clothing and granules. It seems better referred to G. The middle coxae could hardly be regarded as ridged, although shining along the middle; they are certainly not armed. On one specimen there is an obscurely pale ring on each femur, and a few pale scales on the under surface, but on the other the clothing is practically uniformly dark throughout. In general appearance it resembles M. crudus (with mesosternum armed), arciferus and fimbriatus (with abdomen carinated), and piliventris (with densely clothed abdomen). It is close to M. corrugicollis, but the transverse arrangement of the prothoracic granules much less conspicuous; on that species the ridges on the disc are all distinctly wider than the head, whereas on the present one there are many true granules, and no ridge is the width of the head; on the present species also there is an impressed median line, which is absent from corrugicollis, the front tibiae are more arched at the apex, and the clothing generally is darker. Mandalotus incisipes, n. sp. Fig. 7. é. Blackish, parts of antennae and tarsi reddish. Densely clothed with muddy-brown scales and setae, the latter on the elytra almost confined to a row on each interstice. Rostrum with median carina exposed throughout. Scape rather long and thin (the rest of antennae wanting). Prothorax moderately transverse, sides gently rounded, median line slight; with flattened granules transversely arranged and often elongated. Elytra across posthumeral tubercles (which are rather 44 obtuse) the width of prothorax; with rows of large punctures, appearing much smaller through clothing; odd interstices slightly elevated. Metasternum and basal segment of abdomen with a rather deep excavation. Front coxae distinctly but not very widely separated, the middle ones each with an acute ridge, but not dentate, and separated more than the front ones, front tibiae suddenly notched near lower apex. Length, 6 mm. New South Wales: Mittagong, in January (H. J. Carter), Unique. In the key could be associated with M. dentipes, but the notch on the front tibiae is on the lower side of the apex, on that species it is on the upper side. Mandalotus medianus, n. sp. Figs. 8, 19, 20. $. Blackish, parts of antennae and of legs obscurely reddish. Densely clothed with sooty-brown scales, variegated with stramineous, and interspersed with setae, on the elytra almost confined to a single row on each interstice. Rostrum with median carina concealed towards base, but exposed in front. Antennae moderately long. Prothorax slightly transverse, sides rather strongly rounded; traversed hy numerous fine ridges, becoming granules on sides. Elytra slightly narrower than prothorax, base trisinuate, posthumeral tubercles rather feeble; with rows of large punctures, appearing much smaller through clothing, alternate interstices feebly elevated, the apical slope somewhat rough but not tuberculate. Front coxae widely separated, not much less than the least distance between the middle ones, which are obtusely but fairly strongly dentate; front tibiae rather thin, moderately curved at apex, hind ones longer, shining internally and with transverse ridges across the median third. Length, 5-6 mm. @. Differs in being wider in proportion, prothoracic ridges shorter, two basal segments of abdomen gently convex, legs shorter, middle coxae unarmed, and hind tibiae without transverse ridges. New South Wales (C. F. Deuquet), In the key could be associated with M. oayomus, from which it js distinct by the transverse ridges on the inner side of the hind tibiae, somewhat as on AM. niger, It is somewhat like M. contortus, on a reduced scale, but the tibiae are very different. The paler scales are uniform on the head, form a distinct spot at the base of the third interstice on each elytron, clothe most of the sides, and form fecble spots on the rest of the upper surface; on the under surface and legs they cover about half the derm. ‘The basal segment of the abdomen of the male is flattened and depressed in middle, the flattened space being almost glabrous, and margined externally by a curved line, extending from the tip of the seginent to the middle of the coxa on each side, so that at first glance it appears carinated, although it is not really so. On the female the same space (although gentiy convex) is similarly bounded. Mandalotus canalicornis, n. sp. Figs. 16, 24. é. Black, parts of antennae and tarsi reddish. Densely clothed with scales, interspersed with stout setae. Scape very stout, lower surface grooved on apical third. Prothorax moderately transverse, sides rather strongly rounded, granules transversely arranged and many altered to short ridges, Elytra across middle wider than prothorax, base arcuate, shoulders acutely produced; with rows of large, more or less concealed punctures; the odd interstices slightly elevated. Basal segment of abdomen depressed in middle. Front coxae almost touching, middle tibiae with , a slight notch near lower apex, claws distinctly separated. Length, 4-5 mm. 45 9. Differs in being slightly more robust, two basal segments of abdomen gently convex, legs slightly shorter, and middle tibiae feebly incurved near lower apex. q New South Wales: Armidale (C. F, Deuquet). As the prothoracic granules are transversely arranged the species would not go as far as H (the crassicornis group) in the key; but in any case distinguished from all of that group by the normally separated claws (except M. nodicollts, which is structurally very different), and the transverse arrangement of the granules. Referring it to G, it could be associated with M. acutangulus, which has the scape thinner, but heavier than on other species of the genus (except those of I), and the front coxae more distant (about as far apart as the median ones of this species) ; on this species they are almost in contact. On M. crawford, transversus, and setosus, the scape is much thinner. The three specimens taken have the clothing obscured by dried mud, but on scraping some of this away the distinctive sculpture is revealed. The middle tibiae of the male have a slight subapical notch, but it is obscured by the clothing and invisible from most directions. Mandalotus goudiei, n. sp. 3. Black, parts of antennae and legs reddish. Densely clothed with sooty scales, with variable whitish or greyish markings, and interspersed with sloping setae, also varying in colour. Rostrum short, median carina traceable throughout. Scape very stout, except the basal fourth. Prothorax moderately transverse, sides strongly rounded; with numerous round granules traceable through clothing. Elytra subcordate, shoulders rounded, considerably wider than prothorax across middle; with rows of large punctures, appearing much smaller through clothing, or quite concealed ; alternate interstices very feebly elevated. Basal segment of abdomen flattened in middle. Front coxae touching, middle fairly close together. Length, 3 mm. 2. Differs in being slightly more robust, two basal segments of abdomen gently convex, and slightly shorter legs. Victoria: Black Rock, in May (J. C. Goudie). Numerous specimens obtained by sieving fallen leaves. The smallest of all the species with heavy scape, and one of the most interest- ing of the genus. The claws at first appear to be single, but on close examination are seen to be close together (as on most of the species of H, in the key). On several specimens some of the body parts are reddish. The distribution of the paler scales is scarcely alike on any two specimens before me; on the type they cover most of the rostrum, form a distinct line on each side of the prothorax, and a few discal spots, cover about half of the elytra, of which the largest area begins on each shoulder, is obliquely dilated hindwards till it covers most of the apical slope, and cover much of the abdomen (there is a black median vitta on the three apical segments) and legs. On several specimens the shades are much less sharply contrasted, so that the surface appears rather feebly mottled; on several the pale line on each side of the prothorax docs not extend the full length, but is sharply defined; the markings in the scutellar region are particularly variable. The intercoxal process of the mesosternum is briefly subconical, but as it is short, and alike on both sexes, the species could not fairly be referred to C, in the key. Mandalotus granicollis, n. sp. 3. Black, antennae and legs reddish. Rather densely clothed with brown scales, variegated with grey, and becoming sparse on under surface, most of which is shining; with sparse upright sctae, more distinct on elytra than elsewhere ; tibiae with rather long hairs on under surface, sparse on the front and middle pairs, denser and longer on hind ones. 46 Rostrum with median carina distinct only in front. Antennae moderately long. Prothorax slightly transverse, sides rather strongly rounded ; with numerous small, shining granules. Elytra widest slightly behind shoulders, where the width is equal to that of prothorax, base evenly curved, without posthtmeral tubercles ; with rows of punctures, appearing fairly small through clothing; third and fiith interstices wider than the others, but not elevated above them. Basal segment of abdomen rather deeply depressed along middle, Front coxae touching, hind tibiae with a slight notch near outer apex, and several denticles (obscured by clothing) about inner apex. Length, 3°5 mm. New South Wales: Mount Tomah, in October. Type (unique) in Mr. F, E. Wilson’s collection. The hind tibiae are conspicuously fringed, but as only the male is known, it cannot with certainty be associated with M. inusitatus, in which it is fringed in both sexes; it differs from the male of that species in the conspicuous median depression on the basal segment of abdomen, on each side of which there is a swelling (but not a tubercle or carina) ; passing that species, in the key, the feeble markings of the elytra are,not sufficient to associate it with M. maculatus and cordipennis (two smaller species), passing which it is distinct from M. gymio- gaster (also with a shining abdomen), alpinus and muscivorus, by the very different hind tibiae. It is perhaps nearer muscivorus than any previously named specics. In some lights some of the scales have a faint golden gloss. The pro- thorax has numerous small shining granules, which apparently are normally with- out scales. : Mandalotus cinereus, n. sp. é. Reddish-brown, some parts almost black, antennae and tarsi paler. Densely clothed with almost uniform white or greytsh-white scales, with numerous sloping or suberect setae, on the elytra confined to a single row on each interstice ; front tibiae with numerous long hairs on under surface. Rostrum with median carina shining throughout. Antennae long and thin. Prothorax moderately transverse, sides strongly rounded, granules normally almost concealed. Elytra slightly wider than prothorax, base evenly incurved; without posthumeral swellings; with rows of large punctures, appearing much smaller through clothing; interstices almost even. Two basal segments of abdomen flattened and minutely granulate in middle. Front coxae widely separated, tibiae rather thin, front and hind ones moderately arched at apex, and longer than the middle pair. Length, 45 mm. @. Differs in having the elytra wider, two basal segments of abdomen evenly convex, legs shorter, and front tibiae without special clothing. New South Wales: Darling River flood of May and June, 1890 (R. Helms). Types, in Australian Museum; cotypes, in South Australian Museum. With the general appearance of beach-frequenting species of Timareta, but with fairly distinct ocular lobes, and apical incuryature of prosternum well defined. The intercoxal process of the mesosternum, in the female, is wider than the coxae, but the 1926 key deals only with males; on the male of this species the process is scarcely perceptibly, if at all, wider than the coxae. In that key, pass- ing Jf. rufimanus (which has much shorter legs and antennae and different cloth- ing), it should probably be associated with AM. pallidus and blackmorei, which are very differently clothed, and with shorter antennae. ‘The prothoracic granules are feebly traceable before abrasion, but after this they are seen to be moderately large and obtuse, certainly not very minute (as on the species of V) or ordinarily distinct (as on the species of VVV), On two specimens many of the scales have a silvery gloss. Several females were taken, but only one male. 47 Mandalotus modicus, n. sp. 4. Black, antennae and tarsi obscurely reddish. Densely clothed with muddy-brown, feebly variegated scales, and with sloping and mostly pale setae, on the elytra confined to a single row on each interstice; tibiae with rather long hair on under surface, denser, but not very dense, on the front pair than on the others. Rostrum with median carina fine and distinct to base. Antennae compara- tively long and thin. Prothorax moderately transverse, sides strongly rounded, median line moderately impressed; granules traceable through clothing. Elytra at widest the width of prothorax, conjointly arcuate at base, posthumeral swell- ings feeble; with rows of large punctures, appearing much smaller through the clothing, alternate interstices feebly elevated. Metasternum and basal segment of abdomen concave in middle. Front coxae widely separated, almost as widely as the middle pair, front tibiae moderately arched at apex. Length, 4°5-5-5 mm. 9. Differs in having the elytra across middle considerably wider than pro thorax, basal segments of abdomen evenly convex, legs shorter, front tibiae less curved at apex, and without special clothing. Queensland: Maryborough, abundant in flood debris, in January (E. W. Fischer). The middle coxae are slightly more distant than the front ones, but referring the species to NN, in the key, it could be associated with M. raui, which is a smaller species, with much more separated front coxae; it somewhat resembles M. piliventris, but the abdomen is without the long clothing, which is so con- spicuous on the male of that species. Referring it to NNN, the prothoracic granules associate it with VVV, of which M. subglaber is a smaller and more sparsely clothed species; M. angustus is narrower, with front coxae much closer together and paler clothing; and M. ciliatus has much more conspicuous clothing on front and hind tibiae. It differs from M. albonotatus in the clothing at base of elytra and middle of abdomen; M. angustipictus is narrower, with thicker scape ; and M. similis has smaller prothoracic granules and shorter tibiae; and all three species, which it somewhat resembles, have front coxae less widely separated. It is slightly more robust than M. villosipes, the depression on the under surface shallower, and less trough-like in character, and front coxae twice as widely separated. On many specimens the clothing is obscured by dried mud, but even on others in perfect condition it is only feebly variegated. On abrasion the elytral punctures are seen to be distinctly wider than the striae, before abrasion they appear to be much less, they are larger on the male than on the female. The tibiae of the male are shining and with small granules internally, the hind ones from one point of view appear to be feebly dentate at the middle, and gently incurved between there and the apex. On one female the deciduous mandibular processes are present and boomerang-shaped. Mandalotus oculivorus, n. sp. 4. Dark brown, antennae and legs reddish. Densely clothed with greyish scales and with sloping setac. Rostrum without visible median carina, Antennae moderately long. Pro- thorax slightly transverse, sides evenly rounded. LElytra across middle distinctly wider than prothorax, base conjointly slightly arcuate, without posthumeral swellings; punctures appearing small through clothing, or quite concealed; inter- stices even. Two basal segments of abdomen very slightly depressed in middle. Front coxae distinctly but not widely separated, middle coxae obtusely dentate. Length, 2 mm. South Australia: Smoky Bay, in July (H.C, Allen). 48 A minute species, of which two specimens were sent by Mr. Allen as cating the eyes of seed wheat in the ground, and doing considerable damage. The front coxae are not very widely, although distinctly, separated, and in the key the species might be associated with NN, from all the species placed there it differs in being much smaller. If referred to NNN, it could be associated with M. micro- scopicus, from which, as also from M. inconspicuus, it differs in having the scape longer and thinner, sides of prothorax more rounded, and elytra wider in propor- tion, AZ. puncticollis is an even smaller species, and has the front coxae touching. The middle coxae from some directions appear ridged, from another obtusely dentate, and the species is much smaller than any other having dentate or sub- dentate coxae. The clothing is almost uniformly grey, except that the elytra have a darkly-lined appearance, due to the flattening down of the setae on each interstice, but in some lights the elytral scales have a slight golden lustre. Ona female sent with the type, the setae are more conspicuous and on the elytra are not flattened down. ie ae eG | 91 A STUDY OF THE VEGETATION OF THE LAKE TORRENS PLATEAU, SOUTH AUSTRALIA. By B. Jean Murray, B.Sc. (Communicated by J. G. Wood, M.5c.) [Read August 13, 1931.] PLate IV. CONTENTS. Page. I. Intrropuction +t iy 7 eu 4 ye ts 2 13. 91 YI. Crurmatic Factors” ..- WA 35 , - AA oh eA 8 92 TTI, Epapuic Factors A ns _ a os TARE cud ” _ 93 IV. ANALYSIS OF THE FLoRA OF THE REGION—FLoRISTICS .. vy te a 96 V. ANALYSIS OF THE VEGETATION OF THE REGION-—-ECOLOGICAL .. te vd 98 (a) Life-forms ts es - 5 ei £3 ie i 98 (b) Main assaciations ¥ ie 4 eas Me io . 99 (1) Vegetation of the tableland .. vs a3 ‘a ne 99 (2) Vegetation of the sandhills .. a at ni _ 104 (3) Vegetation of the flats .. a As “ss 7 ate 106 (4) Salinas and fresh-water swamps ¢: ae es Bh 109 VI. Veceration Mar or tHe Laxr Torrens PLATEAU .. aA ee a 109 VII. Summary ar 7 is ne “s _ at as oo Sy 109 LITERATURE CITED is an se on we ie 110 I. INTRODUCTION. The Lake Torrens plateau and surrounding country is an interesting region to a botanist, as it has not been studied ecologically before, although plant coliec- tions were made by some of the early exploring expeditions, notably those led by Babbage and Stuart. Accounts of these expeditions and pioneers of the North- West are given by Threadgill (23) and Richardson (21). The botanical collec- tions made by Hergolt, of the Babbage Expedition, in 1858-9, were named by von Miieller (19). In more recent years Cannon (3) made a brief ecological study of the vegetation round Port Augusta and Tarcoola; Adamson and Osborn (1) described the vegetation at Ooldea, a station on the East-West Trans- continental line, 427 miles west of Port Augusta; and Cleland (4) gives an account of a journey through part of the Lake Torrens plateau and farther North-west to Mount Eba and Tarcoola; it contains plant lists and a brief description of some of the flora, including that on Andamooka. The present paper is based on observations and collections made from 1927 to 1930, during which time all the stations between Carriewerloo and Arcoona (as far north as Young’s Lagoon) were visited and the vegetation studied in different seasons. The country consists of pastoral districts mostly devoted to sheep, although, until recently, cattle predominated on the northern stations on the plateau. The Lake Torrens plateau lies on the stable foreland to the west of the Lake Torrens faults and consists of Ordovician (?) or, possibly, older rocks which have been eroded to form flat-topped hills and stony or “gibber” tablelands; the altitude is from 500-1,000 feet above sea-level. It is surrounded by Recent to Pleistocene country (Jess than 500 fcet) consisting mostly of loamy flats inter- sected by sand-ridges, salt-water lakes and lagoons and fresh-water swamps. The whole area lies between the 5- and 8-inch rainfall isohyets and is part of Tate’s Eremian Region (22). Descriptions of the topography and geology of this district are given by Fenner (6,7), Howchin (12, 13, 14), Gregory (10), and Jack (17). Ward (24) gives the composition of the beds of the plateau as, ‘‘dolomitic limestone; quartzite with shaly bands; greenish shale weathering brown; massive siliceous 92 conglomerate. No fossils known”; and that of the sedimentary beds surrounding it as, “Unconsolidated sand of coast and interior; consolidated sand dunes ; saline, gypseous and calcareous earths; travertine limestone; river, lake and swamp deposits ; lateritic ironstone ; vegetable earths; gravels of outwash fans.” The eastern boundary of the plateau is Lake Torrens, which Madigan (18) has proved to be quite dry, with patches of glistening white salt. After heavy rains it may hold a little water, in places, but this is soon evaporated, Il. CLIMATIC FACTORS. This region is situated within the typical flask-shaped arid belt determined by the trade winds, and converges towards the southern with the southern boundary of this arid belt (11). The country dealt with in this paper lies between the 5- and 8-inch isohyets to the north and south, respectively. The rainfall over this area 1s uncertain and influenced both by monsoonal summer rains and winter rains of Antarctic origin; there are no regular wet and dry seasons, and rain is almost as likely to fall in one month as in another; there is perhaps (particularly towards the southern end of the region) a slight preponderance of winter rains with June as the wettest month, whereas in Central Australia it seldom rains during the winter, TABLE I. Mran Monrury Rainrart rn Points (100 = 1 Incn). jan. Feb. Mar, Apr. May June July Aug. Sept. Oct. Niov. Dec. Year Purple Downs 25 47 56 45 35 89 95 42 41 76 37 68 54 685 Arcoona ... 39 55 51 38 41 79 88 39 33 61 38 46 38 607 Oakden Hills 46 49 45 35 58 74 92 49 56 68 55 64 47 692 South Gap ... 46 47 50 42 52 85 86 48 54 59 46 57 50 676 Whittita ... 45 51 48 43 58 84 99 57 63 66 48 50 50 717 Farina .. .«. 50 53 53 68 42 67 90 37 42 47 48 48 58 653 Port Augusta 69 54 50 72 77 114 117 71 87 95 8&5 69 59 950 [Figures kindly supplied by Mr. E. Bromley, Government Meteorologist. Yearly records up to 1917 are given by Hunt (16).] The average rainfall on most of the sheep stations visited is from 6:5 to 7 inches per annum; this is to some extent misleading, for should the country be assured of its average annual rainfall it would be in a much better condition, but good seasons alternate with periods of severe drought when the rainfall may drop to Jess than 2 inches in a year, to be succeeded by years of greater rainfall until a climax is reached. These very wet years restore the ravages of drought, fill the dams, wells, and waterholes. Thus the reliability of rainfall is poor throughout the area which lies between the isopleths showing 20% to 30% variation from the average amount of rainfall and verging towards 40% variation at the north end of the plateau. Another feature of the rainfall is the manner in which it falls; in some instances the whole year’s rainfall may fall within a few days, while in others it may be scattered through several months in a number of ineffectual falls (3). TABLE IT, DIsTRIBUTION OF ANNUAL RAINFALL. Arcoona. Cane South Gap. Whittita. Highest number of wet days per annum 26 43 38 4] Lowest 7 a. re r sd 9 16 16 13 Mean Te on r x 3 17 30 24 31 Number of Years in Mean ute — 20 34 34 - 32 Mean annual rainfall (in points) .... 607 692 676 717 Highest annual rainfall (in points) .... 1392 1403 1368 1845 Lowest 45 ” ss sid 138 182 168 173 (Figures for the first four items only available up to 1915; last three up to 1928.) 93 No reliable figures for the temperatures are available. The area lies between the 65° and 70° isotherms, and the nearest meteorological stations are at Port Augusta and Farina which have annual mean maximum and minimum tempera- tures of 77:2° and 55°3° Fahr., and 80°6° and 54°2° Fahr., respectively. Iso- therms showing mean monthly temperatures throughout this region are given by Hunt (15). February is the hottest month. Even.in summer there are con- siderable extremes of temperature, the intense heat of the day being rapidly lost at night by radiation, which is very great in such exposed regions ; while in winter the extremes of heat and cold are still greater, and ground frosts fairly common on clear nights during one to four months of the year. Owing to the high tem- peratures throughout the year, the dryness of the air, and the prevalence of drying winds, the rate of evaporation is very great, this area lying between the 60” and 80” isatmics. Managers of some of the stations visited estimate the evaporation of the big dams and waterholes as at the rate of 12 feet per annum; and Howchin (13) gives 100 inches per annum for the amount of evaporation in the arid country round Lake Eyre. Where the variation in temperature is negligible the main characteristics of the vegetation of a region are determined chiefly by the amount of rainfall and its reliability. By these Tate determined the outlines of his Eremian or arid region; and Wood (26) gives the 8-inch isohyet as the lowest margin for the occurrence of mallee in South Australia. In the Lake Torrens District mallee is a climatic indicator only found towards the south-western boundaries of the southern stations, such as Carriewerloo and Yudnapinna, through which the 8-inch isohyet passes. Between the 8- and 5-inch isohyets no mallee occtirs, and the vegetation, which is of the arid type, is fairly uniform in character, differences in the asso- ciations being due mainly to edaphic and not to climatic conditions. Perhaps the reliability of the rainfall has more effect in determining the constitution of the vegetation than the actual amount; a region with an average rainfall of 6-8 inches of normal reliability could support a much less arid vegetation than one where, as in the case of this district, the reliability varies up to 40% from the normal, and wet seasons are widely separated by long dry seasons, tending to increase the xerophytic character of the flora. The vegetation, particularly the therophytic or annual, is noticeably affected by the season in which rain falls. ‘The wealth of herbage produced by winter and spring rains differing largely in constitution (being chiefly composed of Cruciferae, Leguminosae, Zygophyllaceae and Compositae) from that (mainly grasses) grow- ing after summer rains. Certain plants will only develop or flower if rain falls at the particular seasons favourable to them, e.g., the parakeelya and Crinum pedunculatum. Ill. EDAPHIC FACTORS. (A) Hapsirats. There are three main types of habitat in the district :— 1. The gibber tableland; also characteristic of the tops and sides of the flat-topped hills. 2. The red sandhills. 3. The loamy or sandy flats and claypans between the sandhills and sur- rounding the plateau. The gibber tableland comprises most of the plateau and corresponds in extent and boundaries very closely with the area marked “Ordovician” (?) by Ward (24). ‘The gibbers are reddish-brown in colour, of all shapes and sizes, but usually flat and angular, about four inches across by a half to one inch thick; with age and weathering they tend to become rounded and smooth; in some parts younger 94 Eyrian gibbers may be found (17). The mantle of gibbers acts as a natural mulch; bencath them the soil is red, argillaceous or loamy and of considerable depth (some feet). It collects, without gibbers, in depressions to form ‘‘crab- holes” or small swamps. The sandhills on the plateau may be from 30-40 feet high or less and usually lie at right angles to the prevailing winds, generally in a north-easterly to south- westerly direction; they consist of fine to coarse siliceous particles, red in colour. They are separated by loamy or sandy flats or claypans of varying size, The latter have a retentive floor and may hold fresh or salt water for a short time after rain; some are more extensive and form “playa”-lakes or salinas. The district surrounding the plateau on the south and extending to and around the numerous lakes and lagoons around the Island Lagoon, which marks the western boundary of these investigations, consists chiefly of undulating country or flats of light red sandy loam or red sandy soil, between red sandhills, or with patches of scattered sand-ridges. Both these sandhills and flats resemble those to be found in certain localities on the tableland (see map), but they are usually of greater extent and the soil on the flats is usually deeper. There are no permanent water-courses throughout the district, which is out- side the great artesian water system (17); but some of the larger gum-creeks, such as the Elizabeth, Pernatty, and Whittita Crecks, may hold water for a con- siderable time after heavy rains. The tableland and numerous hills and bluffs are drained by smaller stony acacia creeks, which only flow after rain but do not retain water for long. B. Soits. The soils of the Lake Torrens plateau and environs have not been properly investigated yet, but appear to be the typical red soils developed under insufficient moisture conditions characteristic of similar dry regions in sub-tropical latitudes and warmer regions of the temperate zone in other parts of the world, z.¢., South Africa, Spain, and the dry steppes of Russia and round the Caspian Sea (8). As Glinka (8, 9) pointed out, whatever the origin of the rocks subjected to sim‘lar climatic conditions the soils eventually produced from them will be very similar, and this is the case in this region. The water-retaining capacity of the various soils in this region is considerably less than that given by Wood (26) for typical saltbush, bluebush, and mallee soils outside the 8-inch isohyet. The amount of water at saturation in typical soils from Oakden Hills Station is given below -— Tasie ITT. WATER AT SATURATION, Locality. Amount of Water, Tableland (saltbush soil) Er shes a side 240% Gypsum (growing saltbush) .... ety Bs 30°5% Limestone knoll (myall and bluebush) sav ste 22°0% Loam flat (myall and bluebush soil) . on jee 19-0% Claypan he on ae wi uh; 22°5% Claypan, Pernatty “Station | = Bh beg nee 22°5% Dry swamp (lignum) .... eal, rhs ule ea 18°5% Sandhill (mulga soil) .. ~ a shee rats 16°5% Because of the small amount of moisture the vegetation is not well developed, and consequently there is little humus in the soil, which is always alkaline. The author is indebted to Professor Prescott and his staff at the Waite Institute for the analyses of the soil samples given in the following table :-— 95 TABLe IV. ANALYSES OF SOILS. Tableland—Saltbush and samphire. 1. Woocalla: (a) Gibbers over dark red clayey soil .. (b) Ironstone gibbers, red clayey soil .. 2. Top of South Oakden Hill, samphire and saltbush, gibbers over red clayey soil ie 3. Pernatty, gibber plain, no vegetation, polished gibbers over red sand to depth of 4 cm., then clay .. Loamy flat, Whittita—Saltbush, mostly dead. Profile taken from sides of dry sandy creek, washed out in loam flat. (a) Coarse red sand over light bright-red sandy soil, 3 cm... (b) Dull, darker reddish- brown soil, ‘fine particles, 20-29 cm... (c) Subsoil, dark bright- red, hard, some scattered stones ‘of various sizes up to 4 cm. long; patches of white throughout, over 40 cm. (d) Below, damp dark red soil 4 ats 1 Loam flat, Yudnapinna—Fremophila glabra, saltbush and bluebush (Kochia sedifolia). Sandy loam, coarse sand over fine particles, 13 cm. Loam flat, Qakden Hills—Myall and bluebush (K. sedifolia and K. pyranudata) 2 +f ‘3 Toby’s Swamp, Pernatty — Canegrass flat (Glyceria ramigera). Red sand to depth of 4-6 cm.; black soil below Edge of swamp, Pernatty—Gums (Eucalyptus rostrata). (a) Whitish sand (b) Hard greenish- white patch j in sand, no vegetation Claypan, Pernatty—-No vegetation. Hard light red clay... . Pe i an Swamp, Oakden Hills—Canegrass, samphire and typical mixed swamp flora. Clay, brown and cracked on surface, 1-2 cm.; red clay below, 4-6 cm. Lake Torrens, South Gap. 1. Sample taken on lake, 100 yards from shore. (a) To 24 cm. .. (b) Below 24 cm. \ 2. Sample taken on shore two yards from ‘edge of lake—Samphire (Arthrocnemum halicnemoides var. pergranitlatum). Sand - a, ld A _ 7 3. Sample taken on shore, 60 yards from edge of Jake—Samphire, saltbush and blucbush (K. fomen- tosa var. appressa). Sandy soil pH. Value. 7-9 8-6 77 7-8 8-1 8:5 8-3 Total Salts. 0-59 0-18 1-23 1-13 0-45 0-09 Large amount of gypsum 8-6 8-0 8-2 The following was obtained and analysed later by the author :— Sandhills, Oakden Hills—-Mulga. Bright red sand for a considerable depth . high 1-00 0-05 0°50 5-30 0-63 0-01 0-03 Tt will be seen that the alkalinity of the soils is high; and in many cases the percentage of total salts is high also, possibly owing to the frequent occurrence of gypsum throughout the district as well as the large amount of sodium chloride present in many soils, accounting for the halophytic character of the vegetation, t.e., the general occurrence of samphire on the tableland; the saltbush (Atriplex wesicarium) is also more tolerant of salt than the bluebush (K. sedifolia), which usually only occurs in crab-holes or towards the edge of the tableland. 96 IV. ANALYSIS OF THE FLORA OF THE REGION—FLORISTICS. (A) Oricin anp DISTRIBUTION OF SPECIES. This region is part of Tate’s Eremian Region of the Australian Flora (12, 22). In Cretaceous times, when the endemic flora of Australia arose, the Lake Torrens plateau was doubly cut off from the migration of species from the west (25); when the Cretaceous seas receded, the climatic changes of the Tertiary period were already bringing that increasing aridity which has since proved such an effective barrier against invasion from the south-western flora. It, therefore, seems apparent that at least for long periods of time migration from the west into this limited area must have come from the south by way of the narrow coastal belt. On the other hand, eastern communications with this area were never com- pletely severed. The relationships between the percentages of species found also in West and East Australia are almost identical to those quoted by Wood (27) for the Flora of the Gulf Region of South Australia, thus pointing to the same sources of origin. These figures are all the more striking because the individual composi- tion of the floras is very different. Three hundred and eighty-seven plants are recorded for the Lake Torrens Region. Fifty-six Natural Orders are represented, including 177 genera con- taining 372 species, of which 17 were only found in a varietal form; in addition there were 15 varieties of species already found there, and 15 were introduced plants. ‘TABLE V. DISTRIBUTION OF SPECIES THROUGHOUT AUSTRALTA. a Region. Total number of species found in Lake Torrens region .... 387 — Number of species found also in Eastern Australia .. 316 80 8&4 Number of species found also in Western Australia wv. 172 44:5 46:5 Western Australian species found also in Eastern Australia 156 40 41 Species found in Eastern Australia but not in Western Australia... 150 39 40 Species found in Western Australia ‘but not in “Eastern Australia... a. tke ais aye TH 45 3 Species endemic to South Australia ... bed shes ws BL 13 Species found also in Central Australia... we «72 185 Species found also in North and Tropical Australia... 32. =«8 — Species found throughout Australia, chiefly in warm dry parts of Temperate Australia .... Ives aa we «80 205 + Cosmopolitan species... sie Se sng sa ses 5 1 — *In the last column of this table corresponding figures for the Gulf Region based on Wood's figures (27) are given for comparison. Although having most in common with the arid flora of the Far North and with that of the adjacent Flinders Range, it will be seen from the following table, based on localities given by Black (2), that most of the species found on the We eal plateau are widely distributed throughout the drier parts of South ustraha. TABLE VI. DISTRIBUTION OF SPECIES THROUGHOUT SouTiIr AUSTRALIA, Total. %. Total number of species in Lake Torrens region ies vee 387 —_ Number common with Far North .... tis he ub wae 237 6] Number common with dry northern Gsiriets os a we 43 11 97 Total. %. Number common with Far North-West . i 99 25 Number common with western region from Port Augusta ta Ooldea .... used 1 ‘nie wee 132 34 Number common with Flinders Range aks test tev w+ 170 44 Number common with Murray Lands _.... IT am greatly indebted to Mr. Harold Snow, of Rochester (on whose properties most of these exposures occur) for kindly piloting me over the area and also noting on the map the respective positions of the principal occurrences. 134 EYRE PENINSULA Hamley Bridg®ys Wi SCALE 21510 5 9 20 30 MILES Lee eee eee ee ee ee nner eens! Fig. 4. Map of part of South Australia on which is shown, in red, the probable direction of some of the main river courses [the Western Group] before they were truncated. 135 covered by this patch of great stones is unploughable for 16 chains in one direction by an average width of 4 chains. At a mile eastward from the group just described [in Section 522] are several exposures. The three principal ones measure in circumference, respectively, 6 chains, 8 chains, and 10 chains. Another somewhat isolated occurrence is seen in Section 454 [not included in the accompanying map], on the Michel Bros. land, about 14 miles north of the most northerly exposures of those shown in fig. 3, situated directly on the crest of the water-parting between Sandy Creek and Magpie Creek, and, therefore, forms a link between the occurrences found in these respective valleys. This great field of ancient consolidated alluvia is unique. The local streams are small and at a juvenile stage. One has a small permanent spring, the others are mostly dry. The consolidated sediments vary in structure from fine sand up to coarse gravel, and are often current-bedded. The lower portion of Magpie Creek is cut in deep argillaceous alluvium of Recent age. So extensive an area of water-borne heavy material is suggestive of very strong transporting currents. From Yacka, southwards, there are indications of a great river, now extinct, that can be associated with the ancient Lake Frome and Orroroo trunk-waterway that drained portions of western Queensland, and found its discharge by union with another trunk line that came down the wide Koolunga and Snowtown plain to the southern coast. As an indication of the thick alluvial that exists in the Snowtown plain the following quotation may be given:—“In August, 1886, I observed the flood waters of the River Wakefield coming down after a long continuance of dry weather, and the progress in its channel across the plain near Balaklava was only at about the rate of half a mile in 24 hours, and frequently the head of the flood remained stationary for half an hour whilst it poured into one of the many large fissures in the river bed.”—Thos. Parker, C.E., “The Underground Waters of S. Aust.,” Trans. Roy. Soc. S. Aust., vol. x., p. 84, 1888. DESCRIPTION OF PLATE V. Fig. 1. View of Ancient Consolidated Alluvia now undergoing erosion by the Walloway Creek. Fig. 2. Ancient river terrace of Consolidated Alluvia exfoliating into large siliceous masses. Another terrace, at a greater elevation, can be seen in the distance. Hundred of Caltowie. ADDITIONS TO THE FLORA OF SOUTH AUSTRALIA. NO 29 BY J. M. BLACK, A.L.S. Summary Eragrostis japonica (Thunb.) Trin. in Mem. Acad. Petersb. ser. 6:1:450 (1830) instead of E. tenella, Bentli. Fl. Aust. 7:643 (1878) non Beauv. Or E. interrupta, Stapf non Beauv. var tenuissima, Stapf in Fl. Brit. Ind. 7: 316 (1897) ; Black in Trans. Roy. Soc. S. Aust. 48: 253 (1924) ; Poa japonica, Thunb. Fl. jap. 51 (1784). Miss C. D. Niles and Mrs. Agnes Chase in “A Bibliographic Study of Beauvois’ Agrostographie,” published in Contrib. U.S. Nat. Herb. 24:6:135- 214 (1925) show that E. interrupta, Beauv. Agrost. 71 (1812) was based on Poa interrupta, R. Br. Prodr. 180 (1810), which is classed by Bentharn as var. interrupta of E. Brownii, Nees, and cannot therefore be applied to the minute-glumed species which extends from Australia to Japan. Besides, Poa interrupta, R. Br., is invalidated by the earlier name Poa interrupta, Lamk. Tabl. Encycl. 1:185 (1791), which is perhaps the same as the still earlier Poa japonica. (Fig. I.). 136 ADDITIONS TO THE FLORA OF SOUTH AUSTRALIA. No. 29. By J. M. Brack, A.L.S. {Read October 8, 1931.] Piate VI. GRAMINEAE, Eragrostis japonica (Thunb.) Trin. in Mém. Acad. Pétersb. sér. 6:1: 450 (1830) instead of FE. tenella, Benth. Fl. Aust. 7:643 (1878) non Beauv. or E, interrupta, Stapf non Beauv. var tenuissima, Stapf in Fl. Brit. Ind. 7: 316 (1897) ; Black in Trans. Roy. Soc. S. Aust. 48: 253 (1924) ; Poa japonica, Thunb. Fl. jap. 51 (1784). Miss C. D. Niles and Mrs. Agnes Chase in “A Bibliographic Study of Beauvois’ Agrostographie,” published in Contrib, U.S. Nat. Herb. 24:6: 135-214 (1925) show that FE. interrupta, Beauv. Agrost. 71 (1812) was based on Poa interrupta, R. Br. Prodr. 180 (1810), which is classed by Bentham as var. interrupta of E. Brownii, Nees, and cannot therefore be applied to the minute- glumed species which extends from Australia to Japan. Besides, Poa interrupta, R. Br., is invalidated by the earlier name Poa interrupta, Lamk. Tabl. Encycl. 1: 185 (1791), which is perhaps the same as the still earlier Poa japonica, (Fig. 1.) Eragrostis confertiflora, n. comb. instead of F. inlerrupta, “Beauv.” var. densiflora, J. M. Black in Trans. Roy. S. Aust. 48: 253 (1924); Fl. S. Aust. 674 (1929). In raising this variety to specific rank it became necessary to change the varietal name, because of the existence of E. densiflora, Rendle, Cat. Welw. Afr. pl. 2:244 (1899) a species of tropical Africa. (Fig. 2.) Gramen gracile glabrum, 20-40 cm. altum, culmo binodi; foliorum laminae 3-6 cm. longae 1-3 mm. latae, vaginis longiores; ligula brevissima, truncata, ciliolata; panicula erecta, spiciformis, 6-11 cm. longa, 5-8 mm. diam., basin versus interrupta, ramis erectis, revera solitariis sed interdum glomeratim approximatis, 5-20 mm. longis, usque ad basin dense vestitis; spiculae subsessiles, confertae, 2-3 mm. longae, 4-5-florae; glumae acutae, exteriores uninerves, prima 3 mm., secunda { mm. longa, florifera 3-nervis, 1 mm. longa, palea omnino glabra; stamina saepius 2; caryopsis ovoidea, brunnea, nitens, 4 mm. longa. Eragrostis Kennedyae, F, Turner in Proc. Linn. Soc. N.S.W. 19: 535 (1894). Wirraminna, between Lakes Gairdner and Hart; May, 1931; ‘coll. Hon. G. F. Jenkins. A slender grass, with narrow and spike-like panicle, or the branches sometimes spreading slightly, and minute purplish 3-5-flowered hyaline spikelets, the flowers almost globular and only 4 mm. long; the outer glumes 1-nerved, obtuse, ciliate on the nerve and at apex, the lower one 4 mm. long, the upper one 3mm. long. The terminal flower is very small, apparently always barren, and falls off early. The type came from Wonnaminta, near Broken Hill, and specimens have been collected on the River Darling and in the Murchison district, W.A. The species has, therefore, a wide distribution, although it has not been previously recorded in South Australia. The only difference perceptible in the West Aus- tralian specimen (which is in the Sydney Herbarium), is that there are some long hairs at the orifice of the leaf-sheath and the blades are rather flatter. (Fig, 3.) 137 The relative position of these three species may be explained by the follow- ing key :— Rhachilla fragile, disarticulating between the flowering glumes, which fall off with their paleas and ripe grain; panicle-branches mostly solitary, divided and clothed with spikelets to base; glumes and grain minute, the latter brown, shining, barely % mm. long; palea-keels glabrous. (Section Cataclastos). Panicle rather loose; spikelets 4-12-flowered, 2-3 mm. long; flowers oblong; outer glumes subequal; stamens usually 2 .. a: a Panicle usually dense, spike-like; spikelets densely crowded; outer glumes unequal. Panicle always dense, 5-8 mm. diam,; ligule short, ciliolate ; spikelets 4-5-flowered, about 2 mm. long; flowers oblong; flowering glumes twice as long as adjacent outer glumes; stamens usually 2 .. E. confertiflora Panicle 3-5 mm, diam., or sometimes broader, owing to the slightly spreading branches; ligule a minute rim of hairs; spikelets 3-5-flowered, 1-12 mm. long; flowers almost globular; flowering glumes not or scarcely exceeding the adjacent outer glumes; stamens usually 3 Ke x Ra hg + rr E. japonica E, Kennedyoe Eragrostis infecunda, n. sp. Gramen perenne, stoloniferum ; culmi ascen- dentes, rigidi, longi, glabri, 6-20-nodes, ad basin bulbosi ; foliorum laminae involuto- subulatae, infra glabrae, erectae, 3-8 cm. longae, vaginis glabris multo longiores, inferiores planiusculae, 2-3 mm. latae; folia infima ad vaginas rigidas albas reducta; ligula ex annulo brevium pilorum constans ; panicula coarctata, 3-10 cm. longa, 1-2 cm. diam., ramis solitariis, fere simplicibus, puberulis; pedicelli subnulli vel usque ad 2 mm. longi; spiculae paucae in quoque ramo, 3-8-florae, 7-12 mm. longae, circ. 14 mm. latae, floribus saepe distantibus in rhachilla plus minus flexuosa; glumae exteriores hyalinae, uninerves, in carina scaberulae, prima circ. 3 mm. longa, secunda 33 mm. longa; glumae florentes fuscae, glabrac, ovatae, 34 mm. longae, 3-nerves, nervo mediano percurrente vel brevissime excurrente, lateralibus dimidio brevioribus; palea fere aequilonga, non persistens, carinis glabris; antherae 3, lineares, vix 2 mm. longae; styli ad basin coaliti; caryopsis adhuc non inventa. Along the Gilbert and Wakefield Rivers near Riverton, and apparently propagating itself by the long rooting surface runners rather than by grain. The collector, Mr. Worsley Johnston, after careful search during March and April of this year, was unable to find any fruits. Until ripe fruits are discovered, it is difficult to say how the rhachilla breaks up, although it is apparently not per- sistent. In its straggling, wiry stems, and its rather loose spikelets, this grass bears some resemblance to a slender form of Glyceria ramigera, but in the latter species the midnerve of the flowering glume ends at some distance below the summit, while in Eragrostis infecunda it 1s always percurrent, (Fig. 4.) Agrostis limitanea, n. sp. Gramen caespitosum, perenne, 30-45 cm. altum ; culmi graciles, rigidult, erecti, 2-4-nodes; foliorum laminae erectae, superne involuto-subulatae, 4-12 cm. longae, subtus minute scaberulae, vaginis plerumque duplo longiores, summa basin paniculae amplectans; ligula lanceolata, 4-6 mm. longa; panicula diffusa, 8-20 cm. longa, ramis capillaribus, verticillatis, divisis, scaberulis, pedicellis, 3-6 mm. longis, distantibus; glumae exteriores inaequales, acutac, divergentes, secus carinam scaberulae, prima 3 mm. longa, secunda 24 mm. longa; gluma florens 1 mm, longa, glabra, truncata, apice denticulata, 4-5-nervis, mutica; palea paululo brevior; rhachilla in setam glabran vel parce pilosam dimidio vel minus quam palea breviorem producta, ceteroqui nuda; caryopsis conico-oblonga, 14 mm. longa. Near Riverton, March, 1931. According to the collector, Mr. Worsley John- ston, it grows in tussocks inside the railway fence. This might suggest an intro- duced alien, but the railway reserves also serve as sanctuaries for native plants. 138 Although this species has the rhachilla produced in a small bristle behind the palea, it seems, from the absence of hairs at the base of the flowering glume, to be better placed in Agrostis than in C alamagrostis. It differs from C, aequata in the follow- ing characters: stiff stems and involute, not flat leaves, outer glumes about twice as long and much longer than the flowering glume, which has no tuft of hairs at its base. From the Mediterranean Agrostis maritima, Lamk. it differs in the long loose panicle, longer pedicels, palea scarcely shorter than the flowering glume and the bristle at its back. (Fig. 5.) *Schismus barbatus (L.) Jucl, Hort. Linn, 13 (1913) instead of S. calycinus (Loefl.) Coss. et Dur. The name Festuca barbata, |.. (1756) is two years earlier than that of Loefling. *Sphenopus divaricatus. Bute; previously recorded from Port Adelaide, Eragrostis Brownii, Nees. William Creek and Irrapatana (Far North); coll. J. B. Cleland. These northern specimens have the spikelets at first green or purplish, finally straw-coloured, 5-10 mm. long by 2 mm. broad, 10-20-flowered, rarely more. They resemble FE. setifolia, but the spikelets are sessile or almost so, and the base of the stem is neither bulbous nor woolly. Deschampsia caespitosa (L.) Beauv. var. macrantha, Hackel, On wet rocks beside waterfall on Upper Hindmarsh River; coll. J. B. Cleland. Hitherto only recorded (in South Australia) from the South-East. Differs from the typical European form by the involute-subulate, not flat leaves, and by the longer outer glumes—43-5 mm. instead of 3 mm. long. These characters seem to occur also in the East-Australian and New Zealand plant, and justify the varietal name which Cheeseman states was given to it by Hackel. In the Fl. Cap. 7: 587 (1900) Dr. Stapf says that Sporobolus elongatus, R. Br, (S. indicus, R. Br. var. elongatus, Bailey) is a diandrous species distinct from S. indicus and extending from Australia to Japan, Ina specimen from the Finke River, C.A., with panicles to 25 cm. long and much interrupted, I found, however, out of 14 spikelets examined, 10 flowers with 3 stamens, 2 with 2 stamens, and 2 with 1 stamen each. *Ehrharta calycina, Sm. Established on light sandy land at Cockatoo Valley, Barossa Goldfields; Aug., 1931; per A. J. Warren, Department of Agriculture. Native of South Africa. Like all Ehrhartas, a good fodder grass. CYPERACEAE, Cladium Gunnii, Hook. f. was found in 1930 by Prof. J. B. Cleland at Cleland’s Gully, Square Waterhole. Schoenus tesquorum, J, M. Black. Back Valley, near Encounter Bay; coll. J.B. Cleland. Hitherto only recorded from the South-East. CHENOPODIACEAE, Atriplex crassipes, J. M. Black (1918). Mr. R. H. Anderson, in Proc. Linn. Soc. N.S.W. 55:5; 494 (1930) considers that this should be retained as a distinct species, instead of being united to A. elachophyllum, F. v. M, (1869). An examination of the type of the latter species shows that there are almost always 3 small protuberances or tooth-like appendages on one or both faces of the fruit- ing bracteoles. These teeth are lacking in A, crassipes. A, elachophyllum = A. varium, Ewart et Davies (1917) from Central Australia, A. acutibracteum, R. H. Anderson, lc. 500. A new species created for the form of A. leplocarpum var, acuminatum, which I described in Fl. S. Aust. as having “2 small hard dorsal tubercles at the base of the {fruiting bracteoles,”’— Murray Flats; Ooldea; Hughes, 139 PAPAVERACEAE. *Roemeria hybrida (L.) DC. Growing wild near Riverton; coll. Worsley C. Johnston. Mediterranean region. Not hitherto recorded. CRUCIFERAE. Cardamine tenuifolia, Hook. Nonning, E.P.; coll. R. H. Pulleine. An entirely new district for this species, hitherto found only in the South-East. Hutchinsia eremaea, J. M. Black. Wangianna, north of Marree, Aug., 1931, coll. J. B. Cleland. These are much better specimens than those on which the species was founded in these Trans. 47: 369 (1923). Stems ascending, about 25 cm. high; leaves 14-3 cm. long, including the petiole into which they taper; sepals about 3 mm. long; petals 5-6 mm. long, with a bright yellow, almost orbicu- lar, lamina; stamens 6, with anthers 14 mm. long; pods from almost orbicular to ovate and sometimes only 5 mm. long; style exserted beyond the pod to a length of 14 mm.; cotyledons strictly incumbent. *Sisymbrium Irio, L. “London Rocket.” Near railway station of Owen, Oct., 1931, coll, Worsley C. Johnston. First record for South Australia. LEGUMINOSAE. Pultenaea quadricolor, n. sp. Fruticulus erectus, gracilis, ramosus, 20-30 cm. altus, ramis pubescentibus; folia alterna, oblanceolata vel superiora lineari-lanceolata, uninervia, 6-14 mm. longa, 14-2 mm. lata, infra pubcrula, supra glabra et marginibus incurvis concava, apice in mucronem recurvum desinentia ; stipulae subulatae, 2 mm. longae, petiolum brevissimum superantes ; flores axillares, solitarii; pedicelli filiformes, 6-7 mm. longi sed foliis breviores; bracteolae herbaceae, lineares, basi duobus lobulis stipuliformibus instructae, sub imo calyce insertae, calycem aequantes vel paulo superantes; calyx 4-5 mm. longus, parce puberulus, lobis lanceolatis tubo paulo longioribus, duobus superioribus breviter coalitis; vexillum calyce duplo longius, flavum, in medio rubrum, alas flavas carinamque coccineam paulo superans; ovarium pubescens, biovulatum ; legumen ignotum. Back Valley, near Encounter Bay; coll. J. B. Cleland, Nov., 1930. The specific name refers to the green of the leaves, the red and yellow of the standard, the yellow of the wings and the crimson of the keel. Section Coelophyllum. Differs from P. elliptica, Sm. in the broader ‘eaves, flowers all axillary and bracteoles much longer; from P. villifera, Sieb. var. glabrescens and P. trinervis in the fewer and less conspicuous nerves of the leaves and in the flowers on rather long pedicels, not almost sessile. (Fig. 8.) *Alhagi camelorum, Fisch, “Camel Thorn.” This spiny, deep-rooted perennial was sent from cultivated land near Jamestown to the Agricultural Department. Said to occur also in water-channels at Berri. Recorded from Rutherglen, Victoria, in 1919. Its native country extends from Southern Russia to North-western India. *Trifolium lappaceum, L. Echunga—Mediterranean region. ZYGOPHYLLACEAE. Specimens of Zygophyllum ammophilum, from between Coward Springs and Edward Creek, have the capsule only 3 mm. long, with 1 seed in each cell, 4 stamens and petals minute and obovate. 140 RUTACEAE, Phebalium brachyphyllum, Benth. Sherlock (Pinnaroo railway) and Warooka, Y.P., 1930. This dwarf shrub had remained undetected since the original specimens were collected at Encounter Bay and Coffin Bay over 70 years ago. EXUPHORBIACEAE. *Chrozophora tinctoria (L.) Juss. A weed at Appila, An annual of the Mediterranean region, sometimes cultivated for the blue dye which it yields. FRANKENIACEAE, Frankenia annua, Summerh, in Journ. Linn. Soc. 48: 379, t. 17 (1930) var. orthotricha, n. var, Variat pilis patentibus, interdum leviter curvis sed nunquam uncinatis, caulibus pilosioribus et petalis paulo latioribus (4-5 mm, latis). Diamantina River, S.A., May, 1931; coll, L. Reese. UMBELLIFERAE, Eryngium supinum (F. v. M.) n. comb. Caules prostrati, rigidi, costati, fistulati, simplices vel parce ramosi; foliorum laminae flaccidae, cuneatae, com- plicatae, acute trilobae, 1-2 em. longae, 3-6 mm. latae, reticulato-nervosae ; petiolus 3-6 cm. longus, fistulatus et parce septatus, basi dilatatus ; capitula axillaria et radicalia, brevissime pedunculata, adulta oblongo-cylindrata, 12-17 mm. longa, 8 mm. diam.; involucri bracteae 5-8, lineari-lanceolata, 7-8 mm. longa, fere pungentia; petalorum apex inflexus, fimbriatus; receptaculi squamae conico- acuminatae; mericarpium apud commissuram rotundatum; vittae 5, duae com- missurales approximatae.—BE. plantagineum, F. v. M. var. supinum, F. v. M. in herb. S. Aust.—Diamantina River, coll. Dr. Morgan; near Innamincka, coll. kk. Cockburn. Qld.—Wills Creek. Dr. Murray, of Howitt’s Expedition. Differs from E. rostratum and E. plantagineum in its prostrate habit and very short peduncles; from E. vesicwlosum in its rigid stems, 3-lobed, not many-toothed leaves, and shorter peduncles, the radical heads being practically sessile. (Fig. 6.) It has been stated, apparently in all Australian floras (my own of South Australia included), that our species of Eryngium have no vittas. All our local species have 5 vittas. The mericarps of Australian Eryngia are orbicular in trans- verse section and have a narrow commissure, so that the two commissural vittas are close together; in European species the mericarp, when cut across, is almost triangular, with a very broad commissure, so that the two commissural vittas are far apart. *“Bupleurum subovatuan, Link (1818) has appeared at the Grange, near Adelaide; a new record.—Mediterranean region, More usually known as B. pro- tractwm, Hoffmannsegg et Link (1820). ERICACEAE, *Erica arborea, L. Roadside, near Aldgate. Flowering Oct., 1931; coll. J. B. Cleland. This is the “White Heath” or “Tree Heath” of gardeners, the “bruyére arborescente” of the French, from whose root “briar pipes” are made. Mediterranean region. Recorded as an escape in Victoria, EPACRIDACEAE, Leucopogon collinus (Labill.) R. Br. Bangham Forest Reserve, near Frances; 1930; coll, J. B. Cleland. 141 BoRRAGINACEAE, Embadium, n. gen. (From Greek embadion, a little slipper, to which the nutlets bear some resemblance.) Calyx 5-sectus; corolla 5-loba, inappendiculata; stamina 5, inclusa; stylus inter 4 lobos ovarii insertus, stigmate capitato; nuculae 4, suberectae, ovatae vel fere triangulares, superne liberae et stylum multo superantes, dorso margine tumido inflexo crenulato circumdatae, cum parva gibba oblonga tumida mediana semen tegente ; nuculae interne convexae, pilis minutis uncinatis conspersae, areola mediana ad gynobasin pyramidalem affixae, ab areola usque ad apicem partis seminiferae carinatae. Herba annua; pedicelli fructiferi recurvi. Embadium stagnense, n. sp. Herba annua, pilis appressis e tuberculis ortis scabrida; caules rigiduli, ascendentes, 5-12 cm. longi, parce ramosa; folia radicalia rosulata, 1-2 cm. longa, longe petiolata, caulina rigidula, sessilia, dis- tantia, oblonga vel lanceolato-ovata, 5-15 mm. longa, in bracteas florales transeuntia; pedicelli fructiferi valde recurvi, 4-8 mm. longi; calycis segmenta lanceolato-ovata, 2 mm. longa, sub fructu patentia; corolla 24 mm. longa, sine squamis in faucibus, lobis tubo brevioribus ; nuculae paululo infra medium parva areola ad gynobasin affixae. On recently flooded land at Arcoona, west of Lake Torrens, Sept., 1927; coll, Miss Beatrice Murray. (Fig. 7.) Mr. Ivan M. Johnston, of the Gray Herbarium, Harvard University, and a specialist on Borraginaceae, considers that although this little plant approaches Eritrichiwm in the attachment of the nutlets to the receptacle or gynobase, it has other peculiarities which necessitate the creation of a new genus. After examining a specimen which was forwarded to him, he writes:—‘The tumid margin combined with the medial crest of the nutlets is unique in the family. Such excessive developments of nutlet margins are usually found in the Cynoglosseae. The uncinate pubescence on the fruit is quite characteristic of that tribe also. In fact, | might say that in every positwe' character, except nutlet attachment, it fits into that tribe best, and close to Omphalodes. 1 have, on various occasions stated that I believe that the strict and sole use of nutlet attachments in defining the tribes of the Borages leads to unnaturalness in classification. I am inclined to believe that your plant is a case in point, and that although its technical characters may place it near Hackelia and Eritrichium, in all probability its nearest relationships are in Omphalodes. If this is the case, your plant is a curious Australian development of the Cyno- glosseae springing from the same immediate stock as has Omphalodes, but which, although developing curious marginal structures, has persisted in a primitive attachment of its nutlets.” SCROPHULARIACEAE. Limosella Curdicana, F.v. M. Beresford (between Marree and Oodnadatta). Flowering and fruiting Aug., 1931; coll. J. B. Cleland. A much more northerly site than any yet recorded. LABIATAE, Prostanthera aspalathoides, A. Cunn, Nonning, E.P.; coll, R. H, Pulleine. A new district for this species. Leaves 4-14 mm. long; calyx 12 mm. long, purplish. *Salvia lanigera, Poir. (1817). Netherton, near Tailem Bend. An almost ~ woolly weed not hitherto recorded, Southern Italy, Spain, North Africa, Syria. A synonym is S. controversa, Ten, (1830). COMPOSITAE. Myriocephalus rhizocephalus (DC.) Benth. var, pluriflorus, J. M. Black. Beresford (between Marree and Oodnadatta); coll. J. B. Cleland. A more 142 northerly site than any previously recorded. The uppermost leaves are sometimes no longer than the general involucre. *Matricaria multiflora (Thunb.) Fenzl. First record of this rather showy South African annual, which has established itself over a considerable area near Calomba, a railway station about 7 miles north-west of Mallala. It has numerous small bright-yellow homogamous-discoid heads arranged in dense corymbs. Minuria rigida. Leaves sometimes nearly all opposite, and the plant may be not more than 8 cm. high. Diamantina River, S.A.; coll. A. M. Morgan. DESCRIPTION OF PLATE VI. Fig. 1. Eragrostis japonica: a, spikelet; b, flowering glume and palea; c, grain. Fig. 2. E. confertiflora: d, leaf and ligule; ¢, spikelet; f, flowering glume and palea;: g, grain. Fig. 3. E. Kennedyae: h, spikelet; i, lowering glume and palea; j, grain. Fig. 4. E. infecunda: k, spikelet; 1, pistil; m, base of stem, Fig. 5. Agrostis limitanca: n, spikelet; 0, flowering glume and palea; p, abaxial face of grain, showing small embryo; q, adaxial face, showing longitudinal groove. Fig. 6. Eryngium supinum: r, flower; s, petal; t, cross section of ripe mericarp; u, blade of leaf. Fig. 7. Embadium stagnense: v, outer or dorsal face of nutlet; w, inner or ventral face; +, tubercle-seated hair. Fig. 8. Pultenaca quadricolor: y, flower; 2, leaf. POLLINATION OF CALADENIA DEFORMIS, R. BR. BY R. S. ROGERS, M.A., M.D., F.L.S. Summary It is a singular fact, that although considerable attention has been directed to the pollination of Australian orchids, very few observations have been recorded in regard to the large and conspicuous Caladenia. By far the most important paper that has yet appeared is that of Oswald H. Sargent ‘'? on the "Pollination of C. Babarossae,” Rehb. f., a Western Australian plant, in 1907. 143 POLLINATION OF CALADENIA DEFORMIS, R. Br. By R. S. Rocers, M.A., M.D., F.L.S. [Read October 8, 1931.] It is a singular fact, that although considerable attention has been directed to the pollination of Australian orchids, very few observations have been recorded in regard to the large and conspicuous gents Caladenia. By far the most important paper that has yet appeared is that of Oswald H. Sargent on the “Pollination of C. Barbarossae,” Rehb. f., a Western Aus- tralian plant, in 1907. In his great and well-known work, Fitzgerald makes three brief and rather unconvincing contributions to the subject. In the first of these, when describing C. dimorpha, he states: “This is the only species of orchid I have known, when placed in a room, to be fertilized by insects. A house-fly, lighting on the lip, was carried by its spring against the column and, becoming entangled in the gluten of the stigma, and struggling to escape, removed the pollen in its masses from the anther and smeared them on the stigma. Such rather large insects arc, I believe, the principal agents of fer- tilization in the genus, the species of which without such agency, never produce seed.” His next reference is to C. tesselata, Fitzg.: “On examination of the plant as it grew, the pollen was found to be drawn out of the anther and attached to the centre of the stigma by a little group of chaffy scales of some plant, which helped to form a cocoon. This cocoon belonged to a dipterous insect, and the flower must have been fertilized by the efforts of the inmate to get rid of its covering. A method of fertilization that may frequently occur, as the dorsal sepal presents a suitable shelter for an insect about to undergo a change, but a method that would hardly be conjectured if not observed.” His third observation is as follows: “On one occasion I had the pleasure of seeing C. alba actually fertilized by an insect. A flower was observed to tremble and, on examination, it was found that a fly had alighted upon its labellum, and was by its spring carried against the stigma and, adhering to it, struggled violently to escape, and thereby withdrew the pollen-masses from.the anther and smeared them on the stigma. This instance, in my opinion, goes far to show that though the pollinia in this and many other species may, without fertilizing the flower, be easily removed by touching the discs, the operation is not by any means so neatly performed by an entrapped insect, and the consequence is that the flowers are impregnated by their own pollen.” While admitting that pollination may possibly have been effected in the manner described, it is suggested that such instances are of a fortuitous or acci- dental nature and lack the purposive character which usually marks the act of pollination in orchids under normal conditions. This act is often intricate and, as far as our experience goes, never clumsily performed. The mechanism involved is beautifully adapted for its intended purpose, and wherever success- fully investigated has proved to be unencumbered by superfluous or unessential parts. It frequently happens that some minute structural detail, apparently too trivial to claim the attention of the observer, ultimately proved an important factor in the process. @ “Journ. Nat. Hist. Soc. of W.A.,” No. 4 (1907), p. 6. () “Australian Orchids.” 144 More or less complicated structures are present in the three species of orchid to which Fitzgerald refers. That they perform some definite function in the normal pollination of the plant there can be no doubt—a function which remains unexplained and valueless in the instances recorded by him. The patient and careful observations of Sargent form a much more valuable contribution to the literature of the subject, They are confined to the investigation of a single species endemic to his State, and they extended over a period of two or three years. He furnishes satisfactory evidence that this orchid, which is remark- ably distinctive in structure, is pollinated by an unidentified wasp. Whether this is the sole agent is, of course, uncertain. Mrs. Edith Coleman“ recently reported that on two occasions she had seen this particular orchid visited by a small bee resembling Halictus, sp. No pollen was removed on either occasion, but the movements of the insect led her to suspect that it might be an active agent in the fertilization of the plant. Early in August, of the present year, Mr. Harold Goldsack, of Coromandel Valley, informed me that he had seen a small native bee crawling on the labellum of Caladenia deformis, an orchid well represcnted in this locality. As the bee had several pollinia adhering to ils back, he collected it, as well as the plant, and took them home for further observation. The following morning, on opening the box in which it was confined, he found the insect wedged tightly in the tubular space formed between the column and labellum of the orchid. When disturbed it hurriedly backed out on to the free extremity of the lip. During this move- ment the pollinia were rubbed hard against the stigma, but no further masses were removed from the anther. Later it returned to “its repast,’ and it was possible to follow its subsequent movements more carefully. It was able, without any exertion, to penetrate the space referred to, until a point was reached when the mesothorax was on a level with the stigma. Then, in its effort to reach the calli at the base oi the labellum, it pushed the latter away from the column, at the same time exerting strong pressure with its back against this structure and incidentally pressing the polien on to the stigma. This moyement enabled it to lever the labellum further outwards, and so penetrate a little lower into the tube. The calli were then apparently seized with its fore-legs and the tongue of the insect was protruded, but further observations were, unfortunately, prevented owing to the structure of the flower. The bee was dead when the specimens reached me by post, the wings were raised and more or less parallel to cach other. On the dorsum of the thorax, lying nearly transversely across the insertions of the wings, were two pollinia, and probably a portion of a third. Examination of the anther of the flower showed that the cells were intact and still contained their normal complement of pollen- masses, two in cach cell. The stigma was thoroughly dusted with pollen, and an almost entire pollinium was adherent to it. The condition of the anther made it manifest that the masses attached to the insect were derived from another flower. Towards the close of August we made an inspection of many of these plants in the National Park, The day was dull and cloudy, and although not very cold was unfavourable for our investigation. I discovered a specimen of the bee in ques- tion crawling up the ovary of a flower. The flower was collected and the sub- sequent movements of the insect watched. As in the case of Mr, Goldsack’s bee, it carried two pollen lamellae attached exactly in the same manner to the dorsum of its thorax. It crawled on to the recurved tip of the labellum and then on to the dense, shaggy calli, where it found a firm footing. It purposively pursued its way into the tube formed by the column and labellum, with its back to the former. It was able to enter this space without coming into contact with the anther or @ “Vict. Nat.” xlvi., 1930, p. 204. 145 pollinia, and penetrated the space deeply until only the tip of its abdomen was visible at the orifice. It then exerted hard and persistent pressure with its feet against the labellum, as though endeavouring to push the latter away from it. The results of this pressure were plainly visible to us in the movements of the labellum causing a widening of the “space” and pressure of the thorax of the insect against the stigma. It was not possible to observe exactly what took place at the base of the column. After a short interval the insect emerged by backing out, and was captured. It still had a couple of pollen lamellae attached to its thorax, but apparently not the same ones, as two masses had been removed from the anther of the flower and pollinia had been successfully deposited on the stigma. A little later in the day my wife discovered another bee with pollinia adherent as in previous cases. Unfortunately, a capture was not effected in this instance. On September 2 my wife and I again visited the Park. The day was again unsuitable for our quest, rather cold and rain threatening, Only a single Halictus was seen, towards the close of the afternoon. Lamellae were adherent in the usual manner. As the light was beginning to fade, the insect was captured and placed in a container w:ith the flower on which it was found. The following morning it was found within the flower. It liberated itself as previous bees had done, but emerged without pollinia on its body. Again the flower was pollinated, but the contents of the anther were intact. Possibly the flower was no longer fresh and, in consequence, removal of the masses had not been accomplished. Another flower, stripped of its perianth, was introduced into the bottle, but a successful entry was not made. Caladenia deformis is a very common orchid and is distributed throughout the entire breadth of Southern Australia. It also extends north along the east coast as far as latitude 33°, It this State it is the first Caladenia to appear during the season. A few isolated specimens may be seen during the last few days of July, but it becomes fairly prolific during the early part of August. The flowers are solitary, deep blue in colour, but occasionally an albino may be seen. The segments of the perianth are spreading, with the exception of the dorsal sepal which is usually more or less hoodlike behind the column. The labellum is not attached by the usual mobile claw, but is rather rigidly erect against the column, the sides of which it clasps, thus forming a tube; the apical part is triangular with fimbriated or dentate margins, and recurved so that its tip is in contact with the anterior surface of the vertical portion, thus facilitating access of the pollinating agent to the tube. Except for a small nude glandular area at the extreme tip, the recurved portion is covered with dense, shaggy calli arranged in six rows; behind these and within the mouth of the tube the calli are smaller, less crowded and reduced to four rows, while towards the base of the lamina there is an area of densely packed, large, colourless, glandular-looking calli covered with stellate hairs. When the labellum is in its normal position these large calli are in close contact with the anterior surface of the lower half of the column, thus preventing further ingress of an insect, unless force is exerted by the latter to displace the labellum outwards. These basal calli are apparently the tissues sought by the bee, and it is during its efforts to reach them that its thorax is pressed against the stigma and pollination is thus effected. The column is about 1-3 cm. long, erect in its lower part and rather abruptly incutved near the apex. It is winged throughout, the wings being much wider on either side of the stigma, thus limiting lateral movement of the agent and ensuring direct contact with the stigmatic surface. It is blue or purplish in tint with darker dots or transverse markings, especially on the wings. 146 The anther is attached almost horizontally to the summit, its mucrone being rather long and acute. It is bilocular, each cell being subdivided by a vertical dissepiment and containing a pair of free pollinia. Each pollinium consists of a flat, rather bluntly falcate lamella composed of dry, mealy pollen. {t is wider at the base than at the apex, and has a convex anterior margin. As the anther matures, its valves slightly retract by curling backwards, thus exposing the convex margins of the pollinia a little above the rudimentary rostellum, the surface of which is very viscid. The pollinia are devoid of caudicles and viscid disc. Immediately below the anther is the funnel-shaped stigma, which like the rostellum is extremely viscid. The pollinia are easily withdrawn when any of this sticky material is brought in contact with their exposed edges. No nectar is secreted by the flower. The base of the labellum and its adjacent parts are quite dry, and no injury to any of these parts can be detected after the act of pollination. The actual source of attraction to the insect is not apparent. It is noticeable, however, that a positive glucose reaction is yielded by the column and to a lesser extent by the labellum. The bee was subsequently identified as Halictus subinclinans, Ckll., by Mr. Tarlton Rayment, the well-known expert in this group. It is smaller than a house-fly and is admirably adapted, both in shape and size, to perform the service which it renders to this particular Caladenia. In his letter to our Museum staff, Mr. Rayment briefly outlines its life-history, which is of extreme interest and forms one of the romances of science. Inter alia he states that the first females emerge early in August, but he has never been able to secure any males. He has reason to suppose that the males only appear in the late autumn. Their homes are in well-drained sandy banks, the shafts have a diameter of two or three millimetres and go down almost vertically. PELECYPODA FROM THE ABATTOIRS BORE, INCLUDING TWELVE NEW SPECIES BY NELLY HOOPER WoobDs, M.A. Summary The shells of which those cited in this paper farm a part were collected by Sir Joseph Verco after the sinking of the Abattoirs Bore in 1919. Unfortunately, the material had been heaped beside the bore before any opportunity could be gained of ascertaining the depth from which the various fossils were obtained. It is impossible, therefore, to assign new species to any definite horizon; one can only remark that the fossils were taken from a depth between 400 and 500 feet, and are probably of Janjukian to Werrikooian age. 147 PELECYPODA FROM THE ABATTOIRS BORE, INCLUDING TWELVE NEW SPECIES. By Netty Hooper Woops, M.A. [Read October 8, 1931.] Puates VII. ann VIII. INTRODUCTION. The shells of which those cited in this paper form a part were collected by Sir Joseph Verco after the sinking of the Abattoirs Bore in 1919, Unfortunately, the material had been heaped beside the bore before any opportunity could be gained of ascertaining the depth from which the various fossils were obtained. It is impossible, therefore, to assign new species to any definite horizon; one can only remark that the fossils were taken from a depth between 400 and 500 feet, and are probably of Janjukian to Werrikooian age. It has been interesting to compare many of the fossil species with the recent shells, and to observe in some cases gradations between the fossil and the recent. This is so in the case of Nuculana crebrecostata T. Woods and Nuculana vercoms Verco; specimens intermediate between the two were found. In the case of Limopsis beaumariensis Chapman, the juvenile shells bear a marked resemblance to Limopsis eucosmus Verco. Many thanks are due to Mr. F. Chapman for his assistance with difficult and new species. Nucula venusta, n. sp. (PI. vii. figs. 1 and 2.) Solid, ventricose, inequilateral, ovate; umbo very prominent, inclined markedly to posterior ; posterior margin short, curved evenly from dorsal to ventral border; anterior margin longer than posterior, curving evenly though slightly more sharply from dorsal to ventral border. Interior of shell smooth, nacreous, ventral margin flattened, without denti- culations; surface smooth, shining, with fine concentric growth-lines of varying strength. Cardinal line with 17 teeth anteriorly, slightly uncinate, and 6 posteriorly; teeth strong, high. Length, 5-6 mm.; height, 4°8 mm. Observations —Unfortunately there are only two right valves of this beautiful little shell, which, though resembling it in some respects, differs markedly from N, obliqua Lamarck. The shell is much more tumid, especially in the umbonal region ; it is more produced anteriorly, and has fewer teeth. The junction between the anterior and posterior row is worn in both specimens, making it impossible to tell the nature of the resilifer. Rochefortia macer, n. sp. (PI. vii. fig. 3.) Thin, white, medium, somewhat flattened, inequilateral, posterior side longer than anterior and more sharply produced; anterior margin roundly curving to ventral border; interior of shell smooth, ventral border without denticulations; surface smooth, shining, with fine concentric striae of varying prominence, Hinge 148 with one strong cardinal tooth inclined posteriorly and one small, more depressed tooth inclined anteriorly. Length, 11:1 mm.; height, 9-3 mm. Rochefortia tellinoides, n. sp. (PI, vii, fig. 4.) Small, thin, moderately convex, inequilateral; posterior margin slightly more sharply curving than anterior; umbones small, situated anteriorly. Hinge line without dentition, consisting of two small plates on either side of the umbo, leav- ing a space beneath the umbo. Shell longer than high; interior smooth; adductor impressions and pallial line distinct ; surface ornamented with fine concentric striae. Length, 5°7 mm.; height, 3-6 mm. Dosinia grandis, n. sp. (Pl. vii, figs. 5 and 6.) Large, thick, solid, several thick layers being revealed when the shell is broken; area inside pallial line thicker than that outside (this may possibly be due to weathering of specimens) ; outline of shell indefinite, as the shells have been broken in taking them from the bore; hinge plate very heavy, bearing in the right valve two strong, high cardinal teeth and a deep depression for the insertion of a large tooth of the left valve, and from the umbd to the posterior edge a sub- triangular depressed area followed by a triangular area for the ligament and a deep and narrow sulcus; lunule deeply situated; adductor impressions deep and clear. Sculpture consists of numerous thin, fine concentric striae. Measurements cannot be accurately determined owing to fragmentary nature of specimens, but approximate to :— Length, 70 mm.; height, 70 mm. Gafrarium perornatum, n. sp. (Pl. vii. figs. 7 and 8.) Medium to small, solid, ovate, longer than high, posterior side produced somewhat; umbones high, acute, situated in front of the middle axis of the shell; both posterior and anterior borders roundly curving to the ventral margin. Interior smooth, pallial sinus distinct. Surface of shell ornamented with numerous regular radial striae which, on certain umbo-ventral lines, are crossed by short oblique striae making V-shaped patterns, pointing both ventrally and dorsally. Also occasional and varying concentric striae of growth. Type: Length, 9°6 mm.; height, 7°5 mm. Larger specimen: Length, 13-5 mm.; height, 10°8 mm. Antigona pernitida, n. sp. (Pl, vit, figs. 1 and 2.) Small to medium, sub-ovate, lengthened anterior-posteriorly; umbones prominent and acute, situated to the anterior side of the central axis; anterior margin roundly curving to the ventral border; posterior margin longer than anterior, curving sharply to ventral edge. Shell moderately inflated, interior finely crenulate; surface closely ornamented with regular growth lines, the interspaces of which are crossed by fine radial costae. The type is a small, neat shell which seems to differ from A. dennantt Chap. and Crespin in the greater number and prominence of its concentric growth lines, and the greater length of the valve in relation to its height. 149 A fragment of a larger shell of the same species also came up from the bore, dimensions being about twice those of the type. Length, 12°3 mm. ; height, 9-4 mm. Pseudoarcopagia detrita, n. sp. (PL. vii., fig. 9.) Small, solid, trigonal, moderately convex, particularly in the umbonal area: anterior side slightly longer and more rounded than posterior. Interior of shell smooth, ventral border without crenulations; pallial line indistinct owing to the weathering of specimens. In right valve two cardinal teeth, and one lamellar tooth on anterior side—deep socket on posterior; in left valve two cardinal teeth fitting into corresponding sockets of right valve; one large posterior lateral tooth and two anterior laterals fitting into sockets of right valve. Exterior of shell sculptured with numerous fine radial striae bifurcating towards the ventral border. Length, 4°8 mm.; height, 4:2 mm. Diplodonta solitaria, n. sp. (PI. viii, fig. 3.) Holotype—one rather worn specimen only of left valve. Orbicular, subequilateral, moderately convex; ligament groove long, narrow ; umbo subcentral, slightly incurved ; lunule lanceolate, slightly sunken. Valve with three cardinal teeth, one bifid; pallial line punctate marked; ventral margin rounded, smooth. Sculpture—fine concentric lines of growth with occasional broader lines. Length, 22°8 mm. ; height, 21:7 mm. Codakia salebrosa, n. sp. (PI, viii, figs. 4 and 5.) Very thick, rude, interior inside pallial line covered frequently with thick concretions ; inequilateral, equivalve, sharply curving on anterior margin, semi- circular on posterior side. Ligament pit internal, long, deepening from umbo to top of posterior border ; umbones acuminate, teeth embryonic or obsolete. Ventral margin without crenulations. Shell surface very rough with numerous concentric growth lines of varying prominence, crowded near the border. Length, 27-5 mm.; height, 26°7 mm. Cryptodon sinuatum, n. sp. (Pl. viii, Fig. 6.) Medium to small, thin, triangular-ovate: inequilateral, very swollen, par- ticularly in the dorsal region; umbones prominent, high, placed a little in front of the central axis. Anterior border curved, lower half coming almost at right angles to the ventral margin, posterior part of shell with deep told. Interior of valve smooth; ventral margin sharp, without crenulations; surface of shell with fine concentric striae of growth in varying prominence. Holotype: One left valve only. Length, 8-1 mm.; height, 8:2 mm. Solecurtus subrectangularis, n. sp. (Pl. viii, Fig. 7.) Small, thin, inequilateral, posterior side longer and broader than anterior, oblong, very slightly gaping at both ends; posterior side about 2 of the whole length; posterior dorsal line straight, nearly parallel with ventral margin; anterior dorsal margin inclined to horizontal; anterior margin more rounded than 156 posterior margin, which is nearly vertical, Umbo small. One strong cardinal tooth beneath the umbo; two lateral sockets in front and behind. Pallial line and adductor impressions indistinct. Surface of shell rudely sculptured with concentric striae of growth crossed by fine bifurcating radial striae ; umbonal area showing buff colour, ventral margin white. Holotype: Left valve. Length, 7°7 mm.; height, 4°6 mm. Corbula equivalvis, n. sp. (Pl. viii., figs. 8 and 9.) Solid, equivalve, inequilateral, ventricose, ovately-triangular, rounded anteriorly, beaked posteriorly, posterior side longer than anterior. Umbones prominent, incurved, especially in right valve; right valve with sharp teeth situated in anterior; left valve with large flattened tooth in posterior side of shell. Surface of shell sculptured with many fine concentric striae, varying in thickness and often discontinuous, particularly in centre of shell, where in one senile ‘specimen a slight furrow is produced; posterior side carinated from umbo to ventral margin; sculpture behind the carina same as rest of shell. and adductor impressions distinct. Pallial sinus Type: Length, 14:2 mm.; height, 8-4 mm, Larger specimen: Length, 16-8 mm.; height, 9-3 mm. LIST OF PELECYPODS OBTAINED FROM THE BORE. Order PRIONODESMACEA. Family NUCULIDAE. Nucula obliqua Lamarck. Nucula morundiana Tate. Nucula venusta Hooper Woeds. Family NUCULANIDAE, Nuculana woodsii Tenison Woods, Nuculana crebrecostata T. Woeds. Nuculana verconis Verco. Family PARALLELODONTIDAE. Cucullaea corioensis McCoy. Family LIMOPSIDAE. Limopsis beaumaricnsis Chapman. Limopsis maccoyi Chapman. Limopsis affinitalis Chapman. Family ARCIDAE. Lissarca rubricata Tate. Arca navicularis Tate. Arca (Barbatia) pistachia Lamarck. Glycimeris convexa Tate. Glycimeris tenuicostata Reeve. Family PTERIIDAE. Pinctada crassicardia Tate. Family OSTREIDAE. Ostrea hyotidoidea Tate. Family TRIGONIIDAE. Neotrigonia acuticostata McCoy. Family PECTINIDAE. Pecten consobrinus Tate. Chlamys peroni Tate. Chlamys antiaustralis Tate. Amusium hochstetteri Zittel. Hinnites corioensis McCoy. Family SPONDYLIDAE. Spondylus arenicola Tate. Family LIMIDAE. Austrolima bassi T. Woods (Lima bassi) Family ANOMIIDAE. Monia ione Grey. Family MYTILIDAE. Trichomya hirsuta Lamarck (Brachyodon- tes hirsutus). Family THRACIIDAE. Thraciopsis clongata May. Family MYOCHAMIDAE. Myodora ovata Reeve. Myodora tenuilirata Tate. Myodora corrugata Tate. Family CLAVELLIDAE. Humphreyia strangei Adams and Angas. Family CUSPIDARIIDAE. Cuspidaria subrostrata Tate. Order TELEODESMACEA. Family CRASSITELLITIDAE. Crassitellites oblonga T. Woods, Cuna polita Tate. Family CARDITIDAE. Cardita compta Tate. Cardita preissi Minke. Venericardia spinulosa Tate, Venericardia pecten Tate. Venericardia subcompacta Chapman and Crespin. Family CHAMIDAE. Chama lamellifera T. Woods. Family LUCINIDAE. Lucina leucomorpha Tate. Lucina affinis Tate. Lucina projecta Tate. Lucina nuciformis Tate. Lucina fabuloides Tate. Loripes icterica Reeve. Codakia salebrosa H. Woods. Divaricella quadrisulcata D’Orbigny. Family DIPLODONTIDATE. Diplodonta solitaria H. Woods. Family CRY PTODONTIDAE. Cryptodon sinuatum H. Woods, Family LEPTONIDAE. Lepton trigonale Tate. Lepton crassum Tate. Erycina micans Tate, Montacuta sericia Tate. Rochefortia anomala Angas. Rochefortia donaciformis Angas. Rochefortia ovalis Tate. Rochefortia macer H. Woods. Rochefortia tellinoides H. Woeds. EXPLANATION Prate VII. Fig. 1. Nucula venusta, n. sp. X8. Fig. 2. N. venusta, n. sp. X8. Fig. 3. Rochefortia macer, n. sp. X4. Fig. 4. R. tellinoides, n. sp. X8. Fig. 5 # natural size. Fig. 6. D. grandis (right valve), n. sp. Nearly natural size, Fig. 7. Gafrarium perornatum, n. sp. X5. Fig. 8. G. perornatum, n. sp. X5. Fig. 9. Pseudoarcopagia detrita, n. sp. X9. - Dosinia grandis (left valve), n. sp. Family GALEOMMIDAE Sportella jubata Hedley. Family CARDIIDAE., Cardium cygnorum Deshayes. Cardium hemimeris Tate. Family VENERIDAE. Dosinia johnstoni Tate. Dosinia grandis H. Woods. Gafrarium perornatum H. Woods. Macrocallista submultistriata Tate. Antigona propinqua Tate. Antigona striatissima Tate. Antigona dimorphylla Tate. . Antigona pernitida H. Woods. Clausinella subroborata Tate. Bassina allporti T, Woods. Callanaitis cainozoica T. Woods. Paphia fabagelloides Tate. Family TELLINIDAE. Psammobia hamiltonensis Tate. Psammobia equalis Tate. Tellina masoni Tate. Pseudoarcopagia detrita H. Woods. Family SEMELIDAE. Semele vesiculosa Tate. Family SOLENIDAE. Solecurtus dennanti Tate. Solecurtus subrectangularis H, Woods. Family MACTRIDAE. Mactra howchiniana Tate. Zenatiopsis angustata Tate. Family CORBULIDAE. Corbula ephamilla Tate. Corbula equivalvis H. Woods. Family SAXICAVIDAE, Saxicava australis Lamarck. Saxicava subalata Gatliff and Gabriel. OF PLATES. PiLate VIII. Fig. 1. Antigona pernitida, n. sp. X4. Fig. 2. A. pernitida, n. sp. X4, Fig. 3. Diplodonta solitaria, mn. sp. X2. Fig. 4. Codakia salebrosa, n. sp. X°/z. Fig. 5. C. salebrosa, n. sp. X°/s, Fig. 6. Cryptodon sinuatum, n. sp. X5. Fig. 7 X7, Fig 8. Corbula equivalvis, n. sp. X3. Fig. 9. C. equivalvis, n. sp. X3. . Solecurtus subrectangularis, n. sp. AUSTRALIAN FUNGI: NOTES AND DESCRIPTION - NO.8 BY J. BURTON CLELAND, M.D. Summary The last paper of this series, No. 7, appeared in these Transactions and Proceedings, vol. li1., 1928, pp. 217-222. In the present paper, a number of new species of Agarics and Clavarias are described. Colour tints when specifically noted in capital letters are based, unless otherwise stated, on Ridgway's "Colour Standards and Colour Nomenclature," 1912 edition, references to the plates therein being given. 152 AUSTRALIAN FUNGI: NOTES AND DESCRIPTIONS.—No. 8. By J. Burton CLeLranp, M.D. [Read October 8, 1931,] The last paper of this series, No. 7, appeared in these Transactions and Pro- ceedings, vol. lii., 1928, pp. 217-222. In the present paper, a number of new species of Agarics and Clavarias are described. Colour tints when specifically noted in capital letters are based, unless otherwise stated, on Ridgway’s “Colour Standards and Colour Nomenclature,” 1912 edition, references to the plates therein being given. Lam much indebted to Miss E. M. Wakefield, of the Royal Botanic Gardens, Kew, England, for kindly examining a number of specimens and water-colour drawings of Australian Clavarias forwarded to the Director, and for comparing these critically with the world-wide collections there and for expressitig opinions on these species. Extracts from her report are appended to the descriptions of the species concerned. 499, Amanita subalbida, n. sp—Pileus 1% in. (4°4 cm.) in diameter, irre- gularly convex, then nearly plane with the centre somewhat depressed, mealy with the remains of the universal veil, pallid brownish to nearly white. Gills just adnexed, close, rather narrow, white. Stem 1 in. (2°5 cm.) high, elongating from the bulbous base to 13 ins. (4:4 em.), rather short, $ in, (10 mm.) thick, submealy, nearly equal when expanded, solid, white. Ring superior, when young well marked, membranous, marked above with gill-lines, tending to disappear. Volva disappearing, as a slight friable rim-like edge to the bulb. Spores white, obliquely elliptical, 9-5 > 5:5 ». Half-buried in sandy soil. S.A.—Kinchina, June 8, 1925. 500. Amanita conico-bulbosa, n. sp—When young 2 ins. (5 cm.) in diameter, plano-convex with a deep rounded border and edge turned in, slightly viscid when moist, finely villose, greyish white; base of the stem very bulbous (13 in., 3-7 cm. thick), the root conical and 24 ins. (5-6 cm.) deep. When adult, pileus 2 to 4 ins. (5 to 10cm.) in diameter, slightly convex to a little upturned or convex with the centre depressed, slightly viscid when moist, in places smooth and shin- ing, in others subvillose, with scattered warty patches often villose at the base, or the whole surface dull with no flakes, cuticle peels, white with a slight biscuity tint or chalky white, occasionally with a greyish-brown tint. Gills just reaching the stem, attenuated towards it, close, } to 4 in. (6 to 12-5 mm.) deep, ventricose, short ones at the periphery, creamy white, when old with a buffy tint in some lights, when dry brownish. Stem up to 3 to 5 ins. (7°5 to 12°5 cm.) high, ¢ in. (21 cm.) thick, slightly attenuated upwards, gill-marked above, fibrillose-scaly to matt below, solid, white or whitish, bulb 14 to 14 in. (3-1 to 3:7 cm.) thick, root up to 3 ins. (7-5 cm.) long, conical downwards. Ring superior to median, ample, dependant marked with gill-lines above, evanescent. No obvious volva, or volva as a mealy-evanescent rim when young. Flesh white, moderately thick (4 in., 6 mm., or more), attenuated outwards. Smell sometimes slightly fragrant, when cut somewhat phosphorus-like. Spores elliptical, hyaline, 9 to 11-5 %& 5-5 to 7p». Partly buried in sand or in the ground. S.A—Kinchina, Beaumont, June, July, August, October. This species is characterized by being greyish-white when young, later white with a slight biscuity or greyish-brown tint, and by having, usually, scattered villose warty patches, no strong smell and a very long conical root. 153 501. Amanitopsis sublutea, n. sp—Pileus 14 to 13 in. (3:7 to 4:3 cm.) broad, convex, becoming depressed a little in the centre, sticky when moist, pale buff [a little deeper than Warm Buff (xv.) ]. Gills just reaching the stem, moderately close, becoming slightly ventricose, white. Stem up to 12 in. (4°3 cm.) high, in. (10 mm.) thick, equal, mealy above, smooth below, white or a little buff-tinted below, bulb up to # in. (19 mm.) thick, spherical, the colour of the pileus, edge just free. Flesh of pileus rather thin, white, attenuated out- wards. Spores oblique, 13 to 13°5 & 7 uw. In sand. S.A—Encounter Bay, August, 1929, 502. Lepiota fuliginosa, n. sp—Pileus up to in. (2°2 cm.) in diameter, slightly convex, then plane or a little upturned, in the centre sooty-brown from minute fibrous scales, almost villose, the scales scantier and paler round the periphery, leaving a pallid, slightly sooty zone 4 in. (3 mm.) or more in diameter. Gills barely free, narrow, close, creamy white. Stem 4 to 4 in. (3 to 19 mm.) high, short, equal, smooth, solid, pallid whitish. Ring distant, as a narrow membranaceous ring, evanescent. Spores oblique with an apiculus, 5°5 X 4°5 p. On the ground. S.A.—Kinchina, June 8, 1925. ; 503. Lepiota nigro-cinerea, n. sp.—Pileus 4 in, (1'2 cm.) in diameter, convex, subumbonate, dark grey from floccose scales, Gills barely reaching the stem, moderately close, cream-coloured. Stem § in. (16 mm.) high, slender, a little fibrillose below, whitish. Ring (?) evanescent. Spores very oblique, sometimes nearly triangular, not thick-walled, 5°5 x 3°5 mw. On the ground. S.A— Encounter Bay, May 24, 1928. 504. Lepiota cinnamonea, n. sp.—Pileus 4 to 1 in. (1:2 to 2°5 cm.) in diameter, at first almost campanulato-convex, then expanding to convex, often more or less broadly umbonate, slightly floccose to flecked with minute scales, Light Pinkish Cinnamon (xxix.) to Cinnamon or near Tawny (xv.), sometimes when dry near Apricot Buff (xiv.), Gills just free or barely reaching the stem, rather close to moderately distant, rather narrow, ventricose, cream. Stem 1 in. (2:5 cm.) high, rarely 14 in. (3°7 cm.), rather slender (5 mm. thick), slightly attenuated upwards, fibrillose to fibrillose-scaly up to the veil attachment which is superior, stuffed or slightly hollow, paler than the pileus to near Tawny, sometimes Cinnamon Rufous (xiv.). No definite ring. Flesh thin, whitish, in the stem with a cinnamon tint and white in the centre. Spores elliptical, slightly oblique, not thick-walled, 5-5 to 7-5 & 3:7 to 4. S,A.—On the ground in a glade in stringy-bark forest, National Park; in Pinus radiata Don. (P. insignis Douglas) forest, Mount Burr (S.E.), May, 1931, 505. Lepiota subcristata, n. sp—Pileus 1 to 14 in. (2°5 to 3:7 cm.), at first conico-convex, then convex to nearly plane, with an obtuse umbo, densely covered with small brown fibrillose scales, darker and closer at the disc. Gills free, moderately close, white. Stem to 14 in, (4°4 cm.) high, rather slender, stuffed tending to be hollow, shaggy with fibrils up to the veil attachment, smooth above. No definite ring. Spores elliptical, not thick-walled, 5-5 to 3-7 ». S.A—In Pinus radiata Don. (P. insignis Douglas) forest, Mount Burr, May, 1931. Resembles L. cristata (A. and S.) Fr., but differs in the shaggy fibrillose stem without a definite ring. L. cristata grew in the same locality. 506. Lepiota discolorata, n. sp—Pileus 14 to 2 ins. (3 to 5 em.) in diameter, nearly plane, a little upturned, subumbonate, covered with dense very dark reddish-brown scales, fewer near the periphery, sometimes slightly striate at the periphery. Gills just free, close, white or cream-coloured. Stem 14 to 2} ins. (3:1 to 6°2 cm.) high, slender, hollow, white above, pale brownish below or pallid with minute brownish flecks. Ring distant, membranaceous, evanescent. Flesh white stained reddish. Smell strong, radishy. The whole plant when dry Fuscous to Fuscous Black (xlvi). Spores elliptical, oblique, not thick-walled, 5 to 6°5 X F 154 3°5 p, sometimes 7°35 to9 X 4°54. Onthe ground, S.A.—Mount Lofty; National Park; in Pinus radiata Don. (P. insignis Douglas) forest, Kalangadoo (S.E.). April to June. A moderately small species, recognised by the dark reddish-brown scales in the pileus and the discolouration of the whole plant on drying. 507. Lepiota haemorrhagica, n. sp—Pileus 4 to 14 in. (1°8 toa 3-1 cm.) in diameter, convex, sometimes irregular, covered with reddish-brown fibrillose scales thicker and darker at the disc. Gills free, close, creamy-coloured turning reddish like a fresh bloodstain when bruised. Stem 2 ins, (5 cm.) high, rela- tively rather stout, attenuated upwards, slightly hollow, clothed with reddish- brown fibrils even above the distant definite membranous pale to reddish ring. Spores elliptical, slightly oblique, not thick-walled, microscopically slightly tinted, 6 X 3:5 pw. On the ground in Eucalyptus forest. S.A—Mount Burr (S.E.), May, 1931. 508. Lepiota umbonata, n. sp—Pileus 4 in. (1:2 cm.) in diameter, slightly convex with a broad obtuse umbo, pallid whitish with a buffy tint. Gills just free, moderately close, pallid flesh-coloured. Stem 14 in, (3:1 cm.) high, slender, flesh-coloured. Ring distant. Whole plant when drying brownish. Spores not thick-walled, 5-5 & 3:5 ». S.A.—In Pinus radiata Don. (P. insignis Douglas)! forest, Kalangadoo, May. A small species with a whitish umbonate pileus and slender moderately long stem. , 509. Lepiota albo-fibrillosa, n. sp—Pure white. Pileus ¢ in. (3-5 mm.) in diameter, convex, subumbonate, mealy, dotted with white fibrils continuous with the veil and clothing the stem below the attachment of the veil, no definite ringy Gills free, many short, edges rather thick, white. Stem % in. (10 mm.) high, slender, base a little swollen. Spores elliptical, 6:2 to 7-5 & 3-75 ». On the ground, under a rock. 5.A—Mount Lofty, May. A minute white delicate short-stemmed species with white fibrils on the pileus and clothing the stem without a well-defined ring. 510. Lepiota bulbosa, n. sp—Pileus 1 in. (2:5 cm.), convex, pale earthy brown with scattered villose scales. Gills barely reaching the stem, close, slightly ventricose, creamy white. Stem 3 in. (18 mm.) with the bulb, slender, under Zin. (6 mm.) thick, bulb 4 in. (12°5 mm.) thick, white and striate from the gills above the median fixed definite membranous ring, slightly fibrillose and whitish with a brownish tint below, slightly hollow. Spores elliptical, oblique, not thick- walled, 9°3 to 10°5 & 5:5 to 7 w. On the ground. S.A—Inman Valley, Sep- tember 5, 1925. S11, Clitocybe straminea, n. sp—Pileus 1 to 14 ins, (2°5 to 3-7 cm.) in diameter, irregularly convex, centre usually depressed, thin, nearly semi-trans- lucent, pilose in the centre, fibrillose peripherally, slightly striate, edge radiately splitting, centre blackish-brown, the remainder smoky yellowish-brown, the smokiness due to fine fibrils. Gills slightly but definitely decurrent, moderately close, straw-coloured to pale egg-ycllowish. Stem 14 to 2 ins. (3-7 to 5 em.) high, equal, slender, somewhat flexuous, twisted, slightly striate, mealy fibrillose above, less so below, hollow, the colour of the gills. Spores subspherical, 4 to 5 p. Densely caespitose at the base of stumps. S.A.—Mount Lofty, March, April. The specific name has reference to the straw colour of the gills and stem. 912. Clitocybe eucalyptorum, n. sp—Pileus 6 ins. (15 cm.) or more in diameter, irregularly convex with the edge turned in when young, then expanding, the centre finally more or less depressed, repand, innately fibrillose to sub- tomentose with occasionally small circular patches of thickened cuticle, the edge slightly sulcate, Drab (xlvi.) when young to browner than Tawny Olive (xxix.). Gills moderately decurrent, moderately close, up to $ in. (10 mm.) deep, attenuated at the periphery, cream-coloured, assuming a slight fleshy tint, becoming yellowish 155 round the edge when old. Stem 4 ins. (10 cm.) long, stout, up to Lin. (2:5 cm.) thick, swollen below when young, marked above with lines of the gills, sub- fibrillose below, pallid with tints as on the pileus, with white mycelium mixed with earth at the base. Shed spores subspherical, pear-shaped, slightly irregular, hyaline, 5°5 to 6°5 X 4:5 ». On the ground amongst leaves, etc., under Euca- lypius. S,A—National Park, July. 513. Clitocybe campestris, n. sp—Pileus up to 1 in. (2-5 cm.) in diameter, slightly convex, irregular with a depressed centre, slightly shiny, the edge turned in when young, pallid stone colour and slightly mottled, faintly obscured by a minute white pile (near Avellaneous, xl.; Light Buff, xv.). Gills adnate, close, rather thick, rarely forking or with buttresses, pallid brownish white (Avellane- ous, xl.; near Vinaceous Buff, xl.). Stem up to $ in. (1°8 cm.) high, stout, some- times flattened, slightly fibrous, tough, hollow, mealy, pallid, or the colour of the pileus. Flesh white. Smell strong. Spores 4:5 to 4°8 X 2°2 to 3-2 yw, . In grassy places, Beaumont Common, May, June; Eagle-on-the-Hill, June (Miss Fiveash, Watercolour No. 25); Noarlunga Hill (spores 5°5 X 3°7 ). A small species somewhat resembling small specimens of Hebeloma hiemale Bres., characterised by its pallid buff pileus with darker tints of avellaneous and wood brown appearing as if under the surface, the avellaneous gills, short stem and occurrence in grassy places. 514. Clitocybe pascua, n. sp—Pileus 1 to 1) in, (2°5 to 3:7 cm.), rarely 2 ins. (5 cm.) in diameter, irregularly convex, soon becoming depressed in the centre and sometimes infundibuliform, edge often irregular and wavy or slightly lobed, sometimes lacerated, smooth, when moist between Sudan Brown and Brussels Brown (iii.) and semi-translucent, when dry opaque whitish or buffy. whitish. Gills slightly decurrent, rather close, moderately narrow, many short, greyer than Pinkish Buff (xxix.). Stem short, $ to 1 in. (1:2 to 2°5 cm.) high, slender, equal or sometimes attenuated downwards, fibrillose, hollow, brownish when moist, pallid when dry. Flesh watery brownish when moist, whitish when dry. Smell a little strong. Spores obliquely elliptical, 7 & 3°7 p. Gregarious on grassy hills. S.A—Near Noarlunga, June 25, 1927. 515. Clitocybe australiana, n. sp—Pileus up to 14 to 4 ins. (3°1 to 10 cm.) in diameter, irregular, somewhat convex, centre depressed, edge rather irregular and broken up, dull, smooth, pale biscuit colour (near Pinkish Buff, xxix.), paler than Mikado Brown (xxix.) and near Vinaceous Cinnamon (xxix.), soapy- looking when moist, near Sayal Brown (xxix.) when dry, Gills adnato-decurrent to decurrent, narrow, moderately close, near Pinkish Buff. Stem up to 14 ins. (3:7 cm.) high, slender to stout, up to § in. (15 mm.) thick, slightly attenuated downwards, dull surface, solid or slightly hollow, with fluffy mycelium at the base, white, Flesh white, thick over the stem, attenuated outwards, Spores 3-2 to 5-6 X 1°6 to 3'2 pw. Single or two or three together or subcacspitose in sandy soil under trees. 5.4.—Kinchina, Monarto South, and Enfield. N.S.W.— Bumberry and Manildra, July, August, September, October, 516. Collybia subdryophila, n, sp—Pileus up to 14 in. (3*1 cm.), slightly con- vex, sometimes eventually a little upturned at the edge, irregular, matt, near Pinkish Buff (xxix.), Gills adnate to adnexed (once apparently sinuate), close, narrow, creamy white. Stem up to 14 in. (3'7 cm.) high, rather slender, some- times flattened, sometimes slightly attenuated upwards, smooth or matt, hollow, flesh confluent with but heterogeneous from that of the pileus, reddish-brown (near Verona Brown, xxix.). Shed spores with one end more pointed, 4 to 4:2 32, S.A.—Mount Lofty, July, 1921, and April, 1924 (spores 5:6 X& 3°75»); Mount Compass, October ; Kinchina, July (spores 3:2 X 2»); near Happy Valley, September ; National Park; Hope Valley. 156 517, Collybia deusta, n. sp—Pileus 2 to 3 ins. (5 to 7°5 cm.) in diameter, irregularly plane to slightly depressed with a trace of umbonation, edge somewhat undulatory, surface matt to subtomentose, smoky brownish to scorched brown. Gills adnato-adnexed with occasionally a decurrent tooth, close, narrow (4 in. +, 6°5 mm. deep), pallid dingy greyish to pallid dingy buff. Stem 14 to 2 ins, (3°7 to 5 cm.), rather slender (4 to } in., 6 to 9 mm., thick), fibrillose, tough, solid, base slightly swollen into a knob ending abruptly, dark smoky brown. Flesh of stem cartilaginous differing in texture from the flesh of the pileus, which is white and thin. Spores elliptical, 8°5 & 5:2». No obvious smell. S.A.—In sand under Melaleuca halmaturorum IF. y. M., Inman River, Victor Harbour. May. 518. Collybia alutacea, n. sp.—Pileus 3 to 14 in. (1-8 to 3-7 cm.) in diameter, more or less plane becoming upturned-repand, sometimes subumbonate, smooth, rich salmony-buff and moist-looking, sometimes reddish-brown at the periphery, drying opaque matt and a paler pinkish-buff. Gills adnexed, narrow, close, creamy white. Stem # to 1 in. (1°8 to 2°5 cm.) high, rather short, somewhat slender, sometimes flattened, equal, smooth, slightly hollow, pallid with a slight or definite tint of the pileus. Flesh of the stem cartilaginous differing from the thin white flesh of the pileus which is attenuated outwards. On the ground. S.A.—Back Valley, off Inman Valley. May, 1929, Characterised by the rich salmony-buff pileus becoming pinkish-buff, con- trasting with the close white gills and short pallid stem slightly tinted like the pileus. 519. Collybia abutyracea, n. sp.—Pileus up to 43 ins. (11-8 em.) in diameter, at first convex with the edge turned in, then expanding, irregular and repand and more or less subumbonate, at first slightly velutinate, finely somewhat shining and subfibrillose, when young pallid or Cream Buff (xxx.) with a smoky brown tinge, then pallid biscuit-coloured, sometimes with a smoky or scorched tinge, sometimes with the umbo approaching Saccardo’s Umber (xxxix.). Gills slightly sinuate to adnate, close, rather dingy creamy white, becoming more biscuit-coloured. Stem up to 13 in. (3-7 cm.) high, rather slender to moderately stout, 4 to 4 in. (10 to 12-5 mm.) thick, coarsely fibrillose, equal, not rooting, tough and cartilaginous but with the flesh not very clearly distinct from that of the pileus, solid, not stuffed, breaking up into tough fibrils, dark smoky brown to pallid brownish, base whitish when young. Spores elliptical, 7*5 to9 X 5 to 5-5 w. No special smell. Amongst grass. S.A.—Beaumont Common, Pinnaroo, Belair, June, July, August. 520. Collybia eucalyptorum, n. sp—Pileus 3 to 14 in. (1-6 to 3-7 em.) in diameter, broadly conico-campanulate to nearly plane, then slightly upturned, smooth, with the surface dull from innate fibrils, edge slightly striate, Pale Pinkish Buff becoming Cinnamon Buff (xxix), or Ochraceous Buff (xv.) and darker in the centre, becoming pallid towards the periphery. Gills adnexed, close, narrow, with short ones at the periphery, creamy-white or approaching Warm Buff (xv.). Stem | to 25 ins. (2°5 to 6:2 cm.) high, relatively slender (4 in., 3-5 mm., or morc thick), flexuous, smooth or subfibrillose, barely striate, hollow, cartilaginous, differing from the flesh of the pileus, reddish-brown (between Tawny, xv., and Russet, xv.; Mikado Brown, xxix.). Flesh thin, slightly brownish. Smell moderately strong. Spores pear-shaped, hyaline, 5 to 5-5 & 3-5. Caespitose at the bases of old trunks of Eucalyptus or stumps. S.A—Mount Lofty Summit, June. 521. Mycena subgalericulata, n. sp—Pileus 4 to 1 in, (1-2 to 2:5 cm.) in diameter, 4 to fin. (0°8 to 1-8 cm.) high, conico-campanulate, somewhat expand- ing, umbonate, dry, smooth, submembranaceous, somewhat striate to the umbo, near Olive Brown (xl.), occasionally paler (Buffy Brown, xl.), sometimes Mummy Brown (xv.), during drying becoming paler from above from Olive Brown to Butty Brown, when young with a pallid peripheral ring. Gills adnate, sometimes 157 with a slight decurrent tooth, sometimes connected by veins, whitish, sometimes flesh-tinted or greyish when old. Stem 4} to 2 ins. (1°8 to 5 cm.) high, often curved, smooth, polished, somewhat fragile to rather tough, base somewhat strigose, whitish to pallid, sometimes brownish especially below, Shed spores elliptical, oblique, 9 to 13 X 5°5 to 8:5 p. No cystidia seen. No smell. Caespitose on trunks. S.A—Mount Lofty (on trunks of Eucalyptus obliqua L’Heérit.), National Park. April, June, July, August. This is evidently a variable species. It differs from Rea’s description of Mycena galericulata (Scop.) Fr. in being of smaller size, with the cap apparently darker, the gills sometimes becoming greyish when old, and in the stem often being nearly pure white. The characteristics of the species are the caespitose habit on trunks or stumps, the dark fuscous brown to pale smoky brown umbonate pileus, the gills adnate sometimes with a decurrent tooth and whitish becoming flesh-coloured or greyish, and in the whitish or pallid stem sometimes brownish below. 522. Mycena australiana, n. sp-——Pileus 4 in. (1:2 cm.) high, # in. (1:8 cm.) broad, broadly conico-campanulate, slightly striate, Buffy Brown to Clove Brown (xl,) or Wood Brown (xl.), apex darker. Gills adnate, with no decurrent tooth, moderately close, pure white becoming creamy. Stem about 3:7 cm. high, slender, polished, a little mealy at the base but without strigose hairs, apex whitish, Buffy Brown towards the base. Spores 8:5 to 11 X 6 to 7‘5 pw. Gre- garious of caespitose on fallen log. S.A—National Park, Mount Lofty. May, June, July. 523. Mycena vinacea, n. sp.—Pileus 3 to 14 in. (1°8 to 3-7 cm.) in diameter, conico-hemispherical or broadly conical to convex, then expanded, sometimes with an acute or obtuse umbo, matt or smooth, slightly shining, striate at the periphery when moist, edge slightly incurved when young, Pale Vinaceous Drab to Vinaceous Drab (xly.), Light Cinnamon Drab (xlvi.), near Sorghum Brown (xxxix.) or yellower than Vinaceous Brown (xl.), sometimes Fuscous (xlvi.) when old, drying to near Pinkish Buff (xxix.), paler than Avellaneous (xl.) or between Avellaneous and Olive Buff (xl.). Gills adnate or slightly sinuate with a decurrent tooth, moderately close, ventricose, many short, edges tending to be irayed, Pale Vinaceous Drab, Pale Brownish Drab (xlv.), Pale Greyish Vinaceous, or Vinaceous Fawn to Fawn Colour (xl.). Stem 1 to 23 ins. (2:5 to 6°8 cm.) high, slender to a little stout, equal or slightly attenuated upwards or downwards, smooth, hollow, base pallid and tending to be villose, Dark Vinaceous Drab (xlv.) when young, Light Greyish Vinaceous (xxxix.), near pale Brownish Drab or Wood Brown (xi.). The pallid brownish flesh of the cartilaginous stem hetero- geneous from the white flesh of the pileus. Spores obliquely elliptical, 7-5 to 13 & 4 to 8:5 ». Caespitose or subcaespitose on fallen wood on the ground, at the base of stumps, or amongst fallen leaves and grass or pine needles, S.A.— Mount Lofty, National Park, Baker’s Gully near Clarendon, Kuitpo, Kinchina, Kalangadoo (under Pinus), Caroline State Forest (near Mount Gambier—under Pinus), N.S.W.—Cambewarra Mount. May, June, July, August. Readily recognised by the lilacy or vinaceous tint of the whole plant and the caespitose habit. 524. Mycena subalbida, n. sp—Pileus up to 4+ in. (6°2 mm.) in diameter, usually less, conico-campanulate to convex, sometimes dimpled, sometimes gibbous or umbonate, ribbed, mealy or scurfy to glabrous, white with a greyish-brown or creamy tint. Gills adnate, attached to a collar, ascending, slightly ventricose, rather narrow, about 12 to 14 in number with shorter ones interposed, pallid greyish white. Stem °/,, to $ im. (4°5 to 10 mm.) high, curved, very slender, mealy to smooth, white to pallid, sometimes slightly brownish below, attached by a minute slightly strigose disc. Spores subspherical, 9 to 11 », 10 * 8:4 yg; the 158 cells on the edges of the gills bristling with minute processes; cystidia, 25 « long, with tapering apices and ventricose bases seen in one batch of specimens. S.A,—On mossv bark of elms (Ulmus campestris L.), North Terrace, Adelaide, June, July; on bark of Schinas Molle L., Fullarton, July; on trunk, National Park (spores 9°5 & 6°5 yw). The species seems to be related to M. corticola Fr. and M. hiemalis (Osb.) Fr., but differs and belongs to the section Basipedes by having a definite though small disc. We cannot find a description to fit it. 929. Leptonia fusca, n. sp—Pileus 3 to 14 in. (1°8 to 2°8 cm.) in diameter, slightly convex, umbilicate, radiately fibrillose, between Natal Brown and Bone Brown (xl.). Gills sinuately adnexed, moderately close, edges not dark, near Vinaceous Buff (xl). Stem 14 in. (3-1 cm.) high, slender, sometimes flattened, polished, brittle, hollow, cartilaginous, near Dusky Drab (xlv.), base whitish. Flesh very thin. Spores angular, tinted, 11 to 13 % 7°5 pw. On the ground. 5.A.—Encounter Bay. May 24, 1931. Characterised by the dark dusky brown pileus and stem, whitish base to the stem, vinaceous buff gills and rather large angular spores. 526. Clitopilus prostratus, n. sp-—FPileus 2 to 1 in. (1°8 to 2°5 cm.) in diameter, very irregular, more or less convex with the centre depressed, somewhat rugose, somewhat fibrillose, edge sometimes crinkled, colour of dead grass. Gills decurrent, moderately close, relatively deep, pallid salmon-colourd. Stem short (1 cm.), central to excentric, slender, surface matt, whitish. Spores angular with a central yellowish gutta, tinted, 9°5 to 10°5 & 7:5 p. Nearly prostrate on bare sandy soil in heathy scrub. S.A.—-Near Mount Burr (S.E.). May, 1931. 527. Clitopilus subfrumentaceus, n. sp—Pileus 14 to 4 ins. (3 to 10 em.) in diameter, irregularly convex, then more expanded or with the centre depressed, often distorted, sometimes with a small umbo, subfibrillose, edge turned in when young, not shining, somewhat hygrophanous, Pinkish Cinnamon, Cinnamon, Sayal Brown, or Mikado Brown (xxix.) becoming paler. Gills adnate to adnate- decurrent, narrow, moderately close, edges sometimes irregularly serrate, rarely forking or anastomosing near the stem to form long narrow cells, Light Pinkish Cinnamon (xxix.). Stem 14 to 24 ins. (3°7 to 6°2 cm.) high, stout (up to % in., 2:2 cm. thick), base swollen (1 in,, 2°5 cm. thick), sometimes a little excentric. somewhat mealy or fibrillose, solid, pale fawny or biscuity whitish or white. Flesh watery semi-translucent becoming whitish. Slight smell of radishes. Spores obliquely pear-shaped, rather irregular, definitely tinted, 6-5 to 8-5 & 4:2 ta 6 p. Densely cacspitose under trees or amongst grass. S.A.—Mount J.ofty Range, National Park. Vict-——Ararat. April to August. The specific name has reference to its resemblance to Entoloma frumentaceum ( Bull.) Berk. 528. Clavaria vinaceo-cervina, n. sp—Plants 4 to 2 ins. (1:2 to 5 em.) high, nearly vertical or slightly spreading, from a short stem-like base very irregularly branching, sometimes with only a few branches or prong-like divisions, sometimes with a number of small branches, ultimate divisions short, prong-like, mostly blunt, sometimes acute and thorn-like, sometimes awl-like or finger-like, often fastigiate, the branches often irregularly flattened and the whole plant rugose, usually relatively slender but in some collections stouter and more knobby, Vina- ceous Fawn (xl.) to Fawn Color (xl.), near a pale Vinaceous Russet (xxviii.), deeper than Vinaceous Buff (xl.), between Vinaceous Buff and Avellaneous (xl.), Vinaceous Pink (xxviii.) at the tips with the stem Vinaceous Fawn (xl.), greyer than Buff Pink (xxviii.), or Pinkish Cinnamon (xxix.) with a fine bloom giving a vinaceous pink colour tinge on the pinkish cinnamon, base of stem pallid. Spores subspherical 7-5 to 9 », 8 & 6°5 », 9 to 10 & 8 to 8B w. On the ground under 159 trees amongst shrubs, S.A.—Stirling West, July 23, 1927; Mount Lofty, April, June (Kew No. 86), July; Belair, July; Clare, August. Specimens of this species (Kew, No. 86) were sent to Miss E. M. Wakefield, who reported: “Probably new. Not European or American.” 529, Clavaria australiana, n. sp—Densely branched, up to 4 ins. (10 cm.) high and 5 ins. (12:5 cm.) broad, the branches between Vinaceous Buff (xl.) and Avellaneous (xl.), their tips near Vinaceous Fawn and Fawn (xl.). Con- tracting uniformly from above to a broad conical base of several stout compacted stems, The thick main branches spread somewhat and divide rather sparingly and very irregularly till the last 2 in. is reached. Here they divide frequently into numerous blunt irregular prongs, often at wide angles, the prongs often divided again and flattened. The stottt main branches and the branchlets are definitely rugose. Spore mass slightly but definitely coloured. Spores microscopically slightly coloured, elongated, oblique, mummy-shaped, not striate, 13 to 16 & 4:5 to 5°5 ». S.A——On the ground, Mount Lofty, July, 1927. 530. Clavaria corallino-rosacea, n. sp.—Clubs simple, occasionally forked several times, up to 14 to 24 ins. (4 to 5-6 cm.) high, prongs when present up to 1 cm. long, slender, attenuated downwards and also sometimes upwards, some- times rather flattened and grooved, solid, coral red or rosy pink (when moist a little pinker than Morocco Red, Dauthenay, Pl. 95, Ton. 1; when drying shades of Coral Red, Pale Scarlet, Salmon Pink, Pl. 76), often pruinose above, when buried under leaves base whitish. Flesh light coral red. Spores somewhat pear- shaped, 6 & 3:4 to 4». On the ground, sometimes under Lantana. N.S.W.— Mosman (Kew, No. 81; D.I.C., Water-colour No. 54) and Neutral Bay, Sydney, April and June. Miss E, M. Wakefield, in reporting on No. 81, says :-—‘‘Probably the same as the Brisbane specimen (Bailey 241) on which the Australian record of C, miltina was founded. The true miltina from South America is stouter and has no distinct stem. Unfortunately, the type shows no spores, but it seems unlikely that the Australian species would be the same.” 531. Clavaria complana, n. sp—Forming a mass 3 ins. (7°5 cm.) high and 5 ins. (12°5 cm.) broad. From the solid base dividing repeatedly into slender branches which then become flattened and expanded, and then again divide into slender digitate processes 4+ in. (6 mm.) long, pale pinkish tussore, becoming brownish salmon, when damp staining paper pinkish salmon. Spores hyaline, subspherical, 5*2 to occasionally 7 ». N.S.W.—Sydney suburb, probably Hornsby, June 13, 1916 (Kew, No. 68). Miss Wakefield reported as follows:—“The habit is like that of C. flabellata Wakef. from New Caledonia, but the colour is different and the spores larger. It differs from most of the other large branched forms in its hyaline spores.” 532. Clavaria sinapicolor, n. sp—Densely branched forming masses up to 24 2h ins. and 3 X 3ins. (5°6 & 5:6 em. and 7:5 & 7:5 cm.), near Mustard Yellow (xvi.) or yellower, Straw Yellow (xvi.) and Colonial Buff (xxx.), Naples Yellow (xvi.) or dingier, or Light Orange Yellow (iii.), when old near Chamois (xxx.) but yellower towards the tips or near Cinnamon Buff (xxix.), the bases of the branches paler, the stem whitish. The main branches are com- pacted into a broad mass at the base up to 14 in, (3:1 cm.) thick. Dividing upwards repeatedly by very narrow angles into closely pressed nearly vertical more or less rounded rather slender slightly rugose branches, at first 4 in. (6°5 mm.), then $ in. (3°2 mm.) and then less in diameter, the last 4 to 4 in. ending usually in numerous rather blunt prongs, some very short, often with wider angles between them than in the branches. Spore mass slightly but distinctly buff-tinted or old gold. Spores obliquely pear-shaped to elliptical, slightly tinted microscopically, 5:5 to 8 occasionally 10-4 & 3-8 to 4°5, occasionally 5 ». On the ground, usually 160 in Eucalyptus (e.g., E. obliqua) forests. S.A.—Mount Lofty (Kew, Nos. 65, 75, 76), Kuitpo, National Park. N.S.W.—National Park (Kew, No. 66, Miss Clarke, Water-colour No. 126), Kendall, Milson Island in Hawkesbury River (smaller, Kew, No, 67). May to August. Five collections, as above, were submitted to Miss E. M. Wakefield at Kew, who reported as follows:—‘‘Nos. 65 and 66 are apparently the same as No. 75 and 76. The species is not British or North American. There is no specimen of C. Kalchbrennert Miller at Kew, and the meagre description does not fit it very well. It would probably be better described as new. One of Cooke’s determina- tions of ‘C. coralloides, from Ovens River, seems to be the same species. No. 67 has spores similar to the last, but appears to have been a smaller and less branched plant. The material is insufficient to enable me to judge as to habit.” 533. Clavaria ochraceo-salmonicolor, n. sp—Compact, cauliflower-like 14 to 34 ins. (4°4 to 8-7 cm.), usually about 24 ins. (6°2 cm.) high, 2 to 3 ins. (5 to 7°5 cm.) broad in larger specimens. From a thick pallid base up to 1 in. (2°5 cm.) wide, dividing into stout branches (up to 3 in., 10 mm. thick) and these again dividing three or four times to end in blunt prong-like processes capped by several blunt teeth a few mm. long, angles’ rather rounded, branches with longitudinal rugae. Colour Light Ochraceous Salmon (xv.), Ochraceous Salmon (xv.), Light Ochraceous Buff (xv.), or Apricot Buff (xv.) when drying; when young Capuchin Orange (iii.), the tips yellower, which yellow may be lost when older; tips some- times Warm Buff (xv.) or Ochraceous Buff (xv.). Spores elongated pear- shaped with an oblique apiculus, in the mass yellowish-brown, microscopically slightly tinted, 8°5 to 13 & 3°7 to 5 yw, usually about 9 to 10 & 44. S.A—Mount Lofty (Kew, No. 71), Willunga Hill, Second Valley Forest Reserve, MacDonnell B. (in S.E.), April, May, June, July. Specimens from Mount Lofty, June 16, 1917, forwarded to Kew, were returned by Miss Wakefield as “not matched at Kew.” NOTES ON SOME SOUTH AUSTRALIAN AND CENTRAL AUSTRALIAN MAMMALS. PART 2 BY H. H. FINLAYSON Summary 1. Since recording the presence of Thalacomys lagotis in the Musgrave Ranges (Trans. Royal Soc. S. Aust., 1930, p. 178)) further specimens have been obtained from localities in the centre, considerably further north. In August, 1930, Messrs. Hale and Tindale, during the stay of the Adelaide University Anthropological Expedition in the Centre, obtained a male and female from the blacks at Macdonald Downs, about 120 miles north-east of Alice Springs. And again in August of this year, when attached to a similar party, Mr. A. Rau, of the Museum staff, obtained a male at Cockatoo Creek, on the Tanami track, about 150 miles north-west of Alice Springs, in latitude 220 S., approximately. 161 NOTES ON SOME SOUTH AND CENTRAL AUSTRALIAN MAMMALS. PART 2. By H. H, Fintayson. [Read October 8, 1931.] 1. Since recording the presence of Thalacomys lagotis in the Musgrave Ranges (Trans. Royal Soc. S. Aust., 1930, p. 178), further specimens have been obtained from localities in the centre, considerably further north. In August, 1930, Messrs. Hale and Tindale, during the stay of the Adelaide University Anthropological Expedition in the Centre, obtained a male and female from the blacks at Macdonald Downs, about 120 miles north-east of Alice Springs. And again in August of this year, when attached to a similar party, Mr. A. Rau, of the Museum staff, obtained a male at Cockatoo Creek, on the Tanami track, about 150 miles north-west of Alice Springs, in latitude 22° S., approximately. All three specimens present the characters of lagotts, previously noted in the Opparinna example, The largest male, got at Macdonald Downs (S.A. Museum, M 2930), had the following flesh dimensions:—Head and body, 415; tail, 270; pes, 100; and its skull, which is massive and strongly ridged, has a basal length of 98 and zygomatic breadth of 54. The female, taken at the same place, has :— Head and body, 320; tail, 235; pes, 88; and the skull is smooth and devoid of crests, and has basal length 78 and zygomatic breadth 39, Both specimens are aged and show about the same degree of tooth wear; the sexual differences in size and contours are, therefore, extraordinarily marked. The male skull weighs 34 grammes, and the female 15 grammes. Mr. L. Glauert, Director of the Western Australian Museum, has recently examined a splendid series of 40 /agotzs, culled from Western Australian localities and, pending completion of his work, has been good enough to inform me in advance of some of his findings. It would appear that even in comparatively restricted areas the adult size is far from constant and varies sufficiently to embrace all four of the Central specimens I have noted. Under these circumstances, I withdraw the remark that the Central lagotis is a dwarfed form, as being at present insufficiently founded. 2. The disappearance of the short-nosed bandicoot, /soodon obesulus, from the greater part of South Australia is a typical example of a number of similar faunal declines occurring, sometimes over areas almost continental in extent; sometimes, as in this case, in restricted localities, but always without adequate cause being apparent. Originally widely distributed in this State, and in some parts in great numbers, /soodon remained a common and familiar animal long after the begin- nings of settlement, and indeed seems to have received its first serious check not more than 30 years ago. Since then it has dwindled to such an extent that the present generation of settlers has largely forgotten even its name, and when, rarely, one is taken, extraordinary speculations as to its identity are heard. In the last 10 years nearly all specimens obtained at the Museum have come from localities in the South-Eastern district, adjoining Victorian territory, where it is much more common, and Wood Jones’ record of one from Blackwood (Mammals of S.A., vol. ii., p. 138) has remained unique for the Adelaide district till quite recently. 162 The causes usually quoted in explanation of these disappearances are the prevalence of foxes and of feral cats, the occurrence of epidemic diseases, the laying of poison baits, competition with the rapidly breeding rabbit, and the burning off of large areas at frequent intervals, and no doubt all these have played apart. But that these in themselves are inadequate to account for all the observed facts, and that other and more fundamental factors are involved, is shown by the still more mysterious recurrence of “extinct” species from time to time, with no apparent change in ecological conditions. There are now unmistakable signs that the recovery of J. obesulus in the Mount Lofty Range is proceeding apace. During the opossum season of 1930 it was constantly reported as being taken by trappers, and some dozens of pelts were sent into the sale rooms, and during the last few months four specimens have come to hand from localities within 20 miles of Adelaide. The characters of the local race have been fully stated by Wood Jones (loc. cit.) 3. From time to time reports have been received by the writer of a small wallaby or “kangaroo rat,” occuring sparsely in spinifex country in several localities in the far northern areas of the State. Descriptions of its appearance and habits were sufficiently precise and consistent to rule out both Bettongia lesueurt and B. penicillata from its possible identity, and it appeared certain that it was one of the hare wallabies, probably Lagorchestes hirsutus, which was obtained by the Elder Expedition towards the north-west boundary of the State. About a year ago an opportunity occurred of questioning a practised observer who had seen the living animal at close quarters and had handled specimens of it, dead. When confronted with a series of filled skins of L. conspicillatus leichardti, L. c. typicus, L. hirsutus, Bettongia lesueuri, B. penicillata, and Aepyprymnus rufescens, he was quite emphatic that the spinifex wallaby was not represented. Finality has now been reached (September 10), on receipt of a skin and skull of the animal, which proves to be Caloprymnus campestris, described by Gould in 1843, recorded again by Tate in 1878, and since then a “lost” species. Externally it is very distinct both from the other Potoroinae and from the Lagorchestines, and its skull characters are, fortunately, so peculiar and pro- nounced as to remove any element of doubt from the identification. Field work in the locality of the occurrence will at once be undertaken and, pending its completion, a revision of the characters of the animal is deferred. PETROGRAPHIC NOTES ON SOME BASIC ROCKS FROM THE MOUNT BARKER AND WOODSIDE DISTRICTS BY A. R. ALDERMAN, M.Sc., F. G. S. Summary Basic rocks have been recorded from various localities in the Mount Lofty Ranges in South Australia. ‘'’ The rocks described in this paper occur as follow:- I. Basic dyke, sections 5267 and 5269, Hundred of Onkaparinga; about two miles east of Woodside. II. Basic dyke, sections 3828 and 3829, Hundred of Macclesfield; one to two miles north- west of Mount Barker (the hill), and not far from Mount Barker Springs. III. Basic dyke, sections 4213, 4214, and 4216, Hundred of Onkaparinga; near Mount Barker Junction railway station. 163 PETROGRAPHIC NOTES ON SOME BASIC ROCKS FROM THE MOUNT BARKER AND WOODSIDE DISTRICTS. By A. R. Atperman, M.Sc., F.G.S. [Read October 8, 1931.] Basic rocks have been recorded from various localities in the Mount Lofty Ranges in South Australia. The rocks described in this paper occur as follow :— I. Basic dyke, sections 5267 and 5269, Hundred of Onkaparinga; about two miles east of Woodside. II. Basic dyke, sections 3828 and 3829, Hundred of Macclesfield; one to two miles north-west of Mount Barker (the hill), and not far from Mount Barker Springs. III. Basic dyke, sections 4213, 4214, and 4216, Hundred of Onkaparinga ; near Mount Barker Junction railway station. For brevity these occurrences will be referred to as: I., Woodside; II., Mount Barker Springs; and III., Mount Barker Junction. References to these rocks have been made by Howchin,“) Benson,“ and others. The age of the intruded sediments is somewhat doubtful, owing to their metamorphosed condition, but recent work by Prof. W. Howchin“) seems to indicate that at Mount Barker Springs and Mount Barker Junction they are _ Upper Pre-Cambrian (Adelaide Series). The intruded rocks at Woodside may be of the same age or Lower Pre-Cambrian (Barossian). I, The Woodside rock occurs as a broad basic dyke striking in roughly a north-west direction. When examined in the hand specimen the rock appears dark and holocrystalline. It is of porphyritic habit, felspar crystals, up to 5 mm. in length, being embedded in a dark fine-grained groundmass. Some specimens contain far more felspar phenocrysts than others. The rock analysed, and described first, is of the more felspathic type. Microscopic Features——The structure is essentially porphyritic, in which large felspar phenocrysts are embedded in a groundmass consisting mostly of felspar and hornblende. The felspar is a basic labradorite, mostly in oblong forms but occasionally showing square cross-sections. Both albite and pericline twinning is common, Carlsbad twins also being occasionally seen, The felspar crystals contain innumer- able inclusions of all sizes, the larger of these generally being amphibole although sphene sometimes occurs in this way, and, of course, the felspar alteration products. Minute inclusions are extremely numerous and are very often arranged in rows parallel to the sides of the felspar crystals. These inclusions consist of both amphibole and chlorite, the latter evidently being an alteration from the former. The arrangement of these parallel to the crystal faces suggests inclusion of the mother liquors during rapid crystallisation of the felspar. Many of these small () Mawson, D., Rept. A.A.A.S., vol. xviii., 1926, pp. 251-252. ©) Howchin, W., “Geology of South Australia” (sec. ed.), 1929, pp. 61-62. () Benson, W. N., Trans. Roy, Soc, S$, Aust., vol, xxxi., 1909, p. 240, ©) Howchin, W., Trans. Roy. Soc., S. Aust., vol. liti., 1929, pp. 27-32. 164 inclusions are arranged with a similar optic orientation. Some of the felspar is somewhat obscured by dusty kaolinitic material. After the felspar, green hornblende is. the most important mineral, It does not appear to be primary, but to have recrystallized in fine-grained aggregates. No indication of a uralitic nature is shown. It is pleochroic from light to darker green, The smaller individuals, and a few of the larger, show an alteration to chlorite. Scapolite is a notable constituent and in ordinary light it resembles the felspar, but is easily distinguished by its higher colours between crossed nicols. Grains of this mineral occur adjacent to the felspars, from which they are apparently derived by alteration. Grains of zotsite are common, and although they are sometimes found in separate individuals they generally occur with an aggregate of kaolinitic matter and felspar showing a distinct poikiloblastic structure. Some small colourless grains in this aggregate are probably albite, suggesting that the felspar has been saussuritized. The opaque iron ore is ilmenite, which shows an alteration to leucoxene. Sphene is present in dusty grey-brown grains which are occasionally wedge- shaped. A narrow vein of secondary quartz is to be seen in one portion of the section exatined, Chemical Analysis. Percentage. Percentage. Silica (SiO, ) we as «4. 47°63 Carbon dioxide (CO,) .. Nil Alumina (AI,O,) .. .. .. 21°94 = Titanium dioxide (Ti0,) .. 0°77 Ferric oxide (Fe, O, a) ee ee 181 Phosphorus pentoxide (Ps O. =) 0-15 Ferrous oxide (FeO) ety 22 “ATSO Sulphur (S) . 0-15 Magnesia (MgO) . ow. 681 Chromic oxide (Cr, O ) .. Trace Lime (CaO) .. .. .. ) «. 14°08 Manganous oxide (MnO) .. 0:08 Soda (Na,O) wea? oe “ELD Barium oxide (BaO) .. .. Nil Potash (K,O) . ~ .. 0°28 Water (combined) (H, O-+-) 0°68 Total .. .. 99:79 Water (hygroscopic) (H,O-) 0°03 The specific gravity is 2°95. The Norm. Percentage. Orthoclase .. .. .. .. L6/ Salic G Albite ov. La we ae ov. 15°20 I = 67°74 eT TA Anorthite .. .. .. .. 50°87 a Diopside .. .. 1. we. 14°52 t 4. ey, Hypersthene .. .. .. .. 9°64 Besse te QUvine Du cope Be ot Ba 2i42 O = 2:42 : Tmenité 2. 2. 1 aa an D2 } eee. ‘emic Group Magnetite .. .. .. .. 2°55 ee == 31°26% Apatite 2. .. 2. 1. 1. O84 F Pyrite Sal "by aw 8 Bae 0827 } A= 0-61 Water... 0-71 In the C.I.P. W. classification ae position of the rock is, therefore :-— » 5, 4, 4-5. The Anas name is Hessose. 165 The high felspar content is indicated by the high percentages of lime and alumina giving over 50% of anorthite in the norm. Actually a good deal of this lime and alumina must enter into the modal hornblende. Apart from these points the analysis seems quite normal for a rock of this type, and supports the idea gained from the mineral content that it must be classified in the basic division of the calc-alkali series. Another specimen from the same locality differs from the rock analysed mainly in containing a smaller amount of felspar. This mineral, which has the composition of a medium labradorite, again shows twinning on the albite, pericline and Carlsbad laws, and contains the numerous inclusions mentioned in the description of the former rock. These phenocrysts, in some cases, consist of a nuinber of individual crystals which have grown together. The felspars again have suffered an alteration round the edges to scapolite and epidote, but not to the extent noted in the rock analysed. The amphibole, also, is somewhat paler. Il, The occurrence of the Mount Barker Springs rock is described by Howchin, He writes that: “A basic dyke, 28 yards wide, occurs about two miles to the north-west of Mount Barker Hill, with a strike directed towards the mount, and can be traced for nearly a mile; its age is undetermined.” In a more recent paper‘®) the same writer indicates that the intruded rocks are most probably Adelaide Series sediments. In the hand specimen colourless porphyritic felspars, many measuring up to 5 mm. in length, are embedded in a dark, somewhat greenish groundmass. Microscopic F*eatures——The porphyritic texture is again well displayed by the presence of large plagioclase crystals in a groundmass consisting mainly of hornblende and plagioclase. The felspar is a medium labradorite in which albite, pericline and Carlsbad twinning is well developed. Inclusions in the felspar phenocrysts are less numerous, but of the same nature as those in the Woodside rocks. The hornblende also is similar to that in the more felspathic rock at Woodside. Granules of scapalite are associated with some of the felspar phenocrysts. Both zoisite and green epidote are frequently surrounded by fine sericitic material, these aggregates constituting alteration products of the felspar. Small granules of ilmenite, changing to leucoxene are plentiful. A smal! quantity of sphene is also present. Chemical Analysis. Percentage. Percentage. Silica (SiO,) re eal an 4DRD Carbon dioxide (CO,) .. Nil Alumina (Al,O,) .. 0 .. ©. 18°41 Titanium dioxide (TiO,) .. 1:01 Ferric oxide (Fe,O,) .. .. 1°00 Phosphorus pentoxide (P,O,) 0-21 Ferrous oxide (FeO) .. .. 5°87 Sulphur (S) .. 6. 6.2. 0°05 Magnesia (MgO) .. .. .. 8°68 Chromic oxide (Cr,O,) .. Nil Lime (CaO) .. .. 2.) .. 13°08 Manganous oxide (MnO) .. 0-15 Soda (Na,O) ve ee ee) 6118 Barium oxide (BaO) .. .. Nil Potash (K,O) .. ww 1. 0°28 Water (combined) (H,O+) 0-67 Total .. ., 99-87 Water (hygroscopic) (H,O—) 0-04 The specific gravity is 3-00. © “Geology of South Australia” (sec. ed.), 1929, p. 62. () Trans. Roy. Soc. S. Aust., vol, liii., 1929, pp. 27-32, 166 The Norn. Percentage. Quartz .. .. .. «. «. 120 Q = 1:20 Orthoclase .. .. «. «- L167 Salic Group Albite .. 0 1. 2. ee ee) 9°96 F = 55°55 = 56°75% Anorthite sim et tet ot ABO? j Diopside .. .. .. «. +. 16°09) __ ag. Hypersthene .. .. .. .- 22°54 $ Bae BeCOF Ilmenite .. .. .. .. «. 1°98 4 es Femic Group Magitite\ jsf se jg Teese se { M= 337 “49-34% Apatite .. <. 2s. 09+ % O44 A = 0°34 Water... o wy OO In the C.I.P.W. classification the position of the rock is, therefore :— IIL., 5, 5, 4-5. approaching: IIL., 5, 4, 4-5. The magmatic name is Auvergnose-Kedabekase. The analysis shows lower alumina, lime and soda than the Woodside rock. This is reflected in the smaller amount of normative felspar. The ferromagnesian content is correspondingly higher. In these points the norms indicate the actual mineralogical differences between the two rocks. It is probable that the less felspathic type from Woodside would be extremely similar, chemically, to that occurring at Mount Barker Springs. ILI. As mentioned by Howchin,“ the basic dyke at Mount Barker Junction is not seen in outcrop, its presence being indicated by loose stones on the surface of cultivated land. In the hand specimen this rock bears an extremely close resemblance to that from Mount Barker Springs. Microscopic Features——Porphyritic felspars occur in a groundmass of horn- blende and felspar, this structure being similar to that of the rocks described above. The plagioclase phenocrysts again contain numerous inclusions which are mostly chlorite. Extinction angles indicate that the felspar is a medium labradorite, This felspar, although it is corroded, differs from that of the other rocks in that no scapolite has been formed, although both soisite and green epidote are present in small amount. The twinning of the plagioclase is confined to the albite and Carlsbad types, no pericline twins being seen. The hornblende is similar to that of the other rocks and is somewhat chloritized. Sphene is a notable constituent, but iron ores are practically negligible. Chemical Analysis. Percentage. Percentage. Silica (Si0,) we cee as 49209 Water (combined) (11,O0+) 0°89 Alumina (Al,O,) .. .. 18°03 Water (hygroscopic) (H,O-) 0°05 Ferric oxide (Fc,O,) .. «. 3°35 Carbon dioxide (CO,) .. Nil Ferrous oxide (FeO) .. .. 4°56 Titanium dioxide (Ti0,) .. 1°38 Magnesia (MgO) .. .. -. 7°48 Phosphorus pentoxide (P,O,) 0°21 Lime (CaO) .. .. «. -- 12°97 Manganous oxide (MnO) .. 0°18 Soda (Na,Q) eat git et LAA Potash (K,O) epee ee «(O42 Total .. 100-32 The specific gravity is 3°01. @) Loc. cit., p. 31. 167 The Norm. Percentage. Quartz .. .. .. .. 4. 2°58 Q= 2:58 Orthoclase .. .. 2. 0... 2°22 Salic Group Albite .. 6. 2. 0... 1415 F = 56°68 = 59°26% Anorthite Ae tt ee ae Diopside .. .. «2. ..) .. 18°71 _ Hypersthene .. .. ..) 2. 13°41 ' P= 32°12 Magnetite eee ae ee) 487 a. oe Femic Group Imenite .. 1. .. 1. 1. 2-74 ‘ Me tk *eiagioza Apatite 2... 2. 1. 1. 0°34 A= 0:34 Water... .. ww ow. 094 In the C.I.P.W. classification the position of the rock is, therefore :— III., 5, 4, 445. The magmatic name is Auvergnose. The analysis shows the extreme chemical similarity between this and the Mount Barker Springs rock. There are, however, minor variations in the alkalies and the relative amounts of iron and magnesia. CONCLUSION, The mode of occurrence, mineralogical nature, and chemical composition of these three rock types suggest that they are closely related, and that they should be included in that class of rock to which the misleading name epidiorite has been given, The writer would suggest, however, that the compound name dolerite- amphibolite could be applied advantageously to types such as have been described in this paper. ABSTRACT OF THE PROCEEDINGS OF THE ROYAL SOCIETY OF SOUTH AUSTRALIA FROM THE YEAR NOVEMBER I, 1930, TO OCTOBER 31, 1931 Summary A GENERAL DISCUSSION on "Laterite and Lateritic Soils" was introduced by Professor J. A. Prescott, who said that "laterite was originally defined in 1807 from certain soil formations and geological structures in Southern India by Buchanan. Laterites occur throughout a considerable area of Australia, and were first recognised as such in Western Australia by Simpson and by W. G. Woolnough. The essential feature of the laterite consists of a capping or conglomerate of massive or pisolitic iron hydroxide with certain examples of aluminium hydroxide or bauxite. Soil workers have assumed that the process of laterization was a distinct weathering process confined to the tropics, but further investigation has revealed no evidence in support of this, such tropical weathering processes being entirely similar to those occurring in temperate zones. It is suggested, therefore, that laterite, as defined by geologists, is essentially a relic of former soil-forming processes." Professor Prescott then read a letter dealing with the subject by Dr. L. Keith Ward. Mr. R. J. Best then dealt with the chemical aspect, and illustrated his remarks with diagrams. 168 ABSTRACT OF THE PROCEEDINGS OF THE ROYAL SOCIETY OF SOUTH AUSTRALIA (Incorporated). FOR THE YEAR NoveMBER 1, 1930, ro Octoser 31, 1931. ORDINARY MEETING, NOVEMBER 13, 1930. Tue Presripent (Dr. Chas. Fenner) in the chair and 28 members were present. Minutes of the Annual Meeting, held October 9, 1930, were read and confirmed, ELecTIon of FeLtow.—John Irvine Miller, C.E., Crystal Brook,