PROCEEDINGS tril, scarred body, shapely feet, small hand .... Except in colour, they were unlike their neighbours of New Holland (Australia). In hair, in nose, in limb, they differed The lowest down the depths uf barbju'ism, they were neither stupid nor miserable . . . , but had sense and feeling." Bonwick, in his larger work on the " Daily Life and Origin of the Tar-manians " (5) ^ives some anthropometrical measurements of the race, and aiso goes, at much gieater length, into the physical characteristics of the latter, of which the following is a brief summary : — Skin, dark brown, or nearly black, but so disguised with pig- ments as to make it diflScult to state exactly what colo-ur. Hair hangs in cork>-crew appendages aibout the men's faces, is black, and has a crisp, woolly look. The diameters of the hair ellipse are given as 25 is to 15, to which point reference will again be made. The eyes have the iris always dark coloured, whilst the white of the eye is not so clear as in Europeans. Mouth, great width; lips, though full, had not the negro dimensions. Jaws, strongly *et ; chin, inferior to thart: of civilised races, and in the women, particularly, very small and retreating. Nostrils, exceedingh^ wide and full, but the great peculiarity, though not ab>:olutely confined to this people, lay in the depres- sion at the commencement of the organ, giving the feature much of a pyramidal character. Teeth, large and powerful, so much so as to constitute a decided peculiarity. On the question of the teeth, Bonwick ap- parently enlisted the services of a. dentist, for he quotes Pardoe, a Melbourne representative of that profession, at> follows: — • 6 Richard J. A. Berry: " The Tasmanian teeth have laro-e crowns, thickly covered by enamel, more so than in Europeains. Fangs are not so deeply seated in alveolar sockets, nor does ejiiphysis of maxillary bone come so high as in Europe. Gums are much thicker and make up this loss." The colour of the teeth remarkably white, so much so as to have been greatly envied by i-ome of the earliest French voyagers tu the island. Modern methods, with exact measurements and indices, would now enable us by means of Flower's dental index (7) to style these teeth megadont, with an index above 44. Regarding the relative proportions of the hair ellipse, Bonwick gives the following table: — Tasmanians - as 25 is to 15 Negro - „ 20 „ 12 Fiji - - „ 35 „ 20 Malay - „• 22 „ 15 New Zealand „ 24 „ 18 Chinese - „ 33 „ 24 It is not stated whether these proportions are actual measure- ments of the hair, as viewed microscopically in transverse section or not, but ais it would be difficult to obtain «uch proportions without actual measurements, we may assume for the moment that they represent actual measurements, and apply thereto the index mentioned by Duckworth (8), thus : — Incle'c — Llie OB CtlOIl X Length of the Section itli the following results : — Tasmanians - Index (50 Negro - 60 Fiji - 57 Malay - 68 New Zealand 75 Chinese 72 Duckworth says " the numerical value of this index ha^^ been found to vary between 28 and 100, the lowest figure being pro- vided by the curly hair of an Oceanic (Piupuan) Negro, and the highest by the lank and straight hair of Mongolians.'' He also gives a figure of the cross sections of the hair of a Negi'ito Semang from the Malay Peninsula, where the index is in the one 55.2, and in the other 58.9. ^1 Hal/-<:aMe Taxman uui. 7 As it will be presently shown that the consensus of opinion is apparently in the direction of allying the Tasmanians to the Papuans of New Guinea, the discrepancy in the hair index, as worked out from Bonwick's figures, and of Duckworth's state- ment regarding the low nature of the genuine Papuan hair index would be somewhat remarkable could we relv absolutely on Bonwick's proportions representing actuail measurements, but this, as stated, is only conjectural. Roth (12) in his " The Aborigines of Tasmania," published in 1890, has presented us with what is, perhaps, the best general description of this race ; as, however, Roth's account of the physical characteristics of the native Tasmaniaia is a compilation from various authors, whilst Bonwick's account is Irom personal obsei'vation, I have preferred to utilise the latter author only, more particularly as, after all, Roth does not, on this question, differ very considerably, if ait all, from the earlier writer. Roth, however, being of more recent date than Bonwick, is much more precise in his anthropometrical data, and these data I hope to avail myself of in a future communication on this subject. The subject of the present paper is, as already mentioned, a Tasmaaaian half-caste, and a reference to her photograph will show that she bears many striking resemblances to the pen pic- ture quoted from Bonwick, particularly in the colour of the skin, the width of the mouth and nostrils, the weak chin, and the dark eyes. The hair, though distinctly woolly, has departed from the racial type consequent on the admixture of the white blood, though curiously enough, the native type is, on the whole, more marked in the grand-daughter of the aborigmal mother than in the daughter. In this connection, it is interest- ing to see what has been said of the first Tasmanian half-caste, in contradistinction to this, which is the last of such crosses. Evans (9), in his " History and Description of the Present State of Van Dieman's Land" (Tasmania), says: — " The eldest child of this (native Tasmanian) woman, now a fine girl about eleven years old, and the first child born by a native woman to a white main in Van Dieman's Land .... is called Miss Dalrymple, and, like all the other children since produced by an intercourse between the natives and the Euro- peans, is remarkably handsome, of a light copper colour, with 8 lilrliavd J. A. Bki'vi/ : rosy cheekt^, larp:© black eyes, the Avhites of which are tinjxed with blue, and long well-formed eye-lashes, with the teeth un- commonly white, and the limbs admirably formed." Whatever opinion may be formed as to the g-ood looks of the subject of the present paper there can be nn two opinions as to her intelligence. In conversing with her, the two facts which impressed me most strongly were her remarkable in- telligence and the absolute purity «if her English speech, and had I nnt actually heard her, I euuld not have believed that such intelligence could have been derived in one generation from a race, often, but, perhaps, quite errnne- ously, believed to have been one of the most degraded and Ijrutal in the world's race>-. That this opinion is in no way ex;!j^<.'erated is shown by Hallack (4), who says : — " Mrs. William S. is . . . . of a bright and happy dis- position, a most entertaining conversationalist, and, witlial. e.'t- tremely apt at repartee." The records of the deaths of the la-t pure-hrcd Tasmriniiins are as follows : — "The last Tasmtwiian man, William Lanney, Lainiy. "V Jeanne, alias King Billy, died on March 3rd, 1869, aged M. In January of the previous year (1868), he had walked proudly with Prince Alfred, Dvd^e of Edinburgh, on the Hobart Town Regatta gmund, conscious that they alone were in possession of royal blood " (6). Whil>t Lanney was the bust surviving man of the race, he was outlived by a \vom;m named "" Truganini, or Lalla Hoolvh .... who died in May, 1876, and was supposed to be seventy-three years old ' (6). With her, the native Tasmanian race became finally extinct, and there now remain but a few half-castes, of whom the subject of this paper forms (ine of tlie nldest, i!' not actually the oldest, now living. We now pass to the purely controversial side of the question, and though there are many debatable points in connection with the lost Ta^nanian race, attention will only be directed tn three of these problems. These three are, hnwcver, of the very greatest importance, and are as follows : — 1. Is the Tasmanian of remote or recent origin.' 2. With what race is the Tasmanian most closely allied.' 3. Ilow did the Tasmanian reach Tasmania ^ A Half-caste Tasmanian. 9 Regardinfr the first of these questions it may be stated that the few authors who have made any scientific observations whatsover upon this unhappy race, as well as those, who, from personal contact with its then living representatives have been in the best position to judgfe, are all ao^eed as to the ^eat an- tiquity of the Tasmanian aborig-inal. Bonwick, whose excellent description of the Tasmanian aboriginal hais already been quoted, has no doubts upon this point, for he says (5) : — •■ That the Tasmaniiuas .... are of high antiquity, even as regards other inhabitants of the world, can admit of little doubt. A strong argument for their remote age may be gathered from their ignorance of navigation." And, again, the same author says, " No race presents itself to us of a greater relative antiquity (than the Tasmanian). They lived throughout history." Tylor, in his preface to Roth's " The Aborigines of Tas- mania" (12), says: — ■' If there have remained anywhere up to modern time.s men who>e condition has changed little since the early stone ;vj:e. the Tasmajnians seem to have been such a people. They stand before us as a branch of the Negroid race, illustrating the cmi- dition of man near his lowest known level of culture .... it appears that the aborigines of Tasnumia .... by the workmanship of their ^tone implements rather represent the condition of Palaeolithic M;\.n. ' Ihe same author (Tylor l-'^) has elsewhere pointed out that the Tasmanians were representa- tives of the stone age development, and were in a stage lower than that of the (^Kiaternary period of Europe, and hence the distinction may be claimed for them ni beinii" the lowest of modern nomad tribes.'" Howitt (1-i), too, bears witness to the same idea, for he says: — "In considering all the facts before me bearing upon the ques- tion of the origin of the Tasmanians and the Australians, I have been much impressed by the immense periods of time which seem to be essential to any solution of the problem," and, again. "I have said before, and desire to repeat, thait the conclusions to which I have been led as to the origin of the Tasmanians and Australians necessarily demand a vast antiquity on the 10 Richard J. A. Berry: Australian Continent, for the former, and a very lonpr period of at least prehistoric time for the lartter." As examples of scientific evidence the foregoing extracts count for little, but as examples of close scientific reasoning from the known to the unknown they count, or should count, for much, and it seems to me that an antiquity, a great antiquity, must be allowed the now extinct Tasmanian race, for there is no ques- tion that the more one examines the problems attaching to the Tasmanian, the more the opinion forces itself upon one's atten- tion. Concerning the second of these debatable points — " With what race is the Tasmanian must closely allied?" the consensus of opinion appears to be in favour of regarding the Tasmanians as quite distinct from their neighbours of the adjacent Austra- lian mainland, and, second, of allying them to the much more distant Papuans of New Guinea, or, rather, to the primitive stock from which that people may have been derived. As regards the first point, and, incidentally, the second also, Mr. Protector Parker, quoted by Bonwick (5), says: — ■' It is one of the many strange anomalies of Australian geo- graphy that a branch of this Papuan race should have been found in Australia (i.e., Tasmania), whose woolly hair and blacker com- plexion clearly dir^tinguish them from the Continental Austra- lian, and yet that no branch of the same family should be found on the shores of the mainland nearest the presumed locality where the race originated.' Gaa-son, who contributed the osteological chapters to Roth's work on the Tasmanian aborigines (12), says: — '■ The race to which the Tasmanians might naturally be thought most allied from their geographical position is the Aus- tralian, but many points in the physical characters of the two races are so totally unlike as to render this relationship pro- blemartical." Topinard, the great French anthropologist (15), stated that the skulls of Australians and Tasmanians examined by him differed considerably, and he gave it as his opinion tliat these two peoples were distinct races. Huxley (16) points out tluut the type of Australian man is quite distinct from that of the Tasmanian. A Hdlf-cdsto Tiismanian. 1 1 It would therefore appear that, so far from the Tasmanian beintr akin to his nearest neie of islaaids of Torres Straits and of Ne^v Guinea. In the outlying islands to the eastward, especially in the Fijis, the Negritos have certainly undergone considerable intermixture with the Polynesians : and it seems probable that a similar cros.Miikulls aind other osteological remains of the two branches of the race. If stich an examination be not conducted, and that at once, it will be little short of a national disgrace, whilst for the policy which consists in scattering the most valuable Tas- manian nuiteriail et hoc genus omne in European and other foreign nuiseums, I have nothing but condenunition. The 14 Riclidid J. A. lieni/: nmterial havin;jf been sown in Au>tralii<., let Australians see to it that they reap the harvest. Pending- ihe arrival oi this positive proof as to the rela/tion- ship between the Papuan and the Tasmanian, we shall assume, on the opinions of those whose works have been quoted, that such it really the case, and now, therefore, we pass to the third and biiSt question, how did the Papuan get to Tasmania ? If the foregoing view as to the identity of the Papuan and the Tasmanian, using the terms " Papuan '' and " identity " in their very broadest sense, be correct, it implies that representatives of the same Papuan r.tock have become widely sepaa'ated, both geographically and ethnologically ; geooTaphically by the separa- tion of the islands of New Guinea and Tasmania from the Conti- nent of Austrailia ; and ethnologically by the interpolation of a distinct race, the Australian aboriginal. On this point, B(mwick asks the question (5) : — " Ho.v could the woolly haired Papuans of Tasmania get so far >:ep;.irated from the woolly haired Papuans of New Guinea, New Hebrides, etc., whilst having their cousins of more luxuriant hair occupying the Continent of Australia between the two!" He answers his own question l>y assuming that " part^ of New Holland (Aus- tralia) were united to New Guinea, to New Zealand, and to Tasmania," and a little farther on Bonwick adds, "' The Aus- tralians proper are now contined between the two great seats of the so-called Paipuan race, and as there are no evidences of their race dwelling in New Zealand, New Guinea, or in New Caledonia, it is nmch to be doubted whether their advent in their Australian home was not after the separation of those islands. In the same way, it ma}- be thiut they came after Tas- mania became disconnected." It is a somewhat remarkable fact tliat although that ]>art of Bonwick's assumption relative to a, primitive land connection between Australia and New Zealand is not capable of geological proof, the land connection between New Guinea, Eastern Aus- tralia, and Tasmania, may be regarded as c-ertain from the re- searches of Wallace (11), Howitt (U), Spencer (22), and many others. Since Bonwick's time, Wallace (11), working out the problems attendant on the distribution of New Zealand flora, has also endeavoured to show thait during the early cretiiceous A Ildlf-cfiste Taxman io/n. 15 period the present continent of Austrii.lia was divided into two parts, an Eastern and a Western Australia. His precise statement is as follows : — " If we examine the geological map nf Austniilia (given in ■' Stanford's Compendium of Geography and Travel, volume Australasia), we shall see good reason to conclude that the eastern and the western di%'isions of the country tirst existed as separate islands, and only became imited at a comparatively recent epoch. This is indicated by an enormous stretch of cre- taceous ajnd tertiary formations extending from the Gulf of Carpentaria completely across the Continent to the mouth (jf the Murray River At this epoch then .... Aus- tralia may not improbably have consisted of a very large and fertile western island, almost or quite extra-tropical, amd extend- ing from the siliu-ian rocks of the Flinders Range in South Australia, to about 150 miles west of the present west coast, and southward to about 350 miles south of the Great Australian Bight To the east of this, at a distance of from 250 to 400 miles, extended in a north and south direction, a long, but com- paratively narrow island, stretching from far south of Tasmania to New Gtiinea, while the crystalline and secondary formations of Central North Australia probably indicate the existence of one or more large islands in that direction." I am informed that Wallace's contention as to complete sepa- ration of the Australian Continent into two haJlves is geologi- cally untenable, although the wide extension of a cretaiceous sea over what is now Central Australia, as well as the land connec- tions between New Guinea, Eastern Atistralia, and Tasmania are admitted. Notwithstanding that certain parts of Bonwick's assumptions as to land connections have thus been proved to be correct, the theory which he built thereon as to the origin of the Tasmaaiians cannot, in my opinion, be sustained. Bonwick assumed the pre- sence of a large southern continent, by means of which Austra- lia, Tasmania, and New Zealand were connected together, and he thinks thajt both the Tasmanians and the Australians eman- ated primarily from this continent. Tlie submergence of this continent, and the subsequent separation of Tasmania from the Australian mainland resulted in the long isolation of the Tas- IG H/chard J. A. Berry: maniiins. Whilst this theory is liardly caii>able, as stated, of be- inir suslained, it is only fair to mention that in favour of it there is a view that, the aboriginal inhabitants of the southern extremities of the three great continents of Australia, South Africa, and South America, tha-t is the Tasmanians, the Bushmen, and the Fuegians, appear to have some features in common, though even this similarity is much, more closely confined to tlie Ti'isinanians and the Bushmen than to the Fuegians, the last mentioned of which differ very markedly in stature from the first two mentioned races. If, however, we reverse J?onwick"s theory, and make the Tas- manians emanate from the north instead of from the south, it seems to me that we approach much more nearly to the known facts, and, before broaching this theory, which is no new one, it will be well to state what a.re the facts, uieagre at the best, up;in which we have to build. It is certain that the Tasiuanian had no knowledge of naviga- tion ; it is almost certain that he i>; of great anticiuity, and tha.t he is closely allied to the Papuan of the New (luinea district ; it is further known that the Northern or New Guinea section of the family is, or was, until recently, separated from the Southern or Tasniiuiian section by the presence of a different race, the Australian aboriginal; whilst, lastly, it may be taken as fully proved that there was once a land connection between New Guinea, Eastern Australia, and Tasnninia. With these, the nearest approaches to facts available to us, and assuming the land connections to have been somewhat -as sketched, the distribution of the rai}»uan race, or the primitive progenitors of that race, would have extended from what is now New Guinea in the north to what is now Tasmania in the snuth, and this extended range would require no knowledge of naviga- tio:i. Tlie separation of New Guinea and Tasmania would then have broken up the race into three areas. New Guinea, tlie nar- row belt of Eastern Australia, or more likely the present Aus- tralian mainland, and Tasmania. The lack of knowledge of navigation would have confined each section to the area on wliich it then found itself, whilst the subsetpient introduction of a new- race into the Australian continent would, on the assumption of those who liold the Australian aluiriginail to be a homogeneous A H(df-c(isfe Tasonanian. 1 7 race, have resulted in the extermination of the existing Papuan element in the mainland by the new occupants, the Australian aboriginal, or on the view of those who hold the Australian abo- riginal to be an admixture of Negro or Papuan stock with some other race or races, would, by cross breeding with the Papuan, form the present aboriginal race. Thus there would be a detached group of Papuans in the North, a detached group of Papuans in the south, and a; central group differing from the north and south groups, which is exactly what is found. Fur- ther, these races would be of great antiquity, though not neces- sarily of tertiary times, though even this is not impossible, whilst, lastly, the Tasmanian would have reached Tasmania by land in very distant periods. He would, therefore, be of a most ancient race ; he would have no knowledge of navigation ; he would differ from the inhabitants of the adjacent mainland of Australia ; and also from those of New Zealand. He would be closely allied to the natives of distant New Guinea, but would from his isolation amd from the apparent fact that he had not been visited by other and more recent races, retain his primitive manners, and show less signs of advancement than his brothers of New Guinea, all of which coincides in every detail with what we know of this unhappily extinct race, the extinction of which is a blot upon the fair history of British colonisaition. Even though this theory be accepted, the writer has no desire to claim any exclusive rights in, or priority for, the view, for it is merely a revised version of what many anthropologists have already put forward. Howitt (14) has long held the view that the Tasmanians came from the north, and has stated : — "I have long since come to the conclusion that one of the fonda- mental principles to be adopted in discussing the origin of those {Tasmanian) savages must be, that they reached Tasmania ait a time when there was a land communication between it and Aus- tralia." Bonwick (5) has stated that: --"The fact of the crisp-haired Papuans being found in islands all round the New Holland (Australian) coasts, and over so vast an extent of space, ought certainly to indicate their prior migration to that of the Aus- tralians."' Flower (18) has, iu his usual clear and terse way, summed up the whole theory in the words, " they (the Tasmanians; 18 Richard J. A. Berry: reached Van Dieman's Land, by way of Australia, lon^: anterior to the commencement of the comparatively high civilisation of those portions of the race still inhabiting New Guinea and the adjacent islainds, and also anterior to the advent of the existing native race, characterised by their straight hair amd by the pos- session of such weapons as the boomerang, throwing stick, and shield, quite unknown to the Tasmanian.'' De l;)uatefages (21) says: — In Australia there are two dis- tinct types — Australians prcj^^er and Australian Neanderthaloides — the latter a small group occupying the country about Adelaide, and having, among other characteristics, hair which closely re- sembles the woolly hair of the negro This fact can be accounted for by presuming that true negroes formerly oc- cupied the whole or a part of Australia ; that they were invaded by a black race with straight hair ; and that it is to a blood mixture that the differences in the hair must be attributed. It is probable that the Tasmanians furnished this negritic element. Their former existence in Australia has nothing about it which may not be verj^ natural, and their facial characteristics occa- sionally approximate closely enough to those of the Australians to allow of the probability of this hypothesis. An examination of the skulls of Australians with woolly hair from the Southern tribes would probably solve the question. Finally, if my con- jecture be well founded, we must admit that the crossing nmst have taken place at a very remote period, and that the woolly hair could only reappear more or less modified by atavistic phe- nomena.'' Of the objectors to am essential part of the theory, namely, that the Tasmanian ancestry first inhabited, or passed through the Australian continent on their way to Tasmania, Huxley (16) is the most important. He considers that it is " physically impossible that the Tasmanian could have come from Australia, and apparently the only way of accounting for the presence of the Tasmanian was to assume his migration from New Caledonia and the neighbouring islands. It would appear that at one time a low negrito type spread eastwards, and reached Tasmania, not by means of direct and uninterrupted land communication be- tween New Caledonia and Tasmania, but rather by means of broken land in the form of a chain of islands now submerged, A Half-caste 2\ismanian. 19 similar to that which at present extends between New Caledonia and New Guinea." In view of Hewitts subsequent work, to which reference has already been made, it does not appeair to me that Huxley's ob- jection to the land theory of the Tasmaniau migration, can any lonffer be regarded as tenable, and whether the present theory be upheld, or whether it be replaced by some other theory bajsed on a surer foundation of fact than is as yet possible, I am convinced that Howitt's view as to the migi-ation by land will eventually be found to be the correct foundation on which that theory will be built. I can. therefore, only conclude by expressing the hope that this epitome of so much that is at present mere visionary theory will lead to the accumulation of such a collection of material as will enable us to ascertain the true facts of the case, and that the credit of the discovery will belong, as it ought to do, to Australia and Australian scientists. WORKS REFERRED TO 1% THE TEXT. 1. Anonymous. " Gazetteer of the World." Edited by a member of the Royal Geographical Society Vol. iv., page 445. Edinburgh, London and Dublin ... 1856 2. Chisholm, G. 0. Longman's Gazetteer of the World." Edited by G. C. Chisholm. Page 771. London 1895 3. Admiralty Sailing Directions. Australia Directory. Vol. i., page 294. Ninth edition. London ... 1897 4. Hallack, E. H. " Kangaroo Island." Adelaide ... 1905 5. Bonwick, James, F.R.G.S. " Daily Life and Origin of the Tasmanians." London ... ... 1870 6. Ibid. " The Lost Tasmanian Race." London ... 1884 7. Flower. Journal of the Anthropological Institute 1885 8. Duckworth, W. L. H. " iMorphology and Anthropo- logy." Page 358. Cambridge ... ... 1904 9. Evans, G. W. " History and Description of the Present State of Van Dieman's Land." Page 19. London ... ... ... ... ... 1824 2a 20 Richard J. A. Berry: A Half-caste Tasmania a. 10. Prichard, J. C. "The Natural History of Man." Vol. ii., page 467. London ... ... . 1855 11. Wallace, A. R. "Island Life." Third and revised edition. London ... ... ... ... 1902 12. Roth, H. Ling. "The Aborigines of Tasmania. London ... ... ... ... ... 1890 13 Tylor, E. B. "On the Tasmanians as Representatives of Palaeolithic Man." Journal Anthropological Institute, November ... ... ... 1893 U. Howitt, A. W. "The Native Tribes of South-East Australia." London ... ... ... 1904 15. Topinard. xMem. Soc. d'Anth. Vol. iii., page 322 1869 16. Huxley, T. H. Journal of Ethnological Society. Vol. ii. ... ... ... ... ... 1870 17. Hutchinson, Gregory, and Lydekker. "The Living Races of Mankind " ... ... ... 1906 18. Flower, Sir W. H. "The Aborigines of Tasmania." Manchester and London ... ... ... 1878 19. Giglioli, E. H. "I Tasmaniani conni storice ed etno- logici di un pooolo estinto." Page 174. Milan ... 1874 20. Mathew, John. " Eaglehawk and Crow." London and Melbourne ... ... ... ... 1899 21. De Quatrefages. " Introduction a I'Etude des Races Humaines." Paris ... ... ... 1899 22. Spencer, W. Baldwin. Rep. Aust. Assoc. Adv. Sci., President's Address, Sec. D. ... ... 1892 DESCRIPTION OF PLATE L Photograph of half-caste Tasmanian woman from Kangaroo Island (on right) and her daughter, a quarter-caste Tasmanian (on left). Proc. E.S. Victoriii, l'J()7. Pate T. [Proc. Koy. Soc. Victoria, 20 (N.S.), Pt. 1., 1907] A in' II. — Xofr on the De/position of Bedded Tuffs. By T. S. hall, M.A. Melbourne University. (With Plate II.). [Read 11th April, 1907]. In many, it' mil most, of the places in South-western Victoria where tuft's are displayed they are well stratified. These tuffs a/re associated with the basaltic lavas which form such a featm-e of the g-eolog}' of the State, and are referable to various parts of the tertiary period. The volcanic rocks cover, according: to Sel- wyn, somewhere about 9000 square miles, or about a tenth of the total airea of Victoria. The tuffs with which we are at present concerned are usually fawn-coloured, and vary with considerable irregularity in the size of their constituent grains. They show a tendency to split into flags along their bedding planes, and are fairly coherent, 8o that they are used, as in the neighboux'hood of Camperdown, as a rough building stone. The decomposition of the tuft's and of the coarser scoria yields a remarkably rich soil, and the porous subsoil affords both natural drafinage and a capacity for storing up water. The tuffs are then commonly tree-clad. Among the plaices which have come under my notice where well bedded tuft's are to be seen are two belonging to the older volcanic series, which is here seen to underlie the marine Bar- wonian beds. One of these is at Curlewis, about eieht miles east of Geelong, where on the beach platform a continuously dip- ping series is seen extending for about four miles. The basalt of similar age at Airey's Inlet is also associated with well bed- ded tuffs which dip inland from the shore, pointing to the old vent having been out at sea. Among the more recent tuffs which also show this character may be enumerated those of Mount Leura and Bullenmerri, near Camperdown. These beds cover a very large area. On the 22 T. S. Will: flanks of Tower Hill, as the railway runs d()\\^l through cuttings towards Warrnanibool, well stratified tuflEs are extensively dis- played. There is a similar well bedded tuff, though only a few feet in thickness, overlying the Kalininan a-t "McDonald's" on Muddy Creek, near Hamilton. Quite recently I have seen many square miles of equally well bedded tuffs about Mount Gambler in the south-east corner of South Australia, and to these atten- tion will be more fully directed later. Similar, though usually obscure, bedding is shown in the scoria on the Hanks of Mount Leurai. The great banks have been extensively worked for many years for railway ballast and for covering footpaths, so that ever changing sections were displayed. Though bedding appears extremely common in the tuffs, it is not universal, and I call to mind a section shown in a road cut- ting near the Park gates at Camperdown where bedded tuffs show a faulted contact with unstratified ones. The bedding is generally of such a well marked character that rough flags can be quaa'ried almost everywhere. The question arises, To what is this bedding due I Was the deposition subaqueous or merely subaerial? Till recently 1 never thought of the possibility of anything but subaerial de- position being suggested in most of the places mentioned. Pro- fessor J. W. Gregory^ holds that the beds round Camperdown are of subaqueous origin, and that the stratification is due to sorting by water. Of the correctness of this view I have doubts, and I have lately found evidence at Mount Gambler w^hich shows that well bedded tuft's may owe their stratification to subaerial sorting, and hence no reason exists for calling on large lakes or the sea to explain their character. Tte assertion of subaqueous deposition for all these tuff's would demand the existence of large bodies of water, either marine or fresh water, extending over very wide areas and at various periods. The well-bedded tufl's of Ciu'lewis and Airey's Inlet are older tertiary age. Those of K(u-odt, Muddy Creek and Mount Gambler are recent. The BuUenmerri tuff's may be pleistocene. The supposed subaqueous deposition of the tuff's of 1 Geojjfi-aphy of \icloiia, p. 128. Deposition of Bedded Tuffs. 23 various a^es and localities is the nuly evidence in favour of the former existence of these seas or lakes, and that, too, at times in peculiar positions in reference to the modern surface drainage. Then, again, the tuffs, though stratified and very evenly bed- ded, are not of uniform grain. Taking a small piece, stratifica- tion is not evident. Dust and small scoriaceous fragments seem mingled in confusion. It is the fine matter which brings about the fissility, and yet the amount of coumiingled larger material is considerable. There are, of course, well marked, beds of dust, aflid equally well marked ones of coarse grain, but to my mind the sorting is not as thom-ogh as it would have been had water — that is, standing water —been the c-ause of the beddiniJ'. Mud torrents have been suggested, but they also are, I think. out of the question. Such torrents wcmld, if they formed strati- fied deposits at tiill, show false bedding, and not layer after layer through a thickness of many feet, and a lateral measure- ment of scores of yards. However, apart from this, we have at Mount Gambier evidence which, I think, shows clearly that tuffs, as well stratified as any of those of the Camperdown district, may occur under con- ditions which forbid aqueous action. A brief .sketch of the geology of the district is necessary to enable this evidence to be properly weighed. Tlie bed-rock over hundreds of square miles is a white lime- stone mainly composed of polyzoal remains. This is of Bar- wonian age ( ? Eocene;, aind is, as far as can be seen, quite horizontal. It is extremely porous, and water-courses are absent. There are, of course, many swallow holes, caves and underground drainage channels, so that many of the irregularities of the sur- face are undoubtedly due to subterranean solution. This point may be considered unfavourable to my view, so that I wish to be properly considered. The general surface of the country is slightly undulating, and the hills to the north of the town are, for the most part, sandy. They are, in fact, sand dunes of pleistocene or recent age, and vary someAvhat in the amount of lime they contain, and consequently in the amount of con- solidation they have undergone. In places they are loose yellow sands, and in other places consist of the ordinary cross-bedded dune-rock. A lew miles to the south of the town similar dime- 24 T. S. Hall: rock fi>riiis the surface, and the typical form of the cups so characteristic of dunes is easily traced. In other places we find lonandy soil the (luiekliuie produced a heavier crop than any other Plot Clay 0 1291b. 1 1561b. 2 1441b. 3 1541b. 4 1211b. 5 1471b. 6 1321b. 7 1361b. 8 1431b. 9 721b. Soluble Constituents In Alluvial Soil. 41 manure, probably becaaise of its solvent chemical action on the mineral constituents of the soil. The superphosphate, slaked lime, nitrate of soda and blood manure seemed to be ecjuivalent as regards the sandy soil, but the two former were more effi- cient in the clay soil. The star phosphate and bone-dust exer- cised iii stronjjf action on the clay soil, but none on the sandy soil, while the g;ypsum reduced the yield on both. The order of value for the manures, in the proportions given, are as follows: — For the sandy soil — (1) Quicklime ; (2) superphosphate, slaked lime, nitrate of soda, blood manure ; (3) ammonium sulphate ; (4) star phospha/te, bone dust, and no manure; (5) gypsum; For the clay soil — (1) Slaked lime, star prosphaite ; (2) bone dust, nitrate of soda, superphosphate ; (3) blood manure, sul- phate of ammonia; (4) unmanured and gypsum : (5) quicklime. In such cases as these no analyses of the soil, of the crop, or of the manure would enable the results of the application of the latter to be predicted, hence it is essential that the farmer should be guided by local tests rather than by general principles, which are often misleading if improperly applied. Herein lies one of the chief justifications for the existence of experimental plots on farming land throughout the State. a.iid one of the reasons for the avoidance of too much centralisation of experimental field-work in one locality. A point of great interest is to compare the above data with the fluctuations in the soluble water, and of the humus in the soil. The soil saimples were taken from the surface and from depths of 8 and 16 inches, weighed, dried, weighed again, and soaked in 2 litres of distilled water per kilogram of soil. In the first experiments, the clear filtered liquid was boiled down to a small bulk at Burnley, and sent to the University for final testing. Tiie escape of the dissolved carbon dioxide, and the concentration caused, however, a considerable loss, so that all the soil samples were sent to the University, there extracted and fliteied. One-half of the liquid added to the soil was evaporated in the same vessel in which the residues were weighed. The total number of soil samples exceeded 400, and the weight nearly hall a ton. Some idea as to the prevailing conditions in regard to mois- ture and temperature is presented by the following data, giving the percentage of water and the temperatures at dift'erenr depths 42 Alfred J. Ewart : during the median portion of the year. From September on- wards the temperature became higher, and the percentage of soil water decreased, especiaiUy in the upper layers of the soil. Both soils were wettest in July, driest in December. Date Percentage of Water. Clay Soil Fine Sandy Soil Top Sin. 16in. May 18 - 18.2 17.5 17.2 June 20 - 7.75 12.2 11.1 July 20 - 44.5 42.5 44.5 Aug. 25 - 24.25 25.25 23.5 Sept. 25 - 19.0 20.5 22.5 Temperature (Fahr. ) of Each In Situ. Clay Soil Fine Sandy Soil Top Sin. 16in. Top Sin. 16in. Juue 20 - 50 deg. 49.5 cleg. 47 deg. - 51 deg. 50 deg. 51 deg. July 20 - 46 „ 45 „ 45 „ - 41 „ 43 „ 43 „ Aug. 25 - 50 „ 49.5 „ 51 „ - 51 „ 50 ,, 51 „ Sept. 25 - 66 „ 61 ,, 60 „ - 65 „ 60 „ 56 „ Great difficulty was found in obtaining clear watery filtrates from the clay soil without filtering through biscuit porcelain, whicli is tedious with large bulks, and is apt to cause the loss of some of the materials really held, originally, in solution. The acid extracts filter readily, but this part of the work was confined to the sandy soils, «ince the residues from the watery extracts of the clay soil are not at all reliable. The following are two sets of data from the clay soils in May and September, i.e., before and after the main rainfall, the numbers giving the amount of matter dissolved by 2 litres' of water from 1 kilogram of dry soil, plus the amount of non- settling suspended matter able to pass through doubled filter-paper: fop 8 ;iii. Sept. '25. l( ym. May 5. Sept. 25. May 5. May 5. Sept. a.n. Plot 0 0.91) 0.72 0.88 0.69 0.76 0.98 .. 1 0.76 0.61 0.45 1.16 0.49 1.11 2 0.56 1.12 1.11 1.23 1.25 1.21 » '-^ 1 28 1.24 1.31 1.13 1.28 1.15 „ 4 0.81 0.92 1.32 1.38 1.41 1.18 >> 5 0.72 0.64 0.85 1.14 1.25 1.28 „ 6 1.25 1.28 1.11 0.72 1.37 1.38 » 7 1.2 0.72 1.21 1.28 1.28 1.12 ,, 8 1.15 1.28 1.16 0.76 0.96 0.92 „ 9 0.89 0.52 1.13 1.18 1.19 0.81 A verage 0.9(5 0.90 1.05 1.07 1.12 1.11 Soluble Constituents hi All u rial Soil. 43 The data are of value simply as showin cwt. per acre, it is evident that by the time the manures are distributed through the upper 18 inches of soil the anuiunts per kilogram will be too small 2 tons 1201b. 4c wt. 1 ton icwt. 8011.. 2cwt. 2cwt. 2 tons 5.(5 0.148 0.56 2.8 O.oG 0.1 0.28 0.28 5.6 0.70 0.016 0.07 0.35 0.07 0.012 0.035 0.035 0.7 44 Alfred J. Ewart : to perceptibly affect the amount of the water-soluble, and still less of the acid-soluble residues per kilogram. Fluctuations in these of less than 0.02 to 0.01 of a gram appear to be mean- ingless, or, at least, to result from fluctuations or conditions beyo*d control, such as slight differences in the drainaige, in the fineness of the ^oil, and in the slight unevenness in the distri- bution of the vital, physical and meterological conditions which affect it during the period of observation. Even when first ap- plied, and distributed at a depth of 2 inches, the nitrate of soda and ammonium sulphate are barely present in sufficient amount to appreciably affect the soluble extractions from the soil by the method of partial lixiviation (2 litres of water to 1 kilogram) employed. By this method is determined merely the amount of soluble material immediately available for a0f. -JO UnuiainirtHl Top 0.31 0.18 0.41) 8 inches 0.48 0.35 0.48 16 inclies 0.18 0.6S 0.41 Avei'ajj'e 0.42 0 40 0.4(J Soluble Constituents in Alluvvd Soil. 45 The effect of the hciny rains i well shown in the December result. The increased average probably partly results from the attraction of soluble matter from still deeper layers, and partly from nitrification in the soil during warm dry wearther, after the removal of the crop and the absence of rain allowed the nitrates to accumulate. May 18 Sept. 25 Dec. 20 Plot 1.— Air-slaked Lime Surface 0.53 0.41 0.48 (Two tons per acre) 8 inches 0.32 0.40 0.40 16 inches 0.12 0.48 0.52 Average 0.32 0.43 0.47 This plot AVas obviously poorer originally than the unmanured one, the high value for the top layer on May 18 being direct- ly due to the addition of liiue. This appears to keep the soluble matter more uniformly distributed in the upper layers of the soil, and also to cause a greater increase in the December aver- age than occurs in the unmanured plot. The September aver- age also show>; a strong increase, in spite of the presence of a growing crop. May IS Sept. 25 Dec. 20 Plot 9. —Quicklime Surface 0.94 0.ti4 0.4 (Two tons per acre) 8 indies 0.42 0.5 0.4 IG inches U.3B 0.4 U.46 Average 0.57 0.51 0.42 The chemical action of the quicklime results in a liberation of soluble constituents (potash, etc.) in the superficial layers, but so much of this is ultimately washed away thart; the Decem- ber average is below that for the preceding plots. The soluble matter is, however, kept evenly distributed, as in the case of the previous limed plots. The use of powdered quicklime for direct application to the soil is coming into vogue in English agri- culture, a Birmingham firm manufacturing large quantities of a phosphatic and niagnesiau powdered lime for agricultural pur- po!-e.s. The lime needs to be drilled in as though it were so much seed, some time before the crop is planted, and its purpose is obviously to render a large amount of soluble matter im- mediately available tor the use of young seedings. It is evi- dent, however, that the quicklime will be apt to exercise an 46 Alfred J. Ewart : exhausting action on the fertility nf the soil, especially in regions with a high rainfall. Quicklime is often stated to ha"ve a special power of burning out humus from the soil, especially if applied at the rate of one or more tons per acre. This is quite incorrect as regards the ordinary uiode of application of lime in Agriculture. Quick- lime from the kilns, if directly put into the soil, would be rather injurious thaai useful, since all lumps of any size would retain their causticity in the soil long enough to delay seeding, and by their local action would result in very patchy cultivation. In ordinary practice, to secure fine subdivision and even distribu- tion, lump lime must be allowed to slake in heaps on the sur- face, which, when the lime has crumbled down, can be scattered and harrowed in. Dui'ing this process the whole or the grearter part of the lime is converted into carbonate of calcium by the carbon dioxide of the air and soil. The presence of a carbonate of an alkaline base or alkaline earth is one of the conditions for the continuance of the nitrification of humus in the soil, the nitrous and nitric acids produced displacing the carbon dioxide from the carbonates in the soil. In this way the ac- cumulation of acid, which is fatal to further nitrification, is prevented, but it must be remembered that strong alkalies like quicklime are very nearly as injurious to the nitrifying ■and other soil bacteria as are free mineral acids. Hence we should expect to find that the direct application of quicklime would, for a time at least, result in a lessened bacterial oxida- tion of the humus in the soil, as is in fact shown by the follow- ing results, giving the percentages of humus by weight in the dried soil of the various plots at three depths, at the beginning and close of the exoeriments : — Percentages of Humus (line sandy soil). May 5 Sept. 17 Dee. 20 Plot 0 — -Uniiianurecl Surface 1.(3 1.85 1.8 8 inches 1.6 1.7 1.75 16 inclies 1.4. l.« 1.3 Averag-e 1.63 1.65 1.62 Plot 1 — Air-slaked Lime Surface 1.55 — 1.55 (Two tons per acre) 8 inches 1.15 — 0.9 16 inches 1.3 — 0.65 Average 1.33 — 1.03 2.0 — 2.4 1.35 — 1.75 1.5 — 1.45 1.62 — 1.87 l.S — 1.85 1.9 — 1.8 1.9 — 1.8 1.87 — 1.82 Soluble Constituents in Alluvial Soil. 47 Plot 9 —Quicklime Surface (Two tons per acre) 8 inches 16 inches Average Plots 2, 3, 5, 7, 8 Surface (Bulked Average) 8 inches 16 inches Average In all the plots, the fallen dejecta membra of the crop tend to raise the percentage of humus in the surface layers. The air- slaked lime produced a pronounced fall in the pereentace of humus, which, however, increased at a depth of 8 inches with the quicklime, and to a) slight extent also in the unmanured plots. At 18 inches, the humus decreased in all cases, though only to a slight extent, except where air-slaked lime was applied. It must be remembered that the quicklime plot produced the heaviest crop, so that the increase in the percent- age of humus is, to a slight extent, due to the greater develho\vs that the snlvtMit actions in the soil more than hahmced the loss by drainatre and by the crop. May is Plot 3— Star Pliosphate Surface O.'j!) (4cwt. per acre) 8 inches 0.34 1() inches 0.48 Averao^e 0.47 The manure beintr only sparingly soluble, the steady decrease of the averages probably represents soil constituents previously present. These show the usual drop after rain in the surface layer, and no absolute, but only a relative increase or lessened decrease on the surface after dry weather. May 18 Plot 5 — Done Dust Surface 0.51 (4cwt. per acre) 8 inches 0 50 16 inches 0.63 Avera<^e 0.55 This resembles the preceding closely, except thart the concen- tration on the surface in December is better showTi. Both cases indicate an exhaustion of the soluble soil constituents by the crop or by drainage, for if the manures fixed or precipitated the soluble constituents, the first surface ei?timation in May would be a low instead of a high one. May 18 Plot 8 — Calcium Superpliosi^hate Surface 0.54 (2cwt. per acre) 8 inches 0.57 16 inches 0.56 Average 0.55 Although the manure is highly soluble, and the amount of it not too small (0.28 gram per kilogi'am of upper 2 inches when first applied), it does not seom to produce any pronounced direct effect upon the changers in the distribution of the soluble matter in the sdil. which resemble those in the pi'eceding plot. Sept. 25 Dec. 20 0.32 0.44 0.43 0.41 0.52 0.38 0.42 0.41 Sept. 25 Dec. 20 0.36 0.46 0.51 0.40 0.52 0.44 0.46 0.43 .May IS Sept. 25 Dec. 20 io'd. 51 \voi;v 4.84 (4.50) 4.70 (4.29) 4.48 (4.08) The fluctuations at different depths might possibly be the result of imperfect sampling, which is ailways of great import- ance, however homogeneous the soil may aippear to be. The averages, however, show a steady decrease. It is always possible that soluble material from one layer may continually diffuse towards another layer in which it is deposited by some kind of chemicajl precipitation, or as a result of evaporation, or the loss 52 Alfred J. Lhrart : of a solvent gas. This may be the case here, although the fluctuations in the percentage of acid -soluble and water-soluble matter show no apparent relationship. The latter represents, however, merely the condition at the time of taking, whereas the former results from cumulative action prior to sampling. iMa.\- 18 Sept. 25 Dec. 30 Plot 6 — Ammonium sulphate Surface 0.55(6.12) 5.64'(5.24) 6.16(5.82) (80lb. per acre) 8 inches 6.34(5.81) 5.7 (5.21) 5.7 (5.34) 16 inches 6.25(5.7) 6.32(5.91) 4.8 (4.56) Averag-e 6.38(5.88) 5..S9(5.4o) 5.55(5.24) Here the acid- soluble matter undergoes on the average a dis- tinctly greater reduction than in the previous plot, in spite of it-^ lesser yield of crop. The variations closely follow those in plot II., so that it appears as though acid-soluble matter passes downwairds to some extent after prolonged rain, and is slowly drawn upwards during prolonged drought. May IS Sept. 25 Dec. 30 Plot 8— Star Phosphate Surface 5.36 (4.77) 5.06 (4.69) 4.36 (4.0) (4cwt. per acre) 8 inches 4.98 (4.64) 4.76 (4.23) 4.8 (4.36) 16 inches 5.40 (4.92) 5.18 (4.76) 4.72 (4.28) Average 5.25 (4.78) 5.0 (4.56) 4.63 (4.21) May 18 Sept. 25 Deo. 30 Plot 5— Bone Dust Surface 5.3 (4.79) 5.2 (4.68) 5.7 (4.26) (4cwt. per acre) 8 inches 4.92(4.35) 4.86(4.43) 4.58(3.87) 16 inches 4.96(4.33) 4.82(1.3) 4.04(3.66) Average 5.06 (4.36) 4.96 (4.47) 4.67 (3.93) In plots III. and V. the ajmounts of manure added represent 0.56 gram per kilogram of the surface 2 inches, and 0.07 per 18 inches depth. Since only a portion of -each is soluble in the the dilute acid used, it is evident that the steady fall represents mainly materials previously present in the soil. In the i-tar phosphate plot no superficial accumulation of acid-soluble materials takes place in December, but this phenomenon is shown with the bone dui^t plot, possibly because of secondaiy reactions excited by the organic material of the bone dust. May IS Sept. 25 Dec. 30 Plot 8.— Calcium Superphosph. Surface 6.08(5.24) 5.61(5.25) 4.1(3.62) (2cwt. per acre) 8 inches 7.96(7.39) 7 2 (6.69) 4.4(4.0) 16 inches 5.86(4.66) 5.22(4.5) 4.2(3.74) Average 6.63(5.76) 6.01(5.58) 4.2(3.79) Soluhlc Coiisfitat'iifK ill All a rial Soil. Oo The amount of uiamirf added represents 0.28 ^rain i>er kilo- pram of the superficial '_' iiahes. and the hii^h soluljility explains the high iuaiount of water-soluble matter in the superlicial layers on May 18. At the same time the superphosphate seems to exercise a stronu; solvent action on the soil itself, of such character as to strongly increase the amount of matter soluble in dilute acid. The net result is to leave the soil distinctly poorer in matter soluble in dilute acid than any of the previous plots, but if this matter is utilised l)y the crop it represents a.n increased yield and profit instead of so much inert material. Un- fortunately Imt little appeared in the crop as ash, so that super- phospliates by themselves appear to be highly wasteful and exhaust- ing. If these facts hold good for soils in general, the exclusive use of phospliates in Victorian agriculture is likely to rapidly exhaust the fertility of the soil, not only because of the greater demands of the increased crop, l)ut also because the superphosphate exercises a secondary action on the soil, temporarily increasing the amount of water-soluble matter and ako that of materials soluble in very dilute acid. The former are rapidly and the laitter slowly removed from the soil by the action of rain water charged with carbon dioxide as well as by the crop. May 18 Sept. 25 Dec. 20 Plot4-Gypsum Surface 5.2 (4.08) 4.0i (3.68) 4.2 (3.6G) (One ton per acre) H inches 4.4 (3.67) 3.9 (3.18) 3.5 (2.78) Ifi inches 4.15 (3.o(3) 3.4 (2.29) 3.1 (2.19) Average 4.58 (3.77) 3.78 (3.05) 3.6 (2.88) Ma\ 18 Sept. 25 Dec. 20 Plot 0 -No Manure Surface 3.32(3.01) 3.9 (3.72) 4.58(4.09) S inches 3.90 (3.42) 3.4 (3.05) 3.44 (2.96) 16 inches 3.82 (3.34) 3.7 (3.02) 3.06 (2.65) Average 3.68 (3.26) 3.7 (3.26) 3.69 (3.23) Not only did the gj'psum produce a less crop yield than the unnutnured plot and unduly increase the amount of water- soluble niiUerial liable to waste by drainage, but it also ca.used a pronounced decrease in the amount of acid-soluble material by the end of the year. In the unmanured plot the averages remain very nearly constant in spite of an increase towards the surface and a decrease in the deeper layers. 54 Alfird J. Evart: May IS Sept. 25 Dec. 20 Plot 1— Air-slaked Lime Surface 68 (5.77) 5.08 (4.67) 5.44 (4.96) (Two tons per acre) 8 inches 4.6 (4.18) 4.68 (4.28) 5.12 (4.72) 16 inches 4.88 (4.76) 4.46 (3.98) 4.0 (4.48) Averas?e 5.23 (4.0) 4.74 (4.31) 4.85 (4.72) Plot 9— Quicklime Surface 7.38 (5.44) 7.66 (7.' 2) 7.08 (6.68) (Two tons per acre) 8 inches 6.34 (5.92) 6.7 (6.2) 5.52 (5.12) 16 inches 6.38(6.02) 6.5 (6.1) 5.6 (5.54) Average 6.7 (6.13) 6.95 (6.44) 6.U7 (5.78) The acid-soluble materials appear to increase in the quick- lime plot in September, and are throughout high. This is undoubtedly due to the chemicail action of the quicklime on the soil, and this action involves a considerable waste of food materials by the end of the year. This waste is not apparent in the case of the slaked lime plot, in which, as in the unma- nured plot, the total amount of readily soluble mateiial appears to be greater in December than it is after the winter rains in. September. .In all the other plots tlie amount steadily decreases towards the close of the year. For convenience of reference a joint table is given beneath showing the arverages from the upper 18 inches for all the plots in parallel columns. From the totals it appears that the manure added represents the asli of the crop fourfold, and that the total apparent loss from the soil was four times greater than the amount of chemical manure added. In other words, chemical manures do not permanently enrich, but rapidly impoverish, tine soils poor in humus, especially when applied in excess. SUMMARY. As regards the unmanured plot, the sodium nitrate appears to lower the percentage of dissolved matter during the eight months following its applicajtion ; the slaked lime lowers the amount at first, but by the end of the year it is up to the nornuil level again. In all other cases, the amount of water- Boluble matter is increased at first, and lowered below the un- manured level by the end of the year, except in the case of the gA'psum, in which it remains high in spite of a heavy loss, and of the blood manure, in which it does not fall below the amount in the unmanured plot. Sulnhlc doni^tlfUA'iifs III All II rial Soil. 55 In rt'LMrd tn the distribution of the soluble matter in the different hiyers, the dowmvard movement, as the result of con- tinuous rain, and the less marked upward ascent during drouirht. were well shown by the unmanured, slaked lime, gyp- sum, nitrate of soda, bone dust, .-ind sujxTphosphate plots, while the downward movement was nuM-ely lessened in summer, and not actually reversed in the case of the quicklime, ammonium sulphate, star phosphate, and blood manure plots. The averages for the water-soiuVjIe matter in all the plots are remarkably consistent with the conclusions mentioned, the surface showing a fall and I'ise, 8 inches depth a slow fall, and the 16 inches an. almost similar rise by the end of the year. * AvKKAUE Water-soluble Matter for all the Plots. May IS Sept. 25 Dee. 20 Top O.Ol 0.8!) 0.43 N inches 0.-49 0.48 0.45 16 inches U.4() U.47 0.49 The amount of matter soluble in dilute acid underwent a secondary rise at the end of the year in the superficial layers in the case of the unmanured, sodium nitrate, ammonium sul- phate, bone dust, and slaked lime plots, but in all other cases decreased steadily in the surface- laij^ers of soil. The quicklime plot was exceptional in ^showing not only a rise in the superficial layer in September, but also an increase in the average for the whole 16 inches, followed by a pronounced fall in December. As regards the averages, these decreased during the year in all the plots excepting the unmanured and slaked lime [ilots, but the increases in these were very slight in amount. AvERAUE Matter Soluble ix Dilute Acid from all the Plots. May 18 Sept. 25 Dec. 20 Surfac>' - - 4.82 - 4.69 - 4.5G 8 inches - 4.88 - 4.52 - 4.04 10 inches - 4.69 - 4.41 - 3.76 The average acid-soluble matter decreases steadily on the surface and more rapidly in the deeper layers, especially during the summer months. Without claiming • anything more than a preliminary and suggestive value for these data, they nevertheless may be taken 56 Alfred J. Ewart : to apply to the fine alluvial soil occurrinir on so many river flats and valleys in Victoria. None of the manures as ajDplied in the somewhat excessive quantities given would have paid for their application by the increased crop yield. It will fur- ther be noticed that the apparent loss from all the plots except- ing the slaked lime and unmanured ones (where there is a slight gain) is very much greater than can be explained by the ash removed by the crop. To what extent these apparent losses are real ones, and to what extent they are due to decreases in solubility, to increases in absorptive power, or to changes in distribution, must be reserved for further investigaitinii. As far as they go they seem to show that the fine alluvial, -^andy river-flat soils widely cultivated in many parts of Victoria, ap- pear to be peculiarly liable to exhaustion under the action of all chemical manures excepting slaked lime. Even if the maxi- mal apparent loss (851b.) were entirely a real loss, it wnuld take 530 years to remove the top 16 inches, assuming that aul was removed by solution and none by erosion. These soils appear also to be comparatively deficient in humus, but uliei-e this is not so the manurial diagnosis in the case I'f virgin soil would be slaked lime at the rate of h to 1 tim per acre, until the soil begins to sliow signs of exhaustion, then farmyard manure at a probable minimum of 2 to 5 tons per acre, soluble, nitrogen- ous, or phosphatic manures to be used sjjafl-ingly, or not at all unless thie soil shows need fur them. PRACTICAL AXIOMS. Quicklime binds a clay soil, slaked lime ameliorates^ it. Quicklime in excess exercises a v\^asteful solvent action on composite sandy soils. Small (quantities drilled in prior to seeding should, however, stimulate the early growth of seed- lings, ajnd perhaps lessen the danger of infection by fungi. The indirect action of a manure on the soil is usually much more important tha.n its direct chemical value as a nutrient substance. This applies not only to those manures which exert a direct chemical action on the soil, but also to thostr nitrogen- containing, acid or alkaline manures which -affect the activitv of the micro-organisms in the soil. Soluble Constitiienfst hi Alia rial Soil. 57 ClieiJiical lUMiiures, especially solul>le phosphatic ones, should not be applied in any quantity to soils poor in Imnius, except in conipany with farmyard manure or some form of humus. The soil is a changeaible matrix, whose percenrap:e snluiiilitv in water and acid varies appreciably at different depths tlir()U^easonal changes as a whole, especially under the action of chemical manures. The apparent losses from the soil after heavy manuring are many times greater than the ash contained in the crop, and also gi-eater than the amount added to the soil by any of the chemical manures used excepting lime. The oxidation and nitrification of humus in the soil is more favoured by air-slaked lime tlian by the direct application of quicklime, so long ais the latter retains any alkalinity injurious to nitrifying micro-organisms. Postscript. Since the above was written, Mr. Hall has drawn my atten- tion to the fact that Norman Taylor, in Maclvor's " Chemistry of Agi'iculture," 1879, p. 224, suggested that the superficial limestone deposits common in the Mallee may have been pro- duced by the continued drarwing up of chalk in solution by Ciupillary action from the moister layers below. This ex- planation was adopted by Howchin for the desert limestones around Adelaide, and was extended by Gregory (" Geography of Victoria," 1903, p. 93) to the hard siliceous superficial cherts or quartzite beds, and also to the ironstones of superficial gold deposit>'. Recent research has, however, show-u that iron bac- teria may play a most important part in the formation of iron deposits wherever water is present, and such deposits will, in the first instance, be superficial. In any case, the data obtained by me are insufficient to do more than establish the fact that uuiterial may ri^e to the surface in dry weather, but say nothing as to whether wet weather may not waish it down again to an equally great extent. The alkaline ash left after bush fires would certainly tend to carry silica downwards again as soon an anv rain fell. 0» A.ifrea J. jbivart: AUuviai isoii. •ui .10 asuajoafi • HI .10 9sisaja3(j — CO ^ ^ X rt i-H 5^^ O ^ rH '^ rH 35 d d + + O —I 'H rt or •* ■* ^ ec -t lO ^ o o -p CD 05 s 00 03 CO g 1^ X 1^ 2 s CO -f ■* Tfl CO •^ lO -^* i.O •■O 4i O rH M T}< ■* |> odd rH 00 ># CO d d o CO T— 1 CO CO C-l CO Ol CO __l ^ T? ■* -t Tf< 1^ -T< T}< lO » o O o o o o d o o o -*■ ^[od -bs jiad •sr([ ut usti do.i.i 3[od •lis .i.xl 'Sql 111 pi^i.C clo.10 indapsaq.mi 9t aad ■J^o[iji .1 3 d s m B J o H^tdap saqoiii Z .lad "iioiiji J a d s HI ii .1 o ^-^ c« • O ^ 4^ s -H O .t; s X PM M o J 02 M O S ^ K! m o o CO O o o !>• C5 o o O o o o o o c ^ [Proc. Roy. Soc. Victoria, 20 (N.S.), Pt. I., 1907.] Art. VI. — Fossil Fisli Bemaivs from the Tertiaries of Ausfralld. Part II. By F. chapman, A.L.S., etc., Palaeontologist, National Museum, MeU^ourne, AND G. B. PRITCHAPvD, F.G.8., Lecturer on Geology, Sec, Working Men's College, Melbourne. (With Plates V.-VIII.). [Read 13th .June, 1907.] Contents. PAGE I. — Introduction -------- 59 II. — Description of Species ------ 60 III. — Range in time of the Genera - - - - - 71 IV. — Distribution of Species - - - - - - 72 V — Bibliography - - - - - - - - 73 VI. — Corrigenda to Part I - - - - - - 74 VII.— Explanation of Plates - 74 I.— INTRODUCTION. The present paper is ^\Titten as a continuation of the section previously j^ublished,! which dealt entirely with our Tertiary selachiana included in the ^ub-order Asterospondyli. We now describe one of our remaining selachians, as well as the chimae- roids and two members of the Actinopterygii, ailtogether repre- sented by seven species. We have also many specimens of ver- 1 Proc. Ro.v. Soc. Vict., \ol. x\ii., ii.s., pt. i., 1H04, pp. 207-297. 60 aiui.piiian and Pritchar. Pritcliard Coll.), P>eaumaris, Port Phillip, base of the Kalimnan (Nat. Mus. Coll., pres. by the late W. B. Jennings, and Goll. by the late W. Kershanv ; also G. B. Prit<.^hard Coll.). Fossil FisJi Remains. 63 Observations. — As in the case of tlio Diddon presently to Ije described, we are also fortunate in havinp; so complete a series of these fish remains, and although we lack many completely preserved specimens, there is a large number of fragmentary specimens, and these have materially assisted in the ehicidation of their characters as a whole. In all, we have examined about 35 examples of this form. Judging by the general characters shown in the previously illustrated examples of F-daphodon teeth, our forms show some slight divergences from typical specimens ill the number and arrangement of the tritors, apparently mak- ing an approach towards Chimsera.* The recorded range of Edaphodon is Cretaceous to Oligooene. It does not appear to have been noted before as occurring in the rocks of the southern hemisphere. Comparing our fossils with specimens of Edaphodon in our Museum from British localities, and with various descriptions and figures available for our purpose, the nearest ally appears to be Edaphodon bucklandi, Agassiz''^ from the British Eocene ; but this species is a larger form, is more heavily built, and the tritors have a coarser structure. Genus /si'/iyodus, Egerton. Ischyodus mortoni, sp. nov. (Plate YI., Fig. 6.) Description. — The specimen to which we append the above name is part of a left mandibular tooth having a very elongate and robust character, and with a remarkable concavity towards the upper j^art in the posterior region. The extreme anterior portion of the tooth is missing, but the posterior is fairly com- plete, showing part of the smooth superficial bonj' layer of the posterior margin. This marginal layer is marked by a series of slightly undularting thread-like ridges, which are most distinct on the inner third of the i^urface, a stronger ridge marking off this area. It is, moreover, convexly rounded off where it meets 1 Compare diafrrains in Smith Woodward Cat. I'oss. Fishes (Brit. Mus.), pt. ii. 1891, p. u4. •2 Poiss. Foss., vol. iii., 1843, p. 351, pi. .\1., a, figs. 1-4, 9-12, 19-24 ; also E. eurvsrna- thus, Ajf., Dixon, Koss. Sussex, 18.50, p. Ill, pi. x., figs. IS, 19, 22, pi. xii., f. 5. 64 Chapman and Pritcltard : the symphysial facette. The width of the la;tter is 8 nun. near the posterior margin. A narrow elevated ridge runs along the symphj^sial facette about 5 mm. from the margin, making an angle of about 30 deg. with the posterior layer. The tritors consist of very coarsely tubulated material, a.nd in this speci- men have been almost entirely weathered out, only a thin layer remaining, but sufficient t > show the presence of three very long, narrow tritors on the posterior portion of the tooth. Dimensions. — Heiyht, 30.5 mm.; greatest thickness, 16 mm. Locality and Horizon. — Table Cape, Tasmania. From the "Turritella beds." Jan Jukian. Tasmanian Museum Collection, Hobart. Observations. — The above species shows a striking similarity in its general characters to Ischyodus egertoni. Bucklandi from the British Jurassic, but is much' narrower and more elongate, with long narrow tritors showing a more marked parallelism. The genus Ischyodus ha's; been previously recorded from the Tertiary rocks of Amuri Bluff, New Zealand, and identified with Agasisiz's I. brevirostris by E. T. Newton.- The latter species belongs to the Lower and Upper Cretaceous of England. It is unfortunate that a reference to the Amuri Bluff deposits as Greensand'^ should have been made and pei'petuated. consider- ing that the faima is so distinctly tertiary. Subsequent to Newton's description of the New Zealand specimen, three others have been recorded under the same specific name by J. W. Davis, from Amuri Bluff."* It is just possible, however, that since these speciniens are more or less fragmentary, further material may show the New Zealand form to be aj distinct species. With regard to the name I. brevirostris, A. S. Woodward has already shown' that I. thurmauni, Pictet and Campiche. has priority over it. 1 Chimaera ei^ertonii, Buckland. Proc. Geol. Soc. Lond., vol. ii., 1S35, p. 206. Chiinoera (Ischyodus) egertoni, Agassiz. Poiss. Foss., vol. iii., 1S43, p. 340, xl.c, figs. 1-10. 2 Quart. Jour. Geol. Soc, vol. xxxii., 1876, p. 320, pi. xxi., flg. a. 3 Hector, Handbook of N. Zealand, 1883, p. 31 (referred to as Lower Greensand).— Newton, Q.J.G.S., vol. xxxii., 1876, p. 326 (Lower Greensand).- Davis, Trans. R. Dubl. Soc, vol. iv., ser. 2, 1888, p. 42 (Cretaceo-tertiary). Sniitli Woodward, Cat. Foss. Fishes, Brit. Mus., pt. ii., p. 68 (Greensand). 4 Trans. R. Dubl. Soc, vol. iv., ser. ?, 1S8S, p. ii, pi. \ ii.. tigs. 10-13. 5 Cat. Fossil Fishes, Brit. Mus., pt. ii., 18l»l, \>. iu. Fossil Fish Reviains. 65 The Tasnianian specimen, collected from Table Cape, was pre- sented to the Tasmaniam Museum, Hobart, by Chas. Mackenzie, Esq. We owe the opportunity of dealing: with this specimen to the courtesy of Prof. W. B. Spencer, C.M.G., who received it from Mr. Morton, the curator, after wliom we have much plea- sure in nauiin«i' it.^ Order .\CT1N0PTERYGII. Family Labridae. Genus Labrodon, Gervais. Labpodon confertidens, sp. nov. (Plate V., Fig. 7.) Description. — Lower phaiyngeal dentition subtriangular, with a very broad base, and apparently produced in front to a broadly rounded j)oint, but otu* s.pecimen is unfortunately imperfect in this res)tect. Surface strongly convex in the median area and tumid towards the front. Teeth very demsely crowded, normally circular in section, excepting where so closely packed as to be- come compressed into polygonal form, both laterally, aaid from back to front along the lines of gi-eatest convexity. Posteriorly the teeth become distinctly triangular. The largest teeth are situated in the median area, and form aibout 4 rows ; the}^ measure up to 3 mm. in diameter. The smallest teeth are situated anteriorly, and principally along the lateral extremi- ties ; they average about .75 mm. in diaimeter. By the fractured anterior of this pharyngeal, it may be noted that there are five successional series of teeth in addition to the functional layer. The unworn teeth are seen to have perfectly spherical and highly enamelled crowns. Dimensions. — Width of completed specimen, 54 mm. ; width at the base of beak-like projection, about lU mm. ; amtero- posterior diameter, probably about 31 mm. Locality and Horizon. — Grange Burn near Hamilton, Western Victoria. Base of the Kaiimnan. National Museum Collection (presented by A. A. Kelley, Esq.). 1 Since this was written we have heard with the deepest regret of the death of the distinguished curator of the Tasmanian Museum. 66 CJiapm/m and Prltchard : Observations. — The nearest ally to the above species appears to be the North American form described under the name of Pharyngodopilus carolinensis,! from the Tertiary Phosphate beds of South Carolina; but the characters and arraufrement of the teeth are distinct, the latter form having its dentition in more regular series. We might also draw some comparison with L. haueri, Miinster, sp.,^ from the Miocene of the Vienna Basin, Italy, Sicily and Brittany, but this form does not have its teeth so crowded, being usually openly spaced. Labrodon deppessus, sp. nov. (Plate Y., Figs, s-9.) Description. — There is amother specimen of the dentition of Labrodon in our collection which, since it shows considerable divergence from the foregoing species, we have thought ad- visable to separate, noting some of its principal features. This pharyngeal is remarkable for its thin and depressed aspect, nearly equiamgular in outline, with rather irregular and closely- packed teeth, and showing on the edges four successive layers. On the lower surface the baisos of the teeth are well shown, and each possesses a conspicuous and well-developed cavity, which is central and circular. As compared with the previous species, the teeth are rather lenticular than circular. Dimensions. — Diameter of largest teeth, 3.5 mm. : smallest teeth, 2 'mm. Average height of teeth, 1.75 mm. Width of hapryngerl, about 27 mm. ; antero-posterior diameter, Ifi mm. ; total thickness, 8 mm. Locality and Horizon. --Beaum:iii-is, Port Phillip. Kalimnau. (Pritchard Coll.). Family Diodontidae. Genus Diodon, Linnaeus. Diodon formOSUS, sp. nov. (Plate VL, Figs. 1-3 : Plate VII.; Plate VITL, Figs. 1-7.) Description. — Jaws solid, and ajjparently heavier than in other known fossil forms. Upper jaw with a. broadly-angular beak. Lower jaw rounded in front, and more depressed on the inner \ Jourii. Acad. Nat. Sui., -iiid ser., vol. viii., 1S77, p. 2.%, pi. xxxiv., «•,';;. 10--J4 (espeoi- all.v fi^'. 20). •2 I'liyllodus hauoii, von Jlliiister, ISeitr., I'etrefact, pt. vii., 1846, p. d, pi. i., tij,'. 1. F(jsKil F/xli Remains. 67 surface than the correspundinjz paLital surface of the upper jaw. Denticles of the jaw margin comparatively coarse, and irregular in size, with a vermiculately crinkled surface except ■where worn. In the largest specimen the pile of palatal plates numbers seventeen, whilst in the smallest specimens before us there are only five, and avera/gc sized examples show eleven or twelve. These Diodon jaws show very considerable variation in form, both as to the angle of the upper jaw margin, -which ranges from 95 degrees to 110 degrees ; and also in the excavated area between the palatal pile and the margin, which is often deeply concave, while the distance between the anterior margin of the plates and the denticulated border vai-ies between 8.5 mm. and 18 mm. These measurements were made on full gTOwn examples. The shape of the palate is generally suboval, in which the lateral axis is the longer. Dimensions of the Jaws of Diodon formosus, based on the more perfect examples. UPPER iA^^ r_ Lateral Width. Base to Front. Width of Palate. Depth of Palate. No. of Plates. Locality. lomui. 9.5nim.* H.5mm. 7 mm 4 imperfect Beaumaris 31 „ 19.5 „ 20.5 „ 11 „ C^) Beaumaris 36 „ 21 „ 25 „ 14 „ 10 Beaumaris 39 „* 31 „ 3(5.5 „ IS „ 8 Beaumaris 52 ,, 49 „ 34 „ 28 „ s Grange Burn 51 „ 41 „ 32 „ 28.5 „ only 7 visible Grange Burn 56 „ 46 „ 39 „ 26 „ 12 Beaumaris 66 „* 57 37 „ 33 „ only t) visible Grange Biun 70 „ 57 41 ,, 33 „ 17 Beaumaris « (Circ.) LOWER JAW. Lateral Width. Base to Front. Width of Palate. Depth of Palate. No. of Plates. 26.5mm. 17 mm. 17.5mm. 12.5mm. 6 29 „ 15.5 „ 20 ., 11.5 „ 6 32 ,. 20 19 „ 16 „ 6 visible 42* „ 36 30 „ 30 „ -' (Circ.) 12 Locality. Grange Burn Beaumaris Beaumaris Beaumaris 5a 68 Chajwum (trid Fi-itrlntrd : In addition to the above, we have a large number of more or less imperfect palates from which the average number of plates in the jjile was more accurately determined than might be judged from the above table, and the proportion of incomplete to complete specimens we would estimate at subout three to one. The relatively greater abundance of the upper jaw as com- pared with the lower may be due to the fact that the lower jaw soon falls away from the fish after death, and thus runs a greater risk of destruction before coming under the influence of sedimen- tation, and consequent preservation of the remains. Locality and Horizon. — Grange Burn, Western Victoria, from the nodule bed at the base of the Kalimnan (Nat. Mus. Coll., purchaised R. Lindsay ; also 2 fine specimens of upper jaws pre- sented by A. A. Kelley ; also Spry Coll. and Pritchard Coll.). Beaumaris, Port Phillip (Nat. Mus. Coll. specimens collected by the late W. Kershaw, several presented by J. A. Kershaw, an exceptionally fine upper jaw presented by C. P. Smart, a lower jaw presented by the late J. F. Bailey and 8 specimens pur- chased from W. B. Jennings ; also Pritchard Coll., including an extensive series of small specimens ; and Dixon Coll., which has yielded the largest specimen). — Base of the Kailimnan. 01)servations. — The palatal aspect of the upper jaw shows the fossil form to be more angularly pointed at the beak than in the living D. hystrix, L., and more nearly approaching D. blochii, Casteln.. both of which species occur in Port Phillip, the latter beinir the commoner. Lateral margin less curved and more widely divergent tha,n in the living forms, the marginal denticles being generally coarser, averaging ten on each side, whilst our recent forms have twice that number. There is also a greater tendency in the full-grown fossil forms to an extension of the excavated area between the palatai plates and the denticulated oral margin. It appears that the recent form D. hystrix occasionally at- tained dimensions nearly equal to that which is indicated by the size of the jaw in the fossil species, since one individual is recorded from the British Museum Collection^ which has a length of thirty inches, but this is exceptional. Giinther- men- tions the largest form as attaining a length oi two feet. 1 Cat. Fishes ISrit. Mus., vol. viii., 1870, \>. 31IU. 2 -An Inti'oductioTi to tliu Study of Fishes, ISSO, p. (iSi). Fuss/t Fish l{<;iii<(in-s. 69 The present species differs from other described fossils in many particuhirs. Frcnu D. sipna^ it differs in having the dental plates with a subquadrate outline, instead of, as in thai species, strongly rounded sides. From 1). vetus" it niay be readily separated by tlie broad and angular shape of the jaws, and l)y tlieir more massive build in the palatal area. It is noteworthy that of the severajl described fossil forms of Diodon only 1). vetus appears to have been preserved as perfectly as the present species. The only other authentic si>ecies to which we may refer is D. scillae. Agassiz'' frcnn the Miocene of Italy, Sicily and Malta. In this form the palartal plates are thinner and consequently more numerous than in our species, and its lateral boundaries are sinuous and incurved instead of convex. It was in all probability the herein described species that the late Professor Ralph Tate had in mind when he recorded Atopomycterus from the Older Tertiary of Australia in his " Census " of its fauna.^ The reason that Tate assigned this fossil to the genus Atopomycterus may probably have been due to the fact that a fish, recorded b}- Steindachner' as Atopomyc- terus bocagei, had been found in Port Jackson, but this has since been indicated as synonymous with Diodon noveni- niaculatus, Cuviei." The genus Diodon is commonest as a Miocene fossil, and is usually associated with other fish and cetacean remains charac- teristic of phospha/tic beds in various parts of the world, but it also occurs in beds of Oligocene and Eocene age. Diodon connewarrensis, sp. nov. (Plate VIII., Figs. 8-10). Description. — Spine smooth, short and strong, after the type of those of D. hystrix, but not so acutely pointed, and without 1 Martin, Saminl. Geol. Reichsmus. Leiden, ser. 1, vol. iii., p. 16, pi. i., %s. 5, 5(t, oh. ■2 Leidy, Proc. Acad. Nat. Sci. Philad., vol. vii., 1855, p. 397; also Journ. of same Society, ser. '2, vol. viii., 1877, p. -255, pi. xxxiv., figrs. 15-18. 3 Poiss. Koss., vol. ii., pt. ii., p. 274. See also Smith Woodward, Cat. Foss. Fishes Brit. Mus., pt. iv., 1901, p. 572 ; text-figure 20 (p. 573). 4 Jinun. Roy. Soc. N. S. Wales, vol. xxii., pt. ii., 1888, p. 247. 5 Sitzun{,'sb. and K. Ak. Wiss. Wien, vol. liii., 1866, p. 477, pi. vi., f. 3. 6 A. Giinther, Cat. Fishes Brit. Mus., vol. viii., 1870, p. 308; also W. Macleay, Descr. Cat. Australian Fishes, vol. ii., 1881, p. 280. 70 C/iapnian and Pritcltard : the basal grooves. On the other hand, there is evidence of the presence of an anterior ridge sucli as is seen on tlie spines of I), blochii. Both the above-mentioned sjDecies are living in Port Phillip, D. hystrix being readily distinguished by its short spine with a broad base, whilst D. blochii has fewer and more slender spines with a comparajtively narrow base. The section of the fossil spine is subtrigonal, with the inner surface concave at the base, the roots or basal prongs making an angle of about 120 degrees with the main shaft. Dimensions. — Length, 6 mm. ; wddth from point to point of the roots, 5 mm. ; thickness at the base of the spine, 2 mm. Locality and Horizon. — Point Campbell clays. Lake Con- newarre, near Geelong. — Baloombian (Pritchard Coll.). Observations. — No separate spines of Diodou appear to have been recorded in the fossil state except those of the type speci- men of D. erinaceus Agassiz, from the Upper Eocene of Monte Bolca, near Verona. ^ It therefore seems desirable to record the above fossil with a distinctive name, especially since the example was found in beds of older date than those from which the palates have hitherto been procured. 1 Poiss. Foss., vol. ii., pt. ii., p. 271. See also Smith Woodward, Cat. Koss. Fishes, Brit. Mus., pt. iv., 1901, p. 572. Fossil Fish Remains. ? - o a « o o p^ ^; 3 CJ _o ^ > a O o o 1 i j aJ ^ tc "- 1 ^ fl ij o o H 03 o OJ I •■J 1 ^ ■p 1) 1 o 1 _d 'S CO S* ;4 2 ■ ►-S 2 ■ ■ ■ j2 y ^ Oi c a; O .— P-. >> <-— -^ — ^^ j;; •^ "^ 'C '-^ , -X •^ rv v; *< rs; ^ H H i— 72 ChaiJinan and Pritchard ►J m < > ^— ^ ^-i^ 0) » a] ,__. 3j :e " " ^ -O J .« ■•— ^ a ^-.-' *-« X - c j3 3 ^ tc *■ •^ o J3 1-5 a •^ " O ci ^ o ■fc3 cS ^ c5 "^ GC « ^ « 25 -i ■ ■ 9 ■ • • ?■ '■S =* ^ g ^ :: r 3 s ~ 5 ct 5 >~. 1 >3 x^ ^ < C' o I •^ '' o V z O > Q > • .2 • ' • • ' 'u 5 ' ' ' ^ p5 —1 ;^ IJ ' CD ' ' ' tr ' -^ 32 2 1* ' ■ • • 5 o s o -5 fl g IB CO ^ 3 CO _M a X eS pq 'S 'S o 3 03 ir U _0J a; i 5 ^ P C ^ p a) ce :t ^ ^ c^ ^ Ja 0) ^ oS 0 D ce P5 ^K, c-i o CQ 35 >^ cc • ■ ■ • • • *M 'o P. 02 3 S o '■*3 3J -5 CO 1 CO CO CO X CO 0^ CO o ^ Oi CO O q a s s 3 O CO 1 5 J: o '^ r3 o H „ ;© ^ 0 5 O 3 ">> ^ rS T3 '^ S ^ 7J s 2 - Fossil Fish Remains. 7'i V. — Additions to Bibliography. Agassiz, A. — Poissons fossiles. Vol. ii., 1835. Dixon, F. — Geology and Fossils of Sussex. 1850. Eastman, C. R. — Miocene Fishes. In Report of Maryland Geological Survey, 1904. Pp. 71-93. Giinther, A. — Catalogue of the Physostomi in the British Museum. Cat. of Fishes. Vol. vii., 1870. IiL — Introduction to the Study of Fishes. Edinburgh, 1880. Lcidv, J. — Description of Vertebrate Remains, chiefly from the Phosphate Beds of South Carolina. Journ. Acad. Nat. Sciences, Philadelphia. Second ser., vol. viii., 1877. A/acieay, W. — Descriptive Catalogue of Australian Fishes. Vol. ii., 1881. Miinster, G. von. — Ueber die in der Tertiar-Formation des Wiener Beckens vorkonimenden Fisch-Ueberreste, etc. Beitrage zur Petrefacten-Kunde. Heft 7. 1816. Newton, E. D. — On Two Chimaeroid Jaws from the Lower Greensand of New Zealand. Quart. Journ. Geol. Soc. Vol. xxxii., 1876, pp. 326-331. Id. — The Chimseroid Fishes of the British (Jretaceous Rocks. Mem. Geol. Surv. United Kingdom. Monograph iv., 1878. Pliillipi, E. — Ueber Tschyodus suevicus, nov. spec. Palaiontographica. Vol. xliv., 1897, pp. 1-10. Riess, /. — Ueber einige fossile Cliimjeridcii Reste in ^liinchener paljeontologischen Museum. Palseontographica. Vol. xxxiv., 1887-8, pp. 1-27. Stead, D, G^.— Fishes of Australia. Sydney, 1906. Woodward, A. S. — Catalogue of Fossil Fishes, British Museum (Natural History), pt. ii., 1891. 74 Cluvpnian and Pritcltard V[.— CORRIGENDA FOR PART I. p. 280. — 12th line fcoin top, for "pi." I. read "pi. XI." P. 285. — 14th line froiu top, for "Creep" read "Creek." P. 297. — in Explanation to Plates. 14th line from top, delete "[5434]." After Fig. 14 read " Oxyrhina has talis, Agassiz. In tier surface of posterior tooth ; from Beaumaris. Nat- ural size. [5424]." Before Acanthias geelon^ensis insert " Hg. 15 " instead of "14." Delete last two lines on p. 297, " fig. 4 Oxyrhina hastalis" etc. EXPLANATION OF PLATES V.-VIIL V. Fig. 1. Myliobatis niflorabbinensis, sp. nov. Tooth showing articulating and outer surfaces. Beaumaris. Fig. 2. M. moorabhinensis, sj). nov. Lower surface of same tooth. Fig. 3. M. inoorabbineiisis, sp. nov. A tooth of less curvature, showing lower surface. Beaumaris. Fig. 4. Kdapliodon sweeti, sp. nov. Right mandibular tooth, inner side. C-range Burn, near Hamilton. Fig. 5. E. siveeh\ sp. nov. Right vomerine tooth. Grange Burn. Fig. 6. E. sweeti, sp. nov. Left palatine tooth. Beaumaris. Fig. 7. Labrodon conferiidens, sp. nov. Lower phai-yiigeal. Grange Burn. Natural size. Fig. 8. Labrodon depressiis, sp. nov. Phaiyngeal ; up{)er sur- face. Beaumaris. Fig. 9. L. depressus, sp. nov. Lower siii'f.iceof same specinuui. All tigures of the natural size. I'l-oc. U.S. Victoria. I'.KiT. I'latc V. F. ('luilMnaii, ad iiat. ilcl. Teeth of Australian Tertiary Fishes. IVof. E.S. Virtnria. 1007. I'latc VI. V. chapman, ad nat. dt Teeth of Australian Tertiary Fishes. Proc. K.S. N'ictoiia, liiUJ. Tlato VII. F. riiapniaii, ad ii:it. ilcl. Jaw of Diodon formosus, sp. nov. Proc. U.S. Victoria, l!ii>7. Plate VTTT. V. Chaiunaii, ail luit. (lil. Australian Tertiary Fish Remains. FuK.'iil FfsJi ReiiKii U.S. V\. Fig. 1. Diodon forinosiis, %\>. now Outer surface of upper jaw. Grange Burn, near Hamilton. Fig. L\ D. fon/iosus, sp. nov. Inner sui-face of same specimen. Eig. 3. D. Jorniosiis, sp. nov. Profile of same. Fig. -i. Diodoii blochii, Castelnau. Anterior view of upper jaw of one of the species now living in Port Phillip. Fig. o. D. dlochii, Castel. Profile of same. Fig. G. Ischyodus iiiortom\ sp. nov. Left mandibular tooth. Table Cape, Tasmania. All figures of the natural size. VII. Fig. 1. Diodon forinosHS, ii\>. now Outer surface of upper jaw of a full grown e.xanjple (Dixon Coll.). Beau- maris. Fig. 2. D. for/nosus, sp. nov. Inner surface of same .specimen. Both figures of natural size. YTII. Fig. 1. Difldofi for/iiflsies^ sp. nov. Inner surface of upper jaw. Beaumaris. Fig. 2. 1). Jor/nosus, sp. nov. Inner sui'face of lower jaw. Beaumaris. Fig. 3. I). forinosHS, sp. nov. Inner surface of lower jaw of a young example. Beaumaris. Fig. I. 7J. forinosus, sp. nov. Inner surface of upper jaw of a young example. Grange Burn. Fig. 5. D. Jori/iosiis, sp. nov. Inner surface of upper jaw of a full-grown example. 8howing the pile of strong palatal plates and the marginal alveolar ridges. Beaumaris. Fig. G. The same : outer surface. Fig. 7. Tlie same ; edge view. Fig. S. Diodnri coiinetvar?-e}isis, sp. nov. Outer surface of spine. Fig. 9. The same ; inner surface. iFig. 1 0. The same ; basal view. Figs. 1-7 of the natural size ; 8-10 magnified twice. [Proc. Eoy. Soc. Victoria, 20 (N.S.), Pt. 1., 1907.] Ar'I'. VII. — CoiitriJjutions to the Flora of Australio,, No. 0} By ALFRED J. EWART, Ph.D., D.Sc, F.L.S., &c., Government Botanist and Professor of Botany at the Melbourne University. (With Plates IX.-XIII.). [Read 11th Jnly, 1907]. Angianthus HUMtFUSUS, Beiitli., var. gkandiflorus, new var. (Compositae), M. Koch. Woorooloo, W. Australia, 1906. Attention is drawn to this plant on account of its remarkable external re.semblance to large specimens of Myriocephalus rhizocephalus, Benth., forming a striking case of plant mimicry. The two plants are readily distinguished by the pappus, which in M. rhizocephalus consists of a single bristle, and in A. humifusus of five or six fringed ragged scales. [Specimens exhibited]. Bakckea CRisPiFLOKA, F. V. M. Fragm. IV., p. 72, var. tkxuiok (Myrtaceae). Elder exploring expedition No. 2. Kangaroo Hill, R. Helms. 1891. Cowcowing, W.A., xVI. Koch, 1904. The variety is more slender than the type forms and has a shorter pedicel, so that the usually slightly smaller bracts are close under the ovary, which is less urceolate than in the type form. A specimen from .Jibherding, W.A., M. Koch, 190-3, is in some respects intermediate between the variety and type form. (l\r.LiTRi.s MoRRisoNi, R. T. Baker (Coniferae). Linn. Soc. of N.S.W., vol. xxxi., 1906, p. 717. Under this head Baker includes one of Oldtield's specin)ens fi-om W. Australia, which was placed by Mueller as a variety of 1 No. 5 ill Vict. Nat., vol. xxi\ ., 1!)()7, p. .=)6 Flora of Australia. 77 C. verrucosa, R. Br. This latter species is a synonym for U. robusta, R. Bi\, to wliich Baker admits liis species closely approaches. The internodes to whicii Baker attaclies special importance are not any shorter than in other specimens of C. rubusta, and the scales which he gives as obtuse are acute as in C. robusta. There can be no doubt that this species is a varialile one, but variations are shown often on one and the same speci- men, and hence it is necessary to retain for it the scope given by Bentham, and include under it such varieties as microcarpa> verrucosa, intratropica, and possibly also the columellaris of F. M., and the Morrisoni of R. T. Baker. The last-named e.speci- ally seems to come within the range of thn C. robusta type, and a similar specimen was referred to that species by Bentham in the Flora Australiensis, p. 237. Cassinia l.^rvis, R. Br. (Compositae). This plant was recorded by Mueller as new to Victoria (Vict. Nat., vol. X., 1893 and 1894, pp. 132 and 160), on the strength of three specimens, one from Werribee Gorge, A. J. Campbell, 1892, one from J. F. Muldei-, C. Otway, 1893, and the other from C. French, Goulburn R. Mr. Tovey drew my attention to the fact that these specimens were peculiar in .several respects, and on e.xamination the Werribee specimen proves to be C. longifolia, R. Br., and the Ot%vay specimen C. aculeata. These three species are fairly closely related, but the specimens in question are identical with tlie types of their respec- tive species. [Specimens and types exhibited]. Hence C laevis has been wrongly recorded as Victorian. Cassini.a Tiikodoki, F. v. .M. The Victorian specimens in the Herbai'ium all prove to be Cassinia arcuata, R. Br. Hence the former has been wrongly recorded as Victoi'ian owing to incorrect identification. See Vict. Nat, vol. X., p. 160, 1894. Chamaklaucium Halli, n. sp. (Myrtaceae), (after the Secretary of the Royal Society). Cowcowing, W.A., IVI. Koch, Sept., 1904. A small shrub with stiff erect rough greyish branches, the leaves alternate and closely set at their ends in clusters of 78 Alfrrd J. Etvart : nearly 1 to 3 cm. length. The lea\es are terete, mostly half a cm. long, slightly narrowed at the base, the apex curved to a small, usually straight, white point, and .spar.sely covered with glandular spots. The flowers are practically sessile in terminal clusters of usually three or more. Calyx tube wrinkled but not prominently ridged (when dry), dark red, glandular, the five broad obtuse ciliate lobes with a light red border, and with pellucid spots. Corolla twice the length of the calyx, the lobes broad obtuse, pale brownish-yellow, and minutely fringed. .Stamens ten, the anthers adherent to an enlarged glandular connective, alternating with ten staminodes, the whole uniting to form a single distinct tube within the corolla. Ovary of one loculus, with several ovules arising from an erect wavy basal placenta. Style distended below the middle, stignaa globular with a basal fringe of bail's. The plant is allied to C. ciliatum, but its pointed leaves, flowers in terminal clusters, larger and broader unribbed calyces, petals distinctly fringed, at once distinguish it. The latter features show a slight approach to Verticordia, from which genus, however, it differs widely. CoNOSPERMUM Croniniae, Diels. Fragni. Phytog. Austr. Occid., p. 143)=C. amoenum, Meisn. This "species" is merely a depauperated form of C. amoenum, Meisn. With reduced inflorescences, somewhat smaller flowers and bracts, perianth with the external hairs well developed, so that the blue colour is partly hidden and the leaves usually, though not always, horizontally spreading. None of these features is constant, and a specimen seen by Bentham and referred to C amoenum diverges still more widely in the .same direction. In the Flora Australiensis, Bentham apparently described an extreme typo in the direction of luxuriance, and hence for instance exaggei-ates the size of the bracts. All grades of transition exist between the luxuriant and depauperate forms, and Diels' figure of the stamens m the opened corolla is not quite correct, these and the peculiar style being precisely similiar in both the luxuriant and depauperate forms. Diels collected no new material, and apparently saw only two of the extreme types at the Melbourne Flora of Australia. 79 Herbarium, being unaware of the intermediate forms referred by Bentliani and Mueller to this species or of those since obtained. Neither Bentliam nor Mueller considered these forms to be separable as a distinct and fairly constant variety, in which opinion T must emphatically concur, and desire to point out the danger of establishing a new species on a couple of odd forms taken from another Herbarium. Eriostkmox (Piii<:n.\LiuM) gibbosus, Luelini. (Rutaceae). Norse- man, W.A., J. I). Batt, 1S97. This plant was exhibited before the Field Naturalists in 1897 (vol. xiv., p. 18), but no description of it has been published. The specimens are very fragmentary, but the leaves are like those of E. difformis, and the flowers like those of C. obovalis. The filaments are, however, not ciliate, and the anthers not apicu- late, and there is no reason to suppose that the specimens form a hybrid between these two species. The younger branches are minutely pubescent, and the leaves have very prominent glands. The calyx lobes are very short, obtuse, and very slightly ciliate. The petals are glabrous and imbricate. Of the ten stamens those opposite the petals are somewhat longer than the others. The pale glabrous filaments bear reddish spots, and the gynaeceum is glabrous. Gkococcl'S pusillus, J. Drumm. et Harv. (Cruciferae). This curious plant was suggested by Bentliam as being possibly a form of Blennodia with dimorphic flowers and geophilous fruits. This suggestion was revived by Mueller (Vict. Nat., 1892, p. 137), who pointed out that the foliage I'esembled that of Sisym- brium cardaminoides, F. v. M., and that a Brazilian Cardamine sometimes exhibits a similar peculiarity. Geococcus pusillus might possiljly be a geophilous form of Sisymbrim cardaminoides, produced as the result of continued grazing or cropping. 8ome specimens of Geococcus in the Herbarium have the normal flowers of (Sisymbrium, and show great variation in the shape and length of the fruit. The shortened, and some- times almost sagittate, fruit of Geococcus is obviously developed 80 Alfred J. Eivart : in order to penetrate the ground readily. It may even be shorter and l)roader than in the figures given, and may be three or more times longer, and half as broad, thus bringing the fruit near to some of the rather variable shapes assumed by the aerial fruits of Sisymbrium cardamnioides. Mr. Header (Vict. Nat., 1905, p. 177), has, however, watched the growth of the plant, and concludes that it is not a form of 8. cardaniinoides, but is a good species (and genus) usually form- ing hypogeal fruits, but when luxuriant also producing them above ground. The variation in the shape of the fruit would, however, bring it near to S. cardaininoides. The differences in the flowers might be easily the result of their autogamous habit, as in species of Viola or Lamium. Numerous attempts to germ- inate and grow the plant from seed failed. The seed apparently rapidly loses its vitality, presumably in accordance with the fact that normally it is iunnediately planted. The appended figures show that Geococcus differs in many respects besides its general habit from Sisymbrium^ but until the former plant has been, proved to remain true for several generations, the possibility of a relationship between the two remains. Geococcus was omitted from the census by Mueller, but on the present evidence as to its structure must be restored, at least until cultural experiments succeed in showing that it is a form of another plant. GuNNiOPSis INTERMEDIA, Diels. (I)iels and Pritzel, Fragm. Phyt. Aust., etc., p. 197) = Aizoon intermedium, Diels. (Aizoaceae). This new species appears to be the same as the " Aizoon glab- rum " recorded by Mr. Luehmaim, but of which no description was published. In Engler's Pflanzenfamilien, Pax founds the genus Gunni- opsis for the Australian species of Aizoon upon the following characters : — Aizoon, calyx 5 partite, imbricate ; capsule loculicidal. GuNNiOPSis, calyx 4 partite, valvate ; capsule septicidaj. In Gunniopsis, G. quadrifaria (F. v. M.), Pax is included, which is presumably a misprint for G. (juadrilida (A. (piadrifidum, F. v. M.). The capsule is, however, both septicidal and partly Flora of Australia. 81 loculicidal in both the Australian species, the valvate and imbri- cate characters do not appear to be constant, and further, the calyx is sometimes tive partite, as was first noted by Mueller, Frai^m., vol. vii., p. 129. There seems therefore to be no solid reason for founding a new genus for the Australian Aizoons, but preferably to give to that genus the somewhat broader latitude ;ulmitt('(l by Beiitham to include the Australian species, in spite of their additional development of septicidal dehiscence, and usually of four partite calyces. Helipterum Jesseni, F. v. M. M. Koch, W. Australia, 1904. The plant is mentioned on account of its highly misleading external resemblance to Myriocephalus gracilis, Benth. [Specimens exhibited.] Heliciirysum SUBULIFOLIUM, F. V. M. (Compositae). (Syn. H. filifolium, F. v. M.). Various forms of this plant from W. Australia (Cowcowing, M. Koch, 1904) l)ridge the gap to the very closely allied "species" H. Hlifoliuu), F. v. M., which appears to be merely a form of H. subulifolium, and can probably be classed as a variety of that species. The plant is often confused with Helipterum tenellum on account of its aluiost plumose pappus and filiform leaves but differs widely in its involucre. Helichkysum Tepperi, F. v. M. (Compositae). Cowcowing Lakes, W. Australia, M. Koch, 1904; L. Boga, Victoria, H. B. Williamson, 1898. This pretty little Composite described by Mueller in the S. Science Record 1882, p. 1, from 8. Australia, was represented in fieri )ariuui by the type specimens only. The plant from L. Boga was named Podolepis Lessoni by Mr. Luehmann, to a dwarf form of which it bears a fairly close resemfjlance, as noted by JNIueller. The two are, however, quite distinct, and H. Tepperi, though apparently rare has a wide range through Victoria, S. Australia and W. Australia. It has l)een recorded from W. Australia by Spencer le Moore in Journ. Linn. Soc. of London, vol. xxxiv., 1899, p. 198. 82 Alfred J. Ewart : Helipterum Guilfoylei, n. sp. (Compositae) (named after the Director of the Melbourne Botanical Gardens). An annual prostate or ascending, rarely exceeding 4 to 5 cm. in height, covered with long loosely woolly hairs, and with one or more stems branching to form clusters of small ovoid heads. Leaves sessile, narrow, Ihiear, mostly obtusely pointed, and 4 to 5 mm. long, channelled on the upper surface, alternate or opposite. Heads partly within the upper leaves, mostly 5 mm. long by 3 broad, the outer bracts 2 mm., the inner 4 or 5, and with small yellow or brown laminas, the innermost smaller again without any lamina and very thin. All with various entire margins, and twenty or more in number. Flowers all tubular and hermaphro- dite, usually ten, the corolla, with five blunt points, the style swollen at the base, the pappus about the length of the corolla, of usually 8 plumose scales flattened at their bases and united to form a sessile ring easily separated entire. Achenes 1.5 to 2 mm. long, and quite twice as long as broad, reddish-brown, glabrous, the outer layers becoming mucilaginous in water, but with a reticulate surface before swelling. Style l)ifurcate with papillose ends; it and the stamens barely projecting l)eyoiid the tliroat of the corolla. The plant has a close e.\teriial resemblance to H. exiguum, F. V. M., Imt appears to be allied to H. pygmaeum, Benth., and of recently described species. H. verecundum (8. Moore, Journ. Linn. Hoc, vol. xxxiv., 1899, p. 200) is distinguished by its min- ute size, and H. Zacchaeus (S. Moore, Journ. of Bot., 1897, p. 166), by its pappus, achenes nearly as broad as long, and green tips to the involucral scales. The latter species also has presum- ably not the mucilaginous seed coat or peculiar style of H. Guil- foylei. Owing to the former fact the whole cluster of ripe achenes adheres and comes out in one mass, usually with the florets and pappus attached, two or three of the florets lieing usually sterile. KociilA Massoni, n. sp. (Chenopodiaceae) (named after Prof. Masson). Cowcowing, W.A., M. Koch, 1904. A small annual slightly prostate, up to 15 cm. in height, .soft, and sparsely covered with a white oi- brownish wool, less developed Flont of Australia. 83 on the leaves and absent from the fruits. Leaves linear, mostly \h cm. long, narrow without obtuse ends, alternate, closely set, the upper ones with sessile axillary flowers. Fruit sessile, dark greyish-brown, table-like, with ridged sides, 2 mm. high, and 4 mm. broad at the top, 2 mm. at base. The fruit thus has a flat top and broadened rim, but no wing. The latter fact at once distinguishes it from K. huinillima, to which it is otherwise fairly closely allied in habit and general appearance. The plant is much smaller than the Kochia polypterygia of Diels, has smaller fruits with the discoid wing much less developed, and a flattened top to the fruit with the ridges barely showing. Patersonia Dbummondi, F. v. M. (Irideae). Cowcowing, W.A., M. Koch, 1904. The plant appears to be very rare, only three sheets of imperfect specimens being in the National Herbarium, collected by Drunmiond. Koch's specimens have the marginal hairs less prominently developed than the type, but some of Drummond's specimens show the .same peculiarity, the larger hairs apparently rubbing off" readily. A part of Drummond's specimens had evidently been burnt back by a bush tire some time previously to their collection. PoDOLEPis Kendalli, F. v. M., var. nanus, new var. (Conipositae). Height four to six inches. Flowers all terminal and smaller than the terminal ones of the type. Waterloo, W.A., Max Koch, 1906. Champion Bay, W.A., L. Gould, 1890. PoDOLEPis Spencehi, n. sp. (Compositae), (named after Prof. W. Baldwin Spencer). AVoorooloo, W.A., M. Koch, 1906. Annual, 20-40 cm. height, one or more flowering stems from the same root, forming a loose panicle of heads, the flnal forks almost dichotomous. Stems glabrous, leaves hairy, and almost woolly on the under sides. Basal leaves lanceolate, spathulate about 5 cm. long by 1 cm. broad, the upper leaves all alternate, becoming narrower and smaller, and all sessile, with broad slightly-decurrent bases. 84 Alfred J. Ewart : Heads on stalks of usually 5 or more cui., 1 to ^ cm. lon^f, and nearly as broad as long. Basal and outer bracts small, sessile and obtuse, the inner larger, developing pronounced claws with glands on the outer surface, and liecoming more pointed ; all with shining transparent unwrinkled and unfringed laminas. Outer rows of florets, female, ligulate, pale to brownish-yellow, with usually three blunt points, projecting beyond the bracts. Inner disc florets tubular and hern)aphrodite, with five shorf> blunt, equal teeth, 'i'hese in both florets are usually tipped with red. Pappus of about eighc or ten flue bristles, minutely fringed but not plumose, present on all the florets. The plant appears to come between P. Lessoni, and P. rugata. It is easily distinguished from the recently described P. (ieorgei of Diels, by the facts that the outer florets are ligulate, the leaves are never opposite, and the inner scales have curved glandular stalks. The smooth scales distinguish it fi'om P. rugat;i, and its size and the colour of the florets from P. Lessoni. Ptekostylis reflkxa, R. Br., var. intermedia, n. var. (Orchidaceae). This plant has been referred at different times to various species and was finally classed by Baion von Mueller as a variety of P. obtusa. Although closely related to P. ohtusa it differs from that species in various features. The leaves on the evanes- cent basal rosette are three-veined instead of flve-veined, and the two lateral veins are often very faint. The leaves ai-e also smaller and more orbicular. The flowering stem is covered with fine closely-set short papillae, especially short and dense on the stalk and ridges of the ovary, and on the undei- surfaces of the leaves. The upper leaves on the flowering stem are often more than an inch long and nearly (juarter of an inch broad, the edges finely denticulate, and contracted to a subulate brownish, often curved tip usually one-eighth of an inch or more in length, but less developed on the basal leaves. The iabellum is lanceolate. strongly contracted in its upper third to a i'eddi.sh-l)rown entire tip. TIh^ l)asal appendage is curved and irregularly fringed with cilia along its distal third, the terminal cilium being larger than the rest. In other respects the plant bears a close resemblance Flora of Australia. 85 to P. obtusa, from wliicli however its labellum at once distin- guishes it. The labellum a)«d Hower are like those of P. praecox, the leaves, stem, and papillose surface are more like the charac- ters of P. reflexa. Since a perfect series of gradations exist as regard size of flower, length of point of labellum, size and acuminate character of leaves, and scabrous or glabrous character of stem and leaves between P. leflexa and P. praecox, Lindl., the latter species must be reduced to a vaiiety of P. reflexa. P. obtusa, 11. Br., seems to be distinct, especially as regards the obtu.sely oblong shape of its labellum. Mentone, J. R. Tovey and C. French, Juii., 1907 ; Chelten- ham, J. McKibbin, 1893 ; Brighton, C. French, Jun.; Wedder- burn, F. Colvin, 1880; near Beaumaris, C. French, Jun., 1882. Tysonia piiyllostkgia, F. v. M. (Compositae) = Swinbukma PHYLLOSTEfJIA, F. V. M. This plant was descril)ed in the Chemist and Druggist of Australia, Oct. 1, 1896, at the time of Mueller's death. A description but no specimens are in the Herbarium. The latter were apparently claimed and retained by Mueller's Executors. Mueller was evidently unaware of the existence of a prior generic name of Tysonia Bolus, Boraginaceae, represented by one African species. Mueller's name therefore may be replaced Swinburnia phyllostegia, the generic name commemorating the services of the present Minister of Agriculture to Botanical research. Verticordia Pritzkli.i, Diels. Fragm. Phytog. Austr. Occid., p. 401. Under this name Diels and Pritzel include the plant recorded as V. humilis, Benth., of the Elder exploring expedition (Trans. Roy. Soc. 8. Aust. XVI, p. 353). The latter identification was certaiidy incorrectly, since the specimens have bearded and not glabrous styles, but the style is not capitate as shown in Diel's figure but with an obtusely linear point. In other respects the specimens tally closely with Diels' description so that their flgure of the style may have been incorrectly drawn. 86 Alfred. J. Ewart : Unrecorded Naturalised Aliens. Alkanna lutea, D. C. ( Boragineae). Derwent, Tasmania, ex. Herb., Spicer. Alkanna lutea, D. C, var. pakviflora. (Boragineae). Geelong, Victoria, H. B. Williani.son, 1905. Beta vulgaris, L., var. maritima. (Chenopodiaceae). Probably an escape from cultivation. Geelong, Victoria, H. B. Williamson, 1907. CoNiUM maculatum, L. " Hemlock." (Umbelliferae). Portland, 1907, and various other districts in Victoria. EcBALLiUM ELATERIUM, A. Rich. (Cucurbitaceae). Squirting cucumber. ProVmbly a garden escape. Geelong, H. B. Williamson, 1907. Gladiolus cuspidati;s, Jacq. (Irideae). Geelong, H. B. Williamson, 1905. Near Melbourne, F. M. Reader, 1883. Ovens River, A. W. Eu,ston, 1891. The first appearance of this plant was leciirded by Mr. Reader in the Austr. Jour, of Pharmacy, 1887. LoLiUM italicum, a. Br. Italian Rye grass (Gramineae). Various districts in Victoria. Matricaria DiscoiDEA, D. C, "Wild Chamomile." (Conipositae). Widely spread in Victoria. Ranunculus sceleratus, L. (Ranunculaceae). Orbost, Snowy R., C. H. Grove, 1905. Reseda Luteola, L. " Dyi'r's Rocket." (Resedaceae). Vai'ious localities in Victoria. Proc. E.S. Victoria, ]iK)7. Plate IX. Chamaelaucium Halli, n. sp. Trnr. I.'.S. \'irtni-i;i. I'MlT. I'latr X. {« I GeOCOCCUS pusillus, Druniin. ct Ilaiv. (h) Kochia Massoni, n. -p. Pi-oc. E.S. Victoria, 1907. Plate XI. 10 1^ (a GeOCOCCUS pusillus, Drimim. et Hiuv. I'roc. K.S. ViotoriM, 1'.MI7. Plate XI 1. ^n w K-t:'/! e J/^ Helipterum Guilfoylei, n- sp. Proi-. E.S. Yietovia, 1!)07. Plate XIII. W in Podolepis Spenceri, n. sp. Fiord of Australia. 87 ScoLYMUs HisPANicus, L. (Conipositae). Widely spread in Victoria, but not very plentiful. Tkigonella ornithopodoides, D. C. "Fenugreek." (Legum- inosae). Penshurst, H. B. Williamson, January, 1907. DESCRIPTION OF PLATES. Plate IX. Chamaelmiciuvi Halli, n. sp. — (a) Plant somewhat reduced, (b) flower, (c) the same in vertical section, (d) leaf. Plate X. (a) Geococcus piisi/lus, Drumm. and Harv. Plant reduced. (/>) Kochia Massoni, n. sp. Plant reduced. Plate XL Geococcus pusillus. — 1 and 2, flower bud and flower ; 3 and 4, superior and inferior views of flower ; 5 and 6, face and back views of sepal; 7, petal; 8, hair; 9, ovary; 10, 11, 12, 13, stamen and pollen grains; 14, 15, 16, fruit; 17, seed; 18 section of seed ; 19, embryo; 20, radicle; 21, leaf. Plate XII. Hdipteruin Guil/oyiei, n. sp. — (a) Plant somewhat reduced, (l>) a median bract with lamina, (c) an innermost bract with a blunt point but no lamina, (d) flower, (e) pollen grain, (f) pappus. Plate XIII. Podolepsls Spenceri, n. sp. — (a) plant reduced, (l>) ray floret, (c) disc floret, (d) inner stalked bract, (e) and (f) outer sessile bracts. END OF VOLUME XX., PART T. [Published August, 19(>7j. riintcd liy I'ord & Sox, :{72 & 374 Druniinond Street, Cmltoii, .Mull.ounii'. PROCEEDINGS OF 'IHK $0]jal ^oricti) of f irtoria VOL. XX. (Nkw Sektks). PART 11. Edited under the Authority oj the Council. ISSUED MARCH, igo8. (C'tnitainiitg Papers rend before the Society liiiring the uionths of October, November, December, igoj). rilK AUTHORS OK TIIK SRVKRAl, I'Al'KRS AKK 8KVERALLY RKSPONSIBLK KOK TUB SOUNDNKSS OK TIIK OPINIONS OIVKN AND KOR TIIK ACCURACY OK THK STATEMENTS MAUB TIIKRBIK. MKLBOUKNE: K<)i;l) A snX, I'KliNTEKS, DEUMMOND STKEET, CAKLTUN. AdKSTS TO THH ^OClhyi'r : \VIM,IA.\1S & NOKOATK, 14 HENltllCTTA STKEKT, COVENT GAllUEN, LONDON. To wlioni all comiiimiications for transmission to the Royal Society of Victoria, from all parts of Europe, should be sent. 1 908. CONTENTS OF VOLUME XX,, Pt. II, PAGK Art VIII. — Notes on the Geoloijy of Moorooduc in the Morn- inf{ton Peninsula. By Eene«t VV. Skeats, D.Sc, A.E.C.S., F.G.S. (Plates XIV.-XVL). ... 89 Art IX.— On the Validity of Callitris Morrisoni. By R. T. Baker, F.L.S. ... ... ... ... 104. Art. X. — The Formation of Red Wood in Conifers. By Jean White, M.Sc. ... ... .. 107 Art. XI. — Contributions to the Flora of Australia, No. 7. By Alfred J. Ewart, D.Sc, Ph.D., F.L.S. .. 125 Arv. XII. — On the occurrence of a ilarsupium in an Echinoid belonging to the Genus Scutellina. By T. S. Hall, M.A. . ... ... ... ... 14() Art. XIII.— The Coleopteia of King I.sland, Bass Strait. By Arthur M. Lea ... ... ... ... 143 Art. XIV. — New or Little-known Victorian Fossils in the National Museum. Part IX. — Some Tertiary Species. By Frederick Chapman, A.L.S., &c. (Plates XVII.-XIX.) ... .. ... ... 208 Art. XV. — The Anatomy of Some Australian Amphibia. Part I. By Georgina Sweet, D.Sc. (Melb. Univ.). (Plates XX., XXI.) ... ... .. ... 222 .■\kt. XVI. — The Highlands and ilain Divide of Western Victoria. By T. S. Hart, M.A., F.G.S. (Plates XXIL-XXVI.) ... ... ... 250 Annual Report and Balance Sheet ... ... .. 274 Office-Bearers ... .. ... ... ... ... 277 Committees ... ... ... 278 List of Members ... .. ... ... ... ... 279 Index ... ... ... ... ... ... ... 285 Troc. Hot. Soc. Victoria, 20 (N.S.), Pt. II., 1907.] Ai:T. VIII. — Notes on t/ie Geuloyy of Moorooduc in the Morn ingto n Fen insida. By ERNEST \V. SKEATS, D.Sc, A.R.C.S., F.G.S Professor of Geology, University of Melbourne. (With Plates XIV.-XVI.). [Read 10th October, 1907]. Introduction. My first visit to the neighbourhood of Moorooduc was made in 1905, in the compimy of my assistants, Messrs. H. J. Grayson and H. Summers, and the members of the Geological Field Class of the University. On this visit we were concerned mainly with three j^roblems : — 1. The age of the Palseozoic sediments which rise above the mantle of Tertiary rocks. 2. The characters of the granitic mass of Mt. Eliza and of the acid veins proceeding from it. 3. The nature of the metamorphi«m effected by the in- trusion to the granitic rocks into the Palaeozoic sediments. This first visit enabled us to obtain evidence bearing on the two latter questions, but we shared the fate of previous observers in failing to find any fcKssils, sio that the age of the rocks remained in doubt. A second visit made under similar auspices in 1906 was more successful, as recognisable fossils were obtained. The present communication is ba.sed partly on field-work during these visits and a later examination, and partly on a petrological determination of the granitic and metamorphic rock^. 90 Ernrst IT. Sknifs Pkkviohs LiTKKATrm:. Tho fir.^ion of the gra.nite forming these hills." The granites of the district he referred to on page 8 as " presenting no peculiar features, being composed of quartz, reddish-coloured, felspar and black mica; the two latter, how- ever, occasionally vary in colour, the mica, being yellow and the felspar white." On his second map, printed in 1856, Sel- ^^yn records the ca«t of an encrinite stem from Sandstone Island in Western Port, and the rocks are referred to the Silurian period. The next paper bearing on the area was by Mr. A. K. Kitson, F.G.S., entitled, "Report on the coast line and adjacent coun- try between Fraiikston, Mornington, and Droniana," and was pub- lished in March, 1900, in Monthly Progress Report Xo. 12 of the Department of Mines, Victoria. Mr. Kit>on gives an in- teresting and somewhat detiMled account of the geology of the district, and the report is accomi)anied by sections and a geolo- gical sketch map of the area described. Mr. Kitson docs not describe the plutonic mas> of Mt. Kliza, but refers to the Geology of Mooroodnc. Ul acid dykos which penetrate the sediuientaiy rocks in a c[uarry north of Mooroodue Railway Station. He notes that they have indurated the contiguous strata for distances ranging from less than an inch to several feet. He describes 'ino>t of the dykes as aplites, and makes the interesting observation that the musco- vite and biotite in the dykes line the walls, while the centres consist of the more acid quartz and felspiw. He gives a litho- logica/l description of the sediments, and refers to the spotted character of the thin bedded micaceous shales. The rocks are described under the heading "Silurian" by Mr. Kitson, and the same view is expres-sed in the large Geological Map of Victoria of 1902. Mr. Kitson, however, remarks that the rocks resemble in some respects the gi-aptolite^bearing shales of the Lancefield district, and "' they may eventually prove to be of Ordovician age, though the Silurian belt may be the exten- sion of the Upper Silurian of the Melbourne district." In the year 1900, Mr. Evelyn Hogg published a paper en- titled, "The Petrology of certain Victorian gi'anites." i. Mr. Hogg does not discuss the granitic rock at Mt. Eliza, but describes one from an adjoining locality, Frankston, as a medium-grained graiiitite, a rock with pink felspar, orthoclase and plagioclase being about equally represented, quartz and biotite. The rock of Watson's Quarry, Mt. Martha, lying south of Mt. Eliza, is described as a medium-gi'ained syenite. As these are the nearest granitic masses to Mt. Eliza, their composition is of some interest in this connection. It is to be noted, however, that Mr. Hogg defines a granitite as including all holocrystalline quartz-biotite, rocks in which a mouoclinic felspar is not the dominant one, while he defines a syenite as a normal gi'anite with hornblende. Most j^e-trologists would now, I think, describe such a rock as a hornblende gi'anite. In 1901, Messrs. T. S. Hall and G. B. Pritchard published a paperiuthe Proceedingsof theRoyal Society of Victoria, Vol. XIV., N.S,. Pt. 1, entitled " Some Sections Illustrating the Geological Structure of the Country about Morniugton." They go fully into the previous literature of the area, and the greater part of the paper is devoted to the detailed discussion of the Tertiary r(n-ks and fossils of the district. The rocks of the Mooroodue quarry 1 Proc. Roy. Soc. Vic-t., n.a., vol. xiii., 1900, p. 21S. 92 Ernest IF. SkeaU : are described as Silurian or Ordovician. In this paper we have the first indication that the palaeozoic rocks are fossiliferoua They describe a coarse conglomerate which underlies the older bat^a/lt in many places as being " derived in the main from the older palaeozoic i?edimentary rocks of the district, and from the granitic series. In two places — namely, in the first cutting on the coast road south of Frankston, and near the first outcrop of granite rock south again from this place on the >hore, we have found a few graj^tolites in slate pebbles. They are very indis- tinct, and beyond saying that they are species of Diplograptus, we do not at present care to venture. Their evidence, chen, leaves the age of the rocks still open." In the year 1904 the first definite record of fossils found " in situ "' in the older Palaeozoic rocks of the Mornington Peninsula was given by Mr. T. S. Hall, M.A. ' The record does not mention the finder of the fossils, but I understand that it was Mr. W. H. Ferguson, of the Geological Survey of Victoria. The first record is that of the occurrence of Climacograptus and Diplograptus in boulders from Grice's Creek, Mornington, a locality nearer to Moorooduc than the earlier finds of Messrs. Hall and Pritchard. The evidence of these fos^ils, however, still leaves the age of the beds doubtful. A second suite of fossils found " in situ " at Balnarring, and identified by Mr. Hall, showTs clearly that Lower Ordovician rocks occur in that part of the Mornington Peninsula. Mr. Hall records Didyuiograptus, c.f. pritchardi. Tetragr apt us appr oximatus . Tetragraptus ([uadribracliiatu.s. Tetragraptus fruticosus (/). Ostraooda. Mr. Hall states that if the identification of T. fruticosus (?) is correct, the age of the rocks is Bendigonian, and in any case cannot be higher than the horizon of Castlemaine. Another series of fossils from Bulldog Creek, near Dromana, yielded to Mr. Hall the same forms a< those from Balnarring, and in addition undoubted specimens of Tetragraptus fruticosus, thus fixing their Bendigonian horizon. 1 Reports on (iiM^tolites, Hocdrds of Gi()lo<;i<;il Suivey of Victoiiii, vol. i., pt. iii., 11HI4, PI). 220, 221. Geology of Mooruodac. 93 Didymograptus, sp. indet. Temnograptus, sp. Dendrograptus (■/). Rhinopterocaris maccoyi. Braohiopod cast, and indeterminate Hexactinellid sponge spicules were also recog- ni>ed from among the collection made by Mr. Ferguson at this locality. The credit, then, for first finding fossils "in situ" in the Palaeozoic rocks of the Mornington Peninsula belongs to Mr. Ferguson, and for determining their Lower Ordovician (Bendigonian) age to Mr. Hall. The Agk of thh Olokh Skdimkntahv Kocks of Moorooduc. The fore'roiiio' account of the geolotrical literature dealing with the district shows the progi-e-s already made towards deter- mining the age of the Palaeozoic rocks of the Mornington Pen- insula. The records from Balnarring and Bulldog Creek defi- nitely established the Lower Ordovician age of the rocks of the southern part of the Mornington Peninsula, but the age of the serie.i near Moorooduc remained still in doubt, as it lies about ten miles to the north of the localities mentioned above, and, moreover, the graptolites found in the boulders of the conglo- merates of Grice"s Creek and near Frankston showed only that the rocks might be CTrdovician or Lower Silurian. The area near Moorooduc does not look j^romising, as apart from the highly altered rocks in the quarry north of Moorooduc station, rock exposures are veiy few, and several observers had already searched the locality wath negative results. Thi^ also was our experience in 1905, but on the second visit in 190G we were more fortunate. We were searching the hillside about a third of a mile north-east of the large quarry at about an eleva- tion of 350 feet above sea level, and almost due \\&>X of a slight bend in the road which runs north towards Frankston. Here the uprooting of a tree had exposed a very limited area of the elates, and from this and another small exposure close at hand we found a number of graptolites. They were clearly of Lower Ordovician age, a>s forms belonging to Didymograptus and Tetragraptus were recognised. On returning to Melbourne I 94 Ernest W. Skeats .- submitted the collection to Mr. Hall for mure detailed examina- tion, and he has kindly identitied the foUowings forms : — Didymograptus caduceus, Salter. Tetragraptus serra (sensu stricto) Broiign. Diplograptus, sp. Trigonograptus, sp. Lasiograptus, sp. Glossograp'tus, sp. Also specimens of Rhinopterocaris uuiccoyi, Eth. tils., and sponge spicules. Mr. Hall adds the following notes: — "The horizon is that of the Upper Castlemaine series, although the presence of Glosso- gi'aptus is suggestive of the horizon of the Darriwill series. The species of Diplograptus is similar to one which occurs as low down as the Victorian Gully beds at Castlemaine, but is in- distinct. Trigonograptus is known from higher beds, but pos- sibly occurs at Castlemaine. The presence of graptolies in the Eocene (?) conglomerate on the beach near Frankston has be^n recorded by Mr. Pritchard and myself.^ We announced the pre- sence of Diplograptus, but ventured no further. At the same time I found a specimen which I thought might be Didymo- graptus caduceus, but it was so indistinct that I thought it wiser not to mention it, especially as the record of even the genus would have upset the generally-received opinion as to the age of the slates of the district." The discovery of these graptolites provides the evidence previously wanting for determining the age of the Moorooduc rocks, clearly shows their Lower Ordovician character, and Mr. Hall's determinations show that the rocks belong to the Upper Castlemaine, or possibly the Da.rriwill series, a higher horizon than that of the graptolitj bearing rocks previously described irom the southern part of the Peninsula. It is now probable that all the slates and sandstones of the Morn- ington Peninsula belong to the Ordovician series. The grapto- lites were found on a steeply sloping part of the hillside, where no observations of the dip or strike of the beds could be obtained. At a lower level the Ordovician rocks pass below the Tertiary series, the general direction of the eastern boundary of 1 I'roc. Roy. Soc. \ict., xiv., I'.iol, [i. 41. (jcologij of Mooi'oodac. * 95 the hill being X. 60 deg. E. On walking south-westwards to- wards the large qua;rr}% further search gave negative results. The rocks of the quarry are in places much disturbed. On the south-western face of the quarry a steep anticlinal fold is seen, and towards the >s. end an abrupt change of strike to E. and \V., and a dip to N. at 70 deg. is noticed, while at the «outh end, where the beds are less disturbed, the strike is nearly N.E. and S.W., and the dip X.W. at 80 deg. At the opposite or Xurth- easi face of the quarry the strike was observed to be N. -20 deg. E., and dip E. 20 deg. S. at 70 deg. Possibly the rocks oi the quarry belong to the same series as those in which the graptolites were found, as a continuance of the N. 20 deg. E. strike would pass close to the graptolite localities. The Mktamuupiiic Rocks of the Mookooduc Quahky. The rocks consist of sandstones and slates. The sandstones, some of which occur in fairly thick beds, show little visible alteration excejjt that in places they are changed to quartzite. The slates are, however, highly altered. Among the slates are some with alternate dark and light laminae. On splitting a specimen of laminated slate along a bedding plane, elongated colourless prismatic crystals up to an inch in length were seen. A fragment of one of these crystals examined under the micro- scope shows the refractive index, polarization colours, and pink to colourless jileochroism characteristic of andalusite. Thin sections of the slates show the occurrence of two types, the one more, the other less altered. The less altered type is a spotted slate (Sections 505a and 50&I?). Under the microscope crypto-crystalline to micro-crystalline aggregates of a white mica- ceous mineral are seen to form abundant lighter areas with sub- rectangular boundaries, while the fine-grained groundmass con- sists ui biotite, quartz, uniaxial white mica, hematite, limonite, and some dark red-bruwn rutile crystals. The white uniaxial mica is possibly bleached biotite, since some oi the larger cry.-tals have apparently unaltered brown areas parallel to the cleavage traces, while hematite and limonite surround the white mica in such a way as to suggest that th-* iron lias been leached from biotite and deposited as oxide round the 96 Ernest W. Sheds: bleached eystals. Several sections of another iiiiiieral are present. It occurs as colourless prismatic sections, showing minute fluid and other inclusions, with high refractive index and low polarization colours which are grey to yellow, of the first order. Two cleavages are noticed, a well-defined one parallel to the longer axis c)f the crj'stal, and a less well developed one at right angles to this. A few sections showed straight extinction, but the majority extinguished in an oblique position. Th0 nlaximum extinction angle observed was 43 deg. from the longer axis. It is invariably associated with a marginal colourless mineral of lower refractive index and higher polarization colours. This mineral extends inwards from the margins of the crystals, and appears to be an alteration product consisting of a white uniaxial micaceous mineral. Most of the sections of the mineral show the emer- gence of an optic axis in a somewhat oblique position. The mineral is andalusite. The .sections are too thin to cxhiliit the characteristic pleochroism, and the high angle of extinction noticed in some sections is to be connected Avith the large optical axial angle exhibited by this mineral. ■'■ The more altered type of slate (Section 507) shows complete recrystallization of the clastic materials. The rock consists mainly of a number of interlocking qvuirtz granules and mica- ceous minerals. The latter include biotite. muscovite and bleached biotite (.'). No trace of a spotted structure is seen, but the original bedding planes are defined by lines along which there is a greater concentration of biotite and hematite, and larger crystals of the micas occur along these lamin*. Among the minor constituents minute rutiles occur, and a few iileudii-dic granules of tourmaline,* which have been included in the bleached micas. Andalusite is not represented in this rock. These altered rocks, containing an abundance of micas, arc evi- dently rich in alkalies. It is therefore probable that in the formation of the shale the alkali contents were not leached out as sometimes happens. It must be remembered, however, that the alkali contents of the shales may have boon reinforced by thermal solutions passing out from the margin of the gra.nitic intrusion. 1 The miiiilie.'.'s of tlie rock sections refer to tlie riiiversitv colleetioii of rocU slices. Geolugj oj Mooi'oodac. 97 Thk Gkano-Diokite and Apophysks of Mount Eliza. The Apophyses. — The-se have only been noticed in the meta- niorphosed shites and sandstones of the large Moomoduc quarry, about three-quarters of a mile north of Moorooduc railway sta- tion. They consist of acid extrusions from the plutonic maiss, and vary from tine-gi-ainedaplitic- rocks to fairly coarse pegmatites. The largest vein seen measured about three feet in width. They are all somewhat decomposed, and on that account no rock sec- tions have been made from them. In places quartz and felspar alone are present, in others biotite and nmscovite also occur, usually in large flakes up to three-quarters of an inch in length, and in one or two cases black tourmaline was noticed. Mr. Kitson has drawn attention' to the most interesting feature in connection with them — viz., the general concentration of the mica along the walls of the veins, the central parts being relatively free from that mineral. The small sizes of the veins makes it impi'ol)al)le tliat C(jiivection currents have played any pait in the marginal grouping of the micas. This arrangement may be referred to as an illustration of a process tirst investigated by Soret'^ in tiie case of crystallization from aqueous solutions. He showed that if a constant difference of temperature is main- tained between two parts of a ves.sel containing a saturated solution, crystallization will proceed at tirst only in that part of the vessel which is at the lower temperature. Mr. TealP has sought to explain the concentration of the earlier formed basic minerals on the walls of some ingenous intrusions in terms of Soret's principle. The disposition of the mica flakes in the acid veins of the Moorooduc quariy may probably be referred to the same cause. Tlie Grano-diorite. — The plutonic mass of Mt. Eliza extends as a somewhat elliptical shaped mass ju.-t over two miles long from N.E. to S.W., and about a mile across in the widest part in a N.W. S.E. direction. Most of this area is covered with a mantle of granite detritus, and only two or three limited out- crops are seen of the rock " in situ. ' The best exposure occurs 1 Op. cit. 2 Ann. Chim. Phy., Paris, 1881, (.'>) 22, p. 203. 3 British Petroijraphy, p. 402. 98 Ernest W. Skeafs : in a shallow quarry near the summit of the hill. The rock is grey, fairly even grained, and felspar, quartz, blaick biotite and a little hornblende are visible in the hand specimen. Its specific gravity is 2.69. Under the microscope (Section oOi) it is no- ticed that both plagioclase and orthoclase are present, that some of the biotite has been altered to chlorite, abundant needles of apatite are included in the generally ragged crystals of biotite, and a little rutile ii5 ^^robably present. The symmetrical extinc- tion angles of the plagioclase lamellae ra-nge from about 11 deg. to 17 deg. The crystals are frequently zoned, the margins being invariably more acid, and are sometimes untAvinned. The cen- tral parts of the crysta)ls correspond to andesine of composition AbjAn.,, the mafl'gins to oligoclase of composition Ab^Auj. The average composition of the plagioclase as a whole is probably near AbgAug. The plagioclase is generally someiwhat kaolinised, and is usuallj'' idiomorphic. The orthoclase, containing some minute irregular intergTowths with albite is, however, fresh and moulded on the plagioclase. The structure of the rock, as a whole, i* hypidiomorphic, and the average grain-size is 1 mm. Petrographically, it should be classed with a number if other Victorian granitic rocks as a grano-diorite, on account of the large amount of quartz present, the considerable quantity of an alkali felspar, and the relatively acid character of the plagioclase.- present in this group of rocks. Professor Gregory,-"- following American usage, has suggested the application of this term in preference to Quartz-mica-di(jrite, to which group Dr. Howitt ha.> referred some of them. No chemical analysis of this rock is available, but an attempt has been made to determine, quantitatively, its mineral volume composition. From this the bulk mineral composition is found by multiplying the percentage volume of each mineral by its specific gravity. Finally, by accepting analyses of minerals having similar optical properties, an attempt has been made to determine approximately the chemical composition of the rock. The method followed in determining the volume percentage of each mineral in the rock is due to Rosiwal.- He has used a 1 The Geology of Mount Macrcli.n, VictoriiL, I'mi'. Roy. Soc. Victona, 1-1 (l!)0-i), i>. VM. 2 Verhanrtl d.k.k. Ocol Kfifli.s:uis),. ISils, pii. 14.S, ^•t^^■i\. Tlu' t^uiuititativi.' flii.ssitiiMtioii of Igneous Itooks, l!)0:i, p. '201. J. I'. l(ldiiij;s, Joiinial of (ieolo.M , vol. \ii. (l!)04>, p. 2.')2. Geohxji/ oj Mooroodac. 99 travelling stage or eye-piece micrometer to obtain a number of traverses acres* a microsection, and has shown that the volumes of the different minerals are proportional to the sums of their intercepts or any line or lines dra^vn across the rock, if the number of minerals traversed be sufficient. Applying this method, it wa« found that, out of a total length of 1035 units, the sums of the intercepts of the different minerals were as follows : — The percentage volumes are .shown in tlu' second column : — Plagioelase - - - 11 4 - 10 Quartz - - - - 3.)5 - 29.47 Orthoclase - - - IDS - 19.13 Biotite - - - . 113 - 10.91 Hornblende ... 5 - 0.48 Apatite (estimatedj - - 2.5 - 0.24 100.23 The specific gravities of the minerals is taken to be as follow : — Plagioclase (Abj, An,) - = 2.ti5 Orthoclase - - - - = 2.55 Quartz = 2.65 Biotite^ - - - - ^ 2.99 Hornblende - - - - = 3.28 Apatite - - - - = 3.20 Multiplying the percentage volumes of the minerals by their densities we ol)tain the proportions by weight whicli are then recalculated as percentages. Plagioclase Quartz Orthoclase Biotite Hornblende .Apatite 1 The specific s;ravity of the Biotite was deteniiined from Hakes b.v immersing them in Sonstadt's heavy liquid, and determining- by the Westphal balance the specific gravity of the liquid in which they floated in any position. The composition of the Biotite is assumed to be similar to that of C'hebarkul of similar specific gravity (see Dana's System of Miner- alogy, p. 630), while ttie Hornblende is assumed to be similar to that from a Vesuvian localitv. Gravimetric Pe rcentage Mineral proportions. Composition. 10(3.00 - 39.87 78.09 - 29.37 46.78 - 17.59 32.62 - 12.23 1.57 - .50 .so .30 265.86 99.95 100 Ernest W. Skeats : The plagioclase is assumed to liave the composition Abg Auj. Distiibutiiig the 39.87 per cent, among these two molecules we obtain Albite 25.63 per cent., Anorthite 14.24 per cent. Knowing the chemical composition of all the minerals and the percentage of each mineral present we arrive at the ultimate chemical composition of the rock. c S -J 1 c a 3 O < — 17.59 Ik 24 25.63 29.37 12.23 .59 .30 — SiO^ 11.38 6.15 17.62 29.37 4.70 .24 — 69.46 Al.,03 3.21 5.24 5.00 — 1.76 .09 — 15.33 Fe,0, — — — — .66 .01 — .67 FeO — — — — 1.80 — — 1.80 M^O — — .._ — 2.00 .07 — 2.07 CaO — 2.86 — — — .08 .20 3.14 K,0 2.96 — — — 1.00 — — 3.96 Na,0 — — 3.00 — .07 — — 3.07 P.O, — — — — — — •10 .10 H.,0 — — — — .13 — — .12 Total - 99.72 In this mineralogical analysis, ajpart from the small experi- mental error in the traverses and estimation of the volume com- position of the minerals of the rock, there are two sources of ambiguity. The Biotite analysis from Chebarkul, chosen for comparison on account of simiUur specific gravity, is that of a variety in which ferrous and magnesia oxides are present in almost equal amounts. The biotite in thi« grano-diorite may not have these oxides present in similar proportions. The other ambiguity rises from the fact that no allowamce has been made for the Ismail quantity of albite irregularly intergrown with the orthoclase. If this could be allowed for, potai^h would be slightly diminished and soda correspondingly increased in amount. Apart from these possible sources of error, it is believed that the figures fairly represent the chemical composition of the rock. Although tlie indicated silica percentage is higher than in some of the Victorian gi-ano-diorites, the high total of the alkaline earths and the lack of preponderance of the potash Geobxjn of Mnornod (ic. 101 in the alkalies, shows thii-t it should be grouped with the graiio- diorites rather than the gi'anites or gi'anitites. An analysis of the granodiorite from two miles N. or Dandenoiig township is appended for comparison.' (Jraiio-dioi i-ito (Ji •aMo-dioiite .Mount Eli Zll. N. 0! f OaiKleiioiiy:. SiO, - - 69.46 - f)3.38 Al.,6, - 15.33 - 17.36 Fe'.O; - - .67 1.61 FeO - - 1.80 - 1.98 MgO- - - 2.07 - 1.80 CaO - - 3 11 - 4.18 K.,0 - - 3.915 - .31 NaJO - 3.07 - 4.07 P.O5 - .10 - .54 H,0 - " .4:i - CO., Fes 1.13 3.38 99.72 99.74 SuMiM.VltY .\N1) (yONCLUSIOX. 1. This paper discusses the sedimentary, igneous and metar morphic rock* of Moorooduc, in the Mornington Peninsula, Victoria. 2. Tte previous literature on the area is discussed. Chrono- logically arranged, the salient features so far as they bear on this communication aire as follow : — 1856. Selwyn finds a oast of an eucrinite stem on Sand- stone Island, Western Port, and indicates the age of the older sedimentary rocks of the district as Silurian on his geological map. 1900. A. E. Kitson suggests a lithoLigical resemblance between the older sedimentary rocks of Moorooduc and the L. Ordovician rocks of Lancefield. He also notes the localization of the micas in the acid veins from the granite to the walls of the intrusion. 1900. Evelyn G. Hogg describes petrologically the granite rocks of areas adjoining Mt. Eliza. Some are de- scribed as " granitite," others as " syenite." 1 Geology of .Mount .Macedon, Proc. Koy. Soe. Viet., 14 (1!>02), p. 2(il. 102 Ernest W. S/,-eATK XIV. Geological sketch map of the Moorooduc district, reduced from the Geological sketch map of tlie Mornington district, by A. E. Kitson, F.G.S. Pl.mk XV. Upper figure. Grano-diorite, Mt. Eliza (No. 504) -f Xicols X 17 diameteis. Lower figure. Metamorphic spotted Ordovician Slate, N. of Moorooduc Railway Station (No. 506B), + Nicols. x 48 diameters.. H. L. Grayson, Photo-micro. Plate XVI. Kev to plate XV. Upper tigure. Q -= Quartz, Bi = Biotite, H = Hornblende, And = A°Klesine, Ab = Albiteand Albite-Oligoclase. Lower tignre. M = Micaceous alteration product, Q = Quartz, bI= Biotite, M. A. = Mica aggregates. Hem = Hematite. [Proc. Eot. Soc. Victoria, 20 (N.S.), Pt. II., 1907.] Akt IX. — On the Validitij of Callitrls Morrisoni. By R. T. baker, F.L.S. Curator Technological Museum, Sydney Communicated by Professor Alfred J. Ewart, Ph.D., D.Sc, F.L S. [Eead 14th November, 19D7.] Ill Vol. 20 (N.8.), part T., 1907, p. 76, of these Proceedings, Professor Ewart under " Contribu'ions to the Flora of Australia, No. 6," expresses an opinion as to the specific rank of this species. Inter alia he states, " Thei-e can be no doubt that this species (C robusta) is a variable one, but variations are often shown on one and the same specimen, and hence it is neces.sary to retain for it the scope given by Benthani, and include under it such varieties as niicrocarpa, verrucosa, intratiopica, and possibly also the columellaris of F. v. M. and the Morrisoni of R. T. Baker."^ Unfortunately no facts are educed to support the statement that there " can be no doubt that C robusta is a variable species," or that it is necessary to retain for it the scope given by Bentham and include niicrocarpa, verrucosa, intratropica, and possibly columellaris and C Morrisoni. That a variation of fruits can be found on the same twig is common in most species, but the point is, can fruits similar to C verrucosa, C. intratropica, C. Drummondii and C. calcarata be found on one and the same speci- men, for my species is allied to these two last and not C. robusta, which was a misprint in my paper ? I maintain, No. The establishment of C. Morrisoni was made only after 1. A thorough examination of all Callitiis material in the principal herbaria of Europe and Australia. 2. A thorough morphological examination of living viaterial of nearly every known Callitris species of Australia and Tasmania. 3. A macro- and microscopical examination of their timbers, barks, leaves and fruits. 1 Proc. Linn. Soc. N.S.W., vol. xxxi., 1906. Validity of Callitris Morrosoni 105 4. A cliemical investigation of their oils, saiidaraclis, camphors tfec. f). And lastly the aid of the physicist (wliich supports this differentiation) has been laid under tribute. Tlie result of ail this has been the accumulation of specific data that leave no alternative but to differentiate all these species enumerated (supra). That is, if differences constitute a species, as T believe they do. It would be asking the Royal Society too much to publish here all the results obtained in this connection, in order to prove the case, but it is hoped they will be in print next year. I might, however, state en passant tliat, concerning two of the proposed varietal forms, i.e. verrucosa and columellarLs, these two morphologically, cortically, ligneously, chemically, and physio- logically are quite different, and again any one who has compared only the timbers of intratropica and microcarpa would hardly be prepared to say they also are one and the same species. Unfortunately Professor Ewartdoes not say to which C. robusta his x'emarks refer. It was to clear the identity of this species more especially that the European herbaria were visited by me, for in my opinion it was hopeless to do it in Australia, and this was especially impressed on me after reading De Candolle's list of doubtful and excluded species of Callitris, [Prodromus. vol. IG. pp. 451-3]. To place ;i]l those Callitris enumerated by Professor Ewart under one species would be a parallel case to that of Baron von Mueller, who when dealing with Eucalyptus amygdalina, Labill, synonymised at least half a dozen good species under this name, which can all be shown to pos.se.ss distinctive nioi'phological, cortical, chemical, and other physiological differences from La Billardiere's species. The.se two ca.ses are only another illustration of the failure of morp]wh\^y alone in the determination of species in Eucalypts and Callitris. In this connection no better example can be quoted than that of Eucalyptus niaculata and E. citriodora. Both species were established by Hooker, and later weie synonymised l>y Baron von Mueller because morphologically the leaves and fruits were identical. .-> 106 M. T. Baker: Validity of CalUtris. Recent research has shown (1) that the two trees ditier in facies, being easily distinguished in the field ; (2) that they (lifter in the quality and texture of their timber and hark; (3) most decidedly in the chemical constituents of the leaf content. In all probability E. citriodora will be a source of considerable commercial enterprise in the future when it will be known as such, and not as E. niaculata, var. citriodora, of recent botanists, and the same remarks will also apply to the several Pines it is now proposed to classify as Oallitris robusta. [Phoc. Koy. Soc. Victoria, 20 (X.S.), Pt. II., 1907.] Anr. X. — 2Vif Voi'inatlon of Red Wood in Conifers. By jean white, M.Sc. [Read l-4th November, 1007]. A series of experiments was carried out by Professor Ewart and Mr. Mason-Jones, on Pinus contorta, and P. cembra, and on Cupressus nutkaensis and C. LaAvsoniana, in which certain of the lateral branches were curved round and tied securely for some time, until new wood could be formed. The results of their experiments, published in the " Annals of Botany," April, 1906, led them to conclude that in all probability the formation of red wood was a " morphogenic response to a g^'avitational stimulus." Professor Ewart suggested that I should carry the investiga- tions further by noting the effects of diffusing the actiiui of gravity, by causing a plaint to rotate on a klinostat. The plants experimented 'on were growing in flower pots, and included Araucaria excelsa, Callitris Gunnii, Cedrus deodara, Cr\'"ptomeria elegans, Cupressus sempervirens, Juniperub phoe- nicea, Pinus strobus, Podocarpus elata, Sequoia sempervirens, Taxus baccata. Thuja orientalis. On 20th September. 1906, the Cupressus was put on to the klinostat, which was set rotating at the rate of one turn in four hours. The remaining plants were laid down horizontally on their sides, in a glas>-house, the upper .side of each pot being marked. The plants had been in-evimisly tied to long stakes, so as to relieve the pressure on the under side. The plants were all examined on 28th November, 1906. and the following results were noted: — In Araucaria excelsa, Cryp- tomeria elegans, Juniperus phoenicea, Pinus strobus, Podocarpus ©lata. Sequoia sempervirens, and Taxus baccata, there was an extremely well-marked layer of red wood developed on the under side of the stem, as it lay horizontally. AI^o the red wood was very conspicuous on the under surface of all the lateral branches which were examined. -is 108 Jean White: In Callitris gunni, Cedrus deodara a.nd Thuja orientalis, there ■\va« a less oonspicuous layer of red tracheides produced on the under siu'face of the main stem and its branches. The Cupressus sempervireiis was ako removed from the klinostat on 28th November, 1906. During the two months' rotation, there were two stoppages of the klinostat for a possible duration of 16 hours and 3 hours respectivel3^ Examination showed a uniform exceedingly fadnt layer of red wood round the main stem and lateral branches. The Cupressus plant had two similar main i>tems, one of which v.as examined when it was first removed from the klinostat, and the other one after it was removed for the second time. The Cupressus plant was replaced on the klinostat on :29th November, 1906, the speed of rotation being changed to one revolution in two minutes. It was kept on the klinostat till 28th December, during which time there was a stoppage of the machine possibly for 30 hours. On examination of the stem, after removal of the plant from the klinostat on 28th December, no red wood was visible. Evidently, therefore, to produce any per- manent impression upon the developing cambial segments, the gravitational stimulus must last at least 1 to 2 hours. One minute's stimulation is either not perceived, or leaves the seg- ment cell in a labile condition, continuallj^ reversed by the completion of each rotation without producing any permanent and definite morphogenic response. The pot containing the ('upres>us plant was laid on its side in the glass-house on 25th January, 1907, the upper side of the flower put ))eing marked. It was left in this position till 18th October, 1907. On stripping oft' the bark, a thick layer of red wood, altout 20 tracheides deep, was observed on the under surface of the main steam and lateral branches. The above results serve to strengthen Professor Ewart's and Mr. Mason-Jones' conclusions as to the primary stimulus le- sponsible for the production of the red wood, being a gravita- tional one. Diameter of the Xijh i/i ]'rss<-/x, (mil 'Jlnrkiirss of Their Walls. The diameters of the cavities of the vessels, and also the tliickness of the vessel walls were measured by meaoas of the Red Wood 111 Colli f(')'s. 109 screw micrometer eyepiece, in both the red and white wood. Sections were cut of the main stems of several of the plants, Avhich had been growing in the pot^ placed on their sides from 20th September to 28th November, 1906. Taking the averages of the thickness of the walls in the red and white wood, the two sets of readings being taken from the same section as nearly diametrically opposite as possible, in practically eveiy case, the walls of the white tracheides were found to be thicker than those of the red, whilst the cavities of the traeheides of the white wood were smaller in diameter than those of the red tracheides. These results are not in accordance with those previously recorded by Son n tag, ^ who found that the walls of the tracheides in the red wood were thicker than those of the white. Experimental Results. A number of readings were taken of the internal diauietors and the thickness of the walls of the tracheides, and the results are given in the form of averages of sets of tive readings, fol- lowed l)y a\erages of tliese again. Taxus baccata. In the thickest part of the i-ed wood in the sections exannned, the tracheides were 20 deep. Section I. Red Wood White Wood Internal Diameter Thickness of Wall Internal Diameter Thickness of Wall .008 mm. .004 mm. .008 mm. .008 mm. .009 •, .004 ., .008 ,, .008 ., .011 ,, .004 ,, .008 ,, .007 ,, .012 .. .005 ., .009 ., .007 ,, Average .010 mm. Average 004 mm. Average .008 mm. Average .008 mm. 1 Jahrl). fur vviss, Bot., Bd. xxxix., p. 71. 110 Jean White : Section II. Eed Wood Internal Diameter Thickness of Wall .011 mm. .008 „ .010 „ .010 ,, .009 ,, .011 „ Average .009 mm. .006 mm. .006 ,, .006 „ .006 ,, .006 ,, .007 ,, Average .006 mm. White Wood Internal Diameter Thickness of Wall .007 mm. .008 ,, .008 „ .007 ,, .008 ,, •Oil ,, Average .008 mm. .005 mm. .006 .007 .005 .008 .007 Avei rage .006 mm. Section III. Red Wood Internal Diameter Thickness of Wall .006 mm. .007 ,. .008 ., .005 ., .005 „ .004 „ Average .006 mm. .008 mm. .008 .011 .008 .007 .010 Average .008 mm. White Wood Internal Diameter Thicknes of Wall .006 mm. .009 ,, .007 ,, .008 ,, .009 „ .008 ,, Average .008 mm. .008 mm. .008 ,, .006 ,, .007 ,, .007 ,, .006 ,, Average .007 mm. Pinus stpobus. Section I. Red Wood Internal Diameter Thickness of Wall .016 mm. .012 „ .012 ,, Average 0l;5 mm. .005 mm. .003 ,, .004 ,, Average .004 mm. White Wood Internal Diameter Thickness of Wall .014 mm. .012 „ .014 „ Average .013 mm. .005 mm. .005 ,, .003 ,, Average .004 mm. Red Wood Section II. Internal Diameter .015 mm. .014 ,, .012 ., Average .014 mm. Thickness of Wall .004 mm. .004 ,, .005 ,, Average .004 mm. White Wood Internal Kianieter .014 mm. .015 ,, .012 „ Average .014 mm. Thickness of Wal .005 mm. .005 ,, .005 ,, Average .005 mm. Red Wood in Conifers. Ill Araucania excelsa. Section I. Red Wood White Wood Internal Diameter Thickness of Wall Intertial Diameter Thickness of Wall .017 inm. .007 mm. .017 mm. .009 mm. .017 ,, .007 ,, .018 „ .009 ,, .020 ,. .007 ,, .016 ,, .009 ,, A\eraj;e Average Average Average .018 mm. .007 mm. .017 mm. .009 mm. Section II. Red Wood White Wood Internal Diameter Thickness of Wall Internal Diameter Thickness of Wall .016 mm. .006 mm. .020 mm. .009 mm. .018 ,, .008 ,. .017 ,, .009 ,, .017 ,. .007 .. .016 ,, .008 „ Average .Average Average Average .017 mm. .007 mm. .018 mm. .009 mm. Podocappus elata Section I. Red Wood White Wood Internal Diameter Thickness of WuU Internal Diameter Thickness of Wall .014 mm. .007 mm. .008 mm. .007 mm. .011 ., .005 ,, .011 „ .007 ,, Average Average Average Average .012 mm. .006 mm. .009 mm. .007 mm. Section II. Red Wood. White Wood. Internal Diameter Thickness of Wall Internal Diameter Thickness of Wall .012 mm. .009 mm. .008 mm. .007 mm. .012 ,, .005 ,, .011 ,, .007 ,, Average Avei"Bge Average A\erage .012 mm. .007 mm. .009 mm. .007 mm. Summary of Precedin^^ Results. The pliiiit.s were placed hoi-izontally, the upper side being tlie moi'e sti-ongly iUuminated. 112 Jean White : Taxus baccata. Thickness of Walls Internal Diameter Red Wood White Wood Ked Wood White Wood .004 mm. .008 mm. .010 mm. .008 mm. .006 ,, .006 ,, .009 ,, .008 ,, .006 ,. .007 ., .008 ,, .008 .. Average Average Average -Average .0053 mm. .0070 mm. .0090 mm. .0080 mm. PlNUS STUOBUS. .004 „ .004 ,, .013 ., .013 ., .004 „ .005 „ .014 ., .014 ,, Average Average Average Average .0040 mm. .0045 mm. .0135 mm. .0135 mm. Araucaria EXCELSA. .007 „ .009 ,, .018 ,, .017 .. .007 ,, .009 ,, .017 ,, .018 ,, Average Average Average Average ,0070 mm. .0090 mm. .0175 mm. .0175 nun. fODOtARPUS ELATA. .006 ,, .007 .. .012 ,, .009 ., .007 ,, .007 ,, .012 ,, .009 ., Average Average A\erage Average ,0065 mm. .Ot)70 mm. .0120 mm. .0090 mm. The above results appeared to indicate that the thickness of the tracheide walls might be influenced by either pre.ssure or illu- mination, or both. In order to investigate this matter further, some of the lateral branches of Cedrus deodara.. Thuja orientalis, Callitri.s (lunnii, Cryptomeria elegans, and Pinus strobus, were curved rnund and tied in the >ame manner as were those described by Professor Ewart and Mr. Mason-Jones. ^ The plants were set upright in the glass-house on 21st May, 1907. Parts of the lateral branches of Taxus baccata, Podocarpus elata, a-iid Araucai-ia excelsa were bound round with tinfoil, and the pot« were laid hcrizontally in the glass-house, and the uppermost part of the pot marked, also on 21st May, 1907. 1 Annals of Botany, vol. xx., |). 202. Red Wood in Conifers. 113 On 10th August, 1907, some of these plants were examined. Very conspicuous layers of red wood were developed on the under surface of both parts of the curve, just as described in the " Annals of Botany."' Sections were cut from the parts of the curve where the development of red wood was greatest, and the internal diameters of the red and white tracheides, and also the thickness of their walls, were measured. Experimental Results. Pinus strobus. Sectiox I. — (From upper purtion of curve). Red Wood White Wood Iiittiiial Diameter Thickness of Wall Internal Diameter Thifkiiess of Wall .008 mm. .003 mm. .012 mm. .006 mm. .010 ., .002 ,, .012 ,, .006 ,, .009 ,, .003 ,, .010 ,, .006 „ .010 ,, .002 ,, .010 ,, .007 ,, Average Average Average Average .009 mm. .003 mm. .011 mm. .006 mm. Section 11. Red Wood White W GOD Internal Diameter Thickness of Wall Internal Diameter Thickness of V .009 mm. .003 mm. .008 mm. .005 mm .012 ,, .003 ,, .010 „ .006 „ .012 ,, .002 ,, .009 „ .006 ,, .009 ,, .002 „ .010 „ .007 ., Average Average Average Average 010 mm. .003 mm. .009 mm. .006 mm Section III. Red Wood White Wood Internal Diameter Thickness of Wall Internal Diameter Thickness of Wall .009 mm. .002 mm. .013 ,, .008 ,, Average .009 mm. .003 ,, .002 ,, Average .002 mm. .008 mm. .010 „ .009 „ Average .009 mm. .006 mm. .006 !> .006 )) Average .006 ram . 114 Jean White Callitrls gunni. Section I. Red Wood Internal Diameter Thickness of Wall .012 mm. .006 ,, .009 ,, .009 ,, Average .008 mm. .002 mm. .002 ,, .003 . . .002 ,, Averaj^e .002 mm. White Wood Internal Diameter Thickness of Wall .009 mm. .008 ,, .008 „ .008 „ Average .008 mm. .004 mm. .003 ,, .005 ,, .003 ,, A\erag;e .002 mm. Section II. Eed Wood Internal Diameter Thickness of Wall .004 mm. .002 ,, .001 ,, .002 ,, Averaj;e .002 mill. .009 mm. .006 M .008 .008 ) J Average .008 ram White Wood Internal Diameter Thickness of Wall .009 mm. .006 ,, .006 ,, .008 ,, Average .007 mm. .005 mm. .004 ,, .003 ,, .003 ,, Average .003 mm. Cryptomeria elegans. Section I. Eed Wood White Wood Internal Diameter Thickness of Wall Internal Diameter Thickness of Wall .006 mm. .002 mm. .006 mm. .007 ,, .001 ,. .006 ,, .008 ,, .001 ,, .008 „ .008 ,, .002 ,. .009 ,, Average Average Average .007 mm. .002 mm. .007 mm. .OCo mm. .004 ,1 .006 J J .005 )> Average .005 mm. Section II. Red Wood Internal Diameter Thickness of Wal .007 mm. .009 „ .008 ,, .010 ,, Average .009 mm. .002 mm. .002 ,, .001 ,, .002 ,, Average .002 mm. White Wood Internal Diameter Thickness of Wall .006 mm. .006 ,, .008 ,, .009 „ Average .008 mm. .005 mm. .006 > > .005 )> .003 ,, Ave rage .005 mm. Red ^Voo(l in Conifers. 115 Sununary oj Preceding Results. The stems were bent in curves. The white wood side was under tension, and was the more strongly illuminated side. PiNUS STROBU8. Thickxess of Walls Red Wood White Wood Internal Diamet?:k Ked Wood White Wood .003 mm. .006 mm. .009 mm. .011 mm. .003 „ .006 ,. .010 ,, .009 ,, .002 ,, .006 ,, .009 ,, .009 ,, Average A\eratfe Average Average ,0028 mm. .0060 mm. .0091 mm. .0093 mm. C'ALLITRIS GUNNI. .002 „ .003 ,. .008 ,, .008 ,, .002 ,, .003 ,, .008 ,, .008 ,, Avera^je Average Average Average .0020 mm. .0030 mm. .0080 mm. .0080 mm. Cryptomeria ELEOANS. .002 ,, .005 ,, .007 „ .007 „ .002 ,, .005 ., .009 ,, .008 ,, Averaire Average Average Average .0020 mm. .0050 mm. .0080 mm. .0075 mm. On August 1 6th 1907, the lateral branches which had been covered with tinfoil wei'e examined. .^ layer of red wood was observed on the under side of the branches which had been covered with tinfoil, as before. Hence Sonntag is incorrect in supposing that heliotropic or pressure stimuli are responsible for the formation of redwood. Sections of the lateral bi'anches which were covered were cut, and the thickness of the trache- ide walls and their inteiiial diameters were measured. EAperiiiicuiaJ Results. Taxus baccata. Section I. Red Wood Internal Diameter Thickness of Wall White Wood Internal Diameter Thickness of Wall .010 mm. .008 ,. .009 ,, .009 ,, Average .009 mm. .001 mm. .002 ,. 004 .. .004 ,. Average .003 mm. .008 mm. .005 ,, .007 „ .006 ,, Average .006 mm. .005 mm. .002 ,, .002 ,, .003 ,, Average .003 mm. 116 Jean White : Section II. Red Wood Iiitei-iial Diameter Thickness of Wall .009 mm. .007 „ .008 ,, .008 ,, Average .008 mm. .004 mm. .004 ,, .004 „ .005 „ A\ era^je .004 mm. White Wood Internal Diameter Thickness of Wall .008 mm. .005 ., .006 ,, .006 ,, Averas^e .006 mm. .005 mm .005 >5 .005 J J .004 .> A\erage .004 mm Podocappus elata. Red Wood Internal Diameter Thickness of Wall Section I. White Wood Internal Diameter Thickness of Wall .010 mm. .002 mm. .009 mm. .001 mm. .016 ,, .004 ,, .009 ,, .004 „ .012 „ .006 ,, .012 ,, .005 ,, .009 ,, .005 ,, .010 „ .005 ,, Averafie Average Average A\erage .011 mm. .004 mm. .010 mm. .004 mm. Section II. Red Wood Internal Diameter Thickness of Wall White Wood Internal Diameter Thickness of Wall .010 mm. .003 mm. .012 mm. .003 mm. .009 ,, .007 ,, .013 ,, .005 „ .008 ,, .006 ,, .012 „ .005 „ .009 .. .005 „ .012 „ .008 ., Average A\erage Average Average .009 mm. .005 mm. .012 mm. .005 mm. Red Wood Internal Diameter Thickness of Wall Section III. White Wood Internal Diameter Thickness uf Wall .013 mm. .015 ,. .010 ,, .008 ,, Average Oil mm. .003 mm. .005 ,, .006 „ .003 ,, Average .004 mm. .013 mm. .015 ,, .010 „ .008 „ Average .011 mm. .003 mm. .005 „ .006 ,, .003 „ Average .004 mm. Red Wood In Conifers. II Araucaria excelsa. Section I. Kki) Wood Interiiiil Diameter Thickness of Wal .013 mm. .010 ,, .010 „ .010 „ Averaj^e .011 mm. .1)03 mm. .001 „ .002 ,, .003 ,, Average .002 mm. Whitk "VVoud Inteiiial Diameter Thickness of Wall .022 mm. .016 „ .015 „ .013 „ Average .014 mm. .003 mm. .004 „ .003 „ .003 „ Average .003 mm. Section II. Intern; .016 mm .021 ,, .017 ,, .018 ,, Average .018 mm Red "Wood Diameter Thickness of Wal' .005 mm. .004 „ .005 ,, .006 ,, Average .005 mm. White Wood Internal Diameter Thickness of Wall .020 mm. .016 ,, .016 „ Average .017 mm. .005 mm. .003 „ .006 ,, Average .005 mm. Section III. Eed Wood Internal Diameter Thickness of Wall .013 mm. .018 ,, .006 ,, ■016 „ Average .013 mm. .005 mm. .008 ,, .005 ,, .006 „ Average .006 mm. White Wof)D Internal Diameter Thickness of Wall .013 mm. .023 ,, .013 ,, .020 „ Average .017 mm. .004 mm. .006 ,, .003 ,, .005 „ Average .005 mm. SHiiiifiary of Preceding Results. Both .side.s were; ecjually cUirkened. The red wood side was under compression, find the white wood side was under tension. Taxus baccata. Thickness u y Walls ileii Wood White Wooii .003 mm. .003 mm. .004 . .004 ,, Average Average .0035 mm. .0035 mm. Intkknal Diameter lied Wood White Wood .009 mm. .006 mm. .008 „ .006 „ Average .\verage .0085 mm. .0060 mm. 118 Jean White : POUOCARPUS ELATA. .004 „ .004 .. .011 ., .010 ,. .005 ., .005 ,, .009 ,. .012 .. .004 ., .004 ,, .011 ,. .011 ,, Avei'ai,'e A\'eraj;e Averaf^e Average .0043 mm. .0043 mm. .0101 mm. .0110 mm. Araucaria excelsa. .002 ., .003 ,. •Oil .. .014 ,. .005 ,, .005 ,, .018 ,, .017 ,, .006 ,, .005 ,, .013 ,, .017 ,, Average Avei'aye Average A veray e .0043 mm. .0043 mm. .0140 mm. .0490 mm. The^e sections, taken from the curved lateral liranehes. were all cut from the upj^er portion of the curve, so that tlie red wood vessels were subjected to compresision. and the white wood vessels were subject to tension. Concerning the thickness of the walls, the ratio of the thickness of the white to the red ti-acheides is not very different from their ratio when they were not subjected to any special pressure, and so, presumably, the thickness of the w^alls does not to :iny appre- ciable extent depend on pressure effects of the intensity JjJ'o- duced by forcibly bendinp- a stem into circular form or laymc; a vertical stem in a horizontal poi>ition. Also in those .sections, cut from the parts of the branches covered with tinfoil, in prac- tically every case, it was found that the thickness nf the walls of red wood vessels and white wood vesseli? was the same, which indicates that photomorphic stimuli take an important part in the regulation of the thickness of the walls. In this respect my experiments; appear to agree with those of KiiiL'ht,^ who found, for instance, that roots freed from soil a-nd exposed to light formed firmer wood. The C'ryptomeria, Callitris and Pinus had some of their lateral branches curved and tied round in the manner described pre- viously, the upper part of the curve being covered over with tinfoil. Ttiey were placed upright in the glass-house nn 16th August, 1907. Measurements of the diameters of the red and white wood vessels, and of their Avails, were taken on 4th November, 1907. IPfeffer's Ph.vsiolojr.v of Plants, Riij^lish Translation. Vol. ii.. page 88. Red Wood in Conifers. IIU Ked "Wood Internal Diaiuetei' .009 mm. .008 ,, .008 ,, .009 „ Average .008 mm. Experimentd/ Results. Callitpis Gunnii. White Wood Tliickiiess of Wall Iiittirial Diameter Thickness of \V;ill .005 mm. .008 mm. .005 mm. .005 ,, .009 ,, .008 ,, .006 ,, .008 ,, .005 ,, .005 ,, .006 ,, .005 ,, Average Average A\ernge .005 mm. .008 mm. .006 mm. Cryptomeria elegans. Red Wood Internal Diameter Thickness of Wall .005 mm. .006 mm. .008 ,, .006 ,, .005 ,, .004 ,, .006 ,, .004 ., Average Average .006 mm. .005 mm. White Wood Internal Diameter Thickness of Wall .006 mm. .005 mm. .005 ,, .005 ,, .005 ,, .004 ,, .005 „ .004 ,. Average Average .005 mm. .005 mm Pinus strobus. Red Wood Internal Diameter Thickness of Wall White Wood Internal Diameter Thickness of Wall .010 mm. .006 mm. .009 mm. .008 mm. .010 ,, .008 ,, .008 ,, .008 ,, .012 ,, .005 .. •013 ,, .005 ,, .009 „ .006 ., .010 ,, .005 ,, Average .\verage Average Average .010 mm. .006 mm. .010 mm. .006 mm. Summary of Preceding Experiments. Callitrib ounni. Internal Diameter Thickness of Walls Red Wood White Wood Red Wood White Wood .008 mm. .008 mm. .005 mm. .006 mm. .006 CrYPTOMERIA Er,ECiANS. .005 .. .005 , .005 120 Jean White : .010 mm. PiNUS STROBUS. .006 mm. .010 mm. .006 mm. A large branch of Cupressus sempervirens with sufficient bark and phloem to cut off nearly all light from the cambium was removed from the tree on which it was growing normally. An extremely thick layer of red wood was visible on the under side of the branch as it grew. Sections of the red and white wood were examined, and the diameters of the cavities, and the thickness of the walls of the xylem vessels in each kind were measured, -wath the following results : — Bed Wood Internal Diameter Thickness of W:ill White Wood Internal Diameter Thickness of Wall .012 mm. .004 mm. .016 mm. .006 mm. .016 „ .004 ,. .016 „ .005 ,, .012 „ .005 ,, .013 „ .005 ,, 015 ,, .005 ,, .018 ,, .005 „ Average Average Average Average .0143 mm. .0045 mm. .0156 mm. .0052 mm. In this case the difiereiice between the intensity of the illum- ination on the upper (white wood) and the lower (red wood) surface must be extremely small, and the average thickness of the white wood walls is little or not at all greater than in the red wood. Sum/nary of Preceding:; Experiments. Thickness of Walls Internal Diameter Red Wood White Wood lied Wood White Wood .0160 mm. .0052 mm. .0140 mm. .0045 mm. Pinus strobus. Plant grown horizontally. White wood on illuminated side. Thickness of Walls Internal Diameter Red Wood White Wood Ked Wood White Wood Average .0040 mm. Average .0045 mm. Average .0135 mm. Average 0135 mm. Red Wood in Conifers. 121 Plant grown veitically. Curved stein. Red wood compressed. White wood stretclied and illuminated. Thickness of Walls Internal Diameter Red Wood Average .0033 mm. White Wood Average .0060 mm. Red Wood Average .0093 mm. White Wood Average .0096 mm. Curved stem. Both sides equally darkened. Thickness of Walls Internal Diameter Red Wood White Wood Average .0060 mm. Average .0060 mm. Red Wood Average .0100 mm. White Wood Average .0100 mm. Araucaria excelsa. Plant grown horizontally. White wood on the illuminated side. Thickness of Walls Red Wood Average .0070 mm. White Wood Average .0090 mm. Plant grown upright. Both sides equally darkened. Thickness of Walls Red Wood Average .0043 mm. White Wood Average .0043 mm. Internal Diameter Red Wood Average .0175 mm. White Wood Average .0175 mm. Internal Diameter Red Wood Average .0140 mm. White Wood Average .0160 mm. Podocarpus elata. Plant grown horizontally. Whited wood on the illuminated side. Thickness of Walls Internal Diameter Red Wood Average .0065 mm. White Wood Average .0070 mm. Red Wood .■\verage .0120 mm. White Wood Average .0090 mm. 3 122 Jean White Plant i^rowii upright. Both sides equally darkened. Thickness of Walls Red Wood White Wood Average .0043 mm. Averasje .0043 mm. Internal Diameter Red Wood White Wood Average Average .0103 mm. .0110 mm. Tax US baccata. Plant grown horizontally. White wood on the illuminated side. Thickness of Walls Red Wood Average .0053 mm. White Wood Average .0070 mm. Internal Diameter Red Wood Average .0090 mm. White Wood Axerage .0080 mm. Plant grown vertically. Both sides equally darkened. Thickness op Walls Red Wood White Wood Average Average .0035 mm.. .0035 mm. Internal Diameter Red Wood White Wood Average Average .0085 mm. .0060 mm. Cpyptomepia elegans. Plant grown vertically. Lateral branches curved. White wood on illuminated side. Thickness of Walls Red Wood Average .0020 mm. White Wood A\eragc .0050 mm. Internal Diametkr Red Wood Wliitu Wood A\erage Average .0080 mm. .0075 mm. Lateral branches curved. Both sides e(iually darkened. Thicknkss of Walls Internal Diameter Red Wood Average .0050 mm. White Wood Average .0050 mm. Red Wood Average .0060 mm. White Wood Average .0050 mm. Red Wood in Conifers. 123 Callitpis Gunnii. Plant grown vertically (lateral branches curved). White wood on the illuminated side. Thickness of Walls Intkrnal Diameter Red Wood White Wood Ked Wood White Wood .008 mm. 008 mm. .008 ,, .008 ,, Average Axerajje Average Average .0020 mm. .0030 mm. .0080 mm. .0080 mm. Lateral branches curved. Both sides equally darkened. Thickness of Walls Internal Diameter Red Wood Wliite Wood Red Wood White Wood .0080 mm. .OOGO mm. .0080 mm. .0050 mm. Conclusion. So far as my results dealing with this matter go, they point to the conclusion that the formation of red wood is primarily due to a gravitational stimulus, while the lesser thickness shown by the wall of the red wood tracheides, as compared with that of the white wood tracheides. appears to be largely the result of a photomorphic stimulus, the response being somewhat akin to etiolation in character. The preceding averages all agree in showing that the thickness of the tracheide walls on the more strongly illuminated side ex- ceeded that of the tracheide walls where the illumination was less intense. The same uniformity, under similar conditions, does not apparently prevajil in the size of the internal cavities of the tracheides ; thus as a general rule- along a single radial row of tracheides, isolated cases occurred in which the internal cavities were of abnormal size in either direction, while such aibrupt variations did not appear to occur to any marked extent in the thickness of the tracheide walls of either kind of wood. Considering the cases in which both sides of the branches were equally darkened, the assumption is :-trengthened by the fact that in eveiy case tested except one. the thickness of the tracheide walls in both red and white wood tallied exactly. 124 Jean White: Red Wood in Conifers. As in some of the above cases, the parts of the ?tem which were equally darkened were subjected to unequal pressure or tension, due to the curving round of the branches experimented upon, the thicknes,s of the tracheide walls in red and white wood was the saaiie, it seems to be improbable that pressure and ten- sion of the intensity produced by forcibly curving the branch, exert any pronounced influence on either the thickness of the walls, diametei' of the tracheides, or formation of red wood. The latter is purely a response to gravity, and is only accompanied by an increase in the thickness of the walls when the red wood side is the more strongly illuminated one, which is unusual. The minimal period for perception and response, as tested by the method of summation, is two hours. Exposures to gravity of less than two minutes' duration produce no permanently lasting effect. In conclusion, I wish to record my sincei'e thanks to Prof. Ewart for his assistance, and also for allowing me the use of the Botanical laboratory at the Melbourne University. [Proc. Eoy. Soc. Victoria, 20 (N.S.), Pt. II., 1907.] AiiT. XI. — Contributions to the Flora of Australia, No. 7} By ALFRED J. EWART, D.Sc, Ph.D., F.L.S., Goveniment Botiinist and Professor of Botany at the Melbourne University. [Read 14th November, 1907.] Latin in Systematic Botany. At the last Botanical Cnugress, held at Vienna in 1905, on the whole a salutary check was administered to the objectionable tendencies of modern systematists in certain quarters, especially as regards frivolous changes of name, and it is, in fact, a matter of regret that the list of protected nameis was not greatly in- creased. On the other hand, it is impossible to follow !Mi\ Maiden" when he states that bntanists are as bound by these laws as by those of their own country, and must follow these laws whether they approve of them or not. "For this to be requisite the Congress would need to be a really representative one, to which all botanists sent elected representatives. At present it is a fortuitous concourse almost solely of systematists, among whom the local interests of the country in which the Congress is held are always unduly strongly represented. So far as I am aware, botanists from the south of the Equator were entirely unrepresented, and plant physiologists and anatomists were conspicuous by their absence. Yet the man who has intimately investigated the structure and properties of a plant has a greater claim to decide that its name shall not be aJtered than the systematist whose interest in the plant largely ceases as soon as it is labelled, and is often only revived when a chance of relabelling it occurs. 1 No. C in Proc. Roy. Soc. Vict, 1907, vol. 20, p. 76. 2 Jour. Roy. Soc. N.S. Wales, vol. xl., 1906, p. 74. 126 Alfred J. Eivart : Until the Congress if; a thoroughly representative one, it must remain a purely voluntaiy matter with each botanist as to whether he follows its rules or not, and the power of the Con- gress to enforce its rules will depend solely upon the number of botanists who elect to follow them. Under these circumstances I must take strong exception to Art. 36, and, by disobeying it, adopt the best plan to have it rescinded or altered. Art. 36 reads : " On and after Jamuary l«t, 1908, the publica- tion of names of new groups will be valid only when they are accomjoanied by a Latin diagnosis."' In Art. 13 a group i-s defined as including a species. Any practice which tends to render a science luinecessarily inaccessible to the general public is bad in principle, and ultimately reacts injuriously upon the science in question, and upon the eclectic few connected with it. Latin is thoroughly discredited as a scientific language, and in re-adopting it systematists are taking a step back to the middle ages. If the rule had been to the effect that diagnoses not ^vrit- ten in Engli^^h, French, or German, or unaccompanied by diag- nostic figures must be written in Latin, less exception could have been taken to it, although it would have been more satisfactory to state that diagnoses not accompanied by analytic figures, must be written in English, French or German. A good diag- nostic figure is worth a dozen pages of the average systematist's dog Latin, which at its best would hardly satisfy even Tacitus, and at its worst is sufiicient to make Cicero turn in his grave. To describe plants both in the author's language and in Latin would be to unnecessarily increase the already enormous bulk of systematic literature, and to swell its pouring torrent to a permanent flood level. To avoid this, and as a protest against the rule, the plants, in the present and subsequent papers, will be given, as hitherto, with diagnoses in Englisli, and if neces- sary with explanatory figures. Any Latinist who would like to see his initials after a plant name is at liberty to acquire this right by publishing a translation in Latin of the plant diagnosis here given, and thus following the rule laid down by the last Congi'css. I shall make no complaint, and am willing to take this risk in order to get an absurd law altered. It is a pity the rules were not submitted to some well-known authority on jurisprudence before publication. Thus the oniis- Flora uj Australia. 127 sion of the word Latin in Art. 37 renders Ai'ts. 36 and 39 invalid, or at least renders their interpretation doubtful in many cases. By means of Art. 37, it would be possible in a round- about way to force the acceptance of a new species according to Congress rules without a Latin diagnosis. Further, to change the name or authority for a new species because it had not been published with a Latin diagnosis would be to act in flat defiance of Art. 50, and other instances of rules whose effects are difficult to harmonise might be given. Nomina Couservanda. — It is gi*eatly to be regretted that the time limit for fhiuiige of name was put so far back as 1753, and that the list of nomina consei-vanda was not greatly extended. To give an instance. Anthistiria L. (Gniiiiinae), 1779, is changed to Themeda, Forst, 1775, by Haeckel in De Candolle's Monograph and in Engler's Pflanzenfamilien. It is impossible to accept any such change of a Linnean name on such slender gi'ounds as a four years' priority, when a name has been univer- sally accepted for over 120 years. Questions of general con- venience oveiTide any such claim in a case of this kind. Acacia accola, Maiden and Betche. Proc. Linn. Soc. N.S.Wales, 1906, p. 73-1. (Leguniinosae). This appears to be .a narrow-leaved and broad-fruited form of A. neriifolia. A specimen from Bailey resembles Maiden's form more closely as regards the fruit and the funicle of the seed, but has the broader phyllodes of A. neriifolia. Probably the future discovery of other intervening forms will render advisable the reduction of this species to a variety. ADiiXAXTHos rv(;xoKU.M, Diels. Fragin. Phyt. Aust. Occid., p. 1 38. (Proteaceae). This " species " is made to include the A. apiculata of Meiss- ner, and the Drummond specimens of A. sericea. The species is, however, undoubtedly the same as A. *fericea, Benth., a.nd if Dr. Diels had seen No. 788 as well as No. 787, he would probably not have made this error. There can be no doubt that many accepted species of this genus will be ultimately reduced to varie- ties as the result of cultural observations, and hence r rsat care 128 Alfred J. Ewart : should be exercised to avoid creating useless synonyms by con- clusions made without such observations in the case of highly plastic genera of this character. It is also doubtful whether the A. Drummondii, Meisn., revived by Diels, represents more than a vai-iety of A. apiculata, R.Br. AlZOON INTERMEDIUM, Diels, and AlZOON GLABRUM, n. sp. (Aizoaceae). The former sipecies is distinguished by Diels from A. zygophylloides (F. v. M.), by the shape of the leaves, longer pedicels and nan'ow calyx lobes. It comes very close to i?ome nearly smooth stemmed specimens included by F. v. Mueller in A. zygophylloides, and may ultimately prove to have not more than ai varietal significance. It is, however, quite distinct from Luehmann's undescribed xVizoon glabrum. This is a rather small plant, spreading more or less from a single root, the slender wiry glabrous stems, 2 to 6 inches high, simple or branching one or more times, bearing terminal fiowers in loose cymes on short pedicels, one or two pairs of linear leaves being close under the flower, which is sometimes an inch across when fully open, but usually less. Calyx 4 partite, usually divided nearly to the base, enlarging during flowering to neai'ly ^ incli in length, in large, fully-opened flowers, the lobes more or less acuminate, usually lanceolate, but not always of equal breadth in the same flower. Stamens numerous. Styles 4. Capsule dehiscing into 8 valves. Seeds numerous, almost black, shaped like the head of a mace and covered with small tuberculate spines. Murchison R., I. Tyson, 1898/ Mt. Caroline, 1891, Miss Sewel ; Salt Lakes, Martha Heal. AlZOON KODWAYI, n. S]). Plant 3 to over 8 inches high, stems more or less decumbent at base, and spreading. Leaves in opposite pairSi, soft, fleshy, with scattered wairty, transparent tubercles, ovate or linear, mostly ^ inch long, but beneath each flower usually a larger paiir more pointed and with broader bases. Plant glabrous throughout, the stems more slender than A. (piadrifldum, but stouter than A. glabrum. Flowers large terminal, 1 to 1^ inches Flora of Aa.sli-allii. 129 diameter when fully expanded. Calyx divided to about the middle, the five segments with brnad ba>es and more or less bluntly acuminate tips. Other fea^tures much as in A. quadri- fidum. Seeds apparently reddish-brown, but other^\ase as in A. glabrum. The absence of any t^curfy tomentum at once distinguishes these two species from A. quadrifidum. In addition, A. Rodwayi has broader ovate or lanceolate leaves, the calyx is less deeply divided, the flower larger and more bulky at its ba^e. This, with the less deeply divided calyx and the shape of the leaves dis- tinguishes the plant from A. zygophylloides. A. glabrum is told by its glabrous wiry stems, smaller flowers a;nd calyx deeply divided to the base. I. Tyson, Salt Marsh, W. Australia, 1893; F. A. Rod way, M.B., dried up salt lake, De^demona, W. Australia, 1907. Angiantiius iiUMiFi'HUS, Bentli., var. grandiflorus. (Compositae). In the last contribution to the Flora of Australia, No. 6, this was erroneously given as a new variety by the accidental omis- sion of a proof correction. Cas.sixia i.aevis, 1\. Br. (Compositae). The record from C. French, Goulburn R., under the above heading, in Contributions to the Flora of Australia, No. 6, should apply to Cassinia arcuata, R.Br., ^\Tongly recorded as C. Theodori, F. v. M. CONOSPKKMUM POLYCKPII ALUM, Meisn., val'. LKIANTHUM, Benth. (Pi-oteaceae). Diels and PritzeU raise this variety to specific rank as C. leianthum, Benth. The material at the Nati'jnal Herbarium, a part only of which appears to have been examined by Diels and Pritzel, i^hows conclusively that there is no reason for this change. It is impossible to lay down any clear line of demarca- 1 Fragni. Phyt. Austr. Occid., p. 141. 130 Alfred J. Eivart: tion ba.sed on a L'Toup of constant characters, for Diels' dis- tinctions do not apply to all the specimens between this variety and the type species. The typical form of the closely allied C. Toddii of F. Mueller^ shows a trifling difference in the size and acuminate character of the bracts, while the perianth tube is rather more slender, is longer in proportion to the lobes, and somewhat more slender, and is more pubescent outside. Even these characters do not apjiear to be quite constant, so that C. Toddii may also ultimately prove to be a variety of C. poly- cephalum when more intermediate material is available. Daviesia coitYMBOSA, var. St. Johnii = D. corymbosa, var. viRUATA. (Papilionaceae). This plant was recorded in the Victorian Naturalist, Nov.> 1906, p. 133, and specimens have since been received from Mr. C. French, Jr., collected at Ringwoiid. They are identical with the D. virgata of Cunningham, which Bentham refers to D. corymbosa, var. mimosoides. The condensed clusters of small flowers and the very narrow leaves would, as suggested by Mr. W. R. Guilfoyle, justify the recognition of a second variety differing more widely from the type than var. mimosoides, even though transition forms occur, but the name should be variety virgata, in recognition of the old specific name for the variety. DiPLOTAXis MUUALis, D.C., the wall or sand mustard. (Cfuciferae). This introduced alien has been variously referred to as D. nmralis and D. tenuifolia, D.C. It has, however, the small flowers and less divided leaves of the former. Tlie Australian specimens have the leaves less exclusively radical and more on the stem, and are often double the normal height, frequently reaching 12 to 16 inches. The plants also r^how a gi-eater ten- dency to be perennial, but these changes ai'e probably climatic ones, not necessarily of varietal significance, although they are a[)pr()aches towards I), tenuifolia, D.C. 1 Fragm., vol. x., p. 20. Flora of Australia. 131 Eriostemon gracile, R. Grah. (Proteaceae). In Edinb. N. Phil. Journ., xvi., 1834, p. 175 = Eriostemon difformis, A. Cunn. The former name is given as that of a valid species in the Kew Index, iind the Nat. Herbarium possesses specimens from the Grampians, which are indenticaJ. with forms of E. difformis, the glabrous petals and slightly longer flower stalks of the specimens being variable features in E. difformis. The speci- mens do not exactly tally with the description given in Mueller's Plants Indigenous to Victoria, L, 1860, p. 125, but our specimens appear to be authentic. E. gracile is the older name, but to change the current one would be a frivolous interference with established nomenclature. Bentham seems to have entirely overlooked E. gracile, and makes no mention of it in the Flora Australiensis. Eriostkmon intermedius. (Proc. Roy. Soc. Vict., 19, 1907, p. 40 = E. DESERTi, Pritzel (Fragm. Phytog. Austr. Dec, 1905. p. 320). The plant was described before Diels' and Pritzel's work was available. Their description is exceedingly condensed, and im- perfect in several respects, but specimens of their plant since received show that the two species are identical, the older name standing. Pritzel seems to have overlooked the fact that the plant is an interesting connecting link, especially as regards the stamens between the Leionema section of " Phebaliuiii " and EriostenK^n proper. The close resemblance to E. Brucei, which misled Mueller, and to which Pritzel attaches undue importance, is mainly external. Euphrasia collina, R. Bv. (8yn. E. Jjrowmi, F. v. M., Fragin., v. 88. (Scrophulariaceae). There can be no doubt that Mueller wa's correct in placing four of R. Brown's species in one, but as was pointed out by Bentham, the proper course was to extend one of them to include the others, and so avoid a new name. Even considered as varieties, the line of demarcation is not distinct in all cases, and the type forms show a regular gradation from dwarf, small- 132 Alfred J. Eivart : flowered form* to taller, more luxuriant and larger-leaved and flowered forms, in the following sequence : — E. collina, K.Br., var. striata. (E. striata, R.Br. ; E. alpina, var. humilis, Benth.). E. collina, R.Br., var. alpina.. (E. alpina, R.Br. ; E. die- menica, Spreng.). (Type form) var. typica. (E. collina, R.Br. ; E. tetragona, R.Br. ; E. multicaulis., Benth.). E. collina, R.Br., var. paludosa. ,, ,, var. speciosa. (E. speciosa, R.Br.). Further, the size of the flowers tends to increase in cultivated specimens, and the colour is highly variable. KocHiA viLLOSA, Liudl., 1848. (Salsolaceae). Among some stored specimens at the Herbarium, probably derived from the Souder collection, one was found from A. de Jussieu, dated 1832, E. Nova Hollandia, and named Rhogodia, Billardierii, R.Br., which proves to be the above. Hence this Kochia reached Eurojje long before it wa« described by Lindley, and the present is possibly the oldest Herbarium specimen of the plant. The label and specimen are pasted on the sheet, and hence there is no possibility of accidental transference having occurred. Lyonsia stkamixea, R. Br. =L. stra.minea (R. Br.). Benth. and Mueller. (Apocynaceae). In pursuing some interesting archa?olooical but hardly botanical, studies, Britten^ concludes that the L. reticuhita of F. v. Mueller, is the true L. straminea of R.Br., and proposes a new name (L. Brownii) for the plant, supposed to be BruwTi's L. straminea by Bentham and Mueller. A more confusing and unnecessary addition to synonymy could hardly be proposed, and it is in- teresting to note on p. 240, that Britten sharply criticises Druce for a similar addition to synonymy based on no more certain grounds. Britten admits that " Brov>'n published no de- tailed description of the species," but cons.iders that de Can- dolle's description of L. straminea referred, "ait any rate in 1 Jourii. of Bot;uiy, vol. \1\ ., 1!)07, p. 235. Flora of Amtralia. 133 part," to Mueller's reticulata, and that Bauer's figure was named, "doubtless on Brown's authority," L. straminea. Vague assump- tions of this kind afford no gi'ounds for troublesome changes of long-standing names. Indeed, a work of this chai'acter tends to bring systematic botany into bad odour with workers in other branches, who suffer from such changes, and if there is any diffi- culty in regard to the specimens at the National Museum, Lon- don, sm'ely the proper c(jur^6 is to add explanatorj* labels to them, as in the above heading. Archa^olog}^ and botany are sepa- rate subjects, and should be kept a2>art. Article 50 of the International Rules of Botanical Nomencla- ture, 1905, says: — ''No one is authorised to reject, change or modify a name (or combination of names) becjuuse of the exist- ence of an earlier homonym which is universally regarded as non- valid, or for any other motive either contestable or of little import." Hence the names should remain as before, L. Brownii Britten being a synonym for L. straminea (R.Br.), Bentham and Mueller. Medicago hispida, Gaertn., var. inebmis, Urb. (Papilionaceae). (Syn. MiCDiCAGO RETICULATA, Beuth.). Determined at Kew Herbarium, England. Dimboola Shire, F. M. Reader, October 16th, 1898. Geelong and Penshurst (1906), H. B. Williamson. This Medick was recorded by Mr. Reader in the Victorian Natm-alist, vol. 19 (1903), p. 159, as Medicago turbinata, Willd., but M. turbinata is quite a distinct plant from Reader's speci- mens. It was also known here under the names of M. striata and M. nummularia (M. cretioa), but dift'ers from both of the«e. As no specimens of the above variety were in the National Her- barium, the plants were sent to the Kew Herbarium for verifica- tion, and determined as above. It is a naturalized alien from Southern Europe. Olearia hqmolepis, F. v. M., var. PILOSA, new var. (Compositae). Cowcowing, West Australia. Max Koch, No. 1087 (1904). The variety differs from the type in having slightly longer 134 Alfrfd J. Ewart: jDeduncles, the flowers sometiiiie^^ more than three together, the bracts usually >oinewhat shorter and more pointed. The leaves shorter (about 1cm. long), and the whole plant covered with a more or less well-developed pubescence, the -s^caibrous hairs less developed. From 0. strigosa, Benth., it differs in its twenty or more ray florets, equal 23appus and short nonseptata scabrou'> hairs. It bears some resembla)nce to 0. adenolasia (F. v. M.), but is dis- tinguished by its more numerou-s ray florets, larger heads, more pointed ajnd usually coloured bracts. Phymatocarpus. (Myrtaceae). Tlie leaves of this plant are given as opposite in Bentham's Flora, as in Beaufortia and Regelia. Examination shows that they are all ajlternate in Phymatocarpus, though closely set in P. porphyrocephalus, the bases are all at different levels, and in P. Maxwellii the internodes between the separate leaves are of some length. This gives an easy mode of distinguishing roughly Phymatocarpus from Beaufortia and Regelia. The only exception to the inile of opjD^osite leaves in the last two genera is in Beaufortia squarrosa. This ha« mostly opposite leaves, but in some of the shoots the leaves, though closely set, are alternate, the bases being all at different levels. This is prob- ably an instance of partial reversion to the more primitive type. PouoLKPis Spknckki, A. J. E. (Compositae). This plant bears a close external resemblance to P. aristata, Benth., Fl. Aust., III., 605, from which, however, the blunt outer bracts, the less deeply lobed ra,y florets, and the flowers white or pale instead of yellow distinguish it. Mr. Max Koch, its discoverer, also informs me that P. spenceri is only found in ■damp places near rivei" flats, whereas P. aristata grows in drier situations. PtKKOSTVLIS OONCINN.'V X P. REFLRXA, var. INTEHiMEDIA. A hybrid Orrhid. In all large genera (.Salix, Eucalyptus, Acacia, Rubus, Hierac- iuni) the imperfect segregation of certain species may result in the Flora, of Ausfrdliu. 135 production of hybrids, some of which in tiaie obtain the to the production of hybrids, some of which in time obtain the fixity of species. The same applies to many genera of less ex- tended scope. Tlie present case of the occurrence of a natural hybrid in the genus Pterostylis (Orchidaceae) is, so far as I am aware, only the second instance recorded for that genus in Aus- tralia. The plants were found by Mr. J. R. Tovey at Mentone, \ ic- toria, 1907, growing among patches of Pterostylis conoinna and of P. reflexa, var. intermedia. Externally they resemble the latter 2:)lant, except that the basal rosette of leaves persists in some cases until flowering. The labellum, how^ever, instead of hanging an entire obtusely-pointed tip, is broader and faintly but dis- tinctly bifid at its extremity, in this respect, being exactly intermediate between the two forms. Some specimens show signs of reversion to one or the other parent. Bentham gives the scape of P. concinna as rarely above 1 inch. It is usually 3 to 5 inches long, and may bear 1, 2, or even 3 bracts, the lower ones always empty. The wings of the column are marked in white, green and purple, but the intensity, especially of the latter coloration, varies. The possibility of hybridisation must be borne in mind in future studies of this genus, and this ex- planation may apply to some of the species already described. In Fitzgerald's Australian Orchids mention is made of a supposed hybrid between P. cm'ta, R.Br., and P. pedunculata, R.Br. PuLTENAEA STRiCTA, Sims. In Bot. Mag., 1588 (1813). (Leguminosae). Synonyms: P. maideni, F. M. Reader, in Vict. Nat., xxii., 158 (1905); P. largiflorens, F. v. M., in Benth., Fl. Austr., ii., 13-1 (1864); P. gunnii, Benth., in Ann. der Wien. Mus., ii., 82 (1839). As the result of a close investigation of the numerrius forms of these highly variable and closely related '" species," it can only be concluded that Ave are dealing with forms of one large, extremely pleomorphic species. The original description of Pultenaea stricta in the Botanical Magazine, 1813, page 1588, was made from a plant flowering in England, and naturally refers to that specimen only, Sims being unaware of the varied funns assumed by the species in its native habitat. 136 Alfred J. Ewart: The following description tallies in all esf-entials with the original one, but includes the other species mentioned. P. stricta, Sims. An erect spi'eading or somewhat decumbent shrub of 1 to 3 feet ; the slender young branches minutely hoary or more or lei~s silky-pubescent, sometimes somewhat angular and becoming glabrous when old. Leave* varying greatly in shape and size, sometimes on the same plant, from about 3 to 12 mm long, ovate, oblong, cuneate or linear, obtuse or with a small straight or recurved point, nearly flat, but with the margin usually slightly recurved, shining and glabrous above, paler and hairy or silky pubescent beneath, especially when young. Mid- rib prominent, stipules small, narrow or lanceolate, and ap- pressed, the nari'ower stipules often spreading. Flowers very Schortly stalked, usually in small terminal heads of 2 to 8, but sometimes laterally aji'ranged, and then usually axillary. Bracts imbricate, the outer ones small, the inner ones, when present, larger, 3 or 4 mm. long and either entire, bilobed, or with a hairy point between the two apical lobes, varying in these re- spects in the same head. Bracteolea lanceolate or nearly linear, usually about 3 mm. long, and more or less hairy on the back, inserted on the calyx tube, usually near its base. Calyx about 4: mm., pubescent or silky villous the three lower lobes pointed — ^lanceolate, about as long as the tube, the two upper lobes broader, usually more or less falcate and united to about the middle. Standard twice as long as the calyx, the wings and keel a little shorter than the ^standard, the keel deeply coloured, the ovary villous, the style filiform, but ^^lightly thickened towards the base, where a few scattered hairs may be seen. Pod obliquely or almost triangular, ovate, more or less flattened and hairj% or silky, pubescent, usually 4 to 5 mm., long and projetting beyond the calyx. Variety Maideni (Pultenaea maideni, Reader). The stipules more lanceolate, the inner bracts usually hairy on the back, as well as the edges and tip, and slightly shorter. The " trifid " or bilobed apex of some of the inner bracts is evi- dence of their stipular character, and is not peculiar to this variet)"", which is very close to the type form. Flora oj AuMralia. 137 Variety GuNNii (P. Gunmi, r>entli.). In Ann. der Wien. Mus., ii., 82 (1839). This has naiTOwer, usually spreading stipules, the inner larpje bracts are usually absent, but in all the forms, including the type, the bracts usually fall as flowering advances, and some forms of our Gunnii, have much larger' bracts than others. In the typical forms the leaves are usually broader at the base and taper more or less towards a usually pointed apex. In the typical A. stricta the leaves are usual^^ broader near the apex' then suddenly contracting to a distinct point. The difference is more constant on the larger t^tem leaves. Variety Larc.iflohens (P. laugiflorens), F. v. M. In Fzienth. Fl. Austr., ii., 134 (1864). The flowers may be either axillary or lateral, or in terminal clusters, the bracteoles are usually inserted higher up on the calyx tube, and the two upper calyx teeth are less or not at all falcate, and the fruits usually smaller. Variety Ixcurvata, new var. Locality, Frankston, Coll. J. W. Audas, 1907. This has the leaves with hard, minutely-pointed, recurved tips, giving the plant a peculiar harsh feel when drawn through the fingers. In some respects it is intermediate between the variety Maideni and the type form. Pultenaea retusa, Sm., comes near to some forms of P. stricta, l)ut the calyx teeth are of more equal shape and length, and the calyx is hardly bilabiate. The usually straight upper calyx teeth of variety largiflorens show an approach to this species. It may seem a bold course to reduce these three well-kno-\\Ti species, but the numerous connecting links leave no other course possible, and there is no evidence as to the existence of hybridi- zation between these four species. Variety largifloi'ens, shows the largest, variety Gunnii a lesser, and variety Maideni the least divergence from the type, but the same reasons that could be urged for their maAntenance as distinct species could be used to found at least 12 species out of the numerous connecting forms. It may be taken as a general rule that in all large genera the 4 138 Alfred J. Ewart: term " species " should be given as broad a scope as possible, not merely for reasons of practical utility, though these are of value, but because it is precisely in such genera that groups of varieties as yet imperfectly segregated into species are most likely to occur, and by recognising such varieties as species too hastily we render it more difficult for the workers of subse- quent centuries to obtain evidence of evolution in such cases. In addition, the synonymy is less likely to become so extensive as at present. Thus it is doubtful whether the genus Pultenaea contains many more than 60 valid species, although over 150 have been recorded, and the same proportion holds for most large genera. A specimen of P. Williamsoni, Maiden^ was referred at Kew to a vaa-iety of P. -stricta. The National Herbarium contains both under P. stricta and P. paleacea specimens examined by Bentham or by Mueller, which come very close to, or prac- tically match specimens of P. Williamsoni. Altogether there can be no doubt that the genus will not be on a satisfactory basis until cultural experiments under varying conditions have been performed with all its supposed species, and tte result of such experiments will probably be to give the tfelected species in the genus a much wider range than they have hitherto enjoyed. Sporobolus Benthami, Bailey = S. vihginicus, Kunth, var. PALLIDA. (Gramineae). Queensland Flora, p. 1880, Bull. Dept. Agric, Queensland, xiii., p. 16. The 22 sheets of this variety in the National Herbarium show a far greajter range of variation than Bailey's specimen from the type, and yet have no constant character of more than varietal significance. Although the (inter glumes are u>ually about e<:;[ual, the lower one iis occasionalh' slightly longer than the upper, and sometimes, ospeciallj' on the basal spikelets, not more than half its length. In this respect, in the more hyaline outer glumes, and in the longer spike the variety shows am approach to S. indicus, K.Br., from which, however, the vegetative habit differs. It is, in fact, possible that cultm-al experiments might show S. virginicus to be a marsh and maritime foini of variety of S. indicus, developed in brackish situations. 1 Vict. Nat., vol. xxii,, p. C, 100,=i. Flora of Australia. 139 Bailey admits thart. his S. Benthami, and his var. minor of S. virginicus probably foiiu the var. pallida of S. virginicus, recog- nised by Bentham, and even a cursor}'- examination of the material at the National HerbaJi'ium would "have shown that the new specie.- was untenable. Given as new to New South Wales (L. Cudgellic-o) by Maiden and Betche, Proc. Limi. Soc. N.S.Wales, 1906, Vol. XXXI., p. 739. Tunica prolifera (L.) Scop., var. velutina (T. velutina, Fisch. and Meyer). (Caryophyllaceae). This naturalized alien was recorded by Mueller as T. velutina in Vict. Nat. X., p. 145, 1893, and by Reader as T. polifera, in Vict. Nat., XX., p. 88, 1903. Both species are given as valid in the Ke.v Index, and in Boissier'i^ Flora Orientalis. T. velutina dif- fers from T. prolifera mainly in having leaves with smooth edges (instead of minutely toothed), hairy internodes (instead of glabrous), longer leaf-sheaths; and smaller seeds. None of these features are constant ; haiiy specimens may have rough-edged leaves, and some specimens of T. velutina have the lower leaves minutely toothed entirely or in part. The length of the leaf- «heath may vary on one and the same specimen, as may also the size of the seeds. Hence the species must be reduced to a variety of T. prolifera, joined to that species by intervening forms. Most of the Victorian specimens belong to the variety velutina, but some of Mueller's are intermediate in character. Mount Ararat, Nov., 1883, D. Sullivan ; Upper Murray River, C. French, 188'6 ; Clyde Mts., N.S.W., Oct., 1888, W. Bauerlen ; Delatite. 1890 and 1891, Rev. R. Thorn, Goulbourn River, 1892, W. F. Gates; near Lake Urana, N.S.W., 1894, G. Luehmann, Jnr. ; near Seymour, 1902, Mrs. F. M. Reader. [Proc. Eot. Soc. Victoria, 20 (N.S.), Pt. II., 1907.] Art XII. — On the occurrence of a Marsitpium in an, Echinoid belonging to the Genus ScuteUina. By T. S. hall, I\r.A., Melbourne University. [Eead 14tli November, 1907.] On a visit to the mouth of the Glenelg River, in the west of Victoria, I collected about twenty-five specimens of ScuteUina. They were found in the ^oft white polyzoal limestone which occupies such a large area of the south-east of South Australia aind the south-western borderland of Victoria. The age of this formation is Banvonian, and may be Eocene. On cleaning them with a dental engine a deep depression was displayed in some examples on the actinal surf ace, between the peristome and the anterior maa*gin. The size and shape of the pit varies somewhat in different individuals. It is generally so deep that its upper surface is almost, if not quite, in contact with the abactinal surface of the test. The pit is very shallow near the peristone, but deepens as it runs forward. Its front and lateral walls are vertical. A rounded, but distinct median ridge slightly divides the pit into two halve.-. ScuteUina sp., actinal and abactinal views, profile, and section tliroug-h marsupium and moutli x l^. The only suggestion that I can make as to the function of the pit is that it is a marsupium for the protection of the young. Marsiiinum in Echinoid. 141 The only group of Echinoids in which a definite marsupium has been recorded, as far as I am aware, is that of the Spatan- goids. In them those forms with i-unken petals, such as Hemi- aster and Schizaster, the pits in some cases, and perhaps in all, function as brood pouches. In Hemiaster cavernosus, the pits are present in the female, absent in the male, so that they furnish an external sexual character. Eleven of my specimens have a marsupium, while the re- mainder are without it. Its presence, then, if we may argue on the analogy of Hemia-ster, indicates the female. In some of the Cidaroids a temporary protection is afforded to the young by the tent-like arrangement of the spines, but there is no pitting in the te«t, as in the case of Spatangoids, or as in the present specimens. It is consequently of interest to find the permanent n^arsupium present in a second order of Echinuids, the Clypea^teroida. The question as to the name of the species is not easy to settle. The amount of specific vairiation amongst echinoids is considei'- able, and there is a growing tendency to limit the number of specific forms. F. Jeffrey Bell is one of the most eminent of those who hold this view.-l We have already two species of Scutellina described from our Australian older tertiary — namely, S. patella, Tate^ and S. mor- gani, Cotteau.^ Although there are certain details of Cotteau's species that I cannot decipher in specimens from Mount Gam- bler, the locality of the tjrpe, yet I have no doubt that Tate's and Cotteau's species are identical. Tate in his description gives Mount Gambler as one of the localities from which his species was obtained. The species is widely spread, being found in almost all our tertiary limestones. The question of priority is not easy to settle, for both paper's are dated 1891. Professor Tate, many years ago, when acting as editor of the publications of the South Australian Society, told me that the publications for the year always appeared in that year, so that though his present paper was read only in October, it almost certainly appeared in 1891. Cotteau's paper 1 Marine Investigations in South Africa, vol. iii. 2 Trans. Roy. Soc. S. Australia, 1891, p. 279. 3 Mem. Soc. Zool. d'i France, pt. iv. (1891), pp. 629, 630, pi. 19, figs. 10-14. 142 T. S. Hall : Marsiqnum in Echinoid. appeared in part 4, the finajl part, of the volume- for 1891. Both species are recorded on the same page of the Zoological Record. The fact that S. patella has been familiar to Australian geolo- gists a« a manuscript name of Tate's is no argument for its uae, but till the question of priority is settled I shall use Tate's name. The present specimens, with the marsupiura, are not, I think, separable, though in most of them the pentagonal outline is decided, and I think they may be regarded as S. patella. I have found one or two specimens from Mount Gambler also showing the marsupium. [Pboc. Roy. Soc. Victoria, 20 (N.S.), Pt. II., 19o7.] Art. XI ir — Tlic Culeopteva of Kiixj Island, Baid. SlLPilSDAK. 55. Ptomaf^hiia /ni://ry//it>sii. Sell. TiuciioPTKin ciiiAi;. 56. Aitiiiop/ervx aiishalis, .Ma.ttli. NlTJDULlUAK. 57. Brachypef^lns /'as a /is, Er. 58. Haptoficitrd meyricki, Blackb. 59. Cryptarcha elegantior, Blackb. Th or; OS IT I DAK. 60. Leperiua decurata^ Va\ C'OLYDIIDAK. 61. Penihelispa /iilioi/iosa, Ei-. 62. P. secii/a, Pasc. Cue LM I DAK. 63. Prostoiiiis (i/kins(>fn\ ^^'atll. 64. P. cor/ii//i{s, W'alli. 65. Hyliota australis, Va\ 66. Cryptaiitorp/ia ollilPu Blackb. 67. Myraboliix groirvi'/Iiana, \\i\\ 68. M. loHgtcorfiis, IJlackli. CuYPTOKlIAfJIDAK. 69. Cryptop/idj^its ias/iiniiicus, Blackb. Latiuudiidak. 70. Lathriauis apicalis, JJlackb. 71. Z. nigroinacitlaliis, lilackb. ('oh'opfrrii of King Ishuul. 155 7'i. Corticaria adelaidae^ lihicklj. 73. C tiiisfm/is, Blackl). 1 )i:i>oo(jf/!\ lilacklx 7'k r. /imrio^ l^r. 77. 7! ^AV'^''? '^^'■• 7S. Der//iesk'S cddave} iiii/s,^ Fab. I^YUKHID.AK. 79. Miciochaetes scoparius, Ei-. 1 Trans. Hoy. Soc. South Austialia, 1891, p. 129. 156 Arthur M. Lea : Parnidak. 80. Eiviis fasmanicna, Blackb. LUCANIDAE. 81. Syndesiis Cflrmttus, Fabr. 82. Ceratognathus niger, Westw. 83. Lissotes cancroides, Westw. 84. Mastochilns po/itus, Burm. ScARABAEIDAIi;. 85. Onthophngiis mistralis, Cluer. 86. O. }/ii{tattis, Har. 87. O. posticus, Er. 88. O. promts, Er. 89. Aphodiiis granarius, Linn. 90. Saprosites niendax, Blackb. 91. Diphucep/ia/a pulchella, Wath. 92. D. colaspidoides, Gryll. 93. Scitala languida, Er. 94. Heteronvx obesiis, Bui-m. 95. H. striatipemiis, Blanch. 96. H. tempestiviis, Er. 97. Atttomoliis bicolor, Blackb. 98. Adoryphorns coiiloui, Burm. 99. Pi me I opus porcel/us, Er. 100. Cheiropiaiys vioelius, Er. Buprp:stidae. 101. Stignwdera flavopicta, Saiind. 1 02. Meiobasis fiilgurans. Thorns. 103. M. hypocrita, Er. 104. M. prisca, Er. Elateridae. 105. Monocrepidius fabrilis, Er. 106. Elaler gramilatipennis, n. sp. Colcopb't'd ()/ Kin;/ /sldiul. 157 Black or blackish ; antennae (ba.sal joint sometimes infuscate), palpi and legs (femora more iir less infuscate) reddish. Rathc'r densely clothed with fine whitish pubescence. Head convex ; denseh' and rather coarsely punctate. An- tennae extending to m&tasternuui. Pruthni-ax as long as wide, but apparently slightly longer than wide, strongh^ convex, sides rounded in front, basal two-thirds subparallel. hind angles moderately produced, mcdinn lino almost absent, with a wide shallow basal impression on each side; punctures as dense as on head, but rather shallnwer and smaller. Scutelhnn gi'anulato. Elytra (by measurement) about twice and one-half the length 01 prothorax. gently decreasing in width from near base to apex, apex obtusely pointed : with naarow, apparently ini- punctate, striae ; interstices with small dense rounded granules. Under surfaice with dense punctures, becoming granules on apical segment, and subgranulate on basal segments of abdomen. Length, 8 — 9i mm. A beach frequenting species \ also occurs near Sydney. In general appearance somewhait resembling Acroniopus rugosus, Cand. but with tarsi (except that they are longer) as in Elater perplexus, Cand. On two of the five specimens before me the elytra are piceous brown instead of black. 107. MelanoxiDitlins (juadriguttatus, Er. 1 OS. Cardiophori/s huniihs, (^and. 109. Ccry/ii/'itt's siiavis, Caiid. 110. Hapa/t'Siis hirtiis, Cand. The specimens from the island seem to represent a variety of this species, as they differ from typical ones in being smaller, with the clothing denser and Ionizer, and the punctures in the elytral striae more pronounced ; they also have the elytra more convex, and the median line of the prothorax more noticeable. I should probably have rejjfarded them as belonging to a distinct species, Vuit that a specimen before me has these differences even more pronounced, and was returned to nie by Monsieur Candeze as var. minor of hirtus. One of the speci- mens wa.s taken under Ijark, but seven others were taken at 158 Arthur M. Lea : the roots of beach gi-owing plants, and on which their larvae proliably feed. 111. Crepidfl/iit'ii/(s aberratis^ n. sp. J Piceous-red, auti^iinae, scutelhini, prosternuni and .sides of meso- and of metasternuni black or blackish ; legs obscurely variegated. Rather densely clothed with short, silvery pubes- cence; on the upper surface variegated with irregular spots of rusty or golden pubescence. Head densely punctate, with a wide, feeble depression be- tween eyes. Antennae extending to hind coxae. Prothorax ap- parently twice as long as wide, but by actual measurement not once and one-half as long as wide, sides subparaillel to near base, hind angles acute and embracing shoulders ; median line rather deep and wide in middle, becoming obsolete towards apex and subobsolete towards base ; punctures rather smaller and not quite so dense as on head. Srutelluw subcordate. Elytra (by measurement) not thrice the length of prothorax, each semicircularly notched at inner apex ; striate-punctate, punctures in striae small, but deep, interstices with moderately dense minute punctures. Under surface rather sparsely punctate along middle, but densely at sides ; base and apex of prosternum with coarse punctures. Tarsi thin, fourth joint nan-ower than third. Length, 14 — 18 mm. ? Difler.s in being niiu-li \vi(h^i-, liotli jH'othorax and elytra less parallel-sided, antennae not passing hind angles of prothorax, and legs shorter. Also from Tasmania (Frankford). The long prothorax of the male and the narrow tarsi are at variance with others of the genus, amd in fact at a glance the species looks like a Chrosis. The only female before me is somewhat abraded, but all of its clothing appears to be more golden than silvery; whilst on the upper surface it is decidedly golden, with a feeble mottling of sooty. 112. C. aiisffd/is, l)<)i. 113. C. dao/af//s, Er. lU. C. fu/i^idits. Va: Ciiit'Dph I'd of Kniij Island. 159 Dascillidak. 1 I ■'>. MacroJielodes i/i^i^cr, n. sp. Deep black : parts nf inoiith apjiendaiies a.nd of Siterna flavnus ; second and tliird joints of antennae, knees and parts of tarsi obscurely diluted with red. Upper surface oflabrous, lower with fine pubescence, except in middle nf metaslernuni. Head with dense and fine punctures. Second and third joints of antennae combined shoi-ter than fourth. Frothora.r with sparse and vei'v small punetures. becomina- denser and larger at sides, but even there smaller and sparser than on head. Elytra with dense a.nd not very fine punctures, smaller along suture than elsewhere. Len^ith. Sj, mm. Differs from the descriptions of princeps and lucidus in its entirely black upper surface (ineludinji- the sides), and almost entirely black antennae and legs ; jn-inceps is also said to have the elytral punctures " sparsim." those on the elytra of lucidus are not mentioned, Init the s})eeie.s is said to have " cetera ut M. princeps." On the present species the punctures are denser than in tasmanieus, but siimewhat smaller ; and they are denser than in crassus. On the type b^th antennae have the three termin.il joints missing. IK). Helodes vic/ofiiie, l>laekl>. 117. Cvplion oveiisciisis, l>liiekl>. 1 18. C. pit Ins, Hlaekb. 1 lit. C. spilotiis. Hlaekb. Malacodkiohdak. 120. TricJialus kershaivi, n. sp. J Black : suture and margins of elytra reddish. Antennae senate, extendiiii; to middle of elytra. Prothorax moderately transverse, hind angles acutely produced ; with fairly numerous and rather large punctures in front, and a row of somewhat larger inineturos behind. Sciifcll inii concave. apex gently arcuate. Elytra parallel-sided to near ajjex, with double rows of lai-ge transverse punctures ; each elytron with three strong costae, except near base, where there are four. 160 Arfhnr M. Lea : Penultimate sejj:ment of ahditmeii feobly mitched. Length, 10—13 mm. $ Dirters iu being more robust, with shorter and less strongly serrated antennae and simple abdomen. The antennae of both sexes are much as in ampliatus ; the entirely black prothorax will readily distinguish it from insignis, which otherwise it strongly resembles. 121. Me(rior?-]iyn, lilts kiiii^eiisis, u. sp. ? Black, shoulders very feebly diluted with red. Eosfrunt very short. Antennae strongly serrated, scarcely extending to basal third of elytra. Protliorac triareolate, middle areolet narrowly open in front, rather more widely open behind, middle of apex deeply notched. Scutellum concave, apex strongly notched. Elytra wide, subparallel to near apex ; each with four fairly strong costae, and with double rows of large sub- quadrate punctures. Length, 12 J mm. The combination of triareolate prothorax, very short mstrum and double rows of elytral punctures will readily distinguish from all other black species hitherto describe-d. The antennae are much as in the male of atratus. 122. M. ohscuripeiinis^ Lea (m.s.). 123. J/, riifipeniiis, Fab. 1 24. Teiephorus iiflbilitatns^ Er. 125. 7: pii/che/Iiis, W. .S. Macl. I 26. Heteri'iiiasdx apicifliwtis, n. sp. Black, middle of prothorax, tips of elytra, apical half ecies with the ]iri thorax black, have the head very differently sculptured. Clkkidak. 1:^0. Opilo sexNoiin/s, Westw. Apteropilo, n. g. Prothi rax without longitudinal aaid ti-aiisverse impressions. I'^lvlra ol)iivate. Metasternum short, .\ptcrous. Other charac- ters mostly as in Opilo. 1 AN(i :i\vait'ni;;' piililicatiiin in iii\ ri'\ii'\v of .MaUu-odcniiiilac. ('(>h'(>i>hiii i>f Kniij Isliiiid. lOS In Imth JMac-klnirirs ami Gurham's tables of C'hridae this lan(l). 164 Arthur M. Lea : 134. Paratilliis cams, Newni. 135. Leniidia cicatricosa, Lea (m.s.) 136. Z. nii^rovaria, Lea (m.s.). 137. L. sii/isofii, Lea (m.s.). 138. Z. niieiis, Newm. CioiDAi:. 139. Cis kani/s, Black b. BOSTRYCHIDAE. 140. Xylobosca iu'spinosa, Mac). TEMiBHIOMDAK. 141. Caediflinoi pha Iietcroinera, Kin^. 142. I^rionotiis senicollis, Hope. 143. Hyocis cancell'ita, ii. sp. Black ; muzzle, front marmaller iind differently coloured, the ]irothorax with more evenly rounded sides, smoother surface and narrower hase ; the punctures are also everywhere smaller, and nn the elytra more decidedly Coleopteni <>J K'nuj liland. 167 seriate in a.rran!/ Ixlond. 169 iniddU\ shoulders feebly intlated : with dense and line inincture'.s, and each with traces of three very feebly raised lines. Lfffn Innir. tibial spurs short but distinct. Lenofth, 7^ — 9 mm. There are usually four black spots on tlie prnthorax — a fadrly large one on each side near the middle of the base (but not on the extreme base), and a much smaller one on each side about one-third from apex ; these latter are often reduced in size and occasionally are absent ; on an occasional specimen there are also two or three more small spots. Numerous specimens were taken cl )se to the sea beach. In Blackburn's table of the Australian Oedemeridae this species would be placed in his typical section of the genus Copidita. The claws are slightly swollen at the base as in Kershuwi. The eyes are not so coarsely faceted as in punctum, still the facets are much larger than in Ischnomera sul)lineata.^ 1 70. Pseiidolychus liaeniort hoidalis. Fab. Twelve sjiecimens from the island are before me, three have the t}ijica)l red tip of the elytra, two have the red tip almost absent, whilst the others have the elytra entirely dark. I ha.ve seen no similar specimens as the latter from Tasmania or Aus- tralia. 171. P. margiuatus, Guer. CUHCULIONIDAE. 17l'. Prosay/eus hopei, Sch. 17-"l Rhadinosonnis lacordairei\ Pasc. 174. Pimareta subtcrtanea, n. sp. Dark reddish brown, appendages paler. Densely clothed with white scales, usually more or less feebly mottled with brown ; with dense, tine, white setae. Eyta proujinent, coarsely faceted, and rather small. Scrobes distinct from above. Antennae extending to base of prothorax, scape about the length of funicle and club combined, first joint of funicle slightlv longer than second. Prothorax moderately 1 There is coiisidenililc difference in the size of the facets of snblineata and atkiiisoni, aw! accordiiis; to the talile tliese would cause the species to be ijenerically separated. 170 Arthur M. Lea : transverse, side^i reirularly munded, median line feeble ; with dense, ra.ther small punctures ; and small, irregular flattened grajiules. Elytra ovate, conjointly arcuate at base ; striate — punctate, punctures rather large, becoming smaller posteriorly ; interstices gentlj'' convex, regular and distinctly wider than striae. Under surface with dense, rather small and partially concealed punctures. Abdomen with basal segment slightly con- cave in male, slightly convex in female. Femora stout ; tibiae suddenly inflated at apex ; claw joint long. Length, 4 — 5 mm. The sculpture is described from abraded specimens, as the clothing is so dense as to entirely conceal the derm of the prothorax, and to cause the elytra to appear feebly striate — punctate, or even fee'bly striated only. The scales are sometimes entirely white, but they are usually mottled with very feeble brown or smoky spots on the elytra, and on the prothorax with feeble stripes. From some directions the first joint of funicle appears to be slightly shorter than the second. The gra.nules of the prothorax are variable, as on complete abrasion of two specimens they are seen to be fairly dense and regular on one specimen, and entirely absent from some parts of the other ; on another specimen they can just be traced, but the punctures are always distinct though small. The males are usually smaller than the females, and are slightly narrower, but the sexual differences are not very pronounced. In appearance it is close to some of the varities of crinita, but is riither mure robust (the male is fully as wide as the female of that species), the setae on the prothorax and elytra decidedly finer and more numerous, and the abdominal punctures smaller. Numerous specimens were obtained amongst the roots of beach-gi'owing plants. 175. Maiidalfltiis caviventris, n. sp. Black ; antennae, tai-si, knees and parts of tibiae more or less reddish. Densely clothed with greyish — white scales, occasion- ally fee})ly spotted with pale brown : and with fairly dense thin setaiS. Head with small partially ciUK'ealed granules between eyes; base finely corrugated. Rostrum with granules as on head ; ('olp(>/)teni, of KiiKj /shiinl 171 with a thin and (.'ontinuous median carina. Scape the length of funicle and clul) cunibined : first joint of funicle once and one- luvlf the lenu'th of second. Prothorar, about once and one-third as wide as long ; with dense and more or less flattened granules. Elytra not much wider than prothorax. parallel-sided to near the middle, thence regularly decreasing in width to apex ; striate — punctate, punctures partially concealed, interstices wide, with numerous suuull seta-bearing granules. Front coxae widely separated. Intercoxal process of niesosternum simple. Metas- ternum transversely corrugated. Abdomen indistinctly wrinkled ; with dense, minute and subobsolete granules. Femora stout, tibiae bisinuate beneath. Length, 5J — 8 mm. The male differs from the female in being smaller and nar- rower, Avith thicker antennae and femora, and with a large ex- cavation common to the two basal segments of abdomen ; these being gently convex in female. The claw joint from its base is as long as the three basal joints coml)ined. Each seta, except some on the appendages, arises frum a araaiule. One specimen has the legs entirely of a dull red. In general appearance much like many species of Polyphrades, but the tarsi are not soldered together at the base. Tlie setae and granules of the prothorax are much as in seticollis, and the abdomen and legs, etc., are nnich the same: but the pro- thorax, although without scales in the middle, is densely clothed on the sides; and the elytral granules, although small, are (juite conspicuous. 176. M. ani/ents, J^ea. 177. M. c nidus, E rid is. \~i'!^.. M. vodralis, Jjlackb. 179. Leptops tribulus, l^'abr. 180. Ferpenis costirosiris, n. sp. P,hu-k, antennae tarsi and ocular h)bes obscurely diluted with red. Densely clothed with small white scales, and with numer- ous more or less decumbent whitish setae. Head with small dense punctures and with a few larger (but still small) ones scattered about. Rostrum with a narrow acute 17^ Arthur M. Lea : costa, commencing between the eyes and terminating- at the apex in the form of a narrow triangle, apical half of sides flattened, glabrous and with sparser punctures than elsewhere. Scrobes deep and curved aibout antennae, but disappearing half-way be- tween them and eyes. With feeble sublateral sulci. Antennae short ; first joint of funicle distinctly longer than second, second longer than third, fourth-sixth sub-globular, seventh feebly trans- verse. Prothorax transverse, convex, sides evenly rounded ex- cept at almost extreme base and apex, usually with a feeble median impunctate line ; punctures as on head. Scutellum small but distinct. Elytra elongate — subcordate, conjointly arcuate at base, with rows of fairly large but usually concealed jDunctures ; interstices gently convex, the alternate ones very feebly raised, with dense and very small punctures. Under sur- face with small and dense punctures. Legs rather long ; front tibiae denticulate below. Length (excluding rostrum), 8^ — 10|^ mm. The male differs from the female in being smaller, with narrower and more parallel-sided elytra and longer legs. The acutely carinated rostrum and first joint of funicle de- cidedly longer than the second readily distinguish from most previously described species of Perperus ; the sides of the ros- trum in front are reminiscent of Rhinaria. In some specimens (usually females) the derm is entirely of a dark reddish brown. The scales are so readily abraded that the disc of the prothorax usually appears to be> glabrous, and on the elytra large irregular patches are frequently denuded ; on the elytra the scales fre- quently have a golden gloss ; on them also they aa-e everywhere dense, but they are rather denser on the odd than the even interstices. On the upper surface the scales are more numerous than the setae, but the reverse is the case on the under surface and legs. Tho hind femora are usiuilly feebly ringed, and traces of still more feeble rings cam sometimes be seen on the others. 181. Perperus coiiloni, n. sp. Black, appendages and ocular lobes more or less i-ed. Densely clothed with small, rounded scales, vaiying on individuals from fawn-coloured to muddy brown, and occasionally with a faint Colcopfera of h'tn;/ IkJoli. 187. O. /epidoiits, \^v. 188. Rhaciodes bicaudatus, Boi. 189. R. orauuUfer, Chev. 190. Eristiis pallidiis, d. sp. R(;d(lisli-fl;ivou!i, metasteriiuin wwi^. liasal seginent of al/do- luen sometimes somewhat darker. Clothed with fairly stout, whitish pubescence, denser at base of prothorax and sidc8 of metasternum than elsewhere, on the elytra more or less seriate in arrangement. Head with numerniis punctures on lower portion of forehead. Rostrum wide, flattened, feebly curved ; in male about once and one-half as lonj^ as wide, in female al)out once and two-thirds ; with numerous more or less concealed punctures on basal por- tion but sparse elsewhere. Profhnra r moderately transverse, apox narrower than base, sides rather stronuly rounded ; with numerous punctures, but which are concealed towards base and sides. Elytra suboblontf, considerably wider than prothorax, parallel-sided to near apex : with series of large punctures in rather feeble striae ; interstices feebly convex, each with a row of small punctures. Under surface with rather small but dis- tinct punctures. Abdomen with third and fourth segments feeibly curved throughout. Legs ratlier stout. Length (excluding rostrum), 2 mm. Smaller and veiy differently coloured to the two species (setosus and bicolor) hitherto descriljed ; but there are several closely allied undescribed species. 191. Cxttalia sydiicve/isis. lUackb. 192. Misaphrice a/'/oiii^^a, llbu-kl). 193. Enidpea siibciwriilca, \\. sp. Black ; rostrum and appendages (parts of tarsi infuscate) red- dish ; elytra usually reddish, but frequently the sides and the suture near base stained with black ; prothorax also often reddish, Moderately densely clothed with short stout pubescence (scarcely scales, except on the imder surface), varying from white (usually with a bluish or greenish tinge) to brown. Coleopterti of Kin;/ Is/ihh/. 177 Ji(tic)ii of the AustnilUui Crvptorliv iirlii(lf>. Coleoptera of King Island. 183 Head at extreme base with densf and not concealed ininc- tures, these concealed elsewhere. Knstruni ra.ther stout, shorter than prothorax, sides feebly incurved to middle : with dense punctures, concealed on Ijasal third in female, on l)asal two- thirds in male. Antennae moderately stout; scape inserted two- fifths fritm apex of rostrum, the lenut once and one-half the length of greaitest depth ; Avith rows of large, round but partially concealed (less so on sides than on disc) punctures or foveae, somewhat interrupted by interstices; these usually narrowei than iniuetures; l)ut subtuberculate in places ; a small shining granule on each side of suture a.t b;i5e ; apex trisinuate. Punctures of undei- xurfare entirely concealed, but second segment of abdomen shallnwly transversely im- pressed. Legs rather short and stout, fourth joint distinctly longer than first, claAvs feebly separated. Length, 5 — 6J/ mm. The male has the rostrum shorter and stouter than in the female, clothed to a greater exteht, and with the antennae inserted rather nearer the apex. On the elytra the posterior declivity has the scales more grey than brown, and at the liasal tliird tliere are also some obscure greyish spots; there is usually an obsciu'e pale strijie along the middle of the prothorax and a similar one on the abdomen. There are also obscure greyish rings on the legs. In addition to the ordinaiy scales there are some stouter setose ones, rather more numerous on the abdomen and legs than else- where, l)ut causing a fasciculate appearance on the prothoracic aiul elytra! tubercular elevations. The elevaitions on the ])ro thorax are very obtuse, and appear to be placed in two or three feeible transverse series, but the individual tuljercles themselves are often obliquely placed. On each elytron there is a larger (but still obtuse) tubercle than elsewhere on the third inter- stice, amd a somewhat smaller one on the fifth, forming (on both elvtva) a transverse series of four at the summit of the 184 Arthur M. Lea: posterior declivity ; this is rather abrupt and thickly studded with small tubercles, the larj^est of which are almost apical ; there are other obtuse tubercles on the third, lifth and seventh interstices. Most of the specimens before me are encrusted with mud. '204. roroptenis lubetcr, Erichs. I Acalles rjibetra, Erichs). Referred by Erichson to Acalles,^ but belongs to the group of Poropterus represented by such species as exitiosus aaid bisignatus ; although in its deeply sulcate basal segments of abdomen 2 it is unique in the genus. There is usually a small shining tubercle on each side of the scutellar region, and the elytra when abraded aippear to be vermiculate-tuberculate. The derm obliquely behind the shoulders is occasionally diluted with red. The two spots on each side of the head and the four luteous spots placed transversely on the prothorax are usually indistinct ; and the median line is so faint as to be practically invisible. The apex of the prothorax appears to be feebly bitid, but this is due almost solely to the clothing. The male has the rostrum stouter than in the female, with denser and coarser punctures, and hais scales almost to the antennae instead of at the base only. Specimens are to be taken under logs, or crawling over them at night time. I have specimens from Frankford, Ulverstone, Wilmot and Stanley in Tasmania, as well as from King Island, and have seen the type. 205. P. conifer, Uoli. 206. F. succisus, Er. 207. Microporopterus funii/Iosus, Pasc. Roptoperus, n. g. flead moderately large, not concealed. Eye,'i ovate, widely separated. coarsely faceted. Rostrum raither short and wide, feebly curved. Antennae moderately stout ; scape inserted closer to base than apex of rostrum and much shorter than funicle ; two basal joints of funicle elongate ; club ovate, much wider than funicle. Prothorax slightly longer than wide, or 1. I'iiMOc tlioiiKlit it lieloiifjed to I'aleticn.s. ■1. A clmractiTOVLTlooktMlby Kriclisoii, l>nt (■onniuiifr.l ui>(.n l>.v lilackl.tirii. Coleoptera of King Island. 185 slii.'htly \\ ider than lon<^, base bisinuate, constriction feeble, ocular lolies subobtuse. Srutellam not trajceable. Elytra elonirate-ovate', considerably wider than and about twice the length of pruthorax. Pectoial canal deep and wide, terminated between front ]i:irt of middle coxae. Mesosttrnal receptacle. feebly raised in fmnt, about once and one-half as wide as long, emari^nation semicircular ; cavernous. M etasternum moderately lontr but uiuch shorter than the following segment; episterna narrow, but distinct thi'oughout. Abdomen large, sutures dis- tinct and deep except that between first and second segments ; first as long as second and third combined, intercoxal process wide ; third and fourth naiTow, but with deep' and wide sutures, the distance between second and fifth equal in length to that of either. Ler/s of moderate length ; femorai stout, not grooved, edentate, posterior terminated before apex of abdomen ; tibiae feebly compi-^ssed and feehly bisinuate beneath, in addition to the terminal hook with a small subapical tooth ; tai'si shining, thin but not veiy long, third joint feebly bilobed and very little wider than second, fourth elongate. Elliptic, modei'ately con- vex, squamose. fasciculate, apterous. This genus appears to be intermediaite in position between the Chaetectetorus and Poropterus groups, but it may be placed with the latter on account of the head being depressed at the base in all the species, -l- and on account of the narrow glabrous tarsi — so suggestive of affinity with Methidrysis. The suture between the first and second abdominal segments is deep and distinct at the sides, but (unless the clothing be removed) not traceable across the middle. "208. Roptoperus tas/iianiensis, n. sp. Dark brown, antennae and tarsi of a rather pale red. Very densely clothed with rather ding}' fawn coloured scales ; with stout-er scales rather thickly scattered about and forming ten fascicles on the prothorax and about twenty on the elytra ; femora and tibiae with indistinct pale rings and with rather numerous elongate scales. Head slightly convex, base depressed ; puncttires concealed. Rostrum the length of prothorax, slightly longer in female than 1. Two others are known to nie in ;ii.: Cryptor]iy)icltiis si/idi/s, i*]r.' ; Hexymus suhplatiatiis, Lea.) Dr. Bolsduval's description is quite worthless for the identifica- tion of this species, but I have examined his tvjie (now in the lirussels Museum), and it is certainly a IJexyiiiiis, and the species described by Erichson a.s (' ri/ pforJn/urlntx xolidii.^ ~ and by my- self as^ Hexymus suhjjlanatus. 1 Wiegni. .\rcli., 1S4-2, p. 205 (omitted from Master's Catalogue). 2 I have exauiiiiefl a speciiiieii from the Herlin Museum marked " Civptorhynehus NoHUus, Er.; Type 3.')937." It is, ho\ve\ er, prohahly the specimen of which Krichsoii said " Variat foi-jiore toto fusco-squainoso." But, except for the colours of its scales and that the rostrum is almost entirely hlack, it ajjrees with his description. 'A From a t^reatly abraded specimen. Coleoptera of Kim/ Island. 187 The species is variable in the cnlmir of its eIothiii>:, and also of its rostrinn. Erichson described the rostrum as riifa. Imt in most specimens it is reddish at the tip only. On the prothorax there are usually ei^ht fascicles placed in two trans- verse series, bur they aa-e not always clearly defined, and oi"ten appear as if but four in number. When perfectly fresh the prothoracic carina is usually covered with scales, althoufrh always distinctly traceable. On the elytra there are usually four {but sometimes only two or three) sliininu' aranules on eadi side of tlie suture about the middle. I have si>ecimens from New South Wales (Nepean Kiver and Burrawanir) and Tasmania, as well as from Kinjr Island. 210. Decilans major, n. sp. Black : antennae and tarsi reddish. Densely cluthed with soft, pale brown scales : im tlie elytra variepr,ts of paler and darker scales. Head with sculpture entirely concealed. Rustrum with dense l>unctures. Antennae in.serted about one-third from apex of rostrum in male, two-lifths in female; scape the length of five basal joints of funicle : of these the first is as long as the third and fourth combined and slightly longer than the second. Pi-ofliora.r about once and one-half ais wide as long, sides strongly diminishing in width from near base to apex ; with dense, fairly large, round punctures, uniform in size except at apex. Ehjtra with outline almost continu(uis with that of pro- thorax ; with roAvs of large somewhat rounded, but almost en- tirely concealed punctures; each interstice with m row of iMuud. shining and very conspicuous granules. Ahdomeii with dense and fairly large punctures on two basal segments ; the ^econd not much shorter than first alonir the middlo. Length. 7 !) mm. The scales on the pntlliorux aro stout, each is set in a puncture and rises above ilir derm : --o^ mm. On the prothorax the scales are stout and each arises from a puncture. On the elytra the scales are smaller and uniform in size, and mostly sooty brown, but with numerous irregulai'ly defined spots or patches of pale brown or ochreous, and with snow^r white scales scattei'ed singly or in small spots, causing a speckled appearance. On the prothorax the scales also vary in colour, but they are not condensed into spots. On the under surface the white scales are absent, but there are a few on the legs. Where the clothing has been abraded minute granules ciui sometimes be found on the elytra, but they are quite concealed by the clothing ; the derm, both tlicre and on the prothorax, appears to be very finely wrinkled. In general appearance somewhat close to apicatus, but the scales much smaller and the punctures totally different. Ovatus has much denser clothing, and its sculpture is very different. Coryssopus is more densely and differently clothed, and lias armed femora ; from squamipennis it differs in being larger, punctures of prothoraix more concealed by the scales (which are individually larger) and by its unarmed femora. From all tho other described species it is very distinct. 213. Decilaus mollis, n. sp. Black or l)lackish ])rown, elytra reddish brown, rostrum antennae and tarsi paler. Densely clothed with large soit scales ; interspersed with numerous stout suberect setae. 100 Arthur M. Lea: Eodruin wide, feebly curved, shining ; with nuiuerousi small punctures. Scape stout, inserted almost in exact middle of side of rostrum, much shorter than funiole. Prothorax not much wider than long-, sides strongly rounded, apex less than half the Avidth of base ; with dense, lartre, round, concealed punctures. Elytra, subcordate. base ahuost truncate, rather strongly in- flated near base and then strongly diminishing in width to near aipex ; with rows of large, rormd. concealed punctures ; inter- stices rather strongly and almost equally convex. Abdomen with large, partially concealed punctures. Length, 1| - 2J mm. The clothing is so dense that the derm is almost everywhere concealed, and the elytra appear to be finely striated only. The scales, however, are atbsent from all but the base of the rostrum. The scales on the prothorax and abdomen are larger than else- where. Ijut on the prothorax they are wider and more closely applied to the derm than on the a.bdomen. Most of the scales are of a pale uiuddy grey, but on each elytron there is usually an irregular triangle of black scales, the base of which is on the side, and the apex nearly touching the suttu-e about its middle ; but the triangle is sometimes broken up into small and irregular spots, or appears as an irregular fascia. There are usually some snowy white scales on the elytra. The legs are usually feebly annulated. On the elytra the darker setae usually form two loose fascicles on the third interstice — one near the base, the other median. I have a paiir taken //; cop., but cannot detect any sexual differences, apart from a thickening of the male femora. Nearer noctivagus than any other described species, but smaller, with more variegated clothing, and which on the under surface is sparser ; the scape shorter, stouter and more median, and the mesosternal receptacle less raised and thinner. '214:. Decilaus auriconius, Lea., \ar. ijisitlaris, n. var. A single specimen frcm ilio island evidently represents a variety of this species ; it differs from the types in having Ihe Vjody (l)ut not the apptMuhigcs) entirely black ; the clothing is more variogated, and on the elytra the scales are distinctly less ri'unded; this lartter character would probably have been Colcojilcra of K'ukj Island. 191 reirarded as of specific importance, but that the clothing of the abdomen is of the saane remarkable nature as in the types. 215. Decilaus acerosiis, .Er. Referred by Erichson to Acalles, but belongs to this genus. It is a commut derm usually concealed. Upper surface with dense scales, varying from dingy white to sooty-black ; scutellum with white scales ; under sur- face and legs with sparser scales than on upper surface, ^.he scales mostly white; basal third of rostrum squainose. Pro- thorax with eight fascicles : two at apex, two at base, and four across middle, the two apical and two mediolateral usually cnm- posed of reddish-brown scales, the others of blackish scales ; each elytron with about six fascicles, and with sca.ttered erect scales. Rostruiii feebly curved, slightly increasing in width to apex, apical two-thirds finely punctate. Scape stout, inserted nearer base than apex of rostrum, half the length of funicle and club combined. Prothorar gently convex; punctures entirely con- cealed ; apex more than half the width of l)ase. Elytra about once and one-third the width of prothorax, shoulders strongly rounded ; striate punctate, striae distinct, hut punctures con- cealed. Under surface Avith moderately dense and strong but partially concealed punctures. Leys rathei- long ; femora edentaite. Length, 1 5-6 — 2 mm. Coleopteiu of Kukj IsIk ml. \\yi Also from Tasmania (Holjaa't, Bruni Island and Huon River;. The fascicles of the prothorax are sometimes very ill-defined ; on the elytra there is usually a more or less distinct parch nf reddish scales on the suture, behind the scutellum. The species is the smallest of the Chaetectetorus ofroup know-n to me. 218. Alenios sordida/NS. ii. sp. Red. but colour (except of rnstruiu iuid antennaej concc-aled ; rostrum shininpr towards apex. Densely clothed with s.ift slaty- brown scales ; under surface and fenmra with din^ry whitish scales. Prothorax with six fascicles : two at apex and fnur across middle ; suture, third and fifth interstices witii rather numerous small fascicles. Head depressed between eyes. Rostrum straight, sides feebly incurved to middle : apical half feebly punctured. Scape in- serted almost in exact middle of side of rostrum. I'mtlxirdv moderately transverse, apex much narniWL'r than base, sides rounded and increasinp: in ^^'idtk to bivse. base bisinuate ; with dense but concealed punctures. Elytra closely applied to pro- thorax and very little wider, base trisinuate ; striate-punctate, striae distinct, but punctures almost concealed, third and filth interstices feebly elevated towards base : preapical callus scarcely traceable. Under surface with dense concealed punctures. Femora moderately strongly and equally dentate, the front pair from some directions apparently edentate, l^enirth, 4 — 4i mm. Also from W. Australia (Albany) and New South Wales (Sydney). On one of the specimens there are a few nbsciu-e whitish spots on the elytra. ■J II'. PhiaoglyiiiDia mixta. \\. sp. Dark reddish-brown, in plact^s beconiinti' black; antennae (club excepted) and claws reddish. Densely cinthed with scales vary- intr fr^im white to black, and furming feelile fascicles in places. IJtad with dense concealed punctures, liostrum rather wide and lightly curved, slightly shorter than prothorax ; with dense punctures, concealed rm basal third in male, un basal fourth in female. Antennae inserted nearer ba.se than apex of rostrum, scape alxjut half the length of funicle and club combined ; two 194 Arthur M. Lea: liasal joints of funiele the length of four followino- eouibined, third to seventh transverse. Prothorarc about once and one- third as wide as lone, apex much narrower than base; with dense and fairly larpe, but quite concealed punctures. Scutellum small but distinct. Elytra elnngate-subcordate, shoulders feebly produced ; with rows of large, more or less concealed punctures, in feeble striae; interstices with dense, concealed inmctures. and subtuberculate beneaith fascicles. Under surface with dense more or less concealed punctures. Femora acutely dentate ; tibiae angular at external base. Length oh-^%\ mm. The clothing is so dense as to entirely conceal the derm. On the head and base of rostrum the s-'cales are mostly pale ochreous ■with numerous black scailes inters2:)ersed ; on the prothorax the scales are somewhat similar, but wider, and there is usually a I»ale median line, on eacli side of the apex of which is a feeble black fascicle. ()n each elytron there is a pale (sometimes almost white) oblique stripe fi'om in line with the shoulder to near the suture at about the middle, but^ touching neither suture nor shoulder (the two to the naked eye appearing like a feeble V) : parallel with this and about half way between it and apex aa*e traces of another feeble stripe, and thei'e is usually a small whitish spot close to airex. There are fe'eble black fascicles on the second and fourth (and sometimes on the sixth) interstices about the middle, on the third and fifth near the base, and a few still more feeble ones elsewhere. The clothing of the under surface and legs is paler than elsewhere, amd the black scales are entirely absent. In sluqie it closely resembles alternans, but is con.-^iderably larger, with denser clothing (without lineate arrangement of colours except the veiy indistinct median line of prothorax), and with tlie rostrum decidedly shorter and wider. IVlicpocpyptoi''hynchus, n. g. Head large, invisible from above. Eiiex small, ovate, widely separated, coarsely faceted. Rostrum short, stout aind almost straight. Aiitmnae rather st subcontinuous with funiele. Prnthorox longer than wide, sides slightly rounded, base and apex almost Colcoi>b'i'n (>/ KiiKi I .si and. * 105 equal in width, ocuhw lobes obtuse. SruteUum not traceable. Elijtra isliiihtly wider than prothorax, oblong-elliptic. I'ertoral nnial deep and wide, terminated between intermediate coxae. Jloiosfrrnal rerepfarle scaireely raised, emarg:ination semi-circu- lar ; cavernous. M it astern urn slightly shorter than the follow- ing segnuent ; episterna not traceable. Ahdomtn moderately large, tw(j basal segments large, the three apical depressed. Lrgs moderately lon^ ; femora not grooved or dentate, pos- terior not extending to ajiex of abdomen ; tibiae stout, almost straight : tarsi short. 3rd joint wide and deeply bilobed, 4th elongaite. Subcylindrical, elongate, squainose, apterous. In addition to the species described below, two others are known t^) me. I do not know amy closely allied genus and its position in the ("lyptorhynchides is very uncertain. For the present it may be placed at the end of the allies of Poropterus, althouLrh the a.ppearance of the head and rostrum is not unlike many •i the allies of Chaetectetorus. ■J20. Microcrvptorliynclins pyo/nacus, n. sp. Dull red or brownisli red. Densely clothed with muddy scales : and with numerous semierect setaie scattered about. Html with rather coarse but concealed punctures. Kostrum with distinct punctures on aiiical h;ilf in female, on aipical third in male : elsewhere concealed. I'rotliorai: verv' little wider than long, sides moderately rounded, apex about two-thirds the -vWdth of base : with dense and coarse but concealed punctures. Fjhjtra elongate-cordate, gently elevated to a,bout the middle, thence strongly rounded to apex ; with rows of large concealed punctures, interstices as wide as and slightly narroAver than punctures, the alternate ones distinctly raised. Two basal seg- ments of ahiloiiKu with dense, large, concealed punctures.. Length. \\ — 1^ mm. Tht smallest Australian spwies of the sub-faauily a.s yet described. Before abrasion the sculpture is almost entirely con- cealed. The derm is sometimes of a dingy brown, especially in the males. The scales are always muddy looking, and not in- dividually traceable. The setae are st-out and more or less erect, but not long, but longer on the elytra than on the prothorax; they are nowhere condensed into fascicles. Both 196 * A, ■linn- M. L'n: scales and ^^etae ajiiieur to be easily abraded, and specinieus ore usually very dirty when obtained. The sexes are readily dis- tinguished by the clothing of the rostrum. Two specimens from Tasmania (Mount Wellington) may repre- sent a variety ; they differ in being almost Ijlack except fr>r the antennae, tarsi, and part of the rostrum. Wibupdia, n. g. Head rather large. Eyes rather small, distant, finely faceted. Ilostrum rather short, stout, feebly curved : scrobes considerably widened posteriorly and partially visible from above. Antennae rather stout, submedian ; first joint of funicle moderately long, the seventh widely transverse and aijJiiarently forming portion of club. Prothorax transverse, apex narrow and subtubular. base bisinuate, ocular lobes almost rectangular. Scutellum distinct. Elytra subcylindrical. baise trisiuuate. Pectoral canal deep and wide, terminated before middle coxae. MesO'Sternal receptacle thick, not raised and slightly concave. Metasternum elongate. Abdomen with all sutures distinct. Femora edentaite.-^^ not grooved ; tibiae with subai>ical tooth as well as with terminal hook; third tarsal joint wide, deeply bilobed, fourth elongate. Winged. In general appearance resembling Mctyrus iind the genus to which Cryptorhynchus sirius. Er. belongs. Init with the mesos- ternal receptacle^ totally different to any of the allies of Chaetectetorus and someAvhat resembling that organ in Therebus, aaid some of the other allies (if Psephnlax ; for the present, how- ever, it may be placed near Metyrus. The seventh jex (not '" apice leviter dilataitum.") Erichson also makes no inention of the conspicuous rostral carinae, and the size he men before me.^ At a glance it appears to be close to Entromus gives (2^7 German lines) is less than that of the smallest speci- dorsoplagiatus, but the rostrum and prothoraicic carina are very different from those of that species.. On rhe basal half of the rostrum most of the pubescence is white, and the clothing of this colour extends backwards on to the head in the form of a trident, the outer prongs of which margin the eyes. On the prothorax there are numerous scat- tered spors of Avhitish and yellowish pubescence. The scutellar clothing is entirely pale. On the elytra there is a large sub- quadrate pale patch extending from about one-fifth from the base to near the middle, elsewhere there are numerous spots of varialile colours. The legs ai-e prettily variegated with red and black, and with rings of black and white pubescence. Between the district prothoracic carina and the base another but much more feeble one can be traced, and between these two there twe traces of two still more feeble ones. In addition to the type and above described specimen thei-e are three others before me. Of these one from Jenolan (New South Wales) is slightly smaller (Of mm.) than the type and the subquadrate patch of paile scales on the elytra is much smaller and much less distinct. One from Mount Kosciusko (New South Wales) is still smaller (8 mm.), and the patch can scarcely be 1 Since this was written I ha\e examined the t\i>e of rioi>ideies nmsivu.s, V.v. : it cer- ainly belonKS to Kpargenius, and in fact is very close in appearance to tlie Huon Kiver sjicoinicn, but is smaller, less robust and with the rostral carinae (if present) quite concealed by tlie dotliing-, the prothoracic carinae are identical. Krichson's description of the rostrum is misleading, as it is quite stronjjly dilated towards the apex. Colcoptcra of Kiny Idmid. 201 traeed. The last from the Huon Kiver (Tasmania) is smaller still (7-^ mm.), the patch is also very indistinct, the antennae (exceptint; the club) are entirely pale, the legs are also pale with the exception of the tips of the tibiae, and the prothorax and elytra are reddish. On all four specimens the suture, near and on the posterior declivity, is alternately marked with black and white spots. 'l-~. Xvnotropis niicmis, Blackb. Ci:j{amhycii)Ai:. 22S. Toxeiites d/citaiits, Fahi-. 'I'lS). Enneaphyllus aeneipeniiis, Watli. 1*30. Pliacodes ohsaims^ Fab. 2H1. /-". personatiis. Ki-. 232. Epithora dorsa/is, W. S. Maci. 233. Callidiopsis scntellaris, Fab. 234. Grdciliit pvi^^/iuiea, Kab. 235. Plerasteiuis amco/or, W. S. INF. 236. P. sitiuralis^ 01 i v. 237. Amphirlwc decora^ Newm. 238. Macrones piirpiireipes, u. sp. Black, ill places blackish brown ; appendages with a decided bluish or purplish gloss; elytra whitish and semi-transparent, but with the thickened parts blackish brown ; hind tarsi with first and second joints flavous, the third dark brown, and the fourth reddish. Under surface with dense, fine, greyish pubescence. Head with numerous regularly distributed punctures ; with a deeply impressed median line from near base to near clypeus. Antennae extending to second segment of abdomen, first joint as long ais three following combined, third longer than fourth, the others regularly decreasing in length, but eleventh once and one-half the length of tenth. Prothorax longer than wide, irregularly transversely ^\Tinkled, with three tubercles (of which one is lateral and the median one is veiy feeble) transvei-sely placed at the basail third, and a feeble tubercular elevation on 202 Arthar M. Lm : each, side of middle, at apical third. ScuttUuin subtriano;ular, with irreofuliu- punctures. Elytra passing base of penultimate seg-ment of abdomen, strongly nsirrowed to basal third, thence line-like to apex ; each with two punctate or gi-anulate discal oostae, which towards the base cm've rcnmd to and become con- joined by rugulosities on the shf>ulder ; sides and margins raised ; semitrajnsparent portion with shallow obscure punctures. Under surface with dense minute punctures, and dense fine transverse impressions. Length, 30 mm. Also from Tasmania (Hobart). A large species second only in size to ritfu^. The rugose parts at the shoulders are less in area than in that species, and the scupture of the prothorax is very different. In general ap- pearance (except that it is much larger) it somewhat resembles exilis, but the femora are not reddish at the base. I have described a Tasmanian specimen, as the only one from King Island before me is evidently immaitiire. 239. M. sitbclavatiis Pasc. 240. Ancita iiiai-^^inicollis, Boi. Chrysomklidai-:. 241. Cryptocephalus pal lens. Lea. Numerous specimens obtained frtjm Melaleuca and Lcptasper- mxmi scrub. In some of the females the whole of the under surface, the head, scutellum aind legs are pallid ; and in some males the abdomen, except at apex, is almost entirely infuscate. Tlie second joint of the antennae is distinctly shorter than the third, not "' almost as long," as previously descril)ed; in some specimens, however, it is slightly longer than in others. 242. C. subfasciatiis, Saund. 243. Cadmus a2ts trails, Boi. 244. C. coi^iiatits, Saund. 245. Loxflpli'itriis vhidls, Saund. 246. Lachnahotliya saiindcrsl, \\;\\y. 247. Toinyris I'lrlitiilo, Kr. 248. Paropsls accllvls, Blackl). Cob'optern of Kivf/ Island. ?03 249. P. s///>/ascia/(i, Clip., var. />/ij/iior, lilaokh. 250. r. aoricoia. Clip. 2oOa. p. ai^rico/a, Cli})., \;ir. 251. P. dehilis, Chp. 252. P. /a//ax, Newiii. 253. P. hiten, Marsh. 254. P. ohlileraia, Er. 255. P. (uphaiia. Er. 256. P. porosn, Ei'. 257. P. reticulata, Mar.sh. 258. Chaltolampra hitrsii, lilnckb 259. Arsipoda variei^nia, W'atli. var. kingensis, HIackb. 260. A. erichsotii, Baly. 261. Haltica gravida, Blacklj. 262. Moiwlepta sordidula, Blackl). Ekotyllidae. 263. Tiiallis vinula, Y^v. Cocci NKi.Li DAE. 264. Leis con form is, IJoi. Four specimens from the island have the markings covering a greater area than any others 1 have seen, the spots on the elytra are all more or less conjoined, and the prothoracic mark- ings are conjoined on the basal half. 265 Halyzia >iie//yi. .Muls. 266. A^t/vius cardinalis, Muls. 267. Sc\inin/s cortica/is. n sp. Hlaok; ai wide median stripe nn each elytron, tarsi, til)iae antennae and paJpi more or less red. Moderately chjthed with short, whitish pubescence, on the elytra sinuously disposed. Upper surface with dense minute punctui-es, larger and sparser on elytra than elsewhere. Intercoxal pi-ocess of prosternuni almost parallel-sided, sides very finely carinarted. Metasternum and abdomen with dense, small punctures, sparser in middle 204 Arthur M. Lea: than elsewhere ; lamellae touching suture, the latter very feeble across middle. Length, 1§ — ^2| mm. Also oommon under bark in Taismania (Hobart and New Nor- folk). The reddish elytra! stripes commence near the base and be- come conjoined near the apex, on their outer margins their out- line is regular, but on their innei' sides they are sometimes angu- larly encroached upon abriut the middle. Usually the front angles of the prothorax are reddish at their tips and occasionally the extreme apex is reddish. On a> small specimen from Hobart the elytra' are mostly red, with a fairly large oval piceous spot extending from the base to the middle, and with the margins very narrowly infuscated on the basal half. Tlie tibiae r.re usually somewhat infuscated. A depressed species close to description of yarrensis, but larger and mostly deep black (including the head and femora). In colour and size it is somewhat close to vittipennis, but the stripes- do not commence at the base itself as in that species, and meet across the suture (except for the tinely raiised jiortion of the suture itself) instead of tei-minating ])efore it. It is also flatter than that species, with denser punctures on ehi:ra, wider pro- tliorax, darker legs and epipleurae entirely dark. 268. .V. flavitrons, I'.l.u-kl.. 269. Rliizohius /lii^roiuvi/is, n. sp. Fla/vous with l)lack ( r infuscate markings. Moderately clothed with tine whitish pubescence. Head and prothorax with minute punctures ; elytra with small punctures, but, except when concealed by clothing, clearly de- fined. Intercoxal process of prosternum wide, gently convex, dilated to apex, sides ver}' finely curinated. Sides of metaster- num and of al)domen with distinct punctui'es, elsewhere shining and ahiiost or (piite iinpuiiclale; lamellae extending rather more than half-way to suture. Length, 1| — H mm. Also from Tasmania (Frankford, LTlverstone, Laamceston and New Norfolk). Although there are .33 specimens before me, hardly any two are identical in all their markings. Tlie head is sometimes en- Coleoptrrtf of Ki inj Idand. 205 tirely pale, sometimes infusca'ted and sometimes almost entirely black. Tlie prothorax usually has a larj^e infuscate blotch in the middle, the blotch occasionally occupying the entire surface ex- cept for a ver\' narrow border, whilst sometimes a very faint stain only can be traced. The elytral markinp^s are very variable and not always clearly defined ; the suture appears to be alwjws narmwly infuscated throuo-hout, at about its basal third there is a blackish blotch (in some specimens this blotch is heart-shaped, in others it is ccmnected with discal markiniis, whilst in a common form it is represented by a rounded spot on each side close to. Init not of, the suture), and at about its apical third it is a.gain, but less strongly dilated : in uumy specimens, however, the subapical dilatatifm is entirely absent. On the disc there is usually a sinuous line extending from near the base to one-third from the a-pex, where it becomes trans- versely dilated and termin.Tjtes ; sometimes after proceeding a shr»i-t distance it bifurcates, but the two arms in such cases become conjoined at one-third from the apex. The meso and metasternum are always more or less dark, but the abdomen varies from entirely pale to entii'ely infuscate. On one specimen the elytral markings consist of a conspicuous zig-zag fascia at the basal third (extending across the suture but not to the margins), and a. feebly infuscated spot at about one- third from the apex. On several there is a feebly infuscated spot on each side of the suture at its basal third, and a very feeble oblique stripe between this and the margin. Usually, however, the sinuous line can be traced in parts. The specimens from the island, a.s a rule, are less distinctly marked than tho^e fnim Tasmania. In general appearance somewhat like alphabeticus, but smaller, comparatively wider, with smaller punctures' and different mai'k- ings on eh-tra. In size and shape it is close to occidentalis, but the elytral punctures are much more distinct than in that species. "270. Rhizobiiis hiackburni, n. sp. Black or bhickish, head (infuscated posteriorly) front and sides of prothorax, sides a,nd apex of elytra, abdomen (the base in- fuscated) and appendages more or less reddish. Clothed with short pale yellowish pubescence interspersed with subsetose but similarly coloured pubescence. 206 Arthur M. Lea: Head and 2>rothorax with small dense punctures. Elytra with slightlj- larger and sparser punctures, interspersed with numer- ous larger (but still small) punctures. Under surface with sparse and small punctures, beco-ming very dense at sides. Inter- coxal process of presternum moderately convex, carinaite at apex but not at sides. Lamellaie extending to about one-fifth from suture. Length, 3^ — 3-| mm. Also from Tasmania (Hobart). A greatly depressed species apparently close to aurantii, but the prosternum convex instead of concan-e along the middle, and apparently with smaller jjunctures, those on the elytra being decidedly smaller and denser than on discolor, and uneven in places; the punctures on the prothorax are rather denser than on the elytra and are decidedly small. The pale portion at the apex of the prothorax is very narrow; on each elytron it commences ait the base, close to but not on the margin itself,' and dilates till it becomes marginal, and still dilaites till it occupies about one-third of tlie apex. On one specimen, however, it commences behind the middle and be- comes marginal only near the apex. On the elytra the pubescence is somewhat sinuously disposed in places, and on abrasion very faint traces, as of striation, become visible. 271. Rliizobius ki/igcfisis, n. sj), Black, elytra with a coppery gloss ; head, apex and sides of prothorax, apex of abdomen, tarsi, tibiae (the four hind ones somewhat infuscated), knees, trochanters, antennae and palpi reddish. Moderately clothed with rather long, whitish, curved pubescence, interspersed Avith suberect fine brownish setae. Upper surfaice with small punctmTS of even size, but denser on head than on prothorax, and on prothorax tli '.n on elytra. Intercoxal })rocess of prosteriiuni flat, sides scarcely cai'inated. Metasternum and abdomen with small and spaa-se punctures in middle, becoming dense at sides ; lamellae extending to about one-fourth from suture. Length, 2 mm. Close to lindi, but smaller and darker, i)ubesccncc lunger and seta.e shorter and sparser. From plebejus (except that it is 1 Oil one spec-imen, however, it. is iiiartjinal at the base. Coleoiitera of K'uxj Island 207 nbout the same size) it differs in the same |)artieuhirs. Cnui- pared with a specimen of hirteUus of the same size it diflfe-rs (apart from colour) in havin. Class — Spongida. Order^ — Moiuxctinellida. Cenus — Cliona^ (! I'aiit. Cliona mammiliata, sp. no\. (PI. XVI 11., Fig. W). Specific Characters. — The chambers e.xtavatcd hy the ^pmiLie are comparatively large, irregularly 'ipheroidal and depressed. Vn-lorian T'Ws/V.s', Part IX. 209 111 nearly all cases they liorc smaller loculi on their lateral walls, and these appear in the casts as luaiiniiillate protuberamoes. Cavities connected by rather long and and conspicuously curved stolons. Averaire diameter of chambers, 4 mm. ; length of con- necting stolons, about 3 inm. ; width, 0.5 mm. These borings occur on the surface of the internal cast of a Voluta having a length of 16.5 cm., and the Cliona crypts en- tirely cover the spire and a large part of the body -whorl. Observations. — In the absence of spicules it is difficult to separate the fossil casts of the boring sponge Cliona by charac- ters which may be regarded as specific. In the present in- stance, however, certain features are exhibited which we can use for future reference, and we may therefore reasonably give it ai distinguishing name. As an example of Cliona borings already specifically described we may refer to Cliona ("Ento- bia") cretacea, Portlock^ a common form in Cretaceous shells in Britain and elsewhere, which is I'ecogiiised by its regularly spheroidal form, crowded clianibers and comparatively fine, radiating system of stolons. Locality and Horizon. — Swan Reach, Bairnsdale Lakes. Ter- tiary (Ivailimnan). Pres. by Mr. H. J. Hauschildt. [9146]. ? Cliona peregpinator, sp. nov. (PI. XVI II., Fig. 4). Specific Characters. — Crypts globular to pyriform, sometimes united into a more or less lengthy tube. The passages from chamber to chamber ;ire often reduced to a mere constriction and there is also evidence of occasional, long slender stolons. Diameter of an average-size globular chamber, 2.5 mm. ; length of pyriform chambers, rather less. The habit of this organi:Hm in the wandering manner of its growth is unlike the majority of Clionae. The fossil occurs on the surface of a limestone cast of a coT*al, Comoseris, into the coenenchyina of which it had bored in the errant manner described. Locality and Horizon. — Valley of the Moorabool at Maude. Tertiary (Barwonian). Coll. Geol. Surv., Vict. WT.M2. [9153]. 1 (Jtnl. Loiidondrrry, 1813, Ji- •'*>*'• *^'^'<-' 'il*^<' Clioiiites coiin lieari, Morii-^ : Ami. .Mu^. Nat. Hist., \ nl. \ Mi., 1851, pi. viii., fi^-. 9. 210 F. Chapman: Order — Tktraotinkllida. Genus — Ecionenia. Bowerhjiuk. Ecionema newberyi, McCoy sp. (PI. XVI r., Figs. 1-13). Tethya newberyi, McCoy, 1877, Prod. Palaeont. Vict., Dec. v., p. 31, PI. XLVIIL, Fig. 1. Observations. — Having recently examined the above type speci- men [9145], I am able to record the presence of typical tetr.ictin- ellid spicules (protriaene), in reference to whicli McCoy remarked' as follows : — " I have not seen any triradiate terminations to ainy of the spicules such as occasionally occur with the simple forms in the recent Tethya, but thej^ are «o brittle that such may yet well be found." Jn liis description, McCoy compares this fossil sponge with Tethya cranium, which species is now removed to the genus Craniella, Schmidt. Among the spicules of the Victorian fossil sponge are numerou>^ micruscleres, which are alj- sent in all the forms of Craniella referred to by Sollas,''' except- ing C. schmidtii. This species alone possesses sigmaispires : the microi-cleres of our fossil, however, are represented, amongst other forms, by the simpler modification, the microstrongyles. The known species of Craniella are distinguished by numerous megaloscleres of the form anatriaene, but these are absent in our specimen. With regard to Tethya, the definition of the genus as now restricted and given by Sollas (op. cit. p. 427) is as follows : " Tethyidae of more or less spherical form, in which the rhabdus is a strongyloxea. The chamber-system is diplodal." This definition excludes our fossil, since all the oxea are bluntly pointed, in contradistinction to the cylindrical strongylo.xe.i. The genui5 with which the Victorian fossil appears to show most agreement, both in regard to form and spicular structure, is Ecionema, which includes at least two specii>s found in southern Australian waters — viz., E. .australiense, Carter sp. ,uul E. bacilliferum, var. robusta. Carter var. 1 Loc. cit., p. "1. •2 Chall. Kep., vol. \\v., ISSS. l{fi)oi-t on tlu' Toti-a<'tiii«lli(la. ]>)>. .-{(i^l. Victorian Fossils, Pari IX. 211 Tlie genus Ecionema is delined hy SoUas' as '' Jilijilidastrose Stellettidae. in which the ectosome does not form a cortex, with two fiirnis of microscleres, one of them being a microrabd, derived either from an anthaster or a chiaster by reduction in the number of the actines to two." In the present specimen there are at least four types of microscleres ; spherasters, sterrasters, microstrongj'les and the microrabds (probably derived from a chiaster). It may sul)se- queutly be found necessary to form a new genus for the reception of this sponge should other specimens occui-, but for the present it may be referred to Ecionema. Extended Description. — In addition to the characters noted by McCoy, we may state that the spicules consist both of the large (megalo*cleres) and the small types (microscle'res). The former consist of — (1) long arcuate or sigmoidal spicules pointed at both ends (oxea), generally smooth, sometimes slightly spinose ; and (2) tetraradiate spicules of the form protriaene, with the three short rays directed away from the main axis, sometimes cun'ed, but more often straight, forming an angle of about 45 deg. from the axis of the rhabdus produced. There are also occasional dichotriaene, in which the three radial cladi are bifurcate, and with the main actines suppressed, after the manner of Ecionema nana. Carter sp.- The microscleres con- sist of — (1) arcuate or open V-shaped microrabds, cylindrical and with rounded ends (micro-trongyles), bearing surface tuber- culations and depressions; (2) a Ispiraster, with blunt spines, especially neair one extremity ; (3) a microxea with whorls of spines ; (4) a spheraster. with moderately long arms carrying two or more spines at the exti'eme tips ; (5) a depressed ellip- soidal sterraster. with .'hilum nearly central ; and (6) a sani- (laster slightly tapering to one end. and armed witli numerous short spines. Dimensions of the Spicules. — The chief skeletal spicules are the oxea. which are nearly ahvayi? slightly curved : the greatest length they appear to attain is about 5 mm., although McCoy- says " some apparently about 1 inch long."'' They are massed 1 Loe. supra eit., p. 195. 2 Annals and Man'. Nat. lUst., ^ii'V. \'., vol. \i., 18>i0, pi. vii., f. 48. 3 Loc. cit., p. 31. 212 F. Chapmaii: together in a closely fasciculivte manner. The examples now figured measure as f()llf)vv.s : — (PI.XVTr., I<'i^. 1). Len;L,'th, 2.346 mm.; greatest breadth, 0.0721 mm. A slightly sig-moidal spicule (PI. XVII., Fig. 2), length, 1.6. mm. Protriaene. — A variety v/ith straight cladi, 0.423 mm. long ; length of cladi, 0.154 mm. Cladi making ajn angle of 48 deg. with the produced rhabdui^. A variety with curved cladi having a length of 0.481 mm. : cladi forming an angle of 30 deg. A variety with the cladi sigmoidally cun'ed, 0.461 mm. in length; width of chord, 0.25 mm. Dichotriaene. — Rays of the trivium lying nearly in the same plame. That which would ordinarily be considered the prin- cipal actine is almost entirely suppressed. An example from this sponge has an extreme diameter of 0.4S mm. Microstrongyles.-- Length of an average example, 0.423 mm. ; width, 0.0384 mm. The ?spiraster. — Length. 0.346 mm. Microxea with spine.- in whorls. — Length, 0.25 mm. ; width, 0.1 mm. Spheraster. — Diameter of centrum, 0.0576 ; length of longest rays. 0.0432 mm. An ellipsoidal sterraster. — Longer diameter. 0.153 mm ; i-horter diajmeter, 0.11 mm. A sanidaster with a length of 0.336 mm. Class — Anthozoa. Family — Astraeidae. Genus — Heliastraea, Ed. iuid Haime. HeMastraea tasmaniensis, Duncan. H. tasmaniensis, Duncan, 1876, Quart. Jnurn. Cleol. Soc, Vol. XXXIL, p. 342, PI. XXIL, Figs. 1-3. Obi-ervations. — An examplo of this coral occurs as a cast in ironstone, and is sufficiently well preserved to furnish a sharp wax iinpressiuii, clearly showing tlic number of primary and secondary stpta ar.d their quaternary arrangement, as described by Duncan. The corallum measures about 4 cm. square, whilst the oaliees have n diameter of about 4 nnn. Victorian Fossils, I'arl IX. 213 Near to one side of the coralluiii in this specimen there occurs what is evidently a malformed calice of the same stock, form- ing a funnel-shape depression aibout 15 mm. across, and sur- rounded by a ring of calices of the norinal form. The malformed calice suggests at first sight tliat of an Agarieia, but a cast of the bottom of the calice shows it to be similar to that of the smoller corallites of the group. Locality arnd Horizon. — Flemington (" Royal Park '). Pro- bably from the Vict. Geol. Surv. coll. Tertiar\' (Barwonian). [915o]. Family — 'J'li a.mna.sthakidak. Genus — Comoseris, D'Oil)igny. .Sub-Genus — Oroseris, Ivlsvards and Hainie. Comoseris (Oroseris) australis, sp. nov. (PI. XVI II., Figs. 1, -l). Description. — The present example occurs in the form of a ferruginous limestone cast. Base of corallum encrusting. Calices measuring about 6 mm. in diameter ; arranged in a widely flexuous series, and divided by moderately high, rounded, flexuous ridges. Septa (traibeculae) sinuous, strongly curved or angulate, granulate on the sides, and united by synapticula ; about 20 main septal plates, some of which branch into two, usually at a di>tance of about one and a hailf millimetres from the centre of the calice, continuous with the costae of the ridges. Sometimes the branching of the septa occurs nearly at the summit of the ridge. Columella small, formed of the united ends of the septa. Depth of calices about 5 mm. From top of ridge to bottom of calice, 9 mm. Observations. — The corallum of the type species ha« been ex- tensively invaded by a boring sponge (? Cliona), the casts of whose orypts stand up prominently on the fossil coral. The coral before us bears >ome resemblrnce to certain forms of Stylomaeandra and Latimaeandra, both of which have the calices situated between collines or ridges ; the former genus having a styliform columella, whilst the latter is deficient in that respect. A closer examination of the septal arrangement 214 F. (.Uience of a rudimentary or papillose columella, show its affinity with the Thamnastraeans. The subgenus Oroseris is distinguished from Comoseris by the limited extent of the col- lines, which do not traverse the entire length of the colony ais in Comoseris, and in this respect our specimen is in agreements A closely allied species to ours is Comoseris (Oroseris) regularis, Fromentel, which, however, has fewer septa, and a more pro- nounced papillate columella.- This subgenus is represented in the Jurassic, Neocomian, Cretaceous, Eocene and Miocene forma- tions. In the Eocene it is known from Europe, a.nd in the Miocene from Italy. Locality and Horizon. — Valley of the Moorabool, at Maude. '' From irregular bands of limestone not more than 2 ft. thick, interstratified in the upper part of the older basalt." C. S. Wilkinson, Dec, 1865. Coll. Geol. Surv., Vict. (WTM2). Ter- tiary (Barwonian). [91 oS]. Class — EcriiNoiDKA. Family — Cassidulidac. Genus — Studeria, I )uncan. Studepia elegans, Laube sp. Catopygus elegans, Laube, ISGO, Sitz. d.k. Akad. d. Wis- sensch. Wien, Vol. LIX. p. 190, PI. Figs. 8, 8 a-c. Tris- tomantlius elegans, Bittiiei-, 1892, Sitz. d.k. Akad. d. Wissensch. Wien, Vol. CI., p. 352, PL IV., Fig. 3. Observations. -Hitherto this echiiioid has been recorded t'oi- Victoria only Irom the mouth of the Clenelg River, near tlie S. Australian Border, and from Apsley. In S. Australia it occurs at the Murray River and Mt. (Tnmbicr. It is therefore in- teresting to record its occurreiu-c at aiiiillier, widely removed, localitv in Virtorin. Tlic specimens, of which there are six 1 See Dnneaii, "Revision of the Families and (ieneia of the .Macheporaiia." Jourii. Liini. Soc. Lond., Zoolo<]. Genus — ENpa/cigi/s, Agassiz. Eupatagus potundus, Duncan. Eupatagus rotundus, Duncaai, 1877, Quart. Journ. Geol. Soc, Vol. XXXIII., p. 53, PI. III., Figs. 14-17. Observations. — This species i* not very abundant in our Ter- tiary beds. It is readily recognised by its exceptionally large size compared with the other Australian examples of the genus, the almost circular ambitus, the gi'eater proportional height of the vertex, which is | the length of the test, the nearly centric position of the apical system, and the sharply angulated peri- petalous fasciole. A fine i^peeiraen of this ecliinoid has been presented by Mr. F. P. Spry to the Museum collection [915GJ. The test is part- ly encrusted by a hard pink or reddish brown limestone, and the fossil itself is of a brick-red colour. Tliis specimen wa..s said to be from Muddy Creek, but the exact locality was open to doubt. During a recent vii«it to the Hamilton District I was able to locate the bed of limestone from whence the present example was obtained. It is best developed at the junc- tion of the Muddy Creek with the Grange Burn, and this par- ticular fossil must have come from near the junction or below, on on the Grange Burn, since it is there that the reddish - coloured limestone occurs. The latter occurs as a very thick bed of foraminiferal and polyzoal rock (Amphistegina and Cellepora being the predominant genera), and throughout the bed are scat- tered numerous tests of echinoids, chiefly of Eupatagus rotundus, I also finmd a portion of a very large echinoid, probably refer- able to Linthia gigas, McCoy sp. Tliis bed of foraminiferal and pnlyzoal limestone occupies a position immediately over the richly fos«liferous clays best seen elsewhere at Clifton Bank ; 218 F. Chap man: and it can be traced up the f7ran-tignathidae, have been un- touched heretofore. These three families are well represented in Australia, especially the Cystignathidae. I ha;ve therefore endeavoured to till the gap by this work on the following eight forms : — - HvLiUif;. Hyla aurea. H. ksueurii BUFONIDiE. Notaden hennetti. Pseiidophryne ausiraiis. Aimtomy of Avs/r(i/!e being doubtless the passage of fluid material from the coelom to the exterior. Moreover it is just as certain that they do not normally persist in the adult of the higher Vertebratai. It becomes desirable then to ascertain just where these structures disappear as a feature of the adult, and what changes take place in their r^ lationships and function during their disappcaa-ance. Historical. The history of the discussion as to the presence and relation- ships of the nephrostomes in the Anura is a very interesting one. I give it here in brief outline: — 1874-. Heidenhain. [Ecker, pp. 327, 336] was unable to find them. 1875. Spengel. [Spengel. "77, and Marshall and Bles, '90. p. 147] stated that the nephrostomes open on the 224 Georglna Sweet : surfiMio of the kidney. He found the-iu in Rana, But'o, Bomljinator and Discoglofsns. There may be one nephrostonie to two tubules, or one to four nephrostomes to one tubule. TTiey are connected with the fourth part of the uriniferous tubule. 1875. Meyer. [Ecker, pp. 328 and 336, and Marshall and Bles, '90, p. 147]. Quite independently and un- known to each other, Meyer confirms Spengel's work. He found 250-360 in Rana. 1877. Nussbaum. [Farrington, '93, p. 309], confirmed pre- vious work as to the internal opening. 1880. Nussbaum. [Ecker, p. 328, 336 : Nussbaum, '80], stated that the nephrostonie is connected with the neck of the tubule in the larva, but opened into the Renal Portal Vein in the adult. Weidersheim, according to Haslam [Ecker, p. 336], at one time stated that the nephrostomes had no openings ait all on the surface. 1886. Nussbaum and Wichmann. [Marshall and Bles, "90, p. 150]. Tliese found that in Rana fusca, R. e.sculenta, Biifo calamites, and Alytes obstetiicans, they open into the Renal Veins and so to the Inferior Vena Cava. 1886. Hoffmann. [Hoffman, '86], asserted that they end blindly in the adult, though connected with the neck of the capsule in the larva. 1886. Wiedersheim. [Wiederslieim, 86, p. 756], accepts Nussbnum's work of 1886 with the remark that the peritoneal fluid is no longer lost, but is re- turned to the general circulation like the re«t of the lymph. 1889. Hiijslam. [Ecker, p. 336], states that he could not find any trace of them, and that if present (1) they are very difficult to find, (2) they do not form a free communicating path between any part of the uriniferous tubules and the abdominal cavity, and (3) their superficial terminations have no free cilia Andtomii of Aiistrailaii AmphUjla. 225 1890. Marshall and Dies. [Marshall and Bles, '90, p. 133]. They are easily seen, though not in everj' sec- tion in a series; also, the whole length of a nephrostome is rarely seen in one single section. The nephro8i:ome>-tubule has no relation except of apposition with the urinary tubules, and opens by a conspicuous aperture through which a tuft of flagella projects into the Renal Vein. 1893. Farrington. [Farrington, '93], states that in Rana catesbiana, and R. virescens, they may open directly inwards, or take a short horizontal first. He could not trace the internal opening with certainty : though ciliary action was seen at the external openings. By injection, he obtained almost conclusive proof of their connection with the Renal veins near the ventral surface. 1898. Bles. [Bles, '98], finding considerable scepticism re- garding the point, exhibited before the Cajii- bridge Philosophical Society 4 sections, '" show- ing a nephro.stome tubule opening into a narrow space lined with endothelium and containing a blood-corpuscle, the space being continuous with venous spaces in neighbouring sections." 1898. Frankl of Vienna. [Frankl, '98], attacking an allied problem finds incidentally by injection that there is no connection between the nephrostomes and urinary tubules. 1898. Beissner. [Beissner, '98], confirms the statements of Nus^baaim [1886] Marshall and Bles. 1902. Marshall's "' Frog." In the 8th edition of this work, the Editor confirms and accepts Marshall's and Bles' findings in 1890 and 1898, i.e., that the nephro- stomes open into the Renal Veins. 1902. Howes. [Howes, '02, PI. VII., Figs. XXXV.. ;,nd XXXVI.], shows clearly the opening into these blood-vessels. 1905. Sedgwick, in the new edition of his text-book of Zoology ['05, p. 295], Avrites : " In the Anura nephrostomes are present. ... It has be«.'n 226 Georgina Sweet : asserted that ther open into the Renal Veins. This statement mu>t be accepted with caution. It appears more probable that they have lost their connection with the renal tubules, and per- sist as ciliated cups on the surface of the kidney." 1906. Holmes, ['06, p. 204] accepts the internal opening of the nephrostomes as into the branches of the Renal Vein. It was then with the hope that our Australian forms might throw some light on this problem, that this part of the Avork has been done. Sh lecture. In general external form and position, the kidneys of the forms examined do not differ materially from those of Rana, being flattened bodies, three to four times as long as they are broad, and one-third to one^quarter of their width in thickness. [Hee Pi. XX., fig. 1.] Situated just ventral to the dorsal l)ody wall, in the a,bdominal lymph space, they are covered ventrally by the peritoneum which keeps them in position agaimt the dorsal body wall. The ventral surface is generally flat or occasionally distinctly concan^e, while the dorsal surface is always more or less convex. The outer edge of each kidney i.s formed by the ureter which arises by branches in the substance of the kidney, and runs back behind the kidney dorsal to the large intestine, to open into the roof of the cloaca. Blood-vessels.— The Kenal Arteries vary in number, being generally in 5 or 6 pairs. They enter the kidney usually at about one-third of its width fr(un the inner edge, and break up i(it once into numerous branches. Somewhat dorsal to the ureter runs tlie Renal Portal Vein often receiving one or more lumbar veins fr(un the body wall. This vein breaks up into numerous branches running inwards acr.)ss tlie dorsal surface of the kidney, breaking uj) ais they do so. The Renal Veins arise on the ventral surface of the kidney, sometimes nearer to the inner edge than the entrance of the Kenal Arteries (e.g.. in Crinia signifera). More often these veins leave the kidney on the outer side of the arteries (e.g., in Hyla aurea, Pseudo- pliryne australis, Heloiojiorus pictus, Notaden bennetti, and Anaforii)/ of AiidraJiav AinpliiJiin. 227 Chiroleptes ailbofnittatus), in which latter they arise from the luiddle lit' the kidney. In a few fornir, these veins are very short, the kidneys being so closely apposed ais to appear as one mass ventrally, as in Pseudophryne australis and Notaden benuetti. Rarely the kidneys are distinctly unequal in length as in Chiroleptes alboguttatus, where the right kidney is fully 2 mm. longer than the left. In the more detailed structure of the kidney, considerable variations are f)ii}/ of Australiun Aiiqjhibia. 'Z'6\ described lur H. aurea, the carmine particles were found in the nephrostomial tubes and throughout all the blood-spaces of the kidney and in the Renal and Renal Portal Veins, in which the particles were embedded in the mass of coagulum, but none in the uriniferous tubules or ureter. Apparentlj' the pressure in the Posterinr Vena Cava Avas so great in this instance that the carmine found it ea>ier to spread back into the branches of the Rfual Portal Vein tlian to pass on into the Posterior Vona Cava. BUFOXIDAE. Pseudflphryne australis. The kidneys in this form are much more triangular in trans- verse section than those of Hyla aurea, the outer edge being formed by the ureter ventrally and Renal Portal Vein dorsally. The Renal Veins emerge at the ventral edge of the inner side, while the Renal Arteries enter the kidney just internal or dor>al to the exit uf the Renal Veins, and the Vasa effereutia enter immediately to the outer side of these veins. The general arrangement of the uriniferous tubules seems to be as in Hyla aurea, the difference in character between the glandular and conducting parts of the tubules being specially well-marked. The Malpighian bodiei^ are almost spherical and somewhat less numerous than in Hyla aurea. There is but little supportive fibrous tissue, though the blood-spaces are still small and normal in relationship. The nephrostomes are most numerous posterior to the plane of the hinder- end of the Testes, and from the median line of each kidney outwards. Their funnels run more lengthwise and obliquely in the kidney in this form than in the previous forms, so that they are less often cut longitudinally in transverse sections of the kidneys. However, here, as in Hyla aurea and H. lesueurii, they open into the blood-spaces directly connected with the Renal Veins, their internal ends being alway;> surrounded by a mass of blood corpuscles. Notaden bennetti. The kidneys of this form show the same tendency to adpres- sion of the inner part of their dorsal surfaces as has already been found in PseudophrAiie australii>. Here also the Renal 232 Georgiria Sioeet : Veins are short and enormously large, causing often deep de- pressions on the ventral surface of the kidney. The ureter in some specimen-^ of this species lies right outside the kidney in the parietal peritoneum. Seen in transverse section [see PI. XX., fig. 2], especially in the posterior half, the kidney of Notaden is conspicuously unlike any of the forma so far described. The vertical disposition of the tubules is very strongly marked, in places forming radiating lines from the midventral line of the kidney. The tubules have often pig- mented walls. There is practically no " kidney-parenchyma," the whole kidney being extremely vascular, more so than in any other form of which I have any knowledge, though Chiroleptes alboguttatus, and Heleioporus pictus are also re- majrkably vascular. The extreme posterior end has compara- tively small blood-spaces, but they increase very rapidly in size and number forwards from this point. Along the midventral line of each kidney is developed as a core or "pelvis" occupying one-third to one-half the thickiifss of the kidney, a series of large venous .spaces traversed or subdivided by a network of trabeculae, the blood-spaces in which are connected on the one hand with the Renal Veins, and on the other with the radiating blood-spaces of the general kidney-substance. The general appearance of the kidney microscopically is that of a groundwork of corpuscles in which the tubules and Malpighian bodies are embedded. The Malpighian bodies are normal in number, round and some- what fimall in compai'ison with the size of the kidney. Those in the outer half are often quite close to the ventral surface of the kidney, while those elsewhere form two or three irregular rows art: aboiit the middle of the kidney thickness. The dif- ferences in structure and appearance between the necks of the Capsules, the conducting, glandular and collecting tubules, though i^imilar in character to that found typically as in Hyla aurea, ai-e very much more strongly marked. The nuclei of the cells forming the necks, and the conducting tubules stain very deeply indeed with nuclear stains, so that it is only by careful tracing of the tubules along their length that one can believe that these parts and the glandular parts are really connected. hi n(^phrostomes al>o, Notaden bennetti is quite unlike pre- viously-described forms. They are extremely numerous ix)s- Av<(t(jiiii/ of A a^traliaa Aui'pldhid. 233 teriorly where the venous spaces form nearly half the thickness of the kidney, and diminish in number somewhat irre<:n^ilarly towards the anterior end. In one kidney alone I counted 1067 external openings of nephrostome funnels : I have seen sus many as 1(1 external openings in a single thin section across one kidney. There are here several totally different types of nephrostumes. The first are the normal ones like those found most frequently in Hyla aurea. which are wide, Sihort and un- Ijranched, and run almo>t hnrizonta*lly beneath the kidney sur- face, and emjftyinu' directly into the main venous spaces, very much like that figured for another form in Fig. 4. except that there is no supporting tissue in Notaden bennetti. The second set, although resembling some of those in Hyla aurea in that they branch, are quite distinct from those in the sjtructure of the " funnel.'' This is long, narrow and more tubular than in any form previously described : it branches freely, running a considerable distance into the centre of the kidney. The branches, of which there may be as many a.s five from one nephrostome, run generally along the trabeculae and then leav- ing them, end in a blood-.- pace. There I believe them to open, though I have not been able to detect the actual aperture. These nephrostomes are especiaiUy numerous on either side of the main venous space. Just within the inner and outer edges of each kidney, es- pecially in the outer edge, are here and there coils of small thin-walled tubes, whose cells have deeply staining nuclei, resem- bling generally the second or branching type of nephrostomial- tubule. Sometimes these open clearly to the exterior — ^ome even on the dorsal side of the outer edge and anteriorly (though still through the peritoneum which often is continued on to the d')rsal side of the kidney for a short distance) — at other times they do not open, but are still connected with the surface of the kidne}% and may end blindly internally in a swollen mass of cells — or. one, two, or three nephrostomes may open into a sinirle uriniferous tubule in its ith part — or, yet again, may apparently come into relationship with a smaller tvpe of Malpighian body than that usual elsewhere, while in yet others, the Malpighian body is still there, but is ver^- de- generate. 10 234 Georgina Siveet : It is, I think, evident that here we have exactly what Spensjel [Spengel, '77] and Meyer have described in Rana, viz., that the nephro^tomes open into the ith part of the uriniferous tubules, and that two or more funnels may open into one tubule and vice versa. It is curious that after 20 years, during which time no one has confirmed Spengel's and Meyer's work, but on the con- trary everyone has shown it not to be true in the generality of cases, one should come across a similar condition evidently as a passing stage in a form such as Notaden belonging to quite u different group of the Anura. Forms of Notaden bennetti have been examined from New South Wales as well as Central Australia. The de>criptin]i above given refers to the Central Australian form. There is ))i) comparison in the amount of blood supply in the two sets of forms, the New South Wales form Ijeing practically normal as regards its general vascularity, though there is still a lack of supporting tissue, and a tendency to a central arrangement of large venous spaces such as are so nuiirked in the Central Aus- tralian form of Notaden, and to a less extent in Heleioporus, as will be i-een later. It should be added, however, that in the New South Wales, as well as the Central Australian forms, the various types of nephrostomial tubules are present, although the total number of external nephrostome openings is very much less. Notaden, it may l)e rememl)ered. is one of the burrowing forms met with frequently in Northern Central Australia, where during the drought season they remain underground, in per- manent burrows, having first filled themselves out with water [Spencer, '96, pp. 159, 1().'), etc.]. This water is a.pparently taken in through the mouth, and jirobably through the skin also, during the time of plentiful water, being then absorbed into the vai^cular system, and excreted V\y the kidneys, passing into the urinar}^ bladder. It will be found that in the Report of the Horn Expediton [loc cit.] I'rofessor Baldwin Spencer has described this water as l)eing in the body-cavity of these frogs, but he informs me that on subsequent visits to Central Australia and di>section of a considerable number of forms, he has discovered that it is stored in the urinary bladder and not in the bod)' cavity. In Notaden bennetti there is always a con- Analoniii (>/ ^[iislralidii Amphibia. 235 siderable uiununt of coajiiiluin alono much water in the bladder is related to the tremendous development of the nephrostomes in this form is not at all easy to see, though there is certainly an intimate relationship between the two facts. It seems most probable that the water from the extremely thin- wa.llrd bladder soaks out into the body-cavity, and is passed back by the nephrostomes into the blood vascular system whence what is required may be taken by the organs of the body, the surplus being again excreted into the bladder, and so on; thus maintaining a constant circulation of this water fur the benefit of the body generally. This return of waste with the water from the bladder would be less injurious than in the ordinary frog, since in these aestivating frogs oxidation of the tissue will be at a mininmm, probablj' only sufficient to maintain life. Cystignat}iii)ai:. Crinia si«;nifera. The general kidney arrangement does not call for any special comment, the relationships of the kidney tubules, blood-spaces and supportive tissue resembling those found in Hyla aurea. The glomeruli are spherical, very few in number, and situated close to the ventral surface. The nephrostomes also are very few in number, what there are being chiefly at the anterior end. Their walls, however, are veiy easily distinguishable from those of the uriniferous tubules. They iwe often not much more thau a slit, in ^ome cases no cavity or cilia being visible, but wherever determinable, they open into the venous blood-spaces on the ventral surface. The body cavity of this form also was injected, with the result that the carmine was drawn through the nephrostomes into the blood-spaces of the ventral one-third of the kidney, though to a very much less extent thiun in other forms similarly treated. The small number of nephrostomes. their frequently diminished ■cavity, and their apparently smaller functional activity would seem to indicate that in Crinia signifera, they are rapidly losing IOa 236 Georgritd Sicret : their Tunction and ceasing to exist, compared ^nth otlier forms herein described, unless possibly Hyla lesuenrii. CJuroleples albo^^^uttatiis. Thiis form, like Notaden bennetti, is a burrowing^ one which stores up water in its body while aestivating. Its kidneys are almost oval in transverse section, and seem peculiarly liable to be folded back against one another, their inner edges ■with the Renal Veins forming the ventrai edge of the mas's (Cf. also Pseudophryne australis., and Notaden bennetti). The dorsal sur- face of the kidney is the more convex. Tlie kidneys resemble those of Notaden in having the minimum of fibrous tissue and very large blood-space*^, though the large central venous space found in Notaden is lacking here, the Renal Veins arising in the usual way in Chiroleptes alboguttatus. The general vertical (dorso-ventral) arrangement of the tubules and blood-spaces is very strongly marked as seen in transverse s.©ctions, the tubules being much pigmented and the blood-spaices crammed full of corpuscles. The Malpighian bodies are rounded, ver}' few in number, and remarkably small in comparison with the size of the kidney. Indeed one often comes across v-eotions in which no sign of Malpighian bodies is to be seen. They are found in the ventral one-third of the kidney tliickness. The nephrostomes, on the other hand, are numerous and well-dc>- veloped, though not neajrly to such a,n extent as in Notaden bennetti. In number I found in one kidney, 210 e\i;ernal open- ings, the number diminishing from the anterior end back- wards. There is hardly a section in a full series through the whole length of the kidneys, in which the nephrostomes are absent, while there may be as many as six in one section. To a certain extent they resemble Notaden in having two kinds of " funnelT^," though the branched forms are much less developed than in Notaden. These slope inwards as a rule, at an angle of 20 to 30 deg. for a sliort distance, and then branch : the branches coil more or less through the substance of the kidney, but always end in blood-spaces, where their cilia protrude among- the corpuscles which are so densely packed around these in- ternal openings. There is also, neau- the median edge of the Anatomy of Aa.-ifnditui Amphibia. 237 kidney a series of large short nephrostoinial funnels which open immediately without branching into the main Regnal Veins. Tlie strenirth of the blood pressure in the kidney, as well aa a corroboration of the connection of the funnels with the blood- spaces, is shown in the fact that in two or three cases the red corpuscles had been forced through the internal opening of the neplu'ostome funnel, and were lying entangled among its cilia. As may be inferred from this statement, the cavity in many of these funnels is much greater than in some othea's of the previous genera. Here, too, although the development of nephrostomes is not so great as in Notadeu bennetti, the association of intense vascularity of the kidney with aestivation is very evident. Heleioporits piciiis. Here also the kidneys are almost oval in transverse section. The Ureter and Renal Portal Vein lie on the dorsal surface, near but not at the outer edge. Heleioporus pictus is another of the burrowing aestivating forms, and, as in Notaden bennetti and Chiroleptes alboguttatus, we have here a very vascular kidney somewhat resembling Notaden in type, but much less developed. As in those forms also, the connective tiss.ue is verj" small in amount, and the blood-spaces are so crammed full of corpuscles that no definite walls are often to be found. The regular dorsal- ventral arrangement of the kidney is interfered with somewhat by the greater convolution of the glandular part of the kidney tubules. The glomeruli are spherical and much more numerous than in the last two forms. The neck of the tubule opens dor- sally from the Malpighiau body, while the blood vessels enter and leave the outer side of the glomerulus [see- Text figure and PI. XXI., fig. \\. The nephrostomes are not as numerous as in the last numbering in each kidney 105. They ai-e almost entirely absent at the anterior end, gradually increasing in number to the beginning of the posterior one-third of the length of the kidney, and then diminishing very rapidly to the posterior end. They have been found to open some into the general venous blood-spaces of the kidney, wliere their cilia may be seen pro- truding inwards and surrounded by blood corpuscles: others lying on either side of the main branches of the Renal Veins 238 Georghia Siveef : may open directly into thein. The funnels are large and long [see PL XXI., tig. 4], and in the case of the former, which are branching forms, after entering the kidney obliquely they run horizontally for some distance and then branch, their branches running along the trabeculae far into the ventral half of the kidney thickness. These branching form.s of nei>hrostominaI tubules resemble those of Hyla aurea, and Chiroleptes albogut- tatus, rather than the more strongly defined type found in Notaden. I have examined specimens from Centrail Australia and from Victoria, and find very little difference in the kidneys of the forms from the two areas. Liinnodynastes dorsalis. The kidneys ai-e here much fiattened ventrally and convex dorsally, the adrenal body forming in transverse section- a con- spicuous structure along the middle of the kidney. In general the internal structure is very similar to that found in Hyla a.urea, the connective tissue being considerable in amount [see Plate XXI, fig. 3], and the blood-spaces small and empty and well-defined compared with those of the last three forms^ — the glomeruli are r(unid and fairly numerous. The nephrostome funnels are short and unbranched and somewhat larger than in Rana catesbiana (.035 mm. in diameter ac- cording to Farrington ["93, p. 310]), while those of Limno- dynastes dorsalis are .037 to .04 mm. They have a well- marked cavity, their internal ends projecting into the bloud- spacas [PI. XXL, tig. 3] among the corpuscles when these are pre- sent. I have not detected ainy funnel>* opening into the main branches of the Renal Veins as in some forms previou>ly de- scribed herein. In specimens injected from the Renal Portal Veins under pressure, the carmine was found to be present throughout the blood-spaces, and had been forced out by the prei^sure into the funnels where the particles were found en- tangled among the cilia. B. — TiiK Connection of thk \'asa kkkkhknti.a wnii tiik TvlDNKV. Here, as in Part A, the object in view is to find a sequence of forms in this case illustrating the ma.nner in which in the course of the evolution of the group, the male reproductive ducts have Anaforitij of Axi.stndian Aiaphibni. 239 been {rradually separated off i'rdui tho kidney tubules. In the Fishes ivnd in the Urodeles and Coecilidae, the niide reproductive ducts are very closely connected with the anterior sexual part of the kidne^'. In Bufo and Kuna esculenta, they are closely connected with the Malpif^hian b:)dies of the urinary tubules of the kidney ; in Riwia fusca the connection is less close, being only with the cullecting' tubules. The severance increases in Bonil)inator and DiscoglossuSs till in Alytes the male ducts open (|uite independently of the kidney, into the ureter : i.e., a portion of the mesonephric duct separates off as a duct for the testis, and at the level of the wider end of the kidney this joins the remnant of the original mesoiieplirie duct which functions as a kidney duct. The question of the relationships in Rana has been a much vexed one from the time of Bidder's work in 1846, but as most of it has arisen through the confusion of the two species R. esculenta and R. fusca, its results may be sunnncd up as above. Nussbaum's work ['97, p. 425.], and that of Beissner ['9S. p. 168.] practically settle the main connections as j^iven above for those two species, the only variation between the two being that Xussbaum has only found the loncitudinal Bidder's canal in K. esculenta, while Beissner describes it in R. fusca also. Gfiwral Slnictitre and Relationships of Ducts. The testes lie ventrally to the anterior portion of the kid- ney, being kept in position In- the mesorchium, the fold of peritoneum which encloses them entirely, except for one part of their inner surfaces where the blood-vessels and ducts enter or leave them. [See PI. XX., tig. l.| In shape and size, they vary greatly in diffex'ent individuals and at different times of the year — and as most of my material was siiirit-preserved, except Hyla aurea, Crinia signifera and Limnodynastes dorsalis, due allowance has to be made for distortion by pressure of other organs. I I \ MDAK. Ifyla auiea. [PI. XX., fig. I]. When fully developed, the testes are long, whitish cylindrical bodies, each end being rounded. They average 10 or 11 mm. in 240 Georgiiui Sweet : length and 3 mm. in diameter. The Vasa efferentia run straight from the testis to the kidney, and then entering tlie latter spread directly dorsalwards, branching to enter the ventral ends of the long Malpighian caipsules, as in Bufo [Spengel, '77] and Rana esculenta [Nmsbaum, '97, 1 and 2, and Beissner, '98]. These canals have narrow cavities, and thin walla of small cubical cells with large darkly staining nuclei, and, as a rule, they stand out conspicuously in sectiom^ across the kidney of this form. I have not been able to make certain of the existence of a Bidder's canal in Hyla aurea, comparable to that described for Bufo cinereus by Spengel ['77], and by Nussbaum ['97, p. 425] for Rana esiculenta, and by Beissner ['98, p. 168] for R. fu>:ca also. There is no doubt, however, that the vasa efferentia do open into the Malpighian Capsules, unlike R. fusca, Bombinator, Discoglossus., and Alytes [Wieder- sheim, '86 p. 784], where they open either into the collecting tubules (Cf. R. fusca), or into the ureter itself. As the testis in all male specimens examined by me was comparatively little developed, and no spermatozoa were present in the vasa effer- entia, either inside or outside the kidney, it is possible tha.t the Bidder's canal may be present, but small and contracted, and so evade recognition. It is chiefly the Malnighian capsules near the inner edge of the kidney which are thus connected with the vasa eft'erentia. Ik'FOXIDAK. Fsciiiiopliryne austral is. Here the testes were large, flat, irregular in outline, touether hiding fully two-thirds of the kidneys when viewed from the ventral surface, extending also anteriorly and laterally beyond each kidney. Here, as in all other forms of which male speci- mens were examined, the Vasa etfereutia of each testis run in the mesorchium dorsalwards to tlie kidne}' of its own side. They then in Pseudophryne, enter the kidney at the outer edge of the Renal Vein- without any previous branching. They aijjiear to rim straight in and without forming a Bidder's canal enter the Malpighian capsules at their ventral edge as in Hyhi aurea. Curiously, although the testes in the specimens examined are large and w^,^ll-developed, I could find no sperm Aiialoiiii/ uf Aaalrdlian Amphibia. 241 in the vasa eflferentia, the kidney or ureter. I have, however, no doubt as to the connection of the branches of the Vasa efferentia with the Malpighian capsules as alcove described. [Cf. PI. XX., fig. 1.] Noiadi'N hoinetli. In the male specimens of this species available, the testes were spherical bodies having about the same diameter as the kidney itself, but quite unsymmetricall}' placed — the one at about the middle of the length of its kidney, the other at ihe posterior end of its kidney — each lying laterally to the kidney of its own side, the mark * [in PI. XX., (ig. 2] indicating the inner edge of the testis lying in the mesorchium. As stated in Part I. A., the ureter lies often in the peritoneum lateral to the kidney, and often, near the hinder end of the kidney and posterior to this, it swells out to form a large glandular Vesi- cula seminalis. The Vasa efferent ia. pass in along the mesor- chium, and entering the kidney run straight out to open into the Bowman's capsules of the Malpighian bodies in the inner one-third (if the kidney. I have lieen unable to find any longitudinal Bidder's canal. Chiroleptes alboguttatus. In this form the testes are long and thin, and in the speci- mens examined very feebly developed, i)robably owing to the season of the ^^ear. Tbey were approximately one-half the length of the kidney and one-third its average width. The Vasa efferentia pass into the kidney at its inner edge from the inner side of the testis as usual, and spread out at unce into the kidney substance without forming any longitudinal canal. Apparently they enter the ventral part of Bowman's ca[)sule, as in previous forms. Owing to the extreme vascularity of the kidiiey and the great number of corpuscles present in all the blood--spaces, it is difficult to make out the relationship of these ducts further than as indicated al)Ove. Heleioporns pictus. The testes of H. pictus are very irregular in shape, and unequally develoijed on each side — that on the one side being 242 (jf'oiy/'iKf Sivecl : nearly the full lenoth of the kidney, and that on the other only half that length. The path followed by the sperm on its wa,y to the exterior is more clearly seen here than in any other form examined by me, as not only were the testes very larg*; and well developed, but the >perm could be traced right through the kidney to the ureter. The Vasa efferentia leaving the testis run dorsally, forming a netwiirk in the mesorchium, to open into a longitudinal Bid- der's canal. Both network and canal are continued posteriorly and also anteriorly to the plane of the testis. The canal lies further from the middle line than the Renal Arteries, and both are nearer the middle line than the Renal Veins. From the ampullae on this canal ducts pass both dorsally a.nd laterally, dividing up greatly — one ventral branch runs ventrally towards the outer edge of the kidney forming the ventral transverse canal, as seen in the Text-figure. The branches of these canals open into the ventral part of the Maliiighian Cap- sules. In many cases, thi- Bowman's capsule is greatly distended by the masses of sperm [See PI. XXI., tig. 5J, the glomerulus being pushed quite to one side. All the Malpighian bodies are not so ciinnected with the sperm ducts, and there i.- certainly a relationship between the position of the Malpighian body and its connection or otherwise with the sperm duct^^. Thus no sperm is to be found in the Malpighian bodies occupying the outer one-third of the width of the kidney eve-n when sperm is pre.-ent in the tubules near by. One is apt to be misled as to the existence of such connection, since it does not follow invariably that even when the sperm ducts, and the uriniferous tubules are both full of sperm, that any will be found in the Bowman's capsule to and from which they are clearly open. Nevertheless, I have not been al)le to Hnd any sperm ducts opening into the Capsules along the miter edge of either kidney (Cf. previous forms, e.g., Hyla aurca, amd Notaden bennetti). From the Malpighian bodies, the sperm passes by the ordinary ui-iiiifcrous tubules into the tran>vers© collecting tubes whii'h run outwards parallel with the dorsal surface o\' the kidney to pass their contents into the ureter where tlie sperm may be seen in ureat numbers. Anafonri/ of AusfrdlUiiK Amphibia. 243 Lyinu; in the coeloiu ventral to the kidney is a flat branchinp^ structure nearly coextensive in length with the te.-tis on either side. It is attached to the niesorchiuni on the outer side of the latter, by connective tissue across which small arteries pass from the Renal Arteries, and small veins to the Renal Veins. In appearance it is somewhat lymphoid, and might be regarded as a rudimentaiy fat-body, this being absent in the specimens examined, but for its position. That frtructure is attached normally to the anterior end of the testis, while this lies between the kidney and the testis for almost the whole extent of the latter. A very rudimentary condition of what is apparently the same structure was >een in one instance in Notaden bennetti, where it is in umch closer relationship to rhe kidney surface. I am unable at this stage to add any fr.rther evidence as to its homologies or function. Lininodyiuxsfes dorsalis. Here the testes are very similar to tho^e of Hyla aurea, except that occasionally the testis may lie quite anterior to the kidney. The arrangement of the Vasa efferentia, of the sperm ducts in the kidney and their entrance into the Malpighian capsules, are all >imilar to that already described in previous forms. As in Hyla aurea, Pseudophryne australis and Notaden bennetti, I have not been able to find any longitudinal Bidder's canal, the Vasa efferentia apparently going straight into the substance of the kidney. Sum inn r\. The results of this enquiry may be briefly summed up thusr: — 1. Nephrostomial openings from the coelom are present in each of the eight species examined — viz., Hylai aurea and H. lesueux'ii ; Pseudophryne australis and Notaden bennetti ; Crinia signifera, Chiroleptes alboguttatus, Heleiporut« pictus and Limnodynastes dorsalis. 2. There are five main types of nephrostomes and nephros- tomial. tubules. 3. The first, which never branch, open directly into the main bramches of the Renal Veins. Tlie.se are present in all fornis. 244 Georgind Sweet : 4. Those of the second type fire unbranched nephro-toine*! openino; into the uriniferous tubules, as first described by Spengel in Rana. Tliese are found in Notaden bennetti only. 5. The third type consists of branched nephrostomial tubules opening into the venous spaces. Tliese are found in all forms, and especially well developed in Notaden bennetti. 6. The fourth type is also branched nephrostomiajl tul)es opening however into the uriniferous tubules. This is found in Notaden bennetti only. 7. The fifth type is a^ third form of branched nephrostome tube, which is closed at either or both ends. Tbese are only known in Notaden bennetti. 8. It will thus be seen that in all forms, nephrostomes open- ing into the Renal Veins are present, these being the only type present in most; in one form Notiiden l:'ennetti, all five kinds of internal connections are found. 9. It appears evident that these structures are undergoing very rapid modification at the present time. 10. In Notaden bennetti and Chiroleptes alboguttatus, the vascularity of the kidney is very strongly marked — the uriniferous tubules appearing to lie in a series of much branched sinuses — the epithelial lining being in many cases almost im- possible to determine. Tlie same is true, though to a less ex- tent, in Heleioporus pictus. The vascularity would thus ap])ear to be associated with the capacity of these frogs for storing water in the urinary bladder while aestivating in their burmws during the dry seasons. 11. There is also a marked difference in the number of nephrostome openings in one kidney from a minimum of 3U in Hyla lesLieurii, to a maximum of 1067 in Notaden bennetti. 12. Hyla lesueurii and Crinia signifera are tlie most do generate in respect of their nephrositomes. 13. Evidence from frogs injected with carmine indicates that the nephrostomes do functinn for the conveyance of material from the body cavity inro the kidney. 14. In all forms, the Vasa eti'erentia enter the kidney just external to the Renal Arteries. 15. In Heleioporus pictus a well-marked limgitudinal Bid- der's canal is present, as possibly also in Ilyla aurea, though A)iato)iu/ of Aiislrtdlaii A ihj>Ii tlna. 245 not in Pseudophryne australis. Chirolepteo alboijuttatns, Nota- den bennetti, or Liiimodynastes dor.salis. IG. In all forni>> examined the Vasa effcrentia In-a.uch and enter the ventral part of Bowman's capsule?, in the inner one- third or two-thirds of the kidney. C(>u<://tsuf/is. It must be conceded that there is C(^nside'ra.ble evidence given by the forms lierein described, tliat in the course of their disappearance in the adult condition during- the evolution of the CToup, the nephrostomes have be-en subjected to well- marked modification — their original connection with the kidney tubules being transferred to the Renal Veins, with a correlated change of function from the passage of fluid to the exterior from the body cavity, to that of lymph vessels. Moreover their degree of development seems to be to a great extent individual or characteristic of the species, varying greatly in harmony with their functional importance both in turn being associated apparently with differences in the hajbit of the animal. In Notaden bennetti for some rea-on or other, alongside the cTcater development in one direction, there seems to have been a check to the harmonious development of these structures in all parts of the kidney, since there are still present along the edges, nephrostomial tubules in various stages of modification, as to their iriternal connections. In all the species here described, of which male specimens were obtained, we find that the separation of the male re- productive ducts from the excretorv' ducts has not yet begun, the condition being comparable to that found in Rana esculenta, the higher stage found in Rana fusca not being present in the Australian species so far examined. They are therefore far less specialised than are the corresponding parts in Alytes obstetricans, the most specialised known in the Anura. I have again to thank Professor Baldwin Spencer for the use of the Biological Laboratory in the University of Melbourne, where this work has been done, and for the use of his collec- tion of specimens, and those of the Biological Museum, as well as for much kindlv interest and valued advice. 246 Geoiyhia fSweet : Literature. Beissner, H. — Der Bau der samenableitenden Wege bei Rana fusca und R. esculenta : Archiv. fur Mikroskopische Anatomie: Bd. 53: 1898—9: p. 168. Bles, E. J. — Correlated Distribution of Abdominal Pores and Nephrostomes in Fishes : Jour, of Anat. aaid Phys. : Vol. XXXII. 1898,3 p. 484. Bles, E. J. — On the connection between the peritoneal cavity and Renal Veins through Nephrostoniial tubules in Frog: Proc. Camb. Phil. Soc, Vol. 9, Pt. 2, 1895-8, p. 73. Brauer, A. — Zur Kenntnis der Entwiekelung der Excretions organe der Gymnophioneii : Zool. Anzeiger. Bd. 23, 1900, p. 353. Ecker. — Anatomy of Frog (English translation) by Haslam ; Oxford, 1889. Frankl, 0. — Die Ausfuhrwege der Harnsamenniere des Frosches ; Zeitschr. wissen. Zoolog . Bd. G3, Heft 1 : 1898, p. 23. Fa.rrington. 0. C. — The Nephrostomes of Rana. Trans. Con- necticut Academy, Vol. VIII.. Part 2 (1893), p. 309. Hill, J. P. — Abnormal Connection of Renal Portals in young Limnodynastes peronii : Proc. Linn. Soc, N.S.W., Ser. 3. Vol. 8, Part 2. p. 222. Ilcffiiiann. - " Zur Entwickelungsgeschicte der Urogenitalorgane bei den Anamnia "' — Zeitschrift wissen. Zool. Bd. 44, 1886, p. 573. Holmes. — Biology of the Frog. New York, 1906. Howes. — Atlais of Practical Elementary Zootomy, Plate VII., Figs. XXXV. and XXXVI. Mnrsliall aaid Bles. — Development of Kidneys and Fat-bodies in the Frog-studies from the Biological DepartmtMit, Owe-n's College, Vol. 2. 1890. p. 13:",. N'uss])aum. — Ueber die Endigung der Winipcrtrichter in der Niere der Anuren. Zool. Anz., 1880, p. 514. Nussbaum. - Der (ieschlechtsteil der Froschiiiere Zool. Anz., 1897. 1). 425. Proc. U.S. Victoria, 1907. Plate XX. Proe. K.S. Victoria, Htu7. I'latr XXI. lo.n ur.t Anatomy of Aii-^fral/aii A mpli/ihia. 247 Nussbaum. — ^Notiz zu dem AutVatze— 0. Frankl's. '* Die Ausfuhr. der Ham. des Frosches " — Archiv. f. Mikr. Anat. Bd. 51, 1897.2 p. 213. Sedowifk.— Text-book of Zoology, Vol. 2, 1905. Shore T.W. — Renal Portal Veins in the Frog — Jour. Anat. and Phys., Vol. 34, 1900, p. 398. Slipace. For the sake of clearness, the blood corpuscles have been represented as much fewer than they really are, the whole space being ci'ammed full of tlicni in this form. x 'JOG. FuiUHK 5. Section similar to that in Figure 4. though much anterior to it, and less magnified, showing the gland (?) (gl.), and the con- nection of a Vas eft'erens w^ith the ampulla (a), on Bidder's longitudinal canal, and especially the opening of a branch from the ampulla directly into the Malpighian capsule, where the mass of sperm has pushed the glomerulu;? (g.) quite to one side. The presence of sperm in the uriniferous tubules is also shown. xfiO. [Proc. lioY. Soc. Victoria, 20 (N.^.)- ^''^'- ^J^- l-'O?.] Akt. XVI. — The His XXII.-XXVI.). [Read 12tli December, 1907.] The highlands of Western Victoria form an area mostly occu- pied by ancient rocks between the north-western and south- western plaiins'. The line of Division on these highlands be- tween the north and south flowing streams is variously sp(.)ken of as the Main Divide or the Dividing Range; to the latter name the words " Main '' and " Great " are often prefixed. These names are also applied to the rest of the main watershed line throughout the State. The name Greait Dividing Range had a very simple origin. In the days of ^arly settlement exact description of localities was desirable, and the colony was divided into counties. For the most part the boundary linos of these counties are the streams, as being easily located natural boundaries ; hence the main waitershed became the boundary between the counties of the south slope and those of the north. Tliis watershed is an actual fact on the land, usually easily located, though not always conspicuous. For the most part it forms a range in the popular sense of the Avord. It divides adjacent valleys, amd from the fact that it forms the dividing line between a series of northert and a series of southern valleys, it easily became known as the Great Dividing Hange. The Hist official use of the term was in connection with the definition of county boundaries,' purely as a descriptive term without reference to its vai'ving character. But apart from any geological examination the early surveyors must have known that its aspect varied considerably, and that occasionally it recpiired careful observation to exactly decide 1 N. S. Wales Oovernnient Gazette, 184S. Hi< of WcMem Victoria. 251 its pof-'ition. The use of the term did not imply thut it was ;i Mountain Raniro in an exact geological sense (the date of in- troduction of the term should be remembered), nor do geologists ever seem to have regarded it as such. Rather strangely Professor Gregory' lias described Mr. Reginald Murray as supporting the term, and connects with this supposed support its frequent use. Murray's " Geologv' and Phy^^ical Geography" was published in 1887, nearly forty yea is after the term had first been otfieially used. But we hud on reference to the book that Murray does not use the term Great Dividing Range, but consistently speaks of the "Main Divide." Apparently his supposed support consisted in describing a main divide in Victoria running from east to west, whereas Brougli Smyth" had previously described the principal dividing line as running south to Wilson's Promontory. But Sehvyn'' had already de-alt with this idea, tracing it to Count Strzelecki in a map published in 1845. Every objection which is urged against the Great Dividing Range can be used with at least equal force against this line to the Promontory. It does not conform to the arrangement of the ancient folded rocks ; it is composed of residual ridges of ■denudation, and further it crosses the Mesozoic trough (a feature which cannot be paralleled on the Main Divide) ; so that on this southern line the continuity of any early high land area was soon interrupted. Brough Smyth himself uses the tenns '" Dividing Range " and " Great Dividing Range," and applies the shorter term even to the Mt. Ararat ridge far distant from his main dividing line, and subsequent to the date at which he had ■described the latter.' On the geological map of Victoria the term Gre*\t Dividing Range appears, but its use is not due to Murray. He distinctly says' that " the latest Geological Sketch map is— with the addi- tion of being geologically coloured -the topographical map is- sued from the Crown Lands Department." Even here the con- 1 The Geography of Victoria, 1903, p. 62. •1 GoUlfields aiifl Mineral Districts of Victoria, 1809. 3 Notes on ttie Physical Geography and Geology of Victoria, 1807. ■t Report of Progress of the Geological Survey of Victoria, II., 1874, p. 18; III., 1875, p. 17. ij Op. cit., p. 3. 11.4 252 T. S. Hart : spicuousiiess of the Divide i^^ largely due to the fact that in addition to the hill-shading there is the broken line used to denote the county boundary. The boundiu'v is an actual fact in the contiguraition of the surface, but without the hill-shading would haA'e been shown in the same way as the point-to-point lines which have to do duty as boundaries in >ouie parts of the plains.-!- North-west of Ballarat is a. part of the range which Professor Gregory particularly criticises. He presents what is said to be an actual view of the country, and states that a number of persons would vary considerably in their location of the Divide at this point. I have put the question to a class of students on the road between Blowhaird and Ascot, and though most of them were quite unacquainted with the place they had no diffi- culty in determining its position, and were all in agreement. I have, however, good information that the photograph reproduced in illustration was not taken on the Divide at all. Certainly it does not truly represent the character of the Divide at this point. Enough ha>^ probably been said to show that Murray cannot be regarded as in any way responsible for the use of the term, and that its use and the jjrominence assigned to it by tht Lands Department is not, from their point of view, unreasonable. Professor Gregory goes further, and says that the Great Dividing Range is " a misleading geographical myth." We have seen that the Divide is certainly a.n actual fact ; the name may be badly chosen, but it is ordinary current language, and majkes no claim to be a scientific term. It does not seem to have misled many scientific investigators, though it may have been misleading in the teaching of geography by teachers with little scientific knowledge. Professor Gregory attacks the biological evidence. He depreciates the support of the^ biologists by hinting that it biassed and selects out of the mass of evidence, two items ivv his argument of disproof, the dis- tribution of the eel and of the varieties of magpies. He says that he has heard from fishermen of eels being taken from the tributaries of the Murray ; so have most people, but unfor- 1 See boundaries of the County of Kipon on the (Jcoloyical .Map of \ictoiiii. H'njhhtiuJs i)f Wt'stmi Victor HI. 253 tunately fcir the ;irs' Strait dates from within the human period in Victoria. 2 The .Main Divide, from a geographical point of A-iew, is a watershed line of composite character between the north and south flowing streams. Biologically it is an area of highlands sufficient to offer some direct obstruction to the migration of ])lants and animals, and to establish a climatic difi'erence which further affect*; their distrilmtion. Commercially and indus- trially it is important, not only for this climatic difference, but because it is a suihcient barrier to have determined trade routes by its easiest passes. Politically it has become incorporated as a boundary line of districts in much of our administrative system. I propose now to consider the character of the Avestern high- lands as a whole, then of the Divide as we now see it, then its origin and early history. The rocks of the western highlands are for the most part coloured as Ordovician on the geological maps, though direct evidence of fossils has not vet been obtained over the greater 1 A. H. S. Luc^as, " On some facts in the f!eoi,''rap)iical Distribution of Land and Fresh- water \ertel)rates in Victoria." l^roe. lloy. Soe. Victoria, IX., new series, 18ft7. 2 A. W. Howitt, Aiistralasiaii .Association for the Advancement of Science, Sydney, 1898. Presidential Address, Section G. 254 T. S. Hart: part of the area. With these are aissociatod granitic and metamorphic rocks and a few small patches of more basic igneous rocks. In the extreme west a considerable area of sandstones usually regarded as Upper Palaeozoic occurs, resting on the granitic, metamorphic and other old rock^. A few scat- tered patches of the Permo-carboniferous glacial series also occur, though these are absent from the greater part of the area. The Mesozoic rocks we may regard as outside our present sub- ject. The supposed occurrence of this series at Skipton requires further evidence before it can be accepted as definitely of this age. 1 Overlying the older rocks on the highlands are fluviatile, lacustrine and volcanic rocks of Cainozoic age. On the margins of the highlands some of these beds may be litoral or estuarine. The fluviatile depositTs are in some cases remnants, and then usually at a high level ; in other cases they are well preserved continuous valley deposits, forming deep leads either above or below the present vailley levels. The present surface config-uration is not determined by the folding of the older rocks. To quote Selwyn, " the strike of the older rocks constituting tlie mass of the main range is at right angles to the axis of the range itself, and quite uninfluenced by the granitic and other plutonic or basaltic rocks occasionally met with equally on the range as on either side of. and remote from, its axis." From almost any eminence one of the first features of the landscape which attracts attention is the occurrence of long lines of nearly level-topped or undulating ridges. Occasionally these ridges may abruptly end or bo continued at a lower level. Here and there an isolated volcanic hill rises, or it may be a grouj) of such hills, and more rarely there are solitary hills and ranges of other appearance. The general character is that of a plateau which has been deeply trenched by a series of valleys. Between these valleys are the residual ridges, the remnants of the old high plain. If we imagine the high plain restoi'cd following the line of the present nearly level hill crests, we would have a plain 1 R. A. V. Murray, Report on tlie Skipton Coal Seams. Report of Progress Geol. Surv. Victoi-ia, Vn., ISS Hhjhlduch of WeMern Victoria. 255 often with an appreciable slope and "with marked difference of level at different parts: Abrupt inequalities would be found at places, as, for instance, on the east fa-ce of the Grampians, the south face of the Pyrenees, and to the south-west of Bacchus Marsh. In other cases the fall would be gradual as from Daylesford south-westerly. We must bear in mind, howe\'er, that it is possible for a long gradual slope of the summits to be a result of denudation of a once level surface; as the lower parts of the valleys are likely tO' be deeper and wider, the ridges between them, if narrow, may be reduced in height. The plain clearly does not conform to the folds of the under- lying rocks, and is a plaiin due to excavation, not accumulation of material. As the superficial deposits of the plain are of terrestrial origin we may regard the plain as due to subaerial denudation, and as representing a peneplain formed by long continued erosion. Possibly ai few low ridges older than the peneplain may still be recognised. The present highest point in Western Victoria i^ Mt. William, 3827 feet above sea level ; Mt. Buangor in the Pyrenees reaches 3247- Mt. Buangor is, however, simply a part of the sloping plateau stretching far to the north. If it were part of an older peneplain we would expect more advanced dis- section of the mass, Avhereas the steep valleys of the south slope of the Pja-enees are clearly of no great antiquity. Also if Mt. William and the accompanying ridges had existed before the peneplain was established, we Avould expect them to be either more di.s.sectefl, or that we would find marine deposits extending into their deep valleys. Both of these are best regarded as most elevated parts of the peneplain itself. The peak of Mt. Ararat is in the hai'd contact rocks adjacent to the granitic rock, and with the present small width of the ridge would naturally result from denudation of a late date, and still in progi'ess. The granitic hills of Mts. Beckworth. Bolton and Misery appear to be possible peaks risinses to the north of tlie Divide, crossing it in a gap at an elevation of 1104 feet; it thus enters the valley of the Upper Wimmera, and follows it up eastward through the I'yrenees to Mt. Direction, where by another pass at an elevaticm of 1:214 feet it enters tlie Glenlo<>-io vallov and thence to the Avoca. H'nihliDKU n/ Wrsfrrn Vicforia. 259 One of the most reiiiark.ible of these meridional valleys lies between Larne-Gerin ajid Mt. Buantior. To the north the Mfc. Cole Creek runs down to the Wimmera. But us one stands on Ben Xevis, some miles to the north of the Divide, there is an uninterrupted view down the valleys of Middle Creek and Fiery Creek into the south- westei'n plains. The Divide in this valley needs lookinu' for. Stamdins; on Ben Nevis more than a thou- sand feet above it, it sinks into insignificance. From Mt. Cole lookino- south similar meridional valleys and ridges are seen on the lower country The Larne-Gerin range continues south of the railway line, and i^ sufficiently important to cause the road and railway to run to the same low notch close up to La.rne- Geriu. East of the Mt. Cole Ranges the same north and south ridges continue, but v.ith less elevations. The road from Chate to Lexton crosses a low Divide, but is flanked on either side by more prominent ridges. In fact this portion of the Divide, from the heads of the Glenelg to the head> of Trawalla Creek, is composed of two differing constituents. It is high where it crosses the meridional ridges or continues along them for some dista.nce ; it is usually low where it crosses the intervening valleys. A> the head of each valley is to some extent indepen- dent of the next one, the Divide sometimes acquires a distinct- ly angular character, most marked in the rectangular portion at the head of Mt. Cole Creek. But a short distance to the south-east of Lexton the chara.cter of the Divide changes. The old rocks disappeair, and, instead, the itummit of the watershed is composed of Volcanic rocks. Out of a gi'eat. area of volcaxiic rocks there sta.nd up the peaks of Mt. Misery, Bolton and Beckworth, none of them actually on the present Divide, but forming the highest points of a meridional ridge buried by the volcanic rocks. If these rocks are supposed removed we would have two more great valleys. We may call the western one, from the parish name, the Ercildoun Gap, and the eastern, similarly, the Ascot (iap. The latter is the lower, and even with its extensive lava streams and volcanic hills is still the easiest and lowest level pass across the Divide be- tween Kilmore and the neighboiu'hood of Ararat. Here four madn roads and two railwavs cross the Divide from Ballarat, and 260 T. S. Hart: the situation of Ballarat is at the entrance to a large area of north-central Victoria just as Ararat stands at the gate of the north-west. Further east the Divide is again composed of Ordoviciau Ranges, but the meridional ridges are prominent even in them in spite of the existence of the west flowing heads of the Burrum- beet and Yarrowee Creeks. From west of Creswick a ridge runs south to far beyond Buninyong. The complete infilling of the valley on its east side by basalts has diverted waters across this ridge into the Yarrowee, but this is clearly a late modification. In this eastern valley the Divide is again buried under basalts in what we may call the Dean Gap. The area south of this is commonly known in Ballarat as the Eastern Plateau. Mt. Warrenheip stands on it, but contributes veiy little to it'. The plateau lavas are from the north near the hills on the present Divide, and its surface falls with an un- broken slope past the east side (jf Warrenheip. Further east the Divide follows Ordoviciaai rocks with gradually rising levels to the vicinity of Daylesford. I have attempted to illustrate these featui-es of the Divide by the accompanying plans and sketches. On the general ])lan of the Divide I have indicated some of the north and south ridges. I have had to compile this from various sources. T\vo early maps of Ripon Count}^ issued by the Lands Department both note the absence of a distinct ridge west of Mt. Ararat. One of these marks the ridge south of Larne-Gerin, and states that it was noticed by Mitchell. From these maps also I have obtained the position of the volcanic hills in the Ercil- doun (lap. I have obtained other information from the Geological maps of Ararat and of Learmonth. The view taken from the summit of Mt. Buninyong shows the south ends of a succession of ridges and the intervening valleys. The line of sight to Larne-Gerin crosses the Divide so that the slope of that hill facing tlic observer drains north by Mt. Cole Creek. The view from Mt. JMowhard shows the series of Volcanio hills which form the Divide in the Ascot Gap. The broken line round the base of each liill on the ma.p shows the aj^iroxi- mate extent of the slope from that hill. The Divide is formed by the coalescence of the bases of the volcanic hills, and hence H'i(/l(/(iin/s of Westrrii V'lchirla. 261 may be at any hei'her level than close to the Divide, fallinof from 1508 feet at Waubra Junction to 1350 at Leitrmonth, then crossino- the Divide at abovit 13G0 and ascendiup- a valley to Waubra. Tliere are i^everal swamps close to the Divide where the slopes of the volcanic hills meet. Lake Learmonth only differs from these swamps in beinjr larger and practically permanent, which has been helped by building up its outlet and lirintrinp' in water across the Divide. Beneath the extensive volcanic area both north -west and north- east of Hallarat there is still some uncertainty as to the cour.-es of the old Divide and of the old valleys. We may safely say that the whole area which now drains north also drained north before the lava flows. In the Ercildoun Gap is ai lead falling north ; in the Ascot Gap another lead also falling to the north, and a north falling lead exists not far north of the present Divide at Dean. Bat south of the Divide there are three areas which present difficulties, the Burrunibeet Basin, the western lead>j of Ballarat and the Haddon leads, and the area about Warren- heip and Bnngaree. Investigations have usually proceeded on the assumption of the non-disturbance of relative levels at different points, but we have sufficient evidence that this is not absolutely safe. We must take account of local disturbances^- and of unequal movements affecting large areas, as will be shown below. The question is further ciuiiplicatcd liy changes of the tiow of streams by the ordinary process of river capture, and by the fact that various lava .streams which have altered the flow or in- fluenced it are not absolutely, and sometimes not approximately, contemporaneous even within the one drainage area. By this mean* a great change might be produced in one part of a valley while another part of the same valley was unaffected. 1 Notes on the Stony Creek Hasiii, Daylesford, and references there. Proc. Roy. Soc. Victoria, XVII., pt. II., New Series. 2m T. >S'. Hart: From Sniythesdale eertaiiily a lead was worked with a fall northward under the present snuth flowinir Sniythes Creek. The alteration was probably due to the lava streams. From near Staffordshire Reef a. large valley falls to the north to join the Yarrowe© Creek, and its waters iire turned south iioain down that creek. The Yarrowee valley itself west and south of Bunin- yonn', and the old Durham Lead which preceded it, are compara- tively narrow. The l^al La-1 JJasin cannot have drained south at the west end of Mt. Doran. ^^~e find thus a considerable east and west Divide many miles south of the present Divide. (Such, a Divide is mentioned without the evidence being quoted in Professor Grep'ory's Geoi^raphy. It can, however, scarcely turn to the north-ea.-t as shown in the ligure there.) At Sniythes- dale it has long been recognised. It, must not be ajssumed that this was a Main Divide from which the streams flowed north to the Murray. North of Mt. Doran we find an outlet to the east. .Vlluvial sands just show below the basalt at the fo.it of the Lai Lai Falls. Half a mile east, just below^ the little falls of the Western Moorabool, the ba*-alt for a short distance comes down to the bed of the river. On a creek a little further east a considerable width of sands is exposed and not bottomed. On the Eastern Moorabool ;it Bungeeltap, they are much wider, and show also in some of the creeks in the parish of Bungal. Thence the volcanic plain is unljroken till we reach the Par wan valley, where the estuarine beds appear and are well exposed as far as the steep descent on the Rowfiley fault. There is little doubt that this is the original outlet from the Lai Lai Basin. (The a.ciual area of the brown coal at Lai Lai nuiy l)e regarded as a local subsidence.) This eastward valley could scarcely have drained any appreciable area west of Mt. Buninyong. It received, no doubt, a part of the drainage of the country buried under the southeani edge of the Eastern Plateau. Further north a part (tf the area north of Warrenheip may have discharged its waters past (Gordon and thence also to the Parwan Estuary. To return to the lead at Smythesdidc. The main Trunk Lead has been worked for some distance north, but not far enough to leave its final course without doubt. From the Ballarat Connnon westward, the railway follows down the gentle slope of a lava II l;/li/(i ikJs of Wrshrn Victorht. 263 stream tn the IJurruinheet Creek. Beneath it is a valley into which the Trunk Ijead flows, hut opinions have been divided as to whether the outlet of tin- valley is towards Burrumbeet or in the opposite direction, and thence to the Ascot Gap. If the fall is into the Burrumbeet IJasin. we are still in doubt as to the outlet from that Basin. Parts of that area may drain northward by the Ereildoun (Jap. north-east to the Ascot Gap or south-west by a route near the present outlet. The western leads of Bnllarat aie subject to the i-ame uncer- tainty as the Trunk lead. But even with regard to the Golden Point Gutter itself there is still some diversity of opinion. Close to the south limit of Balhirat City a point is reached at which there is a broad lead tn tiie west, and a comparatively narrow lead to the south. Jioth have been Avorked. Tlie ques- tion as to which was the real outlet of the stream above v/a-s discussed by Murray, l and he decided for the southern — that is, that the lead followed the saiuc' valley as the present Yarrowee. Btit the decision was based mi small diiYerences of levels. Either way the averaige gradient fm* some distance is much less than in the lead upstream. Against the southern otitlet are ilie change in width and perhaps some minor featm-es of the lead itself, the narrow valley of the Yarrowee downstream and the decided north fall from near Staffordshire Reef. Recent bores- show that an outlet is possible to the Avest. There is thus a double uncertainty in the course of the old valley, both as to the direction the waters took at Ballarat and as to the subsequent course of the AVe.stein Loads. Probaibly at Ballarat both outlets have been used, the head waters of the lead having been captured and diverted. The present drainage system at Ballarat is determined largely by the volcanic centres. From the Ballarat Common extensive flows of lava have run to north, south, and west. Murray con- sidered, from the records of the rock passed through in the ]ionshaw shaft, that the uppermost lava stream or "first rock" is here misr^ing. This shaft is in a valley at the south end of Sebastopol. In the present condition of the creek, the second rock, much decomposed on its surface, is seen exposed under 1 Report of I'l'Di^ress Geol. Surv. \'k'., I. '2 .Viinual Hcport, Mines Depai-tment, Aictoria, 1892. 264 T. S. Hart: tlie first rock. The fir.st rock is inissinii' from the shaft, but it continues on the opposite side of the little creek, and is tra.ce- able to the south end of the Buninyonu' Estate. There is no reason to regard Lake Wendouree as a crater ; it is only a shallow depression on the ed determined liy some slight and almost accidental undidation, that may either h;ivc pre-existed on the old sea bed, or been produced during one of the earliest brond and cquabii' upheavals, tlunt resulted in a dry land surface.'' Professor (iregory regards it as connected with the intru.-ion of a series of uranitic masses forn»ing a Pri- Highlands of Wesfrrn Victoria. 205 iiiirive Mountain Cliain. Apparently this is regarded as Devonian in ape. Mr. T. S. Halli has ascribed it to cross folding connected with tho i)itL'h cdnnaonly observed in the folds of the older rocks. I think it can be shown that any feature produced prior to the deveh'pment of the peneplain must be of minor importance in determining the position of the Main Divide. The gradient of the streams which formed the peneplain must ha^-e been very i^light by the time that operation was finished. Without takincr tho estimated grade as low as 1 in 50, 000, » we may say that if as low a.s 3 feet to the mile (about the gradient of a large part of the Tliames), it would be quite inadequate to account for even the more modera.te inequalities of level of the peneplain. Some other cause has established far greater differ- ences of elevation than those of the peneplain as formed, and this cause must have operated subsequent to the formation of the peneplain. At Warrenheip the peneplain level is about 1750 feet above the sea level. Thirty miles to the south are contem- poraneous marine beds which even allowing for depth of water may be stated as elevated less than 700 feet. This gives a difference in elevation equal to 1000 feet in 30 miles. From the south-west of Daylesford to Warrenheip the general level of the pen;^plain falls from aibout 2350 to 1750, or about 30 feet to the mile. These would give slopes quite sufficient to overcome the slopes of the original peneplain, though it would not neces- sarily reverse the original slope in the vicinity of a I'idge. And it must be remembered that the ridge remaining on the peneplain would only be a very nmch modified remnant of an older eminence. The more marked differences of elevation about the Grampians and Pyrenees would be correspondingly more effective in over- coming older inequiulities. Selwyns suggestion leaves the question very open if we substitute peneplain for original sea bottom. Professor Gre- gory's Primitive Chain, if it existed, would have been reduced to insignificance in the formation of the peneplain, but there are gi'eat difficulties in supposing its existence. In the first 1 A'ietoriau Year Book, 190;^-C. 2 Gregory, oj) cit., \i. 7S. 12 266 T. S. Hart: place our granitic rocks do not, as a rule, present the cliai'acter of axos of even small mountain masses. The dip and ^trike of adjacent rocks are little affected by them (except perhaps as will be noticed below). He states ^ that the Warrenheip granite affects the direction of the beds at Ballarat, but he- has stated the direction of the grainite boundary \*rongly, and missed the abundant evidence of folding. His argument on this point at Ballarat completely fails on examination. It is far more pro- bable that most of our frranitic intrusions Avere introduced by a " stoping " process with foundering and absorption of the ad- jacent rock overhead. In addition a Primitive Divide as early as the granitic intrusions does not i>rovide for the southern origin of the glacial series.^ With regard to Mr. T. S. Hall's suggestion, we require a good deal more evidence on the matter of pitch, particularly as to the extent to which it is persistent, and how it varies from place to place. It majy be supposed to originate in many ways, and may be consistent or inconsistent in neighbouring folds. It is liable to be inconsistent if it is due to the making and dying away of individual folds, or if due to local disturbance as by a fault affecting a smaiU area. Besides these it is possible that pitch may originate by varying intensity of the folding from place to place, so that the fold is sharper at one jilace than at another, and is curved in its strike. Or it may be due to the fact that compression in a solid produces a tendency to expand in a direction a/t right angles to the pressure, and this, if pre- vented, may give rise to a simultaneous transverse folding. Or it may be due to subsequent crossfolding. Or to the settlement of an imperfectly supported area over an invading gi'anite. Or finally to the tilting of folded blocks the folds themselves being inert. Settlement on an invading granite might be suggested iij the case of the southward pitch from Bendigo, and the northward pitch from Keilor, But there are other cases which cannot be so explained. Mr. W. Bara.gwanath, jun., has called my at- tention to the pitch at Ballarat East, northerly at Black Hill, and southerly at Magpie Gully, and neither of them near granite. 1 Memoirs Geol. Survey, Vic, No. 4, 190". 2 Ottifer and Ho!?ection near the north end. The latter shows one of the minor folds. He also shows the granitic rocks as intrusive, though in his description he re- gards this point as doubtful. Two causes may have led to his favouring the idea that the granodiorite was intrusive. In the area afterwards worked for gold at Mt. William sandstone from the high hills close by is common. In its natural condition it would be difficult to see that this area was granodiorite covered with a thin layer of debris from the hills, and he >eeuis to have mapped it as sandstone with dykes. Also there are undoubted dykes in the sandstone. Those with which I am acquainted at Hall's Gap are more probably allied to the Coleraine trachyte than to the granitic rocks. The strike of the Grampians sandstone varies considerably, being distinctly west of north in the northern parts, but more meridional or east of north in the vicinity of Mr. William. The dips are usually low, but near Hall's Gap Krause records 77 deg., 1 Profjress Report, Geol. Surv., Vic. I. 270 T. S. Hart : and I have observed 60 degi-ees, both on the outer ranpre. A dip of 51 degrees is shown on Stony Ci-eek by Krause. The greatest elevation is at Mt. William, in the centre of the eastern ranges. Neither of the earlier sections show any faulting, though the descriptions in both cases give estimates of thickness which are quite inadequate, apart from faulting. The main faults have not been observed, and I have no definite information to assign them a probable inclination. I have observed a nearly vertical strike fault in Stony Creek, near Hall's Gap. There is, however, little doubt that the great valleys here are determined primarily by faults allowing a succession of parallel ranges to be formed of the same beds, and it is probable also that many other parallel faults occur allowing repetition of beds in the long gentle slopes of the hill towards the synciine. The average in- clination of these slopes is much more gentle than the dip, though in a hill south of the junction of Fyan's and Stony Creeks at Hall's Gap it is possible to walk a long distance down the bare rock following the bedding plains. A feature of Fyain'« Creek valley, which indicates late move- ment on this fault, is the form of some of the tributary valleya At the Silver IJand fall the stream comes out from a gap in the east face of the range and drops into the valley below. We may ascribe the Grampians to the unequal elevation iuid tilting of a number of fault blocks, in which the principal faults were approximately meridional. The synciine was probably pre- ©xistent and sandstones already somewhat denuded, forming a part of the peneplain, so that they already were reduced in thickness on their eastern edge. Another fault may exist under the Mt. William Creek valley outside the ranges, and the whole series may be associated with similar faults determining the west end of the high lands as a whole. The faults need not necessarily be altogether of late date ; all that is requisite is movement subsequent to the peneplain fdriuation, it nuiy be on old faults, and only on some of these. Wo may extend this idea of block elevation to the rest oi the highlands. A striking example is seen in a side view of the Pyrenees. Such a view is obtained from many points ; that from Maiden Hill at Waubra is pjwticularly good. At the south end Highlands of Western Victorut. 271 the ranges drop suddenly to the lower country about Beaufort. There is no doubt that this southern end of the Mt. Cole Ranges is a fault scarp, and forms the south end of a great tilted block. It is al>o pro'bably nj fault line of early date on which move- ment has been renewed, as the gi'anite is apparently absent from the peneplain below. The south end is deeply scarred with short steep valleys (not visible fron;i this point), and the edge of the granite is in part buried under the talus fans. Northward the ranges present a long even slope. Buangor is 3247 feet above sea level ; Ben Nevis, 2875 ; Avoca Hill, 2464 ; and Lands- borough Ilill, 190.3. The Avoca River runs north down the east side of the highest ranges. Further north the Richardson and the Avon drain the north end of the' block. The summits of the Pyrenees are to be regarded as the same peneplain as at Ballarat, but elevated more than 1000 feet higher. The parallel range of Ben Major and Ben More is on this side the Avoea River between it and the Bet Bet Creek. Possibly the Avoca followTs down an east boundary fault, for the Ben Major rang© is much lower, and the levels fall still more to the east under the lava filled valleys. There is one great interruption to the regular slope of the lino of the Pyrenees summits. Behind Lexton we look into a great gap in the majss ; in the gap is the pointed summit of the Sugarloaf, and beyond is the shoulder of Ben Nevis. We look down the valley of the Upper Wimmera. Two explanations are possible. There may have been an original slight hollow pro- duced in the elevation of the block or a change of slope. Tlie sniaill dift'erence in height between Ben Nevis and Avoca Hill seems to favour this as well as the extent of country whose drainage is diverted west and the fact that the diversion is not very new. As an alternative it may be simply a case of river capture, perhaps helped by an original steep fall at a fault on the west of the bli:ck. Some amount of capture from the Richardson and Avon has probably taken place, but an original inequality of elevation very likely started this course of the Wimmera. The Pyrenees may be taken as representing a range, due to the same causes as ordinarily produce the Pennine type, but with a gentle or slightly undulating long slope and now much modi- fied by denudation. It is evident that in such a series of eleva- 272 T. S Hart: tions and tilting- there may be more than one east and Ave«t crest produced. This would explain the character of the drainage system initiated in the Ballarat area, Mt. Doran, if previously formed, helping somewhat to constitute the southern crest. From the high land before mentioned running from east of Cres- wick to beyond Buninyong there seems to be a general fall to the Avest. Much of this may be due tO' the later denudation, but it is not unlikely that a fault exists under the Ascot Gap or at least a line of an original minimum elevation. A fault might have contributed to the preservation of the small patch of glacial rocks known to exist at the Midas Mines. That east and west crests would be ait first produced might be expected from the general trend of the Mesozoic trough, the tertiary trough and Bass Strait, all of which may be regarded as a series indicating a prominence of movements on east and west fractures since the time at which the active folding of the older rocks ceased. Some of the faults suggested may seem to have little to sup- port them, but I think there is sufficient evidence to sum up the character of our western highland and Divide as due to unequal block elevations of a Mesozoic or early Tertiarv pene- plain, with subsequent extensive modification by denudation and volcanic activity. DESCRIPTION OF PLATES XXII.-XXVI. Fig. 1. The Main Divide and some of the transverse i-idges, from near Lexton westwards, compiled from various sources. The names of the principal north flowing streams are shown. The south slope, except close to the Serra Range, drains eventually to the Hfipkins. The numbers indicate hills as follows: — 1, Mt. Ararat; 2, Larne Gerin ; 3, Mt. Buangor ; 4. Ben Xevis ; 5, Sugarloaf ; 6, Avoca Hill ; 7, Ben More ; 8, Ben Major ; 9, Mt. William, a is the gap by which the railway goes from Ararat to Stawell ; at h the Ararat-Avoca railway crosses the Divide; and at c passes from the Wimmera to the Avoca valley. The Main Divide is indicated by a broken line. Fig. 2. Hills on and near the Divide at the Ercildouu and Ascot Gaps. 0 indicates Ordovician amd G granitic artxis. The full Pruc. h'.S. \ictoriii, 1!»U7. I'lat.' XXII. / I / v£> 3l .#' '^ w:' ivv ,Tl|Wt^ -^: *'»"' "5: C) -J UJ «4J c< $ 9 1 > Uj K "-*^ :::> 'o P^ ^ ^ ,§ ^ ^ R ^ ^ ^ kj \J h5 < "^ ^ ^ fevS^rs. ^ Proe. R.S. Victuiia, VM)7. I'lat.- XXIir. I'ruc. K.S. Vi.'t.uia, I'.ioT. I'l.it.' \ \ I V K ^ ^'^ ^ ::i^ ^ a: o I oq 10 LO l~~, 2: lf is indicated by the heavy broken line, and the fainter broken lines mark the approximate limit of the slopes from each vol- canic centre. A number of swsiimps are shown by the letter S. The volcanic hills are as follows: — On the Divide: 1, Brown's Hill; 2, Bankin's Hill; 3, Coghill's Hill, 1630 feet; 4, Mt. Cavern, 1588; 5, Mt. Hollowback, 1842; 6, Mt. Pi^gah, 1771. South of the Divide: 7, Mt. Blowhard, 1664 ; 8, McLean's Hill; 9, Morton's Hill; 10, Saddleback Hill, 1548; 11, Weatherboard Hill, 1826 ; 16, Mt. Rowan. North of the Divide : 12, Tinkler's Hill; 13, Webster's Hill; 14. Vaughain's Hill, 1611; 15, One Mile Hill, 1443. The information is largely from Mr. Norman Taylor's Geologi- cal Map of Learmonth. Fig. 3. Profile of the Divide in Ascot Gap, Granite and Ordovician Ranges in the background shajded ; the numbers have the same significance as in the preceding diagram. Sketched from Mt. Blowhard. The level topped Ordovician Ranges seen in the distance are partly west and partly east of the Dean Gap, a few volcanic Hills in this Gap are shown. Fig. 4. Sketch of the Hills west and north-west from Mt. Buninyong. The distances of several of the lulls are given under their names. Distances from point to point are also shown. The Serra Range is often visible as a more continuous line. Fig. 5. Diagram to illustrate the possible courses of the leads south of the present Divide. 0 indicates areas of Ordovician rocks close to the surface ; G of granitic rocks. The interven- ing areas are mainly volcanic or alluvial. The direction of the fall of some of the leads 'v? sho^vTl by arrows. Fig. 6. Diagrammatic cross section of the Grampians, south of Hall's Gap. The full lines show the probable position of the fault lines along two valleys, the dotted lines probable other faults whose number and position are uncei'tain. The depth lo which the sandstone extends is also uncertain. Fig. 7. Profile of the Pyrenees as seen from Maiden Hill near Waubra. Mt. Mitchell, in the foreground, is a volcanic hill. The increasing distances of the hills are shown by fainter lines. ANNUAL REPORT OF THE COUNCIL FOE THE YEAE 1906. The Council herewith j^reseiits to ^Members of the Society tlie Annual Report and Details of Receipts and Expenditure for the year 1906. The following meetings were held : — March 8. — Annual Meeting and Election of Officers. Ordi- nary Meeting. Mr. J. A. Smith exhibited and described a new method of testing lenses during their figuring. Mr. A. E. Kitson exhibited some snakes from Western Australia which were stated to have killed themselves by self-inflicted bites. Also some dried plants which were reported to have grown several inches in length while in the Herbarium cases. April 19. — Mr. J. A. Smith delivered a lecture on "The flow rf fluids, illustrated by stream-line methods." The lecture was illustrated by experiments. May 10. — Papers read : 1. " Some little-known Victorian Decapod Crustacea, with descriptions of new species, No. 3," by S. W. Fulton and F. E. Grant. 2. " Census of Victorian Decapod Crustacea, Part 1 : Brachyura ; " by S. W. Fulton and F. E. Grant. 3. " Xew Species of Victorian Marine Mollusca," by J. H. GatlifF. June li. — Paper read: " Micrometric measurements by a projected scale," by Dr. F. Clendinnen. Illustrated by experiments. July 12.— Lecture by Prof. W. C. Kernot, M.A., M.C.E., on " Balloons and Airships." Illustrated by lantern slides. Aug. 9. — Lecture by Kerr Grant, M.Sc, " The Vibrations of Jets." Illustrated by numerous experiments. September 9. — Lecture by W. N. Kernot, B.C.E., " Some applications of the Electro-magnet." Illustrated by lantern slides and by numerous e.xperiments. October 11 — Papeis read : 1. "New or little-known Victorian Fossils in the National Museum ; Part 8. — Some Palaeozoic Brittle Proceedings of the Royal Society of Victoria. 275 Stars of the Melbournian series," by F. Chapman, A.L.8. 2. " Note on Caligorgia flabellum from Port Pliillip," by Prof. Sydney J. Hickson, D.Sc, F.R.S. 3. "New or rare Australian Plants," by Prof. A. J. Ewart, D.Sc, Ph.D. November 8.— Lecture by Prof. E. W. Skeats, D.Sc, "The life history of a crystal. ' Illustrated by lantern slides. December 13. — Papers read: 1. " Remarks on some Sub-fossil Bones from King Island," by Prof. W. Baldwin Spencer, O.M.G., F.R.S. 2. "Surface-tension as an aid in Canyon-formation," by J. A. Leach, M.Sc. 3. " Description of a new species of Cypridina from Hobson's Bay," by F. Chapman, A.L.S. 4. "Four new Echinoids from the Australian Tertiary," by T. S. Hall, M.A. The series of lectures delivered during the year was a great success and large numbers of members and their friends attended. During the year four Members and six Associates have been elected, two Associates have resigned, while one memV)er and three Associates have died. Mr. H. Moors, for many years member of Council, Treasurer, and more lately Auditor, of the Society, passed away at an ad- vanced age, regretted by a large number of friends to whom his kindly nature had endeared him. The three Associates whose deaths we have to mourn are Messrs. F. J. Odling, C. Stewart, C.E., and G. J. Bolton, M.A. The Proceedings of the Society, Vol. XVIIL, Part 2, and Vol. XIX., Part 1, were published during the year, and owing to the. low state of our finances in a most attenuated form. Owing to the generosity of the Treasurer of the State our grant has been increased to £100, and we hope to increase our output of printed matter. The Librarian reports 1454 additions during the year, which is the largest number yet recorded. The sum of £25 has been voted by the Committee for Vjinding. 276 Froceedinys of the Royal Society of Victoria. ^ ^ |3t I— I 1— ( ao o to o cc rH lO X -f( rt o 71 :o i> 05 O ' '-H C-l c-l o O 05 O c. -^ •-1 t^ 30 1—1 «o O «3 1—1 >— 1 CD = ce : • •o • ■ fl -3 ■ 2 - a; • rt to to '^ -^i ^H O to JS 02 Q o bx) — I oj ri oi g r^ TO to f^ >H g^ O' S -^ .2 * 2 -3 £ * t; S S; « o o .-I o O O O O O O M 00 -rfi CO a; oi O 1-1 fcO >o J^ CC CC 1— I >» .2 g 0) o -(-^ s 2 S to o Members Country ; Associate Arrears .. Advance c o M en o 3 s CM O >^ c 0^ etf o 3 v ' fT^ WCiJOJ « CQ n P4 o H^^iH .^ s i e w > en CO u e-i S O as o M ^> M » ii^ §.opl c^odctn af ^ictaria. 1907. JJ.iUon : SIR KKCJiNALI) TALP.oT, K.C.B. JJrceibrnt : C. E. ()l,I\F.i;, M.C.K. X)iCC-iJlC6itlCUtS : p. BAKACCHI, F.R.A.S. E. J. WHITE, F.R.A.S. ^Sjoii 'iEvias.unr : J. SHF,PHAR1>. ^jou. ILibi.ui.in : R. I>. BOYS, B.A. lljou. ^fcvctaru : T. S. HALL, M.A. 3/\ (£ouiu'il Pl!(i|\ R. J. A. BERRY, M.K., F.R.S.E. J. r)F:XNANT, F.G.S., F.C.S. PP>OF. A. .1. EWART, D.Si-., Pll.l). P. i.K .1. (iRlT. 1>R. .1. .lA.MIESOX, M.D. ,1. A. KERSHAW, F.E.S. PROF. T. i;. I,Y1,E, M.A. PRoK. E. W. 8KEATS, D.Si. E. .(. DIXX, F.G.S. I'RUF. W. C. KERNOT, M.A., M.C.E. I'ROF. W. BALDWIN SPENCER, C.M.G., M.A., F.R.S. G. SWEET, F.G.S. R. H. WAIX'OTT, F.G.S. (Eommitti'fs of tljf (Houiuil : %nusc liloiumittcc THK HON. TKKASL'KKK (Convknkk). P. DK J. GRUT. G SWEET, F.G..S. ^Jiiutiiig *i"iimiiuttcc : THE HON. TREASUREK. THE HON. SECRETARY ((Jonvknkr). PROF. W. BALDWIN SPENCER, C.M.G., M.A., F.R.S. ri.iiinr.ui) ^utiitovs : J. E. GILBERT. P. DE J. GRUT. JiSioiunniu ^3vrchitcct : AV. A. M. BLACK ETT. '(Trusters •• R. L. .). ELLERV, C..M.(i., F.R.S.. i(. E. J. WHITE, F.R.A.S. PROFESSOR W. C. KEUNOT, .\l..\., .M.C.E.. ic. 1907. LIST OF MEMBERS, WITH THEIR YEAR OF JOINING. Patron. His Excellency Sir Reginaild Talbot, K.C.B 1904 HoNORART Members. Forrest, The Hon. Sir J., K.C.M.G., West Australia ... 1888 Hector, Sir James, K.C.M.G., M.D., F.K.S., Wellington, 1888 N.Z. Liversidge, Professor A., LL.D., F.R.S., University, 1892 Sydney, N.S.W. Neumayer, Prof. George, Ph.D., F.R.S., Neustadt a.d. 1857 Haardt, Germany Russell, H. C., B.A., F.R.S.. F.K.A.S., Observatory. 1888 Sydney, N.S.W. Scott, Rev. W., M.A., Kunajong Heights, N.S.W 1855 Todd, Sir Charles, K.C.M.G., F.R.S., Adelaide, S.A 1856 Verbeek, Dr. R. D. M., Buitenwrg, Batavia. Java 1886 Life Memhers. Butters, J. S., F.R.G.S., Empire Buildings, Comns-sireet 18C0 west Eaton, H. F 1857 Fowler, Thos. W., M.C.E., Colonial Mutual Ch., 421 Col- 1879 lins-street, Melbomme 280 Lif^t of Members. Gibbons, Sydney, F.CS., 31 Gipps-street, East Mel- 1851 bourne. Gill)L'rt. J. E., "Melrose," Glenferrie-road, Kew 1872 Love, E. F. J., M.A., F.R.A.S., 213 Victoria Terrace, 1888 Royal Park Nicholas, Williajm, F.G.S 1864 Kusden, H. K., Glenhuntly-road, Eisternwick 1866 Selby, G. W., 99 Queen-street, Melbourne 1881 White, E. J., F.R.A.S., Observatory, Melbourne 1868 Ordinary Members. Balfour, Lewis, B.A., M.B., B.8., Burwood-road, Hawthorn 1892 Baracchi, Pietro, F.R.A.S., Observatory, Melbourne 1887 Barnes, Benjamin, Queen's Terrace, South Melbourne ... 1866 Bavay, A. F. J. de, Foster Brewery, Collingwood 1905 Boese, C. H. E 1895 Boys, "R. D., B.A., Public Library, Melbourne 1903 Berry, Prof. K. J. A., M.D., F.R.S.E., University ... 1906 Cherry, T., M.D., M.S., Department of Agriculture, Mel- 1893 bourne Cohen, Joseph B., A.R.I. B.A., Public Works Department, 1877 Melbourne Dciinant, John, F.G.S. , F.C.S., Stanhope-grove, Camber- 1886 well Dunn, E. J., F.G.S., " Rosoneath," Pakington-street, Kew 1893 EUery, R. L. J., C.M.G., F.R.S., F.K.A.S., Observatory, 1S56 Melbourne Ewart, Prof. A. J., D.Sc, F.L.S., University, Melbourne 1906 Fox, Dr. W. R., L.R.C.S., L.R.C.P., 90 (V.llins-stroet 1899 Mclhoiiriic Fryett, A. G., care Dr. F. Bird. Spring-street, Melbourne 1900 (lault. Dr. E. L., M.A., M.B., B.S.. Collins-street, Mel- 1899 bourne Gillott, The Hon. Sir S., K.C.M.G., " Edonsor," lirun.s- 190;") wick-street, Fitzroy Grut, P. de Jersey, 125 Osborne-street, South Yarra ... 1901 LiM <>J M<-,i>h('r^. 281 Hake, C. N., F.C.S., Mellmurue (Uul), Melbourne 1890 Hall, T. 8., M.A., University, Melbourne 1890 Hartnell, W. A., " Irrewjwra,' Burke-road Camberwell ... 1900 Harvey, J. II . A.K.V.I.A.. 12S IN.wlett-street, East Mel- 1895 l)ournc lleftVrnan. Iv U., M.D., U.S.. 10 Hruuswirk-street, Fitzroy 1871) Hoo[)er, Dr. J. W. Dunbar, L.H.C.S., etc., Collins-street, 1904 Melbourne Janiieson, James, M.D., 9() E.\liibitiiin-street, Melbnurne 1877 Kernot. Professor W. C, M.A., M.C.E., University, Mel- 1870 bourne Kernot, W. N., B.C.E., Workino- Men's College, Melbourne 19UG Kershaw, J. A., F.E.S., National Museum, Swanston-street 1900 Kitson, A. E.. F.GS 189-1 Lvle, Prof. T. R.. M.A., Universitv, Melbourne 1889 Li.ughrey, B.. M.A.. M.D., Ck.B.,'M.C.E.. 1 Elgin-street, 1880 Hawthorn Masson, Prof. Orme, M.A., D.Sc, Univer.'antula, 35 Scyoiiiaoims kiutji, 15:? 'I'iinarcta ifi'.hterranca, li'i'.l Si-yiiuni.s r(iriic Shells, iiiarine, IJl Triffonella oriiithopodoides, 87 Siphoiiaria stowa.e, 35 'I'l-igi^nothops vittipi'.miii, 14') Skeats, E. W., Hi) Tio^odciiiia lilackhurni, 155 Soil constituents, 39 Tutis, iiedded, 21 Soil, seasonal variations, 38 Tunica prolifera, 13!) Spoi-obolus Bonthami, 13S velutina, I3i) viri«inicus, 138 'I'ysonia pliyllostci^'ia, 85 Studeria cicoans, 214 Verticordia' I'ritzrdli, 85 Stylifci'ina translucida, 31 White, J., 1()7 Sweet, ii., 222 WUnirdia arrohiruhila, l!)(J Switihurn'Kt phyll"stegia, 85 Wood, l^ed, 1<>7 END OF VOFATME XX [Part It. Issued March, 1908.] PRINTKD BY Konn Asn SOS, MKT.liOI'liNK. -^t A^^ UJ LIBRA R PROCEEDINGS OF THE fapl ^amti) of ^irtonH, VOL. XX. (New Series). PART I. Edited under the Authority of the Council. ISSUED AUGUST, 1907. (Containing Papers read be/ore the Society during' the months of April, June, Jidy, igoy). THK AUTHORS OK THK SEVERAL PAPERS ARE SEVERALLY RESPONSIBLE FOR THK 80ri«DNE68 OK THE OPIKIONS GIVEN AND FOR THE ACCl'RACY OF THE STATEMENTS MADE THEREIN. MELBOUKNE : FOKD L SON, PKINTERS, DRUMMOND STREET, CARLTON. AGESTS TO THE SOCIETY: WILLIAMS & NORGATE. 14 HENRIETTA STREET, COVENT GARDEN, LONDON. To whom all coumiunications for transmission to the Royal Society of Victoria, from all parts of Europe, should be sent. 1907. Publications of tlie Royal Society of Victoria, and of the Societies amalgamated with it. Victorian Institute for the Advancement op Science. Transactions. Vol. 1. 1855. PiiiLosoPincAL Society of Victoria. Transactions. Vol. 1. 1855. These two Societies ilien ainalgamaled and became: — Philosophical Institute of Victoria. Transactiuns. Vols. 1-4. The Society then became : — Royal Society of Victoria. Transactions and Proceedings (V^ol. 5, entitled Transac- tions). Vols. 5-2-1:. Transactions. Vols. 1, 2, 3 (part one only), 4. 1888-95. Proceedings (New Series). Vols. 1 . 1888 . Microscopical Society of Victoria. Journal (Vol. 1, Pt. 1, e'w/zV/df^/ Quarterly Journal). Vol. 1 (Pts. 1 to 4), 2 (Pt. 1), title page and index {all published]. 1879-82. {The Society then combined unth the Royal Society of Victoria]. Note. — Most of the volumes published before i8go are out of print. PROCEEDINGS ioDHi .^arietD af §ixtonH. VOL. XX. (Nkw Skkies). A\: PAHT II. Edited tinder the Aiitliority of the Cimncil. ISSUED MARCH. igoS Qd :^ -^•-•k. .«\< ers rend he/ote tite Si'ciety during the tnuiiths uj October, November, December, igoy). IMK AtrilOHS UK IIIK SKVBRAI. I'APRHS AKK 8KVKKALLY KRBruNHI RLK KOK I UK HOUNDNESS OK TKK OPINIONS OIVKN AND FOR TIIR ACC'IHACV OF TIIK 8TATK.MRNTS MADK TMKKKIN. MELBOUKNE . KURD & SOiN, PKINTEKS, DKUMJMOND SIKEKI'. CAKL'l'ON. AGKsrs TO run sociE'iy : WILLIAMS i NORGATE, 14 HENKIETTA STREET, COVEN'T GARDEN, LONDON. To whom all coniinunications for tiansniissioii to the Royal Society of Victoria. from all parts of Europe, should be sent. 1908. MHI WHOI I Ihrnry 5 WHSE 00604