JOURNAL OF THE ROYAL SOCIETY OF WESTERN AUSTRALIA VOLUME 57 PART 1 MAY, 1974 PRICE: TWO DOLLARS REGISTERED FOR POSTING AS A PERIODICAL-CATEGORY B THE President Vice-Presidents Past President Joint Hon. Secretaries Hon. Treasurer Hon. Librarian Hon. Editor ROYAL SOCIETY OF WESTERN AUSTRALIA PATRON Her Majesty the Queen COUNCIL 1973-1974 A. F. Trendall, B.Sc., Ph.D., A.R.C.S., F.G.S. G. A. Bottomiey, B.Sc., Ph.D. B. E. Balme, D.Sc. R. M. Berndt, M.A., Dip.Anth., Ph.D., F.R.A.I., F.F.A.A.A. M. Perry, B.Sc. (Agric.) (Hons.) G. Perry, B.Sc. (Hons.) S. J. Curry, M.A. A. Neumann, B.A. A. J. McComb, M.Sc., Ph.D. S. J. Hallam, M.A. L. J. Peet, B.Sc., F.G.S. P. G. Quilty, B.Sc. (Hons.), Ph.D. J. A. Springett, B.Sc., Ph.D. G. M. Storr, B.Sc., Ph.D. J. C. Taylor, B.Sc., Ph.D., A.R.C.S. P. G. Wilson. M.Sc. P. R. Wycherley, O.B.E., B.Sc., Ph.D., F.L.S. 1. — Australian Aborigines : Research and Welfare Presidential Address, 1973 by Ronald M. Berndt i * 1 Delivered 16 July, 1973 Abstract Reference to an article in the first issue of this Journal points up changes in Aboriginal living conditions and in Australian society generally, and also in research methods. The main features of traditional Aboriginal society are outlined. They influenced the Aborigines’ reactions to the European invaders as well as vice versa. Only when these features were modified was rapprochement possible, but it was a one-way process. Today, despite appearances to the contrary, the Aborigines are actually more dependent on others than ever before, and welfare policies and practice take even less account of solid research findings. Moreover, the new Aboriginal identity is being shaped in a context where the traditional past is a source of inspiration but provides few guide lines for the future. I A presidential address is not intended simply to illuminate a particular topic from the angle of a particular academic discipline. It should also say something about the contribution of the Society (in this case, the Royal Society of W.A.) in relation to that topic. It is interesting therefore to note that when this Society emerged (in 1914) from the Natural History and Science Society, under the guiding hand of Professor W. J. Dakin, Vol. I of its Journal and Proceed- ings for 1914-15 contained an article on Sunday Island by W. D. Campbell and W. H. Bird <1916: 55-82). Bird was a teacher at the mission settle- ment which had been established among the Bard people in 1899. It is not a professional anthropological contribution, but it points up obliquely some of the things I shall be talking about. For instance, it underlines the tremen- dous changes in research methods that have taken place over the years. It reflects the con- trast between then and now in another sense too. The small, quiet settlement at Sunday Island was run primarily on private funds, with a government allowance of blankets and nine- pence per day for the aged and infirm. The State Education Department made a grant of £100 a year to the mission school. In that span of almost sixty years, it is as if another world has been superimposed on the old, but in such a way that the old one has not been entirely eradicated — at least, not yet. In general, those years have seen much fluctuation both in policies and in practice. They have also seen the movement of Aborigines from being an inarticulate, almost invisible minority, to a vociferous, highly visible and expanding popu- lation. One focus of attention in this State, both then and now, has been the economic activities i Presidential address delivered July 16th, 1973. of the Bard. As Campbell and Bird noted, these people were concerned with marine products. Today, the Ecology Unit under Commonwealth financial support and guidance is endeavouring to establish a turtle-farming venture, because the Bard are heavily dependent on outside help. But in 1914 and before, they were economically viable, with their pearlshell and beche-de-mer fishing. The theme of outside concern is the same, but the local circumstances are not. II If the gap between the early Bard example and the contemporary situation is so consider- able, the gap between traditional Aboriginal life and what survives now, not only among the Bard but also more generally, is even wider. We could almost say that it was a world apart from Aboriginal life as it exists today. Aside from the question of attitudes on the part of early European settlers toward non- Europeans in general, attitudes which are fairly well documented, the immense difference in life- styles between the newcomers and the Aborigines made any real rapprochement between them very difficult indeed, if not actually impossible. These difficulties became increasingly evident soon after initial contact, as both peoples be- came more conscious of pervasive incompatibili- ties. They have been modified only through an ironing-out or blurring of traditional Aboriginal elements. For a long time, these differences were seen as a contrast between “civilized” and “uncivilized”, between “sophisticated” and “primitive” man. But this was a biased and faulty interpretation, one that could not stand up to closer scrutiny. Traditional Aboriginal societies and cultures were highly complex. Their members were ordinary, intelligent human beings, guided by their own belief systems and their accepted behavioural patterns. The organ- ization of their societies was different from ours, and so were their values: their aims were not the same. The positive qualities of their way of life were not immediately obvious to outsiders who were used to a very dissimilar social and economic setting, with its stress on material goods and its thing-oriented tech- nology. The Aborigines were on the whole a deeply religious people. Religious feeling was mani- fested through ritual observance and through mythic expression: it was really something that was taken for granted. In essence, it was based on a deep and emotional attachment to the land. That land was to them full of signs, which had a direct relevance to socio-economic living. Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 1 It was a land made familiar and intimate to them through mythic beings who were believed to be manifested at specific sites, beings who were believed to be always present and to be approachable through the medium of ritual. These were eternal or enduring elements, under- lining continuity sustained through religious practice. They emphasized the fundamental unity of the human and the natural-physical environment, a kind of empathy between man and all natural species and phenomena. Every- thing was, at one conceptual level, brought to- gether into one expanded socio-cultural environ- ment, so that man was not seen as opposed to nature but as working in harmony with it. This was expressed through the concept of the Dreaming, as it is sometimes called in transla- tion, in which all the most significant aspects of life were believed to have been set in motion by the primary mythic and spirit beings. The deities were manifested through man, and other living things and other features were selected as intermediaries or symbolic representations: all drew on the same life force. This particular relationship to the land and all within it was phrased as a total dependence upon it: and the way this was wrapped up in socio-cultural terms provided an emotional assurance that helped people to cope with such natural crises as drought or floods, and the human crisis of physical death. It was a screen between them and the stark reality of such crises, helping them to feel that their survival did not rest solely on their skills, their few weapons and techniques: the mythic beings stood protectively between them and any potential disaster, as a buffer or an intermediary, a source of confidence as well as a source of traditional guides to prac- tical procedures. But living and making a living involved also relations between people. In short-range terms, any Aboriginal man and his immediate family could live off the land quite capably under normal seasonal conditions. However, the fuller round of events called for a wider range of people. This meant that the range of depend- ence was extended, highlighting the issues of reciprocity and responsibility for others that were the basis of group existence. The large kin-oriented structures which were usual in Aboriginal Australia had a specifically utilitarian (or socio-economic) value. They represented a buffer of a different sort, a non-mythic or non- ritual buffer, between every Aboriginal person and the demands of his physical environment. Co-operation with others was an essential part of Aboriginal semi-nomadic living. Kinship net- works linked people together, in combinations that sometimes took the form of conflict but more often emphasized collaboration and mutual help. Within a certain regional span, a person could always be sure of having relatives who could more or less be relied on to take his part — not necessarily the same people in all circumstances; but there were always some to defend or support him. The genius of traditionally-oriented Aborigines rested primarily on their ability to organize, providing an assured though reasonably flexible programme for co-operation in everyday affairs, and a religious belief system which substantiated a life within surroundings that were familiar but full of interest and meaning. It is a mistake to believe that this life was unduly monotonous or consisted of repetitive action within a circum- scribed and limited frame of belief. It is true that traditional Aboriginal life was cast within the mould of the past; what had been proven then, or believed to have been proven, was considered to have a direct bearing on the present: the lessons learnt from the past could be applied effectively to the present and to the future, if for no other reason than that solutions to specific problems of living within the Australian environment could not be varied radically — not without risk. Outside the dynamics of social living — in, for example, domestic relations, marriage, children growing up and being initiated, confronting the inevitability of death — outside of these, religious rituals were essentially concerned with renewal, with spiritually stimulating environmental fer- tility, and with sharpening intellectual faculties. In that respect, ritual provided the main stimulus to enquiry and speculation. Such en- quiry took place within what can be called a closed system, but it was not entirely straight- jacketed. It nurtured and enriched the Abor- iginal arts — music, song-poetry and oral litera- ture generally, dancing, painting and sculpture. Evaluating a society or a culture solely in terms of what people do to gain a livelihood provides only a one-sided appraisal. In all human societies there are particular imperatives which cannot be evaded if survival is to be ensured. However, what people do outside that sphere of necessity is especially significant. As far as the Aborigines are concerned, the great mythic epics and song cycles demonstrate beyond doubt a high level of cultural attainment: they thought and felt and expressed themselves poetically in ways which were mediated not only through religion but also through ordinary living. I am reminded of what Strehlow (1971: 247) has pointed out, and others too: that anyone con- versing with fully-initiated Aboriginal men “trained in speech by means of the sacred myths and songs” cannot fail to be aware that he is in the presence of men of education and culture. Against this picture of relative harmony and environmental adjustment and intellectual de- velopment, there are inevitably many examples of human fallibility. Life could be harsh, social relations were fraught with difficulties, and the ordinary course of living was punctuated by interpersonal dissension. It was not a utopian existence. It does seem that the basic needs of Aboriginal man were reasonably satisfied, that people were able to achieve a fair degree of happiness and comfort. But limitations were imposed. Even if they were not directly recognized, they were definitely present. Inde- pendence — individual independence — was played down, or undervalued, because group co-opera- tion was an economic necessity; and speculation Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 2 and experiment could go only so far, because the status quo depended on common expecta- tions in belief and in action. In balance, how- ever, there were more positive attributes: values related to a perceived affinity with the land, harmony with nature, co-operativeness, a love of beauty and aesthetic appreciation, a respect for the past which provided an assurance for the future, and recognition of personal rights viewed as affecting others — recognition that a person is responsible to others as those others are respon- sible to him. These values were either ignored or not appreciated by the European newcomers. Ill The traditional Aboriginal aspects which I have noted in summary must be taken into account if we are to appreciate the results of alien impact. They have to be understood, too, in relation to the creation of a social identity for Aborigines in today’s scene. It was this traditional world of the Aborigines which, in so many instances, received a death- blow when it came into contact with outsiders. In the southern and south-eastern areas, where European settlement expanded rapidly, it meant the complete destruction of the Aboriginal way of life and, in a number of cases, of the people themselves as well. That history is so much a part of our own that there is no need for me to sketch it out in any detail. Three points only need be mentioned. One: there was, as soon as Aborigines realized that the aliens had come to stay, a great deal of opposition to this intrusion, especially when their lands were appropriated without reference to them and their food resources were threatened. Literally, many were obliged to fight for their very existence. But they were ill- equipped for that purpose and lacked one of the fundamental pre-requisites — that is, political strategy and an overarching institution which would have enabled them to muster a large opposition force. I said that their genius rested on organization — but not on such a large scale, and not for collective aggressive acts of offence and defence. They depended instead on skirmish and on ruse, on guerilla warfare. What occurred in the Swan River Colony between 1830-1840 bears this out, as it does in other areas. The “Battle of Pinjarra”, as it has been called, was not a battle at all. The Aborigines concerned, including women and children, were not ready to fight. It was an ambush on the part of the Europeans, and there was little opportunity for any Aborigines to escape through the cross-fire of the two parties. Such instances were dupli- cated, in one way or another, so that capitula- tion was only a matter of time and was the only course open to Aborigines — not in the spirit of “if you’re being licked by them, join them”, but simply because no other alternative existed. In the long run, those who did survive “joined them” (that is, the Europeans). This brings me to my second point. To “join them” meant that Aboriginal tradtional life had to be considerably modified. For instance, by the 1880’s, in the south-west of this State, that traditional life had disappeared as a living, functional reality. In the process, the indigen- ous population was almost entirely replaced by a part-Aboriginal population — a few of them directly descended from the original local people, but most of considerably mixed Aboriginal affinity. Again, that situation was duplicated in New South Wales, Victoria, South Australia, and so on — except, of course, in Tasmania: the story there is not all that different, only more extreme. The third point relates to the unevenness of alien contact. Not all Aborigines were affected in this way. But eventually, as time went on, all were exposed in varying degrees to external pressures, and their reactions to these differed. However, because of this unevenness, much of traditional life survived in some areas — at first in its traditional form, but increasingly with considerable modifications. The last great socio-cultural reservoirs, so to speak, have been Arnhem Land and the Western Desert — or they were, until just after the second world war. Wherever contact with outsiders took place, and irrespective of policies promulgated or put into action — whether on government or mission settlements, on pastoral stations, in country or “fringe” towns or in cities — wherever such con- tact took place, the theme of “civilizing” was emphasized: and to Europeans, “civilization” meant “Europeanization”. Accordingly, Abor- igines of all kinds were persuaded, directly or indirectly, to become more Europeanized, and, what was most important in this process, to learn new work patterns and adopt a new economic pattern of living. Although welfare policies varied over the years, from advocating separate development or from “smoothing the dying pillow” to inevitable assimilation, all in- volved increasing European control and influ- ence — most of which ignored or discredited the importance of Aboriginal life or what remained of it. The Aborigines, so it was said, had little to offer. Europeans, on the other hand, had everything to give — but at a price. This state of imbalance was actively encouraged, and coloured virtually all of Aboriginal-European interaction. It developed a pronounced state of dependence. It also had repercussions which have extended into the present-day scene. What emerged from this — with only a few, very few, exceptions — was that the Aborigines were re- duced to an almost invisible, almost inaudible, segment of the Australian population. The radical diminishing of their independence, the removal of land from their control, the down- grading of religious belief: all of these, along with others, led inevitably to their socio-cultural impoverishment. In the “outside world”, being of Aboriginal descent had no positive value at all, only a negative one. The hard road toward a “new” culture with its new social implications was strewn with obstacles, tangible and intangible, which most Aborigines were unable to overcome. There is no need for me to spell this out, and in relation to specific local groups the space- time component varied considerably. For in- stance, although this state of affairs existed Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 3 almost from the onset of alien contact, it is still observable today in a number of country towns in Western Australia, among other places. Opportunities for breaking this vicious circle are much greater now, but the process is still a traumatic and difficult one for the people concerned. The extension of Australian citizen- ship to Aborigines came only gradually, and for a long time meant very little to them. Special regulations affecting them were originally de- signed to protect and safeguard them, as a people in special need of protection and guid- ance, but too often they became almost ends in themselves. Welfare and advancement ideals became bogged down in a welter of prohibitions. And it was only too clear that protective policies were not there solely to protect Aborigines; they were also designed, or used, to protect non- Aboriginal interests. I won’t speak of exploita- tion in this respect, except to say that it was not only present but was also actively encour- aged in a number of instances — especially in some pastoral areas of the Kimberleys and the Northern Territory, as well as elsewhere. The problems vis-a-vis Aborigines, resulting on one hand from external contact and on the other from their own attempts to sustain rapidly changing traditional patterns — these problems ramified and could not be resolved without drastic action which, in turn, had further reper- cussions. To mention only three of these: (a) economic deprivation; (b) restricted educational opportunities and restricted opportunities for acquisition of basic skills; and (c) the eroding influence of drinking to excess. These issues alone were sufficient to define the Aborigines’ position within the wider community; (a) and (b) were complementary, one upholding and reinforcing the other. Low socio-economic status confined groups of Aborigines to particular urban settings, or forced them to the fringes of country towns. This set up or augmented social barriers which already existed in other forms, and which only a few of them were able to cross. The same was the case in the north. For example, on pastoral stations the Aboriginal camps were spatially separate from the areas in which Europeans resided. On government and mission reserves, the same patterns were visible. People living in such conditions were caught in a trap of increasing — conspicuously increasing — poverty and squalor. The only Aborigines who escaped were those who still remained tradi- tionally-oriented. The problems of housing which in recent years have received so much publicity as a primary symptom of Aboriginal deprivation, were and are only part of this wider syndrome. Lack of communication between Aborigines and other Australians was, and is, perhaps much more significant. The school was for a long time, and still is in many cases, an outstanding example of minimal communication, and of puzzlement on the part of educational authori- ties as to how to remedy that situation. The remedy, of course, did not lie in tackling only one aspect and leaving the rest in a kind of social vacuum, as was often the case in the immediate past. The approach had to be — should have been — in terms of the total con- figuration: but this has only recently been possible, and then only up to a point. Among other things, the continuing influence of drinking to excess, which over the years of contact has become virtually endemic — or, to put it in another way, has become patterned behaviour linked to particular positive values that were or are regarded by many Aborigines as being desirable and part of an expected way of life — that state of affairs has become, as more opportunities are offered to people of Aboriginal descent, an inhibiting or retarding factor. The “right to drink’’, which so many of us supported in the 1950’s (and probably would still support on the platform of equality), can, in perspective, be viewed as one of the most obvious ingredients in social and moral de- terioration. The present situation at Kalgoorlie, at Wiluna, or in Derby, Wyndham, and Alice Springs, for example, underlines that point. Even more disastrous is the situation at Gove in north-eastern Arnhem Land, near the new town of Nhulunbhuy, or at Oenpelli in western Arnhem Land. The solution does not lie in “teaching Aborigines to drink” or in gaoling Aborigines for drunkenness, but in education and rehabilitation. And, of course, it cannot be isolated from the picture of what is happening in the wider Australian scene — the patterns of expected and actual behaviour among Austra- lians in general. This issue has not yet been seriously tackled by State or Commonwealth authorities. It is tempting, at this juncture, to comment on Aborigines and the law. I shall not do so, except to say that legal representation is by no means all that is required. It is true that negative discrimination against Aborigines is apparent in this sphere, even though virtually all prejudicial legislation in this respect has been lifted. The problems involved are not simply within the courts; they are to be found em- bedded in social situations, the informal, human situations in which people of Aboriginal descent are involved. Many of the conditions I have mentioned stem at least to some extent from the past: but they have their repercussions in the present, and influence future trends. Changing them in a positive sense means re-programming or re- directing the course of events. And to know what to do in this respect requires, initially, research. This is or should be a significant component in all welfare developmental pro- grammes, but the need for it is being recognized far too slowly. In any re-programming, account must be taken of what can be called “the Aboriginal heritage”. Broadly, it has two inter- related facets. One concerns the “traditional, specifically Aboriginal heritage”. This is relevant in different ways to all Aborigines, whether they remain traditionally-oriented (as some still are) or are to all intents and purposes ordinary Aus- tralians (that is, culturally speaking). Secondly, there is the traumatic history of past and near- present contact. Ideas about this are com- municable to on-coming generations. I mean, Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 4 here, views and feelings related to being an Aboriginal, being defined as an Aboriginal per- son, by oneself or by others, in a society domin- ated by non-Aborigines: restrictions on access to potential advantages, restricted opportunities; and discriminatory attitudes and behaviour. In combination, the two facets of the Aboriginal heritage provide a formidable emotional frame against which to measure virtually the total range of experience, from adaptation and ac- ceptance to opposition and rejection. IV The break with older, negative policies and attitudes did not really get under way until just after the second world war. The main motivators were not, at first, Aborigines themselves. But these motivators did include anthropologists. The nature of their work, which involved intimate and sustained relationships with Aborigines, placed them in a strategic position. They were, simultaneously, both preservational- ists and activists, concerned with recording and understanding socio-cultural life in its tradi- tional perspective as well as under conditions of extreme change; and they were also con- cerned with the implications of what they observed in relation to human aspirations, human dignity, and human satisfactions. The first effective and systematic anthro- pological field research did not take place until the late 1920’s and early 1930’s. At that time, when Aboriginal opinion was largely inarticu- late, anthropologists (and there were only a few of them) served as intermediaries between Aborigines and administrations, as well as mis- sionaries. At that time, they were almost the only people other than Aborigines themselves who had any real knowledge of what was hap- pening in Aboriginal areas and what Aborigines felt about it. I do not, of course, want to under- estimate the influence of welfare agencies. What I am saying refers to social-scientifically- informed knowledge. Without anthropologists (plus a very few missionaries and others), virtually no information about Aboriginal life in the immediate past would be available today — and not just in reference to traditional life. Without them, our understanding of present- day conditions would be considerably limited. And, as you will realize, the collection of such material has a direct bearing on contemporary ideas about social identity. This aside, anthro- pologists have influenced administrative policy at all levels. Moreover, they were instrumental in achieving an almost complete reversal of the older, outmoded policies, as well as helping to turn public opinion toward a more positive appreciation of Aborigines and their culture. This is no exaggeration : it is a matter of history, as yet unwritten. However, this development must be seen in context, and in relation to current socio-cultural trends within the wider Australian society. The first major break-through occurred with the establishment, during the last war, of army settlements in the Northern Territory. There, Aborigines came into close contact with a variety of different kinds of Australians — and not just administrators and missionaries, station-man- agers, stockmen, and so on. Conditions on those settlements were in marked contrast to what existed in their “home” areas. A wage economy was introduced, whereas on many pastoral sta- tions at that time no such payments were available. Army settlements provided housing, showers, latrines, beds and other amenities: on the stations, humpies and huts, and “native camps” were the norm, usually relegated to the local creek bed or some such site. Additionally, the local Aborigines were viewed by the owners and managers of many stations as part of the natural environment, which was there for them to exploit. Movement outside the confines of such stations and other settlements brought growing awareness among Aborigines themselves of barriers block- ing social and spatial mobility where they were concerned. This same upsurge of interest was apparent also in the towns and in the cities, where persons of Aboriginal descent were be- coming more vocal, and as a result more visible. A movement had begun which increased in momentum, and brought in its train radical policy changes. A number of years were to elapse before the majority of Aborigines felt its impact. However, in contrast to what had been going on before, those changes were very rapid indeed; and eventually, with varying de- grees of effectiveness, they succeeded in partially rechannelling the course of events. It was in this new climate of opinion and unrest that the ideal of assimilation was forged as far as Australia was concerned, although of course it had been suggested before. Its original premise rested on non-discrimination and on equal opportunities within the wider Australian society for all people of Aboriginal descent. It assumed that traditional Aboriginal life would become a thing of the past, and that socio- economic viability could be achieved through some effort on the part of all those involved and through common consensus. However, the history of past contact militated against that ideal, at least for a substantial part of the Aboriginal population. Contrary forces were at work; and social protest, which previously had been regionally confined and easily dissipated, crystallized and took on wider political signifi- cance. At the same time, the Aboriginal popula- tion explosion became much less localized than it had seemed to be at first. Spatial mobility increased, there was a stepping-up of educa- tional programmes, and, most importantly, the gradual shaping of a new identity. Recognizing that attitudes and aspirations were changing in these directions, the assimilation aim was officially modified in 1965 to permit a greater emphasis to be placed on traditional Aboriginal culture. From that point in time, there was no turning back. At the administrative level, further far- reaching changes took place — but not without the prodding of social protest (see R. Berndt 1971: 25-43). Social Service benefits for all Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 5 Aborigines, and the right to vote and to drink intoxicating liquor, were introduced unevenly among the States. The Pastoral Award, designed to come into effect in the Northern Territory by 1968, had to be moved forward to 1966 be- cause of the Guirindji (Wave Hill) strike. In 1967, as you all know, a Referendum was held which concerned, on one hand, the inclusion of all Aborigines in the Commonwealth Census and, on the other, the assignment of powers to the Commonwealth to enable it to legislate for Aborigines. As it was framed then, in its two ambiguous questions on which the Australian public went to the poll, the implications were not clear; they have become much clearer over the years. Policy became more realistic and more reflec- tive of what was happening in the various Aboriginal communities themselves. It was also more receptive to the demands of protest. Of these, two outstanding cases of recent years — the Guirindji sequence, and the Gove Land Rights dispute — had the greatest public and political repercussions. The first, in the Wave Hill pastoral area, concerned employment and independence, coupled with land rights. The other, at Gove, in north-eastern Arnhem Land, was more far-reaching in its significance. It was a direct attack on mining exploitation in that area, which was regarded as taking place at the expense of local Aborigines; and in the litigation which followed, the Aborigines con- fronted the combined opposition of the Nabalco mining complex and the Commonwealth govern- ment. The Aborigines sought to establish owner- ship of “tribal” lands within the context of Australian law. We all know about the negative judgement in this long-drawn-out case. One of the first tasks of the present Federal government was to establish an Aboriginal Land Rights Commission — not to debate the legal question of whether or not land rights should be given, but to determine how they should be given, and to whom (that is, to what groups of Aborigines). External intervention and stimulus were ap- parent in both of these instances — in bringing the issues to a head, in sustaining public atten- tion, and in instigating political action. Further, the issues were raised at a particularly oppor- tune time, when the public was receptive, and when the “Aboriginal cause” was considered to be worth taking up. Aborigines had become good politics, and had attained respectability. But this creation of a congenial atmosphere, with its encouraging possibilities for better conditions, had been preceded by a multitude of processions, sit-ins, student involvement and other forms of protest. A lot of hard work in these and other directions had already gone into it, on the part of Aborigines and non-Aborigines. The Can- berra “Embassy” was a highlight of this series, which effectively, in its repercussions, dissipated any hard, overt resistance to Aboriginal rights writ large. It was followed by other manifesta- tions, such as the North Adelaide tent and the Western Australian Parliament Stone (which, incidentally, remains unresolved). By this time, the public had “got the message”, and people of Aboriginal descent had firmed up their aspira- tions. However, as I have said, many of these protests were taken up or actively encouraged by non-Aborigines — for valid reasons, because Aborigines were, collectively speaking, disadvan- taged. They were also taken up for political and other reasons, and this has been quite ap- parent in a number of instances. A measure of the importance of Aboriginal affairs today is what could well be called the “bandwagon approach”. A large number of people, drawn from various academic disciplines as well as from the non-professional public, became involved. What most had in common was an ignorance of Aborigines and Aboriginal life, as well as of the problems of change facing these people. It stimulated consultant firms to mount government-sponsored surveys, which meant big money for their personnel. The truth of the matter was (and is) that many Aborigines were not always in a position to act for themselves, and this was specially the case in northern areas, though much less so in the south. This meant that they were, and often still are, vulnerable and subject to manipu- lation. That phase is gradually passing, or, rather, is being redirected into different hands. There is always a danger in this respect, for all peoples — but more so when a people like the Aborigines are concerned, a people who are struggling for equal rights and opportunities and for an effective voice in their own affairs. V The public was receptive. In one sense, we can speak of the Aborigines being re-discovered by other Australians (see C. Berndt 1969: 16-34). Often what was sought was a highly romantic picture of traditional Aboriginal life, an en- capsulation of exotic elements which could be translated and transformed by novelists, poets, artists, musicians, dancers, and so on. But that transformation, when it was made, had little resemblance to the reality of traditional life; and they used it, not so much to understand it, as to provide an extra dimension to their own work. Counterbalancing this trend were the hard- core anthropological and social scientific studies which were reasonably objective and had an entirely different aim. Research meant learning for a purpose, not just idle curiosity, and not solely for academic ends. In this way, anthro- pologists not only provided a detailed record of living traditional life, but explored all aspects of change wherever persons of Aboriginal des- cent were to be found — in the bush or in the city, to note only two contrasts. What they learned could be applied in relation to Aboriginal advancement. This research was appreciably stimulated by the establishment in 1961 of the Australian Institute of Aboriginal Studies, which sponsors a wide range of research. Then came the Commonwealth Department of Aboriginal Affairs (as it is now called), which is specifically concerned with welfare-oriented research. Ad- ditionally, there are the universities, and the Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 6 State departments of Aboriginal welfare and planning. It is within this frame of research that, increasingly, contemporary social issues are be- ing reviewed. There is no need for me to emphasize the significance of all this, except to repeat, what I seem to have said so often: that we cannot afford to neglect learning more about the society in which we live, and becoming better informed about the various forces that are at work within it. A “commonsense” ap- proach based on one’s own experience is inade- quate: it can lead, as it has done in the past, to more difficulties. And, as part of this broader scene, though only one part of it, an anthro- pological and social scientific approach to the understanding of problems facing persons of Aboriginal descent is essential if we seek their ultimate well-being. If other Australians have re-discovered the Aborigines, people of Aboriginal descent are now engaged in the process of re-discovering them- selves. This is not so much the case among tradi- tionally-oriented Aborigines, especially now that policy changes have provided them with an opportunity to sustain and maintain substantial areas of their own culture. The extent to which this will be possible is another matter, and it seems to be a highly selective business — par- ticularly when it is supported by, for example, the Australian Council for the Arts through its Advisory Committee on the Aboriginal Arts; or by official emphasis on being taught in and through their own vernacular languages, with only hazy ideas about the kind of content that this could entail as far as their traditional culture is concerned. How much of that culture can survive, and for what functional reasons, is a subject I shall not go into here — although it is crucial to this particular issue. Opportuni- ties do exist: but it is also true to say that what will survive will be radically different from what it was in the purely Aboriginal situation — and what there was before cannot be artificially resuscitated. The process of people of Aboriginal descent re-discovering themselves, is something else again. Out of a long history of dependence and subordination, paternalism and protection, mal- treatment and neglect, and even worse — out of all of that and more, has arisen a resentment which has become increasingly pronounced. This has resulted, as I have said, in protest — some of it mild, some of it aggressive. And it is within this context that the new image of Aboriginality is being formed. This has been projected on to the wider Australian community in two ways. One takes the form of demands that people of Aboriginal descent should be able to make decisions for themselves about their own affairs, and the corollary (insisted on by some of them) that nobody else should be allowed to do so. This has stimulated the emergence of Aboriginal spokesmen and leaders on the State and national levels. In this respect, the southern urbanized people of Aboriginal descent have had a con- siderable influence. The other, related to the first, is manifested in a concern for cultural preservation and Aboriginal revival. A wave of feeling for “Aboriginal” identity — which could eventually lead to pan-Aboriginality — seeks to establish a common socio-cultural heritage. It is the “idea” of traditional Aboriginal life which is used for this purpose — and not the reality of what was once traditional semi-nomadic exist- ence. This has been expressed through high- lighting Aboriginal religious features, especially in regard to secret-sacred material and sacred sites, “law-carriers” or “elders” and male authority in the ritual sphere. With this has come, or has been more clearly stated, justifica- tion for land ownership, and not necessarily in economic terms, but in terms of the spiritual and emotional linkages which were pivotal features of traditional life. It is, in fact, a pseudo-renaissance. It is important not to underestimate the significance of the movement toward Aboriginal identity, because this is used to define persons of Aboriginal descent in contrast to non- Aborigines. It may also be framed in terms of “moderates” versus “extremists” — Black Theatre as a medium of protest, versus Black Power; or, put simply, “Black” versus “White”. Views are hardening, as one might expect them to do — always bearing in mind the history of Aboriginal emergence as a political force in Australian society. That identity, whatever its outward manifestation, has political implica- tions; and Aborigines are well aware of these, as is the current Federal government. Within that picture is Aboriginal identity as a positive expression of a pride in being Aboriginal and in having a common background, however far that may be removed from the actualities of the past. That identity must be seen in a dual sense, as having something to do with the tradi- tional past and also something to do with the struggle for equality, against what appeared at times to be insurmountable odds. Such an identity can help to provide emotional security and a sense of belonging which, outside the traditional Aboriginal scene, has been sadly lacking. This is probably one of the most signi- ficant developments that have taken place over the years — much more important, in my view, than the upsurge of political awareness which, however, can be viewed as part of it. Contrasts between “Black” and “White” are becoming increasingly irrelevant and outmoded in this present-day world, in spite of numerous examples to the contrary. Such catchwords point to supposed physical characteristics and say little about mental ability and cultural attainment. Further, they point to political dis- crimination and to prejudice from either side of the ethnic fence. Such contrasts are not anthro- pologically sound, and they never were. As far as the Australian Aborigines are concerned, they are not “black”, even the darkest of them in the northern coastal regions. What does make good sense, anthropologically, is cultural diver- sity and the sustaining of particular heritages. The “new” Aborigines— and I use that term generally to refer to all those persons who are Journal pf the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 7 of Aboriginal descent or who identify themselves in that way — the new Aborigines are in a posi- tion, and have not hesitated, to carve out for themselves a particular niche in Australian society, and in the process to develop an identity which should, ideally, symbolize two features: (1) a distinctive contribution to Australian society generally and (2) a particular way of tackling their own problems and their own projects. By this last, I mean that so many projects which are being set up today are stimu- lated from the “outside”, and their organization and motivations are in fact non-Aboriginal. Their aim is to achieve socio-economic viability in both short and long range terms. But just because these may be run or operated by persons of Aboriginal descent, that does not automati- cally make them “Aboriginal”. They must also be fitted into a particular ethos, into a par- ticular framework of ideas which could be defined as Aboriginal. Aboriginal identity, and what is meant by that label, could provide that ethos. It is within this context, as in others, that research — especially anthropological research — becomes vitally significant. As far as Aborigines are concerned, systematic research is needed into all aspects of Aboriginal life: traditional, and in terms of current and past changes, and in relation to all conditions of living wherever these are to be found. It is necessary to have a detailed understanding of a large range of social situations, so that that knowledge can be ap- plied practically. I am not, here, emphasizing its significance in purely professional terms: that is another matter. I am concerned, though, that the results of such work should be available to all who are involved in Aboriginal advance- ment, including Aborigines themselves. For one thing, a meaningful framework for social identity can be sustained only through such knowledge. Various administrative policies and their translation into action must rest on a firm basis of understanding what is being done and what can result from it. Too often, in my experience, such research has been ignored, or hasty surveys by commercially-oriented research consultants have been made and ventures initiated with little or no awareness of what the possible im- plications might be for the people themselves. The States and the Commonwealth have both erred in this respect. Human beings are too valuable a commodity to be treated so brusquely. It is not money alone which will transform the Aboriginal scene. Rather, that hinges on how money is spent; and how it is spent should rest on systematic research, with proper attention to the needs and wishes of the people involved. They require personal attention, and local situa- tions require local consideration. In regard to the last, centralization in respect to Canberra or elsewhere, almost inevitably overshadows local perspectives. It can spell impersonalization. It also means that more controls are likely to be imposed — together with more stress on uni- formity and less on diversity, which (within a certain range) is a necessary aspect of ordinary living. Anthropologically, cultural diversity has almost an intrinsic value of its own, as some- thing which is of immense importance to man- kind, just as are individual variations. In rela- tion to people, centralization could mean less, or less effective, management in their own affairs. This point is quite vital. Aborigines are only now being really involved in processes of decision-making. Only now are alternative choices available to them. There are different ways of achieving similar goals; and those different ways, or the choices relevant to them, should be kept open. Moreover, a reasonable choice from among a range of possible alterna- tives can be made only if the persons involved are aware of the consequences. One of the major tasks of the social sciences is to supply that information in a form which can be used by people who do not have particular training in that direction. It is the responsibility of all Aboriginal administrative agencies to seek out that knowledge and to apply it. And it is to the advantage of all Aborigines to be able to draw on such knowledge. Hopefully, also, more Aborigines will come to have a greater apprecia- tion of social science research, in its theoretical as well as its applied aspects, and some of them will themselves carry out such research, not only among their own people but in the wider Australian scene and beyond. References Berndt, C. H. (1969). — A Time of Rediscovery. In “Aboriginal Progress. A New Era?” (D. E. Hutchison, ed.) University of Western Aus- tralia Press, Perth. Berndt, R. M. (1971). — The Concept of Protest within an Australian Aboriginal Context. In “A Question of Choice: an Australian Aboriginal Dilemma.” (R. M. Berndt, ed.) University of Western Australia Press, Perth. Campbell, W. D., and W. H. Bird (1916). — An account of the Aboriginals of Sunday Island, King Sound, Kimberley, Western Australia. Jour- nal and Proceedings of the Royal Society of Western Australia I: 1914-15 (published 1915). Strehlow, T. G. H. (1971). — “ Songs of Central Australia”. Angus and Robertson, Sydney. Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 8 2. — Clastic dykes at Albany, Western Australia by John G. Kay 1 Manuscript received 22 August 1972; accepted 20 February, 1973. Abstract. Four clastic dykes outcrop on the northern shore of the entrance channel of Princess Royal Harbour, Albany. The largest dyke is about 1 metre wide, the others less than 10 cm wide. All are made up of well lithified, quartz-rich, wacke-type sediment and are enclosed by biotitic granitic gneiss. Introduction Clastic dykes have been observed at several widely scattered localities in Western Australia, for example at Watheroo, the Billeranga Hills, Puntapin Hill, Dillon Bay and Albany. Of the known examples, only those at Watheroo have been described (Logan 1958). The purpose of this paper is to record the occurrence at Albany. Description of dykes The clastic dykes at Albany outcrop on the northern shore of the entrance channel of Prin- cess Royal Harbour, as shown in Figure 1. Access is by bitumen road from the Albany townsite. The dykes cut across a flat gneissic pavement which is about 10 metres wide, and continue northward beneath unconsolidated sand and southward under the harbour waters; it is, therefore, not possible to see their full length. These dykes trend approximately 300°, almost at right angles to the shore line, and dip 80° to 85° W. A third dyke extends south - i Department of Geology, The University of Western Australia, Nedlands, Western Australia 6009. Figure 2. — General view looking south along the strike of the largest dyke. This dyke is about 1 metre wide. ward from the shore line and at very low tide its northern few metres are exposed. A fourth dyke occurs on a small, rather rough gneissic head- land immediately east of the pavement. The widest dyke is about 1 metre in width and is shown in Figure 2; the others are narrower, averaging less than 10 cm in width. They are well lithified and are as resistant to marine erosion as the adjacent gneiss. The dykes simulate normal intrusives, and their clastic lithology is not obvious unless looked at closely. Included fragments of gneiss have the appearance of xenoliths. Margins, particularly those of the widest dyke, are pos- sibly offset, but foliation patterns in the Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 9 gneissic country rock are not sufficiently dis- tinctive to allow indisputable matching. The dykes are enclosed by biotitic granitic gneiss which has a foliation trending 250° and dipping 65° N. Numerous joints cut the dykes in a diagonal pattern and many continue without refraction across the dyke boundaries into sur- rounding gneiss. The dykes are made up of well-lithified, quartz-rich wacke-type sediment. Detrital par- ticles include quartz, quartzite, microcline, plagioclase, pinitized cordierite, zircon, apatite and rare fragments of gneiss; quartz grains make up approximately 70 per cent. Size sorting is very poor, and grains range from less than 0.05 mm to more than 4 mm across. Roundness and sphericity are also variable, and all gradations from well-rounded, sub-spherical grains to sharply angular fragments are present. The boundaries of many grains are irregular and embayed as a result of marginal solution and matrix encroachment. Between 5 and 15 per cent of the rock is made up of matrix, originally clay, but now chlorite, biotite, muscovite and limonite. Matrix in the narrow dykes is mostly fine-grained dark green chlorite, with rare flakes of muscovite and biotite, whereas in the widest dyke it is limonite and muscovite. This difference is expressed in the greyish green colour of the narrow dykes and the dark purplish brown of the widest dyke. Seem- ingly, the original clay matrix was altered to chlorite, which in turn partly altered to biotite and muscovite in the narrow dykes and almost completely altered to muscovite and limonite in the widest dyke. The alterations are probably authigenic, but the possibility of slight meta- morphism cannot be excluded. Reasons for the matrix differences are unknown. Secondary muscovite flakes range in size up to 0.05 mm in thickness and more than 0.8 mm across, and commonly are moulded around detrital grains. Penetration into detrital grains is rare. The lithification and secondary mica development suggest that the clastic dykes are not recent, and may be of considerable age. Sediments comparable with the dykes are not known else- where in the Albany area. Specimens of the Princess Royal Harbour clastic dykes are housed in the rock collection of the Department of Geology, The University of Western Australia; the largest dyke is rep- resented by specimens 44933 and 48211, the smaller dykes by specimens 44934 and 48210. Reference Logan, B. W. (1958). — Clastic Dykes from Watheroo, Western Australia. J. Roy. Soc. W. Aust. 41: 27-28. Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 10 3. — The pouch of Planigale subtilissima and other dasyurid marsupials by P. Woolley 1 Manuscript received 22 May 1973; accepted 17 July 1973 Abstract Planigale subtilissima is little known. In the two female specimens available for study the structure of the pouch was different from that of all other dasyurid marsupials for which information is available. The structure of the pouch is described and compared with that of other dasyurids. Information obtained at the time of capture of the animals suggests that P. subtilissima breeds in the summer months. Introduction Two female Planigale subtilissima collected by the Combined Museum’s Expedition to the Ord River, Western Australia in mid-January, 1972 were maintained at the Western Australian Museum, Perth by M. Archer until mid-Septem- ber and then at La Trobe University, Melbourne by the author until their deaths in November, 1972. When captured one female (number 1) had young on the nipples but there were no young in the pouch of the other female (number 2). Both females were sent to Perth soon after capture; the young of female 1 were lost in transit. Previously this species was known only from the type specimen,* * a male, caught by Dr. Mjoberg’s Swedish Scientific Expedition to Aus- tralia 1910-13 (Lonnberg 1913) and from six animals obtained by B. Rudeforth, in December 1949. Four of these animals, including both sexes, were maintained alive in the Zoology Department of the University of Western Aus- tralia, and general observations on their biology were made. Over a four month period in cap- tivity “no young were observed, although the pouch of the female seemed to change in size as though in preparation for carrying young during that time” (Rudeforth 1950). The pouch of Planigale subtilissima On arrival in Perth the pouch of 9 1, which was now empty, was stained, the nipples elongated and the pouch hairs long and stained (M. Archer in. litt .) . The appearance of the pouch of $ 2 suggested to Archer that this female also was in breeding condition. The pouch hairs were slightly stained and slightly longer than the surrounding body hair and the nipples were well developed. By 10th February 1 Department of Zoology, La Trobe University, Mel- bourne, Australia. * Phascogale subtilissima, transferred to Planigale by Troughton (1928). the pouch hairs had doubled in length. How- ever, no young appeared and within a month the pouch hairs had become less prominent and remained so. When the animals arrived in Melbourne the pouch of each animal was inconspicuous and the entrance to it partly covered by long hairs. Because of the difficulty in handling these very small animals (body weight 5.0 to 6.0 g) no detailed examination of the pouch was at- tempted while the animals were alive. The superficial appearance of the pouch throughout the period the animals were alive in Melbourne can be seen in Figure 1. Examination of the pouch following the death of each animal revealed a structure different from that recorded for any other species of dasyurid marsupial. When the hairs covering the pouch region were clipped a fold of skin forming an anteriorly directed pouch with the opening at the rear could be seen, but no nipples could be found on the abdominal skin beneath the overlying fold. The pouch skin was dissected away from the body and two pockets were seen projecting forward from the anterior margin of the skin fold (Figure 2). After everting the pockets 5 nipples could be seen around the antero-lateral margins of each pocket (Figure 3). In each animal the hairs in the pockets and on the skin lining the entrance area were reddish brown in colour and there was an accumulation of dry red secretion around them. The pouch of other dasyurid marsupials The pouches of other species of dasyurid mar- supials for which information is available appear to be of three general types: — Type 1. The mammary area has no covering fold of skin. Marginal (usually lateral) ridges of skin develop during the breed- ing season. Type 2. The mammary area is partially covered by a crescentic antero-lateral fold of skin. The fold is usually deepest anteriorly. Type 3. The mammary area is covered by a circular fold of skin. These three types of pouch are shown diagram- matically in Figure 4, together with a diagram of the pouch of P. subtilissima (Type 4) for comparison. The type of pouch found in vari- ous species of dasyurid marsupials is listed in Table 1. The typical pouch condition may not always be apparent ; it is seen only in the breeding season in all species with a Type 1 pouch, and Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. ~ Figure 1. — Ventral view of female 1 showing the entrance to the pouch which is partly covered by long hairs. Figure 2. — The pouch and cloacal region of female 2 dissected from the abdomen. The V-shaped indentation (arrowed) in the skin fold covering the entrance to the pouch was probably an artefact caused by contraction of the loose skin. It was not present in female 1. Figure 3. — The pouch and cloacal region of female 1 dissected from the abdomen. The skin fold has been turned forward and over and the two pockets everted to show the interior of the pouch. Arrows point to two of the five nipples in the left pocket. TYPE 3 TYPE 4 Figure 4. — Diagrammatic representation of the types of pouch found in dasyurid marsupials. The broken lines indicate the limits of the pouch area. The solid lines indicate, in Type 1, the marginal ridges of skin, and in Types 2, 3 and 4, the free edge of the fold of skin at the entrance to the pouch. Where the nipples are exposed they are shown as solid rather than open circles. Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 12 Table 1 The type of pouch in various species of dasyurid marsupials Species Pouch Described by Type Planigale subtilissima 4 Woolley, this paper. Planigale ingrami 2 Troughton (1928), Heinsohn (1970). Planigale tenuirostris 2 Troughton (1928). Planigale gilesi 2 Aitken (1972). Antechinus maculatus 2 Pocock (1926) — as Phascogale minutissima, Johnson (1964). Antechinus stuartii 1 Pocock (1926) — as Phascogale unicolor, Horner and Taylor (1959) and Marlow (1961) under misnomer A. flavipes, Woolley (1966a, b). Antechinus flavipes flavipes 1 Pocock (1926) — as Phascogale flavipes, Fleay (1949), Woolley (1966b), Wakefield and War- neke (1967). Antechinus flavipes leucogaster 1 Woolley (1966b). Antechinus godmani 1 Wakefield and Warneke (1967). Antechinus apicalis 1 Woolley (1971a). Antechinus rosamondae 1 Hide (1964). Antechinus macdonnellensis 1 Spencer (1896), Woolley (pers. obs.) Antechinus swainsonii 1 Pocock (1926) — as Phasogale swainsonii, Fleay (1932), Wakefield and Warneke (1963), Woolley (pers. obs.). Antechinus minimus 1 Wakefield and Warneke (1963). Antechinomys spenceri 2 Pocock (1926), Woolley (pers. obs.). Antechinomys laniger 2 Lidicker and Marlow (1970). Sminthopsis crassicaudata 3 Pocock (1926), Fleay (1929), Smith and Godfrey (1970), Woolley (pers. obs./). Sminthopsis larapinta 3 Godfrey (1969), Woolley (pers. obs.). Dasycerus cristicauda 1 Spencer (1896), Pocock (1926), Jones (1949). Fleay (1961), Michener (1969), Woolley (1971b). Dasyuroides byrnei 1 Spencer (1896), Jones (1923), Pocock (1926), Woolley (1971b). Phascogale tapoatafa 1 Fleay (1934). Satanellus hallucatus 1 Pocock (1926) — as Dasyurus hallucatus, Fleay (1962), Johnson (1964). Dasyurus viverrinus 1 Pocock (1926), O’Donoghue (1911), Hill and O’Donoghue (1913), Fleay (1935a), Green (1967). Dasyurus geoffroii 1 Pocock (1926). Dasyurus maculatus 2 Pocock (1926), Fleay (1940). Sarcophilus harrisii 2 Pocock (1926), Fleay (1935b), Green (1967), Guiler (1970). Thylacinus cynocephalus 2 Pocock (1926). Myrmecobius fasciatus 1 Jones (1923), Calaby (1960). N eophascogale lorentzii 2 Pocock (1926) — as Phascogale lorentzii. Phascocolosorex dorsalis 2 Pocock (1926) — as Phascogale dorsalis. Myoictis melas 1 Pocock (1926) — as Phascogale thorbeckiana. it may not be present in immature females of species with another type of pouch e.g. An- techinomys spenceri (Type 2 pouch). The skin fold in this species does not develop until the approach of the first breeding season (Woolley, pers. obs.) and this may account for statements that a pouch is absent in A. spencer i (see Lidicker and Marlow 1970 p. 219). The appearance of the pouch is known to change during oestrus, pregnancy and lactation. The changes that occur during oestrus and pregnancy have been described in detail for Antechinus stuartii (Type 1 pouch) by Woolley (1966a, b), for Dasyurus viverrinus (Type 1 pouch) by O’Donoghue (1911) and for Smin- thopsis larapinta (Type 3 pouch) by Godfrey (1969). Other species with a Type 1 pouch in which changes similar to those observed in A. stuartii during pregnancy, but not during oes- trus, have been recorded include Antechinus fiavipes flavipes and Antechinus flavipes leuco- gaster (Woolley 1966b), Antechinus apicalis (Woolley 1971a), Dasycercus cristicauda (Mich- ener 1969; and Woolley 1971b) and Dasyuroides byrnei (Woolley 1971b). Smith and Godfrey (1970) noted changes in the pouch of pregnant Sminthopsis crassicaudata (Type 3 pouch) similar to those seen in Sminthopsis larapinta. Identical changes in the pouch are known to occur in females of many of these species kept isolated from males during the breeding season. Among species with a Type 2 pouch, changes in the pouch during the breeding season have been observed in Planigale ingrami (Fleay 1965), Antechinomys spenceri (Woolley, pers. obs.), Dasyurus maculatus (Fleay 1940) and Sarco- philus harrisii (Fleay 1935b). The changes in the pouch that occur during lactation involve mainly enlargement of the mammary area, nipples and skin folds. Staining of the pouch skin and hairs may also occur. When the young are weaned the pouch slowly regresses to approximately the condition seen immediately prior to the commencement of the breeding season, except that the nipples remain slightly elongated and the pouch hairs some- times lightly stained. The appearance of the nipples and pouch hairs provide a means of distinguishing between females that have reared young and those that have not. Discussion The pouch of P. subtilissima differs from that of other dasyurid marsupials in that the mam- mary area is more fully enclosed. The nipples are not located on the abdominal skin beneath the skin fold as in species with an enclosed mammary area (Type 2 or Type 3 pouch), but in two anteriorly directed pockets which are extensions of the area covered by the skin fold. Archer was able to see the nipples in both females when he received them but at the time of their deaths the nipples could not be seen until the pockets were everted. This suggests that the proportions of the pouch of each Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 13 animal were different at these times. Female 1 was known to have been suckling young and the nipples may have been visible as a result of general enlargement of the pouch to accom- modate the young. The fact that he could see well developed nipples in female 2, together with the similarity in the appearance of the pouches of the two females, suggests that female 2 also had been suckling young which may have been lost at or shortly before capture. These observations on pouch development in the two females, together with that of Rudeforth (1950) mentioned above, suggest that P. sub- tilissima breeds in the summer months, unlike the majority of dasyurid marsupials (Woolley 1973). A long period of lactation, ranging from 10 weeks in some species to 5 months or more in others (Woolley 1973) is characteristic of dasyurid marsupials. Like most other mar- supials the young are born at a very early stage of development by comparison with placental mammals. The young suckle continuously for some weeks and then intermittently for the remainder of the period of lactation. When suckling becomes intermittent the young are not carried about by the mother at all times. Dur- ing the period when suckling is continuous and the young are carried at all times the amount of cover afforded the young by the pouch is different for each type of pouch. In those species with a Type 1 pouch the young are almost completely exposed from the time of birth. The young of species with a Type 2 pouch may be covered by the skin fold only during the very early part of the period when suckling is continuous, as in Antechinomys spenceri (Woolley, pers. obs.) or for the entire period, as in Sarcophilus harrisii (Fleay 1935b). Fleay records that in this species suckling is continuous for 15 weeks, and 4 young can still be carried completely inside the pouch at 15 weeks. The young of Sminthopsis larapinta, which has a Type 3 pouch, are fully enclosed within the pouch for about 37 of the 40 days during which suckling is continuous (Godfrey 1969). Similarly, in Sminthopsis crassicaudata, which also has a Type 3 pouch, the young are fully enclosed while suckling is continuous (Ewer 1968). While no information is available on the development of the young or the duration of lactation in P. subtilissima it seems likely that the more fully enclosed mammary area would provide complete cover for the young for some time during lactation. The different types of pouch found in dasyurid marsupials must pro- vide different environments for the young, at least during the early part of lactation, but what effect this may have on the development of physiological processes in the young is not known. No obvious correlation can be seen between the type of pouch found in different species and the habitat of the animals, body size, the num- ber of young per litter and the duration of the period of continuous suckling or the total period of lactation. Acknowledgements . — The author wishes to thank Dr. D. Kitchener, of the Western Australian Museum, and the Department of Fisheries and Fauna, Western Aus- tralia, for the opportunity to study these animals, and Mr. M. Archer, of the Queensland Museum, for permis- sion to refer to his unpublished observations. References Aitken, P. F. (1972) .— Planigale gilesi (Marsupialia, Dasyuridae); a new species from the interior of south-eastern Australia. Rec. S. Aust. Mus. 16: 1-14. Calaby, J. H. (1960). — Observations on the banded ant- eater Myrmecobius /. fasciatus Waterhouse (Marsupialia), with particular reference to its food habits. Proc. zool. Soc. Lond. 135: 183-207. Ewer, R. F. (1968). — A preliminary survey of the be- haviour in captivity of the dasyurid mar- supial, Sminthopsis crassicaudata (Gould). Z. Tierpsychol. 25: 319-365. Fleay, D. (1929). — The fat-tailed pouched mouse. Viet. Nat. 45: 278-280. Fleay, D. (1932). — Swainson’s phascogale (the “bush mouse”). Viet. Nat. 49: 132-134. Fleay, D. (1934). — The brush-tailed phascogale. First record of breeding habits. Viet. Nat. 51: 89-100. Fleay, D. (1935a). — Breeding of Dasyurus viverrinus and general observations on the species. J. Mammal. 16: 10-16. Fleay, D. (1935b). — Notes on the breeding of Tasmanian devils. Viet. Nat. 52: 100-105. Fleay, D. (1940). — Breeding of the tiger-cat. Viet. Nat. 56: 159-163. Fleay, D. (1949). — The yellow-footed marsupial mouse. Viet. Nat. 65: 273-277. Fleay, D. (1961). — Breeding the Mulgara. Viet. Nat. 78: 160-167. Fleay, D. (1962). — The northern Quoll, Satanellus hallu- catus. Viet. Nat. 78: 288-293. Fleay, D. (1965). — Australia’s “needle-in-a-haystack” marsupial. Vicissitudes in the pursuit and study of Ingram’s Planigale, the smallest pouch-bearer. Viet. Nat. 82: 162-167. Godfrey, G. K. (1969). — Reproduction in a laboratory colony of the marsupial mouse Sminthopsis larapinta (Marsupialia: Dasyuridae). Aust. J. Zool. 17: 637-654. Green, R. H. (1967). — Notes on the Devil (Sarcophilus harrisii) and the Quoll (Dasyurus viverrinus) in north-eastern Tasmania. Rec. Queen Viet. Mus. No. 27: 1-13. Guiler, E. R. (1970). — Observations on the Tasmanian Devil, Sarcophilus harrisii (Marsupialia: Dasyuridae). II. Reproduction, breeding and growth of pouch young. Aust. J. Zool. 18: 63-70. Heinsohn, G. F. (1970). — World’s smallest marsupial. The fiat-headed marsupial mouse. Animals 13: 220-222. Hill, J. P., and O’Donoghue, C. H. (1913). — The repro- ductive cycle in the marsupial Dasyurus viverrinus. Quart. J. microsc. Sci 59: 133-174. Horner, B. E., and Taylor, J. M. (1959). — Results of the Archbold Expeditions. No. 80. Observations on the biology of the yellow-footed mar- supial mouse, Antechinus flavipes flavipes. Am. Mus. Novit. No. 1972: 1-24. Johnson, D. H. (1964). — Mammals of the Arnhem Land Expedition. In “ American- Australian Scien- tific Expedition to Arnhem Land. Records. 4. Zoology”, 427-515. Melbourne University Press. Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 14 Jones, F. W. (1923) —"The Mammals of South Aus- tralia.” Part 1. The Monotremes and Car- nivorous Marsupials. Government Printer, Adelaide. Jones, F. W. (1949). — The study of a generalised mar- supial (Dasycercus cristicauda Krefft). Trans, zool. Soc. Lond. 26: 409-501. Lidicker, W. Z., and Marlow, B. J. (1970). — A review of the dasyurid marsupial genus Antechinomys Krefft. Mammalia 34: 212-227. Lonnberg, E. (1913). — Results of Dr. E. Mjoberg’s Swed- ish Scientific Expeditions to Australia 1910- 13. I. Mammals. K. svenska Vetensk Akad. Handl. 52 (1): 1-10. Marlow, B. J. (1961). — Reproductive behaviour of the marsupial mouse, Antechinus flavipes (Waterhouse) (Marsupialia) and the develop- ment of the pouch young. Aust. J. Zool. 9: 203-218. Michener, G. R. (1969). — Notes on the breeding and young of the crest-tailed marsupial mouse, Dasycercus cristicauda. J. Mammal. 50: 633- 635. O’Donoghue, C. H. (1911). — The growth-changes in the mammary apparatus of Dasyurus and the relation of the corpora lutea thereto. Quart. J. microsc. Sci. 57: 187-234. Pocock, R. I. (1926). — The external characters of Thylacinus, Sarcophilus and some related marsupials. Proc. Zool. Soc. No. 68: 1037- 1084. Ride, W. D. L. (1964 ). — Antechinus rosamondae, a new species of dasyurid marsupial from the Pil- bara district of Western Australia; with remarks on the classification of Antechinus. W. Aust. Nat. 9: 58-65. Rudeforth, B. F. (1950). — Some notes on an interesting marsupial. Scope No. 5: 10-11. Smith, M. J., and Godfrey, G. K. (1970).— Ovulation induced by gonadotrophins in the marsu- pial, Sminthopsis crassicaudata (Gould). J. Reprod. Fert. 22: 41-47. Spencer, B. (1896) .—Mammalia. In “ Report on the work of the Horn Scientific Expedition to Central Australia. Part 2. Zoology”, 1-52. London, Dulau and Co. Trough ton, E. Le G. (1928). — A new genus, species, and subspecies of marsupial mice (Family Dasy- uridae). Rec. Aust. Mus. 16: 281-288. Wakefield, N. A., and Warneke, R. M. (1963). — Some revision in Antechinus (Marsupialia) — 1. Viet. Nat. 80: 194-219. Wakefield, N.A., and Warneke, R. M. (1967). — Some revision in Antechinus (Marsupialia) — 2. Viet. Nat. 84: 69-99. Woolley, P. (1966a). — Reproduction in Antechinus spp. and other dasyurid marsupials. Symp. zool. Soc. Lond. no. 15: 281-294. Woolley, P. (1966b). — Reproductive biology of Antechinus stuartii Macleay (Marsupialia: Dasyuridae). Ph. D. thesis, Australian National University, Canberra. Woolley, P. (1971a). — Observations on the reproductive biology of the Dibbler, Antechinus apicalis (Marsupialia: Dasyuridae). J. Proc. R. Soc. West. Aust. 54: 99-102. Woolley, P. (1971b). — Maintenance and breeding of laboratory colonies of Dasyuroides byrnei and Dasycercus cristicauda. Int. Zoo Yb. 11: 54, 351-354. Woolley, P. (1973). — Breeding patterns, and the breeding and laboratory maintenance of dasyurid marsupials. Exp. Animals 22 Sup- plement, 161-172. Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 15 4. — Cainozoic stratigraphy in the Perth area 1 by Patrick G. Quilty 2 Manuscript received 22 May 1973; accepted 17 July 1973 Abstract The Cainozoic sequence in the vicinity of Perth is described. The two major sedimentary cycles recognised are Late Palaeocene-Early Eocene and Early-Middle Miocene. The former includes the Kings Park Formation (including the Mullaloo Sandstone Member, defined here- in). The later cycle includes the Stark Bay Formation, also defined herein, and possibly an overlying poorly-known carbonate unit. The Mullaloo Sandstone Member is interpreted as a marine-deposited sand from the early Swan River, and the rest of the Kings Park Forma- tion as a more southerly estuarine to marine unit. The Stark Bay Formation is a carbonate- chert formation deposited on the inner con- tinental shelf. Age and ecological results are based on planktonic and benthonic foraminifera. Lepidocyclina aff. liowchini Chapman and Cres- pin is recorded from Gage Roads No. 2. It indicates that there were warm waters in the Early Miocene all around the Australian coast and also that the warm water planktonic foraminiferal zonation may be used in the Miocene of the Perth Basin. A fairly widespread Pliocene-Quaternary mar- ine incursion is represented in the area by thin, discontinuous sediments with characteristic molluscan faunas. Introduction History of study Several published papers have been concerned with the Tertiary palaeontology of the Perth Basin, mainly involving the Kings Park Forma- tion onshore in the vicinity of Perth. Several unpublished results are also worthy of recogni- tion. Recently, there has been an upsurge in interest in the younger Cainozoic and several preliminary papers have appeared. Tertiary sediments are not known from outcrop in the area under review, all evidence presented here having been gleaned from bore material. The earliest work in the area was reviewed by Coleman (1952) and that will not be repeated here. The major palaeontological results are contained in papers by Parr (1938), Coleman (1952), Cookson and Eisenack (1961), and Mc- Gowran (1964). Relevant unpublished works are those by Pudovskis (1962) and Cockbain and Ingram (1967). Wells (1943) made a brief comment on a coral fauna and Glaessner (1956) described a crustacean from one of the bores. Parr’s (1938) paper was the first major con- tribution to the understanding of age relation- ships in the area. He examined in detail the foraminiferal fauna from the Kings Park No. 2 Bore and assigned an Eocene age by comparison with contemporary American studies, mainly by Cushman. This Eocene age was accepted for quite some time, although some aspects of the fauna were confusing. 1 Published with permission of the management of West Australian Petroleum Pty. Limited. 2 West Australian Petroleum Pty. Limited, Box C1580, G.P.O., Perth, Western Australia 6001. Coleman (1952) examined more material from a total of six bores in the Perth metro- politan area. His paper carried two important footnotes. One foreshadowed the application of the name Kings Park Shale for the sequence studied. The other hinted at a possible Pal- aeocene age by comparison with contemporary studies in Sweden. Coleman recorded a more diverse foraminiferal fauna and several more groups of fossils than did Parr. He noted difficulties in biostratigraphic correlation in the Perth metropolitan area bores. Cookson and Eisenack (1961) expanded the area of interest to the Rottnest Island Bore. Two samples from the Kings Park Formation interval were described as younger than Vic- torian time equivalents of the Kings Park Formation in its type section. Pudovskis (1962), while still accepting the Late Eocene age for the Kings Park Formation, produced what is still the most detailed litho- logical correlation of wells in the area. His report used information from 38 wells spread from Pt. Peron and Rottnest Island in the west to the Midland Railway workshops and Helena Vale in the east. McGowran (1964) restudied Parr’s samples and concluded that the age of the Kings Park Formation in its type section (Kings Park No. 2) is Late Palaeocene, Globorotalia pseudomen - ardii Subzone of the G. velascoensis Zone of Berggren (1965) (G. pseudomenardii Zone of Bolli, 1957). This is the same as zone P4 in the comprehensive scheme of Blow (1969) as detailed by several authors including Berggren (1971). McGowran suggested that it was un- likely that the material in the Rottnest Island Bore (284-666 m) would be younger than that in Kings Park No. 2. Cockbain and Ingram (1967) examined foraminifera and palynomorphs from the Rottnest Island Bore and recorded a Late Palaeocene to Early Eocene age for the sedi- ments there. No foraminiferal investigations have been made on bore material from the post-Miocene in the Perth area. However, Mr. G. W. Kendrick, W. A. Museum, has commenced detailed studies of molluscs (mainly bivalves) from shallow bores and from outcrops of the Coastal Limestone (Darragh and Kendrick, 1971; Kendrick, 1960). Enough information is avail- able from his work to suggest that there are at least two post-Miocene episodes of marine sedimentation in the Perth area. Scope of this investigation All wells drilled offshore in the Perth Basin have been drilled for West Australian Petroleum Pty. Limited (WAPET). Study of the bores drilled to-date permits this summary to be made. Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 16 Figure 1. — Tertiary section Kings Park No. 2-Rce No. 1. Figure 2. — Tertiary section Quinns Rocks No. 1-Warnbro No. 1. SECTION TO HORIZONTAL SCALE Figure 3. — Isopach and palaeogeographic map of the Kings Park Formation. There is a large area of Tertiary, between the sequence in the Carnarvon Basin 1000 km to the north and in the Plantagenet Group, 400 km to the south from which little has been studied. Apart from the Kings Park Formation, Figure 4. — Tertiary isopach and depth to base Tertiary. no proven Tertiary marine faunas or lithological units have been defined yet from the Perth area. This report deals mainly with the Palaeocene, Eocene and Miocene, but the younger sediments are noted briefly for completeness. Character- journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 19 istic foraminifera are illustrated on two plates. Figures 1 and 2 are compilations of the information detailed in the major part of this paper. Figures 3 to 5 show the palaeogeography and distribution of the units mentioned. Figures 6 and 7 give electric log and lithology of the new units defined and Figure 8 is a summary of the Cainozoic stratigraphy of the Perth area. Sample reliability The most reliable samples are from con- ventional cores (C) or sidewall cores (SWC). These are taken from the depth indicated. All available samples were used in this study. In many cases, the only samples available are from ditch cuttings (DC). These samples are circulated to the surface in drilling mud and removed from the mud on vibrating screens. While they are usually fairly representative of the fauna at the depth taken, they may contain much downhole contamination, or rarely, may be entirely composed of contaminants. Ditch cuttings samples are of a much lower order of reliability than core samples. Depth Measurements . — Throughout this work, depths mentioned are drill depths; this is, depths below the Rotary Table (R.T.). All measurements were originally recorded in feet, and have been converted to the nearest metre. Repositories . — Rock specimens are held at the Geological Survey of Western Australia and at the Bureau of Mineral Resources, Geology and Geophysics, Canberra. The foraminifera figured on the plates are housed in the Palaeontological collection of the Geology Department, University of Western Australia and the number following the initials U.W.A.G.D. is the catalogue number in that collection. Discussion of the sections examined Kings Park area No new material has been examined from the type section of the Kings Park Formation, as the age, fauna and lithology have been described adequately by Parr (1938), Coleman (1952) and McGowran (1964). The Kings Park Shale was described in McWhae et al. (1958, p. 130) as consisting of “grey calcareous shales and claystones . . . .”. Pudovskis (1962) stated that it “consists pre- dominantly of grey calcareous, glauconitic siltstones, shales and some sandstones. Some thin hard limestone beds are present in the lower part of the formation”. The latter seems the better definition. Samples have been examined cursorily from excavations at the Narrows Interchange Project and also from the old Celtic Club Bore on the corner of St. George’s Terrace and Irwin St., Perth. Although faunas vary a little from those in the type section, the lithology appears iden- tical and the age of P4 for the Kings Park Formation onshore is substantiated. Overlying the Kings Park Formation at the Narrows Interchange site (Fig. 1) is a sequence of Quaternary deposits. At the base strati- graphically is the Coastal Limestone, which has Figure 5. — Distribution of the Stark Bay Formation and of the Pliocene onshore. been dated on Rottnest Island at about 100 000 ± 20 000 years BP (Teichert, 1967). This formation (aeolian in the immediate vicinity) has been eroded by the Swan River and two sequences of estuarine muds succeed it. The older sequence is the more indurated and Journal of the Royal Society of Western Australia. Vol. 57, Part 1. May. 1974. 20 is referred to as the “Blue Mud” by site engineers. It is distinctly older than the present sequence and contains abundant bivalves attesting to a marine or estuarine environment. No foraminifera were recovered from this material. The younger sequence (deposited by the modern erosional cycle) is referred to as the “Black Mud”. The Quaternary sequence extends to approxi- mately 37 m below sea level. The relationships are illustrated diagrammatically in Figure 1. Figure 6. — Lithological and log characters of the Mullaloo Sandstone Member of the Kings Park Formation. METRES 400 FEET 1*00 Figure 7. — Lithological and log characters of the Stark Bay Formation. Figure 8. — Stratigraphic column for the Cainozoic of the Perth Area. AGE ROCK UNIT MILLS YEARS 0 m 1 ATFRNARV lii’** -y 77 \Jy JH I cniiHnT ll II COASTAL LIMESTONE ^DS N 21 PLIOCENE N 20 MARINE SEDIMENTS IN VARIOUS N 19 WATER BORES AROUND PERTH N 18 N 17 N 16 1 AGE AND STRATIGRAPHIC RELATIONS UNKNOWN N 15 10 N 14 UNNAMED CARBONATE FORMATION N 13 ‘ROTTNEST SANDSTONE' N 12 MIOCENE N 11 N 10 N 9 STARK BAY FORMATION ^ 1 N 8 N 7 y 20 N 6 N 5 N 4 N 3 30 OLIGOCENE N 2 >- QL < N 1 h- Qd LU L - P 18 r— P 17 40 P 16 P 15 P 14 P 13 EOCENE P 12 P 11 P 10 50 P 9 P 8 P 7 P 6 P 5 KINGS PARK FORMATION \ PAI AFHPPNP P 4 MULLALOO SANDSTONE 60 P 3 P 2 P 1 70 — CRETACEOUS Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 21 The Rottnest Island Bore Gage Roads No. 1 The Rottnest Island Bore (Fig. 1) was drilled in 1913 in an unsuccessful quest for artesian water. No samples were preserved between the surface and 390 m. Below 390 m cores were preserved representing the intervals marked on Figure 1. 284-666 m: Kings Park Formation. All cores recovered are from a lithology identical with the Kings Park Formation in its type section, although there is a slight but significant age difference between the formation on Rottnest Island and in its type section. Although there is the possibility of a stratigraphic break between the two sections, it seems probable that they are continuous and the name Kings Park Formation is taken as valid for the section in the Rottnest Island Bore. The top of the formation (284 m) is based on drillers’ records. The upper part (284-451 m) consists of grey argillaceous sandstone, and the lower part (451-666 m) of grey shale. Only one sample from this well is present in the WAPET collections — that from the interval 486-616 m. Through the courtesy of Messrs J. H. Lord, B.S. Ingram and Dr A. E. Cockbain, of the Geological Survey of Western Australia, I have been able to examine material from the following four intervals in the Rottnest Island Bore: 451-470 m; 480-486 m; 486-616 m; and 617-666 m. The shallowest sample contains an unidenti- fied Pseudohastigerina ( wilcoxensis or pseudoiota ), Globorotalia aequa Cushman and Renz, and Globigerina of the linaperta group. This fauna indicates a Globorotalia rex Zone (P6) age. The fauna from 480-486 m and 486-616 m is virtually identical. That from 617-666 m is a little different but there is no recognisable age difference. The main forms present are Pseudogloboquadrina primitiva Finlay, Globi- gerina of the linaperta group, Globorotalia aequa and G. broedermanni Cushman and Bermudez. This also supports a P6 age. The benthonic content of all faunas is typical of the Kings Park Formation and many other sections in Australia of Palaeocene age. 71-284 m: “Rottnest Sandstone ”. McWhae ( pers . comm.) informally used this name for . . . 700 feet thickness of friable, fine to coarse, and conglomeratic sandstones, red and brown in colour and unfossiliferous, probably continental in origin”. This formation is not known outside the type locality. It also seems that no representative collections of it are available now. Thus, the name remains informal and the quotation marks used are appropriate. It could be part of the Kings Park Formation, an equivalent of the Stark Bay Formation or even younger. 0-71 m: Coastal Limestone. This record also is based on drillers’ records and no samples are known from the well. Teichert (1967) recorded an age of 100,000 ± 20,000 years BP for marine fossils from this formation on Rottnest Island. Gage Roads No. 1 was the second offshore oil well drilled in the Perth Basin. For this study, the following samples have been examined: Depth Sample Type 107-116 m DC* 116-125 m DC 301 m Q,** 302 m C 303 m c 329-338 m DC 393-402 m DC 415 m SWCf 430 m SWC 442 m SWC 463 m SWC 466-475 m DC 472 m SWC 486 m SWC 502 m SWC 512-521 m DC 527 m SWC 542 m SWC 583 m SWC 594-604 m DC 604-610 m DC 620 m * Ditch Cuttings ** Conventional Core t Sidewall Core SWC From these samples, from log interpretation and from wellsite sample examination, the following intervals are recognised in the well. Faunas in general are poor. The best samples for dating purposes are from 502 m and 604-610 m. 80-?100m: Coastal Limestone — Pleistocene: The well spudded in Coastal Limestone whose thickness is unknown but is probably thin, certainly less than 30 m as the ditch cuttings at 94 m are of the underlying formation. 9100-389 m: Unnamed Limestone Formation — Pliocene-Miocene: The formation consists of yellow and pink calcarenites, clearly of bio- logical origin in the upper region. Recrystalli- sation has destroyed most fossils but the shallow ditch cuttings contain abundant large foraminifera Amphistegina lessonii d’Orbigny and a species of Operculina which Barker (1960) refers to O. ammonoides (Gronovius). The species found here is the same as that figured by Barker but whether or not it is O. ammonoides is open to doubt, as was pointed out by Barker. This species makes up 60-70% of the foraminiferal fauna. The age of the fauna cannot be defined more accurately than post-Oligocene. The sediment accumulated in warm, shallow marine con- ditions. The relationship of this formation with the overlying Coastal Limestone is prob- ably unconformable. The nature of the contact with the sediment below is unknown. 389-619 m: Stark Bay Formation (new formation, defined later) — Early to Middle Miocene : The upper limit is taken at the marked log changes (resistivity) at 389 m and the lower limit at another marked change from the South Perth Formation (Cretaceous). Thirteen samples have been examined from this section. The entire section is basically dolomite or dolomitic limestone, and can be divided into Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 22 two subintervals on the basis of lithological variation. (a) 389-442 m (approximately). Dolomite unit. This interval consists of dolomites, grey in colour, barren of foraminifera and devoid of any primary features. (b) 442-619 m Cherty unit. This unit also contains much dolomite but grey and brown chert is dominant in the samples seen and lithological characters are more variable than in the above section. The chert contains abundant sponge remains and silicified calcareous fossils such as foramini- fera. The age control for the whole formation is based on the sidewall core at 502 m which contains a poorly preserved but dateable fauna. Forms present include Globigerinoides sicanus de Stefani, G. quadrilobatus trilobus Reuss, Globigerina cf G. euapertura Jenkins and Globorotalia obesa Bolli. These place the fauna in the Burdigalian (upper half of the Early Miocene) or lowest Langhian (N8-N9). The fauna is 25% planktonic species suggesting deposition in 30-60 m of water. This is sup- ported by the presence of more than 10% of bolivinid species. An interesting benthonic form in this fauna is Pavonina triformis Parr. Another sample yielding a good fauna is that from the interval 604-610 m. This contains Globorotalia barisanensis LeRoy, Globigerina woodi woodi Jenkins and Sherbornina cunei- marginata Wade. The age is slightly older than other Miocene records, probably about N7 (lower half of Globigerinatella insueta Zone of Bolli, 1957). The lower boundary of this section is marked by the log change at 619 m which represents a transition from dolomite and chert above over- lying a barren, fine sandstone of Cretaceous age below. Gage Roads No. 2 Samples examined from this well are as fol- lows: Depth Sample C. 305 m Cutting: 366 m SWC 369-372 m DC 396-399 m DC 402 m SWC 424-427 m DC 430 m SWC 451-454 m DC 461 m SWC 488-490 m DC 515-518 m DC 527 m SWC 543-546 m DC 549 m SWC 579 m SWC 579-582 m DC from blade of 20” bit Lithology. — From examination of logs, ditch cuttings and sidewall cores, the following litho- logical divisions are recognised in the well. 1. Ground Level — ? “Coastal Limestone”. The well spudded in carbonate sediment, probably still forming under marine conditions and identical with marine members of the Coastal Limestone. Its thickness is unknown, but is probably not great. Its base is not indicated on the logs. A thickness of the order of 30 m or less is envisaged as this is a “normal” thickness for the formation. 2. ? — approximately 362 m. Unnamed carbon- ate formation. Only a single sample is available from this interval. It is a bulk sample taken from the blades of the bit at about 304 m. While a poor sample, it gives some indication of the lithology at this depth. It consists of recrystallised red to yellow limestone, slightly indurated with almost all organic structure obliterated. Some forms are recognisable as foraminifera. None are identifiable. It is the same lithology as that occurring at about the same depth in Gage Roads No. 1 and Roe No. 1. The lower boundary is taken at 362 m as there is a marked sonic log change there, and the sample at 366 m is different. 3. 362-577 m Stark Bay Formation. 3a. 362-427 m. Several samples in this interval show that at least in part the sediment consists of white bryozoan calcarenites with rich, well preserved foraminiferal faunas. There is also grey to pale brown dolomite in most samples. This member seems absent from Gage Roads No. 1, but a much thicker sequence is found in Roe No. 1. It may represent the original lithology of the formation, diagenesis being responsible for the lithology below. 3b. 427-514 m. Sample control in this interval is poor, the only sidewall cores being at 430 m and 461 m. The litho- logy of these cores is identical and consists of saccharoidal, friable brown crystalline dolomite with minor pyrite. No fossils are identifiable. The upper and lower boundaries are selected from sonic logs. 3c. 514-544 m. This interval is known from a sidewall core at 526 m which consists of brown, friable, saccharoidal dolomite, markedly coarser than in the interval above. 3d. 544-577 m. Lithology here is recorded in a sidewall core at 549 m and in a very distinctive ditch cuttings sample at 579-582m. Lithology in the interval is variable. The sidewall core at 549 m is a mixture of powdery white un- altered calcarenite with brown dolomite and chert. Thus, diagenetic change in this interval is not so complete as above. The sample has minor iron oxide staining. The ditch cuttings sample from 579-582 m consists almost entirely of downhole contamination, prob- ably from just above the Cretaceous-Tertiary con- tact, here taken as 577 m — a sonic log pick. A side- wall core at 579 m is part of the underlying Cretaceous section. The ditch cuttings sample is partly of lightly iron-stained calcarenite with a well preserved foraminiferal fauna. The lithology and fauna are very characteristic, and have not been encountered above. They are thus probably very close to in situ. The faunas. — No satisfactory faunas occur in samples above 366 m, the top sidewall core. Also, faunas in ditch cuttings between 427 m and 544 m must be suspect and probably represent downhole contamination. Thus faunas from sidewall cores at 366 m, 402 m and 549 m, as well as ditch cuttings between 366-427 m and 544- 577 m, can be taken as representative of faunas at those depths. All faunas appear to belong to the N8-N9 interval of Blow (1969). Samples from above 427 m and possible con- taminated samples as deep as 518 m contain Orbulina universa d’Orbigny, Globorotalia archeomenardii Bolli, Globoquadrina dehiscens dehiscens Chapman, Parr and Collins, and occasionally Globigerinoides sicanus. Thus these Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 23 Figures 9, 10. — Globigerina mckannai White, Warnbro No. 1, 552 m (ditch cuttings), X120. UWAGD 70421. Figures 11, 12. — Globorotalia pusilla laevigata Bolli, Warnbro No. 1, 552 m (ditch cuttings), X180. UWAGD 70422. Figures 13, 14. — Globorotalia chapmani Parr, Warnbro No. 1, 552 m (ditch cuttings), X100. UWAGD 70423. Figures 15, 16. — Globorotalia dolabrata Jenkins, Quinns Rock No. 1, 263 m (sidewall core), X130. UWAGD 70424. Figures 17, 18. — Globorotalia velascoensis parva Rey, Warnbro No. 1, 552 m (ditch cuttings), X130. UWAGD 70425. Figures 19, 20. — Pseudogloboquadrina primitiva (Finlay), Warnbro No. 1, 552 m (ditch cuttings), X165. UWAGD 70426. Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 24 Figures 21, 22. — Pseudohastigerina sp. Rottnest Island bore, 486-616 m, X190. UWAGD 70427. Figures 23 24. — Globorotalia pseudomenardii Bolli, Warnbro No. 1, 552 m (ditch cuttings), X250. UWAGD 70428. Figure 25 —Globigerinoides sicanus de Stefani, Gage Roads No. 2, 582 m (ditch cuttings), X140. UWAGD 70429. Figure 26 .—Praeorbulina transitoria (Blow), Gage Roads No. 2, 582 m (ditch cuttings), Xc.150. Specimen lost since photography. Figure 21—Orbulina universa d’Orbigny, Gage Roads No. 2, 454m (ditch cuttings), X110. UWAGD 70430. Figures 28 29. — Globorotalia archeomenardii Bolli, Gage Roads No. 2, 399m (ditch cuttings), X150. UWAGD 70431. Figures 30-32 —Lepidocyclina cf. howchini Chapman and Crespin, Gage Roads No. 2 (ditch cuttings). 30, 31. — 399m (ditch cuttings). Equatorial section, UWAGD 70432. 30 x 40; 31 x 90. 32.— 427 m (ditch cuttings). Vertical section, X40. UWAGD 70433. Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 25 samples all appear to belong to N9 (approxi- mately the Globorotalia fohsi barisanensis Zone of Bolli, 1957). The deepest sample (579-582 m) is different, with Globigerinoides sicanus and Praeorbulina transitoria Blow. An age of N8 is thus indicated. A most interesting aspect of the faunas from the upper part of the section (especially 350-425 m) is the presence of Lepidocyclina in ditch cuttings. The significance is discussed later. Roe No. 1 The samples studied are from sidewall cores at 399 m 439 m, 498 m, 537 m and 600 m. No samples are available above these depths. The well spudded in calcarenites identical with parts of the Coastal Limestone. As in the other wells in this vicinity, it seems that the Coastal Limestone may still be forming here under marine conditions, as samples brought to the surface by divers consist of weakly consoli- dated shell fragments and quartz grains. The shells still retain their original colour. The formation’s thickness is unknown, again probably of the order of 30 m or less. The top four sidewall cores all contain abundant, well preserved foraminiferal faunas in a clean, white bryozoan calcarenite. There may be very minor development of chert and dolomite. This unit is the same as that seen in Gage Roads No. 2 but not in Gage Roads No. The sidewall core at 600 m has a much more poorly preserved fauna in a dolomitic unit with chert. This unit also occurs in Gage Roads Nos. 1 and 2. The boundary between the two units is not well marked lithologically but could be taken at 552 m on a marked sonic log change. The base of the Tertiary section is well marked at 616 m by sonic and electric log changes. The faunas. — The upper calcarenite unit con- tains abundant planktonic foraminifera includ- ing Orbulina universa (to 537 m), Globoqua- drina altispira globosa Bolli, and Globigerinoides quadrilobatus immaturus LeRoy. Globorotalia seems to be absent. By comparison with Gage Roads No. 1, this interval can be expected to be N9 in age. The lowest sidewall core contains a different, poorly preserved, small fauna with Globigerina euapertura. Also by comparison with the lower reaches of Gage Roads No. 2, and by virtue of the apparent absence of Orbulina , this sample may be taken as N8 in age. Quinns Rock No. 1 The following samples have been examined from this well: Depth Sample Type 243 m SWC 263 m SWC 314 m SWC 337 m SWC 358 m SWC 335-344 m DC Lithology. — The lithology encountered in the Tertiary section of this well is different from that encountered in any other sections discussed in this report (Figs. 2, 6). There is no “Coastal Limestone” section and the well spudded in quartz sandstone. 65-356 m: Mullaloo Sandstone Member of the Kings Park Formation (defined below) — Early Eocene. The entire Tertiary section consists of sandstone which Bozanic (1969) described as “. . . . characterized by the abundance of dis- crete fine to granule (2-4 mm) sized, well to very well-rounded quartz grains. The colour of the quartz is mostly clear and milky white with minor amounts of pale yellow and pale brown. Many of the grains are frosted. Sidewall cores indicate that most of the free quartz grains as observed in drilling samples are lightly cemented by argillaceous material before being drilled”. “A number of mostly fine grained but locally ranging from fine to coarse grained sandstone interbeds are present. The quartz grains vary from subangular to subrounded. Fine to medium sized grains of black glauconite are present and fine mica flakes locally common. The sandstone is variably cemented by calcareous, kaolinitic, argillaceous and pyritic material.” “Carbonaceous and lignitic streaks and fragments — often pyritized — are not uncom- mon.” “Siltstone zones are also developed. They are light grey, with fine and very fine black glau- conite grains and cemented by calcareous and argillaceous material.” The Faunas. — All samples examined but one, are sidewall cores and thus the samples can be regarded as in situ. Only two samples contain any very significant foraminifera. They are at 263 m and 314 m. The sample at 263 m contains two identifiable specimens. They are Globocassidulina subglobosa (Brady) and Globorotalia dolabrata Jenkins. Thus this sample is Early Eocene ( Globanoma - lina wilcoxensis to Globorotalia crater crater zones of Jenkins, 1971). Equivalence between Jenkins’s and Blow’s (1969) scheme is not yet precise, but the age of the sample in Blow’s scheme is P6-P7 approximately, about the same age as the Eocene sediments in the Rottnest Island Bore. The fauna at 314 m is less well dated. Again it consists of few specimens. Present are Pseudogloboquadrina primitiva and G. cf. taroubaensis Bronnimann, which support the Early Eocene age indicated for the sample above at 263 m. The other sidewall core samples are barren except for a single “ Cibicides ” umbonifer Parr at 337 m. The ditch cuttings sample from 344 m contains a small, well preserved Tertiary fauna including Globigerina triangularis White, which is consistent with the ages determined above. The sidewall core sample at 360 m contains a Cretaceous fauna and the base of the Tertiary is taken at the electric log-lithology change at 356 m. Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 26 Charlotte No. 1 The section in Charlotte No. 1 (Fig. 2) has been examined from the following samples: Depth 329 m 332 m 332-335 m 354 m 363-366 m 375 m 393-396 m 396 m 424-427 m 462 m Sample Type SWC SWC DC SWC DC SWC DC SWC DC SWC Lithology. — Lithology from sidewall cores and ditch cuttings is uniform between 328 m and 396 m. The electric log characters of this inter- val are consistent up to 209 m. and down to 411m. Thus the interval 200-411 m can be taken as one lithological unit. The sidewall core at 462 m is a clean, angular, feldspathic sand shown palynologically to be Lower Cretaceous, and the ditch cuttings at 427 m (in an interval of severe circulation problems) contain much clean rounded quartz sand. Thus the Cretaceous-Tertiary boundary could be taken at the log change at 411 m. How- ever, it is possible that the interval 411-451 m, a distinct part of the logs, is Tertiary sandstone, equivalent to the Early Eocene sands in Quinns Rock No. 1. This is unproven as no early Tertiary fossils are known from this interval. Thus the Cretaceous-Tertiary boun- dary could be taken alternatively at 451 m. The following lithological units can be recognized in the well. 1. 73m (Seabed) — 93m (approximately). “Coastal Limestone”. Material brought to the surface from the seabed consists mainly of shell fragments and may be presently forming Coastal Limestone. 2. 93 m— 209 m. Lithology unknown. This interval may be equivalent to the “Rottnest Sandstone” but no samples were taken during drilling. 3. 209-411 m Stark Bay Formation. Early to Middle Miocene. Samples are very uniform in lithology throughout this interval. They consist of moderately recrystal- lised friable white to grey calcarenites with no terrigenous content. Much of the recrystallisation is to pale translucent rhombs which react readily with cold HC1. Much of the original sediment appears to be echinoderm and bryozoan debris. Chert is virtually absent from the section except perhaps in the lower portion. Chert fragments do occur in the ditch cuttings at 427 m. The Faunas. — Samples from the Miocene section in this well all contain almost identical faunas. Cibicides (sensu lato ) makes up 50-90% of all faunas, planktonic species (undiagnostic) occur in only one sample (354 m) where they constitute only 3% of the fauna, and buliminid species make up to 24%. Other elements are chiefly cassidulinids. Benthonic species are identical with those from the Miocene sections identified elsewhere in this work. Thus a com- bination of lithology and benthonic fauna indicates that this is part of the Miocene sequence in the area. The following time division is therefore pos- sible in the well: 73-93 m (approximately) Quaternary “Coastal Limestone” 93-209 m Unknown (? “Rottnest Sand- stone”) 209-411 m Early to Middle Miocene 411-451 m ? Early Eocene Sandstone. Warnbro No. 1 The following samples have been examined: Depth 512-515 m 515 m 546 m 549-552 m Sample Type DC SWC SWC DC Lithology. — All samples are typical Kings Park Formation (Fig. 2) of calcareous shales and siltstones with abundant quartz residue. Glauconite is very common. The sonic log character is consistent to 337 m, where the formation top is placed. An alter- native top, based on gamma ray logs, could be 241 m. The lower boundary at 552 m is based on sonic logs and on lithological and faunal change. The faunas. — The best fauna is in the side- wall core at 515 m which, while diverse, con- tains few age diagnostic species. The age diagnostic species Globorotalia pseudomenardii Bolli and G. chapmani Parr occur in ditch cut- tings at 552 m. Thus a P4 age is indicated. Sugarloaf No. 1 Sugarloaf No. 1 is not listed on the accom- panying diagrams as no in situ Tertiary material was examined from the well. The uppermost sidewall core in the well is at 321 m and this, from palynology, is a Cretaceous argillaceous sandstone. The only evidence of Tertiary seen in material from this well is in ditch cuttings at 366 m. The sample is heavily contaminated with rocks from uphole, including pale brown chert, identical with that common in Unit 1 of the Stark Bay Formation (see below). Thus, while no Tertiary fossils have been identified from the well, the Miocene sediments of the Perth area can be inferred to be present this far south. Stratigraphic units Rocks of at least three different cycles of deposition occur in the area and are best dis- cussed in the context of these cycles. The Palaeocene- Eocene cycle One formation with one new member is recognised in this cycle. The Kings Park Formation (Kings Park Shale of Fairbridge, in Coleman, 1952) occurs quite extensively onshore and was defined from the interval 37-302 m in Kings Park No. 2 Bore. Where present in cores in the metropolitan area, the thickness is usually of the order of 200-250 m but in South Perth Bore it reaches about 450 m and may be even a little thicker in the Claremont area (Fig. 3). Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 27 Offshore thicknesses are 216 m in Warnbro No. 1 and 382 m in the Rottnest Island Bore. Lithology is variable, and for this reason, the name formation is preferred to shale. The dominant rock type appears to be calcareous shale and siltstone, usually glau- conitic, but in places limestone and sandy facies are dominant. Examination of lithology logs presented by Pudovskis (1962) shows that sand- stone is more abundant than indicated in other literature. The fauna includes bryozoa, foraminifera, small molluscs, rare ostracods, common sponge spicules, etc. The age varies from place to place. All on- shore samples seen are Late Palaeocene (P4) on the basis of planktonic foraminifera. This is also the age in Warnbro No. 1 offshore. In the Rottnest Island Bore it is Early Eocene (P6). No rocks of P5 age are yet known and it may be that what is included here in the one for- mation could belong to two deposition cycles with a minor unconformity or disconformity between. However, until this possibility is proven or disproven, all is included in one formation. P5 could very easily be represented by a thin section and be missed in this analysis. Whether P4 or P6 in age, the lithology and fauna are almost identical indicating a similar environment of deposition. Planktonic foramini- fera are not abundant and all other faunal ele- ments are consistent with shallow (perhaps of the order of 30 m or less) water depths, probably in a large embayment or estuary. The Mullaloo Sandstone Member (new name) Type section: WAPET’s Quinns Rock No. 1 in the interval 65-356 m. Coordinates: Latitude 31° 48' 01" S; Longitude 115° 30' 52" E. Derivation of name: The suburb and beach of Mullaloo, approximately the closest land to the well. Thickness : In the type section: 292 m. Else- where: Charlotte No. 1 (411-451 m) 40 m. Lithology: The member consists of poorly sorted, fine to very coarse quartz sandstone with angular to rounded grains. The rock has an argillaceous matrix and commonly is slightly glauconitic. Further details of the lithology are included in the discussion of Quinns Rock No. 1. Extent : The member is so far identified posi- tively only in Quinns Rock No. 1 although the identification in Charlotte No. 1 is also probably correct. The sandstone in that section is of rounded quartz grains, very similar to the type section. As the lithology has not been shown to be widespread, the unit has been defined as a member only. It is the same age as the younger part of the Kings Park Formation and is continuous with it. The consistent, distinctive lithology supports its recognition as a new member. Fauna and age: The only fossils so far re- covered are rare foraminifera, including Globro- talia dolabrata Jenkins which indicates an Early Eocene (P6-P7) age. It is thus a lateral time equivalent of part of the Kings Park Formation. Relations of boundaries: The upper boundary is an unconformity with the Miocene carbon- ates (Stark Bay Formation). The lower boundary is an unconformity with the Lower Cretaceous sediments below. The Miocene Carbonate cycle The full extent of this cycle of deposition is not known. One formation (Stark Bay Forma- tion) is defined to include the white calcaren- ites, brown dolomites and cherts common north and west of Rottnest Island (see below). The relationship of the overlying red and brown calcarenites is unknown, and a new formation name may eventually be necessary to describe that unit. The Stark Bay Formation (new formation) (Fig. 7) Type section: WAPET’S Gage Roads No. 2 in the interval 362-577 m. Coordinates: Latitude 31° 57' 05" S, Longitude 115° 21' 45" E. Derivation of name: Stark Bay, Rottnest Is- land, one of the closest named geographical features. Thickness: In the type section 215 m. Else- where: Gage Roads No. 1 (389-619 m) 230 m. Roe No. 1 (389-616 m) ?227 m. Charlotte No. 1 (209-411 m) ?202 m. Lithology. — The Stark Bay Formation consists dominantly of friable white bryozoan and echino- dermal calcarenite, altered diagenetically in places to brown dolomite and chert, especially in the lower parts of the formation. It formed under marine conditions with virtually no ter- rigenous component. Three subdivisions of the formation can be recognized, each occurring in two or more well sections. These units probably have gradational boundaries. Thicknesses given are approximate only. Unit 1 occurs in Gage Roads No. 1 (442-619 m) and Gage Roads No. 2 (514-544 m). It is dark brown in colour and consists of chert and dolomite in equal parts, or with chert pre- dominant. Unit 2 occurs in Gage Roads No. 1 (389-442 m), Gage Roads No. 2 (427-514 m) and Roe No. 1 (552-616 m). It consists of brown dolomitic limestone with subordinate chert. Unit 3 occurs in Gage Roads No. 2 (362-427 m), Roe No. 1 (9389-544 m) and Charlotte No. 1 (9209-411 m). It is in places an unaltered white friable calcarenite, probably representing the original condition of the whole formation. Else- where, as in Charlotte No. 1, it is somewhat recrystallised. Fauna and age. — The formation contains in places a rich foraminiferal fauna with abundant diagnostic planktonic species. N8 and N9 are the most common ages, but the bottom of the section in Gage Roads No. 1 may be as old as N7. Other elements of the fauna seen are echino- derm and bryozoan fragments. Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 28 Relations of boundaries. — The lower boundary is everywhere an unconformity with either Cre- taceous sediments or the Palaeocene-Eocene Kings Park Formation. The upper boundary is not well known. The formation is overlain by a red and brown re- crystallised, marine limestone of indefinite age and relationship. It may be part of the same cycle of deposition but it may be separated by an unconformity. Repository of comparative material. — Repre- sentative portions of this formation from all sections listed are in the collections of the Bureau of Mineral Resources, Canberra, and the Geological Survey of Western Australia. Unnamed Carbonate Formation Overlying the Stark Bay Formation in Gage Roads No. 1, Gage Roads No. 2 and probably elsewhere, is a poorly sampled, partly recrystal- lised red to brown limestone or dolomitic lime- stone unit. It is known from ditch cuttings at 107-137 m and in a conventional core (Core 1) at about 302 m in Gage Roads No. 1. Between these intervals is a zone of lost circulation so nothing is known from this. This lithology has also been recovered from rock chips on the bit at 305 m in Gage Roads No. 2. The formation is marine, containing foramin- ifera including Operculina and Amphistegina. Its exact age, and the relationships with underlying Miocene carbonates are unknown. It is dis- cussed further below in connection with post- Miocene sedimentation. “Rottnest Sandstone ” McWhae (pers. comm.) used informally the name “Rottnest Sandstone” for sediments be- tween 71m and 284 m in the Rottnest Island Bore. His definition is listed under discussion of that bore. The relationships of the formation are un- known. Drillers’ records (1913) of the drilled interval record “red and brown sands”. This is the record that may be expected of drilling in the unnamed red and brown carbonates above the Stark Bay Formation early in this century. It may be that the “Rottnest Sandstone” has been misinterpreted and it may prove to be an extension of the unnamed carbonate formation or its lateral equivalent. Its possible relation- ships are explored further below. Post-Miocene sediments Scattered deposits in the vicinity of Perth Post-Miocene sediments in the Perth Basin have been little studied to date, but Kendrick (in Darragh and Kendrick, 1971, and pers. comm.) has examined molluscan faunas in the vicinity of Perth from shallow water bores in such places as Bullsbrook, Redcliffe, Kewdale, Gosnells, Jandakot, Peel Estate and the Gnan- gara district. All contain marine faunas indicat- ing at least two periods of sedimentation, none older than Pliocene. The materials from Lake Gnangara is only questionably Pliocene and that from Jandakot is Pleistocene. All the Pliocene localities listed above contain the pelagic gastropod Hartungia typica typica Bronn which seems certainly pre-Pleistocene (Kendrick, pers. comm.). The species ranges back to the Miocene and Kendrick believes that the accompanying mollusc fauna is post- Miocene. Thus, a thin Pliocene marine incursion seems reasonably widespread. The lithology is usually “a grey to yellowish calcareous sand- stone, often with a high proportion of quartz grains and even pebbles of crystalline rock”, (Kendrick, pers. comm., 21/12/1971). No formal stratigraphic terminology has yet evolved. Following the Pliocene, probably disconform- ably, is Pleistocene sediment with the bivalve Zenatiopsis ultima Darragh and Kendrick. The extent of the Pleistocene is as yet unknown and its relationship to other young sediments (e.g., Coastal Limestone) is not yet known com- pletely. For purposes of this work, all the Pleistocene and Pliocene is taken as a single depositional cycle, in turn consisting of several minor episodes of sedimentation. Coastal Limestone Overlying the Tertiary section in coastal regions and to sea is the Quaternary Coastal Limestone consisting of aeolian and marine cal- careous sandstones. This reaches a maximum thickness of 70 m in the Rottnest Island Bore. Elsewhere, it is about 30 m or less. The maximum age so far recorded is 100 000 ± 20 000 years BP (Teichert, 1967). Palaeoecology and palaeogeography Post-Miocene relationships It is clear that the two sequences of river muds are younger than or equivalent to the Coastal Limestone and that the Coastal Lime- stone is very Late Pleistocene to Recent (prob- ablly still forming in several places). It is thus distinctly younger than the Pliocene calcareous sandstones although the environment of forma- tion of the marine Pliocene may be very similar to that of the marine part of the Coastal Lime- stone. The relationship of the Early Pleistocene at Jandakot (Darragh and Kendrick, 1971), the “Rottnest Sandstone” and the unnamed carbon- ate unit overlying the Stark Bay Formation are completely unknown and there is the possibility of lateral equivalence in part. Palaeoecology of the Stark Bay formation While original lithology seems constant over the area covered by this formation, the con- tained faunas indicate quite marked environ- mental differences. Faunas in almost all sam- ples from each well have constant characters but these are different from the characters in samples from neighbouring wells. Figure 5 shows the extent of the formation. In Charlotte No. 1, planktonic percentages are very low (0-3%) throughout the section. In Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 29 Gage Roads No. 2 they are usually in the range 20-30% and in Roe No. 1, 40-70%. Figures for Gage Roads No. 1 are more difficult to obtain accurately but are low, considerably lower than in Gage Roads No. 2. Thus, some consistent bathymetric relationship holds, with the forma- tion in Gage Roads No. 2 and Roe No. 1 representing deeper water, more open marine facies. From planktonic percentages above, the 30 m isobath can be placed roughly between Gage Roads Nos. 1 and 2 and seaward from Charlotte No. 1. Percentages of Bolivina (sensu lato ) in a fauna also provide a depth clue. The only wells analysed with many bolivinids are Charlotte No. 1, Gage Roads No. 2 and Roe No. 1. In Gage Roads No. 2, percentages are between 1 and 3; in Charlotte No. 1, 1-6; and in Roe No. 1, normally 20-40. Again the depth gradient exists between Gage Roads No. 2 and Roe No. 1, with the latter considerably deeper. The high bolivinid and planktonic percentages in Roe No. 1 would indicate outer continental shelf depths of the order of 100 m or more. The fauna in Gage Roads No. 2 is of shallower aspect, but a depth more than 30 m would be expected. The area at Charlotte and Gage Roads No. 1 was certainly inner continental shelf. Palaeoecology in the Palaeocene-Eocene In the Late Palaeocene and Early Eocene, sands with some marine indicators accumulated in the vicinity of Quinns Rock No. 1 (Fig. 3). At the same time, shales, limestones, etc., were deposited in a deep embayment under Perth and to the west of the mouth of this embayment at Rottnest Island and in Warnbro No. 1. It may be coincidence, but the Swan River now, and probably then, crosses the Darling Scarp in Walyunga National Park, due east of Quinns Rock No. 1. It is now, and probably was then, a more significant stream than either the Helena or Canning Rivers which are further south. It is here suggested that in the Late Palaeo- cene-Early Eocene, a significant Swan River flowed west from Walyunga and deposited sandy sediments in the vicinity of Quinns Rock No. 1. At the same time, the Canning and Helena Rivers were smaller streams flowing into a deep embayment and depositing more argilla- ceous sediments. The Swan would have migrated to its present position later. Palaeontologically significant results Large Foraminifera in the area A very interesting feature of the faunas in the upper part of the Stark Bay Formation in Gage Roads No. 2 (and in that well only) is the presence of a few specimens of Lepidocyclina ( Eulepidina) from samples at 399 m, 427 m, 454 m and 491 m. Unfortunately, it is not known from sidewall cores so its exact faunal associates are not known. The specimens are characteristiclly very small — up to about 1.5 mm. It occurs in association with N9 planktonic foraminifera. The signifi- cance of this record of Lepidocyclina is mani- fold. It means that warm water seas existed in the Perth area at the time. This is an extension south of about 1 000 km of the range of this genus and these conditions in W.A. The age of the occurrence seems the same as in south- eastern Australia for L. howchini Crespin, and the two species are very similar. This indicates strongly that warm water seas existed all around the Australian coast at this time. L. howchini has seemed for some years to be isolated from any other Australian occurrences of the genus and its path of migration to south- eastern Australia has been conjectural. One suggestion has been that the tropical climatic zone moved south and that migration was down Australia’s east coast. It is now just as likely that migration was from the west and south coasts. This fits well the present current pat- tern, which probably also existed in the Miocene. The presence of the genus means that the warm water planktonic foraminiferal zonation scheme outlined by Blow (1969) can be used in the Perth Basin at this time. The Age of the Orbulina Datum in Western Australia Much has been written about variations in age of the influx of Orbulina into various sedi- mentary sequences on a worldwide basis. Its absolute base is N9 (Blow, 1969). Its occurrence with Globigerinoides sicanus in Gage Roads No. 2 (BDC at 399 m) shows that the age there is close to the base of its absolute range and the presence of G. sicanus with Praeorbulina tran- sitoria at 579-582 m in the same well indicates that N8 is also present. O. universa may occur in the well as deep as 518 m, suggesting strongly that the absolute base of the range is almost surely represented in the offshore Perth Basin. The presence of Operculina, Amphistegina and Lepidocyclina indicates a warm water environ- ment, so O. universa could be expected to be here at the base of its range. Carter (1964) observed that in Victoria, L. howchini appears stratigraphically below Orbulina universa. In the section studied here, they are coeval and thus three possibilities arise: 1. L. howchini occurs earlier in the Perth Basin than in Victoria. 2. The species is the same age in both localities but the O. universa influx in Victoria is later than in W.A. 3. That the absence of O. universa in Vic- toria is due merely to the normal mutual exclusion of planktonic and large benthonic species. Note added in proof . — Since presenting this paper for publication, the Mines Department of Western Australia has drilled the Claremont Asylum No. 2 bore and recovered six sidewall cores in the Kings Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 30 Park Formation. Three of these sidewall cores (68 m, 391 m and 453 m) contain planktonic foraminiferal faunas. Those from 391 m and 453 m are typical P4 Kings Park Formation faunas but that from 68 m is different. It contains Globorotalia rex, G. cf quetra, G. aequa, and the important zonal species G. pseudo- menardii and G. chapmani are absent. It is probable that this fauna is P6 in age, the same as the Kings Park Formation in the Rottnest Island Bore and in Quinns Rock No. 1. It is thus the first record of rocks of this age onshore in the Perth area. The fauna occurs in a well developed sandstone unit at the top of the Kings Park Formation. This sand- stone probably can be referred to the Mullaloo Sand- stone Member. Acknowledgements. — I must express my indebtedness to the management of WAPET for permission to publish this paper, some material of which would not otherwise be available. Within WAPET, I must mention particu- larly Drs R. G. Alexander and R. A. McTavish and Messrs M. H. Johnstone, P. W. Nygreen and H. L. Ott for constructive criticism and encouragement. Mr Wayne Copley of WAPET is responsible for the photo- graphy using WAPET’S JSM-2 Scanning Electron Micro- scope. Mr J . H. Lord, Geological Survey of Western Australia, and Mr B. S. Ingram now of Burmah Oil Company (Australia), then of the Geological Survey, helped me obtain material from the Rottnest Island Bore. References Barker, R. W. (1960). — Taxonomic notes. Spec. Publns Soc. econ. Paleont. Miner, Tulsa, 9. Berggren, W. A. (1965). — The recognition of the Globoro- talia uncinata Zone (Lower Paleocene) in the Gulf Coast. Micropaleontology 11 (1): 111-113. (1971). — Multiple phylogenetic zonations of the Cenozoic based on planktonic foramin- ifera. Proc. 2nd InternaVl Conf. Planktonic Microfossils, Rome, 1: 41-56. Blow, W. H. (1969). — Late Middle Eocene to Recent planktonic foraminiferal biostratigraphy. Proc. 1st Internat’l Conf. Planktonic Micro- fossils, Geneva, 1: 199-422. Bolli, H. (1957). — The genera Globigerina and Globoro- talia in the Paleocene-Lower Eocene Lizard Springs Formation of Trinidad, B.W.I. Bull. U.S. Natn. Mus., 215: 61-82. Bozanic, D. (1969). — Well Completion Report, Quinns Rock No. 1. Petrol. Search Subsidy Acts Publn, Bur. Min. Resour. Geol. Geophys. Aust. Carter, A. N. (1964). — Tertiary foraminifera from Gipps- land, Victoria and their stratigraphical sig- nificance. Mem. Geol. Surv. Viet. 23. Cockbain, A. E., and B. S. Ingram (1967). — The age of the Kings Park Shale in the Rottnest Island Borehole. Palaeont. Rep. Geol. Surv. West Aust. 29/67 (unpublished). Coleman, P. J. (1952). — Foraminiferal investigations in the Perth Basin, Western Australia. J. Proc. R. Soc. West. Aust. 36: 31-43. Cookson, Isabel C., and A. Eisenack (1961). — Tertiary Microplankton from the Rottnest Island Bore, Western Australia, ibid. 44: 39-47. Darragh, T. A., and G. W. Kendrick (1971). — Zenatiopsis ultima sp. nov. terminal species of the Zenatiopsis lineage (Bivalvia: Mactridae). Proc. R. Soc. Viet. 84: 87-92. Glaessner, M. F. (1956). — Crustacea from the Creta- ceous and Eocene of Western Australia J. Proc. R. Soc. West. Aust. 40: 33-35. Jenkins, D. G. (1971). — New Zealand Cenozoic plank- tonic foraminifera. Palaeont. Bull., Welling- ton, 42. Kendrick, G. W. (1960). — The fossil Mollusca of the Peppermint Grove Limestone, Swan River district of Western Australia. West. Aust. Nat. 7 (3): 53-66. McGowran, B. (1964). — Foraminiferal evidence for the Paleocene age of the Kings Park Shale (Perth Basin, Western Australia). J. Proc. R. Soc. West. Aust. 47: 81-86. McWhae, J. R. H., P. E. Playford, A. W. Lindner, B. F. Glenister, and B. E. Balme (1958). — The Stratigraphy of Western Australia. J. Geol. Soc. Aust. 4 (2): 1-161. Parr, W. J. (1938). — Upper Eocene foraminifera from deep borings in Kings Park, Perth, Western Australia. J. Proc. R. Soc. West. Aust. 24: 69-101. Pudovskis, V. (1962). — Subsurface geology of the Perth Metropolitan area. WAPET Report (unpub- lished). Teichert, C. (1967). — Age of Coastal Limestone, Western Australia. Aust. J. Sci. 30 (2) : 68, 69. Wells, J. W. (1943). — Note on fossil corals from Langley Park Bore, Perth. J. Proc. R. Soc. West. Aust. 27: 95-96. Journal of the Royal Society of Western Australia, Vol. 57, Part 1, May, 1974. 31 INSTRUCTIONS TO AUTHORS Contributions to this Journal should be sent to The Honorary Editor Royal Society of Western Australia , Western Australian Museum , Perth. Papers are received only from, or by communication through, Members of the Society. The Council decides whether any contribution will be accepted for publication. All papers accepted must be read either in full or in abstract or be tabled at an ordinary meeting before publication. Papers should be accompanied by a table of contents, on a separate . sheet, showing clearly the status of all headings; this will not necessarily be published. Authors should maintain a proper balance between length and substance, and papers longer than 10,000 words would need to be of exceptional importance to be considered for publication. The Abstract (which will probably be read more than any other part of the paper) should not be an expanded title, but should include the main substance of the paper in a con- densed form. Typescripts should be double-spaced on opaque white foolscap paper; the original and one carbon copy should be sent. All Tables, and captions for Figures, should be on separate sheets. Authors are advised to use recent issues of the Journal as a guide to the general format of their papers, includ- ing the preparation of references; journal titles in references may be given in full or may follow any widely used conventional system of abbreviation. Note that all illustrations are Figures, which are numbered in a single sequence. In composite Figures, made up of several photographs or diagrams, each of these should be designated by letter (e.g. Figure 13B). Illustrations should include all necessary lettering, and must be suitable for direct photo- graphic reproduction. To avoid unnecessary handling of the original illus- trations, which are usually best prepared between li and 2 times the required size, authors are advised to supply extra prints already reduced. Additional printing costs, such as those for folding maps or colour blocks, will normally be charged to authors. It is the responsibility of authors to adhere to the International Rules of Botanical and Zoological Nomenclature. Palaeontological papers must follow the appropriate rules for zoology or botany, and all new stratigraphic names must have been previously approved by the Stratigraphic Nomenclature Com- mittee of the Geological Society of Australia. Thirty reprints are supplied to authors free of charge, up to a maximum of 60 for any one paper. Further reprints may be ordered at cost, provided that orders are submitted with the returned galley proofs. Authors are solely responsible for the accuracy of all information in their papers, and for any opinion they express. Journal of the Royal Society of Western Australia Volume 57 1974 Part 1 Contents 1. Australian Aborigines: Research and Welfare. Presidential Address, 1973. By R. M. Berndt. 2. Clastic dykes at Albany, Western Australia. By J. G. Kay. 3. The pouch of Planigale subtilissima and other dasyurid marsupials. By P. Woolley. 4. Cainozoic stratigraphy in the Perth area. By P. G. Quilty. Editor: A. J. McComb The Royal Society of Western Australia, Western Australian Museum, Perth 34999/1/74—625 WILLIAM C. BROWN, Government Printer, Western Australia JOURNAL OF THE ROYAL SOCIETY OF WESTERN AUSTRALIA VOLUME 57 PART 2 AUGUST, 1974 PRICE: TWO DOLLARS REGISTERED FOR POSTING AS A PERIODICAL-CATEGORY B THE ROYAL SOCIETY OF WESTERN AUSTRALIA PATRON Her Majesty the Queen VICE-PATRON His Excellency Air Commodore Sir Hughie Edwards, V.C., K.C.M.G., C.B., D.S.O., O.B.E., D.F.C. COUNCIL 1974-1975 President Vice Presidents Past President Joint Hon. Secretaries Hon. Treasurer Hon. Librarian Hon. Editor G. A. Bottomley, B.Sc., Ph.D. B. E. Balme, D.Sc. P. R. Wycherley, O.B.E., B.Sc., Ph.D., F.L.S. A. F. Trendall, B.Sc., Ph.D., A.R.C.S., F.G.S. M. Perry, B.Sc. (Agric.) (Hons.) G. Perry, B.Sc. (Hons.) S. J. Curry, M.A. A. Neumann, B.A. A. J. McComb, M.Sc., Ph.D. R. M. Berndt, M.A., Dip.Anth., Ph.D., F.R.A.I., F.F.A.A.A. C. E. Dortch, B.S., M.Phil. L. J. Peet, B.Sc., F.G.S. P. E. Playford, B.Sc., Ph.D. P. G. Quilty, B.Sc. (Hons.), Ph.D. J. C. Taylor, B.Sc., Ph.D., A.R.C.S. P. G. Wilson, M.Sc. 5. — The petrology of the Wooramel Group (Lower Permian) in the Lyons River area, Carnarvon Basin, Western Australia. by Gerard J. McGann 1 Manuscript received 18 July 1972; accepted 19 February 1974 Abstract In the Lyons River area, about 100 miles east of Carnarvon, Western Australia, the Wooramel Group (Lower Permian) comprises the Moogoo- loo Sandstone and the Billidee Formation. The Moogooloo Sandstone, overlying the Callytharra Formation with a conformable, disconformable or faulted contact, consists of orthoquartzite, often ferruginized, with minor orthoconglome- rates. Gypseous carbonaceous shales are present towards the top of the formation. The Billidee Formation, resting conformably on the Moogooloo Sandstone consists of fine to medium grained orthoquartzite and calcite- cemented sandstone, interbedded with gypseous carbonaceous shale. A laterally-persistent, fos- siliferous unit is present at the top of the formation. Introduction The name “Wooramel Group” was first pro- posed by Condit (1935) and revised by Konecki et al. (1958) who described it as . . predo- minantly arenaceous sequence disconformably or unconformably overlying the Callytharra For- mation, Lyons Group or Precambrian schist, or overlain conformably by, or changing laterally into the Byro Group.” In the Lyons river area, only two formations of the Wooramel Group, the Moogoolo Sandstone and the Billidee Formation are exposed. No previous detailed work has been done on the Wooramel Group in the Lyons River area, although the area has been re- gionally mapped by the Bureau of Mineral Re- sources (Condon 1954, 1962, 1967). An area about 110 miles east of Carnarvon, Western Australia (Fig. 1) was mapped in 1970. Rocks exposed are the Precambrian rocks of the Weedarra Inlier, and Permian rocks — the glacigene Lyons Group, the Callytharra Forma- tion and the overlying Wooramel Group. Map- ping was done by numerous ground traverses with the aid of aerial photographs. Twelve stra- tigraphic sections were measured. About 70 thin sections were examined with a petrological mi- croscope. Some friable rocks were impregnated with plastic or Canada balsam before sectioning. Sediments are classified according to Pettijohn (1957). Topography The topography of the Lyons River area closely reflects the distribution of rock types. The Moogooloo Sandstone outcrops strongly as a strike ridge and dip slope, or as mesas and 1 Geology Department, University of Western Aus- tralia, Nedlands, W.A. 6009. Present address: Exploration Logging of Australia; Burswood Avenue, Victoria Park, W.A. 6100. buttes. The sandstone lithologies of the Billidee Formation outcrop as low strike ridges, and the interbedded shales do not outcrop. Stratigraphy The Lower Permian stratigraphy of the Lyons River is summarised in Fig. 2. The outcrop pat- tern and inferred subsurface extent of the Wooramel Group which, in the Lyons River area, comprises the Moogooloo Sandstone and the Billidee Formation, is shown on Fig. 3. Structure and Tectonic History The area mapped is in the south-easterly corner of the Merlinleigh Basin and in the west of the Bidgemia Basin, both sub-basins in the eastern Carnarvon Basin. The Weedarra In- lier (Precambrian) is the dominant structural element in the area mapped. Permian sediments are down faulted against the inlier by major faults. The Permian sediments strike north- south and dip to the west at about 5 degrees. Faulting is common in the Permian sediments. There are two dominant trends: strike faults, and faults trending at about 280 degrees. Most of these faults have a throw of less than 46 m. Petrology of the Wooramel Group In the Lyons River area, the Wooramel Group comprises the Moogooloo Sandstone and the overlying Billidee Formation. The group is dated as Artinskian (Early Permian). Moogooloo Sandstone The Moogooloo Sandstone consists of red, brown and white fine to coarse-grained ortho- quartzites; subordinate arkoses (feldspathic sandstones) and orthoconglomerates. Minor in- terbedded silty shales are present in the upper part of the formation. Orthoquartzite is the dominant lithology, grad- ing compositionally into subarkose. Grain size ranges from 0.05 mm to 1.8 mm, with the majority of grains in the 0.2 to 0.5 mm range. Secondary overgrowth has resulted in many ori- ginally rounded grains becoming subangular. Sorting is moderate to good. The dominant mineral in the sandstone suite is quartz, commonly with undulose extinction. Syntaxial growth on quartz grains is present in every sample examined. Pressure welding is ob- served in some juxta-positioned grains. Inclu- sions in quartz include tourmaline, zircon and muscovite. Microline is the only feldspar pre- sent, and ranges in abundance from almost nil Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. 33 Figure 1 — Locality map. Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. 34 STRATIGRAPHIC COLUMN LYONS RIVER AREA Coyrie Formation Billidee Formation Moogooloo Sandstone Callytharra Formation Lyons Group Precambrian FEET 1-0 KEY SANDSTONE LIMESTONE CONGLOMERATE SILTSTONE SHALE ERRATICS ED ED S El Figure 2 — Stratigraphy of the Lyons River area. to 9% of the detrital fraction. Lithic fragments are present in over 60% of the orthoquartzites, and in all of the subarkoses examined, fine- grained metaquartzite being by far the most abundant. Muscovite is a common detrital grain, in plates up to 1.3 mm long. Light-brown bio- tite is a common accessory. No systematic study of the heavy mineral suite was undertaken, but a number of heavy mineral samples were examined. Euhedral to subhedral zircon and tourmaline are the most abundant minerals. Tourmaline is present in pleochroic yellow-green and dark grey-pink va- rieties. Both pink and colourless garnet are pre- sent. Rutile, sphene and staurolite are minor accessories. The detrital opaque minerals are ilmenite, with partial leucoxene coatings, and minor limonite. The matrix is dominated by kaolinite, either coating grains or filling all interstitial spaces. In some samples, the clay has been completely ferruginized, resulting in a rock with a high proportion (up to 22%) of hematite with some limonite. Chlorite, sericite, microcrystalline quartz, gypsum, collophane and calcite are all minor, probably authigenic, constituents of the matrix. Orthoconglomerates are volumetrically a very minor constituent of the Moogooloo Sandstone, occurring in beds up to 30 cm thick which per- sist laterally for up to 120 metres. The conglo- merates are poorly sorted, the grain size ranging from 0.02 mm to 4 cm. The dominant detrital mineral is quartz, with minor secondary over- growths. Other minor detrital grains are fresh microcline feldspar and lithic fragments. The matrix is made up entirely of hematite and minor limonite. Dark-grey to brown, friable, carbonaceous silty shales occur near the top of the formation in beds up to 1.8 metres thick. The shales are interbedded with coarse-grained orthoquartzites and conglomerates containing the trace fossil Palaeophycus sp. The shales are poorly lamin- ated and contain up to 15% carbonaceous matter. Silt-sized detrital grains, mainly quartz, are concentrated in lamellae up to 0.1 mm thick. Gypsum accounts for up to 12% of the mode, in laths up to 50 mm long, and is thought to be a primary mineral, although some recrystalliza- tion has occurred. Slight ferruginization has partially welded the clay-sized particles together. Diagenesis is here discussed under three headings: (a) Constructive diagenesis. (b) Destructive diagenesis. (c) Compactional effects. The discussion is here confined to the more abundant sandstone suite. (a) Constructive diagenesis. Under this cate- gory are grouped all diagenetic mineral growths and enlargements. The development of syntaxial Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. 35 rims on quartz is the most common of these effects. Other diagenetic minerals are kaolinite, hematite and limonite, sericite, collophane, mi- crocrystalline quartz, chlorite and calcite. He- matite and limonite were deposited after kao- linite, growing preferentially in the kaolinitic clay matrix. (b) The principal destructive effect is the etching of quartz and feldspar by kaolinite. Hematite and limonite, where abundant, also etch quartz and feldspar. Tourmaline suffers noticeable solution, only fragments of once large detrital grains remaining. (c) Compactional effects observed are bend- ing of muscovite plates around detrital grains, and pressure welding of quartz grains at points of contact. The material which has been re- moved from areas of solution is thought to have been redeposited as syntaxial growths. Billidee Formation The Billidee Formation conformably overlies the Moogooloo Sandstone, and consists of ortho- quartzites and calcite-cemented sandstones. Gypseous, carbonaceous silty shales, similar to the shales of the Moogooloo Sandstone, make up about 45% of the formation. Rare siltstones, often calcareous, and calcareous conglomeratic sandstones are also present. None of the units within the formation has any lateral persis- tence except the uppermost calcareous sandstone unit which contains abundant pelecypods. This is the only unit which can be traced along strike for more than 460 m. The uppermost unit is up to 6 m thick, persists throughout the whole of the area mapped, and is the only mappable unit in the formation. Fossils (apart from fos- sil wood) are found only in this unit. All of the sandstones are thinly bedded, with bedding thicknesses ranging from 2.50 to 60 cm. They are predominantly fine to medium grained, the mean grain size being 0.2 mm. Nearly all of the sandstones are moderately well sorted, with moderate to poor rounding and low sphe- ricity of detrital grains. Quartz, often with undulose extinction, is the dominant mineral, averaging 60% of the mode. Fresh and slightly altered microcline comprises up to 40% of the mode, in grains up to 1.5 mm in diameter. Muscovite is common accessory, in plates, often bent by pressure, up to 0.4 mm long. Light-brown, slightly pleochroic biotite is present in some specimens. In one fine-grained orthoquartzite, recrystallized primary gypsum is interbedded with detrital quartz. Rare limonite pseudomorphs after pyrite are present in grains up to 1 cm in diameter. Oligoclase is present in one thin section. Lithic fragments present in the sandstones in- clude chert and occasional fragments of quartz- muscovite schist up to 14 cm in diameter, in the uppermost unit of the formation. Rare car- bonate lithic fragments occurr, some containing unidentified Bryozoa, punctate brachiopods (perhaps Permorthotetes sp.) and also an un- identified calcareous alga. The lithic fragments are partially replaced by sparry calcite in iso- lated, often lenticular, patches. The carbonate lithic fragments contain between 3% and 20% sandsized detrital grains admixed with the car- bonate, and are extensively ferruginized and cut by ferroan calcite veins. Rare ferroan calcite lithic fragments containing quartz and authi- genic microcline are present in some calcite-ce- mented sandstones. The non-opaque heavy mineral suite, in order of abundance, is tourmaline, zircon, colourless and pink garnet, rutile, sphene, staurolite and barite. The opaques are limonite and minor ilmenite. There is great variety in the matrix of the sandstones of the Billidee Formation, with cal- cite, ferric oxides, clay minerals and chlorite all being major constituents in different rocks. Calcite cement is present in a large number of the sandstone bodies within the formation. The cement consists of crystalline ferroan cal- cite, with patches of non-ferroan calcite and admixed clay minerals. The ferroan calcite ce- ment consists of interlocking crystals, some up to 9 mm in diameter. In many of the calcite- cemented sandstones, the detrital grains are “floating” unsupported in the calcite matrix. Evidence of pressure solution is widespread, and suggests that the sediment was grain supported when deposited, with a relatively high proportion of detrital skeletal fragments. Pressure solution has been applied, either by sedimentary load or by tectonic stress, and has resulted in the solu- tion of the carbonate skeletal fragments. Dis- solved carbonate has been reprecipitated as fer- roan calcite, producing the unsupported “float- ing” texture. The iron necessary for the pro- duction of ferroan calcite may have been derived from clay minerals, as outlined by Oldershaw and Scoffln (1967). Some skeletal fragments have escaped solution, especially in the upper- most unit of the formation. Non-ferroan calcite is a minor constituent of the matrix, occurring as coatings, up to 0.03 mm thick, on detrital grains, and as isolated patches within ferroan calcite. There are two possible explanations for the origin of this non-ferroan calcite, the first being that the patches are rem- nants of original skeletal grains that have escaped pressure solution, and thus have not been converted to ferroan calcite. The second alternative is that calcite has been dissolved by pressure solution and reprecipitated, but because of an oxidizing micro-environment around de- trital grains and in isolated patches, non-fer- roan calcite has been deposited instead of fer- roan calcite. Ferroan calcite can be deposited only in reducing conditions (Evamy, 1969). The second explanation is favoured by the author. Ferric oxides are abundant in the matrix of some orthoquartzites, with hematite and limonite occupying up to 35% of the rock. The average is, however, about 4%. Minor clay minerals are common in the matrix, occurring in isolated patches, often partially ferruginized, Colourless chlorite is a common matrix mineral, in scalar aggregates up to 1 mm in diameter, occupying up to 10% of the rock, the average being about 1%. The matrix also contains minor gypsum, in scalar aggregates. Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. 36 Pressure effects are abundant in the Billidee Formation, with pressure welding of quartz and microcline grains, bending of micas and pres- sure solution of calcite. Quartz, and some mi- crocline grains show secondary overgrowths, the material for the overgrowth probably being de- rived from pressure solution. In sandstones with abundant calcite cement, authigenic secondary growths are rare, and etching of quartz is common. The fossils of the uppermost member of the Billidee Formation are recrystallized into blocky calcite, sparry in places. Dark lamellar organic matter (conchilium) is still present in many val- ves of Oriocrassatella sp., and is seen to pass through the sparry calcite, indicating that the spar is formed by replacement. Oriocrassatella sp., is also partially replaced by ferroan calcite in quite irregular patches. Replacement of cal- cite by ferroan calcite has, to the author’s knowledge, not previously been recorded in the literature. Borings, perhaps algal, are common in some pelecypods in the uppermost unit in the forma- tion, penetrating up to 0.4 mm into the valve. The borings are infilled with ferroan calcite, of different composition (less iron) than the re- placement ferroan calcite mentioned above. Late-stage diagenetic effects observed in the Billidee Formation are ferruginization, sparry calcite veins and sparry infillings of fossils, together with the development of calcrete. Acknowledgements. — The author is deeply indebted to WAPET Pty. Ltd. and to the University of Western Australia for funding the project. Sincere thanks must also go to Dr. B. W. Logan and members of the Uni- versity staff, fellow honours students, and also the manager and staff of Lyons River Station. References Condit, D. D. (1935). — Oil possibilities in the North- West districts, Western Australia. Econ. Geol. 30 (8): 860-878. Condon M. A. (1954). — Progress report on stratigraphy and structure of the Carnarvon Basin, West- ern Australia. Bur. Miner. Rescur. Geol. Geophys. Aust., Report 15. Condon, M. A. (1962). — Kennedy Range, W.A. — 1 : 250 000 Geological Series. Bur Miner. Resour. Geol. Geophys. Aust. Explan. Notes. 28 pp. Condon, M. A. (1967).— The Geology of the Carnarvon Basin, Western Australia. Part 2: Permian Stratigraphy. Bur Miner. Res. Geol. Geophys. Aust. Bull. 77. Evamy, B. D. (1969). — The precipitational environment and correlation of some calcite cements de- duced from artificial staining. J. Sediment Petrol. 39 (2): 787-793. Konecki, M. C., Dickins, J. M. and Quinlan, T. (1958). — The Geology of the coastal area between the lower Gascoyne and Murchison Rivers, West- ern Australia. Bur. Miner. Resour. Geol. Geophys. Aust., Report 37. Oldershaw, A. E. and Scoffin, T. P. (1967). — The source of ferroan and non-ferroan calcite cements in the Halkin and Wenlock Limestones. Geol. J. 5 (2): 309-320. Pettijohn, F. J. (1957). — “Sedimentary rocks”. Harper Row, New York. Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. 37 6. — An annotated list of lichens from the coastal limestone near Perth, Western Australia by N. C. Sammy 1 and G. G. Smith 1 Manuscript received 18 July, 1972; accepted 16 April, 1974. Abstract Records of distribution are given for thirteen lichen species of the genera Aspicilia, Buellia, Caloplaca, Dermatocar-pon, Diplo schist es, Ful- gensia, Lecidea, Leptogium, Toninia, Verrucaria and Xantlioria occurring in the limestone belt along thirty miles of coastline near Perth, Western Australia. The habitat preferences and frequency of each species are evaluated. Introduction The Swan Coastal Plain is a narrow strip of sandy country along the coast of South-Western Australia. It is composed almost entirely of fluviatile and eolian sediments derived originally from the PreCambrian land mass, the western edge of which is called the Darling Scarp. The western border of the plain is composed largely of the Spearwood Dune System of Pleistocene age and is partially overlain on its seaward edge by a very narrow coastal strip of Post-Glacial and contemporary sands of the Quindalup Dune System (McArthur and Bettenay 1960, Fair- bridge 1950). The Spearwood Dunes have a core of cal- careous eolianite (locally called the Coastal Limestone) and residual brown and yellow sands resulting from leaching of the carbonate from the rock core. The limestone is heterogeneous in that it has varying proportions of siliceous grains and calcareous fragments of marine organisms bound by fine calcareous cement. Much of this limestone is porous, friable and often stratified and cross-bedded (Figure 1). Fine grained travertines have been secondarily deposited in the eolianite and are exposed abundantly along the coast as karst features in the forms of hard, gray mantles, solution pipes and pinnacles (Figure 2). Aerial erosion of the eolianite and the travertines results in pitted and jagged rock surfaces. The sea cliifs mostly have the more highly incised form of weather- ing, especially in the extreme upper littoral where both physical and biological agencies of marine erosion are strongly active (Figure 3). This much-dissected form of the Coastal Limestone with its richly convoluted surfaces provides a large variety of ecological niches for lichens occurring upshore from the splash zone of the extreme upper littoral where Verrucaria maura Wahl., occurs. The limestone exposures of the metropolitan sector of the coast were investigated for their lichen flora, the region sampled including some thirty miles of the coast 1 Botany Department, University of Western Australia, Nedlands, W.A. C009. between Cape Peron South (locally called Point Peron) and North Beach. Also included in the survey were limestone sites at Garden and Rott- nest Islands near Fremantle, limestone cliffs of the Swan River Estuary and a single outcrop at Point Mount Henry on the Canning River (Figure 4). Some of the species on limestone also occur on wood and bark of trees and shrubs in the vicinity, and these occurrences are also recorded. Thirteen species of lichens were collected from the study area, and of these the most abundant were Buellia alboatra (Hoffm.) Branth. and Rostr., Fulgensia bracteata (Hoffm.) Ras., and Xanthoria ectanea (Ach.) Ras. ex R. Filson. The voucher specimens cited here are kept in Figure 1. — Limestone cliffs at Point Mount Henry, showing bedding planes. Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. Journal of tlie Royal Society of We. tern Australia, Vol. 57, Part 2, August. 1974 39 Collemaeeae Leptogium phyllocarpmn (Pers.) Mont., Ann. Sci. Nat. Bot. Ill, 10: 134 (1848). Ccllema phyllocarpum Pers. in Gaud. Voyag. Uranie, 2G4 (1826). Rare. Rottnest I., on limestone at Lake Her- scheil (UWA 1096) ; North Beach, in mess swards on limestone (UWA 717); Point Mount Henry, on trunk of Tuart tree (Eucalyptus gompho- cephala DC.) (UWA 607) ; Kings Park, on trunks of Tuart trees (UWA 1009); Yanchep, in moss swards on limestone at Silver Stocking Cave (UWA 1030). Common on friable limestone and on shallow soil overlying limestone, usually in shaded sites. Limestone cliffs of the Swan River Estuary at Kings Park (UWA 733), Rocky Bay (UWA 793, UWA 1029) and Point Roe (UWA 794). Rott- nest I., common on limestone about the salt lakes (UWA 1097, UWA 1087). Diploschistaeeae Diploschistes scruposus (Schreb.) Norm, in Nyt Magazin for Naturv. 7: 232 (1853). Urceolaria scruposa (Schreb.) Ach. in Lich. Suec. 32 (1798). Lichen scruposus Schreb., in Spic. FI. Lips. 133 (1771). Common on limestone cliffs of the Swan River Estuary at Kings Park (UWA 626) and Rocky Bay (UWA 793, UWA 795). Figure 2. — Travertine limestone at North Beach with Buellia albcatra. This limestone is a few feet above high water mark of winter storm tides (beach in the background) . the Herbarium of the Botany Department, Uni- versity of W. A. (UWA). The classification used below is that of Mat- tick (1954). ASCOLICHENES V errucariaceae Verrucaria maura Wahl., in Ach. Meth. Lich. Suppl. 19. (1803). Occasional, as a black to dirk green, encrusting growth immediately above the littcrinid zone of Melaraphe unifasci- aia (Gray) in the splash and spray zone of the extreme upper littoral of limestone sea cliffs. Cape Peron South (UWA 666, UWA 624); Gar- den I., at the southern end (UWA 1044) and at Entrance Pcint (UWA 1049, UWA 1050); Rott- nest I., at Radar Reef (UWA 1093); Cottesloe, at Mudurup Reef (UWA 1397); North Beach (UWA 714). Lecideaceae Lecidea decipiens (Ehrh.) Ach. Method. Lich. 80 (1803) and Synops. Lich. 53 (1814). Psora decipiens (Ehrh.) Hcffm. in Descr. PI. Crypt. 2: 63 (1794). Biatcra decipiens (Ehrh.) E. Fries in Lichen. Europ. Reform. 252 (1831). Lichen decipiens Ehrh. in Hedw. Descr. et Adumbr. Musccr. Frcndos 2: 7 (1789). Occasional on soil overlying limestone out- crops along the Swan River Estuary, Rocky Bay (UWA 798) ; Rottnest I., on indurated limestone at Lake Herschell (UWA 1086). Preiss collected Dermatocarpaceae Dermatccarpon hepaficum (Ach.) Th. Fries in Nova Acta Reg. Soc. Scient. Upsalla, ser. 3, VII: 355 (1861). Endzcirpcn hepaticum Ach. in Kgl. Ve\ Akc.d. Nya Handl., 156 (1809); and Lichen. Univers. 298 (1810). Figure 3.— Sea cliff at Mudurup Reef, Cottesloe, at low tide, showing incised form of weathering. Verrucaria maura occurs sparsely in the supra-littoral below the visor or overhang of the cliff. Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. 40 this species from sandy soil near a lake at Rottnest I. (Fries 1846). Lecidea plana (Lahm) Nyl. in Flora 55:552 (1872). Lecidella plana Lahm in Koerb. Par. Lich. 211 (1865). Frequent on coastal limestone, but easily over- looked. Point Roe in the lower Swan River Estuary (UWA 799) ; Cape Peron South (UWA 1051); Rottnest I., on indurated limestone along the banks of Lake Herschell (UWA 1085). Toninia cumulata (Sommerf.) Th. Fries in Lich. Scand. I: 341 (1874). Lecidea cumulata Sommerf., in Suppl. Flora Lapp. 157 (1826). Lecidea conglomerata Sommerf., in Egl. Vidensk. Shrifter II; 54 (1827). Occasional on limestone at Cape Peron South (UWA 1052) ; North Beach (UWA 1053); Garden I.; Rottnest I., on friable limestone sea cliffs at Radar Reef (UWA 1089, UWA 1054); Point Mount Henry (UWA 606). Lecanoraceae Aspicilia calcarea (L.) Mudd, Manual Brit. Lich. 161, tab. Ill, (1861). Lecanora calcarea (L.) Sommerf. in Suppl. Flor. Lappon. 102 (1826). Lichen calcareus L. in Sp. PI. 1140 (1753). Garden I., abundant on sea cliffs at Collie Head (UWA 1036, UWA 1040, UWA 1041); Rott- nest I., sea cliffs at Geordie Bay (UWA 1083), limestone outcrops at Serpentine Lake (UWA 1098) and indurated limestone at Lake Herschell (UWA 1084). Apparently rare on the mainland. Cape Peron South, on limestone cliff (UWA 664). Caloplacaceae Caloplaca aurantiaca (Lightf.) Th. Fries in Nova Acta Reg. Soc. Scient. Upsalla, ser. 3, III: 219 (1861); and Lichen. Scandin. I: 177 (1871). Lichen aurantiacus Lightf., Flora Scotica II: 810 (1777); edit. 2, 810 (1789). Common on limestone and also on bark and wood of coastal shrubs. North Beach (UWA 716); Scarborough (UWA 1027); limestone cliffs of the Swan River Estuary at Point Roe (UWA 796) ; Blackwall Reach (UWA 1028) ; Rocky Bay (UWA 795); Kings Park (UWA 849); Garden I., on cliffs at Beacon Head (UWA 1047, UWA 1048) ; Rottnest I., abundant on limestone about the salt lakes (UWA 1092) and on sea cliffs at Radar Reef (UWA 1088). Fulgensia bracteata (Hoffm.) Ras., in Die Flecht. Estl. I: 108 (1931). Caloplaca bracteata (Hoffm.) Jatta in Sylloge Lich. Ital., 236 (1900). Psora bracteata Hoffm. in Deutschl. Flora 2: 169 (1796). Common and often abundant on limestone cliffs at Kings Park (UWA 1045) ; Point Mount Henry (UWA 608, UWA 611); Rocky Bay (UWA 793) ; Rottnest I., on friable limestone sea cliffs (UWA 861) and on limestone outcrops at Lake Herschell (UWA 1095). Teloschistaceae Xanthoria ectanea (Ach.) Ras ex R. Filson in Muelleria 2: 65 (1969). Xanthoria ectanea (Ach.) Ras in An. Soc. Scient. arg. Secc. S Fe 131: 103 (1941). Xanthoria parietina var ectanea (Ach.) Kickx, in Flore Cryptog. Flandres, 2: 228 (1867). Parmelia parientina var ectanea Ach. in Lichen. Univ. 464 (1810). Locally abundant on limestone close to the sea and on bark and dead wood of coastal shrubs. Cape Peron South, on limestone (UWA 665, UWA 667, UWA 789) ; Garden I., at South West Point (UWA 1039) ; Rottnest I., abundant on indurated limestone about the salt lakes (UWA 1090) ; North Beach, on dead wood (UWA 715). Xanthoria parietina (L.) Beltr. in Lichen. Bassan. 102 (1858). Lichen parietinus L. in Sp. PI. 1143 (1753). Occasional on limestone near the sea but more frequent on dead wood and bark of coastal shrubs. North Beach, on bark (UWA 694) ; Rottnest I., on dead wood (UWA 308) ; Garden I., on bark (UWA 236). Buelliaceae Buellia alboatra (Hoffm.) Branth. and Rostr., in Botan. Tidskrift, IV: 239 (1869). Verrucaria alboatra Hoffm. Descript et Adumbr., Plant Lich. I, 76, tab. XV (1790), fig. 2 et Deutschl. Flora 193 (1796). Lichen alboater Hoffm. Enum. Lich. 30 (1784). Common on limestone close to the sea, becom- ing less abundant inland. North Beach (UWA 792); Cape Peron South (UWA 790); Garden I., at Beacon Head and elsewhere on limestone (UWA 1038, UWA 1042); Rottnest I., abundant on travertine limestone about the salt lakes (UWA 1094, UWA 1091) ; Point Roe in the Swan River Estuary (UWA 791, UWA 794) ; Point Mount Henry on the Canning River Estuary (UWA 605, UWA 609, UWA 612, UWA 791, UWA 794). Acknowledgements. — We thank Professor W. A. Weber of the University of Colorado Museum, Colorado, for identifying specimens of Aspicilia calcarea and Fulgensia bracteata. Mr. R. Filson of the National Herbarium of Victoria, South Yarra, kindly verified several of our determinations. References Bibby, P. and G. G. Smith (1955). — A list of Lichens of Western Australia. J. R. Soc. W. Aust. 39: 28-29. Fairbridge, R. W. (1950). — The geology and geomor- phology of Point Peron, Western Australia. J. R. Soc. W. Aust. 34: 35-72. Filson, R. (1969). — A review of the genera Teloschistes and Xanthoria in the lichen family Telo- schistaceae in Australia. Muelleria 2: 65-115. Fink, B. (1935). — “The lichen flora of the United States.” Univ. Michigan Press, Michigan. Fries, E. (1846-47). — Lichenes, in “Plantae Preissianae sive enumeratio -plantarum quas in Aus- tralasia occidentali et meridionali-occidentali annis 1834-41”, Vol. II: 140-145, collegit Ludi- vicus Preiss. Edit. C. Lehmann. Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. 41 Lamb, I. M. (1963). — “ Index Nominum Lichenum inter annos 1932 et 1960 divulgatorum” Roland Press, New York. Mattick, F. (1954) .— Lichenes, in A. Engler’s “Syllabus der Pflanzenjamilien”, Band 1, Edit. H. Melchior and E. Werderman. Borntraeger, Berlin-Nikolassee. McArthur, W. A. and Bettenay, E. (I960).— The develop- ment and distribution of the soils of the Swan Coastal Plain, Western Australia. C.S.I.R.O. Soil Publ. No. 16. Zahlbruckner, A. (1921-1940). — “Catalogus lichen univer- salis” Vol. 1-10. Borntraeger, Leipzig and Berlin. Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. 42 7. — New information about the Quaternary distribution of the thylacine (Marsupialia, Thylacinidae) in Australia by M. Archer 1 Manuscript received 22 May 1973; accepted 18 September 1973 Abstract Thylacine remains and associated faunas are reported from cave deposits in the Kimberley Division, the South Wet Division and the Hampton Tableland of the Eucla Division of Western Australia. The thylacine remains from the Kimberly Region are the first reported from northern Australia. Bone from the same deposit in the Kimberley Region is radiocarbon dated at O ± 80 years B.P. (GaK-3890). This date may not apply to the thylacine remains in the deposit. The deposit from the South West Division contains thylacine remains which have been radiocarbon dated at 3090 ± 90 years B P. ( ANU-716). These are the youngest re- ported from the South West Division. The deposit from the Hampton Tableland and other published faunas (e.g., Wakefield 1964) provide evidence for considering that of several possible causes for the thylacine’s decline, the most likely is competition with the introduced eutherian dog. Introduction Thylacinus cynocephalus has been reported from Quaternary deposits in New Guinea and all states of Australia except the Northern Ter- ritory (e.g. Van Deusen 1963, Partridge 1967, De Vis 1894, Macintosh and Mahoney 1964, Gill 1953). It has never been postively recorded liv- ing within historic time from any area except Tasmania, although there are many records of sightings of thylacine — like animals unsupported by material evidence. Thylacines may still be living in certain remote areas of Tasmania but the last living individual captured in the wild was at Mawbanna in 1930 (Marlow 1968). The most recent published dates for Austra- lian mainland occurrences of thylacines are those recorded by Lawton and Twidale (1964) as between 3 240 ± 80 and 3 881 ± 85 radiocarbon years B.P. and by Partridge (1967) as 3280 ± 90 radiocarbon years B.P. The taxonomy used in this paper is that adopted by Ride (1970). Catalogue numbers re- fer to specimens in the collections of the West- ern Australian Museum. Certain caves are listed with registration numbers given by the Western Australian Speleological Group (pers. com. Mr. P. J. Bridge). Western Australian cave names are given without the possessive “ ’s ” in accord with the policy of the Lands and Surveys De- partment of Western Australia (e.g. Murray’s Cave becomes Murray Cave). 1 Western Australian Museum, Francis Street, Perth, Western Australia 6000. Present address, Queensland Museum, Fortitude Valley, Queensland 4006. The Kimberley Division In 1970, Dr. R. E. Lemely, Mr. A. M. Douglas and the author collected bones and other remains within 11 cm of the surface of a small limestone pocket referred to by us as the “Tunnel Creek Carnivore Lair” in the wall of a collapsed doline, above Tunnel Creek in the Napier Range, of the southwestern Kimberley Division. The remains have been registered in the collections of the Western Australian Museum as 71.12.60 - 71.12.119. The fauna includes the following taxa: Dasyuridae Thylacinidae Peramelidae Phalangeridae Petauridae Macropodidae Muridae Pteropodidae Vespertilionidae Megadermatidae Metatheria Phascogale cf. P. tapoatafa Antechinus cf. A. macdonnellensis Dasyurus hallucatus Thylacinus sp. (see Ride 1964 and Archer 1972 for comments about species)* Isoodon sp.* Macrotis lagotis Trichosurus sp. or Wyulda sp.* Petropseudes dahli Peradorcas concinna Petrogale cf . P brachyotis Megaleia rufa* Eutheria Pseudomys nanus P. forresti Zyzomys argurus Conilurus cf . C. penicillatus * Mesembryomys cf. M. macrurus Rattus tunneyi Pteropus sp.* small bat (indet) Macroderma gigas * The specimens representing these taxa probably do not represent undescribed species. They were not specifically identified either because the material is post-cranial in nature (e.g., the thylacine specimen is a humerus, Fig. 1) or the described species of the taxa could not be adequately dif- erentiated using only the characters present in the fossil specimens. Non-mammal remains include two species of lizards, at least one species of snake, a catfish spine and fish vertebrate, fresh-water crabs, mussels, terrestrial snails, insects and spiders. The specimens from this deposit representing Macrotis lagotis, Petropseudes dahli, Peradorcas concinna and Conilurus penicillatus represent range extensions for these species within the Kimberley region of Western Australia (Ride 1970). Pseudomys forresti has not previously been recorded from the Kimberley region. Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. 43 CM Figure 1.— The proximal portion of a thylacine humerus (71.12.119) from the Kimberley deposit. It is compared with the proximal end of a thylacine humerus (Geological Survey of Western Australia Number F6353) from a cave on the Hampton Tableland of the Eucla Basin of Western Australia. There is also a distal portion of a thylacine humerus (perhaps the same humerus) from the Kimberley deposit. The vertebrate remains in the deposit prob- ably represent a carnivore’s accumulation. The larger species of marsupials are represented mainly by juvenile specimens. It is not un- common behaviour for some “scavenger-preda- tors” to specialize on juvenile as well as senile and sick individuals of larger prey species (Estes 1967). In addition the animals present in the deposit, such as the crabs, fresh-water mussels, arboreal Phascogale sp., and plains-dwelling Megalia rufa represent such diverse habitats that it is highly unlikely that their remains could have accumulated in one place without the assistance of some transporting agent. Because the limestone pocket is horizontal, shows no evidence of flooding, and has no entrances from above, contribution to the deposit by means of floods or animals dropping in from above seems improbable. It is most likely that some animal with catholic food habits accumulated the re- mains in one place. The presence of the mussel shells and perhaps also the crab and fish remains suggest Abori- ginal man may have been involved. He would certainly be able to procure mussels and crabs from the wet sand of Tunnel Creek. However, it is also possible that dogs or thylacines might be able to do the same thing. It may be signi- ficant that no burned bones or stone artifacts were recovered from the deposit. The “Tunnel Creek Carnivore Lair” limestone pocket is floored with irregular stones which made it very uncomfortable to sit or recline while trying to reach bone material on the surface which lay under low ledges at the margins as well as in the centre of the pocket. This suggests that Aborigines might not use the small pocket as a shelter. Although Aborigines might have tossed things into this pocket, the absence of burned bones and artifacts suggests that the animal material was gathered by a non-human carnivore. Much of the bone from this deposit is broken. The manner in which bone is broken often sug- gests which carnivore or other agency is res- ponsible (e.g. Brain 1967, Douglas, Kendrick and Merrilees 1966, Lundelius 1966). The bone from the “Tunnell Creek Carnivore Lair” is not as Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. 44 thoroughly smashed as bone from certain un- doubted archaeological deposits (e.g. Puntat- jarpa, in the Warburton Range of Western Aus- tralia, report in press). On the other hand, Douglas, Kendrick and Merrilees (1966) have demonstrated the type of bone-breakage caused by Sarcophilus harrisii in which Bettong-sized and smaller animals are broken up into small pieces in a manner similar to bone broken by humans. The bones from the “Tunnel Creek Carnivore Lair” are not as thoroughly broken as this and may represent the activities of dogs or thylacines. Bone fragments from the sample collected were submitted for radiocarbon dating. The re- sult (Gak-3890) was 0 ± 80 radiocarbon years. It should be pointed out that because of varia- tions in atmospheric C 14 concentrations, this date might represent about 200 calendar years (Radiocarbon 1966). Other uncertainties and variables affecting the relialibility of such young dates are given in Polach and Golson (1966). Because the thylacine remains consist of only a broken humerus, the thylacine bone itself was not submitted as part of the radiocarbon sample. It is possible that the age of the thylacine bone differs from the age of the bone in the sample submitted for dating. The thylacine humerus is not as fresh-looking as some of the other bone in the sample. The possible difference in age of the various bones in the deposit cannot be determined. Brandi (1972) reports Aboriginal rock-paint- ings in Arnhem Land of a striped animal which he interprets as possibly representing Thylacinus. This interpretation is convincing. Although the thylacine bone from the “Tunnel Creek Car- nivore Lair” deposit reported here may not be the same age as the other bone in the deposit, these Arnhem Land rock-paintings reported by Brandi (1972) would seem to add support to the idea that thylacines may have survived in north- ern Australia until late recent time. Wright (1968) reports rock art from the Pilbara Region of northwest Australia which he considers represents a dog-like mammal with vertical stripes and other features suggesting a thylacine. The Pilbara Region is intermediate between the Kimberley area and the South West Division in Western Australia. The South West Division Excavation by Mr. I. Murray, Miss E. A. Jef- ferys and the author in a stratified deposit in Murray Cave (Yn52), formed in what is com- monly referred to as “Coastal Limestone” (Smith 1963) about 40 km north of Perth recovered a fauna that included Thylacinus cynocephalus (to be reported fully elsewhere). Remains of thylacines (e.g. 72.1.1148) were collected from the surface of the cave floor and between the surface and 7 cm in the excavated deposit. A radiocarbon date (ANU-716) on charcoal from 1-7 cm is 3 090 ± 90 radiocarbon years B.P. This is the youngest dated occurrence of thylacines from the South West Division of Western Aus- tralia. Thylacinus cynocephalus is recorded from the Mammoth Cave fauna which may be older than 31 500 years B.P. (Merrilees 1968). This is the only other dated occurrence of thylacines from the South West Division although they are not uncommon in deposits of this area parti- cularly in the Cape Leeuwin — Cape Naturaliste region (Merrilees 1968). The Murray Cave fauna includes the following species (specimens registered as 72.1.1148-72.1. 1170) : Dasyuridae Thylacinidae Peramelidae Phalangeridae Petauridae Macropodidae Metatheria Antechinus apicalis Antechinus flavipes Dasyurus geoffroii Sarcophilus harrisii Thylacinus cynocephalus Isoodon sp.* Trichosurus vulpecula Pseudocheirus peregrinus Bettongia lesueur B. penicillata Macropus irma M. eugenii Petrogale penicillata Setonix brachyurus Vespertilionidae Megadermatidae Muridae Eutheria (indet)* Macroderma gig as Rattus fuscipe s pseudomyine (indet)* * See note at bottom of the “Tunnel Creek Carnivore Lair” list above. In addition there is at least one species of lizard, one species of frog and three species of terrestrial snails. The bone material from Murray Cave is con- sidered to represent a carnivore’s accumulation. As in the “Tunnel Creek Carnivore Lair” de- posit reported above, the larger species are mainly represented by specimens of juvenile in- dividuals. It differs from the “Tunnel Creek Carnivore Lair” deposit in that almost every bit of bone is broken into much smaller pieces, in- cluding most of the jaws. Bone destruction of this sort is typical of Sarcophilus harrisii (Douglas, Kendrick and Merrilees 1966). One of the individuals of Thylacinus cynoce- phalus from Murray Cave consist of a maxilla and dentary representing a very small pouch- young thylacine (Fig. 2). The roots of the crowns had not yet formed and the teeth are stacked in overlapping positions in the den- tary. There is little doubt that this animal was taken into Murray Cave, either by or with its mother, or by a predator. The Hampton Tableland of the Eucla Division In 1969 and 1970 Messers B. Muir, K. Aker- man, and others accompanied the author to Horseshoe Cave (N59) northeast of Madura on the Hampton Table and where three excavations were dug and a large fauna recovered (to be fully reported elsewhere). This fauna included Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. 45 Figure 2 The Murray Cave juvenile thylacine specimen (72.1.1149). The teeth probably would not have pierced the gum at this stage of dental development and the animal would have been totally dependent on its mother. Thylacinus cynocephalus . Specimens (regis- tered as 72.1.1-72.1.1000 etc.) from the same de- posit represent the following species: Metatheria Dasyuridae Sminthopsis crassicaudata Sminthopsis murina Antechinomys spencen Phascogale calura Dasycercus cristicauda Sarcophilus harrisii Dasyurus cf. D. geofjroii Thylacinidae Thylacinus cynocephalus Muridae Hominidae Canidae Vespertilionidae Eutheria Pseudomys cf . P. gouldii P. occidentalis P. desertor P. hermannsburgensis Leporillus apicalis L. conditor Notomys cervinus (and probably other species of the genus) Rattus cf R. vellosissimus Homo sapiens Canis familiaris Small bats (indet)* * See note at the bottom of the “Tunnel Creek Lair” list above. Peramelidae Phalangeridae Burramyidae Vombatidae Macropodidae Perameles cf . P. bougainville Chaeropus ecaudatus Isoodon obesulus Macrotis lagotis Trichosurus vulpecula Cercartetus cf. C. concinnus Lasiorhinus latifrons Caloprymnus campestris Bettongia lesueur B. penicillata Onychogalea lunata Lagorchestes cf. L. hirsutus Laoostrophus fasciatus Petrogale sp. Mcgaleia rufa Sminthopsis murina , Pseudomys cf. P. gouldii , P. desertor, and Rattus cf. R. villosissimus have not previously been recorded from the Hampton Tableland. Thylacine remains have been reported from the Hampton Tableland (e.g. Partridge 1967, Lowry and Merrilees 1969). A date of 3 280 ± 80 radiocarbon years reported by Partridge (1967), was based on dried flesh from a thylacine found on the surface of Murra-el-elevyn cave. Dates of 4 650 ± 104, 4 550 ± 112 and 4 650 ± 153 ra- diocarbon years B.P. are reported by Lowry and Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. 46 Table 1 Thylacine and dog remains from trench excavations in Horseshoe Cave (N 59). T = thylacine, D=dog. There are also dog remains on the surface of the cave floor. The synopsis column gives a summary of the dog , thylacine and relevant dates from particular trench excavations (no. in parenthesis). Depth (cm) Trench 1 Depth (cm) Trench 2 Depth (cm) Trench 3 Synopsis 90-100 110-120 200-220 250-260 T (e.g. 72.1.155) 5630 ±120 (GaK-3814, bone) T (e.g. 72.1.307) 15800 ±1800 (GaK-3815, bone) 40-50 50-53 T (e.g. 72.1.482) T (e.g. 72.1.512) (date from 50-53 level: 7030 ±130, GaK-3888, bone) 30-40 50-60 D (e.g. 72.1.691) 3570 ±100 (GaK-3570, bone) and 4500 ±330 (GaK-3476, charcoal) D on surface D 30-40 (3) 4500 ±330 (3) T 90-100 (1) 5630 ±120 (1) T 200-220 (1) 15800 ±1800 (1) Merrilees (1969) for a thylacine mummy from the surface of Thylacine Hole cave. Thylacine remains from Horseshoe Cave (N59) are sum- marized in Table 1 with radiocarbon dates and the stratigraphic levels containing dog remains. The specimen (72.1.512) from the Trench 2 level dated at 7 030 ± 130 radiocarbon years B.P. is the oldest dated thylacine from the Hampton Tableland. The specimen (72.1.307) from the Trench 1 200-220 cm level may represent an even older thylacine. Discussion It is now clear that during Quarternary time, thylacines existed throughout the Australian continent, as well as in New Guinea and Tas- mania. However, the reasons for the decline of the thylacine are not clear. Fleay (1946) be- lieves that snares and baits laid by European hunters were the cause for the decline of Tas- manian thylacines. Merrilees (1968) and Jones (1968) suggests that Aborigines may have played a major role in the extinction of some of Australia’s Quaternary marsupials. Troughton (1967) and Jones (1970) suggest that competi- tion with dogs may have been a major reason for the decline of the thylacine. Although thylacine remains have occasionally been recovered from excavations that contain Aboriginal remains or artifacts (e.g. Fromm’s Landing, Macintosh and Mahony 1964, Horse- shoe Cave reported in this paper) there is no evidence (such as burned thylacine bones in midden deposits) to suggest Aborigines actively hunted thylacines. Aboriginal man might have been responsible for destroying such vital aspect of the thylacine’s habitat such as the vegetation or food supply. But the fact that thylacines and Aborigines persisted together into historic times in Tasmania suggests co-existance could have also taken place on the mainland. Wright (1971) presented evidence for Aboriginal activity in Koonalda Cave on the Hampton Tableland of the Nullarbor. He states (p. 28) that . .traces of human activity are present from roughly 22 000-15 000 years ago”. Considering the date of 3 280 ± 80 radiocarbon years B.P. recorded by Partridge (1967) for thylacine remains in Mur- ra-el-elevyn Cave on the Hampton Tableland, thylacines and Aborigines probably co-existed on the Hampton Tableland for at least 18 000 radiocarbon years. Therefore Aborigines were probably not responsible for the decline of the thylacine on the Australian mainland or Tas- mania. Dogs and thylacines Tindale (1959) suggests that dogs were ori- ginally brought into Australia by Aborigines The oldest known dog remains in Australia are from a cave at Mt. Burr in South Australia, and are dated at between 7 450 ± 270 and 8 600 ± 300 radiocarbon years B.P. (Mulvaney 1969). Dogs occur in New Guinea (Troughton 1971), formerly occurred in New Zealand (Alio 1971) but are not known to have occurred in Tasmania prior to the arrival of Europeans around 1798 (Jones 1970). Jones (1968) suggests that Tasmania has been separated from the Aus- tralian mainland for about 12 000 years and that therefore (p. 258) it “...seems highly likely that the dingo first appeared in Australia after about 12 000 years ago, and in view of the Mt. Burr evidence some time before 7 000 years ago”. Merrilees (1968) has suggested that a dog tooth in the Western Australian Museum collections from the Mammoth Cave deposit probably did not come from the Pleistocene deposit in that cave (which is dated at greater than 31 000 and 37 000 radiocarbon years B.P.) and Jones (1970) states (p. 258) that “Claims for dog teeth in Pleistocene (Australian) deposits have not been confirmed”. Thylacine remains are known from Austra- lian, Tasmanian and New Guinean late Quater- nary deposits (Merrilees 1968, Van Deusen 1963, Gill 1963, Gill and Banks 1956) and may have been sympatric with dogs that occurred in New Guinea and Australia. Actual instances of the sympatric occurrence of the two species inter- pretable from fossil deposits in Australia have not been clearly demonstrated. At present the only instance known to me is that recorded by Wakefield (1967). He reports dog and thylacine remains from the same “Recent” layer in the McEachern’s Cave in southwest Victoria. This layer is 14 inches thick. Wakefield (1967) also reports dog and thylacine remains from a Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. 47 “mixed” sand at depth of 14 to 30 inches. It is possible that this report indicates the two species were sympatric. On the other hand there are several deposits in which large excavations have revealed super- position of dog remains over those of thylacines (e.g. Wakefield in Mulvaney, Lawton and Twidale 1964). In Horseshoe Cave (reported in this paper and by Archer 1972) on the Hamp- ton Tableland dog remains occur in two of the three excavated trenches but only at levels above levels containing thylacine remains (see above). Lowry and Merrilees (1969) report a radiocarbon date of 220 ± 96 years B.P. for a desiccated carcase of a dog from another cave on the Hampton Tableland. This date is younger than any that applies to thylacine remains in Australia. In fact I am unaware of any radio- carbon dated sequences in Australia which de- monstrate an example of superpositional rela- tionship of thylacine over dog remains. It is possible that the thylacine remains reported in this paper from the “Tunnel Creek Carnivore Lair” Kimberley deposit are contemporaneous with the fauna which has been radiocarbon dated at 0 ± 80 years B.P. (see above). If this is the case it is most likely that dogs and thy- lacines have been sympatric in northern Aus- tralia for some unknown time. But considering that the thylacine remains in this Kimberley deposit are (as noted above) somewhat diffe- rently preserved from the other bones in the de- posit, it cannot be argued on the evidence of this deposit alone that thylacines were sym- patric with dogs in northern Australia. What little is known of New Guinean thy- lacine and dog history suggests a similar situa- tion. White (1971) suggests (p. 190) that dog remains are found in New Guinea shelter sites “. . . only in very recent times . . .”. On the other hand thylacine remains have been recorded by Van Deusen (1963) from the Kiowa Rock Shel- ter in New Guinea above a level dated at 9 920 ± 200 years B.P. (Bulmer 1964 and Van Deusen pers. com.). Tasmanian thylacines are reputed (e.g. Troughton 1967) to have eaten wallabies, smaller marsupials, rats, birds, probably echidnas and possibly lizards as natural foods. Rolls (1969) records (p. 361) that following an active and successful campaign to reduce the numbers of wild dogs (dingo) in southeast Australia in 1863 “. . . there began a startling build-up of wildlife. Kangaroos, wallabies, pademelons, rat- kangaroos, bilbies, and bandicoots which had all been present in insignificant numbers sprang up like grass”. Although it is possible that factors other than the reduction of dog numbers might have contributed to this resurgence of wildlife, it would seem probable that dogs could usurp the thylacine’s natural foods in areas where the two carnivores were sympatric. Calaby (1971) says (p. 90) of the introduced dog that it “...is an opportunist predator with a catholic diet includ- ing virtually anything it can catch, vertebrate or invertebrate, together with carrion and even some vegetable material”. I have recently ex- amined stomach contents of dogs trapped in northwestern Western Australia (by courtesy of Mr. Simon Whitehouse) and identified specimens of feral cats ( Felis domesticus ) in two stomachs. It is therefore not unlikely that dogs might even have preyed directly on thylacines if given the opportunity. The dog and thylacine are of comparable size and partitioning of the habitat without a long historical basis, probably would not have occurred. Further evidence for competition between dogs and thylacines comes from a consideration of the Tasmanian situation. Dogs are not known to have been present in Tasmania prior to his- toric times (Jones 1968). Thylacines on the other hand persisted in Tasmania despite the presence of Aborigines (see above). It could be argued that the introduction of dogs into Tasmania in historic times was a significant reason for the eventual decline of the thylacine in Tasmania. Guiler (1961) examined records of thylacine kills by European hunters in Tas- mania. He noted that the number of thylacines killed remained relatively constant in two areas where killing for bounties had begun in 1888, until the sudden general decline throughout Tasmania around 1909, rather than gradually declining in those areas. He concluded that for this reason hunting by Europeans was not the sole cause of the thylacine’s decline. He sug- gested disease might have been a factor in caus- ing the decline. The same argument could be used to suggest that no single gradual pressure, including pressures caused by the introduced dog, brought about the Tasmanian thylacine’s decline. Guiler (1961) did not however con- sider that predation by Europeans (and or competition with the dog during historic times) may have been responsible for lowering the thylacine populations in Tasmania to a cri- tical level at which disease or some other factor could have dealt a crushing blow. It would seem too much of a coincidence that the Tas- manian thylacine population declined so dras- tically for unknown reasons during the only 11 year period of Quaternary time when it was also actively and methodically predated upon by human hunters. Whatever the cause for Tas- manian thylacine decline, the fact that thyla- cines existed in Tasmania into historic times and evidentely did not on the Australian main- land, suggests that dogs may have been the significant factor which brought about the de- cline of the thylacine on the Australian main- land. There would appear to be no comparable sit- uation within recent time involving introduced dogs and indigenous dog-sized, non-canid car- nivores on any other continent. Dog-sized ca- nids are indigenous to Africa, e.g. jackals, Canis adustus, bush dogs, Lycyon pictus) , South Ame- rica (bush dogs, Speleothos spp., maned wolves, Chrysocyon sp. etc.), and North America, Europe and Asia (grey wolves, Canis lupus). On the other hand a similar situation occurred in South America during late Pliocene (Montehermosan) time. In South America the only large carni- vores were the marsupial borhyaenids which in- cluded thylacine-like and thylacine-sized ani- mals. When faunal interchange became possible Journal of the Royal Society of Western Australia, Vol. 57, Part 2. August, 1974. 48 across Panama, canids, procyonids, ursids, mus- telids and felids entered South America and successfully colonized the continent, whereas the borhyaenids became extinct (Patterson and Pas- cual 1968). It is probably not possible to know what particular role the canids had in displac- ing the borhyaenids. Colbert (1955) suggests that marsupials are “second-class mammals” as compared with eutherians. This impression arises as a result of the apparent competitive inferiority exhibited by many marsupial groups. Storr (1958) has argued that this is because most marsupials evolved in isolation on southern continents. They were not subjected to the constant testing that Holarctic eutherians underwent. Thylacines were in this sense at a disavantage in a confrontation with the dog. The modern dog is a representative of a stock of eutherians that had already successfully confronted the South American marsupial borhyaenids. The Australian thylacine probably had even less of a chance for survival against such a seasoned competitor than the South American Borhy- aenids. Acknowledgements. — Collections upon which this paper was based were made while the author was carrying out research under a Fulbright Scholarship from the Australian American Educational Founda- tion, a grant in aid from the American Explorers Club, and as a research assistant to Dr. W. D. L. Ride who was in receipt of a Research Grant from the Austra- lian Research Grants Committee. Dr. Ride is the Director of the Western Australian Museum. Dr. R. E. Lemley of Rapid City, South Dakota, provided the finances to support Mr. A. M. Douglas, the author and himself on an expedition to northern Australia. Sir Thomas Wardle helped finance certain aspects of the work on the Hampton Tableland in 1970. Mr. I. Murray was instrumental in discovering and with Miss E. A. Jefferys and the author in helping to recover the fauna from Murray Cave. Messrs. B. Muir, K. Akerman, D. Brindle, J. Leahey, and C. Chubbs helped make collections from Horseshoe Cave and other caves on the Hampton Tableland. Many other people and in particular Master M. Hearne and Miss E. A. Jefferys (now Mrs. M. Archer) provided invalu- able assistance in sorting the material obtained from Horseshoe Cave. Drs. D Merrilees, D. Kitchener and Mr. G. W. Kendrick of the Western Australian Museum kindly read and criticised the manuscript. I am grateful to Dr. D. Merrilees for providing many of the macropodid identifications, to Messrs. J. Mahoney and A. Baynes for providing the murid identifications, and to Mr. G. W. Kendrick for providing comments about the invertebrate remains. Dr. Hobart Van Deusen provided comments and information about the age of the Kiowa New Guinea thylacine specimen. References Alio, J. (1971).— The dentition of the Maori Dog of New Zealand. Records of the Auckland Institute and Museum 8: 29-45. Archer, M. (1971). — A re-evaluation of the Fromm’s Landing thylacine tooth. Proceedings of the Royal Society of Victoria 84: 229-234. Archer, M. (1972).— Nullarbor 1969. Western Caver 12: 17-24. Brain, C. K. (1967). — Bone weathering and the problem of bone pseudo-tools. South African Jour- nal of Science 63: 97-99. Brandi, E. J. (1972). — Thylacine designs in Arnhem Land rock paintings. Archaeology and Physi- cal Anthropology in Oceania 7 : 24-30. Bulmer, S. (1964). — Radiocarbon dates from New Guinea. Journal of the Polynesian Society 73: 327- 328. Calaby, J. H. (1971).— Man, fauna and climate in Aboriginal Australia, pp. 80-93 in “ Aboriginal man and environment in Australia” , (D. J. Mulvaney and J. Golson, ed.) 1-389. Aus- tralian National University Press, Canberra. Colbert, E. H. (1955 ) .—“Evolution of the vertebrates”, 1-479. John Wiley and Sons Inc, New York. De Vis, C. W. (1894).— A thylacine of the earlier n D F E • ( F E j 32 Day t ( J F E 36 Day I i C r 40 Day F E < r F E I i 44 Day > 51 Day 59 Day F E? 60 + Day E 3 E 3 83 Day E 3 E 3 E D E 3 E D E D E D 105 Day C i E D * The stages are approximations based on time from date of birth (with a probable error of less than two days). The animal representing the 22 Day Stage was found to be slightly younger than the 21 Day Stage animal on the basis of relative develop- ment. The 60+ Day Stage animal is of unknown age but on the basis of structural development represents a stage between the 59 Day Stage and the 83 Day Stage. The 83 and 105 Day Stage animals were not sectioned. S = initiation; FE — free edge of the dental lamina on the lingual side of the tooth; C = calcification; EG = erupting, i.e. just having pierced the oral epithelium; ED — erupted. Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. 55 Table 2 The development of the cheek-tooth families in the upper jaw.* Stages C 1 (1C 1 P 1 P 8 P 4 dP 4 M 1 M 8 M 8 M 4 4 Day S 10 Day F 12 Day 15 Day 5 “ 22 ” Day F E 21 Day 5 ’ F E F 28 Day C E 5 32 Day F E "'i 36 Day F E 3 40 Day 44 Day e: \'D F E 51 Day 59 Day F E 60+ Day E G y | 83 Day E D E D E D E D E D E D E 5? 105 Day ( E D < * The stages and abbreviations are the same as those given in Table 1. identified with its homologue in each consecu- tive developmental stage either to the stage of eruption, calcification, or in the case of the deciduous canine, resorbtion. The sectioned stages did not provide the later developmental stages of the P4 or M 4 because of the late initiation of these teeth. Generally less than five developmental stages after the initiation of a tooth was observed, the terminal free edge of the dental lamina was again visible at the tooth position, lingual to the tooth bud. The tooth bud thus apeared to have risen up the buccal side of the dental lamina relative to the free edge. With the exception of the deciduous canine, no second generation swellings were observed at established family positions. Therefore, each post-canine tooth family consisted of only one generation of teeth. Although incisor develop- ment is not considered here, it should be pointed out that deciduous incisors were present in association with each incisor position. These generally reached the stage of calcification and were then resorbed. The identification of cheek-tooth families The identification of each tooth family that appeared in sequence along the entire lamina from anterior to posterior end was based on the terminology of Thomas (1887). They are in anterior to posterior order: C; PI; P3; P4; dP4; Ml; M2; M3; M4. Application of these terms to the teeth in question does not mean that I imply any successional relationship be- tween any of the teeth or believe a P2 family is lost in the dentition of Antechinus flavipes . Thomas’s (1887) nomenclature is used simply because it is familiar and widely accepted in connection with Australian metatherians. The sequence of the establishment of cheek-tooth families In the earliest stage (4 days post-birth) two family positions are established: the C and the dP4. In the upper jaw (the anterior end of the dental lamina of the lower jaw was damaged in the youngest stages) the C was a discrete terminal swelling. The first tooth bud is in- terpreted to represent a milk canine (dC) as, in the next developmental stage, there is a swollen free edge lingual to the dC which is the homologue of the erupting C. DP4 was also a terminal swelling. Between these two family positions three ad- ditional positions were seen to be established in later stages in sequential order as summarized in Tables 1-2: the PI, P3, and P4 positions. More or less synchronously with the appearance of the ante-molar positions, four tooth family positions were seen to be established posterior to the dP4; Ml, M2, M3 and M4. 3b 3a Figure 2. — A schematic portrayal of the region of the lower dental lamina at the 40 Day Stage. The P 4 occurs as a swelling of the free end of the dental lamina which is by this stage suspended between the lingual walls of the P 3 and dP 4 . The mass of the swelling is actually slightly closer to the dP 4 than it is to the P.$. The lingual side of the oral epithelium has been slightly displaced dorsally. Note that the dental lamina has lost contact with the oral epithelium in this relatively late stage (compare with Fig. 6 showing the condition in the 28 Day Stage). Dotted lines 3a and 3b indicate approximate positions of transverse section photographs shown in Fig. 3. Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. 56 Figure 3. — Photographs of transverse sections of the 40 Day Stage dental lamina, (a) Section through the P 3 near its posterior end where the dental lamina representing the anterior end of the P 4 adheres to the lingual side, (b) Section through the dP 4 where the P 4 swelling is connected to the dP 4 by a thin band of dental lamina. Abbreviations: oe, oral epithelium; oc, oral cavity; dp, dental papilla; sr, stellate reticulum; dc, cartilage of the dentary bone; fe, free edge of the dental lamina. The form of the dental lamina and its free edge with particular reference to the establishment of cheek-tooth positions In the earliest stage examined the dental lamina posterior to the C position was of uniform depth, continuous, and in contact with the oral epithelium. In subsequent stages, when tooth initiation had occurred, the dental lamina and/or its free edge associated with the de- veloped tooth appeared to be more distant from the oral epithelium than was the free edge of the inter-tooth dental lamina. In addition, as tooth development occurred there was a ten- dency to lose contact with the oral epithelium in the region of tooth development. This pro- duced the appearance, particularly in the lower jaw in later stages, of the dental lamina existing as a ribbon stretched or hung between the lingual walls of the developed tooth buds. When the P4 had initiated, it was flanked anteriorly and posteriorly by two well-developed tooth buds: the dP4 and the P3. In the last sectioned developmental stage (60+ Day Stage), the dental lamina connecting P4 to the surrounding teeth had degenerated, leaving the P4, which was in that stage still an enlarged swelling of laminar tissue, isolated in the matrix of the lower jaw lingual and anterior to the well- developed dP4. In the earliest stages examined, the dental lamina did not extend posterior to the dP4 position. The dP4 itself appears to represent a terminal swelling at the posterior end of the dental lamina’s free edge. One stage later the dental lamina is seen to extend posterior to the dP4 and to be swollen slightly to form the Ml. Figure 4. — Schematic portrayal of the dental lamina and associated structures at the 44 Day Stage to demonstrate the upper molar relationships. The lingual side of the oral epithelium has been slightly depressed ventrally to expose the dental lamina. The dental lamina is in contact with the oral epithelium at its posterior edge. The terminal swelling which represents M 3 is continuous with the free edge of the dental lamina lingual to the M 2 . In a later stage (the 59 Day Stage) the free edge of the dental lamina is again visible at the M 3 position lingual to the M 3 . Dotted lines 5a, 5b, and 5c indicate approximate positions of transverse section photographs shown in Fig. 5. Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. 57 Figure 5. — Photographs of transverse sections of the 44 Day Stage dental lamina, (a) Section through the M- showing its connection with the dental lamina which is in turn connected to the oral epithelium. The free edge of the dental lamina, extending dorsally from the dental lamina on the lingual side of the M 3 , is sectioned anterior to the point at which it swells to form the M 3 swelling, (b) Section through the posterior end of the M 2 and the free edge of the dental lamina which is attached to the oral epithelium. The free edge is thicker than in (a) above but is still anterior to the position of the M 3 swelling, (c) Section through the posterior end of the M 3 swelling. Abbreviations as in Fig. 3. Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. 58 7d 7c 7b 7a Figure 6. — Schematic portrayal of the lower dental lamina and associated structures at the 28 Day Stage. The lingual side of the oral epithelium has been lifted dorsally to expose the dental lamina. The free edge of the dental lamina has been re-established lingual to the Mi and in the anterior region of the M 2 . It has not yet been re-established along the posterior edge of M 2 . The posterior region of the dental lamina is not in” vertical contact with the oral epithelium and is poorly differentiated from the posterior region of the M 2 . The part that is differentiated lingually and extended terminally represents the initiating M :i . Dotted lines 7a, 7b, 7c, and 7d indicate approximate positions of transverse section photographs shown in Fig. 7. At that stage the free edge of the dental lamina is not visible on the lingual side of the dP4. It was, however, visible just prior to the estab- lishment of M2. In the upper jaw the free terminal edge of the dental lamina appears to be more or less vertical in each stage as it is seen to extend farther posteriorly. It maintains continuity with the oral epithelium in the region of posterior growth but further anteriorly, in regions where teeth have been established several stages earlier, the connection with the oral epithelium breaks down. In the upper jaw posterior to the M 1 and in stages after the M 1 is established, the free edge of the dental lamina is seen to be established on the ligual side of the tooth bud prior to, or simultaneously with, the establish- ment of the next molar position. In the lower jaw the dental lamina is seen to extend posteriorly, in stages post-dating the establish- ment of the Mi, such that the ventral free end extends farther posteriorly than the dorsal fixed end in contact with the oral epithelium. This difference in position appears to suggest that the ventral free end extends posteriorly at a more rapid rate than does the dorsal fixed end. As in the case of the upper lamina, in later stages in regions where teeth have been estab- lished in several previous stages, the dental lamina is seen to have lost its contact with the oral epithelium. In contrast, however, with the upper lamina, the free edge of the dental lamina of the lower jaw lingual to each tooth is not visible prior to, or simultaneously with, the establishment of the next molar. It does not normally (except in the case of M 3 ) appear until at least one stage after the initiation of the next posterior molar position. Discussion Tooth replacement and the Zahnreihe theory It is clear that in Antechinus flavipes teeth established posterior to the C position are separate tooth families and each has only one generation. There are therefore no true succes- sional post-canine teeth in the sense of milk and permanent teeth of succeeding generations such as are believed to occur in most eutherians. This supports the observations of Woodward (1893), Engelhardt (1933), Dressel (1931), Lit- tich (1933), and Berkovitz (1966 and 1967) that the P4 develops from the dental lamina between the P3 and dP4 positions and is not a successor to either. It does not support the contention of Kirkpatrick (1969) that P4 is a successional tooth in the same family as the P3, nor does it support the contention of Wilson & Hill (1897) and other earlier workers that P4 was the successor to dP4. Berkovitz (1972) describes tooth replacement in the Guinea Pig (Cavia cobya) . He notes that in the upper dentition the so-called replace- ment premolar develops from dental lamina lying anterior to the so-called deciduous pre- molar. This is a situation comparable with the condition in Antechinus in that the replacing tooth does not develop in a position clearly lingual to the deciduous tooth. However, in the lower dentition of the Guinea Pig, the replace- ment tooth develops as a lingual downgrowth of dental lamina associated with the posterior half of the enamel organ of the deciduous tooth. Unfortunately, the actual homology of these teeth in Guinea Pigs is uncertain (Berkovitz 1972), and all of the teeth, including the molars may belong to one generation. Ziegler (1972) describes tooth replacement in the eutherian Mole Scapanus latimanus and demonstrates that replacement teeth develop from the free edge of the dental lamina that develops from the lingual edge of the deciduous tooth germ. This manner of tooth replacement is unlike that seen in the Antechinus dP4 and P4 in the present study, but is comparable with the situation observed here for the deciduous and replacement incisors and canines. On the basis of the evidence presented in this study, the manner in which the upper M2-4 develop in A. flavipes is not identical with the manner in which the lower M2-4 develop; this difference in ontogenetic behaviour of upper and lower dentitions parallels the condition observed in Cavia by Berkovitz (1972, see above). In the posterior region of the upper jaw the dental lamina is firmly in contact with the oral epithelium as new molar positions are initiated. By the time these initiations occur, posterior to M 1 , the free edge of the dental lamina is established on the lingual side of the previous molar. Consequently it is clear that each molar position is a separate family position which subsequently supports or lies buccal to a lingual free edge of dental lamina. This re-establish- ment of the free edge of the dental lamina appears to be either a passive phenomenon relating to the movement orally of the estab- lished tooth germ or else a positive development of the dental lamina as a means of maintaining the developmental potential continuously along the dental lamina. It does not appear to be a necessary prerequisite to the establishment of posterior tooth positions (as Kirkpatrick, 1969, Journal of the Royal Society of We tern Australia, Vol ol, Part 2, August, 1S74. 59 Figure 7. — Photographs of transverse sections of the 28 Day Stage dental lamina, (a) Section through the Mi showing a small remnant of dental lamina adhering to the lingual side of Mi and not connected to the oral epithelium at this point, (b) Section through Mi and M 2 showing contact between M 2 and the dental lamina and the latter with the oral epithelium, (c) Section through M 2 showing the contact between it, the dental lamina and the latter with the oral epithelium. Note the thickened epithelial cells (which is proliferative dental lamina) on the lingual side of the M 2 which marks the anterior extremity of the M 3 swelling, (d) Section through the M 2 and M 3 in the region where they are side by side. The M 3 is developed as an overgrowth of dental lamina passing postero-lingual to the M 2 . At this point neither the M 2 nor the M 3 are in contact with the oral epithelium. Abbreviations as in Fig. 3. Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. 60 has implied by suggesting that dP4 and Ml-4 are successional replacement teeth of one family) as is clearly demonstrated by its re- tarded re-establishment of the lower dental lamina. This delayed re-establishment of the free edge, as well as the apparent lag in the posterior extension of the junction between the fixed end of the dental lamina and the oral epithelium, constitute the major differences be- tween the development of the lower and upper dental lamina in the molar region. The differ- ence may be caused by the presence of an ascending ramus in the lower jaw. In all later stages, the posterior part of the dental lamina appears to be not only crowded but almost looped. The molars, as they develop, tend to overlap. This could account for the delay in the establishment of a vertical relationship be- tween a lower molar and the oral epithelium which would not or could not occur until such time as the mandible has sufficiently lengthened to permit or facilitate a vertical connection. In the upper jaw, it is clear than bony processes provide no obstacles to development. The de- veloping molars are seen to extend out beneath the orbit where they may be, so to speak, waiting for the maxilla to catch up and provide bony crypts. Churchill (1935) describes molar formation and its relationship to the dental lamina and oral epithelium in Homo sapiens. Except for a slightly more advanced rate of re-establishment of the free edge of the dental lamina, it is a situation remarkably similar to that visualized in this study for the posterior region of the lower dental lamina in Antechinus fiavipes. He even demonstrates the same difference in the rate of posterior development between the pro- liferative terminal free edge and the fixed edge of the dental lamina. The order of initiation of the cheek-teeth in A. fiavipes clearly suggests that there are in fact two distinct series of temporarily related teeth which are also, in many dasyurids (e.g. species of Sminthopsis) , distinct morphological series. The molariform series is the dP4-M4. The two series resemble the Zahnreihen postu- lated by Woerdman (1921). Edmond (1960) de- veloped the theory of Zahnreihen and suggested that some form of pulse passed along the free edge of the dental lamina initiating, as it passes, tooth buds at predetermined tooth family posi- tions. To explain the unique partial-replace- ment condition seen in so many eutherians, he visualized first one pulse passing continuously along the free edge, producing all of the milk teeth and permanent molars. Then a second pulse would sweep along the free edge but only for about half the length of the lamina’s free edge. The pulse would then stop. This would result in the initiation of a series of second generation teeth developing beneath the anterior teeth, and these would of course represent the permanent replacement teeth. In similar terms, the situation in Antechinus fiavipes may be that two cheek-tooth centres for the initiation of Zahnreihen exist (i.e. C and dP4) and that the two rows of teeth initiated by passing waves I 32 DAY STAGE d P4 y c free edge of dental lamina 4 DAY STAGE WfT & '¥fr m4 m3 m2 ml dp 4 P 4 p 3 pi c 60+ DAY STAGE Figure 8. — A schematic portrayal of Zahnreihen as evidenced in the developing lower cheek tooth row of Antechinus fiavipes. Two separate centres for wave initiation are postulated, one at the C family position and one at the dP 4 family position. No second genera- tion waves pass the post-canine positions so that no true replacement teeth are initiated in already existing post-canine families. The possibility that a third Zahn- reihe exists which involves the incisors and the dCl is discussed in the text. Degenerative changes in the dental lamina are not illustrated in this figure nor is the true nature of the posteriorly extending terminal free edge of the dental lamina (see Fig. 6). do not overlap because the waves do not overlap at any one family position. The teeth of ad- jacent families may come to overlap physically after initiation in such a manner that the tooth of one family comes to overlie the tooth of another family giving the appearance of milk- tooth and permanent successor, but this would and does occur only after the sheet of dental lamina has degenerated. In my opinion this is the relationship between the marsupial P4 and dP4. Incisor development in Antechinus fiavipes has not been discussed in the present paper because of difficulties encountered in establishing homo- logues from one specimen to another due to damage in the incisor areas in sectioning which concealed positional relationships, and the added difficulty of the similar morphology of adjacent teeth. It was clear, however, from the material that all incisor families had two generations. The deciduous teeth generally calcified before they were resorbed. It was not clear how these incisor generations are related to the two postu- lated cheek-tooth Zahnreihen because it has not been possible to determine the relative times of development and calcification of the incisors as compared with the cheek-teeth. It is there- fore possible that there are more than two Zahnreihen present in the toothrow of An- techinus fiavipes. The deciduous incisors and deciduous canine may represent one Zahnreihe, the permanent incisors, canine and premolars the second Zahnreihe, and the dP4-M4 the third Zahnreihe. Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. 61 Fosse and Risnes (1972a and b) demonstrate that in the peramelids Isoodon obesulus, I. macrourus, and Perameles gunnii, the I 5 is ontogenetically less developed in the specimens studied than I 4 or C 1 . This suggests the pos- sibility that a separate Zahnreihe exists for the incisors, the posterior member of which is the posterior incisor. Zeigler (1971) has considered the phenomenon of tooth replacement and Zahnreihen in mam- mals. He also concludes that there are two pulses responsible for the production of teeth in mammals but visualizes a greater area of overlap of the two waves than there is evidence for in the present study. Recently the concept of Zahnreihen as visual- ized by Edmund (1960) has been criticised. Osborn (1970, 1971 and 1972) and DeMar (1971, 1973) have presented alternative hypotheses to explain tooth eruption sequences. Osborn (1970) reviews tooth development in some eutherians and concludes that it does not support Ed- mund’s (1960) contention of two overlapping Zahnreihen in mammals. Instead he visualizes between three and six replacement waves. Dif- ficulties which arise from Osborn’s interpreta- tion include the need to allocate homologous teeth (e.g. the permanent canines) of different genera to different Zahnreihen. Osborn believes this may not be a problem providing the Zahn- reihen are not visualized as immutable. Osborn (1971) presents evidence for believing that, in Lacerta, Zahnreihen as visualized by Edmund (1966) are not involved in tooth production. DeMar (1971, 1973), working from data includ- ing Osborn’s (1970) summary, proposes new geometric ways of interpreting tooth eruption. Neither Osborn nor DeMar, however, make reference to marsupial tooth eruption sequences. This is unfortunate, as it is perhaps only in the marsupials that mammalian tooth eruption sequences of the sort proposed by Edmund appear to take place. The data presented in the present study indicate that, contrary to Os- born’s (1970) opinion, the teeth do in fact de- velop and erupt in sequence from anterior to posterior along the dental lamina in the order which is required by passing waves of the Zahnreihe theory. Osborn (1972) has since developed his earlier (1971) idea and suggests that tooth initiation may be a function of released inhibition. Al- though the observations reported here are des- cribed as Zahnreihen, they could equally well be interpreted as the result of released inhibi- tion. More work with other polyprotodont mar- supials will be required before the process of tooth initiation in marsupials is understood. In particular, close attention should be given to longitudinal growth of the dental lamina be- tween established tooth positions. The possibility must remain that tooth erup- tion sequences in eutherians and marsupials are fundamentally different because there would seem to be no diphyodonty in the cheek-teeth of marsupials and the two Zahnreihen have no overlap. It has long been held that the two groups have markedly different patterns of diphyodonty (in that only the last premolar was replaced) as well as differences in cheek- tooth numbers. Premolar number and terminology Although it does seem probably that eutherians and metatherians originally had the same num- ber of premolar positions, there does not seem to be enough information available to determine which premolar position has been suppressed in the metatherians. Archer (in preparation) demonstrates that supernumary premolars are known to occur in dasyurids anterior to the PI, between the PI and P3, and even posterior to the P4. Accordingly, arguments for particular premolar family loss based solely on teratology are unsound. The concept of missing P2 Ride (1964) and Mahoney & Ride (1974) have summarized some aspects of premolar homology in marsupials. The premolar terminology used by Ride is that used by Thomas (1887). Thomas argues that the occasional appearance of an extra premolar between PI and P3 is an atavistic reappearance of a suppressed tooth homologous with the eutherian P2. Accordingly he proposes that the normal marsupial premolars be called PI, P3, and P4. However, no one has demon- strated clear embryological evidence for a sup- pressed tooth family in the P2 position. Wood- ward (1896, p. 184) claims to have found it as “. . . an enormous gap between pml and pm2 both above and below, this is bridged over by dental lamina, which shows a slight indication of being swollen, this probably represents Thomas’ missing pm2, the adult premolars being the 1st, 3rd and 4th.” He notes this in An- techinus and a similar development (p. 286) in Dasyurus. There was however no evidence for a P2 family position in the Antechinus flavipes examined in the present study. The concept of a missing PI Ziegler (1971) has recently re-examined con- cepts of premolar loss in mammals and has concluded that all the marsupials and most eutherians have lost a premolar from the an- terior end of the pre molar row. In this he has reached the same conclusion as Owen (1840-45), but I know of no embryological evidence for this loss in marsupials. Ziegler has pointed out that in mammals premolar loss generally pro- ceeds by loss first of the replacement tooth followed by loss of the deciduous tooth in a tooth family. This loss proceeds from anterior to posterior along the premolar row. I can find no palaeontological or ontological evidence for these trends in marsupials. The concept that the premolariform series is complete but that P4 may occasionally be lost The evidence that is available suggests that C-P4 is a complete unit representing a Zahn- reihe but that in dasyurids there is a general trend towards reduction of the posterior pre- molar family, with loss of this tooth in several Journal of the Royal Society of Weitern Australia, Vol. 57, Part 2, August, 1974. 62 dasyurid genera ( Dasyurus, Dasycercus, Dasy- uroides, some Antechinus, some Planigale, and some Myoictis). This study shows that there are no true pre- molars in Antechinus flavipes if premolars are defined in the terms of Owen (1840-45) as those post-canine teeth which possess milk predeces- sors. This definition is the basis of the modern dental terminology applied to the cheek-teeth of mammals. However, it is clear that there are two different morphological kinds of cheek- teeth in Antechinus flavipes and that the teeth of each of the two kinds (i.e. premolariform and molariform) have a unity in origin best expressed by the concept of Zahnreihen. If studies of other metatherians demonstrate a similar situation, it may be desirable to redefine the kinds of cheek-teeth and to develop a nomenclature that reflects the Zahnreihen. Acknowledgements . — I am most grateful to Dr W. D. L. Ride, Director of the Western Australian Museum, for his helpful criticism and advice. Mr G. Burns of the University of Western Australia helped extensively with the sectioning of the material studied. Dr T. K. Kirk- patrick of the Queensland Department of Primary In- dustries kindly allowed me to examine his doctoral thesis on the development of kangaroo teeth. My wife Elizabeth helped extensively in raising the animals involved in this study. During the course of this research, the author held a Fulbright Scholarship, a grant in aid from the American Explorer’s Club and was a Research Assistant to Dr W. D. L. Ride who was in receipt of a Research Grant from the Australian Research Grants Committee. This work is part of a broader investigation into the phylogeny of the Dasyuridae carried out as doctoral research supervised by Dr W. D. L. Ride. References Berkovitz, B. K. B. (1966). — The homology of the pre- molar teeth in Setonix brachyurus (Macro- podidae: Marsupialia) . Archs oral Biol. 11: 1371-1384. (1667). — The dentition of a 25-day pouch- young specimen of Didelphys virginiana (Didelphidae: Marsupialia). Archs oral Biol. 12 : 1211 - 1212 . (1968). — Some stages in the early develop- ment of the post-incisor dentition of Tri- chosurus vulpecula (Phalangeroidea: Mar- supialia). J. zool. Res. 154: 403-414. (1972). — Ontogeny of tooth replacement in the Guinea Pig (Cavia cobya). Archs oral Biol. 17: 711-718. Churchill, H. R. (1935). — Meyer’s normal histology and histogenesis of the human teeth and asso- ciated parts. Lippincott Co., Philadelphia, I-VIII + 1-305. Clemens, W. A., Jr. (1966). — Fossil mammals of the type Lance Formation, Wyoming. Part II. Mar- supialia. Univ. Calif. Pubis geol. Sci. 62: 1 - 122 . Dawson, A. M. (1926). — A note on the staining of the skeleton of cleared specimens with alizarin red S. Stain Tech. 1: 123-124. DeMar, R. (1972). — Evolutionary implications of Zahn- reihen. Evolution 26: 435-450. (1973).— The functional implications of the geometrical organization of dentitions. J. Paleont. 47: 452-461. Dressel, H. (1931).— Uber die Zahnentwicklung bei Didel- phys. Gegenbaurs morph. Jb. 68: 434-456. Edmund, A. C. (1960). — Tooth replacement phenomena in the lower vertebrates. Contr. Life Sci. Div. R. Ont. Mus. 56: 1-190. Englehardt, H. (1933).— Uber die Zahnentwicklung bei Aepyprymnus rufescens. Gegenbaurs. morph. Jb. 71: 77-94. Fosse, G. (1969). — Development of the teeth in a pouch- young specimen of Antechinus stuartii and a pouch-young specimen of Sminthopsis cras- sicaudata. Dasyuridae: Marsupialia. Archs oral Biol. 14: 207-218. and Risnes, S. (1972a). — Development of the teeth in a pouch-young specimen of Isoodon cbesulus and one of Perameles gunnii (Pera- melidae: Marsupialia). Archs oral Biol. 17: 829-838. (1972b). — Development of the incisors in two pouch-young stages of Isoodon macrourus. Archs oral Biol. 17 : 83-845. Kirkpatrick, T. H. (1969).— The dentition of the mar- supial family Macrcpcdidae with particular reference to tooth development in the grey kangaroo Macropus giganteus Shaw. Unpub- lished thesis, University of Queensland: 1- 128. Littich, F. (1933). — Uber die Zahnentwicklung bei einem 6 cm langen Didelphysjungen. Gegenbaurs morph. Jb. 72: 303-308. Mahoney, J. and Ride W.D.L. — (in press). Osbcrn, J. W. (1970).— New approach to Zahnreihen. Nature, Lond. 225: 343-346. (1971). — The ontogeny of tooth succession in Lacerta vivipara Jacquin (1787). Proc. R. See., Lond. 179: 261-239. (1972). — On the biological improbability of Zahnreihen as embryolcgical units. Evolu- tion 26: 601-607. Owen, R. (1840-5). — “Odontography”. London, Hippolyte Balliere: LXXIV and 1-685. Sinclair, W. J. (1908). — Mammalia of the Santa Cruz Beds. Marsupialia. Rep. Princeton Exped. Patagonia 4: 333-460. Thomas, O. (1887). — The homologies and succession of the teeth in the Dasyuridae, with an attempt to trace the history of evolution of mam- malian teeth in general. Phil. Trans. R. See. 178: 443-462. Wilson, J. T., and Hill, J. P. (1897) .—Observations upon the development and succession of the teeth in Perameles, together with a contribution to the discussion of the homologies of the teeth in marsupial animals. Q. J. microsc. Sci. 39: 427-588. Woerdman, M. W. (1921). — Beitrage zur Entwicklengs- geschichte von Zahne and Gebiss der Rep- tilian. Beitrage IV. Uber die Anlange der Ersatzbegins. Arch, mikresk. Anat. 195: 265- 395. Woodward, M. F. (1893). — Contribution to the study of mammalian dentition. Part 1: On the de- velopment of the teeth of the macrcpodidae. Proc. zool. See. Lond. 5: 450-473. Ziegler, A. C. (1971). — A theory of the evolution of therian dental formulae and replacement patterns. Q. Rev. Biol. 46: 226-249. (1972). — Processes of mammalian tooth de- velopment as illustrated by dental ontogeny in the mole Scapanus latimanus (Talpidae: Insectivcra). Archs oral Biol. 17: 61-76. Journal of the Royal Society of Western Australia, Vol. 57, Part 2, August, 1974. INSTRUCTIONS TO AUTHORS Contributions to this Journal should be sent to The Honorary Editor , Royal Society of Western Australia, Western Australian Museum, Perth. Papers are received only from or by communication through, Members of the Society. The Council decides whether any contribution will be accepted for publication. All papers accepted must be read either in full or in abstract or be tabled at an ordinary meeting before publication. Papers should be accompanied by a table of contents, on a separate sheet, showing clearly the status of all headings; this will not necessarily be published. Authors should maintain a proper balance between length and substance, and papers longer than 10,000 words would need to be of exceptional importance to be considered for publication. The Abstract (which will probably be read more than any other part of the paper) should not be an expanded title, but should include the main substance of the paper in a con- densed form. Typescripts should be double-spaced on opaque white foolscap paper; the original and one carbon copy should be sent. All Tables, and captions for Figures, should be on separate sheets. Authors are advised to use recent issues of the Journal as a guide to the general format of their papers, includ- ing the preparation of references; journal titles in references may be given in full or may follow any widely used conventional system of abbreviation. Note that all illustrations are Figures, which are numbered in a single sequence. In composite Figures, made up of several photographs or diagrams, each of these should be designated by letter (e.g. Figure 13B). Illustrations should include all necessary lettering, and must be suitable for direct photo- graphic reproduction. To avoid unnecessary handling of the original illus- trations, which are usually best prepared between 1J and 2 times the required size, authors are advised to supply extra prints already reduced. Additional printing costs, such as those for folding maps or colour blocks, will normally be charged to authors. It is the responsibility of authors to adhere to the International Rules of Botanical and Zoological Nomenclature. Palaeontological papers must follow the appropriate rules for zoology or botany, and all new stratigraphic names must have been previously approved by the Stratigraphic Nomenclature Com- mittee of the Geological Society of Australia. Thirty reprints are supplied to authors free of charge, up to a maximum of 60 for any one paper. Further reprints may be ordered at cost, provided that orders are submitted with the returned galley proofs. Authors are solely responsible for the accuracy of all information in their papers, and for any opinion they express. Journal of the Royal Society of Western Australia Volume 57 1974 Part 2 Contents 5. The petrology of the Wooramel Group (Lower Permian) in the Lyons River area, Carnarvon Basin, Western Autralia. By G. J. McGann. 6. An annotated list of lichens from the coastal limestone near Perth, West- ern Australia. By N. C. Sammy and G. G. Smith. 7. New Information about the Quaternary distribution of the thylacine (Mar- supialia, Thylacinidae) in Australia. By M. Archer. 8. Petrology of chert artifacts from Devils Lair, Western Australia. By J. E. Glover. 9. The development of the cheek-teeth in Antechinus flavipes (Marsupialia, Dasyuridae). By M. Archer. Editor: A. J. McComb The Royal Society of Western Australia, Western Australian Museum, Perth 38522/6/74—625 WILLIAM C. BROWN, Government Printer, Western Australia JOURNAL OF THE ROYAL SOCIETY OF WESTERN AUSTRALIA VOLUME 57 PART 3 OCTOBER, 1974 PRICE: TWO DOLLARS REGISTERED FOR POSTING AS A PERIODICAL-CATEGORY B THE ROYAL SOCIETY OF WESTERN AUSTRALIA PATRON Her Majesty the Queen VICE-PATRON His Excellency Air Commodore Sir Hughie Edwards, V.C., K.C.M.G., C.B., D.S.O., O.B.E., D.F.C. COUNCIL 1974-1975 President Vice Presidents Past President Joint Hon. Secretaries Hon. Treasurer Hon. Librarian Hon. Editor R. M. Berndt, M.A., Dip.Anth., Ph.D., F.R.A.I., F.F.A.A.A. C. E. Dortch, B.S., M.Phil. L. J. Peet, B.Sc., F.G.S. P. E. Playford, B.Sc., Ph.D. P. G. Quilty, B.Sc. (Hons.), Ph.D. J. C. Taylor, B.Sc., Ph.D., A.R.C.S. P. G. Wilson, M.Sc. G. A. Bottomley, B.Sc., Ph.D. B. E. Balme, D.Sc. P. R. Wycherley, O.B.E., B.Sc., Ph.D., F.L.S. A. F. Trendall, B.Sc., Ph.D., A.R.C.S., F.G.S. M. Perry, B.Sc. (Agric.) (Hons.) G. Perry, B.Sc. (Hons.) S. J. Curry, M.A. A. Neumann, B.A. A. J. McComb, M.Sc., Ph.D. 9. — Amygdaloidal rock from Watheroo in the Permian Nangetty Formation Western Australia by J. E. Glover 1 Manuscript received 18 September 1973; accepted 16 October 1973 Abstract Amygdaloidal igneous rock has been recovered from the glacial Nangetty Formation at Tillite Creek in the Irwin River valley, Western Australia. The amygdales contain quartz, calcite, chlorite and epidote and are the same in other respects as amygdales in dolerite near Watheroo, 150 km south of Tillite Creek. The identity of these distinctive rocks confirms the essentially northward direction of Permian glacial transport suggested for the area, and may help to define the movement rather pre- cisely. Introduction The Permian (Sakmarian) Nangetty Forma- tion crops out in the valleys of the Irwin, Lcckier, Greenough and Murchison Rivers in the northern part of the Perth Basin (Fig. 1). Its glacial origin was recognized by Campbell (1910) and confirmed by Woolnough and Somer- ville (1924), and details of the formation at Tillite Creek were given by Clarke, Prendergast, Teichert and Fairbridge (1951). The last-named authors described many rocks from the hetero- geneous assemblage, and realized that some boulders of quartzite, breccia and chert have counterparts to the south in what has since been named Coomberdale Chert (McWhae et al., 1958, p. 12). Most of the other material at Tillite Creek mentioned by Clarke et al., i.e. “reddish or blackish hard fine-grained mud- stone”, gneiss, pegmatite, quartz-epidote rock, mica schist, quartz schist, porphyry, four variet- ies of granite and three varieties of epidiorite are believed to come from Archaean terrain east of the roughly north-south band of Moora Group rocks. Details of their provenance, how- ever, have not been established. Since the work of Clarke et al., a glacial de- posit at Bindoo Spring, 60 km northwest of Tillite Creek, has been examined by many geo- logists. This deposit contains fragments first identified by Playford and Willmott as Mt Scratch Siltstone, Enokurra Sandstone, Arrow- smith Sandstone, Beaconsfield Conglomerate and Arrino Siltstone, and their derivation from the Yandanooka area to the south-southeast is certain. There are also representatives of the Mcora Group, and shield rocks east of the Darl- ing Fault. The above data provide the main evidence for the generally accepted view that the direc- tion of Permian glacial transport in southwest- ern Western Australia had general westerly and northerly components (see for example Crowell & Frakes 1971, Fig. 10a). Apart from the work 1 Geology Department, University of Western Australia, Nedlands, W.A. 6009. * All specimen numbers refer to the collection of the Geology Department, University of Western Aus- tralia. of Clarke et al., few details of the evidence have been published, though most of it is in an un- published report by Playford & Willmot. This paper records the discovery of two new rock types at Tillite Creek, both of which have counterparts near Watheroo, 150 km to the south. One of the rocks, a red-brown siltstone with coarse dolomite grains resembles part of the Dalaroo Siltstone 2 km northwest of Watheroo. The other, with which this paper is concerned, is an amygdaloidal igneous rock of distinctive mineralogy and texture. Virtually identical rock has been recorded from bodies of rock that in- trude and slightly metamorphose Dalaroo Silt- stone within a strip of country 4X8 km im- mediately northwest of Watheroo by Teoh 1967 (see Fig. 1). The Tillite Creek rock (68275)* * and a specimen of amygdaloidal rock from Watheroo (60145) are described below. Petrography of the amygdaloidal rocks The igneous rock from Tillite Creek is grey- green and fine- to medium-grained and con- tains irregularly shaped to roughly ovoid, black, Figure 1.— Locality map showing Watheroo and Tillite Creek. The small rectangle north-west of Watheroo is the area within which amygdaloidal dolerite has been mapped by Teoh (1969). Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 65 and grey and black bodies up to a centimetre long that resemble amygdales, Under the mi- croscope about 60% of the rock between the amygdales is made up of highly altered plagio- clase laths with no apparent preferred orienta- tion, and the rest is largely chlorite. There ap- pears to be a palimpsest ophitic texture, but pyroxene is absent. There are numerous irre- gularly shaped masses of calcite in the rock: some plagioclase grains are completely converted to calcite, but most are only partly carbonated, and are charged with chlorite flakes. The un- altered plagioclase ranges from oligoclase to andesine. Epidote and sphene are scattered throughout the rock, and a few grains of he- matite, pyrite and a black opaque iron mineral are present. About 5% of the rock consists of the irre- gularly shaped to roughly ovoid bodies visible C in hand specimen. In their most characteristic form they resemble amygdales and have a nar- row discontinuous rim of calcite and quartz with a core of clear green, practically isotropic chlorite containing euhedra of epidote (see Fig. 2). Where the chlorite shows faint anisotro- pism, the anomalous interference tints suggest a micro-drusy structure of which there is little indication without the analyser. There is a tendency for grains of sphene in the rock to be concentrated near the margins of these ap- parent amygdales. In some places the quartz of the rim is lath-like, and seems to have re- placed plagioclase. Eight of the epidote crystals in the bodies described above were measured with the uni- versal stage. Forms identified are {001}, {100}, {110}, {101} and {111}. Six-sided crystals show- ing the front and basal pinacoids and a dome D Figure 2A. Specimen 68275 (altered dolerite) from Tillite Creek. Note the amygdale filled mainly with chlorite (light grey), euhedra of epidote (grey with dark borders), and the discontinuous rim of quartz (white) Small dark grains outside the amygdale are sphene. Plane-polarized light, x 25. Univ. West. Aust Geol Dept negative No. P2612. B.— Specimen 68275, same amygdale illustrated in Figure 2A. The chlorite is black, and some epidote shows twinning. Crossed polarizers, x 25. Univ. West. Aust. Geol. Dept, negative No. P2613. C.— Specimen 60145 (altered dolerite) from Watheroo. Note the amygdale filled mainly with chlorite (light grey), the euhedra of epidote (grey with dark borders), and the discontinuous rim of quartz (white). The small' dark grain in the amygdale (left centre) is sphene: most of the sphene is just outside the amygdale. The resemblance to speci- men 68275 from Tillite Creek is clear. Plane-polarized light, x 36. Univ. West. Aust. Geol. Dept, negative No P2610. D— Specimen 60145, same amygdale illustrated in Figure 2C. Crossed polarizers, x 36. Univ. West. Aust. Geol. Dept, negative No. P2611. Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 66 are common in thin section, and some show twinning on 100. The pleochroic scheme is X r colourless, Y pale yellow, Z = olive-green with absorption X < Y < Z. Dispersion is pro- nounced with r > and r < v in different crystals, suggesting crossed dispersion. XAc (7 grains) measures 2°, 4°, 5°, 5°, 5 h° , 6°, 7°, and ( — )2V (4 grains) measures 68.5°, 69°, 70°, 75°, with measurements reproduc sable to within 2°. These properties are the same as those for epidote in the pistacite range (Winchell & Winchell 1356, Fig. 343). There are other bodies in the rock in which the minerals are arranged differently. In a few the calcite rim is thick and continuous, and chlorite is restricted to a small core. Some of the bodies have a discontinuous rim of quartz, an irregularly shaped inner layer of slightly fibrous chlorite containing epidote euhedra, and a core of granular calcite and very fine grained quartz. Other, commonly complex bodies, have cores either of calcite or microcrystalline quartz. Some very irregularly shaped masses, generally made up essentially of chlorite, epidote and cal- cite, have only vaguely defined boundaries with the surrounding chlorite-rich and calcite-rich groundmass, and have the appearance of having replaced the rock rather than having filled ca- vities in it. Nevertheless, for simplicity all these bodies are from now on called amygdales, and they have one consistent feature: the epidote either penetrates the chlorite or is enclosed by it. The amygdaloidal recks at Watheroo resemble the Tillite Creek specimen described above. Specimen 60145 was collected from a dyke-like body that roughly parallels and locally crosses the Geraldton Highway between about 5 kilo- metres and 10 kilometres north of Watheroo. Outcrop of rock similar to that of specimen 60145 is found 0.15 kilometres north of the Longreach turnoff, 9 kilometres by road north of Watheroo. Specimen 60145 is grey-green, fine-to medium- grained, and contains irregularly shaped to roughly’ avoid, black, and light grey and black bodies up to a centimetre in diameter. In thin section there seems to be a relict ophitic texture, but in places where most of the original minerals have been changed, there is little evidence of such texture. Laths of plagioclase have com- monly been replaced by chlorite, epidote and calcite, and locally by quartz. Interstitial ma- terial consists partly of pale brown and pale green fibrous amphibole, very rarely with a core of augite. There are also irregularly shaped patches of chlorite and calcite, and epidote in the form of clear crystals and cloudy fine- grained aggregates is common. Other consti- tuents include small grains of sphene, an opaque mineral (probably leucoxene) and fine aggre- gates of interstitial quartz. The rock is probably an amygdaloidal quartz dolerite that has been extensively auto-metamorphosed. Some amygdales resemble those in the Tillite Creek boulder very closely (Fig. 2). In general, they have the same mineralogy and show the same range in mineral arrangement and their contained epidote euhedra exhibit the same mor- phology and pleochroism. There is a similar tendsney for small grains of sphene to be con- centrated in the rock near the amygdales. Measurements with the universal stage on epi- dote in specimen 60145 are as follows: XAc 3i°, 5° (2 grains) (— )2V 70°, 71° (2 grains). Dispersion is pronounced with r>v and r < v in different crystals, suggesting crossed dispersion. The epidote is thus indistinguishable optically from epidote described in the Tillite Creek rock. Other similar Watheroo rocks contain amyg- dales with a higher proportion of epidote, and grade into rocks in which the epidote resembles glomeroporphyritic aggregates. Discussion Epidotized amygdaloidal rocks have been re- corded from the Proterozoic Fish Hole Dolerite in the Kimberley Region by Dow & Gemuts (1969, pp. 28, 30-31), but apart from the Watheroo rocks, intrusive amygdaloidal bodies have not been reported elsewhere in Western Australia. The striking similarity of the ma- terial from Tillite Creek and Watheroo makes their common origin very likely. The amygdaloidal dolerite at Watheroo was mapped as dykes cutting the Dalaroo Siltstone both by Teoh (1967), who discussed its petro- logy, and by Low (1969), who did not distin- guish it from the other, non-amygdaloidal dol- erite in the area. The strike of the amyg- daloidal dolerite parallels that of the Moora Group fairly closely, and it may be at least locally concordant. Detailed mapping is neces- sary to check its structural relationships to the Dalaroo Siltstone and to determine its extent. It has not been recognized far north of Teoh’s area, and is known to be absent from the Moora area. If it proves to be restricted to the vicinity of Watheroo, it will demonstrate rather preci- sely the direction of travel of Permian ice erod- ing the Watheroo area, and will indicate the distances of transportation involved. References Campbell, W. D. (1910).— The Irwin River Coalfield and the adjacent districts from Arrino to North- ampton. Bull geol. Surv. West Aust., 38. Clarke, E. de C., Prendergast, K. L., Teichert, C. and Fairbridge, R. W. (1951). — Permian succession and structure in the northern part of the Irwin Basin, Western Australia. J.R. Soc. West. Aust. 35: 31-84. Crowell, J. C., and Frakes, L. A. (1971). — Late Palaeo- zoic glaciation of Australia. J. geol. Soc. Aust., 17: 115-156. Dow, D. B., and Gemuts, I. (1869) .—Geology of the Kimberley Region, Western Australia: the East Kimberley. Bull. geol. Surv. West. Aust., 120. Low, G. H. (1969).— The geology of the Moora Group. Rec. geol. Surv. West. Aust. 1969/5 (unpub- lished) . McWhae, J. R. H., Playford, P. E., Lindner, A. W., Glenister, B. F., and Balme, B. E. (1958). — The stratigraphy of Western Australia, J. geol. Soc. Aust., 4 (2) : 1-161. Teoh, K. T. (1967).— The geology of the Watheroo area, Western Australia. B.Sc. Thesis, Univ. West. Aust. (unpublished). Winchell, A. N., and Winchell, H. (1956).— “Elements of Optical Mineralogy, Part II”. Wiley, New York. Woolnough, W. G., and Somerville, J. L. (1924).— A con- tribution to the geology of the Irwin River Valley of Western Australia. J. R. Soc. N.S.W., 58: 67-112. Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 67 10. — Australites of mass greater than 100 grams from Western Australia by W. H. Cleverly 1 Manuscript received 20 November 1973; accepted 18 June 1974 Abstract Seventeen australites from Western Australia in the mass range 100-437 g and including the most massive australite known have been exam- ined. Most of the other fifteen previously re- corded Western Australian specimens in that range have been re-examined. All are round, oval or dumbbell shaped cores. Where possible, the forms and dimensions of the primary bodies have been assessed and thence the percentage losses involved in forming the remnant cores. Average volume loss was only 46%. The distri- bution of the sites of find has been considered and the south western portion of the state is confirmed as the principal area of infall of australites of mass exceeding 100 grams. The distribution suggests the possibility of mass grading related to the northern boundary of the australite strewnfield. Introduction Australites (Australian tektites) of mass ex- ceeding 100 g are rare, comprising only a few hundredths of one percent of known specimens. The purposes of this paper are to describe 17 Western Australian specimens in the mass range 100 to 437 g and to supply additional details for any other available specimens for which pub- lished descriptions are incomplete. Because all australites under consideration are remnant shapes from which a stress shell has spalled, they have certain morphological features in common. A general procedure was therefore possible when examining the specimens. General principles and procedure Larger specimens were weighed on a Mettler, K-type, top-loading balance which had also arrangements for bottom suspension of the specimen in de-ionized water for determination of specific gravity; for lighter specimens and chips, a more sensitive chemical balance was used (Table 2, cols. 6 & 8). The conventional statement of dimensions in the sequence length, breadth and thickness (Table 2, col. 7) has a simple relationship to australite orientation during flight downwards through the atmosphere. Because nearly all australite primary bodies were either spheres or rotational shapes with considerable sym- metry, they almost invariably adopted a stable flight orientation, generally presenting a broad face forward. Thus the length and breadth (or a diameter) were in a plane normal to the line of flight and the thickness was parallel to the line of flight. The thickness was especially re- duced during flight and the length and breadth 1 W.A. School of Mines, P.O. Box 597 Kalgoorlie 6430. Honorary Associate, Western Australian Museum, Perth. were also affected, but the same relationship of the dimensions to the flight orientation generally applies to the remnant core as for the primary body. The plan view in flight position is thus a view of the posterior surface of flight looking in the direction of flight and this view shows length and breadth or diameter. The plan view dimen ions of australite cores are often but not necessarily defined at the level of the rim (see below) . The shape type of the core (Table 2, col. 2) is the shape seen in plan view. Ideally, a round core is perfectly circular in plan view. Weather- ing may lead to differences of 2-3 mm in the dimensions of a core of (say) 50 mm diameter, but classification as a round core may be pre- ferred. Similarity of the various profiles through the posterior pole of the core is also a criterion which aids the distinction between round and broad oval cores; the distinction is none the less subjective. The arbitrary width/length ratios of Fenner (1940 p.312) were applied strictly in defining broad oval, narrow oval and boat shapes, resulting in a nomenclature change for one pre- viously described specimen (No. 28. Table 2). The dumbbell shape is self-explanatory. Most well-preserved large cores have a rim separating the posterior surface of flight from the surface exposed by loss of the stress shell. The posterior surface is a modified remnant of the surface of the primary or parental body which suffered shape modification during pas- sage downward through the earth’s atmosphere. The form of the posterior surface is thus a guide to the form and dimensions of the primary body. The stress shell was a shell of glass a few millimetres thick immediately beneath the anterior surface of flight where fusion and ablation stripping were active during the earlier hypersonic phase of atmospheric transit. The shell had been first heated and then cooled. Opinions differ as to the timing of the loss of the stress shell — whether in a late stage of flight, on impact, or as the result of terrestrial processes such as diurnal temperature changes. Small cores retaining partially detached por- tions of the stress shell are not uncommon, but specimens of this kind (known as “indicators”) are unknown amongst cores of mass exceeding 100 g, evidently because the greater total amount of expansion and contraction resulted in a more efficient spalling of the shell. The surface ex- posed by loss of the stress shell may be relatively irregular when compared with the posterior surface. Thus, on well preserved cores, the relatively smoothly curved posterior surface terminates at a rim anterior to which the dimensions decrease Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 68 to those of the surface exposed by loss of the stre s shell. Correct interpretation of the “facing” of the rim is critical to identification of the posterior surface of flight and hence to deductions concerning the form and dimensions of the primary body. In a few specimens, ter- restrial weathering has removed or considerably modified parts of the rim, but in only one of the specimens under consideration was the flight orientation in doubt. Anterior to the rim on the surface exposed by loss of the stress shell (the present anterior surface as distinct from that of flight) there may be a distinct shoulder inherited from the ablation form (see for example Fig. 3-17). In such examples an equatorial zone is defined be- tween the natural limits of rim and shoulder. In others, the shoulder is indistinct or absent and the anterior limit of the equatorial zone is then ill-defined, or the zone may exist only in the sense that it usually has a distinctive minor sculpture (Fig. 2-9). The form of the posterior surface was esti- mated by fitting curves to profiles traced on enlarged silhouettes projected with a lantern, the specimen being oriented as in flight. The degree of enlargement used (about X4) was cal- culated for each by comparison of the dimen- sions of the silhouette with those of the speci- men. Profiles were traced on cm-mm graph paper with the rim of the specimen aligned with one set of lines of the paper. When a core is circular in plan view, or nearly so, and the constructed arcs of circles fit closely the various profiles through the posterior pole of the specimen as seen in side elevation and have much the same radius, it is assumed that the primary body was a sphere. When a core is elongate and the posterior surface has distinctly different transverse and longitudinal profiles in side elevation, a simple possibility is that the parent body was a pro- late spheroid which oriented with the long axis normal to the flight path. For such spheroids the radius of the arc fitted to the transverse profile through the posterior pole is identical with the semi-minor axis of the elliptical longi- tudinal profile. The mathematical origin of the ellipse was plotted, and by using the co- ordinates of a point on the best preserved part of the longitudinal profile in the general equa- tion of the ellipse, the semi -major axis was calculated. The positions of the foci were then calculated and an ellipse drawn to test the correctness of fit. Except when badly eroded (e.g. No. 28) these specimens presented little difficulty. When a core is round and all profiles through the posterior pole are closely the same ellipse, the core was probably derived from an oblate spheroid which oriented with the short axis parallel to the flight path. Unlike the previous example, neither semi-axis of the ellipse is directly measurable. The method used was to judge the approximate position of the major axis, calculate the ellipse as before, and draw a trial ellipse on tracing paper superimposed on the diagram. The position of the major axis was then adjusted 8 mm (i.e. c. 2 mm on true scale) and the calculation repeated, in general, three or four trials with reduced degues of adjustment sufficed to produce a well fitting ellipse, the dimensions of which were known to 1 or 2 mm on true scale. Mathematical methods could probably be devised e.g. by fitting a regression line to points selected from the best preserved parts of the profile, but such methods tend to give a rather spurious aura of precision to these estimates based upon pitted and other- wise imperfect surfaces. Neither was a lens measure used in a quantitative way, but it was found to be a valuable aid for the detection of non-spherical surfaces. A narrow, double thick- ness ( 0.05 mm) of transparent adhesive tape with interleaved strip of cm-mm graph paper was attached along the profile to be examined to smooth out minor irregularities before applying the lens measure. Lens measure readings taken along six profiles outward from the posterior pole of specimen No. 5 (Table 1) illustrate the increasing degree of curvature towards the rim of a surface which, to the eye, appears to be part of a sphere (Chapman 1964 Fig. 6A). Table 1 Lens measure readings for posterior surface of australite core from Lake Yealering Distance between mid point of lens measure and posterior pole of core Range of lens measure readings for six profiles Mean lens measure reading for six profiles cm units units 1 10.5-13 12.2 1.5 11.5-15 12.7 2 12-16 14.0 2.5 14-18 16 0 3 15.5-18 5 17.5 Arcs of circles will fit considerable lengths of ellipses to within the thickness of construction lines (see for example Baker 1956 Figs. 21, 31, 32). As ellipticity increases, the possibility of detecting departure from spherical shape im- proves and it is also better in those specimens where losses from the primary body have been relatively small. Blown sand or other terrestrial agents can cause complications e.g. by flattening the polar area of the spherical surface of a core to produce a form approximating an oblate spheroid. If an oblate spheroid with the short axis parallel to the line of flight is mistaken for a sphere, the volume of the primary body will be over estimated and likewise the percent- age losses (see below). A round core calculated as a sphere and having unusually high loss figures may therefore be suspected as having been derived from an oblate spheroid. The writer finds it difficult to believe that a prolate spheroid would be stable with the long axis parallel to the flight path, but if such a spheroid were incorrectly calculated as a sphere, the volume of the parent body would be under- estimated and this could become evident as unusually low percentage losses. Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 69 Table 2 a & m , Oi , wo)'tnMio : X ^ O : •1-05 ■ I>- CO :®XHOXM : d . lO CO CO • ■ 50 : CO r— < ^ Oil . Cl : X rf -f 05 nnxco; -+ 01 05 ■ -f ' X 05 t- 01 : X O : N 05 O 05 ® Tf • X :xn® •tf r- . in -* . r— W : CO Ol :^«0® ^MOXNNM® co co < 1 O CO Cl H T1 - M C X •* 05 05 d x r— : os co x h m in so 01 ■ to to CO X X : -f t> : d -+ 01 m rft 0 : -f :Nh05 05 :*t7??7?7CCiOCO^ : O CO CO : r—i CO : Ol 1-1 !M 1— 1 : 1— l X d -* r-t -f x — — X TO : t- : Cl -f 1 — — & X a 0 i- t> 0 m *+ x -f -f -t m to co x-d X X X X XX X X X ^ X X X ^ g 7” m X T-l g g . TO d . . i-H _• . X t-l rin'T g? d? : * . g £> 1— 1 d X * g®> * t- i- n n • ^ • TO ■ 'Ifl Cj ' -+ TO ■ x ^ t: in rt in -f • s 3 m to CD g Cl ■C Xj- XX'S'C 3 X T3 X r O XX ®xox — X X ■C X X 0 * -f Tt< X Tf ® Tfl N 'f Tt< -ft- X t-l X -f X d m x «-! X co t- c- X X TO 0 X TO TO m to x to 7 5 5 5 mm 1 — m t— to m os -f m x in r- x x m -f t- O XC5Tf050C5C1050t— iCl 0 t- t-i to m d 1 - d 0 01 CO CO CO CO 01 Ol CO CO Cl — T 1 CO CO co 0 01 CO X Cl X Cl X Ol '*i* co Ol -+ -t -f-f-f-f-f-f-f-f-+ : Tf T+ -+ : -f -f -f -f Tf -f -f Tf -f -f -f -f -f -f . -f -+ -f Cl Cl OJ Ol 04 Ol 04 Ol Ol Ol Ol Cl Cl Cl ■ (M d Cl Cl Cl d Cl Cl n X X : X X X X : X X X 0 CO X X X -f X X X X ' X X ' X X X 05 05 -f -f X^ m • t> X CO 05 CO TO t- 05 X X t-l X x X 1- 05 0 m 05 TO X rH TO X X 7 1 x X m TO m 05 Cl m X m to t-i m TO O 0 0 X -Tf CO t— m TO d m m m CO -t m -f m -+ m ~t -f m m m CO *4* -+ -f X m X X t> m to m m m X Tf X X X X m m X X X d m X X X X X X X X X X X X X X X 05 X X X TO 05 05 X X Tf d CO X TO X rf X 05 t- TO Tf T* CO r- CO eo X 30 X 05 Cl (M 1- X Pf ci nwxo-t't’fiux i'- n 01 >- *t wr-ooo!XNto ® o m o Ol Cl Cl OJ —I — — t-l — ■Hr-rT- C1 C --I X O • - X 0000 ;oooooooooooo .0000 r-< ^ o o X X Cl 0000 : X X t— :XNX TfiOOKMOOTf (MiO^O v .00000000000 XXt>XXXXt>Xt-t-lt> t— ( i—t t— — - C 1 -! t— 1 t— t t— t ni »— * : o o o o O o ^ • © >> •d-* £ Tf 0J rH £ 6 •H^Q £ 0 's t-l • w o _Q o'Sg’o 7S £ _ ^ ^ be sc 2 3 ~ ft ^XO^Jrf C3 r— 3 t, X< o ci O C5 o «J !> X s | s _ Jr g m _£ to -=■ d O H m d ffl d g ’ ‘3j cS tS tS cS >s > > ■ > > > "3 ”3 ^ ‘ ‘ O > > >■ O > H 0 0 0 0 0 ,Q O OOO _ O D rW — - •o "w ft ft ft ft • — ^ "w CC "w ft ft ft 2 2 2 2 ; — : ft ft o 3 0 Q S T 3 r- ' Q g TS «S 3 5 2 5 ■ 52 gs 5555335 Sh D. 3 ' cS s t-i O t_ Odd 0 0 0 0 0000 sj O C S 3 O dddOOCC tS O X po (3 ft ftft — — ~ ~ ft ft ^ ft ftftftftftftft PC iO ® X ® O ' rCIW^iO Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 70 The dimensions of a spherical primary body are stated by the diameter; for a prolate spheroid, dimensions are in the form 6x4x4 cm and for an oblate spheroid, in the form 6x6x4 cm (Table 2, col. 9). From the dimensions of the primary body as determined above and on the not inconsider- able assumption of mathematically perfect form, the volume was calculated. The present volume had been determined incidentally to the measure- ment of specific gravity. The total percentage loss could thus be calculated. With the further assumption that the primary body had the same specific gravity as the remnant core, the percentage loss of volume is also the percentage loss of mass. The total thickness loss from the primary body was calculated from a considera- tion of the thicknesses of the primary body and the remnant core. Total losses from the primary body include those of fusion and ablation stripping during flight, spallation of the stress shell, terrestrial losses as the result of both chemical and physi- cal processes, and the artificial damage of some specimens. Losses as a result of certain terrestrial pro- cesses such as chemical etching by soil water and biochemical etching by plant roots and hyphal filaments cannot be estimated quantita- tively. It is believed that such losses will not generally exceed a few percent if expressed as a percentage of the primary body and will usually be of a smaller order of magnitude than flight losses. The losses, whether natural or artificial, of flakes which transect the otherwise smooth form of a core can be reasonably estimated by com- pleting the form with modelling clay and con- verting the added mass of the clay to that of the same volume of glass. This was done where - ever possible and the natural flake losses sub- stracted from total losses. Restoration of arti- ficial losses, which are characterised by the brilliant lustre of the exposed glass, enabled estimation of the mass of an artificially damaged specimen when found. The restored mass of an artificially damaged specimen is shown in brackets immediately preceding the existing mass (Table 2, col. 6) ; likewise a restored dimension immediately precedes the existing dimension (col. 7). The loss figures given in Table 2 cols. 11 and 12 are thus essentially those arising from atmos- pheric flight or its aftermath (stress shell) but including also terrestrial losses which have been minimised by allowance for natural flake losses. Australite primary bodies had a complex in- ternal flow structure, the schlieren differing slightly in their chemistry; they also contained bubble cavities of various sizes and of irregular distribution. Arising partly from the consolida- tion of the initially molten primary body and Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 71 from its heterogeneity, partly from stresses im- posed by atmospheric transit and perhaps also by terrestrial processes, autralites retain various degrees of residual strain. Terrestrial processes developed a variety of minor surface sculptural effects such as gutters (for details see Baker 1959) which often reflect these internal hetero- geneities and strains. The minor sculpture de- veloped on some specimens was partially removed during a later episode of transport or exposure to blown sand. Other specimens which had been abraded became buried and were sub- sequently chemically etched by the constituents of soil water. The better preserved specimens show minor features of the posterior surface such as flow swirls (Fig. 3-12) which were probably original features of the primary body, now accentuated by minor degrees of etching. Only the more prominent surface and sculptural features are noted in the descriptive section below. Descriptive Notes This section should be read in conjunction with Table 2 which shows physical data of the specimens. Figure 1, showing sites of find, if specifically known, should also be consulted. The abbreviations W.A.M., S.A.M., and W.A.S.M. refer to accession numbers in the collections of the Western Australian Museum (Perth), South Australian Museum (Adelaide) and the W.A. School of Mines (Kalgoorlie) respectively. De- scriptions follow in the number sequence of Table 2, col. 1. 1. Fig. 2. Cast: W.A.M. 13238. The original is owned by Mr. P. Repacholi, who found it in 1969 whilst ploughing to a depth of c. 15 cm in the N.E. corner of Avon Location 15085, about 3 km W. of Notting railway siding. It is the most massive australite known. The form is much modified by natural flake losses which have removed the rim except for a 5 cm length along one side and a 1 cm length elsewhere. Resulting also from flake losses the posterior surface is less symmetrical than the anterior. Surface dating from the time of arrival on the earth’s surface is considerably weathered, the sculpture comprising pits of 1-2 mm diameter transitional through oval pits into short gutters, and larger composite pits which contain two or three oval pits within them. Longer gutters are restricted to remnants of the equatorial zone where they are oriented approximately parallel to the flight path and to three larger flake scars in that zone where they are oriented approximately parallel to the rim. The largest and most recent flake loss has removed much of one end of the core and the scar is characterised by circular and lunate bruise marks and is dulled by abrasion. Arti- ficial damage comprises a 2 cm trail of milli- metre sized scars, perhaps attributable to con- tact with the plough. The transverse curvature of the posterior surface can be reasonably estimated, but calcu- lation of the elliptical longitudinal section had to be based principally upon a 5 cm length of pitted surface. Natural flake losses are equiva- lent to c. 3.5% of the mass of the primary body. 2. W.A.S.M. 8925. For a detailed description and illustrations of this specimen from near Warralakin, see Baker (1962). 3. Fig. 2. W.A.M. 12318. Found near Newde- gate. Briefly described by McCall (1965) with illustration of the anterior surface. The form is sub-spherical (thickness/mean diameter ratio 0.91). No rim nor defined equatorial zone is present but there is a central girdle up to 2 cm wide characterised by a complex of gutters. The posterior surface has some patches of short gutters. The anterior surface, which is rather irregular, has some roughly circumferential gutters surrounding a complex of short gutters. The precise flight orientation, and hence the dimensions as conventionally stated, are in some degree a matter of opinion. Arcs of circles fit reasonably well the profiles of the posterior surface. 4. Fig. 2. Found by Mr. H. Biggin in 1940 or 1941 in the N.E. Corner of Avon Location 19835, about 14 km W. of Kondinin. Owned by Mrs. H. Biggin of “Karingal”, Kondinin. The specimen is fairly well preserved except for shallow natural flake losses from the pos- terior surface resulting in a tapered appearance in cross section. The longitudinal profile is not so badly affected and a reasonable estimate of the primary form is possible. There have been no artificial losses. The posterior surface, which is somewhat dulled by abrasion, has some circu- lar and lunate scars and etched flow swirls. The equatorial zone is reasonably defined and has some short gutters, variously oriented. The anterior surface is asymmetrical in cross section and has composite pits containing short gutters and circular and lunate etched scars. Some of the gutters tend to be circumferential near the periphery. This suface is much like that of No. 1 which was found only 11 km distant. 5. W.A.M. 4455. Found on Lake Yealering which adjoins the Yealering townsite. Some de- tails and illustrations of both surfaces were given by Fenner (1955), who noted the deeply pitted posterior surface and the unusual degree of development of gutters on the anterior sur- face (his PI. VII Figs, 2, 1 respectively). Four views were figured by Chapman (1964, Figs. 6A, 7), who noted that gutters had been developed selectively on the surface exposed by loss of the strees shell. The diameter of 67.4 ± 0.5 mm is the largest for any known round core. 6. Owned by Mr. L. P. Berryman, who found it on Lake Ballard in 1968. Described and figured by Cleverly (1971). 7. Fig. 2. W.A.M. 12992. Found by an aborigine about 1920 and given to Mrs. C. Parrot; donated to the W.A. Museum by Mr. G. Woodland in 1969, by which time it was uncertain whether the site of find had been Narrogin or Narembeen (Fig. IB). Extensive but shallow flakes have been arti- ficially struck from posterior and anterior sur- faces, and an earlier, natural flake loss has re- moved the rim from around one . end of the Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 72 Figure 2. — Australites from Western Australia numbered as in text and Table 2. In all elevational views the anterior surface is towards bottom of page. la —Posterior surface. Length 83.7 mm. lb— Anterior surface. lc. — side elevation showing greater regularity of anterior than posterior surface. Id. — End elevation. Width 54.5 mm. 3. — Elevation Angularity at upper left is the result of natural flake loss. Width c.60 mm. 4. — Side elevation. Length 64.8 mm. 7.— Side elevation. Length 67 mm. 9.— Elevation showing wedged anterior profile Width 55.6 mm. 12. — Side elevation. Width 59.9 mm. Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 73 specimen. The posterior surface has unusual pits containing narrow, parallel or divergent, short gutters developed on bundles of more etchable schlieren. The remaining rim is reasonably de- fined and regular. The equatorial zone has some short gutters oriented normal to the rim and also a patch of deep pitting. The anterior sur- face has some pits transitional into short gutters. Because of artificial damage, the form of the primary body can be assessed only very ap- proximately. It is possible that the transverse profile was elliptical i.e. that the primary form was triaxial; if so the mass and percentage losses have been overestimated. 8. W.A.M. 12884. Found about 18 km W. of Notting railway siding. The lustre is dulled by abrasion. This is one of the few specimens examined which tends to bulge slightly ante- rior to the rim. The posterior surface has some lunate scars and small pits. The equatorial zone, which averages 15 mm wide, has abundant gut- ters which are variable both in orientation and form (some parallel to flight path, others ver- miform or branching). The anterior surface has some gutters and bruise scars shallowed by abrasion. Arcs of circles do not fit closely the profiles of the posterior surface and have radii in the range 3. 1-3. 3 centimetres. Because of these un- certainties, a mean figure was used to calculate the primary body as a sphere. 9. W.A.M. 12993. Turned up by a bulldozer from shallow depth in sandy soil about 19 km N. of Ongerup in 1968. This specimen has much the same form and unusual features as No. 3 above. There is no rim or defined limit to the equatorial zone but there is a band of distinctive sculpture. The posterior surface is somewhat irregular as the result of flake losses and has an area of deep pitting. Gutters are present only on the anterior surface with a suggestion of circumferential pattern. The anterior surface is strongly wedged (Fig. 2), thus presenting very different profiles when viewed from different di- rections. Arcs of circles fit profiles of the pos- terior surface fairly well and continue to fit over parts of the “equatorial zone”. 10. Geological Survey of W.A. R.2024. Found near Cuballing. Described in detail and figured by Baker (1966) prior to its donation to the Survey collection. 11. W.A.M. 12843. Found at Graball, E. of Narembeen. Described and figured by Baker (1963) prior to its donation to the Western Aus- tralian Museum. Because the percentage mass loss is the highest for any of these large cores, derivation from an oblate spheroid is suspected and this is supported by lens measure readings. However, because of the extensive etching and minor artificial damage to the posterior surface, no re-estimation of the primary body on that basis has been attempted. 12. Figs. 2 & 3. W.A. Government Chemical Laboratories Mineral Division collection 1678. Found about 5 km S.E. of Corrigin in 1955. The posterior surface is relatively well preserved and shows minor sculpture ranging from barely per- ceptible schlieren through distinctly etched flow swirls to short gutters and a single elongate gutter paralleling the pattern of a flow swirl. The rim is reasonably defined and regular except where affected by minor flake losses and the development of an area of deep pitting. The equatorial zone has some gutters oriented normal to the rim; its limit is ill-defined. The anterior surface is distinctly wedged and shows a few gutters. 13. Cast W.A.M. 13237. Original owned by Mr. R. Kirkpatrick, who found it on Mr. C. Adams’ property “Marambeena”, c. 25 km S.W. of Chil- lilup in 1972. Neither the rim nor the anterior limit of the equatorial zone is clearly defined. The form of the posterior surface is considerably modified by the natural loss of small flakes. The equa- torial zone is c. 10-18 mm wide and the minor scuipture is principally short gutters, variously oriented. The anterior surface is roundly and asymmetrically wedged parallel to the longer diameter. The low specific gravity (2.410) is probably to be ascribed to bubble cavities. The arc of a circle will fit well only to the transverse profile and the radius of that arc has been used in calculating the primary body as a sphere. 14. W.A.M. G8978. Described by Simpson (1939) with two unspecified views, the first an elevation with anterior surface towards top of page, the second the (presumed) anterior surface showing artificial damage. Bowley (1945) concluded that the specimen had been found between Narrogin and Merredin (Fig. 1) and this is feasible because several unusually massive australites have since been found within that general area. 15. In Geological Survey of W.A. collection. Found about 16km N.W. of Ongerup. Fully des- scribed and figured by Baker (1967). 16. Found near Kalgarin in 1960. The finder shattered the core by a heavy blow with a hammer. The larger pieces were given to Mr. R. Pugh, who re-assembled them, but clearly many small fragments were not recovered. Dismemberment disclosed a breached bubble cavity 6 mm diameter located just off the axial line and slightly closer to the posterior than to the anterior surface. Another cavity 2 mm dia- meter and a few smaller ones were also re- vealed. Fragments macroscopically free of frac- tures were selected for determination of specific gravity, the choice being thereby limited to three pieces of total mass 12.5 grams. The spe- cific gravities are in the range 2.435-2.442 with a weighted mean 2 439. These fragments cons- titute only about 10% of the material, and because of exposure of bubble cavities, the spe- cific gravity is biased towards the higher value for australite glass rather than representative of the australite as a whole. The reassembled specimen is sub-spherical (mean diameter/ thickness 0.93) with a defined but rather sinuous rim and equatorial zone. The posterior surface has a meridional strip from rim to rim containing short transverse gutters. Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 74 Figure 3. — Australites from Western Australia numbered as in text and Table 2. In all elevational views the anterior surface is towards bottom of page. 12.— Posterior surface showing etched flow swirls. Diameter 56.9- 55.8 mm 17.— Elevation showing well defined equatorial zone with gutters and natural flake loss from posterior pole. Width c. 53.5 mm. 19.— Anterior surface with artificial damage at upper left and natural flake loss from rim at lower edge. Length 79.6 mm. 21.— Side elevation. Central area is artificially abraded. Width 51.4 mm. 22a. — Anterior surface. Length 61.3 mm. 22b.— Side elevation showing area of deep etching affecting the pos- terior profile. 23.— Elevation showing globular form. Thickness 43.6 mm. 26.— Side elevation. Length 50.4 mm. 30. — Side elevation. Length 54.7 mm. Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 75 The equatorial zone also has gutters, some oi which are meandrine. Because of the history of this specimen, all quantitative estimates of the core and its primary body should be re- garded as very approximate. 17. Fig. 3. W.A.M. 3491. Found near Corrigin. Fenner (1934, 1955) made brief references to this specimen, in the second publication under the incorrect accession number 3441. Chapman (1964, Fig. 11) illustrated a side elevation chosen to emphasise the degree of imperfection (flake losses from posterior pole and rim and an ex- tensive area of deep pitting) when compared with cores from elsewhere. The rim is fairly well preserved and regular. The posterior sur- face has some circular groove structures and some roughly circumferential gutters near the periphery. The equatorial zone has gutters oriented approximately normal to the rim. On the anterior surface, short gutters near the equa- torial zone are oriented more or less circum- ferentially. Except in areas of imperfection noted above, arcs of circles fit fairly well the profiles of the posterior surface. 18. Australian Museum, Sydney DR: 7533. Found at Lake Grace. The form is extremely irregular and asymmetrical as the result of natural flake losses from both posterior and anterior surfaces, and artificial losses caused by a severe blow on the posterior role of the specimen. The rim is represented only by worn remnants and the limit of the equatorial zone is poorly defined. Short gutters, variously oriented, are present on the equatorial zone. A complex of oval pits and short gutters is present on the major anterior flake scar. The specific gravity is well below average for large cores from Western Austra- lia. No reliable estimate of the primary body is possible. 19. Fig. 3. W.A.M. 12264. Found about 26 km E. of Kulin in 1960. The dull posterior surface contrasts with the “lacquered” appearance else- where. Some circular and lunate scars and two small areas of gutters on the posterior surface have been shallowed by abrasion. The rim is ill-defined and sinuous and there is no distinct anterior limit to the equatorial zone, though gutters show the usual orientation approx- imately normal to the rim. A reasonable es- timate of the parental form is possible. 20. W.A.M. 12960 (formerly Geological Survey of W.A. collection 11177). Salient details and an illustration of the anterior surface were given by Simpson (1902, p. 81 and Pl.I), who described the site of find as being “100 miles East by South of Weld Springs, or say about Lat. 25° 30' S, and Long. 123° O' E.”, a point closely coincident with Lake Buchanan. The W.A. Museum catalogue records that the speci- men was found by the Calvert Expedition of 1896-7. The presumed year of find was 1896 because Lake Buchanan was named by Surveyor L.A. Wells of the Calvert Expedition in that year (pers. comm, from W.A. Surveyor General’s Department). An oblique view emphasising the posterior surface was figured by Thorp (1914 PI. XVIII Fig. 6). who added incorrectly that the specimen was found at Weld Springs by the explorer John Forrest; Fenner (1934) repeated the statement regarding the finder. Forrest was not associated with either the Calvert Expedi- tion or the subsequent relief and search efforts; he had named The Weld Spring 22 years pre- viously. The fore-going is given in some detail to em- phasise that the place and circumstances of find of this core are well authenticated. The site of find should not be stated as “Weld Springs” (i.e. The Weld Spring) ; nor should Lake Buch- anan be confused with Lake Buchan, which is more than 900 km distant to the south west. The rim of the core is well defined but there is no distinct limit to the equatorial zone. The posterior surface, where not affected by ex- tensive but shallow artificial flake loss, has some circular pits transitional to circular gutters set in a surface of small scale “hammered metal” appearance. The anterior surface is much like the posterior but with the etching rather more advanced. 21. Fig. 3. S.A.M, T191. Fenner (1955) re- ported briefly on this specimen. Despite the fairly bright lustre, the core appears to have suffered considerable physical erosion. Small natural flake losses, an extensive but shallow artificial flake loss, and artificial abrasion of a small area on the equatorial zone have further contributed to the general irregularity and im- perfection of the form. No rim is present and the assumed orientation depends partly upon the greater degree of asymmetry and the more abundant gutters on that surface chosen as an- terior. Gutters, variously oriented, are also pre- sent on the equatorial zone and on the periphery of the posterior surface. Estimates of the form of the primary body with any reasonable degree of reliability are not possible. 22. Fig. 3. Owned by Mr. F. Davis, who found it on Avon Location 7501 about 8 km E.N.E. of Wickepin East. The lustre is somewhat dulled by abrasion. The rim is irregular and poorly defined as is also the anterior limit of the equatorial zone. Etched strips on the posterior surface contain short gutters, mostly transverse to the length of the strips. Gutters on the equatorial zone are oriented at right angles to the rim and others outline the bottoms of oval flake scars. The anterior surface has a few gutters, etched schlieren and pits. The transverse profile of the posterior surface is symmetrically wedged as the result of natural flake losses (c. 8 g) and the longitudinal profile is also affected by flake loss. The estimates of the primary body are therefore very approximate. 23. Fig. 3. Geological Survey of W.A. collection 1/5327. Found at Narembeen. This core is closely equidimensional, having a thickness/mean dia- meter ratio of 0.95. The posterior surface has a complex of small gutters in depressed, deeply etched areas and other gutters tend to be cir- cumferential near the rim. The rim is well defined. Minor sculpture of the equatorial zone is principally gutters, variously oriented. The anterior surface is asymmetrically wedged and Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 76 has some small gutters with tendency to circum- ferential orientation. Gutters are more abund- ant on the posterior than on the anterior sur- face, an unusual occurrence. The form of the primary body cannot be determined reliably because of the deep etching of the posterior surface. 24. W.A.M. 10613. The posterior surface and equatorial zone are similar to those of No. 21 above. The anterior surface has abundant gut- ters and a minor artificial flake scar. Most gutters are sharply defined and all are narrow (c. 0.2 mm). The forms of the longitudinal and transverse profiles had to be estimated from sur- viving patches of original surface and the re- sults differed by 2 millimetres. The parent mass might have been a slightly prolate spheroid but has been calculated as a sphere using the mean figure as radius. 25. S.A.M. T427. Initially in the collection of Mr. S. F. C. Cook. Fenner (1934, 1955) made brief references to this specimen from Norseman and illustrated the posterior surface. Little can be added because of the extremely corroded state of the specimen. As is well shown by Fenner (1934 PI. IX E2), the posterior surface lost a flake from the polar area and subsequently de- veloped such a degree of pitting as to approach the appearance of hammered metal. Such sculp- ture with patches of interrupted, etched schlie- ren extends over the entire surface. The rim and the limit of an equatorial zone c. 1 cm wide are still recognisable and within the equatorial zone are modified flake scars. The form of the primary body must needs be judged by that of the annular and much corroded remnant of pos- terior surface. 26. Fig. 3. W.A.M. 12090. Found by Mr. F. Basset about 3 km E. of Brookton in 1961. The form is somewhat irregular as the result of minor flake losses. The posterior surface has a narrow band running obliquely from rim to rim containing short gutters and terminating in deeply pitted areas at each end. The rim and the limit of the equatorial zone c. 1 cm wide are poorly defined. The equatorial zone has gutters oriented normal to the rim and outlining the bottoms of oval flake scars. The anterior surface is distinctly asymmetrical in profile; its central area has a complex of short gutters. Flake losses could not be assessed with confidence but the percentage represented would be small. 27. W.A.S.M. 10199. This australite from an unspecified locality in the Eastern Goldfields of Western Australia has been described by Baker (1967). Additionally, an assessment of the primary body has been attempted. 28. W.A.S.M. 8950. For detailed description and illustrations of this specimen from 42 km E. of Narembeen, see Baker (1961). Because of the deeply corroded condition of this core the estimate of the form of the primary body is very approximate. 29. W.A.S.M. 9421. Found by Mr. J. P. Parker about 11 km S.E. of Salmon Gums in 1962. Des- cribed by Baker (1967). An estimate of the form of the primary body has been attempted. 30. Fig. 3. Owned by Mr. C. B. C. Jones of Hampton Hill Pastoral Station. Found on the station about 10 km W. of the abandoned Kurnalpi townsite. This specimen evidently suffered some wear in transport before enclosure in alluvium because natural flake scars are well rounded yet the entire surface now has a uni- formly bright lustre. There has been no arti- ficial damage. Natural flaking has removed a continuous length of half the rim; the remainder is well defined and regular as is also the limit of the equatorial zone. Both posterior and an- terior surfaces have etched lunate and circular bruise scars. The equatorial zone has a few gutters oriented normal to the rim and etched flake scars. A fair estimate of the primary body is possible; flakes account for only about 1% loss. 31. S.A.M. T509. Formerly in the collection of Mr. S. F. C. Cook. Illustrations of this specimen given by Fenner (1955) include an elevational view showing the sharp, well-preserved rim (Pl.VIII-16) . The surface has a bright lustre. The rim is complete except for a minor natural flake loss. The posterior surface has some short gutters with tapered ends and V-shaped cross sections reminiscent of gash fractures and they tend to radial orientation. The Lake Ballard core (No. 6, Table 2) is the only other of these large Western Australian specimens having such gutters on the posterior surface, but their de- velopment is much further advanced in that example (Cleverly 1971 Fig. 1). The equatorial zone has well-defined gutters of U-shaped sec- tion oriented normal to the rim. The anterior surface is almost free of minor sculpture. 32. W.A.M. G7566. This fragment is included for completeness of record. The entire anterior surface and one end have been removed by arti- ficial fracture. The bluntly wedged remnant has angle c.115° and by analogy with other arti- ficially broken specimens the edge was probably immediately beneath the anterior surface as found. Reconstruction suggests that the mass prior to artificial fracture certainly exceeded 100 g, and possibly attained 115 grams. The dull abraded remnant of posterior surface has small depressed areas containing narrow gutters and small areas of deep pitting extending to the equatorial zone, the remnant of which has some gutters oriented normal to the rim. Round core of unknown provenance Fenner (1955) figured the artificially damaged posterior surface and a side elevation (Pl.VII Figs 5 and 6 respectively) and stated that the specimen had been found in the Western Aus- tralian Goldfields. The owner, Miss K. D. Black- ham of Adelaide, is insistent that neither the finder nor site of find is known to her; the specimen might have come from any of the gold mining localities with which members of the family had early associations, viz. Ballarat (Vic- toria), Teetulpa (South Australia), Coolgardie and Kalgoorlie (Western Australia). The dimensions are 52.7 x (51.3)51 x 44.5 mm, the mass 153.96 g (c.168 g prior to artificiai damage), and the specific gravity 2.399. The specific gravity is distinctly lower than that of Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 77 any other australite of mass exceeding 100 g for which values are available. The figure is much closer to average values in Victoria than those elsewhere (Baker 1969 Fig. 2), but the value for an individual australite has very limited significance. Although 2.399 is almost the mean value for Port Campbell, Victoria specimens, it nevertheless lies within the range shown by australites from the Kalgoorlie district, Western Australia (Chapman et al. 1964 Fig. 7). The general aspect of the specimen is that of a water worn pebble. The entire surface is dulled by abrasion and the bruise scars show various de- grees of freshness or of shallowing by abrasion. Possibly only one episode of transport with intermittent bruising is represented. No clear indication of provenance was recog- nised but an area in which true alluvial gravels existed (i.e. Victoria or possibly Teetulpa) seems rather more likely than aria interior Western Australia. Victoria is the more likely because the known area of occurrence of unusually massive australites (Baker 1969) overlaps the alluvial goldfields; Teetulpa is nearly 300 km distant from the Karoonda-Lowaldie area, from which have come the only two cores of mass exceeding 100 g yet reported from South Australia. Discussion The Lake Buchanan core (No. 20) was the first recorded australite of mass exceeding 100 grams (Simpson 1902). The number of such specimens known grew slowly to 24 (Baker 1972). The present known total is 41, comprising 32 from Western Australia (Table 2), six from Victoria (Baker 1969, 1972), two from South Australia (Fenner 1955) and one of unknown provenance (this paper). Most of the recent increase is accountable to previously unpublished specimens in the W.A. Museum collection and to specimens which are privately owned. An australite said to have been found in the vicinity of Eucla and of mass c.142 g (Fenner 1934 p. 78) has not been included in the total because the report is hearsay only, though the locality would fit well the distribution pattern referred to below. The number of additional specimens in private hands is unknown but probably considerable to judge by the response to some very limited publicity. A brief news item submitted to a Kalgoorlie radio station concerning the dis- covery of the 437 g (No. 1) specimen, had the immediate result that No. 16 was offered for examination; an article on australites published for the writer by the Narrogin Observer brought in response Nos. 4 and 22. The sites of find of the Western Australian specimens (Fig. 1) comprise a south west group and scattered occurrences to the east and north east. The south west group lies entirely within the main wheat belt, where there is a settled population and where the land is periodically seen and shallowly embedded objects are brought to the surface during cultivation. To the im- mediate east of this belt, both north and south of the principal Perth-Kalgoorlie road and rail links, there is no permanent settlement and a corresponding gap in the distribution pattern. Further east again is the narrow strip of country containing a few widely spaced centres of population associated with mining in the north and agriculture in the south (Salmon Gums). The two most northerly australites were chance discoveries of a mineral exploration party outside the active mining area and of an exploration expedition in country which is still on the extreme fringe of extensive-type pastoral development more than seventy years later. There is thus such a close positive correlation between the sites of find and the distribution of human activities that it is tempting to dismiss the first distribution as a direct result of the second. However, there is good evidence, at least for the Eastern Goldfields region, that australites of mass exceeding 100 g are extremely rare. The most important evidence is the private collection of Mr. D. L. Tillot on containing nearly 11 000 located australites gathered from a broad belt extending up to 200 km north and south from Kalgoorlie; the most massive speci- men in this collection is of only 58.1 grams. The C. B. C. Jones family collection has been gathered from the country immediately E.N.E. from Kalgoorlie and is numerically at least equal to that of Mr. Tillotson; the most massive specimen is of 101.12 grams (No. 30 this paper). The Cook collection of more than 5 000 speci- mens, now in the South Australian Museum (Fenner 1949), was gathered from a more ex- tensive area also centred upon Kalgoorlie; only two specimens (Nos. 25, 31) qualified for in- clusion in this paper. Smaller official and pri- vate collections could also be cited as well as the thousands of australites which pass through the hands of commercial lapidaries. The most massive of about 5 000 specimens handled by one Kalgoorlie lapidary in recent years has mass 59.1 grams. This general region has twice been intensively prospected, earlier for gold and latterly for base metals. Inclusive of the numer- ous small private collections resulting from these activities it is likely that 40 000 australites have been recovered from the Eastern Goldfields, yet only six are known of mass exceeding 100 grams. Further to the north, Earaheedy and other pastoral stations to the west and south-west of Lake Buchanan are represented in available collections by more than 1 300 specimens of which the most massive is a 74.2 g specimen from Wongawol (No. 22 in E. S. Simp on col- lection held at W.A. Museum). The Western Australian wheat belt extends a further 400 km north-westerly from the aus- tralite occurrences shown in Fig. IB. presum- ably with equally good opportunities for obser- vation of australites, yet no specimen of mass exceeding 1G0 grams is known to have been recovered there. The centering of the infall of unusually massive austrafitss in the south west of Western Australia thus appears to be a reality, but the area of their occurrence is almost certainly larger than is shown in Fig. IB, the eastern boundary having resulted from the circumstances of collection. The coastal strip flanking the wheat belt to the west and south was omitted from considera- tion above. It has relatively high- rainfall and Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 78 much of it is heavily timbered. Though it has thd highest density of human population for any part of the State, it is poorly represented in australite collections. This paucity of known australites is likely to be, at least partially, the consequence of an adequate drainage and vege- tation cover when compared with the extensive cultivated areas and bare, semi-arid terrain further inland. An almost cylindrical australite fragment (W.A.M. 13202) of mass 42 g from the Albany district (Fig. 1A) is the central portion of a stout-' waisted dumbbell; the width and thick- ness at the waist are 29.2 mm and 27.2 mm. The analogous dimensions of the 151 g Ongerup dumbbell are 33.2 mm and 28.8 mm (Baker 1967). It is thus likely that the fragment rep- resents an australite of which the mass attained 100 g, so that the area of infall of unusually massive specimens should be extended to the south coast. Knowledge of australite distribution is very imperfect and the number of specimens con- cerned in the present paper is very small. Extreme caution is therefore necessary before acknowledging the existence of a pattern in the distribution and the following is offered with some diffidence. The six Western Australian australites of mass exceeding 200 g were found in a belt extending S.S.W. from Lake Ballard, i.e., in a direction approximately at right angles to the northern boundary of the australite strewnfield (Baker 1969b Fig. 1). When the distribution of the most massive specimens of other main- land states is also considered (Baker op. cit.) the sites are seen to occur in areas distant from that boundary, the south western portion of the continent being the most distant and extensive and containing the sites of 17 of the 20 most massive australites known. These observa- tions suggest the possibility of a mass grading away from the strewnfield boundary. If the direction of flight is to be related to such a grading, it is at variance with the conclusions of Baker (1969a), McColl and Williams (1970) and Chapman (1971). It would be of interest to know the distribution of a less massive cate- gory, say 50-100 g, and whether there are large overlaps in the distribution of successive cate- gories as would suggest the entry into the atmosphere of a range of masses either con- tinuously or at various points along a flight path. The greater numbers in the less massive categories should provide more reliable data. The mean specific gravity of all 32 speci- mens under consideration has little significance because of the wide area of occurrence repre- sented, but the more circumscribed south west group of 25 specimens warrants brief considera- tion. With the exception of Nos 2, 13 and 18, the low specific gravities of which are prob- ably attributable to bubble cavities, specimens of this group have specific gravities in the narrow range 2.420-2.439, a variation of less than 1%. This degree of constancy might sug- gest a distinct population within the australite shower but another explanation is more probable. The heterogeneity arising from irregular distri- bution of bubble cavities of various sizes in australite glass may be expected to be evident in groups of small australites as a considerable variation in their specific gravities. For groups of increasing size from any one area, the amount of variation arising from this cause may be expected to decrease until the most massive specimens approximate to the bulk specific gravity of the material and only occasional speci- mens show significant departures from the mean value, i.e. the larger samples of a heterogeneous material are more likely to be truly representa- tive. The constancy of values could thus arise from the large sizes of the specimens. The weighted mean specific gravity of the 25 specimens in the south west group of total mass 4.2 kg is 2.427. This is a lower mean value than for other measured groups in Western Australia (Baker 1969b) but the group is located further south than the others. Because these large cores have lost stress shells and are generally quite eroded, they retain virtually no surface glass from which constituents could have been volati- lized during atmospheric flight as occurs from flange glass. The low mean specific gravity cannot therefore be attributed to that cause. Globular, more or less equidimensional cores are well represented in the south west group. They reach their extreme development in Nos. 3 and 9 which lack a rim. Other specimens (Nos. 8, 16, 23), though rimmed, have even higher thickness/diameter ratios and calculated losses are very low. After allowing for a stress shell, thickness losses of 4-9 mm leave little or nothing for ablation losses. It has to be con- ceded as a possibility that these cores, all of which are round or only slightly oval, developed from prolate spheroids which travelled with the long axes parallel to the flight paths. In cal- culating the primary bodies as spheres their primary masses would be underestimated and likewise the losses from them. No detailed analysis of the loss figures has been attempted because the number of reliable estimates available for any one shape type is so small. The loss figures are generally low. The average volume loss is only 46% compared with 56.5% for 23 perfectly preserved buttons from Port Campbell, Victoria (Baker 1962), though the loss of a stress shell did not contribute at the latter locality and terrestrial losses were in- significant. In a very general way, the better preserved cores have the higher loss figures. There may therefore be errors not only in the estimated dimensions of the primary bodies but also in the shape types ascribed to them. Acknowledgments . — For the loan of australites for examination I thank the Trustees of the Western Australian Museum and Dr. C. Pearson, the Trustees of the South Australian Museum and Miss J. Scrymgour, Mr. F. L. Sutherland and Miss J. Hingley of the Aus- tralian Museum, Sydney, The Director of the Geological Survey of W.A. (Mr. J. H. Lord), The Director of the W.A. Government Chemical Laboratories (Dr. L. W. Samuel), Miss K. D. Blackham of Adelaide, Mrs. H. Biggin of Kondinin, Mr. P. Repacholi of Kondinin, Mr. R. Pugh (Headmaster Coolgardie), Mr. F. Davis of Kulin East, Mr. C. B. C. Jones of Hampton Hill Pastoral Station and Mr. R. Kirkpatrick of Mandurah. Dr. George Baker of Melbourne kindly read the first typescript of this paper and it has been distinctly improved by the amendments which he suggested. Journal of the Royal Society of Western Australia^ Vol. 57, Part 3, October, 1974. 79 References Baker, G. (1956).— Nirranda strewnfield australites, south-east of Warrnambool, Western Vic- toria. Mem. Nat. Mus. Viet. 22: 1-172. (1957). — The role of australites in abori- ginal customs. Mem. Nat. Mus. Viet. 22: 1-26. (1959).—' Tektites. Mem. Nat. Mus. Viet. 23: 5-313. (1961). — A naturally etched australite from Narembeen, Western Australia. J. Roy. Soc. West. Aust. 44: 65-68. (1962). — The largest known australite and three smaller specimens from Warralakin, Western Australia. J. Roy. Soc. West. Aust. 45: 12-17. (1963). — Round australite core from Grab- all, Western Australia. J. Roy. Soc. West. Aust. 46: 57-62. (1966).— The largest known dumbbell shaped australite. J. Roy. Soc. West. Aust. 49: 59-63. (1967). — A second large dumbbell shaped australite, Ongerup, Western Australia, with notes on two other large australites. J . Roy. Soc. West. Aust. 50: 113-120. (1969a). — Five large australites from Vic- toria Australia, and their relationships to other large forms. Mem. Nat. Mus. Viet. 29: 53-64. (1969b). — Australites from Mulka, Lake Eyre region, South Australia. Mem. Nat. Mus. Viet. 29: 65-79. (1972) —Largest australite from Victoria, Australia. Mem. Nat. Mus. Viet. 33: 125-130. Bowley, H. (1945). — Australite observed to fall at Cot- tesloe — a correction. J. Roy. Soc. West. Aust. 29: 163. Chapman, D. R. (1964).— On the unity and origin of the Australasian tektites. Geochim. et Cos- mochim. Acta 28: 841-880. (1971).— Australasian tektite geographic pattern, crater and ray of origin, and theory of tektite events. J. Geophys. Res. 76: 6309-6338. Chapman, D. R., Larson, H. K. and Scheiber, L. G. (1964).— Population polygon of tektite specific gravity for various localities in Australasia. Geochim. et Cosmochim. Acta 28: 821-839. Cleverly, W. H. (1971).— An oval australite core frem Lake Ballard, Western Australia. J. Roy. Soc. West. Aust. 54: 14-16. Fenner. C. (1934).— Australites, Part I. Classification of the W. H. C. Shaw Collection. Trans. Roy. Soc. S. Aust. 58: 62-79. (1940).— Australites, Part IV. The John Kennett collection with notes on Darwin glass and bediasites. Trans. Roy. Soc. S. Aust. 64: 305-324. (1949).— Australites, Part V. Tektites in the South Australian Museum, with some notes on theories of origin. Trans. Roy. Soc. S. Aust. 73: 7-21. (1955). — Australites, Part VI. Some notes on unusually large australites. Trans. Roy. Soc. S. Aust. 78: 88-91. McCall, G. J. H. (1965).— The heaviest recorded aus- tralite. Aust. J. Sci. 27: 267. McColl, D. H. and Williams, G. E. ( 1970) .—Australite distribution pattern in southern central Australia. Nature 226: 154-155. Simpson, E. S. (1902).— Notes from the Departmental Laboratory. Geol. Surv. West. Aust. Bull. 6. (1939). — A second australite observed to fall in Western Australia. J. Roy. Soc. West. Aust. 25: 99-100. Thorp, C. G. (1914).— A contribution to the study of australites. J. Nat. Hist. Sci. Soc. West. Aust. 5: 20-43 and Pis. XVIII-XXIII. Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 80 11. — Annual Medicago species with particular reference to those occurring in Western Australia by J. A. McComb 1 Manuscript received 23 April 1974 ; accepted 18 June 1974 Abstract Thirteen species of annual Medicago are recorded from Australia, including Western Australia — M. arabica, M. intertexta, M. laci- niata, M. littoralis, M. lupulina, M. minima, M. orbicularis, M. polymorpha, M. praecox, M. rugosa, M. scuteliata , M. tornata, and M. trun- catula. Ten species are naturalized, and from four species 10 cu.tivars have been developed. A summary of the main generic features is given, together with keys and descriptions for species identification. Introduction The annual species of Medicago, like many other inbreeding plants, may be very variable within species and so have many infraspecific taxa, but show parallel variation and marked similarities between species. These features have resulted in a confused taxonomy. The most recent comprehensive taxonomic work on annual Medicago is by Heyn (1963, 1970) who showed that well over one hundred specific epithets have been applied to the 28 annual species she includes. This paper is not a taxo- nomic revision, and Heyn’s taxonomic treatment is applied to the Australian material. Further information on taxonomy and species descrip- tions is given by Urban (1873, 1877), Sinskaya (1950), Negre (1956, 1959), Bolton (1962), Willis (1972), Lesins and Gillies (1973), and Quinlivan et al. (1974). Descriptions of cultivars registered in Australia and of some naturalized species are available in Barnard (1969, 1972). Medics have been introduced to Australia from the Mediterranean, and the forms found here do not always show the same range of charac- teristics described for the species as a whole, as in Heyn (1963). This paper describes the species found in Australia and provides keys for identification of either flowering plants or burrs alone. The descriptions are based on accessions from various parts of Australia, maintained in a living collection of some 650 accessions of annual medics at the University of Western Australia. The collection also in- cludes examples of all annual species not naturalized in Australia (McComb 1971). For all species some lines have been validated either by C. Heyn (Hebrew University) , or by K. Lesins (University of Alberta). Extensive field collec- tions have been made in Western Australia and herbarium material examined at UWA and PERTH. The synonymy found in Australian lit- erature is listed. Only a summary of the distri- bution in Western Australia is given; similar 1 Agronomy Department. University of Western Austra- lia, Nedlands, Western Australia 6009. information for other Australian states may be found in Amor (1966), Andrew and Hely (1960), and Barnard (1969). In Australia the common name ‘medic’ (earlier medick) is restricted to the annual species while the perennials are called ‘lucerne’ or ‘alfalfa’. In other countries the common names medic, lucerne, or alfalfa may be applied to either perennial or annual species (Bolton 1962). The annual medics are sometimes called ‘yellow trefoils’ but the name ‘trefoil’ is also used for Trifolium and Lotus species, so the term ‘medic’ is preferred. Botanical description of the annual species Habit and vegetative parts Annual medics germinating in late autumn, form a rosette of leaves before producing branches from the crown. The branches, which do not root at the nodes, may be prostrate and up to 1 metre long, or semi-erect and form a diffuse plant up to 40 cm high. Growth forms and branch length depend largely on environ- mental conditions and whether the plants are isolated or growing in a sward. Such features are therefore variable and not particularly use- ful in species descriptions. The outline of the stems in transverse section, as used by Negre (1956) was found to be similarly variable and has not been included. The leaves are trifoliolate and differ from those of clovers (other than hop clovers) in that the stalk of the central leaflet is longer than that of the laterals (i.e. leaves are pinnate rather than digitate). Leaves of seedlings of M. arabica and M. lupulina may have digitate leaves but all later leaves are pinnate. Leaflet shape is variable, even on the same plant; lower leaves are usually cuneate or obovate and those produced higher on the branches more oval. Leaf morphology and measurements given in the species descriptions are based on the central leaflets of the largest mature leaves. Stipules are only very shortly adnate to the petiole as compared with species of related genera Trigonella, Melilotus and Trifolium, and the stipule margin is serrate, incised, or lacini- ate, rarely entire. Vegetative parts may be glabrous (but provided with closely appressed, small glandular hairs with short stalks which are rarely noticed unless an epidermal peel is examined under the microscope) or hairy, with simple or articulated, glandular or non- glandular hairs. Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 81 Flowers The flowers may be solitary, or number to 10, while one species (M. lupulina) has up to 50 flowers per raceme. Flowers are shortly pedi- cellate, bracteate, in loose racemes on an axil- lary peduncle which is most species is produced into an awn beyond the terminal flower. The relative length of the peduncle when it bears open flowers, and the petiole of the subtending leaf, is used as a taxonomic character. The calyx is tubular or campanulate with five equal, or subequal lobes. Petal morphology is shown in Fig. 1. The corolla is always yellow, but there are usually purple -brown stripes in the centre of the standard. The petals occa- sionally have a purple tinge at the tips in young buds, or in herbarium material. ‘Flower length’ in the species descriptions is the length of the standard which, ranging from 2-10 mm, is longer than the wing and keel petals in all species. The relative lengths of wing and keel petals is used as a taxonomic character (c/. Figs. 11A and 16L). Figure 1. — Petal morphology in Medicago truncatula. A: Standard petal. B: edge view and C: inside view of wing petals showing the horn-like protrusion (h). D: side view and E: inside view of keel petals showing the fused edges (f) and the groove (g) into which the horn on the wing is inserted. Nine of the ten anther filaments are fused into a sheath and one is free. Baum (19681 found that anther morphology may be used to separate the closely related genera Medicago and Trigonella. In Medicago, when the staminal sheath is opened out, it can be seen that the filaments arise from an arched apex, the bases of alternate filaments are swollen, and in the fused sheath there is some thickening along the vascular bundles of all, or at least the central three stamens. Trigonella shows two types of androecium, a ‘simple’ type in which the fila- ments emerge from a flat apex and no filaments are swollen at the base; and a ‘medicagoid’ type in which the filaments emerge from an arched apex, and filaments on either side of the cen- tral one have swollen bases. In neither the simple nor the medicagoid type is there any thickening around the vascular bundles. Obser- vations on the medics which occur in Australia show that the anthers arise from an arched apex and that alternate bases are swollen, but these characters are much more strongly developed in some species (e.g. M. truncatula and M. scutellata ) than in others (e.g. M. polymorpha and M. minima) . In some form^ of M. lupulina (which may also lack the ‘tripping’ mechanism — see below), the features are particularly poorly developed (Figure 2). The thickening Figure 2.— Apical part of the 9 fused anther filaments in Medicago species. A: M. truncatula. B: M. scutel- lata. C: M. polymorpha. D: M. minima. E: M. lupu- lina. F: T. S. of fused filaments of M . truncatula and G: M. lupulina. (camera lucida drawings) along the vascular bundles proved difficult to see and it is possible that the ‘thickening’ re- corded by Baum using anther sheaths softened in lacto-phenol and flattened, results from the width of the sheath as seen in T.S. (Fig. 2). The linear ovary has 1-28 ovules. The style which is subulate or filiform and has an oblique terminal stigma, may persist on the fruit but is never hardened into a rostrum as in Trigo- nella. The length of the floral bracts relative to the pedicel length and the calyx, as used by Negre (1956) was found to be too inconsistent, both within species and between flowers on the same plant, to be of taxonomic use. Tripping mechanism The flowers of the annuals are self fertile (with the possible exception of some M. inter - texta (Lesins and Gillies 1973) ), and largely self -pollinating. They lack nectaries (Heyn 1963) but are occasionally visited by insects. The perennial species have an elaborate ‘trip- ping’ mechanism for pollination which is an adaptation to cross fertilization by insects, and this tripping mechanism is retained in the annuals even though they are self pollinating. The mechanism is not found in the related genera Trifolium, Trigonella or Meliloius. In the freshly-opened flower the anthers have dehisced but the pollen is contained by the surrounding keel petals. The style, surrounded by the anther filament sheath, is retained be- tween the keel petals which are fused along their outer edges as well as for some distance along the edges facing the standard, but are free at the tips. The cohesion between keel petals results from interlocking projections on the epidermis (Larkin and Graumann 1954). The strength of the cohesion partly determines the ease with which the flower can be caused to trip. If the keel is touched so that the petals are split apart, the style and anthers are flung forward on to the standard petal. The wing- petals which have horn-like protrusions appressed to grooves in the keel (Fig. 1), may be removed carefully without causing tripping, but usually movements of the wings are transmitted to the Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October. 1974. 82 keel and the style trips (Heszky 1972.). As the style trips a membrane is torn from the stig- matic surface making it receptive, and the pollen is pressed into the stigma (Armstrong and White 1935). The force with which the sexual column snaps over is determined by the 'truc- ture of the region where the base cf the fused anther filaments joins with the base of the keel petals (Larkin and Graumann 1954). When the flower has been tripped the anther filaments are curved and cannot be straightened without tearing the tissue. Ncn -tripped flowers do not set fruit. In the annuals self-tripping and con- sequently self-pollinaticn are most common, although insects sometimes cause tripping. In countries where the species are native a low frequency of cross-pollination occurs and some species hybrids are found (Heyn 1963). In one species only, M. lupulina, are there some forms which lack the explo ive tripping mechanism (Heyn 1963). Fruit The calyx is persistent, and in contrast to most clovers the corolla is deciduous. The de- veloping coiled fruit may be contracted and concealed within the calyx (Fig. 16M) ; con- tracted but protruded sideways from the calyx (Fig. 11B); or it may hang from the calyx in a loose spiral and later become compact (Figs. 5B and 10A). These development types have been used for subgeneric divi ions. One of the characteristics given by Heyn for the generic sub-sections Rotatae and Pachyspirae of the Section Spirocarpos ‘young pods contracted and concealed within the calyx’ was found to be invalid. In M. scutellata and M. rugosa the young pods protrude sideways from the calyx while in M. littoralis, M. truncatula and M. tornata coils could be concealed or protrude sideways in flowers on the same plant. The mature fruit always exceeds the calyx in length and the number of pod coils ranges from 1-11. Fruits may be spiny or smooth and are indehiscent, but may gape open where the seeds are positioned. Much taxonomic weight is placed on the burr morphology and the arrangement of the radial veins on the surface of the burr, and the presence or absence of a submarginal or lateral vein along the dorsal suture (see Figs. 10B and 11F). This venation is sometimes ob- scure, but is best seen on the surface of a middle coil of a full size, but not fully hardened burr. Coiling direction, as determined by the direction of a line running from calyx to style, when the burr is held with the calyx end towards the observer, may be clockwise or anticlockwise. It may be typical of a species or both types may occur in the same species. In the M. trun- catula, M. littoralis and M. tornata species group, coiling direction is controlled by a single gene, clockwise coiling being determined by the dominant allele (Simon 1965; Simon and Mil- lington 1965). The dimensions of the burrs given in the species description are from dry mature burrs, which are smaller than freshly-picked green burrs. Measurements of diameter and height are of the coils without the spines. Seeds M. lupulina has only one seed, but other spe- cies have up to 28. Seeds are yellow or brown (black in M. intertexta) , oval, kidney or horse- shoe shaped, and may be separated in the pod by false parenchymatous partitions. They are commonly impermeable, i.e. hard, when fully mature and dry. The seeds have strophioles (c.f. Hutchinson 1964) and it is usually in this region that cracks form and render the seed soft (Fig. 16K). Seed size usually decreases from the calyx to stylar end with often a small mis- shapen seed at the stylar end. Seed dimensions are based on average sized seeds. Seedlings The expanded cotyledons are oval or oblong- linear, the blade merging gradually into the stalk, in comparison with TrigoneV.a, Meliloths and Trifolium, which have a distinct node between petiole and lamina (Fig. 3). This cha- racter is quite distinct, even though there is no difference in venation between the two types (Heyn 1968). However, I have been unable to confirm the observation that the base of the Figure 3. — Cotyledon morphology. A: Medicago scutel- lata. B: M. polymorpha. C : Trigonella coerulescens, D: T. cylindracea. E: Melilotus italica. F: M. indica. G: Trifolium clypeatum. H: T. subterraneum. Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 83 petiole is “swollen” in Trigonella and Melilotus and not in Medicago (Urban 1873, quoted in Baum 1968; figured in Heyn 1963 and 1970). Species from these genera which I have grown (36 Medicago ; 9 Melilotus ; 11 Trigonella; 11 Trifolium species) show no basal swellings, and at most a slight bulge due to the expansion of the plumule (Fig. 3). In Trifolium the bases of the petioles may be fused and bulge when they enclose the developing plumule, but the thickness of the stalk is not increased at the base. It suffices therefore to distinguish only between a gradual merging between lamina and petiole, and an articulation between the two. The seedlings produce a single unifoliolate leaf, then trlfoliolate leaves. Key to species based on mature burrs (Species descriptions are listed alphabetically) 1. Burrs spineless or with small tub- ercles 2 1. Burrs spiny 2. Burrs single-seeded, black, with distal end coiled . .... 2. Burrs many seeded (rarely only one seed), coiled along full length 3. Coils concave lower coils enclosing upper ones, burr cup- or olive- shaped .... .... .... 9 M. lupulina 3 M. scutellata 3. Coils not as above 4 4. No membranous partition between seeds, burrs hairy 4. With membranous partition be- tween seeds, burrs glabrous or hairy 5. Burr diameter 6-9 mm, central coils lozenge-shaped with raised veins on swollen outer margins, no lateral veins 5. Burr diameter 3-4£ mm, coils lenti- cular or discoid, lateral veins well separated from dorsal suture 6 . No lateral veins, burr diameter 13- 17 mm, margin of coils thin and membranous, seed coat rugose 5 6 M. rugosa M. minima var. brevispina M . orbicularis 6 . Lateral veins present, burr dia- meter less than 10 mm, edge of coil not thin and membranous, seed coat smooth .... ... ... 7 7. No indentations between lateral vein and dorsal suture 7. With indentations between lateral vein and dorsal suture between bases of vestigial spines 8 . Lateral veins close to dorsal suture, groove narrow and deep, 15-20 strongly curved radial veins, burr 2£-8 mm high ... 8 . Groove wider and shallow (in part obliterated in very thickened burrs), about 10 radial veins almost straight, burr 6 £-ll mm high M. tornata var. tornata cv. Tornafield and cv. Murrayland 8 M. polmorpha var. brevis- pina M. truncatula cv Cyfield and naturalized material 9. Burr coiling clockwise 9. Burr coiling anticlockwise 10. Burrs without partitions between seeds 10. Burrs with partitions between seeds 11. Seeds black, burrs 10-14 mm dia- meter, long entangled spines like a sea urchin 11. Seeds brown or yellow, burrs less than 7 mm diameter .... M. truncatula CVS. Cyprus and Hannaford and naturalized material 10 11 13 M. intertexta var. inter- texta 12 12. Burr 5-6 mm wide x 5-7 mm high, broad dorsal suture, radial veins entering lateral veins and a deep groove between lateral veins and dorsal suture .... . .. M. laciniata var. laciniata and var. brachy- acantha 12. Burr 3-5 mm wide x 2£-5 mm high dorsal suture narrow, radial veins enter lateral veins which are separ- ated from dorsal suture by a broad margin (± £ width of coil) across which run the veins to the spines M. minima var. minima 13. Coil edge with 3 grooves and 4 ridges (2 lateral furrows and a central furrow bisecting the dorsal suture); radial veins ± 8, radicle more than \ as long as cotyledons M. arabica 13. Not as above 14 14. Coils strongly appressed when mature .... .... 15 14. Coils not strongly appressed .... 17 15. Lateral veins in the same plane as the dorsal suture on the edge of the coil, no indentations between lateral veins and dorsal suture, burrs glabrous at maturity, spines long (up to 4 mm) inserted at 130° in middle coils, and 90° at apex. M. littoralis var. littoralis 15. Lateral veins on the surface of coil, indentations between lateral veins and dorsal suture between spines bases (especially when young, and seen in all but the most thickened burrs at maturity) .... 16 16. Burrs 4^-10 mm high, 3-7 coils, spines 1-3 mm long, inserted at 120° -90°, burrs with scattered hairs. (pubescent when young) M. truncatula cv. Jemalong, cv. Borung cv. Ghor and naturalized material 16. Burrs, 2£-4 mm high, 3-4 coils, spines 1 mm or less, inserted at 90° to coil surface, burrs glabrous at maturity (sparsely hairy when young) M. littoralis cv. Harbinger 17. Dorsal suture region wide (± 1 mm), adjacent coils not in contact, wide spaces between coils, 10-15 spines per coil, a deep groove between dorsal suture and lateral veins which is seen from the surface rather than the edge of the coil .... M. praecox 17. Dorsal suture region thin (± J mm), coils in loose contact, 15-20 spines per coil, groove between dor- sal suture and lateral vein clearly seen on viewing coil edge-on .... M. polymorpha var. .• vulgaris Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 84 Key to species based on flowering plants Species descriptions without mature burrs 1. Flowers 10-50, peduncle more than 2x the petiole of subtending leaf 1. Flowers 1-10 (rarely — 15 in M. tornata) peduncle at most 2x the petiole of subtending leaf 2. Upper surface of leaves glabrous (check on young leaves) lower surface glabrous-densely hairy .... 2. Upper surface of leaves densely hairy, or with at least sparse hairs when young, lower surface moder- ately-densely hairy 3. Dense glandular and non-glandu- lar hairs on lower side of leaf and elsewhere 3. Not as above 4. Standard 6-7 mm long 1-2(3) flowers, upper side of stipules glabrous or with rare glandular hairs 4. Standard 6-7 mm long 1-3(3) flowers, upper side of stipules with dense glandular hairs .... 5. Young fruit in long spiral side- ways from calyx contracting later, leaves with purple shield-shaped mark M . lupulina 3 9 4 5 M. rugosa M. scutellata M. intertexta 5. Young fruit in contracted spiral sideways from calyx; no leaf mark, or different to the above 6 6. Leaf mark an upper central blotch or a very wide shield and occupy- ing the basal § of the leaf, arti- ticulated hairs on vegetative parts M. arabica 6. Leaf mark none or not as above, simple hairs on vegetative parts 7 7. Leaves oblanceolate, possibly some leaves laciniate .... .... .... .... M. laciniata 7. Leaves cuneate or obovate, margin entire or dentate .... .... .... 3 8. Leaves 8-27 mm long, 7-19 mm wide standard 3-6 mm long, wings longer than keel 8. Leaves 7-9 mm long, 5-7 mm wide, standard 2-3 mm long, wings shorter than keel 9. Stipules entire or weakly dentate, both surfaces hairy 9. Stipules dentate or incised, upper surface glabrous .. 10. Young fruit protruding sideways from calyx in long spiral, con- tracting later 10. Young fruit contracted, usually concealed in the calyx but some- times protruding sideways 11. Flowers 7-10 (15), peduncle longer than petiole (up to 2x as long) .... 11. Flowers 5 or less, peduncle, shorter or longer than petiole 12. Peduncle shorter or longer than petiole, 3-5 flowers 12. Peduncle shorter than petiole, 1-2 (5) flowers 13. Calyx lobes densely hairy, calyx teeth reaching % or more the stand- ard petal length, young pods pub- escent 13. Calyx lobes moderately hairy, teeth at most l length of standard petal, young pods sparsely hairy M. polymorpha M. praecox M. minima 19 M. orbicularis 11 M . tornata 12 M. littoraiis cv. Har- binger* 13 M. truncatula * M. littoraiis* * Mature fruits may be necessary to distinguish these species. Medicago arabica (L.) Huds., FI. Angl. 288 (1762) Synonymy : M. maculata Willd., Sp. PI. 3 : 1412 (1802). Australian representation : The variability ob- served in Australia covers almost the whole spe- cies except that plants with spineless burrs, or without a purple leaf mark have not been reported. Description: Stems, peduncles and petioles sparsely or moderately hairy, mainly articulated hairs on the stems and petioles, both simple and articulated hairs on the peduncles. Central leaflets usually wider than long, 10-25 mm long X 10-28 mm wide, cuneate or obcordate, upper leaves more obovate, apex retuse or emar- ginate, upper surface glabrous with an upper- central purple mark ranging from a small dot to a large inverted V, or a wide basal shield occupying the basal 2/3 of the leaf, no other purple flecks but sometimes small white flecks (the purple blotch may fade in herbarium spe- cimens), lower surface sparsely or densely hairy, Figure 4. — Medicago arabica. A and B: leaf types. C: burr. D: diagramatic T.S. of coil edge showing relative position of veins and grooves. E: seed. simple hairs on the lamina and a few articu- lated ones on the mid rib. Seedling leaves some- times digitate, older leaves clearly pinnate. In more luxuriant specimens the number of leaf- lets may occasionally be 4 or 5. Petioles may be very long (up to 15 cm). Stipules with small sharp teeth, sparse or dense articulated hairs on lower side only. Peduncles much shorter than the subtending petioles (peduncles may be only 1/3 or 1/5 of the petiole) with a slender awn ± 3 mm long with simple and articulated hairs. Flowers 2-4(5) per raceme, calyx with mo- derately dense simple hairs, teeth equal to tube; standard 4-5 mm long, keel ± 1 mm longer than Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 85 the wings. Developing pod contracted and pro- truding sideways from the calyx, coiling anti- clockwise. Mature fruits 1-4 per raceme, barrel- shaped, top coils markedly concave, basal coils less so, 4-6 soft coils not firmly appressed, (4) 5-7 mm diameter, 5-7 mm high, with soft inter- meshed spines up to 4 mm long which have a clear basal groove, inserted at 90-120° to the coil surface, tips straight or slightly hooked, pod surface glabrous. Few radial veins, veins dis- tinct in the centre but anastomosed into a fine net towards the outer edge of the coil, lateral veins in the same plane or protruding slightly beyond the dorsal suture, a deep groove between each lateral vein and the dorsal suture and a shallow one bissecting the dorsal suture. ( A coil viewed from the edge has 4 ridges and 3 grooves). Seeds, 5-8, separated by thin mem- branes, subreniform, yellow to light brown, 2i-3 mm long, ± 1£ mm wide, radicle longer than half the length of the cotyledons, extreme tip of radicle may protrude vertically, hilum small. 2n 16. Common name: “Spotted medic” or “spotted burr medic”. These names refer to the pro- minent leaf blotch, and the resemblance of this species to burr medic, M. polymorpha. Cultivars: None. Distribution : Naturalized but uncommon on fer- tile soils in the lower south west of Western Australia and around Perth. Medicago intertexta (L.) Mill., Gard. Diet, ed. 8, No. 4 (1768) Synonymy: M. ciliaris Willd., Sp. PI. 3 : 1411 (1802) V-V- M. ciliaris All., FI. Pedem. 1 : 315 (1785). M. echinus DC., FI. Fr. 4 : 546 (1805). Australian representation: The description is of naturalized material of M. intertexta var. inter- texta. All naturalized material appears to have the basal leaf mark but this is not a varietal characteristic. Black (1957) records M. ciliaris Willd. (= M. intertexta var. ciliaris (L.) Heyn), frem South Australia but I have been unable to confirm the occurrence of this variety other than in collections of imported material at agri- cultural research stations (Hj. Eichler, pers. comm.). Variety decandollei is endemic to Si- cily and not naturalized in Australia. The va- lidity of Heyn’s (1963) taxonomic treatment might be questioned as in crosses between var. intertexta and var. ciliaris Lesins et al. (1968) showed that leaf markings and pod hairs were probably under the control of single genes but hybrid fertility was only 40%. Lesins considers the two taxa to be separate species. Description: Stems and petioles glabrous, pe- duncles with sparse simple hairs. Central leaf- lets 15-20 mm long x 8-15 mm wide, obovate or oblanceolate, apex apiculate, edge serrate almost to base, upper surface glabrous, lower surface sparsely hairy. Upper surface with a purple, basal shield-shaped mark, no other leaf flecks. This mark may fade in older plants and is not always retained in herbarium material. Stipules deeply dentate, sparse hairs on lower side or confined to margin. Peduncle variable length, Figure 5. — Medicago intertexta var. intertexta. A: leaf. B: young developing burr. C: burr. D: seed. often longer than subtending petiole in flower, equal in fruit, awn from very short to 5 mm, glabrous, often easily detached. Flowers 3-7 calyx with a few hairs at base of teeth, teeth equal to tube or longer. Standard 7-10 mm, keel and wings almost equal or wings longer. Develop- ing pod initially not contracted lut protruding sideways from the calyx in a long spiral coiling anticlockwise. Mature fruits 1-2 per raceme, spherical to ovoid, 10-14 mm diameter, 11-16 mm high, 8-11 coils loosely appressed, glabrous, with long intermeshed spines up to 6 mm long in- serted at 90°-100° to surface of coil, tips not hooked, clear basal groove. No lateral veins , radial veins anastomosed into a net and one vein running from this net into the base of each spine. The dorsal suture has a central groove. Seeds 5-11, no partitions between the seeds, sub- reniform black, 3-5£ mm long x 2-3 mm broad, radicle less than i the length of the cotyledons, tip clear but not protruding, hilum and stro- phiole region brown. 2n 16. Common name : “Calvary medic” — relating to the old myth that the blood red leaf marks resulted from blood from the crucifixion; hedgehog medic — relating to the very spiny burrs. Cultivars: None. Distribution: Only two specimens have been col- lected from Western Australia: from Busselton (1928) and Brunswick (1916). As there are no recent collections, it is doubtful if the species has persisted. Medicago laciniata (L.) Mill., Gard. Diet, ed. 8, No. 5 (1768) Australian representation: Varieties laciniata and brachyacantha Boiss. both occur in Aus- tralia. Description: Stems, petioles and peduncles with sparse simple hairs, central leaflets 8-11 mm long x 4-6 mm wide, oblanceolate, apex trun- cate, retuse and apiculate or tridentate, leaf margins serrate or irregularly laciniate, upper surface glabrous, lower surface with sparse or Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 86 dense simple hairs. Leaves may show purple and white flecking. Stipules deeply incised or laciniate, simple hairs on lower side only. Pe- duncle longer or shorter than subtending pe- tiole, extended into an awn 1-3 mm long gla- brous or with a few hairs. Flowers 1-2(3), calyx sparse to moderately hairy, teeth shorter than the tube. Standard 44-6 mm long, keel slightly longer than the wings. Developing pod contracted but protruding sideways from the calyx coiling anticlockwise. Mature fruits 1(2) per raceme, olive-shaped, globular or cylindrical, 5-6 mm diameter, 5-7 mm high, burrs soft at maturity, 5-8 coils not strongly appressed, spines 2-44 mm long,* inserted at 90°-180° to surface of coils, tips hooked, distinct basal grove, pod surface glabrous or with a few simple hairs. Figure 6. — Medicago laciniata. A-C: var. laciniata. A: vegetative morphology. B: burr. C: seed. D-E: var. brachyacantha. D: vegetative morphology. E: burr. Radial veins S-shaped and not anastomosed until they enter lateral veins which are se- parated from the strongly developed dorsal suture by a deep groove. Dorsal suture region glabrous and in fresh material often glaucous. Seeds (7)9-12(14) not separated by partitions, oval-slightly reniform yellow to brownish yel- low, 2-3 mm long, 1-1.5 mm wide, radicle longer than half the cotyledons, radicle tip clear but very rarely curled out, hilum a small hollow. 2n 16. * Considerable variation in spine development may be observed in burrs on the same branch (e.g., speci- mens from Parkeston railway sheds, PERTH.). It is not certain that this is natural variation, as Mr. A. C. Linto of the W.A. Agriculture Depart- ment has suggested that it may be due to treat- ment of railway reserves with hormonal weedicides. Varieties: 1. Burrs with a 5-8 coils olive-shaped or globular, peduncle longer than petiole, stipules laciniate, at least some leaves laciniate — var. laciniata 1. Burrs with 3-4£ coils cylindrical, peduncle shorter than petiole, stipules incised, leaves dentate — var. brachyacantha Boiss. Common name: “Cut-leaf medic” which refers to the dissected appearance of some of the leaves of var. laciniata. The name is also loosely ap- plied to var. brachyacantha which does not have laciniate leaves. Cultivars : None. Distribution: Of limited occurrence in the lower rainfall areas usually in the same region as M. minima; north-eastern fringe of the cereal- growing areas of W.A. and east of Kal- goorlie. Variety brachyacantha appears the less common. In Israel var. brachyacantha occurs in more extreme desert habitats than var. laciniata (Friedman in Heyn 1971). Medicago littoralis Rohde* ex Lois-Delong., Not. FI. France: 118 (1810) Australian representation: Naturalised M. lit- toralis is rare in Western Australia and only the var. littoralis has been recorded. M. littoralis cv. Harbinger is described separately as it is un- like the typical naturalized M. littoralis. Description: Stems, petioles and pedicels with moderate-dense simple hairs. Central leaflets 7-14 mm long x 7-11 mm broad, cuneate or obovate, apex truncate retuse or obtuse, apicu- late, margin serrate along upper 4, upper sur- face sparse-densely hairy, lower surface densely hairy, purple flecks on upper side, sometimes con- centrated along the mid rib. Stipules deeply incised with sparse hairs on lower side. Peduncle usually shorter than subtending petiole in flower and fruit, with a hairy awn up to 5 mm long. Flowers 1-3, calyx moderately hairy, teeth longer than tube (atypical for littoralis). Standard 5-7 mm long, keel longer than the wings. Developing pod contracted concealed in calyx or protruding sideways, coiling anti- clockwise. Mature fruit 1-2 per raceme, cylin- drical, 4-5 coils strongly appressed, hard at maturity, 5-6 mm diameter, 5-7 mm high, few hairs on young pods, glabrous at maturty. Venation difficult to see, about 10 radial veins on the coil surface, anastomosing and entering a lateral vein separated from the dorsal suture but in the same plane on the edge of the coil, region between dorsal suture and lateral vein flat, spines long (up to 4 mm) hooked at tips, no basal groove, inserted at 130° in middle to 90° on end coils. Seeds 6-8(9), separated by parti- tions, subreniform or reniform, yellowish.. 24-4 mm long x 14-2 mm wide, radicle less or just equal to half the length of the cotyledons, tip clear, rarely slightly protruding, hilum, ob- scure. 2n 16. * The description of this species was sent to Loiseleur- Deslongchamps by MM. Rohde, Bertoloni, Suffren and Requien. In later works the species name is frequently spelled incorrectly litoralis and the authority given as Rhode. Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 87 Figure 7 . — Medicago littoralis. A-B: var. littoralis. A: burr. B: diagrammatic T.S. of coil edge. C-E: cv. Harbinger. C: burr. D: diagramatic T.S. of coil edge. E: seed. Common name : ‘Strand medic’ selected by the Victorian Herbage Liaison Committee to describe the coastal distribution of this species in its native Mediterranean habitat. Cultivars: cv. Harbinger. Distribution: There are two records of M. littoralis . One is a roadside population which has been known for several years, 10 km south of Geraldton on Mr. E. K. Doncon’s property. It was collected by Mr. G. L. Throssell in 1958 and incorporated into the W.A. University Medicago collection as No. 2500. This accession has been used in crosses between M. littoralis and M. truncatula, and M. tornata (Simon 1965; Simon and Millington 1965). The second record is a PERTH specimen “Merredin test rows — seed from J. Suiter, Moor- ine Rock”. In view of the preference of M. littoralis for coastal habitats the Moorine Rock specimen seems unlikely to be from a natural- ized population. Cultivar Harbinger. ‘Harbinger’ has been regis- tered as a cultivar of M. littoralis. M • littoralis is known to hybridize with both M. truncatula and M. tornata in the field (Heyn 1963) and artificially (Simon 1965; Simon and Millington 1965), and cv. Harbinger appears to have had a hybrid origin. Vegetative features such as the lengths of the branches, and shape of leaflets in cv. Harbinger relate to M. truncatula or M. tornata rather than M. littoralis while the size of the leaflets, the number of flowers., and per- haps the peduncle: petiole ratio indicate some introgression to M. tornata. The position of the lateral veins on the surface rather than the extreme edge of the coils, the presence of in- dentations between spine bases and the insertion of the spines at 90° rather than 130° are M. truncatula characteristics. The only features of M. littoralis are the calyx teeth which are broad at the base and about equal to the tube in length, and the pollen which is spindle- cylinder shaped when dry and shows 3 pores when stained, as is typical for M. littoralis (Lesins and Lesins 1963). M. truncatula pollen is triangular to pyramidal or bisphenoid when dry and shows 4, (rarely 5 or 6) pores when stained. I have observed that M. tornata, a species not included in the Lesins’ survey, has the same pollen morphology as M. littoralis. The origin of cv. Harbinger is confused and it is suggested that it may have originated from Iran in 1940 (Barnard 1972). However, Heyn (1963) states that M. truncatula, M. littoralis and M. tornata do not occur in Iran. If the argument for a hybrid origin is accepted, then the source locality for cv. Harbinger must be a region where two or all three of the species grow and may hybridize. Description : as for M. littoralis with the follow- ing points of difference: Central leaflets (11) 16-19 mm long x 8-11 broad, cuneate only in the rosette stage, obovate or oblanceolate, apex obtuse and apiculate, margin serrate in upper half, rare purple flecks on the upper side. Peduncle equal or longer than subtending petiole in flower, equal or shorter in fruit. Flowers 3-5, calyx teeth equal or shorter than tube. Mature fruits (2) 4-5 per raceme, 3-4 coils, 4-5 mm diameter x 2£-3£ mm high. Lateral veins on the surface of the coil rather than the extreme edge, separated from the dorsal suture by a groove in young and mature pods, except in the most fleshy thick- ened burrs. Spines short and straight or tuber- cules, variable in length on the same pod or between pods on the same plant. Seeds 3-4 (5), 3-3| mm long x H-lf broad. Distribution: The cultivar does best on deep yellow sands in the northern cereal areas and is occasionally naturalised along roadsides in the Geraldton district. It is also grown near Salmon Gums. Medicago lupulina L., Sp. PI. 2:779 (1753) Australian representation: The naturalized Western Australian material behaves as an annual (Quinlivan (1965) but Barnard (1969) reports that in other states plants may be bi- ennial or perennial. The species is commonly included in Medicago, and crosses between M. sativa and M. lupulina are possible when M. sativa is used as the female parent (South- worth 1914; Schroch 1943). M. lupulina has non-articulated cotyledons, a generic charac- teristic of Medicago, but it is very like Melilotus in its vegetative parts, inflorescence and nutlet- like pod. Further, some forms lack the typical Medicago floral tripping mechanism (Heyn 1963) and it has been mentioned above that the thickening of the anther sheath and swelling of the bases of alternate anthers may be reduced or absent. Thus some forms of lupulina do not posses all the characteristics necessary for in- clusion in Medicago and Simon (1969) as shown serological differences between M. lupulina and M. secundifl.Gr a (Section Lupularia) and the re- maining species of Medicago. Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 88 Figure 8 . — Medicago lupulina. A: leaves and fruits. B:° fruits (hairs on leaves and fruits not shown). C: seed. Description: Stems, petioles and peduncles densely covered with simple hairs. Central leaf- lets from leaves in rosette 4-£-6i mm long x 4i-7 mm broad, orbicular, cuneate or obovate, sometimes appearing digitate, with petioles up to 7 cm long; leaves on branches 11-15 mm long x 6-10 mm broad, obovate-oval, apex retuse or apiculate, upper half of margin serrate, clearly pinnate, and petioles rarely more than 1 cm long upper surface virtually glabrous to densely hairy, lower surface always densely hairy. No leaf flecks or marks. Stipules entire or dentate, upper surface usually glabrous (rarely a few hairs), lower surface densely hairy. Peduncles longer than subtending petioles in flower, elon- gating in fruit to reach 3-5 times the petiole, awn lacking or small (1-limm long and diffi- cult to see), floral bracts ± equal to pedicel. Flowers (10) 20-50 in dense oblong heads 10 mm or more long. Calyx moderately hairy the 3 anterior teeth slightly longer, and the 2 pos- terior teeth shorter than the tube. Standard 2-2£ rnm long, sometimes only just longer than the calyx teeth, wings slightly longer than the keel. Pod a single-seeded nut with only the distal end coiled. Mature fruits 10-40 per raceme, reniform with convex sides, or discoid, 2-3-2- mm long x 1-2 mm wide, spineless, glabrous or hairy with appressed or erect, simple, or sim- ple and glandular hairs. Pod often blackening when ripe, prominent radial veins in almost concentric semicircles, anastomosing and run- ning into the dorsal suture. Single seeds, round or oval, yellow or yellow-brown, 11 - 2 - 2 - mm long x f-lf mm wide, radicle longer than half the length of the cotyledons, tip distinct and rarely slightly protruding. 2n = 16, 32: it is not known whether both types occur in Australia. Common names: ‘Black medic’ which relates to the black fruits of the species. Cultivars: None, but it is interesting to note that this species was the first medic offered for sale in We ~ tern Australia in colonial times (Quinlivan et al. 1974). Distribution: Rare in Western Australia. Her- barium specimens from Elleker (W. of Albany), Denmark, Yarloop and Nungarin do not dis- tinguish between plants deliberately cultivated and naturalized occurrences. It is a late matur- ing species and in other Australian states is naturalized in the cooler districts to specimens from the first 3 localities are possibly from naturalized stands but it is unlikely that the species is naturalised at Nungarin. It is known to have been established around Deeside at Lake Muir but it does not appear to have persisted (Quinlivan, pers. comm.). Medicago minima (L.) Bart., Cat. Piant. Siena: 61 (1776) Synonymy: M. minima (L.) Grufbg., FI. Angl.: 21 (1754). Australian representation : Most of the range of variability for the species is seen in Australian material. However, the maximum number of flowers appears to be 5, (in var. brevispina) , while in Mediterranean material up to 8 flowers have been recorded. Description: Stems, petioles and peduncles densely covered with simple hairs, or, especially in some var. brevispina, simple and glandular hairs. Central leaflets 5-14 mm long, 3^-7 mm broad, leaflets from the rosette may be orbicular, cuneate or obovate, on branches they are obo- vate or oblanceolate, apex retuse, mucronate or tridentate, upper h of margin serrate, upper and lower surfaces densely hairy with simple, or simple and glandular hairs. Upper surface of leaf lamina may be without flecks or marks or have a dense line of purple flecks along the mid Figure 9 . — Medicago minima. A-D: var. minima. A and B: burr types. C: venation on coil surface. D: seed. E: var. brevispina burr, (hairs on burrs not shown). Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 89 rib. In the rosette, petioles may reach 4 k cm but on the branches they are usually only 1 cm long. Stipules entire or with small teeth, hairy on both sides. The pubescence on the vegetative parts may give the plant a greyish tomentose apperance. Peduncles shorter or longer than the subtending petiole, extended into a short hairy awn (1-2 mm long) which may be diffi- cult to see. Flowers 1-2 (-5 in var. brevispina) , calyx densely hairy teeth longer than the tube the 3 anterior teeth being slightly longer than the 2 posterior ones. Standard 3i-5 mm long, wings and keel of equal length or keel slightly longer. Developing pod contracted, protruding sideways from the calyx, coiling anticlockwise. Mature fruits 1-2 (-5 in var. brevispina ) per raceme, discoid, olive-shaped or barrel-shaped. 31-5 soft thin walled coils not firmly adpressed, 3-5 mm diameter, 2£— 5 mm high, glabrous or with simple, or simple and glandular hairs, spineless, tuberculate or with long hooked spines (up to 34 mm), inserted at 180° to coil surface in centre coils, and about 120° in apical coils, long clear basal groove. Radial veins few (6-8) strongly bent to S-sbaped. not anastomosing, lining a marked laterial vein which is separated from the dorsal suture by a slightly concave margin about 3 the radius of the coil, across which run only the veins to the spines. Dorsal suture marked but not as wide as in M. laciniata ; or M. praecox. Seeds (3) 4-6 (7) per burr, not separated by membranes * oval-subreniform, yellow or yellow-brown li-2k mm long, f-li mm wide, radicle slightly longer than half the cotyle- dons, tin may protrude slightly, hilum obscure. 2n = 16. Varieties (from Heyn, 1963) : 1. Burr discoid or ovate, spines hooked, longer than 2 the radius of the coils — var. minima 1. Burr olive— or barrel-shaped, spines lacking, tub- ercles or short straight spines less than \ the coil radius — var. brevispina Benth. The drawing of var. brevispina (Fig. 9E) is of a line similar to that sometimes referred to as M. sesAlis Peyr. Burrs with no trace of spines and the coil edges completely smooth are also known. Common name: ‘Woolly burr medic’, derived from the superficial resemblance of the burr to burr medic (M. polymorpha) , and the woolly pubescence of the plants. “Goldfields medic” due to its di tribution in Western Australia, or rarely and incorrectly, “Kalgoorlie Clover”. ‘Little medic’ is the common name in Victoria (Willis 1972). Cultivars: None. Distribution: Widsepread but of isolated occur- rence in drier parts of the wheat belt extend- ing into semi-arid sheep country. As far N. as Port Gregory, E. to Norseman and S. to Ra- vensthorpe, most common in the Merredin and Kalgoorlie districts. The spineless var. brevi- spina has been collected from near Merredin (Quinlivan and Francis, pers. comm.). Spread of the species has sometimes been assisted by farmers in the eastern wheatbelt by raking up burrs and distributing them on their properties. * Heyn (1963) states that there are thin interseminal membranes but these ware not seen in Australian material. Medicago orbicularis (L.) Bart., Cat. Piant. Siena : 61(1776) Australian representation: The Australian ma- terial appears to be from the types with large burr sizes. The species is particularly variable and many attempts have been made to create intraspecific taxa. Heyn (1963) does not re- cognize any varieties as she considers that all characters used in previous subdivisions were not sufficiently discontinuous. Description: Stems, petioles and peduncles with sparse to moderately dense simple hairs, or simple and glandular hairs. Central leaflets 11- 17 mm long X 9-13 mm broad, cuneate or obo- vate, apex retuse or apiculate, marginal serra- tion extending almost to base, upper surface glabrous, lower surface sparsely hairy. Stipules deeply incised or laciniate, glabrous or a few hairs on the lower side, mainly along the mar- gins. Peduncle usually shorter than the subtend- ing petiole in flower, sometimes equal in fruit, produced to a long awn (up to 6-2 mm) with a few hairs. Flowers 1-2(5), calyx moderately hairy with teeth longer than the tube, standard 4-6 mm long, keel ± 1 mm longer than the wings, developing pod initially not contracted but pro- truding sideways from the calyx as a long spiral, coiling anticlockwise. Mature fruits 1(2) per raceme, pod shape (W.A. material) lenticular widest coil in the middle 13-17 mm wide, 3^-7 high, spineless, glabrous. Simple or simple and glandular hairs (often early deciduous), 4-7 soft coils not firmly appressed, with a wide thin border often undulating. Straw coloured or blackening to different degrees. No lateral veins , radial veins anastomosed into a net and running to dorsal suture, sometimes thick and raised towards the dorsal suture coil margin. Seeds (10)15-20(26) separated by short partitions Figure 10 . — Medicago orbicularis. A: developing fruit. B: ventral surface and C: side view of fruit (hairs not shown). D: seed. Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 90 which do not extend out to the edge of the coil, orientated so that the radicle is vertical to the axis (rather than tangential as in other Aus- tralian species), surface rough or minutely tu- berculate (all other Australian species are smooth), yellowish-brown triangular in shape, the radicle being almost as long as the coty- ledons, tip clear but not protruding, 3-3 4 mm long x 1J-3 mm bread, hilum small. 2n 16. Common name: “Button Medic” — derived from the pod shape. Cultivars: Ncne. Distribution: A single herbarium specimen from Southern Cress, collected in 1925. It seems un- likely that the species would naturalise in this area as records from other states indicate that it requires a long growing season and is late maturing (Barnard 1969). Medicago polymorpha L., Sp. PI. 2:779 (1753) Synonymy: M. hispida Gaertn., De Fruct. 2:349 (1791). M. lappacea Desr., in Lam. Encycl. Method. 3 : 637 (1772) p.p. M. apiculata Willd., Sp. PI. 3 : 1414 (1802) p.p. M. denticulata Willd., Sp. PI. 3 : 1414 (1802) p.p. M. terebellum Willd., Sp. PI. 3 : 1416 (1802) p.p. M. reticulata Benth., Cat. Pyr. 101 (1826) p.p. M. confinis Koch., Syn. FI. Germ, ed I. 164 (1837) p.p. Australian representation : The species is a wide- spread weed and occasional pasture component and examples are found of all the three va- rieties described by Heyn (1963): var. poly- morpha, var. vulgaris (Benth.) Shin., var. bre- vispina (Benth.) Heyn. The separation of these varieties is unsatisfactory as there is an abund- ance of material intermediate between all va- rieties. Description: Stems, petioles and peduncles vir- tually glabrous or sparse hairs on young pe- tioles and peduncles. Central leaflets 8-20 (27)mm long x 7-15(19)mm wide, cuneate to obovate, apex obtuse or retuse, apex apiculate, margin almost entire or with upper 4 serrate, upper surface glabrous, lower surface glabrous or with sparse hairs. Leaves sometimes with purple and/or white flecks, and/or a basal purple mark which may be either a solid or an empty inverted V. Stipules deeply incised to laciniate, glabrous or a few hairs concentrated along the margin. Peduncle usually shorter, or sometimes longer, than subtending petiole, awn lacking or up to 3 mm long. Flowers (1)2-7 (8), calyx with few hairs, teeth longer than the tube. Standard 3-5 mm long, wings longer than keel. Developing pod initially contracted and protruding sideways from the calyx, coiling an- ticlockwise. Mature fruits 1-5 per raceme, dis- coid, cylindrical, or truncated cone, 44-84 mm wide x 2-10 mm high, without spines or with short or long spines (up to 34 mm), the longer ones usually hooked and with a short groove at the base, inserted at 180° to pod surface on middle coils sometimes 90° on apical coils, 14-64 coils, soft or hard at maturity, not strongly ap- pressed, surface glabrous or sparsely hairy. Many (15-20 per coil) radial veins strongly curved, anastomosing into a net and joining a lateral vein which is separated from the wide Figure 11 . — Medicago polymorpha. A: flower. B. devel- oping burr. C: seeds. D: var. polymorpha burr. E: var. vulgaris burr. F: venation on ventral coil. G: diagrammatic T.S. of coil edge. H and I: var. brevi- spina burr types. dorsal suture by a narrow groove. Seeds (3) 4-8(11) per burr, separated by partitions, oval to subreniform, yellow-yellow brown, 24-4 mm long x 14-24 mm broad, radicle equal or less than half the length of the cotyledons, tip not at all protruding. 2n = 14. Varieties (from Heyn 1963 and 1970): var. polymorpha: Spines of fruit thick and hardened, coils 4-6, hardening at maturity, dia- meter of broadest coil 5-8 (10) mm (coils some- times 34, the diameter exceeding 6 mm) ; inflo- rescence usually few flowered, ( (1) 2-5 flowers), var. vulgaris (Benth.) Shin, in Rhodora 58 : 310 (1956) emend. Heyn, Scripta Hierosolymitana 12 : 75 (1963). Spines of fruit slender, coils 14-34 usually soft, even at maturity, diameter of broadest coil (2) 3-5 (6) mm, inflorescence usually many flowered (5-10 flowers), var. brevispina (Benth.) Heyn, Scripta Hieroso- lymitana 12 :77 (1963). Spines lacking, margin of coils smooth or tubercled, coils (2)3-5 often hardening at maturity, diameter of broadest coil (24) 3-4 (54 ) mm, inflorescence (1) 2-10 flowers. In Western Australian material there appears to be spineless forms of var. vulgaris with up to 5 fruits per raceme and small burrs (4-6 mm diameter x 2-3 mm high with 2-24 coils), and spineless forms of var. polymorpha which have 1-2 fruits per raceme and larger burrs (4-6 mm diameter x 4-6 mm wide with 34-4(5) coils). All the spineless material is classified as var. brevispina. Common names: “Burr medic” which relates to the spiny pods, or, more rarely, “toothed medic”, which is derived from one of the synonyms, M. denticulata, and refers to the very small teeth usual on the leaf margins. Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 91 Distribution: The spiny varieties of the species are established in most of the agricultural and pastoral areas, extend into the Goldfields, and may be found as adventitious weeds in suitable wet places such as wells and tanks, along roads and railway lines as far north as Carnarvon and east to Forrest. The spineless form is rare but is distributed throughout the most of the range of the spiny varieties. Medicago praecox DC., Adnot. Cat. Hort. Monsp. :123 (1813) Australian representation: The description for Australian material encompasses the whole spe- cies. There are no varieties described for this species and it shows less variation than some other Medicago species. Description: Stems, petioles and peduncles with sparse to moderately dense simple hairs. Central leaflets 5.5-9 mm long, 5-9 mm broad, obcordate, or obovate, apex apiculate, upper 1/3 of margin serrate, upper surface glabrous, lower surface densely hairy, no purple flecks or leaf marks. Figure 12 . — Medicago praecox. A: burr. B: seed. Stipules deeply incised, with a few simple hairs on the lower side only. Peduncles much shorter than the subtending petiole in both flower and fruit, no awn, or rarely a very short one (±0.5 mm). Flowers 1-2, calyx with moderately dense simple hairs, teeth ± equal to tube but front 3 teeth slightly longer than back 2. Standard 2-3 mm, keel ± 1 mm longer than wings. Develop- ing pod contracted and protruding sideways from the calyx, coiling anticlockwise. Mature fruits 1 (2) per raceme, coils not appressed, shape (without the spines) a cylinder or truncated cone, 4-5 mm diameter, 4-5£ mm high, spines 2-3 mm long, hooked at tips, strongly grooved at the base, inserted at SO 0 -120° to coil surface, frequently in opposite pairs from the wide dorsal suture, coil surface with sparse hairs, suture region glabrescent. Radial veins 8-12 per coil, strongly curved, anastomosing, entering a dis- tinct lateral vein which is separated from the wide dorsal suture by a narrow deep groove which is visible from the coil surface rather than from the side of burr. Seeds (3)4-6, separated by partitions, oval to subreniform, yellowish, 2-3 mm long X 1-U mm broad, radicle slightly less than half the length of the coty- ledons, with tip closely appressed, hilum ob- scure. 2n = 16. Common name: “Small leaf burr medic” as this species is sometimes mistaken for a small-leafed form of burr medic ( M . polymorpha ) . Cultivars : None. Distribution: The species was previously thought to be absent from Western Australia, but I have found it at 3 widespread localities so it is un- likely to be a recent introduction. (On red soils N. of Mingenew, 20 km N. of Merredin, and E. of Gnowangerup ) Heyn (1963) notes that it is rare, even in the Mediterranean. Medicago rugosa Desr., in Lam. Encycl. Method. 3:632 (1792) Australian representation: The only represen- tative of this species in Australia is the cultivar Paragosa, so this description covers only the variability observed in the cultivar. Description: Stems, petioles and peduncles densely covered with simple and glandular hairs. Central leaflets 10-23 mm long x 7-15 mm broad, obovate or oblanceolate, apex obtuse or slightly retuse, upper half of leaf margin serrate, upper surface with rare purple flecks or none, glab- rous, lower surface with dense glandular and simple hairs, stipules with small sharp teeth, dense glandular and simple hairs on lower side, very rare glandular hairs on upper side. Peduncle shorter than the petiole of the subtending leaf, sometimes becoming equal when fruiting, peduncle with a hairy awn 4-5 mm long. Flowers (1) 3-5 nearly all inserted on the same side of the peduncle. Calyx with dense glandular and simple hairs, teeth shorter or equal to the tube. Standard 3-4 mm long, keel slightly longer than the wings. Developing pod contracted but pro- truding sideways from the calyx, anticlockwise coiling. Mature fruits 1-3 per raceme, 3^-5 concave coils, strongly appressed, 6-9 mm di- ameter, 3-4 i mm high, spinless, pod disc- or lozenge-shaped, surface with glandular and sim- ple hairs which may rub off exposed parts of dry burrs. No lateral veins, very marked radial veins arising in opposite or alternate pairs from the thick dorsal suture, no anastomosis in the outer -3- of the pod diameter. Seeds 1-2 (rarely more), not separated by partitions, strongly 5mm A B Figure 13 . — Medicago rugosa cv. Paragosa. A: burr (hairs not shown). B: seed. Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 92 reniform, glossy, bright- to dark- or brownish- yellow, 3-4-2 mm long, 2-3 mm broad, radicle less than half the length of the cotyledons with a strongly protruding tip, hilum very small and obscure. 2n = 32. Common name: ‘Gama medic’ given to this species by the South Australian Herbage Plant Liaison Committee, as the accession from which the cultivar was selected, (CPI 7791), was originally collected in Portugal, the homeland of Vasco da Gama the navigator. Cultivar : One cultivar, cv. Paragosa. Distribution: The cultivar grows best on heavy alkaline soils, and is rarely grown and not naturalized in Western Australia. Medicago scutellata (L.) Mill., Gard. Diet, ed. 8: No. 2 (1768) Synonymy : M. scutellata (L.) All., FI. Pedem. 1 : 315 (1785). Australian representation: The species appears to have been introduced into Australia at the end of the 19th century and while local strains have developed (Quinlivan 1965), the species is, on the whole, one of the less variable annual species. The description is mainly based on plants sold by seed firms as ‘Snail medic’. Description: Stems, petioles and peduncles densely covered with simple and glandular hairs. Central leaflets 15-22 mm long, 10-15 mm broad, oval, obovate or oblanceolate, apex obtuse only in lowest leaves, mainly acute, upper § of leaf margin serrate, upper surface with rare purple fleck or none, glabrous, lower surface with dense glandular and simple hairs. Stipules with small sharp teeth, glandular hairs on upper side, ( cf . Heyn 1963) glandular and simple hairs on lower side. Peduncle shorter than the sub- tending petiole, with a long (± 5 mm) hairy awn. Flowers 1-2(3) per raceme; calyx with dense simple and glandular hairs, teeth equal to or longer than the tube, standard 6-7 mm long, keel very slightly longer than wings. De- veloping pod contracted and protruding side- ways from the calyx (or sometimes contained in the calyx), anticlockwise coiling. Mature fruits 1-2(3) per raceme, spineless, cup-shaped or olive-shaped, 11-14 mm diameter, 11-16 mm high, 5-7 coils, basal coils enclosing upper ones, dense glandular and simple hairs when young, exposed parts ± glabrous at maturity. Radial Figure 14 . — Medicago scutellata. A: burr (hairs not shown). B: seed. veins distinct, reticulate, no lateral veins. Seeds 4-6, none in the distal 2-3 coils, separated by very thin partitions, reniform, not glossy, yellow or brownish yellow, 4-6 mm long, 2h-Sh mm broad, radicle less than half the length of the cotyledons, tip projecting, hilum distinct. 2n — 32. Common name: ‘Snail medic’, which refers to the pod shape. Cultivar s: None. Distribution: Rarely cultivated and apparently not naturalised in Western Australia. Medicago tornata (L.) Mill., Gard. Diet, ed. 8 : No. 3 (1768) Synonymy : M. obscura Retz., emend. Urb., Verh. bot. Ver. Brandenb. 15 : 66 (1873). Australian representation : M. tornata does not occur in naturalized populations in Western Australia, the specimen reported by Quinlivan (1965) from Ravensthorpe being an incorrectly identified spineless M. truncatula. The cultivar Tornafield is attributable to M. tornata var. tornata, and although cv. Murrayland is pos- sibly of hybrid origin, having some M. littoralis features, it is best also referred to var. tornata. The source material for cv. Murrayland is naturalized at Pooncarie, N.S.W. Description: Stems, petioles and peduncles with sparse simple hairs. Central leaflets 10-17 mm* long X 9-16 mm wide, obovate, angular obovate, or oblanceolate, apex obtuse, apiculate, upper 2/3 of leaf margin serrate, upper surface gla- brous or sparsely hairy, lower surface sparsely hairy. Leaves may have purple and white flecks, and rarely, a narrow ellipsoidal purple mark on the upper surface. Stipules deeply incised, mo- derately dense hairy on lower side only. Pe- duncle markedly longer than petiole of sub- tending leaf (up to twice as long in flower), awn up to 4 mm long with a few hairs. Flowers (4)7-10(15) crowded in the inflorescence, calyx with moderately dense hairs mostly at the base of the teeth, teeth longer than the tube. Standard 5-8 (10) mm long, keel longer than the wings. Developing pod contracted and concealed in the calyx or sometimes pro- truding sideways coiling anticlockwise (in the Australian cultivars). Mature fruits 4-8 per raceme, discoid, cylindrical or a truncated cone 4-7 mm wide (up to 10 mm wide reported in original description of cv. Tornafield) X 2-7 mm high, spineless, glabrous, 2-5 convex coils, appressed only in the centre in cv. Tornafield, and right to the outer edge in cv. Murrayland. Venation difficult to see at maturity, surface of coil with about 10 radial veins which anasto- mose into a net, and run into lateral veins separated from the dorsal suture, no groove between the dorsal suture and laterial veins, edge of coils smooth. Seeds 3-6 per pod, separated by thick partitions, subreniform to strongly re- niferm, yellow or yellow-brown, 2-2-4 mm long X li-2f mm wide, radicle less than half the length of the cotyledons, tip appressed. 2n = 16. * Leaves up to 30 mm long and 20 mm wide are reported by Millington in the description of the cv. Torna- field (Barnard 1972) but I have not seen leaves of this size. Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 93 Figure 15 .— Medicago tornata. A-C: cv. Tornafield. A: leaves (hairs not shown) and burrs. B: burr. C: seed. D and E: burr and seed of cv. Murrayland. Common name: “Disc medic”, chosen by the W.A. Herbage Liaison Committee to describe the fruit of cv. Tornafield which sometimes looks like a series of discs. Barnard has attempted to change this common name to sand medic, arguing that disc medic might be confused with M. disciformis DC. and in any case cv. Murray- land has cylindrical rather than disc-shaped burrs. Cultivars: cv. Tornafield. This was derived from selection among and F 2 of a cross between var. tornata and var. aculeata, selection being aimed at combining the large burr size and absence of soines of the tornata accession with the early flowering of the aculeata accession. The number of pod coils in cv. Tornafield is 2-3£ which is at the lower limit of the range observed in naturally occurring var. tornata (3-6(8) coils). However in view of the parentage, cv. Torna- field can be included in var. tornata. cv. Murrayland. These plants show some features of M. tornata, mcst importantly the many-flowered inflorescence (4-6 flowers) which has a peduncle much longer than the petiole of the subtending leaf. However the pods of cv. Murrayland are typical of M. littoralis var. inermis in their cylindrical shape with the coils appressed to the extreme edge, and with the lateral veins at the very edge of the coil in the same olane as the dorsal suture. Heyn (1963) considers that the flower characteristics out- weight the burr morphology and places material such as this in M. tornata but notes that intro- gression with M. littoralis has occurred. Similar specimens come from the west Mediterranean and are frequently identified as M. striata Bast. The fruits of cv. Murrayland are quite distinc- tive and cannot be confused with pods of any other naturalised or cultivated species in Aus- tralia at present. Distribution: No naturalized occurrence as yet, but it is likely that the cv. Tornafield introduced in 1969 will soon establish naturalized stands along roadsides in the Geraldton district. Medicago tnmcatjila Gasrtn., emend. Urb., Vcrh bot. Ver. Brandenb. 15:67 (1873) Synonymy: M. tribuloides Desr., in Lam., Encycl. Method. 3:635 (1792). A.ustialian representation : Most of the natura- lized material and the cvs. Cyprus, Jemalong and Hannaford are var. truncatula, or inter- mediate between vars. truncatula and longiacu- leata Urb.; cv. Ghor is var. longiaculeata Urb., and cv. Borung is var. tricycla (Negre) Heyn or intermediate between that variety and var. truncatula. The spineless form has not been described as a variety. Heyn (1963, 1970) also notes that intermediate forms occur, and it has been observed that at the beginning of the season a plant may form burrs with thick, straight, strongly appressed spines typical of var. truncatula, but burrs formed later may have thinner, curved spines not strongly appressed more like var. longiaculeata. Description: Stems, petioles and peduncles with dense simple hairs. Central leaflets 6-20(25) mm long X 5-14(20) mm wide, cuneate or obov- ate, apex obtuse, truncate or retuse, apiculate, margin serrate along upper 1/3 surface sparse to densely hairy, lower surface always densely hairy. Upper surface with or without purple flecks, a large central ellipsoidal purple mark, or a small upper central yellow mark faintly bordered with purple. Stipules dentate to deeply dentate, with sparse to dense simple hairs on lower side only. Peduncle usually shorter than petiole in both flower and fruit, a hairy awn up to 5 mm long. Flowers 1-3(4) per raceme, calyx densely hairy, teeth much longer than the tube with the 3 anterior teeth sometimes slightly longer than the posterior ones. Standard 4^-7 mm long, keel just longer than the wings. De- veloping pod contracted and concealed in the calyx or protruding sideways, coiling clockwise or anticlockwise. Mature fruits 1-2 (4) per raceme, a short or long cylinder or a truncated cone, 3-7 coils strongly appressed and thickened at maturity, 4|-8 mm* diameter, 4|-11 mm high*, usually pubescent especially when young, rarely almost glabrous. About 10 radial veins anastomosing and entering a lateral vein which is on the surface rather than the edge of the coil, and which is separated from the dorsal suture by a groove (particularly marked on young burrs End occasionally obliterated in very thickened mature burrs), spines lacking, short or long (up to 3mm), straight or curved, base with or without a short groove, sometimes with base very thickened, inserted at 90° to 120°. Seeds (3)4-8(12), separated bv thin partitions, subreniform to reniform, yellow to brownish- yellow, 2i-4 mm long* X 1 - 2 - 2 - 2 - mm broad, ra- dicle less than half the total length of the cotyledons, tip appressed. 2n 16. Common name: “Barrel medic” which refers to the pod shape. Cultivars: The main features of the six cultivars are: cv. Hannaford (previously ‘Commercial barrel medic’ or ‘South Australian barrel medic’): * The registration description of cv. Cyfield gives the upper limit of pod diameter 12 mm, pod height as 15 mm, and seed length 6 mm. I have not seen material in this size range. Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 94 purple mark Figure 16 . — Medicago truncatula. A-C: cv. Jemalong. A: leaf. B: burr. C: apical coils showing anticlockwise coiling. D: cv. Cyprus apical coils showing clockwise coiling. E and F: cv. Ghor leaf and burr. G: cv. Borung burr. H: cv. Cyfield burr. I: diagrammatic T.S. of coil edge. J: cv. Cyprus seed. K: cv. Cyfield seeds. L: flower. M: developing fruit (hairs on leaves and burrs, except M, not shown). variable burr types with spines strongly ap- pressed or not so. Mainly clockwise, some anti- clockwise, coiling. Burrs of this cultivar com- parable in size with cv. Cyprus and cv. Jema- long: 7-10 mm high X 4|-6 mm wide. cv. Cyprus: burrs with strongly appressed spines, clockwise coiling. Indistinguishable bota- nically but earlier-flowering than similar types in cv. Hannaford. cv. Jemalong (previously ‘173’) large purple central ellipsoidal leaf mark, burrs with strongly appressed spines, anticlockwise coiling. cv. Cyfield : a spineless cultivar with large pods (6-11 mm high), large seeds (up to 4mm long) , and both clockwise and anticlockwise coil- ing burrs. cv. Borung: small burrs (4-6 mm high) with 2-3-2- anticlockwise coils, short spines, 3-4 fruits per raceme. cv. Ghor: leaves with an upper central yellow mark sometimes bordered with purple-brown. Burrs large, woody, anticlockwise coils, large spines up to 3 h mm long, not appressed, seeds large. Distribution: M. truncatula was naturalised in ether states before the introduction of the cul- tivars. In PERTH there is only one specimen of a spiny barrel medic — from Hopetoun in 1932, which predates the introduction of the spiny cultivars, and a spineless one from Ravensthorpe in 1963 pre-dates the introduction of cv. Cyfield. Despite the lack of herbarium records, it is thought that the species was naturalized over a wide area: Merredin, Leonora, Kalgoorlie, Sal- mon Gums, before the 195Q’s (Quinlivan, pers. comm.) . Naturalized material from cv. Hannaford, Cyprus, and rarely Jemalong, is common in the drier cereal and sheep areas, particularly from Merredin to Southern Cross but may also be found along roadsides as far north as Shark’s Bay and Laverton and south east to Salmon Gums. The other cultivars, Cyfield, Borung and Ghor, are recent registrations as yet little used in Western Australian agriculture, and not recorded from naturalised populations. Acknowledgements . — This work was supported by a grant from the W.A. State Wheat Industry Research Committee. Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 95 References Amor, R. L. (1965). — Barrel medic ( Medicago tribu- loides Desr.) in the Australian wheat belt. J. Aust. Inst, agric. Sci. 31 : 25-35. Andrew, W. D. and F. W. Hely (1969). — Frequency of annual species of Medicago on the major soil groups of the Macquarie region of New South Wales. Aust. J. agric. Res. 11 : 705-14. Armstrong, J. M. and W. J. White (1935). — Factors influencing seed-setting in alfalfa. J . agric. Sci., Camb. 25: 161-79. Barnard, C. (1969). — “Herbage Plant Species.” Austra- lian Herbage Plant Registration Authority, CSIRO, Canberra. Barnard, C. (1972). — “Register of Australian Herbage Plant Cultivars.” Australian Herbage Plant Registration Authority, CSIRO, Canberra. Baum, B. R. (1963). — A clarification of the generic limits of Trigonella and Medicago. Can. J. Bot. 46: 741-9. Black, J. M. (1957). — “Flora of South Australia” 2nd Ed. (Govt. Printer: Adelaide). Bolton, J. L. (1262) .—“Alfalfa” (World Crops Books ed. N. Polunin) (Leonard Hill: London). Heszky, L. (1972). — Role played by parts of flower in the tripping mechanism of alfalfa ( Medi- cago sativa L.) flower. Acta Agron. Acad. Scient. Hung. 21: 186-90. Heyn, C. C. (1263). — “The Annual Species of Medi- cago.” Scripta Hierosolymitana 12 (Magnes: Jerusalem). Heyn, C. C. (1263). — An evolutionary study of fruit morphology in the tribe Trigonelleae (Legu- mino:ae). Phytomorphology: 18: 54-9. Heyn, C. C. (1970). — Medicago L. (annuals) In Davis, P. “Flora of Turkey” Vol. 3 (Edinburgh Univ. Press) 483-511. Heyn, C. C. (1971). — Biosystematic approaches to the solution of taxonomic problems in Israel. In “Plant Life of South-West Asia” (Ed. P. H. Davis et al.) (Botanical Society of Edinburgh). Hutchinson, J. (1964). — “The Genera of Fowering Plants” (Clarendon Press: Oxford). Larkin, R. A. and H. O. Graumann (1954). — Anatomi- cal structure of the alfalfa flower and an explanation of the tripping mechanism. Bot. Gaz. 116: 40-52. Lesins, K. and C. Gillies (1973). — Taxonomy and cytogenetics of Medicago. In “Alfalfa Science and Technology” Monograph 15, American Society of Agronomy. Lesins, K. and I. Lesins (1963). — Pollen morphology and species relationships in Medicago L. Can. J. Genet. Cytol. 5: 270-80. Lesins, K., S. M. Singh, and A. Erac ( 1968) .—Relation- ship of taxa in the genus Medicago as revealed by hybridization. V. Section Inter- textae. Can. J. Genet. Cytol. 13: 335-46. McComb, J. A. (1971). — Medicago collections in Aus- tralia. PI. Introd. Rev. 8: 50-8. Negre, R. (1956). — Les luzernes du Maroc. Trav. Inst. scient. cherif. Serie Botanique: No. 5 (trans- lated J. A. McComb, CSIRO translation). Negre, R. (1959). — Revision des Medicago d’Afrique du Nord. Bull. Soc. Hist. nat. Afr. N. 5: 267-314 (translated J. A. McComb, CSIRO Trans- lation). Quinlivan, B. J. (1965). — The naturalised and culti- vated annual medics of Western Australia. J. Agric. West. Aust. 6: 532-43. Quinlivan, B. J., J. A. McComb and A. Devitt. (1974). — Annual medics of Western Australia. Department of Agriculture of W.A. Bulle- tin No. 3874. Schrock, O. (1943). — Beobachtungen an einem Bastard zwischen Luzerne (Medicago media) und Gelbklee (Medicago lupulina) und seiner Nachkcmmenschaft. Zuchter 15: 4-10. Simon, J. P. (1965). — Relationships in annual species of Medicago. II. Interspecific crosses between M. tornata (L.) Mill, and M. littoralis Rhode. Aust. J. agric. Res. 16: 51-60. Simon, J. P. (1969). — Serological studies in Medicago, Melilotus , Trigonella and certain other genera of the Leguminosae. I Quantitative precipitin tests and immunodiffusion tech- niques. Bot. Gaz. 130: 127-41. Simon, J. P. and A. J. Millington (1967). — Relation- ship in annual species of Medicago. III. The complex M . truncatula Gaertn. — M. lit- toralis Rhode. Aust. J. Bot. 15: 35-73. Sinskaya, E. N. (ed) (1950). — “Flora of Cultivated plants of the USSR . XIII. Perennial Legumi- nous Plants. Part 1. Medic, Sweetclover. Fenugreek.” (translated N. Landau, Israel Program for Scientific Translations: Jeru- salem 1961). Southworth, W. (1914). — Alfalfa hybridization. J. Hered. 5: 448-57. Urban, I. (1873). — Prodromus einer Monographic der Gattung Medicago L. Verh, bot. Ver. Prov. Brandenb. 15: 1-85. Urban, I. (1877). — Uber Medicago Ic. 19: 125-137. Willis, J. H. (1972). — “A Handbook to Plants in Vic- toria” 2 (University Press: Melbourne). Journal of the Royal Society of Western Australia, Vol. 57, Part 3, October, 1974. 96 INSTRUCTIONS TO AUTHORS Contributions to this Journal should be sent to The Honorary Editor, Royal Society of Western Australia, Western Australian Museum, Perth. Papers are received only from or by communication through, Members of the Society. The Council decides whether any contribution will be accepted for publication. All papers accepted must be read either in full or in abstract or be tabled at an ordinary meeting before publication. Papers should be accompanied by a table of contents, on a separate sheet, showing clearly the status of ail headings; this will not necessarily be published. Authors should maintain a proper balance between length and substance, and papers longer than 10,000 words would need to be of exceptional importance to be considered for publication. The Abstract (which will probably be read more than any other part of the paper) should not be an expanded title, but should include the main substance of the paper in a con- densed form. Typescripts should be double-spaced on opaque white foolscap paper; the original and one carbon copy should be sent. All Tables, and captions for Figures, should be on separate sheets. Authors are advised to use recent issues of the Journal as a guide to the general format of their papers, includ- ing the preparation of references; journal titles in references may be given in full or may follow any widely used conventional system of abbreviation. Note that all illustrations are Figures, which are numbered in a single sequence. In composite Figures, made up of several photographs or diagrams, each of these should be designated by letter (e.g. Figure 13B). Illustrations should include all necessary lettering, and must be suitable for direct photo- graphic reproduction. To avoid unnecessary handling of the original illus- trations, which are usually best prepared between li and 2 times the required size, authors are advised to supply extra prints already reduced. Additional printing costs, such as those for folding maps or colour blocks, will normally be charged to authors. It is the responsibility of authors to adhere to the International Rules of Botanical and Zoological Nomenclature. Palaeontological papers must follow the appropriate rules for zoology or botany, and all new stratigraphic names must have been previously approved by the Stratigraphic Nomenclature Com- mittee of the Geological Society of Australia. Thirty reprints are supplied to authors free of charge, up to a maximum ot 60 for any one paper. Further reprints may be ordered at cost, provided that orders are submitted with the returned galley proofs. Authors are solely responsible for the accuracy of all information in their papers, and for any opinion they express. Journal of the Royal Society of Western Australia Volume 57 1974 Part 3 Contents 10. Amygdaloidal rock from Watheroo in the Permian Nangetty Formation, Western Australia. By J. E. Glover. 11. Australites of mass greater than 100 grams from Western Australia. By W. H. Cleverly. 12. Annual Medicago species with particular reference to those occurring in Western Australia. By J. A. McComb. Editor: A. J. McComb The Royal Society of Western Australia, Western Australian Museum, Perth 39543/7/74—625 WILLIAM C. BROWN, Government Printer, Western Australia JOURNAL OF THE ROYAL SOCIETY OF WESTERN AUSTRALIA VOLUME 57 PART 4 DECEMBER, 1974 PRICE TWO DOLLARS REGISTERED FOR POSTING AS A PERIODICAL-CATEGORY B THE ROYAL SOCIETY OF WESTERN AUSTRALIA PATRON Her Majesty the Queen VICE-PATRON His Excellency Air Commodore Sir Hughie Edwards, V.C., K.C.M.G., C.B., D.S.O., O.B.E., D.F.C. COUNCIL 1974-1975 G. A. Bottomley, B.Sc., Ph.D. B. E. Balme, D.Sc. P. R. Wycherley, O.B.E., B.Sc., Ph.D., F.L.S. A. F. Trendall, B.Sc., Ph.D., A.R.C.S., F.G.S. M. Perry, B.Sc. (Agric.) (Hons.) G. Perry, B.Sc. (Hons.) S. J. Curry, M.A. A. Neumann, B.A. A. J. McComb, M.Sc., Ph.D. R. M. Berndt, M.A., Dip.Anth., Ph.D., F.R.A.I., F.F.A.A.A. C. E. Dortch, B.S., M.Phil. L. J. Peet, B.Sc., F.G.S. P. E. Playford, B.Sc., Ph.D. P. G. Quilty, B.Sc. (Hons.), Ph.D. J. C. Taylor, B.Sc., Ph.D., A.R.C.S. P. G. Wilson, M.Sc. President Vice Presidents Past President Joint Hon. Secretaries Hon. Treasurer Hon. Librarian Hon. Editor 13. — Sand Fulgurites from Western Australia by J. E. Glover 1 Manuscript received 19 February 1974; accepted 19 March 1974 Abstract Fulgurites have been recovered from sand at Willetton, East Victoria Park and Coopers Sandpit, Canning Vale, all near Perth, and from Thangoo Station, near Broome. The Willetton material consists of tube, wall and flange fragments of very light grey vesicular lechatelierite having a refractive index of 1.461 and a silica content of 99.4%. Partly fused quartz grains are embedded in the rough dull outer walls, but are absent from the botryoidal, shiny, inner surfaces. The data accord with an origin due to fusion of sand by lightning. Irregularly shaped black bodies about one mm in diameter scattered in the glass are higher in Fe 2 0.3, Ti0 2 and Mn0 2 than surrounding lechatelierite and probably repre- sent glass stained by oxides from heavy min- erals in the parent sand. The other fulgurites resemble the Willetton material mineralogically and texturally, but contain a higher propor- tion of cylindrical pieces. An artificial fulgurite from Cottesloe, con- sisting of vesicular lechatelierite, is compared with the natural fulgurites. It is broader, lacks a central lumen and flanges, and most of its embedded sand grains are altered to cristobalite. Introduction The term fulgurite (from the Latin / ulgur, lightning) has long been applied to tube-like glassy bodies found on and near the Earth’s surface and supposedly formed from sand or rock melted by lightning strikes. The term was first used in 1790 (see Harland & Hacker, 1966) but the bodies had been noted earlier. Many fulgurites have been observed since, and there is a comprehensive discussion by Frondel (1962), who lists the significant references to that time. Artificial fulgurites (Petty, 1936; Fenner, 1949; Raeside, 1968), pseudofulgurites of opal (Read, 1951), and palaeofulgurites (Harland & Hacker, 1966) have also been described. In Australia, fulgurites and possible fulgurites are known from Moreton Island and Springsure, Queensland (Connah, 1947; Fenner, 1949), Bondi and Macquarie Harbour, New South Wales (Baker, 1959), numerous localities in western Victoria (Fenner, 1949; Beasley, 1963), Oodna- datta, Farina and Mt Remarkable, South Austra- lia (Fenner, 1949), Tempe Downs, North- ern Territory (Baker, 1953a), and several locali- ties in the Yilgarn Block, Western Australia (Simpson, 1931; Trendall, 1964). Artificial fulgurites from Cottesloe and Welshpool, West- ern Australia, caused by high voltage electric 1 Geology Department, University of Western Austra- lia, Nedlands, W.A. 6009. currents, were collected by Professor E. de C. Clarke and described by Fenner (1949). Previ- ously undescribed natural fulgurites from Thangoo Station, southeast of Broome*, and from Kent Street, East Victoria Parkt, about 5 km southeast of Perth, have been recognized in the rock store of the Geology Department, University of Western Australia. This paper describes lechatelierite fragments discovered by the author at Willetton and at Coopers Sandpit, Canning Vale, both in the Perth area, and also describes the fulgurites from Thangoo Station and East Victoria Park in the Geology Department repository. The material is compared with the artificial Cottesloe fulgurite**, for which some new data are given, and the West Popanyinning fulgurite of Simpsonft. It is shown that all the newly described material can be categorized as classic tubular or sand fulgurites and that they re- semble the West Popanyinning fulgurite, for which an origin by lightning fusion of sand is accepted. Location and stratigraphic position of the fulgurites Some five hundred fulgurite fragments were recovered from a cleared area on the southern edge of Leach Highway where it joins High Road, Willetton, about 10 km southsoutheast of Perth, and one fulgurite fragment was col- lected from Coopers Sandpit in Ranford Road, Canning Vale, about 17 km southsoutheast of Perth (see Fig. 1). At each locality the material was in a fixed dune of the Bassendean Dune System of McArthur & Bettenay (1960). The dunes had been cleared of vegetation and soil to a depth of 30 cm or more, and it is not known whether the clearing, or earlier movement in the dune, caused the fulgurite fragmentation. The abundant Willetton material is being ex- posed by deflation, because more is found with succeeding visits. The smaller flake-like frag- ments are probably moved fairly easily by the wind. The Willetton exposure will be overbuilt shortly, and the discovery site at Coopers Sand- pit has already been excavated and destroyed. * University Geology Department No. 42016 t University Geology Department No. 1158 ** University Geology Department No. 12141 tt University Geology Department No. 12158 Journal of the Royal Society of Western Australia, Vol. 57, Part 4, December, 1974. 97 Figure 1. — Map of the Perth area showin g fulgurite localities. Scale: 10 km -= 6.3 cm. Journal of the Royal Society of Western Australia, Vol. 57, Part 4, December, 1974. 98 The East Victoria Park material is recorded in Geology Department files as having been pre- sented by Mr H. R. Gildard, of 80 Kent Street, in 1935. It was said by Mr Gildard to have been dug out of a hole about 2 metres below the sur- face of his backyard, and to have been in ap- proximately a vertical position. About 30 centi- metres were recovered. The map of McArthur & Bettenay (1960) shows that the Kent Street specimen, like the Willetton and Canning Vale material, was in sand of the Bassendean Dune System. The Thangoo fulgurite was presented to the Geology Department in 1958 by Dr P. E. Play- ford and is understood to have come from a vegetated sand dune on Thangoo Station, southeast of Broome. Petrology The colours and corresponding numerical designations used below refer to the Rock-color Chart distributed by the Geological Society of America (Rock-color Chart Committee, 1963). Petrography of the Willetton material About 100 grams of fulguritic material has been collected at Willetton :!: . Some of the 500 fragments consist of small tubes and some of compressed, highly contorted envelopes open along one or two edges, the latter probably representing broken portions of prominent flanges. Most of the Willetton material, how- ever, encloses no central space and comprises wall fragments about one mm thick and up to about three square cm in area (Fig. 2). All fragments have shiny, somewhat botryoidal inner surfaces and contrasting, dull, jagged exteriors embedded with white, rounded, and subrounded sand grains. The mineral com- prising the walls is very light grey (N8) glass with numerous bubbles and a few irregularly shaped, widely dispersed black bodies about one mm in diameter. Under the microscope the fragments are seen to consist mainly of colourless vesicular glass with bubbles ranging from 0.001mm to 0.04 mm in diameter. The bubbles fall into two categories. The small cavities that have not been breached during sectioning are apparently gas-filled, and their very low refractive index relative to the surrounding glass is emphasized by their black margins. The larger cavities have been breached and filled with a medium of refractive index 1.54, and these cavities show no black borders. Bubbles tend to be oriented with their long axes normal to the plane of the fulgurite frag- ment, or to the lumen where present. Most of the bubble cavities are clear, but a few contain some finely divided, indeterminate pale brown mineral. The photomicrograph illustrating the texture of the fulgurite (Fig. 3) bears a striking resemblance to that used by Julien (1901, Fig. 2) in his illustration of a Polish sand fulgurite. * University Geology Department No. 72124 Figure 2— Fragments of the Willetton lechatelierite, showing (top) the shiny somewhat botryoidal inner surface, and (bottom) the dull rough outer surface. Note the small black portions. The upper fragment is about 2 cm long. The glass shows two types of boundary in transverse section. One boundary, correspond- ing to the inner surface, is smooth, and the other, corresponding to the rough dull outer surface, is uneven. The glass toward the uneven boundary commonly has a rather granular tex- ture and locally grades into faintly anisotropic material. Discrete sand grains are embedded in the outer wall, but are not shown in Figure 3. These grains have an unusual appearance under the microscope and commonly consist of biaxial quartz with curved cracks filled with glass. Some individual grains range from an intimate, turbid mixture of quartz and glass on one side, to clear colourless glass on the other. The refractive index of the glass measured with sodium light is 1.461 ± .002, close to that given by Winchell & Winchell (1956, p. 250) for lechatelierite, and within the range of values reported by Frondel (1962, p. 322) for fulgurite glasses. The black bodies visible in handspecimen are formed of irregularly shaped, brown to black opaque cores and concentrations of wisps that grade outward, with increasing translucency, into brown and pale brown glass. Black bodies of about the same size and distribution are Journal of the Royal Society of Western Australia, Vol. 57, Part 4, December, 1974. 99 evident in some fulgurites photographed by Fenner (1949), notably those from South Australia. They have been mentioned by other authors (see, for example, Petty, 1936, p. 191 and Frondel, 1962, p. 322), and Simpson (1931, p. 146) refers to “small areas . . . darkened by the presence of iron silicate, etc.” They are evidently common in tubular sand fulgurites, but have excited little comment or investigation. Figure 3. — Photomicrograph of thin section of fulgurite from Willetton. The smooth inner surface is on the left, and contrasts with the rough outer surface on the right. Large vesicles tend to be elongated normal to the lumen, and small bubbles show up as almost solid black circles. Length shown 2 mm. Chemistry of the Willetton material The Willetton glass is lechatelierite consisting of 99.4% Si0 2 , with small amounts of titania and other oxides (see Table 1 for comparative analyses of Western Australian glasses). To establish the nature of the scattered black bodies some of the glass was crushed and black material was handpicked for analysis. Few of the selected crushed fragments consisted of pure black material, and most were contaminated by at least an equal volume of light grey glass. X-ray powder pattern photographs showed no lines, confirming the impression gained from micro- scopic examination that the black material is not crystalline. Comparative analyses of the dark concentrate and the light grey glass of the fulgurite by the atomic absorption spectrometer showed that Ti0 2 was concentrated in the dark material by a factor of 2, Fe 2 0 3 by a factor of 4, and Mn0 2 by a factor of more than 6. These oxides were probably supplied by heavy minerals in the fused sand, and doubtless account for the dark colour. Table 1 Analyses of Western Australian vesicular and scoriaceous glasses Willetton* West Popan- y inning** Widgie- moolthaf Widgie- mooltha ft SiO, 99-4 88-46 64-9 59-2 TiO., 014 0-46 0-06 n.d. ALd 3 0-09 6-69 9-12 13-0 Fe,0 3 0012 1-16 5-75 9-4 Mn O 0-002 Tr 0-02 n.d. Mg O 0-0795 0-17 3-27 5-7 Ca O 0-02 0-17 120 10-4 Na,0 0-011 0-01 2-05 2-0 KoO 0-02 2-68 1-64 1-1 P 2 0 5 n.d. n.d. 0-15 n.d. Cu .... n.d. n.d. n.d. 0-01 H»0 + Loss on n.d. n.d. 0-41 n.d. ignition 0-02 — — — — 99-8 (approx.) N = 1-461 99-80 N = 1-465 G 2-21 99-37 100-8 * Analyst Labtech Pty Ltd, Job No 5917, Rack No 6250 (combined XRF, Atomic absorption) ** Reported by Simpson (1931, p.146) t Specimen R851, “fulgurite slag”, Widgiemooltha area, Analyst Govt. Chem. Lab, reported by Trendall (1964, p.7) ft “Fulgurite slag”, near Paris Goldmine, Widgiemooltha area, Analyst C. E. S. Davis, reported by Trendall (1964, p.7). The Willetton sand The fulgurites are mainly found in white to pinkish grey (N9-5YR8/1) sand near the crest of a fixed dune, but winds sweep the bare patch and small fragments are also found in nearby orange sand (see below). The white sand is well sorted (So - 1.22) and consists mainly (about 99.6%) of subrounded to well-rounded quartz grains with a median diameter of 0.28 mm. Some of the quartz is practically free of inclu- sions, and some contains minute inclusions of black opaque minerals, leucoxene, rutile, tour- maline, hypersthene, zircon, probable apatite and fluid. Heavy minerals separated in bromo- form from a sample of sand made up 0.44% by weight and consist mainly of ilmenite and leucoxene (see Table 2). Also near the top of the dune, and containing a few small fulgurite fragments, is some very pale orange to greyish orange (10YR8/2-10YR 7/4) sand. This sand is as well sorted as the white sand, yielded an almost identical crop Journal of the Royal Society of Western Australia, Vol. 57, Part 4, December, 1974. 100 Journal of the Royal Society of Western Australia, Vol. 57, Part 4, December, 1974. 101 of heavy minerals (0.45% by weight), and seems to differ only by the discontinuous film, appar- ently of limonite and hematite, on many of the grains. The concentration of fulgurite frag- ments (particularly the larger ones) in the white sand, and the presence in the glass of embedded white grains but not of orange-coloured grains, shows that the fragments originated in white sand and were recently blown into the orange- coloured sand. Table 2 Mineralogy of white sand from Willetton Mineral Wt% Quartz .... 99-56 Ilmenite .... 0-25 Leucoxene .... 0-09 Magnetite 0-03 Staurolite .... 0-02 Kyanite 0-02 Zircon .... 0-01 Garnet .... Trace Tourmaline .... Trace Rutile .... Trace Spinel .... Trace Andalusite .... Trace Unknown .... Trace Mineralogical source of the Willetton fulgurites The white sand of the Willetton area would yield a highly siliceous glass if fused and chilled, and the high silica content of the Willetton fulgurites, together with their incorporated partly fused quartz grains, accords with such an origin. Ti0 2 (0.14%) is fairly close to the amount calculated for the sand from its mineralogy (0.22%), butFe 2 0 3 (0.012%) is significantly lower than the sand mineralogy would indicate (0.17%). The reason for this apparent discrep- ancy can only be speculated: iron may have migrated quickly when the quartz fused, or may have been leached out of the glass later. The Coopers Sandpit material The material from Coopers Sandpit, Canning Vale*, consists of an irregularly flanged tube about 2.5 cm long with an inner diameter of about 0.7 cm and a variable outer diameter of about 0.9 cm (Figs. 4, 5). The outer walls are rough and uneven, and contain numerous, embedded, white to very light grey (N9 to N8) , rounded to subrounded, fine to medium-grained sand grains. The inner walls consist of smooth, shiny, somewhat botryoidal very light grey (N8) glass with a few black stains about 1 mm in diameter. The weight of the specimen is 0.8 grams. A small fragment was broken from the tube, crushed, and examined microscopically in oils. The material is highly vesicular colourless glass with a refractive index of 1.461 ± .002. * University Geology Department No 72125 Figure 4.— Side view of fulgurite fragment from Coopers Sandpit, Canning Vale. Note the rough dull outer surface, and smooth bright inner surface. Length of fragment 2.5 cm. Figure 5.— End-on view of fulgurite fragment from Coopers Sandpit, Canning Vale. The rough, flanged exterior contrasts with the smooth lumen. The lumen has a diameter of about 7 mm. The East Victoria Park material This material consists of a tube 5.5 cm long and 1.5 cm wide, with walls about 1 mm thick. The outer surface is light grey (N7) and has a rough feel, but is only slightly flanged. Pale yellowish orange (10YR8/6) to very pale orange ( 10YR8/2) rounded to subrounded quartz sand grains embedded in the light grey glass give it a speckled greyish orange appearance. The inner surface of the tube consists of shiny, smooth but crinkled, very light grey (N8) to light grey 4c Hibbertia sp. sp. sp. Pimelea angustifolia R.Br. 4c 4c cf. 4c Pimelea rosea R.Br.. * 4c 4< 4c 4c 4c Pimelea nervosa (Walp.) Meisn * * 4c 4c 4c 4c Pimelea serpyllifolia R.Br. 4c 4c 4c Baecfcea sp. aff. crispiflora F. Muell Beaufortia micrantha Schau 4c 4< 4c 4c Calytrix tetragona Labill 4c 4c cp. Cliamelaucium ax ilia re F. Muell Calothamnus gracilis R.Br. 4c 4c 4f 4c 4c Darwinia diosmoides (DC.) Benth. * 4c 4c 4c 4c 4c Darwinia vestita (Endl) Benth 4c 4c 4c * Eucalyptus angulosa Schau. (? = E. incrassata) 4> 4c 4c 4c Eucalyptus cooperana F. Muell Eucalyptus diver sifolia Bonpl Eucalyptus foecunda Schau. Eucalyptus eremophila (Diels) Maiden .... * Eucalyptus incrassata Labill ? Eucalyptus scyphocalyx (F. Muell.) Maiden & Blakely Eucalyptus uncinata Turcz. Melaleuca elliptica LabiW. * 4c 4c 4> 4c * ( granite) Melaleuca confer ta Benth. 4c 4c 4c ? Melaleuca lanceolata Otto * 4c 4c 4> 4c 4c Melaleuca pentagons Labi 11. * * 4c 4c Melaleuca pulchella R.Br. 4c 4c 4c * V erticordia brownii (Desf.) DC * 4c * 4c 4c 9 Verticordia plumosa (Desf.) Druce 4c 4c 4< * 4c * 9 cf. Acrotriche cor data (Labill.) R.Br (*) 4c Conostephium drummondii (Stschegl.) Gardn 4c 4c (*) Leucopogon aff. squarrosus Benth. (*) Lysinema ciliatum, R.Br * Styphelia hainesii F. Muell. 4c 4c Samolus repens (Forst.) Pers * 4c 4c 4c 4c 4c * 4c 4c Prostanthera sp sp. Halgania lavandulacea Endl * 4c 4c 4c Myoporum insulare R.Br. * 4> 4: cf. 4c Lechenaultia formosa R.Br. * 4c 4c 4c 4c 4c 4c Leclienaultia tubiflora R.Br. 4c 4c 4= 4c 4c Cf. 4c Scaevola crassi folia Labill * 4c 4c 4c 4c 4c Goodenia aff inis De Vriese * 4c 4c 4c Goodenia decursiva Fitzg 4c 4c * (granite) Dampiera parviftora R.Br. 4c 4c ? Stylidium pilosum Labill 4c 4c 4* 4c Calocephalus brownii (Cass.) F. Muell * 4c 4c 4c 4c * 4c Helichrysum obtusifolium F. Muell. & Sond * 4c 4c 4> II elipterum floribundum DC. 4c 4c 4c 4c Olearia axillaris (DC.) F. Muell 4c 4c 4c 4c 4< Western Australia:E— Eremean province (Fig. 1); the lower case letters are the initial letter of the zones within the south-west province (Fig. 1) (Beard 1970). Sand Patch: PC — Point Culver; TO — Toolinna; TC — Twilight Cove (Fig. 2). sp. — unidentified species present, cf. — specimen collected comparable with species name listed, granite — species associated with granite monadnocks (Beard 1970). noted that the Toolinna and Twilight Cove sand patches are very lightly vegetated. This contrasts with the very dense scrub that is found west of Israelite Bay. Species distributions Species distributions are given in Table 2. The collections from Point Culver (by M. G. Brooker in 1973) and from Toolinna (by E. C. Nelson in 1973) are both incomplete and include only the most frequent or noticeable species, but the Twilight Cove collections are much more exten- sive. This latter area has been visited several times recently while the two former regions have only been visited once. Despite these inade- quacies, the information available is considered sufficient to allow comparisons to be made and conclusions to be reached that are unlikely to be altered significantly by further collections. The species lists (Table 2) were examined and the species restricted to deep siliceous sand were noted. The distributions of the species Journal of the Royal Society of Western Australia, Vol. 57, Part 4, December, 1974. Ill within southern Australia were determined using Black 1922, Eichler 1965 and Beard 1970, and data from the Western Australian Herbarium (PERTH). The distributions of some of the species are discussed below. (a) Haemodoraceae Anigozanthos is endemic to the south-west province of Western Australia, where species are found mainly in the siliceous sand plain (Gard- ner 1973). Anigozanthos rufa is the only species that occurs east of Esperance and was previously recorded as far east as Israelite Bay and Mount Ragged. Brooker (pers. comm.) recorded the species on “burnt sandhills, approx. 5 miles north-west of Point Culver about 3-4 miles from the sea”. The species is a typical member of the quonkan. It has not been found at Toolinna or Twilight Cove and therefore it is not strictly disjunct on the sand patches. (b) Proteaceae Both Banksia and Adenanthos are found in the siliceous sand plain areas west of Israelite Bay, and as far as Shark Bay. The genera are mainly restricted to areas where the soils have little or no calcium, and no species of either genus has been recorded on the Nullarbor Plain in calcareous soils. Neither genus is endemic to Western Australia but none of the species found in south-western Australia occur in south- eastern Australia (Rao 1971). Banksia speciosa is a very common shrub near Israelite Bay and at Mount Ragged. It was collected by Brooker at Point Culver, thus ex- tending its distribution further to the east. It probably occurs in abundance along the Israelite Plain, but does not reach Toolinna. Banksia media was collected at Toolinna and also at Point Culver, but it has never been found at Twilight Cove. Therefore the Toolinna popula- tion is the most easterly population of a Western Australian Banksia species. B. media is common in the quonkan between Israelite Bay and Cape Riche (Erickson et a!., 1973), and of mallee scrub further inland (J. S. Beard, pers. comm.). John Eyre stated in his journal for May 1, 1841: One circumstance in our route today cheered me greatly, and led me to expect some important and decisive change in the character and for- mation of the country. It was the appearnce for the first time of the Banksia, a shrub which I have never found to the westward of Spencer’s Gulf [South Australia], but which I knew to abound in the vicinity of King George’s Sound [Western Australia], and that descrip- tion of country generally .... Isolated as it was amidst the scrub and insignificant as the stunted specimens were that I first met with, they led to an inference that I could not be mistaken in ... ” (Eyre 1847, vol. 2, p. 13-14). That night (May 1) Eyre probably camped near the Toolinna waterhole (Fig. 2), having already passed through sand patch which is situated northeast of the waterhole. The description of the “stunted specimens” of Banksia, fits the habit and form of the plants of B.media that were seen recently by the author at Toolinna. On the succeeding day, Eyre was travelling through the Point Culver sand dunes and he later wrote: “We moved through a country which gradu- ally became more scrubby, hilly and sandy .... The smaller Banksia \B. media] now abounded whilst Banksia grandis, and many other shrubs common at King George’s Sound, were frequently met with.” (Eyre 1847, vol. 2, p. 14) The observation that the country became more sandy is exactly what would would be expected in that region; the limestone areas giving way to the dune systems. The increasing prevalence of Western Australian plant species coincides with the situation at Point Culver. Eyre’s reference to Banksia grandis is incorrect as that species is not commonly found east of King George Sound; the species encountered must have been B.speciosa. The most easterly sight- ing of Banksia media is marked on the map published with Eyre’s narrative though the posi- tion is somewhat inaccurate (Eyre 1847). Adenanthos species are not as conspicuous as Banksia, but they form an important component of the quonkan vegetation west of Israelite Bay. A.cuneata has been collect :d at Israelite Bay, Twilight Cove and Mount Ragged (Nelson, in press) and most recently at Toolinna. Brooks collected the species in the late nineteenth century at Israelite Bay, and Batt and Carey collected it at Twilight Cove in 1889 (MEL). The species is abundant at Israelite Bay, but at Toolinna and Twilight Cove it is very rare. Brooker did not collect the species at Point Culver. It is probable that it occurs there today, but it may be rare. Adenanthos forrestii was first collected by John Forrest (Mueller 1882) on his survey in 1870. Two specimens from his journey are extant (MEL) ; one is labelled “near Point Culver 33° 14' S., 124° 2' E.” This location is named ‘Wattle Camp’ on standard topographic maps and is described by Brooks (1894) as an area of dunes consisting of “vary fine sand that has a yellow tint”. The second specimen came from the Twilight Cove area. Thus the species was only known from these two areas before its recent collection at Toolinna. It has not been collected at Point Culver, but it could occur there. Forrest’s location “near Point Culver” has been misinterpreted in the past and should be regarded now as referring to Wattle Camp, about 36 km south of Point Culver. The species is endemic to these sand patches, and to the Israelite Plain; it is not recorded elsewhere. A third species of Adenanthos* was collected at Toolinna in October 1973. It is apparently endemic to that sand patch. It could not be found at Twilight Cove, and no species of Adenanthos were collected at Point Culver by Brooker. The possibility that it occurs in these latter areas cannot be ruled out. * This species has not yet been described. A descrip- tion is to be published soon by E. C. Nelson in a tanonomic revision of Adencnihos. Journal of the Royal Society of Western Australia, Vol. 57, Part 4, December, 1974. 112 Among the other members of the family, Isopogon trilobus and Petrophile teretifolia have been collected at Point Culver, and the former was recently found at Twilight Cove also. Neither species has been collected at Toolinna. Synaphea cf. polymorpha was collected at Toolinna and at Twilight Cove. Several species of Hakea, including H.corymbosa and H.nitida were col- lected at Toolinna and at Point Culver. A number of Grevillea species grow at Point Culver, but they have not been recorded on the more east- erly sand patches. Most of these species are found in the Israelite Bay and Mount Ragged areas, and like the species of Adenanthos and Banksia they are restricted to non-calcareous soils. None of these Proteaceae is found on the dunes near the beach at Twilight Cove which contain small amounts of calcium carbonate. The flora of the dunes near the sea, that have some calcium carbonate incorporated, is very different and consists mostly of littoral plants, with occasional Eucalyptus and Melalauca species. (c) Myrtaceae Apart from some significant species of Eucalyptus that Parsons (1970) has already discussed, the most significant members of the family that have been recorded on these siliceous dunes are Chamelaucium axillare, Darwinia vestita and D.diosmoides, Calothamnus gracilis, V erticordia brownii and V.plumosa. Chamelaucium axillare occurs in the quonkan between Esperance and Israelite Bay. It has not been collected east of Toolinna, where it was found recently. The genus is endemic to south- western Australia, and is restricted to sand plain areas in that region. Calothamnus gracilis is listed by Beard (1970) as occurring in the Eremean Province, but it is usually a sand-plain species (Erickson et al. 1973) Darwinia and Verticordia species fall into similar categories. They are calcifuge species, and while the genera are found in South Australia these species are south-western endemics (Black 1922, Eichler 1965, Beard 1970). (d) Goodeniaceae Lechenaultia species are frequent in the quonkan west of Mount Ragged. L.formosa and L.tubiflora have been collected as far east as Twilight Cove, but they are both infrequent on the sand patches. These species are very com- mon near Israelite Bay and are endemic to Western Australia. Some other members of the family recorded on or near the sand patch are calcicoles such as Scaevola crassifolia (Parson 1970, N. M. Wace, pers. comm.). (e) Other families Other genera recorded on the sand patches but not in the surrounding limestone country include Pimelea, Comesperma and Stylidium. Several different species of Pimelea have been recorded on the sand patches, and representa- tives of the genus are frequent on the younger, coastal sand dunes near Israelite Bay, particul- arly Pimelea ferruginea. Stylidium pilosum was collected by Brooker at Point Culver, and it is known from the area between Esperance and Israelite Bay. Comesperma polygaloides was collected at Twilight Cove as was Stackhousia scoparia ; both species found west of Mount Ragged in siliceous sand associated with granite monadnocks (Beard 1970). The species listed in Table 2 that are restricted to siliceous sand habitats are all endemic to Western Australia. The species that are found either exclusively on limestone or on both silice- ous sand and calcareous soils usually occur in south-eastern Australia and in south-western Australia. Thus the flora of these sand patches is composed mainly of Western Australian cal- cifuge species. Some of the species had not previously been noted in areas east of Israelite Bay and most of them are absent from the limestone plateau of the Nullarbor Plain. The total species numbers in the sand-patches north-east of Point Culver are very low compared with similar areas west of Israelite Bay; the flora of the sand patches is therefore a depauperate form of the quonkan flora typical of the Esperance-Israelite Bay region. However the presence of some of the species at Point Culver indicates that many of the plants may be found on the Israelite Plain, but due to the lack of collections from that area they were not recorded east of Israelite Bay. The flora of the Eyre zone can be said to extend as far as Point Culver, and not to end abruptly at Israelite Bay. It is interesting to note that cer- tain birds which occur in south-western Aus- tralia but which previously had not been recorded north-east of Israelite Bay, have recently been seen in the sand-dune at Point Culver, Toolinna and Twilight Cove (See Appendix 1). There is a gap of about 25 km between Tool- inna sand patch and the major dune system at Point Culver. The Point Culver cliff-top dunes are linked to the Israelite Plains by ramps of sand built up by wind against the Wylie Scarp. Similarly between the Toolinna dunes and the second major system at Twilight Cove there is a gap of about 85 km, which is devoid of deposits of siliceous sand apart from the very small sand patch south-west of Point Dover. The areas between the sand dunes are either bare exposed limestone or have clay soils derived from lime- stone. These intervening areas present no suit- able habitats for calcifuge species. Discussion These isolated siliceous dune systems are col- onised by species which are endemic to Western Australia, and which are considered to be intol- erant of soils containing high proportions of calcium. The sand patches are surrounded by an environment which these species are unable to colonise due to the presence of limestone. The Journal of the Royal Society of Western Australia, Vol. 57, Part 4, December. 1974. 113 climate, inland from the coast, is apparently unsuitable for the growth of many of these plants; it is characterised by high summer tem- peratures and low intermittent rainfall. There are several hypotheses that can be ad- vanced to explain the presence of these species on the sand patches. Firstly, long range disper- sal can be considered; plants may have been transported from the areas to the west over the intervening unsuitable habitats by an agency such as wind or animals. Secondly, migration from the west could have occurred when the conditions, both climatic and geomorphic, were suitable in the past. Finally the species could have had a continuous distribution along the coast, but this has since been fragmented (Par- sons 1970). Long range dispersal of plants has been dis- cussed many times. A propagule can probably be dispersed over long distances even if it is not apparently adapted for dispersal by a specific agency such as sea currents, wind or birds (Good 1947). However long range dispersal is only successful if the propagule reaches a habitat that is suitable for its germination and the survival and subsequent reproduction of the plant (Rao 1971). The species that are found on the dune systems east of Point Culver are not obviously adapted for long range dispersal. Adenanthos fruits do not appear to be adapted in any way for dispersal over long distances. It is usually difficult to find seeds of Adenanthos more than a few metres from a parent plant as the progagules tend to drop directly to the ground on maturity. Banksia species also are not adapted for long range dispersal and the same probably applies to the other genera of the Proteaceae and other families. Rao (1971) argued against ocean dispersal of the Proteaceae as the propagules have flimsy coats and are thus susceptible to toxic effects of sea water. Simil- arly bird dispersal is more probable if species have drupaceous fruits, which none of the species considered here possess (Rao 1971). It should be noted that many of these species do not have seeds which will germinate readily. In the quonkan seedlings of Adenanthos species are very rarely found in areas where there are mature plants established. Very occasionally seedlings are encountered in disturbed ground but only if a major disturbance has occurred, such as “bulldozing” or clearing. Seedlings will be found immediately after a fire has burnt the vegetation in an area. This applies equally to other members of the Proteaceae; Banksia seeds are usually not released from the fruiting cones until after a fire (Gardner 1959) and they fall directly to the ground. It is unusual to find seedlings among mature communities of quonkan vegetation; all plants are apparently of the same age as they probably all germinated after the last fire in the area. In the Proteaceae microbial stimulation may be necessary, as seeds will not readily germinate in sterile soils (Rao 1971, Vogts 1960). When these characteristics are considered, along with the lack of adapta- tion to dispersal over long distances, long range dispersal must be discounted. The extreme sen- sitivity of many south-western Australian species to changes in edaphic conditions (Diels 1906, Speck 1958) and the very restricted dis- tribution ranges of some species argues strongly in favour of this lack of dispersal capacity. The second and third hypotheses differ only in their starting points. Parsons (1970) has suggested that disjunctions in the distributions of Eucalyptus species can be accounted for simply by postulating a continuous strip of siliceous sand linking the Roe Plains and the areas to the west, during the Quaternary low sea levels. “This would provide continuous species distributions which were subsequently fragmented by rising sea levels” (Parsons 1970). In assessing this idea it is important to remember that it seems that the species con- cerned cannot migrate rapidly even over short distances. Thus while there certainly would have been vegetation cover on the exposed coastal lowlands between the Israelite Plains and Twilight Cove, that vegetation is unlikely to have contained all the species typical of the quonkan. Rather, only a small proportion of the quonkan species may have colonised that coastal area, due to differential migration rates and climatic and edaphic barriers. In this context it is significant that certain species are not recorded at Twilight Cove, but have been collected in areas to the south-west. A similar situation relates to the Toolinna sand patch. Thus Banksia speciosa and Anigozan- thos rufa are recorded only as far east as Point Culver; they could have reached that dune system by migrating along the existing Israe- lite Plain. Banksia media and Chamelaucium axillare are found at Toolinna but not at Twilight Cove. Despite the incomplete collec- tions it is noticeable that the floras of each sand patch becomes poorer in species typical of the quonkan the further east they are situated. The Point Culver flora is not as rich in genera and species as that of the Israelite Bay area, and the Toolinna flora is depauperate compared with the Point Culver flora. This suggests that these species have different capacities to migrate, or if they migrate at the same rate that some species have become ex- tinct from the more easterly dune systems. The latter hypothesis is probably partly correct; due to changes in climate some species may have failed to survive. However, both Frank (1971) and Martin (1973) have indicated that there have not been substantial climatic changes during the late Quaternary in this area. While changes have occurred that have caused vegeta- tion patterns to alter, the climate in the late Pleistocene and Holocene was not very much drier, or wetter, than the climate of the present time (Martin 1973). Thus the climatic effects on the vegetation of these sand dunes may have been small. If different rates of migration are involved, those species that could migrate relatively rapidly reached Twilight Cove, while Journal of the Royal Society of Western Australia, Vol. 57, Part 4, December, 1974. 114 those that were slower moving along the exposed coastal plain only reached Toolinna, during the period when sea levels were low. When marine transgression occurred any plants that had reached Twilight Cove were isolated there and those which were at Toolinna became separated from plants on the other dune system. As mentioned above, the cliff-top dunes were probably emplaced by wind-built ramps situated against the cliffs. The plants would have moved from the exposed coastal plain up these ramps on to the dune systems. Marine erosion has removed these basal ramps, since the sea levels returned to the present datum. The absence of endemic species on the sand patches, with the probable exception of species of Synaphea and Adenanthos, indicates that there has been insufficient time for speciation to occur in these isolated populations. The forms of the species that are found on the sand patches are not perceptibly different from phenotypes at Israelite Bay or Mount Ragged. Thus it is probable that the populations at Twilight Cove and Toolinna have only been isolated since the last glacial maximum, and possibly only since the sea levels reached the base of the Baxter Cliffs between 10 000 and 6 000 years B.P. Thom and Chappell 1975; Jennings 1971). It is possible that species which are not potentially very variable, could have been in these areas for longer periods, if the dunes are older than Holocene or late Pleistocene. In the case of Adenanthos, the genus is morphologically variable and shows a very substantial amount of phenotypic plasticity, yet forms of the species found on these dunes do not differ significantly from forms found in other areas where those species occur. It is possible to conclude that the present cliff - top dune systems were probably emplaced in the Holocene, or late Pleistocene, and that the species distribution patterns found in this area have resulted from migrations along a coastal plain during the last period of low sea levels. Subsequent return of sea level to present datum fragmented distributions and isolated species on dune systems disjunct from their main areas of occurrence. Long range dispersal is not thought to account for these disjunctions due to the peculiar propensities of the south-western Aus- tralian flora and to the lack of adaptations of their propagules to dispersal over long distances. One feature of southern Australian phyto- geography is the marked difference between the floras of south-eastern and south-western Aus- tralia. Several workers have discussed this situa- tion and concluded that species similarities can be accounted for by migrations in post-Tertiary times (Burbidge 1960, Green 1965, Marchant 1973). The Nullarbor Plain limestone plateau is considered to present an edaphic barrier to east- west migrations due to the apparent calcifuge propensity of the Australian vegetation (Crocker and Wood 1947). The flora of south-western Australia exhibits a very high rate of endemism at species level (Beard 1969) and only a very few species that are intolerant of lime-rich environments are found on both sides of the Great Australian Bight. That only western calcifuge species have reached Twilight Cove is important. Those species that are found both in South Australia and at Twilight Cove (Table 2) are not typically restricted to siliceous sand plain; an example of this category of species is Eucalyptus diversi- folia (Parsons 1969). It would be expected that an exposed lowland plain, south of the Nullarbor Plain on the continental shelf, would permit migration in either direction; to the west or to the east. The absence of eastern calcifuge species demonstrates that either time was not sufficient for such a migration from the east to occur, or that there was a barrier either edaphic or climatic. Migration of these species requires that the exposed plain be composed of siliceous sands. Lowry (1970) noted that between Israelite Bay and Eucla (Fig. 1), the percentage of calcium carbonate in beach sand increased from 5 per cent in the west to about 50 per cent in the east. This may have been the situation in the past, and the higher calcium carbonate levels in eastern areas would have produced an edaphic barrier (Parsons 1970). Further, if calcifuge species were to migrate from the east leaching of the calcareous sands in the east would have to have taken place. A very long time could elapse before a non-calcareous substrate was available for colonisation and migration. One small area of calcareous sand would constitute a very effective barrier to migration. Such a situation would not have prevented calcicole species from migrating from east to west, or vice versa, as noted by Parsons (1969). While the prevailing direction of dispersing agencies such as winds and currents, are towards the east, they probably would have had little effect on species migrations unless the species were dispersed by these agencies. As indicated above this is considered unlikely. With only small climatic changes dur- ing the Pleistocene period it would appear un- likely that there was a climatic barrier. The evidence presented here suggests that the duration of periods suitable for migration in the Late Quaternary, or at least during the existence of the extant cliff-top dunes, was insufficient for calcifuge species to migrate across the Great Australian Bight in either direction. Species similarities between the south-eastern and south-western regions of Australia among groups of taxa that have similar propensities to those discussed here, may have resulted from migra- tions across the exposed continental shelf dur- ing the Quaternary. Long range dispersal does not provide a reasonable alternative. If the species disjunctions noted on these cliff-top dunes have resulted from migrations during marine regression in the late Pleistocene, a suit- able period for east-west interchange must pre- date the last period of low sea-levels (120 000- 6 000 years B.P.; Chappell 1974). Journal of the Royal Society cf Western Australia, Vol. 57, Part 4, December, 1974. 115 Acknowledgements.— 1 wish to thank Dr. Bob Bar- sons (Latrobe University), Dr. John Beard (Perth) and Michael Brooker (C.S.I.R.O., Division of Wild- life Research, Canberra) for providing me with species lists of collections that they made and for their use- ful comments. Dr. Joe Jennings (Department of Bio- geography and Geomorphology, A.N.U.) discussed the geomorphological situations and commented on the manuscript. My thanks are also due to my supervisor, Dr. Nigel Wace, for his criticisms and suggestions. I am also indebted to several people who assisted me to visit the areas, by providing information on tracks and accessibility of certain features. Mrs. A. E. Crocker (Balladonia) corrected local names of aboriginal origin and provided interesting information on the early settlement of this area. Identification of some of my collections was carried out by the staff of the Western Australian Herbarium, Perth. This work was carried out while the author held a post-graduate research scholarship at the Austra- lian National University, Canberra, and was part of studies for the degree of doctor of philosophy. References Beard, J. S. (1969). — Endemism in the Western Aus- tralian flora at species level: J . Roy. Soc W. Aust., 52: 18-20. (1970). — A descriptive catalogue of West Australian Plants; 2nd edition S.G.A P Perth. (1973).— The vegetation of the Esperance and Malcolm areas, Western Australia; Vegetation Survey of Western Australia- Vegmaps, Perth. Black, J. M. (1922). — Flora of South Australia; 4 pts Govt. Printer, Adelaide. Brooks, J. P. (1894). — Natural features of Israelite Bay; Proc. Aust. Assoc. Adv. Sc., 6: 561-569. Burbidge, N. T. (1960). — The Phytogeography of the Australian region; Aust. J. Bot., 8: 75-212. Chappell, J. (1974). — Geology of coral terraces, Huon Peninsula, New Guinea: A Study of Quaternary Tectonic Movement and Sea- Level Changes; Geol. Soc. Amer. Bull 85 553-570. Crocker, R. L., Wood, J. G., (1947). — Some historical influences on the development of South Australian vegetation communities and their bearing on concepts and classifica- tions in ecology; Trans. Roy. Soc. S. Aust 71: 91-136. Delisser, E. A. (1867). — Survey of a new port on the Great Australian Bight; in South Austra- lian Parliamentary Papers (no. 137 of 1867). Diels, L. (1906). — Die Pflanzenwelt von West Austra- lien; in Die Vegetation der Erde, 7. Engle- mann, Leipzig. Dunkley, J. R. (1967).— Caves of the Nullarbor; Syd- ney Speleol. Research Council. Eichler, Hj. (1965) .—Supplement to J. M. Black’s Flora of South Australia; Govt. Printer, Adelaide. Erickson, R. (1969). — The Drummonds of Hawthorn- den; Lamb Paterson, Perth. Erickson, R., George, A. S., Marchant, N. G., Mor- combe, M. K., (1973). — The Flowers and Plants of Western Australia; Reed, Syd- ney. Eyre, E. J. (1847). — Journal of Expeditions into central Australia and Overland from Adelaide to King George’s Sound in the years 1840-1: Libraries Board of S. Australia reprint (1964). Forrest, J. (1875). — Explorations in Australia; from Perth to Adelaide; Libraries Board of S. Australia reprint (1967). Frank, R. (1971). — Cave sediments as palaeoenviron- mental indicators and the sedimentary sequence in Koonalda Cave; in Aboriginal Man and Environment in Australia, (ed. Mulvaney, D. J., Golson, J.,) Australian National University, Canberra. Gardner, C. A. (1959). — The vegetation of Western Australia; Monographiae Biologicae, 8 - 274- 282. (1965).— in Toxic plants of Western Aus- tralia; W.A. Newspapers, Perth. (1973 ).— Wild flowers of Western Austra- lia; 11th edition, W.A. Newspapers, Perth. Good, R. (1947). — The geography of flowering plants; Longmans, London. Green, J. W. (1965) .—Discontinuous and presumed vicarious plant species in southern Aus- tralia; J . Roy. Soc. W. Aust., 47: 25-32. Jennings, J. N. (1963). — Some geomorphical problems of the Nullarbor Plain; Trans. Roy. Soc. S. Aust., 87: 41-62. (1967). — Cliff-top dunes; Aust. Geog. Studies, 5: 40-49. (1971. — Sea level changes and land links; in Aboriginal Man and Environment in Australia, (ed. Mulvaney, D. J., Golson, J.), Australian National University, Canberra. Lowry, D. C. (1970). — Geology of the Western Austra- lian part of the Eucla Basin; Geological Survey of W. Australia bulletin no. 122, Perth. Lowry, D. C., Jennings, J. N. (1974).— The Nullarbor karst, Australia; Zeit. Geomorpli. 18: 35-81. Marchant, N. G. (1973). — Species diversity in the southwestern flora; J. Roy. Soc. W. Aust., 56: 23-30. Martin, H. A. (1973). — Palynology and historical ecology of some cave excavations in the Australian Nullarbor; Aust. J. Bot., 21; 283-316. Mueller, F. von (1882). — Definition of some new Austra- lian plants; Wing’s Southern Science Re- cord 2 (10): 230. Nelson, E. C. (in press). — The collectors and type locations of some of Labillardiere’s “terra van-Leuwin” (Western Australia) speci- mens; Taxon. Parsons, R. F. (1969). — Distribution and palaeogeo- graphy of two mallee species of Eucalyptus in southern Australia; Aust. J. Bot., 17: 323-330. (1970). — Mallee vegetation of the southern Nullarbor and Roe Plains, Australia; Trans. Roy. Soc. S. Aust., 94: 227-242. Rao, C. V. (1971). — Proteaceae; C.S.I.R., New Delhi. Seddon, G. (1972). — Sense of Place-, University of W. Australia, Perth. Speck, N. H. (1958). — The vegetation of the Darling/ Irwin districts, South-west Australia; un- publ. Ph.D. thesis. University of W. Aus- tralia. Thom, B. G., Chappell, J. (1975). — Holocene sea levels relative to Australia; Search (in press). Vogts, M. (1960). — The South African Proteaceae; the need for more research; S. Africa J., Sci., 56: 297-305. Willis, J. H. (1959). — Notes on the vegetation of the liiucla district, W.A.; Muelleria 1: 92-96 (see also note on Scaevola brooksiana F. Muell., ibid : 91). Appendix 1: Distribution of birds on south-western Nullarbor Plain Reilly, Brooker and Johnstone (in press) have recorded the occurrence of species of birds along the coast of the Great Australian Bight, between Point Culver and Twilight Cove. Nine honeyeaters ( M eliphagidae) were noted, of which two ( Phylidonyris novaehollandiae — New Holland honey eater; GlicipKUa melanops — tawny-crowned honeyeater) were recorded east of their previously known limits of distribution in Western Australia. These two species are Journal of the Royal Society of Western Australia, Vol. 57, Part 4, December, 1974. 116 members of the Bassian avifauna of south- western Australia, and were observed only in the vegetation on the siliceous sand deposits. They are probably dependent for food on plants that copiously produce nectar and which are also restricted to those areas, such as species of Adenanthos, BanJcsia, Grevillea and Anigozan- thos (Reilly et al. in press). Adenanthos spp., Anigozanthos rufa, Gre- villea spp., and certain other plants are pollin- ated by nectar-seeking birds. As these plants are found in soils with a very low calcium car- bonate content, they probably do not occur east of Twilight Cove and on the limestone areas due to the increased proportions of calcium carbon- ate in the soils and sand. Survival of populations of these species requires the presence of suitable pollinators. The presence of the honeyeaters in the sand patch vegetation is presumably import- ant with respect to the survival of these disjunct species populations. The absence of certain honeyeaters from the areas east of Twilight Cove and from limestone areas (Reilly et al. in press) may be due to the absence of populations of nectar-producing plants on which those honeyeaters may be dependent. Reference: Reilly, P. N., Brooker, N. G., John- stone, G. W., (in press) : Birds of the south- western Nullarbor Plain: Emu. Journal of the Royal Society of Western Australia, Vol. 57, Part 4, December, 1974. 117 15- — The development of premolar and molar crowns of Antechinus flavipes (Marsupialia, Dasyuridae) and the significance of cusp ontogeny in mammalian teeth by M. Archer 1 * Manvzcript received 23 April 1974; accepted 17 September 1974 Abstract A series of fourteen transversely sectioned juvenile Antechinus flavipes was studied to reveal information about the ontogenetic devel- opment of the cheek-teeth. Only three pre- molar tooth family positions were found. There was no evidence for milk-teeth posterior to the canine. In the upper molars, the first cusp to initia’e and calcify was the metacone, followed by the paracone. Stylar cusp D developed and calcified before stylar cusp B. In the lower molars, the protoconid developed and calcified first. The metaconid was gen- erally second. In the Mi the hypoconid devel- oped and calcified before the paraconid. Reasons are given for believing that ontogeny of tooth cusps does not necessarily indicate phylogeny, and that ontogeny may be at least partly dependent on the size of the cusps of the adult crown. Introduction The order in which the cusps develop in molar teeth of mammalia is generally regarded as an indication of the identity of the cusps and in particular of the identity of the paracone and the protoconid. In recent years B. K. B. Berkovitz has shown that in some marsupials (e.g. Didelphis and Setonix) there must be some doubt as to the general applicability of using ontogeny in this manner. Here, a study of the developing cusps of the cheek-teeth of Antechinus flavipes is made in order to discover the situation in an Australian marsupial which has a structurally primitive dentition. A series of heads of fourteen juvenile Ante- chinus flavipes of known ages were sectioned transversely. The preparation of specimens us r d in this study is described by Archer (1974). The sixteen developmental stages examined ranged from the 4 Day Stage (i.e. 4 days postbirth) to the 105 Day Stage and included a 4 Day, 10 Day, 12 Day, 15 Day, “22” Day (actually developmentally younger than the next stage), 21 Day, 28 Day, 32 Day, 36 Day, 40 Day, 44 Day, 51 Day, 59 Day, 60+ Day, 83 Day, and 105 Day Stage. The 83 Day Stage and the 105 Day Stage animals were gross skeletal preparations and are registered in the Western Australian Museum mammal col- lections as M 8091 and M 8092 re peetively. 1 Western Australian Museum, Francis Street. Perth. Present address: Queensland Museum, Fortitude Valley, Queensland 4006. Cheek-tooth nomenclature follows that of Thomas (1888) bearing in mind that Archer (1974) has established that dP4 is not a milk- tooth and is the first of the dP4-M4 Zahnreihe. Basic cusp nomenclature is that used by Ben -ley (1903) and Simpson (1936), with modifications (Fig. 1) as presented in Archer (1975). Results The canine The upper canine was slightly more advanced in development in any given stage than the lower canine. Enamel knots (Fig. 2) were present in both teeth in the position of the future paracone and protoconid of both crowns. The canines, unlike any of the other cheek- teeth, both had rudimentary uncalcified milk predecessors which were resorbed soon after development. PI From its initiation Pi was in advance of P l in development; it also calcified one stage earlier. Both teeth however, were in the same state of development in later stages and in the 60+ Day Stage both had well-formed roots and were near- ing eruption. Enamel knots were not observed. P3 The P.i was advanced in development over P 3 in all stages observed, although calcification began at approximately the same time in the 40 Day Stage. Enamel knots were not ob- served. P4 The P 4 and P 4 developed at about the same time, in the 40 Day Stage. Development was very late and calcification was not observed in the sectioned material. dP4 DP i initiation had begun by the 4 Day Stage. DP 4 was slower in development, initiation occurring in the 10 Day Stage. In later stages however the d.P 4 was advanced in development over the dPi. An enamel knot was observed in both teeth above the area of the presumptive paracone and protoconid, the only cusps that developed on the crowns of these teeth (Fig. 3). Journal of the Royal Society of Western Australia, Vol. 57, Part 4, December, 1974. 118 A M2 Ml Figure 1.— The morphology of the cheek-teeth. 1A. RCi-M-*. IB. RM 2 . 1C. RC 1 -M 4 . ID. RM ;i . Abbreviations: a.c., anterior cingulum; a. prcr., anterior protocrista; c.o., crista obliqua; end, entoconid; hycd, hypocris- tid; hyd, hypoconid; hyld, hypoconulid; me., metacone; mec., metacrista; meed, metacristid; vied, meta- conid; mst., metastylar corner of tooth; pa., paracone; pac., paracrista; pacd, paracristid; pad, paraconid; p.c., posterior cingulum; p.prer., postprotocrista; pr., protocone; prd., protoconid; prgd, precingulid or anterior cingu- lum; psgd, postcingulid or posterior cingulum; pstd, parastylid; stA, stylar cusp A; stB. stylar cusp B; stD, stylar cusp D; stE, stylar cusp E. Journal of the Royal Society of Western Australia, Vol. 57, Part 4, December, 1974. 119 Figure 2. — The upper canine at the 15 Day Stage. Ab- breviations; C 1 , upper canine; dC 1 , rudiment of deci- duous canine tooth germ; oe, oral epithelium; dl, den- tal lamina; sr, stellate reticulum; dp, dental papilla; ek, enamel knot; oc, oral cavity; e, enamel; d dentine x 150. Figure 3. — The upper milk premolar at the 10 Day Stage. Abbreviations as in Fig. 2. x 150. There were problematical structures associated with the dP 4 and dP 4 in the 60+ Day Stage animal (Fig. 4). These were concentrically laminated epithelial structures developed in the oral epithelium and the enamel epithelium (which at this stage consisted of the inner and outer enamel epithelial layers pressed together) at the tips of the nearly erupting dP4. A similar but smaller structure was observed above the lingual cingulum of the dP 4 . These structures may be Pearls of Serres (see Scott & Symons 1961). The dental lamina had completely dege- nerated from the region of the dP4 in previous stages. These Pearls of Serres were not observed in association with any other teeth but very few teeth were examined at this relatively late stage of development. Figure 4. — The upper and lower milk premolar at the 60 -f- Day Stage. vs, tissues resembling Pearls of Serres. Other abbreviations as in Fig. 2 x 75. Ml Mi development preceded that of M 1 . The first three cusps visible in the 32 Day Stage in M 1 were the metacone, stylar cusp D and the paracone. The metacone was the largest and had begun calcification. Calcification may have just begun at the tip of the paracone in this Journal of the Royal Society cf Western Australia, Vol. 57, Part 4, December, 1974. 120 stage, but clearly had not proceeded as far as it had on the metacone. In the 36 Day Stage stylar cusp D was large but had not begun calcification and the protocone was only begin- ning to form. The paracone had clearly become calcified. In the 40 Day Stage stylar cusp D was just begin- ning calcification (Fig. 5). Stylar cusp B was developed but not calcified. In the 44 Day Stage the protocone was undergoing calcification (Fig. 6). Stylar cusps B and A did not undergo calcification until the 51 Day Stage. Figure 5. — The upper first molar, section through the metacone and stylar cusp D, at the 40 Day Stage. Both cusps are undergoing calcification. At this stage only the metacone, stylar cusp D and the paracone were calcified, me, metacone; st.D, stylar cusp D. Other abbreviations as in Fig. 2 x 75. Figure 6. — The upper first molar, section through the protocone, paracone and stylar cusp A at the 44 Day Stage. Only the paracone is shown calcified but the beginning calcification of the protocone is apparent in sections posterior to the section shown. At this stage the metacone and stylar cusp D are also calci- fied. pr, protocone; pa, paracone; st.A, stylar cusp A. Other abbreviations as in Fig. 2. x 75. In M 1 the protoconid was the first cusp to become visible in the 15 Day Stage as well as to calcify in the 21 Day Stage. The hypoconid and metaconid were di tinguishable but uncalcified in the 36 Day Stage. All of these cusps had begun calcification by the 40 Day Stage but the hypoconid was the least developed. The entoconid may have been distinguishable in the 44 Day Stage. By the 51 Day Stage the para- conid had begun calcification. The entoconid and hypoconulid may have just begun calcifica- tion. M2 In M 2 the metacone was distinguishable in the 40 Day Stage. By the 44 Day Stage the paracone, protocone, stylar cusp D and possibly stylar cusp B had also appeared but only the metacone had clearly become calcified (Fig. 7a). The paracone may have just begun calcification (Fig. 7b). The paracone had clearly begun cal- cification by the 51 Day Stage and all major cusps had become calcified by the 59 Day Stage. Figure 7. — The upper second molar at the 44 Day Stage. 7A. Section through the well-calcified meta- cone showing enamel as well as dentine, and the pos- teriorly extended dental lamina. 7B. Section through the protocone, paracone and stylar cusp B showing the slight (if any) calcification of the paracone. There appears to be an enamel cord above the paracone tip. ec, enamel cord. Other abbreviations as in Figs. 2, 5 and 6. x 150. In M 2 in the 36 Day Stage, the protoconid, metaconid, paraconid and possibly the hypoconid were present but only the protoconid had begun calcification. In the next stage the metaconid had calcified. The entoconid appears to be dis- tinguishable by the 44 Day Stage. The paraconid did not begin calcification until the 51 Day Stage. The entoconid was calcified by the 59 Day Stage. Journal of the Royal Society of Western Australia, Vol. 57, Part 4, December, 1974. 121 M3 The metacone of M 3 was first distinguishable in the 59 Day Stage. It was uncalcified. An enamel cord was observed over the tip of the cusp (Fig. 8). By the 60+ Day Stage, the paracone, metacone and protocone were distinct but only the first two had undergone calcifica- tion. Figure 8. — The upper third molar at the 59 Day Stage showing an enamel cord over the tip of the metacone. Abbreviations as in Fig. 7. x 75. In M 3 the metaconid and protoconid were present in the 44 Day Stage but only the latter was calcified. The paraconid was distinguish- able in the following stage. By the 59 Day Stage all cusps were present and all were calcified excspt the entoconid and hypoconulid. By the 60+ Day Stage the hypoconulid had begun calcification. M4 The M 4 in the 60+ Day Stage was merely a swelling of the dental lamina. The M 4 on the other hand had all three trigonid cusps formed and calcified in the 60+ Day Stage. Table 1 summarizes the data presented above. Discussion Cusp formation and homology In 1874 Cope proposed that the complex therian molar had developed through addition of cusps peripheral to a single cusp of a primitive tooth. This basic premise was in contrast to that of others such as Bolk (1922) in which the therian molar was seen as a result of fusion of adjacent single-cusped primitive teeth. Cope’s premise has generally been accepted (e.g. see Gregory 1934, Butler 1941, Crompton 1971). However, there has been disagreement about which cusp on the therian molar is the primitive cusp. Osborn (1888) believes it is the lingual cusp or protocone of the upper molar and the antero -buccal cusp or protoconid of the lower molar. This view has been accepted by many later authors (e.g. Gregory 1934, Simpson 1936). Winge (1941) however believes that the variably present external stylar cusps of the upper molars and the three lingual cusps, the paraconid, metaconid and entoconid of the lower molars are the oldest cusps and therefore the central stylar cusp and the metaconid represent the ancestral primitive cusps. Gidley (1906) suggests that the paracone of the upper molars and the protoconid of the lower molars were the first ancestral cusps. Gregory’s view has been sup- ported by Wortman (1902) and Butler (1937) who argue that on the basis of observations of upper premolars that when these teeth become progressively molarized posteriorly along the tooth row through evolution, they often produce secondary cusps serially homologous with the protocone and metacone, suggesting that the ancestral cusp is the paracone in upper molars as well as premolars. Embryological evidence of Rose (1892a), Taeker (1892), Woodward (1896), Kupfer ( 1935) , Marshall & Butler (1966), and Berkovitz (1968) indicates that in marsupi- als and eutherians the paracone and protoconid develop first ontogenetically and Butler (1956) concludes that this supports the view that these cusps are the ancestral molar cusps. However, Rose (1892b) and Berkovitz (1967a) have ob- served that in some molariform marsupial teeth examined, the metacone developed first. The results of Rose (1892b) and Berkovitz (1967a) appear to cast doubt on the concept that position in the ontogenetic sequence of cusp development can indicate the primary cusps. It is also possible, but not likely, that the cusp called a metacone in these marsupials is homo- logous with the paracone of other marsupials (e.g. Trichosurus and Setonix ) and mammals in general. It is also possible, as Butler (1956) sug- gests, that Rose (1892b) may not have used the same criteria as other authors to establish the state of development of a cusp but this is not likely because Berkovitz (1967a) has confirmed Rose’s (1892b) observations in a specimen of the same genus (Didelphis ) . Upper molars In the M 1-2 of Antechinus flavipes examined in the present study it is apparent that the meta- cone precedes all the other cusps in development and calcification. This may also be the situation in M 3 ’ 4 . The paracone is the next cusp to form and calcify, followed by the protocone. This developmental sequence is unlike any other re- ported except for that of Didelphis (Rose 1892b, Berkovitz 1967a). It also appears to contrast with the findings of Woodward (1896) who after examining one specimen of Antechinus sp. states (p. 284) that “The paracone above and the protoconid below develop before any of the other molar cusps”. However, some doubt about Woodward’s conclusion must stem from the fact that he examined only one specimen and, as indicated by the present study, this may not be sufficient. In view of the results reported in this paper there appears to be reason for believing that the paracone may not always pre- Journal of the Royal Society of Western Australia, Vol. 57, Part 4, December, 1974. 122 Table 1 Day Stage post-birth in which initiation (i) and calcification (c) of molar cusps was observed. UPPER M 1 M 2 M 3 M? i c i C i C i c protocone 36 44 44 59 60 + 83 ? ?105* paracone 32* ?32* 44 51 60 + 60* + 9 105* metacone 32* 32* 40* 44* 59* 60* + stylar cusp A 51 51 59 59 ? 105 stylar cusp B 40 51 ?44 59 ? 83 ? ?105* stylar cusp D 32* 40 44 59 9 ' 83 LOWER m t m. m 3 m 4 i c i c i c i c protoconid 15* 21* 36* 36* 44* 44* 60* + 60* + paraconid 51 51 36* 51 51 59 60* + 60* + metaconid .... 36 40 36* 40 44* 59 60* + 60* + hypoconid 36 40 ?36* ?44 59 59 ? 83 entoconid ?44 51 ?44 59 59 83 hypoconulid 51 51 9 ? 59 60+ | * first occurrence. cede other cusps in development. Since other marsupials examined, e.g. Setonix (Berkovitz 1967b) and Trichosurus (Berkovitz 1968) demonstrate that the paracone develops first, this could be interpreted as suggesting the marsupials are polyphyletic with regard to molar formation. An alternative explanation is that the order of cusp development may not be an invariable indicator of the order of cusp evolu- tion of the marsupials in which the metacone develops first. Both Didelvhis and Antechinus differ from other marsupials examined onto- genetically by having a metacone which is the largest cusp on the crown and certainly larger than the paracone (Fig. 1). In Setonix, Tricho- surus and most other phalangeroids, the meta- cone and paracone are subequal. In most other mammal species previously examined the mor- phology of the adult crowns (for examples see Butler 1956) shows that the paracone is sub- equal to or even larger than the metacone. Butler (1967) concludes, from research into the relative growth of the first upper permanent molar in Homo, that the antero-buccal areas (i.e. the area of the paracone) of the crown develop before the postero-lingual areas and interprets this as indicating the probable order of cusp origin in phylogeny. However, Butler (1956) had previously suggested (relying on the research of Canalis 1886 and others) that cusp initiation is the result of a cessation of mitoses at a point on the inner enamel epithelium, the remainder of the crown cusps being subsequently developed in a similar way, while active mitosis, which continues to occur in areas between the cusps, results in enamel deposition in the valleys. Accordingly the cusp destined to be the tallest on the completed crown would presumably be the first to be initiated, and the cusp destined to be the lowest on the crown would be the last to become initiated. Butler’s (1956) suggestion would therefore support the hypothesis that the phylogenetic sequence of development of the cusps in Didelvhis and Antechinus may be modi- fied in ontogeny by the great height of the defin- itive metacone. It is also possible that the rate of development of the cusps of a crown are not identical. If so, a primary cusp could develop first, but by its sur- rounding valley developing more slowly than that of a secondary cusp, appear to have devel- oped after the secondary cusp. For example as noted above, in the M 1 of Antechinus fiavipes, at the 32 Day Stage, the metacone was clearly cal- cified and better-developed than the paracone. Were this the only stage available, it would be open to interpretation that the paracone had been initiated first, but been slower in develop- ment than the metacone. That this is not the case (at least in M 2 ) is demonstrated in this Journal of the Royal Society of Western Australia, Vol. 57, Part 4, December, 1974. 123 series by the fact that the metacone was first distinguishable in the 40 Day Stage, while the paracone was not distinguishable until the 44 Day Stage. The development of the stylar cusps reveals a similar problem in interpreting phylogeny from ontogenetic sequence. In the M l stylar cusp D appears and calcifies prior to stylar cusp B. Many Mesozoic fossil therians have well de- veloped stylar cusps (e.g. Pappotherium) , but the basis for determining the homology of these cusps is dubious. Slaughter (1965), Clemens (1968 and 1971), Kermack, Kermack, & Mussett (1968), and others however all recognize the antiquity of stylar cusp B, often referred to as the stylocone. It is identified topographically as the buccal cusp generally connected by a crest to the paracone (Fig. 1). This cusp is present in Antechinus fiavipes. Stylar cusp D is posterior to but much larger than stylar cusp B. The relatively early development and calcification of stylar cusp D compared with stylar cusp B in A. fiavipes appears to be a situation directly comparable with that of the relatively early development and calcification of the metacone compared with the paracone. Lower molars In the lower molars of the series, the proto- conid invariably developed and calcified first. The metaconid was generally second. In the Mi the hypoconid developed and calcified before the paraconid. In M 2 -4 the paraconid seems to have developed and calcified in advance of the hypoconid. The hypoconulid and entoconid were usually the last to develop and calcify. In the Mi however the entoconid developed before the paraconid. As pointed out by Butler (1956) and sub- sequently shown by Berkovitz (1968), the pro- toconid of mammals appears to develop and calcify first (a possible exception in Setonix is described by Berkovitz 1967b). No exception would have been expected for Antechinus be- cause in that form the protoconid is the largest cusp of each molar (Fig. 1). Rose (1892b) and Berkovitz (1967a, 1968) demonstrate that the metaconid in Didelphis and Trichosurus is the second cusp to develop in the lower molars, as in Antechinus. Butler (1956) reviews other studies (e.g. Woodward 1896, on Setifer ) which show that in some mammals the paraconid is the second cusp to develop. The paraconid is vestigial or absent in Trichosurus. However in Didelphis (Rose 1892b) the paraconid develops late, as in the Mi of Antechinus. The relative size of the paraconid in M m of Antechinus may account for the differences in the rate of de- velopment between Mi and M 2 - 4 . In Mi the pa- raconid is very reduced in contrast to its condi- tion in M 2 _4. As noted above, it would be ex- pected to develop later in Mi. Butler (1956, p. 51) comments in reference to the varying time of development of the paraconid in mammals that “This variation in the time of appearance of the paraconid is in accordance with its relative size . . . Berkovitz (1967b) describes cusp formation in the macropodid marsupial Setonix. He shows that although the cusp he calls the protoconid of the M, develops first, a cusp he identifies as the metaconid on the dP 4 develops first. To ac- count for this he concludes that the larger size of the metaconid of the dP 4 is the reason it develops before the protoconid. Conclusion The results reported in this study for the lower as well as upper molars of Antechinus support the suggestion of Berkovitz (1967b) that the relative size of cusps is at least as important as phylogeny in determining the sequence of development during ontogeny. It may be that if the cusps are equal in height, ontogeny can reveal phylogeny. Since, however, in most living eutherians in which the paracone and metacone are unequal the paracone is larger, belief that early development of the paracone in these forms supports the contention that this cusp is the ancestral cusp is warranted but does not show conclusively that this is so because the reason for prior development may be cusp size alone. Didelphis and Antechinus represent a marsupial lineage in which there was selection for a larger metacone. This innovation probably imposed a practical need in these forms to have the larger metacone develop in advance of the smaller paracone. Acknowledgments . — Research on this project was carried out under a Fulbright Scholarship, a grant in aid from the American Explorer’s Club and as a Research Assistant to Dr W. D. L. Ride (Western Aus- tralian Museum), who was in receipt of a Research Grant from the Australian Research Grants Com- mittee. Mr G. Burns (University of Western Austra- lia) helped extensively in sectioning, Mrs E. Archer helped in raising the colony of Antechinus, and Dr W. D. L. Ride read and constructively criticised the manuscript. Mr Hollywood (University of Queens- land) helped with the photomicrography. References Archer, M. (1974a). — The development of the cheek- teeth in Antechinus fiavipes (Marsupialia, Dasyuridae). J. Roy. Soc. West. Aust., in press. ■ (1975). — Ningaui, a new genus of tiny dasyurids (Marsupialia) and two new species from arid Western Australia, N. timealeyi and N. ridei. Mem. Qd. Mus., in press. Bensley, B. A. (1903). — On the evolution of the Aus- tralian Marsupialia; with remarks on the relationships of the marsupials in gen- eral. Trans. Linn. Soc. Lond. (2nd Series), Zool., 9: 83-217. Berkovitz, B. K. B. (1967a). — The dentition of a 25- day pouch young specimen of Didelphis virginiana (Didelphidae: Marsupialia). Archs oral Biol., 12: 1211-1212. (1967b). — The order of cusp development on the molar teeth of Setonix brachyurus (Macropodidae: Marsupialia). J. Roy. Soc. West. Aust., 50: 41-48. (1968). — Some stages in the early develop- ment of the post-incisor dentition of Tri- chosurus vulpecula (Phalaneeroidea : Marsupialia). J. Zool. Lond. (1968). 154: 403-414. Bolk, L. (1922). — Odontological essays No. 4. Journ. Anat., 56: 107-136. Journal of the Royal Society of Western Australia, Vol. 57, Part 4, December, 1974. 124 Butler, P. M. (1937).— Studies of the mammalian den- tition.— I. The teeth of Centetes ecaudatus and its allies. Proc. Zool. Soc. Lond., (ser. B), 107: 103-132. (1941).— A theory of the evolution of mammalian molar teeth. Am. J. Sci., 239: 421-450. (1956).— The ontogeny of molar pattern. Biol. Rev., 31: 30-70. (1967). — The prenatal development of the human first upper permanent molar. Archs oral Biol., 12: 551-563. Clemens, W. A. (1968) .—Origin and early evolution of marsupials. Evolution, 22: 1-18. (1971). — “Mammalian evolution in the Cretaceous”, pp. 165-180 in Early Mam- mals, ed. by D. M. Kermack and K. A. Kermack. Supplement No. 1 to the J. Linn. Soc. (Zool.), 50. Cope, E. D. (1874).— On the homologies and origin of the types of molar teeth of Mammalia Edu- cabilia. J. Acad. nat. Sci. Philad., 8: 71-89. Crompton, A. W. (1971).— “The origin of the tribo- sphenic molar”, pp. 65-87 in Early Mam- mals ed. by D. M. Kermack and K. A. Kermack. Supplement No. 1 to the J. Linn. Soc. (Zool.), 50. Gidley, J. W. (1906).— Evidence bearing on tooth- cusp development. Proc. Wash. Acad. Sci., 8: 91-106. Gregory, W. K. (1934). — Polyisomerism and aniso- merism in cranial and dental evolution among vertebrates. Proc. natn. Acad. Sci. U.S.A., 20: 1-9. Kermack, D. M., Kermack, K. A., and Mussett, F. (1968).— The Welsh pantothere Kuhneo- therium praecornis. J. Linn. Soc. (Zool.), 47: 407-423. Kupfer, M. (1935).— Beitrage zur Erforschung der baulichen Struktur der Backenzahne des Haustindes ( Bos Taurus L.) Denies, schweiz. naturf. Ges., 70: 1-218. Marshall, P. M., and Butler, P. M. (1966).— Molar cusp development in the bat, Hipposideros beatus, with reference to the ontogenetic basis of occlusion. Archs oral Biol., 2: 949- 965. Osborne, H. F. (1907). — Evolution of mammalian molar teeth to and from the tritubercular type. Macmillan Co., London, 250 pp. Ride, W. D. L. (1964).— A review of Australian fossil marsupials. J. Roy. Soc. West. Aust., 47: 97-131. Rose, C. (1892a).— Uber die Enstehung and Forma- banderung der menschlichen Molaren. Anat. Anz., 7 : 392-421. (1892b).— Uber die Zahnentwicklung der Beuteltiere. Anat. Anz., 7: 639-650, 693- 707. Scott, J. H. and Symons, N. B. B. (1961). — Introduc- tion to dental anatomy. E. and S. Liv- ingstone Ltd., London, 388 pp. Simpson, G. G. (1936) .—Studies in the earliest mam- malian dentitions. Dent. Cosmos, 78: 791- 800, 940-953. Slaughter, B. H. (1965).— A therian from the lower Cretaceous (Albian) of Texas. Postilla, 93: 1-18. Taeker, J. (1892). — Zur Kenntnis der Odontogenese bei Ungulatum. E. J. Karow, Dorpat, 25 pp. Thomas, O. (1887).— On the homologies and succession of the teeth in the Dasyuridae. Phil Trans. R. Soc., 178: 443-462. Winge, H. (1941). — The interrelationships of the mam- malian genera. Vol. 1., C. A. Reitzels for- lag, Kobenhavn. Woodward, M. F. (1896).— On the teeth of the Mar- supialia with special reference to the pre- milk dentition. Anat. Anz., 12: 281-291. Wortman, J. L. (1902).— Studies of Eocene Mammalia in the Marsh Collection, Peabody Museum. Am. J. Sci., 13: 39-46. Ziegler, A. C. (1971).— A theory of the evolution of therian dental formulas and replacement patterns. Q. Rev. Biol., 46: 226-249. Journal of the Royal Society of Western Australia, Vol. 57, Part 4, December, 1974. 125 INSTRUCTIONS TO AUTHORS Contributions to this Journal should be sent to The Honorary Editor , Royal Society of Western Australia , Western Australian Museum, Perth. Papers are received only from or by communication through, Members of the Society. The Council decides whether any contribution will be accepted for publication. All papers accepted must be read either in full or in abstract or be tabled at an ordinary meeting before publication. Papers should be accompanied by a table of contents, on a separate sheet, showing clearly the status of all headings; this will not necessarily be published. Authors should maintain a proper balance between length and substance, and papers longer than 10,000 words would need to be of exceptional importance to be considered for publication. The Abstract (which will probably be read more than any other part of the paper) should not be an expanded title, but should include the main substance of the paper in a con- densed form. Typescripts should be double-spaced on opaque white foolscap paper; the original and one carbon copy should be sent. All Tables, and captions for Figures, should be on separate sheets. Authors are advised to use recent issues of the Journal as a guide to the general format of their papers, includ- ing the preparation of references; journal titles in references may be given in full or may follow any widely used conventional system of abbreviation. Note that all illustrations are Figures, which are numbered in a single sequence. In composite Figures, made up of several photographs or diagrams, each of these should be designated by letter (e.g. Figure 13B). Illustrations should include all necessary lettering, and must be suitable for direct photo- graphic reproduction. To avoid unnecessary handling of the original illus- trations, which are usually best prepared between 1£ and 2 times the required size, authors are advised to supply extra prints already reduced. Additional printing costs, such as those for folding maps or colour blocks, will normally be charged to authors. It is the responsibility of authors to adhere to the International Rules of Botanical and Zoological Nomenclature. Palaeontological papers must follow the appropriate rules for zoology or botany, and all new stratigraphic names must have been previously approved by the Stratigraphic Nomenclature Com- mittee of the Geological Society of Australia. Thirty reprints are supplied to authors free of charge, up to a maximum of 60 for any one paper. Further reprints may be ordered at cost, provided that orders are submitted with the returned galley proofs. Authors are solely responsible for the accuracy of all information in their papers, and for any opinion they express. Journal Volume 57 of the Royal Society of Western Australia 1974 Part 4 Contents 13. Sand fulgurites from Western Australia. By J. E. Glover. 14. Disjunct plant distributions on the south-western Nullarbor Plain, Western Australia. By E. Charles Nelson. 15. The development of premolar and molar crowns of Antechinus flavipes (Marsupialia, Dasyuridae) and the significance of cusp ontogney in mammalian teeth. By M. Archer. Editor: A. J. McComb The Royal Society of Western Australia, Western Australian Museum, Perth 41632/9/74—625 WILLIAM C. BROWN, Government Printer, Western Australia