JOURNAL OF

THE ROYAL SOCIETY

OF

WESTERN AUSTRALIA

VOLUME 47
PART I

PUBLISHED 24th APRIL, 1964

REGISTERED AT THE G.P.O., PERTH FOR TRANSMISSION BY POST AS A PERIODICAL

THE

ROYAL SOCIETY

OF

WESTERN AUSTRALIA

COUNCIL 1963-1964

President
Past President
Vice-Presidents

Joint Hon. Secretaries

Hon. Treasurer
Hon. Librarian
Hon. Editors

C. F. H. Jenkins, M.A.

W. D. L. Ride, M.A., D.Phil.

W. R. Wallace, Dip.For.

J. H. Lord, B.Sc.

Margaret E. Redman, B.Sc.

J. G. Kay, B.Sc.

R. D. Royce, B.Sc. (Agric.).

Ariadna Neumann, B.A.

J. E. Glover, B.Sc., Ph.D. (1963).

R. W. George, B.Sc., Ph.D. (1964).

A. S. George, B.A.

A. B. Hatch, B.Sc., Dip.For.

L. E. Koch, M.Sc.

R. J. Little.

D. Merrilees, B.Sc.

W. C. Packer, Ph.D.

P. E. Playford, B.Sc., Ph.D.

L. W. Samuel, B.Sc., Ph.D., F.R.A.C.I., F.R.I.C.

Journal
of the

Royal Society of Western Australia
Vol. 47 Part 1

1.— The Environment of the Quokka ( Setonix brachyurus) in the Darling

Range, Western Australia

by G. M. Storr*

Manuscript received — 21st May, 1963

The more constant environment of the main-
land quokkas and their maintenance of good
condition through the summer are contrasted
with the seasonal extremes and loss of condition
on Rottnest.

Introduction

Owing to summer drought and the low
capacity of the sandy soil for retaining water,
most Rottnest plants grow only in winter and
spring. They consequently contain much less
water and protein in March and April than in
August and September (Storr 1961). As their
food deteriorates, so the quokkas rapidly lose
weight, and their fur becomes increasingly
frowsy and verminous. In contrast, a small
series of quokkas from the Darling Range north
of Jarrahdale showed no significant seasonal
variation in condition (Sadleir 1959). It there-
fore became important to study the environment
of these quokkas, if only for a better under-
standing of the defects of Rottnest.

Habitat

At Jarrahdale (and apparently everywhere on
the mainland ) quokkas live in densely vegetated
swamps. Some water is always present, whereas
on Rottnest it is restricted in summer to a few
small seepages round the salt lakes and the
coast. Aggravating the prevailing absence of
fresh water on Rottnest is the scarcity of good
shade. On hot days the temperature in a typical
quokka shelter on Rottnest exceeded official
shade temperatures by 9-22 °F, whereas at
Jarrahdale it was 5-12° lower than official read-
ings (Sadleir 1959).

The lower than shade temperature at ground
level in Jarrahdale swamps is due to the great
depth and stratification of the plant cover. The
lowest story consists of tall sedges, Cladium
vreissii and Lepidosperma tetraquetrum, inter-
spersed with shrubs of Acacia diver gens, A.
alata, Oxylobium linear if olium, Mirbelia dilatata ,
Bossiaea biloba, Aotus cor difolia, and Boronia
elatior. Next is a layer of tall shrubs, mostly
Agonis linearifolia, whose canopy is usually
closed, and through which emerge scattered
small trees; Melaleuca parvi flora, Banksia litto-
ralis , and Albizzia distachya. Above all these are
the tall eucalypts, E. patens and E. calophylla.

Quokka runways (more appropriately, tunnels)
intersect the sedge zone, and it is here that the
animals hide up during the day. At night they
come out to feed not only in the swamp itself

* Zoology Department, University of Western Australia.

Present Address: Western Australian Museum, Perth.

but also along its edges, where the ground cover
is much lower and more open. The vege-
tation of these marginal flats consists principally
of small sedges, Cladium laxum, Tetrariopsis
octandra, Cyathochaete avenacea, and Lom-
andra endlicheri; perennial herbs, Thomasia
foliosa, T. pauci flora, and Trachymene com-
pressa ; and the fern Pteridium aquilinum. As
the land rises the ground cover quickly merges
with the highly sclerophyllous undershrubbery
of the jarrah forest.

The vegetation of the swamps and their mar-
gins is thus dominated by sedges, perennial
herbs, and leguminous and myrtaceous shrubs.
In contrast to Rottnest, succulents, i.e. Cheno-
podiaceae, Aizoaceae etc., are completely absent.
Grass is rare and represented by a single species
( Tetrarrhena laevis ) . Indigenous annual herbs
are an insignificant element in the flora, and no
alien species have become established where the
vegetation remains undisturbed.

Seasonal variation in nutrients

The terminal foliage of plants growing in and
beside a quokka swamp near the source of Man-
jedal Brock (3 miles north of Jarrahdale) were
analysed for water and protein on various occa-
sions in 1958-9. The results are set out in Table
1, an asterisk prefixing species known to be
eaten by quokkas (these were found by search-
ing the vegetation along runways for evidence
of browsing).

Almost all the shrubs and herbs growing on
the dry lateritic soil above the swamp have a
winter growth regime. Fragmentary data (not
tabulated) indicate that water and protein levels,
as on Rottnest, are considerably higher in spring
than autumn.

Plants growing on the clayey soil of the mar-
ginal flat have generally a similar regime;
though there is no great contrast between spring
and autumn values of water and protein.
Pteridium alone of the seven species has summer
rather than winter growth, without, however,
marked seasonal variation in nutrients.

Of the eleven species growing in water or per-
manently damp soil, only two, Aotus and Oxy-
lobium, have a winter growth regime; but again
seasonal fluctuation in nutrients is quite small.
Four species, Acacia alata, Albizzia, Eucalyptus
patens, and Mirbelia, have a summer growing
season; while the remainder, including the two
sedges, scarcely change in protein and water
content.

1

TABLE 1

Crude protein content (% dry weight ) t with water content (% wet weight) in brackets , of
terminal foliage of common plants at Manjedal Brook. Known food plants are asterisked.

Swamp Species -

Acacia a lot a
*-4 at) n is I i n ea *■ ifol in
A Ibizi in diatachya

* .1 of n .s' cor difolia .

A star lea fuscic it laris
lltntl-sia littoral is

Cl odium preissii

K u cal t/pt as patens

* Lepiampe.rma tetraquetrum
Mirbetia dil alula

*(hri/lobiuM ! incur if olinm
Margin.il Kbit Species -

Boss uvea bilqba

*Cladium la. mm
Pteridinm nquilin am
7 'el ra r iopsis or fa n <1 m

* Tcharrhena Uivns
Thom a, s in fo/ iosa
Thorn asia pa nr i flora

June and July

October

January

March

10-6 (54)
6*9 (58)

8*1 (58)

118 (01)
6 • 4 (58)

13*7 (58)
8*4 (59)

15-8 (62)
9-5 (72)

16*0 (61)

20*4 (65)
7*3 (59)

5*4 (53)

7 • 0 (56)

4 • 6 (50)

4*8(52)

5*7 (53)

5 -7 (54)

6*0 (56)

4-5 (40)

4*2 (40)

4-0 (43)

4*3 (44)

6*3 (54)

8-3 (66)

7*8 (65)

4*8 (64)

4*7 (61)

5 • 0 (63)

4 ■ 4. (63)

8*7 (68)

9 • 7 (66)

12*2(73)

12*9 (64)

11*6 (59)

12-2 (63)

8*3 (58)

7-7 (45)

6*8 (38)

4*8 (60)

4-8 (64)

3*9 (56)

6*3 (52)

6 • 0 (55)

7*0 (52)

4*7(44)

3*3(41)

9*8 (61)

7*2(62)

6 • 7 (53)

C * 1 (53)

6-7 (49)

8-6 (57)

7-6 (59)

4*8 (50)

5*9 (49)

8*6 (53)

6*2 (56)

7*2 (64.)

At all seasons there are seme plants in or
beside the swamp putting on fresh growth.
Because of this and the small intra-specific
fluctuation in nutrients, there will likewise be
little seasonal variation in the quokkas’ intake
of water and protein. To illustrate this con-
stancy the data in Table 1 have been averaged
and graphed (Fig. 1). Although these averages
are not necessarily the same as those of their
actual food, the latter must vary in the same
way and indeed cannot be greatly different. For
comparison, mean water and protein content of
the food of Rottnest quokkas are also graphed
<data from Storr 1961).

Discussion

The absence of marked seasonal fluctuation in
the condition of Jarrahdale quokkas is undoubt-
edly due to their relatively stable plane of
nutrition. What is not so easily explained is
how T they maintain good condition on a diet that
can be little richer in protein than that of Rott-
nest quokkas at its worst. Perhaps their DMI
(daily intake of dry matter) is considerably
higher than on Rottnest, where it averages 45 g
in adult males (Storr 1963), a much lower rate
than in captive quokkas eating dry food but
supplied with unlimited water (cf. Calaby 1958,
and Bentley 1960). Since DMI in ruminants
falls as the water content of their food exceeds
80% (Dodsworth and Campbell 1952), the intake
of Rottnest quokkas could well be depressed by
the succulence of their food (75-85% water).

A perusal of the data in Table 1 shows that
the mean water content of the food at Jarrah-
dale cannot be much greater than 60% through-
out the year. It is thus possible that a higher
DMI more than compensates for the relatively
low protein content of the food of Jarrahdale
quokkas.

Acknowledgments

The work was financed by research grants
from the CSIRO and University of Western Aus-
tralia. The writer is grateful to Mr. R. D.
Royce (Government Botanist) for the identifica-
tion of plants; to Mr. S. Gorman (Government
Chemist) for protein determinations; to his
colleagues, Drs A. R. Main and R. M. Sadleir,

Fig. 1.— Seasonal changes in protein and water content
of vegetation. Full lines, plants at Manjedal Swamp;
broken lines, food plants of Rottnest quokkas.

for advice and assistance in the field; to Prof.
H. Waring for criticism of the manuscript; and
to Dr Main for supervising the work.

References

Bentley, P. J. (1960). — Evaporative water loss and tem-
perature regulation in the marsupial Seto-
nyx brachyurus. Aust. J. Exp. and Med. Sci.
38: 301-306.

Calaby, J. H. (1958).- The rate of passage of food resi-
dues and digestibility of crude fibre and
protein by the quokka, Setonix brachyurus
(Quoy and Gaimard). Aust. J. Biol. Sci. 11:
571-580.

Dodsworth, T. L. and Campbell, W. H. M. (1952). —
Effect of the percentage dry matter in the
diet on dry matter intake of ruminants.
Nature 170: 1128-1129.

Sadleir, R. M. (1959). — Comparative aspects of the
ecology and physiology of the Rottnest and
Byford populations of the quokka. Setonix
brachyurus (Quoy and Gaimard). Thesis,
University of Western Australia.

Storr, G. M. (1961). — Some field aspects of nutrition in
the quokka (Setonix brachyurus). Thesis,
University of Western Australia.

Storr, G. M. (1963).— Estimation of dry-matter intake in
wild herbivores. Nature 197: 307-308.

2

2. — An Ordovician Cystoid (Pelmatozoa, Echinodermata) from Western

Australia

By Ida A. Brown*

Manuscript received. — 18th June , 1963

The paper describes a new species of Cheiro-
crinus Eichwald from the Ordovician of the
Kimberley Division, Western Australia. This is
the first cystoid (Pelmatozoa, Echinodermata) to
be described from Australia, and the first record
of this genus from the Southern Hemisphere.

Introduction

There are comparatively few records of recog-
nizable fossil Pelmatozoa (Echinodermata) in
Australia, although “crinoidal” limestones are
relatively common within the Palaeozoic
sequence. This is the first description of a
cystoid; a few blastoids have been described
from the Upper Palaeozoic or eastern Australia
(Brown 1941), and a limited number of crinoids
are recorded from the Middle and Upper Palaeo-
zoic of both eastern and western Australia.

Nevertheless some rare and interesting echi-
noderms have been found in Australia and
undoubtedly many await discovery. White-
house (1941), described echinoderms from the
Middle Cambrian or Thorntonia Station in
north-western Queensland (lat 19°30 / S., long,
138°55'E.) which were so unlike any previously
known forms that he placed them in two new
classes of a new sub-phylum — Haplozoa. They
were associated with remains of cystoids, includ-
ing “ Eocystis (?)”, which have not yet been de-
scribed.

More recently Gill and Caster (1960) have
described Silurian and Devonian carpoid echino-
derms from Victoria, which are associated with
“many other echinoderms — cystoids, blastoids,
crinoids, machaerideans, starfish, and brittle
stars”.

The present paper records the occurrence of
a new species of Cheirocrinus from the Ordo-
vician of Western Australia. It is of considerable
geological interest, even though at present it is
represented by only one incomplete specimen,
collected by Mr. D. Merrilees of the Western
Australian Museum, Perth. Western Australia.
It comes from the type-locality of the Emanuel
Formation (Ordovician) in Emanuel Creek, lat.
18 !J 39'S.. long. 125°53'E. The approximate posi-
tion is indicated in Figure 1, and is shown more
accurately on the geological map by Guppy et al.
(1958).

The discovery of a sequence of fossiliferous
Ordovician rocks outcropping in the vicinity of
Prices Creek over an area of about twelve square
miles was first announced by Guppy and Opik
(1950) who described briefly the stratigraphy,
structure and palaeontology of the sequence they
named the Prices Creek Group. The occurrence
was discussed later by Guppy et al. (1958) and
by McWhae et al. (1958, pp. 28-29).

Mrs. W. R. Browne, 363 Edgecliff Rd., Edgecliff, N.S.W.

0 400 mil<zs V

Fig. 1.

The Prices Creek Group consists of the
(lower) Emanuel Formation and the (upper)
Gap Creek Formation. A rich fauna obtained
from both formations includes brachiopods,
trilobites, gastropods and nautiloids which
Guppy and Opik recognised as ranging in age
from Upper Ozarkian to Lower Trenton (Lower
to Middle Ordovician) of the United States suc-
cession. The nautiloids have been described by
Teichert and Glenister (1954). Earlier, Miss K.
L. Prendergast (1935) had described a plectam-
bonoid brachiopod, Spanodonta hoskingae, from
beds now known as the Gap Creek Formation,
which she mistakenly believed to come from the
overlying Devonian. This brachiopod was recog-
nised by Dr. G. A. Cooper of the U.S. National
Museum, Washington, D.C. in 1946 (or earlier)
as an Ordovician form (Browne 1952), but no
details of its field-occurrence were known at the
time.

The presence of Cheirocrinus in the Emanuel
Formation confirms the previous evidence of the
age of the formation, since the genus is known
only from the Chazyan and Mohawkian (Tren-
ton) of North America and from equivalent
Lower and Middle Ordovician formations of
Russia, Norway, Esthonia, Sweden, France, Por-
tugal and Scotland. It has not hitherto been
recorded from the Southern Hemisphere. (See
Bassler & Moodey 1943). Unfortunately it is
not known from which particular bed in the
Emanuel Formation the specimen was derived.

3

I wish to thank Mr D. Merrilees for permission
to examine this specimen.

Systematic Description

Phylum

Sub-Phylum

Class

Order

Super-family

Family

ECHINODERMATA
PELMATOZOA
Cystoidea von Buch, 1844
Rhombifera Zittel, 1879
Glyptocystitida Bather,

1913

Cheirocrinidae Jaekel, 1899

Bather (in Lankester 1900, p. 58) used the old
family name “Glyptocystidae” to include the
genera Giyptocystis, Cheirocrinus, etc. In his
independent work, which was in the press at the
same time as Bather's, Jaekel (1899) proposed
the family name “Chirccrinidae”. Bather (1913)
in his work on Girvan cystoids accepted this
classification, but emended the name to Chei-
rocrinidae to agree with the spelling of Cheiro-
crinus Eichwald, 1856, with the explanation
(p. 434) — “The name Cheirocrinidae has now
been chosen instead of the old-established
‘Glyptocystidae’, so as to avoid confusion with
other senses of the latter name, and in particular
to render it possible for anyone to follow Prof.
Jaekel in transferring ' Giyvtocystis ’ to the Callo-
cystidae without having to alter any Family
name”. Bather (1913, p. 363) used “Glyptocys-
tidae’’ as the name of a super-family, and this
was modified to “Glyptocystitida” by Regnell
(1945).

The family name Cheirocrinidae has been
accepted by Bassler and Moodey (1943, p. 5) and
by Bassler (1943, p. 669) and Regnell (1951).

The arrangement and orientation of the
plates of the theca and the position of the pore-
rhombs in this family have been the subjects
of investigation by a number of writers. Bather
<1900, p. 58) published a figure (reproduced in
Fig. 2) showing the actual distribution of the
pore-rhombs in all known members of the
family, numbering the thecal plates w’hich occur
in five cycles, in the order 1 to 24 as shown in
Figure 2, a scheme which he used again in his
study of the Girvan fauna (Bather 1913).

Foerste (1920, p. 41), Bassler (1943), Sinclair
(1948) and others have followed this scheme of
nomenclature, although Regnell (1951) has pre-

Fig. 2. — Diagrammatic sketch to show the actual distri-
bution of the pore-rhombs in the Family Cheirocrinidae.
(From Bather in Lankester 1900. p. 58, Fig. XX).

ferred Jaekel’s (1899) system of numbering the
plates as basals, 1-4; infra-laterals, 1-5; laterals,
1-5; etc.

To avoid ambiguity for the Western Australian
specimen and for easy comparison with similar
specimens Bather’s system is used in this paper.

Genus Cheirocrinus Eichwald, 1856

Eichwald, E., 1856. Bull. Soc. Nat. Moscou 29
(1): 123. For Synonymy See Bassler and

Mcodey (1943, p. 142). Type-species: Cyatho-
crinus penniger Eichw r ald, 1842.

Eichwald’s original definition (1856) is as
folloW’S; — “Diese Gattung zeichnet sich durch
strahlig gerippte Schilder aus, die parallel den
Rippen eine Reihe durch-gebrochner langlicher
Poren fur die Sauger zeigen und auf den Schul-
terschildern kurze gegliederte Hande tragen, fast
wie der Ichthyocrinus Hall aus Nordamerika: der
Kelch ist gestielt und die grossen Cornuliten
scheinen zu ihm Oder zum Echinoencrinus gig an -
tens zu gehoren, mit dem tiberhaupt diese Gat-
tung viele Veiwvandtschaft zeigt”.

The type-species, Cheirocrinus penniger, is
known to occur in Russia and Esthonia. It has
been studied by various workers including Kirk
(1911), and Bather (1913, p. 436, Text-figure 41)
w r ho has given an analysis of the plates showung
the positions of the pore-rhombs and the shapes
of the individual plates as determined from the
diagrams of Jaekel (1899).

Bassler and Moodey (1943, pp. 142-144) have
recorded no less than 24 species now assigned to
this genus and additional species have been
described by Bassler (1943) and Regnell (1951).
I have been unable to consult the monograph
on Cheirocrinus foreshadowed by Sinclair (1948),
but a study of the literature show r s there is great
variation in the characters of the thecae of these
species and it seems probable that in future
several distinct genera may be recognised. The
Western Australian specimen falls within the
range of the genus Cheirocrinus as understood at
the present time.

Cheirocrinus merrileesi, sp. nov.

(Fig. 3 (A, B, C), 4 and 5)

Holotype. — Western Australian Museum. Perth,
Western Australia. Specimen No. 60.167.
Locality. — Emanuel Creek, Kimberley Division,
180 miles east-south -east of Derby,
Western Australia.

Formation. — Emanuel Formation, Prices Creek
Group. Ordovician. Collected by Mr
D. Merrilees (22. VII. 1960) on an expe-
dition to the Fitzroy Trough. (See
McKenzie 1961).

The specimen consists of the lower portion of
a laterally compressed theca, of which twelve
plates are preserved, comprising four infra-
basals, five basals and three lateral plates,
together with a few’ scattered anal plates and
possibly portion of an ambulacrum adhering to
one of the basal plates; the radiolateral and
deltoid plates are missing and the column and
brachioles are not preserved.

The specimen now’ measures about 18 mm in
width, 6 mm to 7 mm in thickness and 21 mm
in height. The complete thecal height was
probably about 30 mm and the true width and

4

9 § 12 1*5 mm.

Fig. 3 . — Clieirocrinus merrileesi sp. nov. Ordovician, Emanuel Formation, Emanuel Creek, Kimberley District
Western Australia. Holotype, Western Australian Museum, Perth, Western Australia Spec. No. 60.167. Photo

I.A.B. A. Anterior, B. Left lateral and C. Posterior Views.

Fig. 4. Clieirocrinus merrileesi sp. nov. Sketch of theca as shown in Fig. 3 to indicate identification of plates and

position of radiating ridges.

5

thickness were rather greater than at present,
since the plates have sprung apart along the
sutures and seme of the plates overlap, e.g.
plates 1 and 6 overlap plate 5, and 7 overlaps
part of 8. This displacement must have taken
place during preservation, but except for plate
11, none of the individual plates appears to have
suffered fracture or distortion.

Thus the complete theca w T as probably
higher than wide, and considerably wider than
thick, proportions which are characteristic of
Cheirocrinus (Bather 1913, p. 435; Kirk 1911,
p. 19). \

The specimen is preserved in fine yellowish
limestone, and the thecal plates are of crystalline
calcite, the usual preservation of shells of fossil
Echinodermata, /' '

Each of the plates is thick, convex and has
the well-marked radiating ridges characteristic
of the genus, the centres cf radiation usually
being eccentric, as shown in the accompanying
photograph (Fig. 3; A, B, and C) and corre-
sponding diagram (Fig. 4: A, B, and C).

The positions of the ridges vary as indicated
in Figures 3, 4 ,and 5. Sometimes they are
normal to the middle of the sutures of adjacent
plates and appear to have continued across the
suture, e.g . frem plates 3 to 7, 3 to 8, and 3 to
9; in plates 5 and 6 two of the ridges extend to
the angles cf the plates.

The analysis of the theca is shown in Figure 5.

No column is preserved but the scar for its
insertion is clearly marked on the lower surface.

The infra-basals , plates 1 to 4, are variable
in size; plates 2 and 4 are smallest; plate 1 is
rhcmbchedral. modified at the lower edge for
the insertion of the top of the column, while
plate 3, the largest, is almost hexagonal, prob-
ably representing the fusion of two plates.

The basals , plates 5 to 9, are also variable in
size; plates 5, 6 and 9 are hexagonal and plates
7 and 8 are modified for the anal area which is
relatively large.

Of the lower lateral jilates only plates 11, 12
and part of 14 remain. Plate 11 has been
crushed inwards and its sutures with plates 5
and 6 are obscured. Plate 12 is folded about
its vertical axis which is in the plane of com-
pression of the theca, and its contact with the
anal area is depressed. Only portion of plate
14, in contact with plates 8 and 9, is preserved.
It was not considered advisable to remove more
of the matrix on account of the risk cf damage
to the specimen.

No plates of the fourth cycle — plates 15 to 19 —
are preserved, but their existence must be pre-
sumed from the fact that the lower halves of
pore-rhombs occur on plates 11 and 12, requir-
ing plates 16, 17 and 18.

Fig. 5 — Cheirocrinus merrileesi sp. nov. Diagrammatic sketch to show structure and arrangement of the thecal
plates, and the positions of the radiating ridges and pore-rhombs. Infra-basals 1-4, Basals 5-9, Lateral Plates

11. 12 and 14. Anal area, As.

6

There is a small water-worn piece of shelly
material adhering to plate 5 which may be part
of an ambulacral structure.

The arrangement of the pore-rhombs is shown
in Figure 5. By comparison with Figure 2 it
may be seen that it is in agreement with that
of the Family Cheirocrinidae (Bather 1913).
The pore-rhombs are all conjunct, the pore-slits
passing uninterruptedly across the sutures.
Rhombs 1/2 and 1/4 each have about 10 slits;
1/5 and 1/6 about 22 slits; rhomb 5/6 is abnor-
mal in that the pore-slits, about 18 in number,
form a triangular area on plate 6, such that
the longest slit is not central, but at the end
closest to plate 1; poor preservation prevents
determination of all the pore-slits on plate 5.

A condition somewhat similar to that of plate
6 occurs on plates 8 and 9, where the longest
slits are at one side, each against one of the
radiating costae, but the slits pass directly across
the sutures 8/14 and 9/14. Although plate 11
has been crushed inwards it bears its share of
rhombs 11/?10, 11/916, 11/917 and 11/12.

A few scattered small plates indicate the anal
area.

The remainder of the theca is lost.

Remarks. — Cheirocrinus merrileesi sp. nov.
differs from all species of the related Glyptocys-
tites in that plates 5 and 6 meet along a suture
and are not separated by the upward extension
of plate 1 to touch plate 11 (See Sinclair 1948,
p. 310), and in this feature it differs also from
Cheirocrinus penniger the type species of the
genus (See Bather 1913, p. 436, Text-fig. 41).

Bather (1913) suggested a classification of the
Cheirocrinidae based on the character of the
rhombs, whether conjunct, disjunct or multi-
disjunct. C. merrileesi falls within his Group 1
(“Pectini-rhombs all, or at least on the base,
conjunct”.) in which he included only three
species, C. atavus Jaekel, 1899, C. giganteus
(Leuchtenberg, 1843) and C. constrictus Bather,
1913.

I have no specimens of described species of
Cheirocrinus available for direct comparison, but
to judge by the published descriptions and
figures by Jaekel, Bather, Bassler, Sinclair and
Regnell, the specimen from Western Australia
differs from all other species in the arrange-
ment of the plates and pore-rhombs, and there-
fore is regarded as a new species, which I have
named Cheirocrinus merrileesi in honour of Mr.
D. Merrilees.

References

Bassler, R. S. (1943). — New Ordovician cystidean echi-
noderms from Oklahoma. Amer. J. Sci. 241 :
694-703.

Bassler, R. S. and Moodey, Margaret W. (1943). — Biblio-
graphic and faunal index of Paleozoic pel-
matozoan echinoderms. Spec. Pap. GeoL
Soc. Amer. 45.

Bather, F. A. (1900). — The Echinoderma. In Lankester,
E. R., “A Treatise on Zoology”. Part III
(London).

Bather, F. A. (1913).— Caradocian Cystidea from Girvan.
Trans . Roy Soc. Edinb. 49 (2): 359-529.

Brown, Ida A. (1941). — Permian blastoids from New
South Wales. ./. Roy. Soc. N.S.W. 75: 96-103.

Browne, Ida A. (1952). — Ordovician limestone at Bowan
Park. N.S.W. Aust. J. Sci. 15 (1): 29-30.

Eichwald, E. (1856). — Beitrag zur geographischen Ver-
breitung der fossilen Thiere Russlands. Bull.
Soc. Nat. Moscou. 29 (1): 123.

Foerste, A. F. (1920). — Racine and Cedarville cystids and
blastoids with notes on other echinoderms.
Ohio J. Sci. 21 (2): 41.

Gill, E. D., and Caster, K. E. (I960). — Carpoid echino-
derms from the Silurian and Devonian of
Australia. Bull. Amer. Paleont ., 41 (185):
1-71.

Guppy, D. J.. Lindner, A. W., Rattigan. J. H.. and
Casey, J. N. (1958). — The geology of the Fitz-
roy Basin, Western Australia. Bull. Bur.
Min. Resour. Aust. 36: 16-20.

Guppy, D. J, and opik, A. A. (1950). — Discovery of
Ordovician Rocks, Kimberley Division, W.A.
Aust. J. Sci. 12 (6): 205-206.

Jaekel. O. (1899). — “Stammesgeschichte der Pelma-
tozoen”. Pt. 1. Thecoidea und Cystoidea.
(Berlin).

Kirk. E. (1911). — The structure and relationships of
certain Eleutherozoic Pelmatozoa. Proc. U.S.
Nat. Mus. 41 : 19-20.

Lankester, E. R. (1900). — ‘‘A Treatise on Zoology”. Part
III. (London).

McKenzie, K. G. (1961). Vertebrate localities in the
Triassic Blina Shale of the Canning Basin,
Western Australia. J. Roy Soc. W. Aust. 44:
69-76.

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.

Prendergast, K. L. (1935). — Western Australian Upper
Palaeozoic fossils. J. Roy Soc. W. Aust. 21:
9-35.

Regnell, G. (1945). — Non-crinoid Pelmatozoa from the
Paleozoic of Sweden. A Taxonomic. Study.
Medd. Lunds Geol. — Mineral Inst. 108.

Regnell, G. 1 1951 ) .—Revision of the Caradocian-Ash-
gillian cystoid fauna of Belgium. Mem.
Instit. Roy. Soc. Nat. Belg. 120: 1-47.

Sinclair, G. W. (1948). — Three notes on Ordovician cys-
tids. J. Paleont. 22 (3): 301-314.

Teichert, C. and Glenister, B. F. (1954). — Early Ordo-
vician c.ephalopod fauna from northwestern
Australia. Bull. Amer. Paleont. 35: 1-112.

Whitehouse, F. W. (1941). — Early Cambrian echinoderms
similar to the larval stages of Recent forms.
Mem. Qd Mus. 11: 1-28.

3. — Description of a new Stonefish of the family Synanceidae from

Western Australia

By R. J. McKay*

Manuscript received — 18th June, 1963

A new species of the genus Inimicus, belonging
to the family Synanceidae from Western Aus-
tralia is described and compared with other
members of the genus. Notes on the ecology
including an account of the severity of the
sting of the new species are included.

Introduction

During the years 1958, 1960 and 1962

the Western Australian Fisheries Department
obtained a large collection of fishes as a result
of the resource surveys carried out by the re-
search vessels Lancelin and Per on. The majority
of the fish were collected in the Shark Bay and
Exmouth Gulf areas. Over the past five years,
the author has devoted much of his spare time
to the study of certain groups of this collection.

The Western Australian coastline is poorly
known ichthyologically, and therefore it is not
surprising to find that much of the collected
material is not recorded from the State, and
that a few species are found to be new.

This paper presents the description of one of
the species considered to be new. The oppor-
tunity has also been taken to record information
on the ecology of the species.

Genus Inimicus Jordan and Starks, 1904
Genotype Pelor japonicum Cuvier, 1829

Inimicus cirrhosust species nova
Inimicus didactylus; Mees 1960, p. 19.

Differential diagnosis

Differs from all others of the genus in the
possession of yellow spots on the inner surface
of the pectoral fin.

Description

The description is based on the holotype and
fifteen paratypes. The counts and measure-
ments within the parenthesis are those of the
paratypes. D. Ill XIV, 9 (III XIII-XV, 8-10).
A. II, 12. (II-III, 12-13). P, 1, 6, 3 + 2 12

(1 _ 2, 5 — 6, 3 — 4 + 2 — 3), V. I, 5. (I, 5).
C. 2, 9, 3 14 (2 — 3, 8 — 9, 3 — 4 13 — 14).

Gill rakers 2 + 7.

Head 2.9 (2.7 — 3.2); depth of body 3.4 (2.6 —

4.0) ; nectoral fin 2.8 (2.5 — 3.1), all in standard
length. Eye 5.4 (5.0 — 7.0) in head, 1.5 (1.4 —

2.1 ) in interorbital, 2.6 (2.4 — 3.3) in snout.
Interorbital 1.8 (1.4 — 1.8) in snout, 3.6 (3.0

— 4.0) in head. Snout 2.0 (2.0 — 2.5) in head.
Longest dorsal spine 1.5 (about 1.5) in head.

Head about as wide as long, snout greater
than postorbital part of head, and almost half
length of head. Head grotesquely shaped, de-

* Fisheries Department, Perth, Western Australia.
Present address: Western Australian Museum. Perth.

t From the Greek kirrhos, yellow.

pressed. Bony orbits elevated and prominent,
with 2 or 3 low blunt knobs superiorly. A groove
immediately below eye, preorbital pits shallow,
infer orbital space deeply concave with transverse
ridge branching on to orbits at rear. A deep
transverse depression behind eyes, nuchal and
post-temporal spines developed. One anterior
and two posterior spines on suborbital stay.
Coronal ridges on snout more or less developed
in smaller examples, well developed in larger
specimens.

Preorbital with two blunt spines on upper sur-
face, ending in a sharp spine normally covered
with skin. Nostrils low and tubular; anterior
pair situated on each side of nasal hump or pro-
tuberance; posterior pair situated behind this
protuberance. Preopercle with 4 spines; lower
2 blunt and scarcely worthy of being called
spines; upper 2 sharp, superior spine with a
ridge preceding. Opercle terminating in a
fleshy flap; upper portion of opercle with 2 well
developed radiating ridges, uppermost ending in
a spine.

Mouth small, oblique; lower jaw projecting,
with a small symphysial knob at tip. Maxillary
with posterior end expanded; width posteriorly
almost equal to eye diameter. Maxilla ends well
before eye. Mandibulary ramus expanded pos-
teriorly.

Teeth small, sharp, and conical; 4-5 series in
both jaws, fewer posteriorly; similar teeth in
4-5 series on vomer. Palatines toothless.
Tongue triangular, free at tip, upper surface
smooth generally with few low tubercles near
margin.

Gill rakers 1 — 3 + 7 — 8, low, knoblike and
intensely spiniform.

Branchiostegal rays 7. Gill membranes united
to isthmus but not broadly so.

Body elongate, little compressed, tapering pos-
teriorly almost in a straight line along dorsal
surface from 4th or 5th dorsal spine.

The 3 anterior dorsal spines almost detached
from the rest of fin, with membrane almost
to their tips. Remaining 13-15 spines, except
last 1 or 2, have interspinous membrane low.
Dorsal spines sharp, slender and strong, grooved
at tips, skin reaches tips of spines in most
examples. All spines except first 3 and last 3,
almost uniform height. Dorsal rays decrease in
length posteriorly, last ray connected to caudal
peduncle by membrane, only tip free. Anal fin
low, spines sharp but embedded in thick skin.
Tips of anal rays slightly thickened, free for
about one-third of their length, last 3 with
membrane to tips.

8

Pectoral fin large, extending well past origin
of anal. First 1 or 2 and last 5 or 6 pectoral
rays simple, intermediate ones branched. Lower
2 pectoral rays free, for the most part; gener-
ally curved, thickened slightly, and almost equal
in length. Length of ventral fin almost equal
to that of head, adnate to belly with free tips
of the rays reaching well beyond anus.

Caudal fin rounded, length almost equal to
that of head, outermost 2 or 3 rays normally
simple. Caudal peduncle less than interorbital
measurement.

Body and head without scales. Skin loose,
with dermal flaps, filaments, skinny tentacles
and branched appendages, especially on head,
chin, mandibles, margins of preopercle and pec-
toral fin, and upper surfaces of sides and back.
Well developed branched dermal appendages on
each side of chin, placed low on preopercle
margin.

A somewhat obscure lateral line present with
12 to 13 pores along its length. A few pores
above lateral line near dorsal surface; all pores
have small mostly branched dermal appendages.

Colour

Body and head generally brownish, densely
mottled with lighter and darker diffuse blotches
and variable marblings. Holotype deep brown
with areas of lighter brown and patches of grey-
white near 1st, 4th, 5th and 10th to 15th dorsal
spines, extending below lateral line; a pale
diffuse bar through soft dorsal fin. Lower sides
and belly with numerous scattered light and
dark flecks and diffuse blotches. Ventrals uni-
form brown. Caudal with 2 vague narrow
whitish cross-bars or two row's of w'hite spots
(in life, yellowO.

Pectoral fin dark brown, almost black with
indistinct pale cross-bars and variable blotches
on outer surface. Inner surface dark brown to

blackish, with 20 to 35 white (in life bright
yellow) spots of indefinite shape. In larger
examples the yellow’ spots are more diffuse and
frequently in the form of ocelli. (Fig. la).

Base of pectoral fin, and lower sides of belly
above ventral fins, pale with brow’n flecks and
indistinct marblings.

One paratype has body yellowish with scat-
tered brown spots on body and a few r dark spots
on head; dorsal fin plain yellow'. Caudal
brownish with scattered dark browrn and yellow r
spots. Low r er portions of body brownish with
some scattered dark spots. Ventrals brow'nish
with a few' scattered darker spots. Outer surface
of pectoral fin yellow^ with scattered dark browrn
spots. Inner surface of pectoral fin brow'nish
with yellow spots, about eye diameter in size; 2
to 3 row’s of smaller dark browrn spots near mar-
gin of fin, and a few scattered dark spots on
fin, especially near base.

Material Examined

Unless otherwise stated, all Western Australian
Museum material was taken by trawl net on the
State Fisheries Research Vessels Lancelin and
Per on, collected by the author. Western Aus-
tralian Museum and Australian Museum regis-
tered material is abbreviated WAM and AM re-
spectively.

I. cirrhosus

Holotype.— WAM P 4980 164 mm total length,
125 mm standard length, approx. 9 miles north
of Cape Peron Flats, Shark Bay, l.X. 1960.

Paratypes.— WAM P 4983 198 mm t.l., 155 mm

s. l., Exmouth Gulf 14.XI.1960; P 4987 322 mm

t. l., 256 mm s.l., Rat Island, Abrclhos Is.,
2 .III. 1963 hand spear w'hilst diving in 9 fathoms;
13 specimens from Shark Bay collected between
23. V. 1960 and 27.X.1960 WAM P 4981, P4982,

Fig. 1. Inimicus cirrhosus a. Inner surface of pectoral fin showing markings.

9

P 4984, P 4985, P 4986 and P 4988 to P 4995
inclusive, range of measurements 75 to 257 mm
t.L, 56 to 203 mm s.l.

Other species

I. didactylus (Pallas) AM 1B.4439 and 1.1499,
2 specimens.

I. japonicus (Cuvier) AM 1.13702, 1 specimen.

7. barbatus (De Vis) AM 1A.4782. 1A.6137.
1A.6671, 1A.6672, 1.14039, 1.10933 and 1B.1716, 7
specimens.

Distribution

Known from Exmouth Gulf. Abrolhos Islands
and the type locality, Shark Bay.

Comparison with Other Species

Inimic.us cirrhosus appears closest to 7. didac-
tylus and 7. barbatus , their structural details
and bcdy proportions being very similar.

On examining the two specimens of 7. didac-
tylus at the Australian Museum, it w T as found,
as McCullcch (1916, p. 197) noted, that the
coloration and markings of the inner surface
of the pectoral fin agreed with Bleeker’s (1874.
pi. IV, figs. 1-1A. 1878, pi. CCCCXIV. figs. 5-5a>
plates of this species.

The specimens of 7. barbatus have the colora-
tion and markings of the inner surface of the
pectoral fin as McCulloch (1916, pi. L.YIID
figured.

Sleeker (1874, p. 1) and McCulloch (1916)
have remarked on the importance of the colora-
tion and markings on the inner surface of the
pectoral fin in determining species of Inimicus
as these are consistent within a species. This is
true for 7. cirrhosus , as over 200 specimens were
checked from Shark Bay and Exmouth Gulf
during 1958. 1960, and 1962.

The 2 specimens of 7. didactylus examined had
the eye equal to the interorbital space, and in
7. cirrhosus , the eye in interorbital space ranged
from 1.4 to 2.1. Herre (1951, p. 472) records
the eyes as being about a diameter apart in
Philippine specimens of 48 to 128 mm in length.
Fowler (1927, p. 288) gives eye as 2 in interorbi-

tal in a Philippine specimen of 183 mm, while
Day (1878, p. 161) gives eye as 1-1? apart in a
specimen of 54 inches from the Andaman Islands.
McCulloch (1916) records variation in interorbi-
tal space for 7. barbatus, being narrower in
smaller examples: this is generally so in 7. cirr-
hcsus (see Table 1).

Bleeker (1878, pi. CCCCXVI, figs. 3-3a) shows
the pectoral fin markings of 7. cuvieri (Gray)
and on pi. CCCCXIII, figs. 1-la, the pectoral fin
markings cf 7. brachyrhynchus (Bleeker). The
pectoral fin markings of 7. maculatus ( 7.

didactylus), and 7. filamentosus (Cuvier) were
also figured. Published descriptions by Gunther
(1360, p. 150), Herre (1951, p. 470 and p. 473),
and Fowler (1927, p. 289; 1928, p. 299) agree
with Bleeker’s plates.

The specimen cf 7. japonicus examined has
the eye 2.0 in interorbital, snout 2.7 in head,
and the postorbital part of head (25 mm) is
greater than the snout (19 mm). Gunther
(I860, p. 151), gives eye as 2.5 in interorbital,
and an anal ray count of 9, for this species, while
Tanaka (1914, p. 248) records counts of 7 dorsal
and 9 anal rays in a specimen of 220 mm, and
gives snout as 2.5 in head.

Smith (1958, p. 276) records D. as III XII, 8.
A. Ill, 8, ‘‘and pectoral and caudal with middle
parts light” for 7. filamentosus. Fowler (1938,
p. 86) gives coloration of pectoral fin of 7. bifilis
(Fowler) as having “broad white subbasal band,
adjoining blackish area over branched rays and
another terminally; lower detached rays with
dark spots”. 7. bifilis has a dorsal count of XVII.
8, and an anal count of 13.

7. caledonicus (Sauvage) has a dorsal ray
count of 7, and a total anal count of 12. (Fowler
1928, p. 299, (after Sauvage)).

Variation in Pectoral Rays
Paratype WAM P 4994 has 3 free pectoral rays
on the right hand fin, the first 2 are joined for
half their length; the left hand pectoral fin
has the normal 2. The joined pectoral rays in
this specimen number 9 instead of the usual 10.

/

Total length

104

Standard length

125

Length of head

4:5

Length of snout

Ol

Diameter of eye

s

1 iilerorhital

12

Greatest body depth

37

Length of pectoral
fin

45

Leu sit h of upper free
poet oral ray

39

Length of lower free
pectoral ray

40

Soft, dorsal base

32

Anal base

54

Ventral base

35

Depth of oattdal
peduncle

11

No. of dorsal rays

9

No. of anal rays

12

No. o r caudal rays ....

14

TABLE I

Measurements in mm and fin ray counts of 7. cirrhosus

00

-i;

<M

oc

tS

04

A

* 0C

tS

rc

-s

QG

/

•t

aO

35

■*t

r?

'M

pu

CO

1

rf

\T.

191

220

198

257 209

238

322

98

7.5

133

110

99

121

155

130

1 50

1 80

155

203 107

190

250

75

50

1 06

90

77

93

119

106

52

02

52

70 54

00

88

25

21

33

31

25

32

42

37

25

29

23

34 20

28

39

12

10

10

14

12

14

20

15

9

10

9

10 9

1 1

13

4

4

0

5

5

5

6

0

14

18

13

19 16

17

27

7

0

1 L

8

7

10

11

10

43

58

48

79 50

50

81

20

16

32

24

23

23

23

29

41)

00

53

70 59

62

S3

30 |

21

37

32

20

30

45

41

30

35

28

41 32

39

57

15

13

20

16

10

18

17

28

32

32

31

43 31

39

02

10

13

23

19

16

17

19

23

37

47

40

50 47

39

07

20

14

20

22

21

20

28

30

01

78

69

93 80

84

114

35

27

50

37

38

44

53

52

37

45

42

61 47

49

71

20

1 0

27

20

22

25

38

33

1 1

13

12

12 12

14

20

0

5

8

8

(5

7

a

9

9

9

9

8 10

9

9

9

9

9

9

9

8

8

9

12

12

12

12 13

13

12

12

13

12

12

12

32

12

12

13

13

14

13 14

14

14

14

14

14

13

14

14

14

14

10

Paratypes WAM P 4988 and P 4989 have the
two uppermost pectoral rays filamentous at their
tips, the rays extend 8 mm and 12 mm from
the tip of the pectoral fin respectively. Para-
tope WAM P 4922 has only the upper pectoral
ray filamentous; this ray extends 18 mm from
the pectoral fin membrane, 7 mm from the tip
of the pectoral fin. This condition apparently
exists in the juveniles as these 3 paratypes
measure 75, 56, 77 mm in s.l. respectively.

In view of these observations of the juvenile
characters of I. cirrhosus, it is pcssible that
seme “species” with similar characters may be
juveniles of other species. The type of I. bifilis
from the Philippines measures only 57 mm in
length and Fowler remarks “Greatly like
Inimicus filamentosus , but only the uppermost
pectoral ray ending in a filament which reaches
well beyond the depressed pectoral fin”. This
could be a juvenile condition of perhaps I. cuvieri
or I. didactylus which have been recorded from
the Philippines by Herre (1953, p. 576).

General Biology and Ecology

I. cirrhosus was found to be a sluggish,
bottom-dwelling marine form inhabiting open,
randy, or silty types of substrates. Inside Shark
Bay, it has a wide distribution and is commonly
taken on prawn trawling grounds, often in asso-
ciation with Scallops Amusium balloti Bernardi.
Two or 3 per haul are not uncommon. The
greatest number recorded inside Shark Bay
during 1962 was 14 during a 45-minute trawl.

During 1961, F.R.V, Percn used a temperature-
chlorinity conductivity meter, and specimens of
/. cirrhosus were collected in salinities ranging
from 35.7 % n to 44.9% 0 , and temperatures from
18.2 °c. to 26.4 n c. The depths recorded on the
echo sounder were between 7 and 12 fathoms.

Partly digested fish were found in the
stomachs of 2 specimens: these were identified
as Monacanthus sp. and Equula fasciata (Lace-
pede ) .

A near-ripe female, t.l. 215 mm, was trawled
inside Exmouth Gulf early in October, 1958. One
ripening female t.l. 220 mm trawled inside Shark
Bay early in February, 1962, had ova 0.8 mm in
diameter, and 3 females, t.l- 210 to 265 mm were
found to be spent during May, 1962.

Like other members of the family Synanceidae .
they are reluctant to move when approached,
and lie quite motionless, even when spilled from
the trawl net on the deck of the vessel. One
was taken underwater and the fish made no
attempt to move, but quickly spread the dorsal
spines to afford protection frem most directions.
The pectoral fins were extended to show the
markings on their inner surface. The fish was
not easily recognised in its natural surroundings,
as it was covered to some extent by fine sand,
and appeared quite hirsute due to the dense
covering of demal cirri and appendages.

Notes on Severity of Sting

This species is capable of inflicting an ex-
tremely agonising wound. The following notes
were made early in October. 1960, after the
author was stung on the thumb. The pain was
immediate and intense, the wound bled very
freely for a few minutes then ceased. Bleeding
was difficult to restart. Pain became rapidly
unbearable.

1145 hrs. While sorting a trayful of fish
emptied from trawl 31, left thumb deeply pene-
trated by dorsal spine of Inimicus. Bleeding
freely.

1150 hrs. Bleeding stopped.

1155 hrs. Thumb swollen, area around
wound becoming grey-blue in colour, extremities
of fingers becoming very painful, pain in thumb
almost unbearable. Elbow and shoulder now
aching.

12 midnight. Sweating profusely, thumb and
nand swollen, wrist very stiff, area around wound
bluish, threat and mouth dry.

0005 hrs. Became delirious, do not remember
much, but can recollect throwing myself about
the cabin in fearful pain. Crew members re-
strained my movements for the next 30 minutes,
during which I was screaming and complaining
of pains around chest and neck.

0045 hrs. Regained full consciousness, the
pain in arm and shoulder very agonising.

0115 hrs. Hand and thumb immersed in hot
water and sedatives taken.

0200 hrs. Pain still present but greatly
diminished.

Thumb remained stiff and swollen for four
days, but the pain had subsided 12 hours after
encounter. Bathing the wound with hot water
greatly reduced pain, but when the water became
ccol enough to be comfortably borne, the pain
returned. A tourniquet was not applied.

Stings by fishes such as Scorpaenids Para-
centropogon vespa COgilby) , Apistus carinatus
(Bloch) and Minous monodactylus (Bloch &
Schneider), and a stingray Amphotistius kuhlii
(Muller & Henle) were quite common on board
prawn trawlers operating inside Shark Bay.
From personal experience, the sting of /. cirr-
hosus is considerably more painful than any of
these species.

Acknowledgments

I wish to record my appreciation to those
persons who assisted mo in my studies and gave
much helpful advice in the preparation of this
paper: Dr. G. F. Mees, of the Western Australian
Museum and Mr. A. J. Fraser, Director of the
Western Australian Fisheries Department read
the manuscript; Mr. G. P. Whitley, of the Aus-
tralian Museum, gave helpful advice and assist-
ance while I studied fishes under his care: the
skipper. Mr. H. C. W. Piesse, and crew of the
Fisheries Research Vessels Lancelin and Percn,
assisted in collecting fishes during trawling in-
vestigatiens and administered assistance during
the erdeal of the Inimicus sting.

References

Bleaker, P. (1874 k— R evision des cspeces insulin-diennes
de la famille des Synanceoides. Natuurk.
Verh. holland. Maatsch. Wet. 3 ii <3): 1-22.

( 1878) . — “Atlas Ichthyologique ctes Indes

Orientales Ncerlandalses. public sous les
auspices du Gouvernernent Colonial Neer-
landais”. 9 i Amsterdam.)

Day. F. (1878). — The Fishes of India. 2 vols. (London).
Fowler. H. W. (1927). Notes on the Philippine fishes in
the collection of the Academy. Proc. Acad,
nat, Sci. Philad. 79: 255-297.

(1928). — The Fishes of Oceania. Mein.

Bishop Mus. 10: 1-466.

( 1938).— Descriptions of new Fishes obtained

by the United States Bureau of Fisheries
Steamer '‘Albatross'’ chiefly in Philippine
seas and adjacent waters. Proc. U.S. nut
Mus. 85. 31-135.

11

Gunther, A. (1860) .—Catalogue of the Acanthopterygian
fishes in the collection of the British
Museum. 2: 1-527. (London.)

Herre, A. W. (1951). — A Review of the Scorpaenoid
Fishes of the Philippines and adjacent seas.
Philipp. J. Sci. 80 (4).

(1953). — Check list of Philippine Fishes. U.S.

Fish and Wildl. Serv. Res. Dep. No. 20: 1-977.
McCulloch, A. R. (1916). — Report on some fishes obtained
by the F.I.S. “Endeavour” on the coasts of
Queensland, New South Wales, Victoria, Tas-
mania, South and South Western Australia.
Fisheries, Sydney 4: 169-200.

Mees, G. F. (1960). — Additions to the fish fauna of
Western Australia — 2. W. Aust. Fish. Dept.,
Fish. Bull. 9: 13-21.

Smith, J. L. B. (1958). — Fishes of the families Tetra-
rogidae, Caracanthidae and Synanciidae
from the Western Indian Ocean with further
notes on Scorpaenid fishes. Ichthyol. Bull.
12: 167-181.

Tanaka, S. (1914). — Figures and Descriptions of the
Fishes of Japan, including Riukiu Islands,
Bonin Islands, Formosa, Kurile Islands,
Korea, and Southern Sakhalin. 15: 247-262
(Tokyo.)

12

4. — Association of Larger and Planktonic Foraminifera in single samples
from Middle Miocene sediments, Guadalcanal, Solomon Islands, south-
west Pacific

By P. J. Coleman* and R. A, McTavishf

Manuscript received — 22nd October, 1963

Single samples from Miocene sediments in
north-central Guadalcanal, Solomon Islands,
contain both larger foraminifera and planktonic
species of the Sphaeroidinellopsis seminulina
fauna. This association was contemporaneous
during Middle Miocene ( Vindobonian) time.
Lepidocyclina ( ,, Multilepidina ,r ) suvaensis
(Whipple), Lepidocyclina ( N ephrolepidina ) jap-
onica Yabe, Cycloclypeus ( Katacycloclypeus)
martini Van der Vlerk and Miogypsina poly-
morpha (Rutten) survived well into the Vindo-
bonian, at least in this part of the Indo-Pacific
region.

Introduction

Until a few years ago the age-determination
and correlation of Tertiary sediments by means
of their foraminifera was based principally on
the so-called larger foraminifera with but minor
support from the smaller benthonic forms. Over
recent years, however, the use of planktonic
foraminifera for this purpose has become in-
creasingly common, as might have been expected
on purely theoretical grounds, and has been jus-
tified in practice. Studies of the planktonic
foraminifera, however, are usually made inde-
pendent of those on larger forms. There are
reasons for this — these two groups require
different methods of study; more important,
sediments containing abundant planktonic
species of foraminifera are usually barren of the
larger forms, and vice versa, a reflection of their
different life environments. This kind of mutual
exclusiveness, however, carries with it the risk
that age-determinations and correlations based
on the one group will differ from those based
on the other (see Cole, Todd and Johnson 1960;
Glaessner 1960) — a situation which has been
common enough in applied palaeontology. The
chance of such conflict can be reduced if a
sufficient number of mutual occurrences, of
stratigraphically defined normal associations,
can be established. This paper attempts to pro-
vide one such link between larger and plank-
tonic foraminifera which will be applicable to
Miocene foraminiferal faunas, at least in the
Indo-Pacific region.

Larger foraminifera of the Solomon Islands
have been described by Coleman (1963a), and
the planktonic forms by McTavish (1963). In
the Solomon Islands sediments, as elsewhere,
the two groups are not usually found together
in the same formation much less in the same

* Geology Department, University of Western Australia,
Nedlands, Western Australia.

- Formerly Geology Department, University of Western
Australia, Nedlands, Western Australia: now West Aus-
tralian Petroleum Pty. Ltd., Adelaide Terrace. Perth,
Western Australia.

sample. There are two localities, however, in
which both have been found together in single
samples and a third where samples bearing
planktonic forms are only 50 feet stratigraphic-
ally above those with larger foraminifera, in
strata making up an uninterrupted succession
of rapidly-deposited sediments. The first of
these, from Middle Miocene sediments in the
Tangaraisu River, north -central Guadalcanal,
provided the material for this paper. The others
are in Aauitanian (Tertiary upper ‘e’) sediments
on San Cristoval and San Jorge, respectively,
where there are associations of Eulepidina ephip-
pioides, Spiroclypeus leupoldi, Heterostegina bor-
neensis, Miogypsina thecideaeformis , Miogypsi-
noides dehaartii and Cycloclypeus eidae with
Globigerina bollii, G. ( Globigerinita ) dissimilis,
G. (G) unicava, Globigerinoides quadrilobatus ,
Globoquadrina dehiscens and G. Venezuelans
these Aauitanian associations are to be described
in another paper.

Stratigraphical Background

A traverse along the middle and upper reaches
of the Tangaraisu River gives the most complete
and least faulted section yet found in north-
central Guadalcanal, although faults are still
frequent. The sedimentary part of this section
was closely sampled by Coleman in 1961, Three
of the samples contained both planktonic and
larger and small benthonic species of foramini-
fera.

The geology cf the general area was described
by Coleman (1960). Figure 1 shows a map and
section of the Tangaraisu River area, based on
additional fieldwork, and also a general map of
Guadalcanal taken, with slight modification,
from the first regional geological map of the
British Solomon Islands (Coleman 1963b; Cole-
man et. al. 1963). The Miocene sedimentary
succession begins with the Betilonga Limestone,
a fringing-reef formation about 1200 feet thick.
At the base it has a foraminiferal fauna indicat-
ing an Aquitanian age, with Eulepidina, Spiro-
clypeus, Miogypsinoides dehaartii. Heterostegina
borneensis and Cycloclypeus eidae. At the top
it becomes finer grained, lacks reef characteris-
tics, and grades into the overlying Tina Calca-
renite which has a fauna more typically Burdi-
galian. It includes Lepidocyclina (N ephrole-
pidina) martini, L. (NJ radiata, L. ( N .) ferreroi
and Miogypsina polymorpha. The Tina Cal-
carenite is a medium to fine-grained sediment,
coarsely bedded with a minor terrigenous con-
tent of unaltered ferromagnesian and feld-

13

WOE

Pig.

1. Geology of the Tangaraisu River area with reference to a generalized geological map of Guadalcana

British Solomon Islands.

14

Qua ternary

voicar.ics
reef fimfcstanes
conglomerates

r77yVv|

Li.

Pliocene
- U.Miocene

| Middle
Miocene

Lower

Pre -Miocene

volcanic*, largely tuffs
marine, sediments

! Charikange Beds
1 Tanqaraisu Shale

| Tina Caicarenite
\ ftetUonga Limestone

IfSiUto Upp.Oiig. Uitra-basics
volcanic s -Saves, andesitic 5 ^5^
" “ 9 basaltic

gabbro

granitoid plutons

rmsL-i

i V V .

cO>

Pre-Tertiary Basement Complex

( igneous, some schists)

Volcano - • t with crater pres erved - ®

spathic grains derived from the andesitic base-
ment; it is about 1,400 feet thick. It contains
abundant planktonic as well as larger foramini-
fera but the former could not be extracted for
indentification. Towards the top it is marked
by shaley partings which, over a short vertical
distance, become predominant to form the base
of the Tangaraisu Shale. This is at least 2,000
feet thick, is argillaceous and finely bedded for
the most part, but includes many bands of fine-
grained lithic volcanic sandstone and occasional
conglomerate beds. Near the top of the for-
mation the sediment becomes generally coarse-
giained to form the Charikange Beds, a minor
unit about 500 feet thick. These beds are
banded but poorly bedded, show a good deal
cf lithological variation and include intrafor-
mational structures such as cross-bedding,
graded bedding, minor slump folds and displaced
small blocks. The banding is due to alternations
in grain size and in the proportion of shell frag-
ments. The samples to be described come from
near the base of the Charikange Beds. The
remainder of the sedimentary succession is made
up of the Toni Beds and Mt. Austen Beds, of
Pliocene age, and the Honiara Beds of Quater-
nary age. These consist of organogenic and vol-
canic cr volcanically-derived sediments.

In this part of Guadalcanal, therefore, the
sedimentary sequence from basement to the top
of the Charikange Beds is about 4,700 feet thick;
it shews a regular progression from reef to
deeper water sediments with increasing terri-
genous content and without evidence of a sedi-
mentary hiatus. The younger sediments were
deposited rapidly in fault-bounded troughs and
derived from an andesitic terrain which was
being uplifted and eroded rapidly. The Chari-
kange Beds mark a time of acceleration in this
action.

Description of the Samples

The samples have British Solomon Islands
Geological Survey collection numbers 4761, 4762,
and 4763, respectively; splits of them have been
retained by that Survey. They were taken from
an unfaulted cliff face, at about 80 feet vertical
intervals, approximately 550 yards upstream
from the junction of Charikange Creek and the
Tangaraisu River. The locality is shown in
Figure 2. Its co-ordinates on the 1,000 metre
Southern New Guinea Zone British Metric Grid,
Bessel Spheroid, are 4102650/810620, eastings
first.

The grain size in the samples ranges from
interstitial clay to rare rock fragments several
centimetres across. The larger grains are
usually of the order of 2-5 mm across and in-
clude larger foraminifera, shell fragments, frag-
ments of andesite and occasional basaltic rocks
(mostly lava types), dioritic rocks and, especially
important, eroded fragments of Tina Calcarenite
and Betilonga Limestone; of these two, the latter
is the most frequently represented. The rock
fragments are subangular to angular. The
mineral content consists of organic calcite,
hornblende, plagioclase (andesine-labradorite) ,
hypersthene, epidote, clinopyroxene, chlorite,
mica and quartz in that order of abundance.
The grains are fresh and many are euhedral.
Organic calcite varies from 30% to 60% of the
total weight. The cement is calcareous, usually
crystalline, but it may be in the form of clay.
Size analyses were made of two of the samples,
4761 and 4763. The former had a tenacious
calcite cement and could be disaggregated only
after treatment with hydrochloric acid. The
sample was a large block which included part
of a coarse band and part of an underlying fine
band. Each portion was analysed separately, the
coarse part being 4761 A, the fine part 476 IB.
Sample 4763 was broken down with considerable
difficulty, and to judge the effect on the sorting
of the foraminifera and other organic calcite, it
was examined in its original state (4763 A) and
after treatment with hydrochloric acid (4763B*.
The resultant distribution curves are shown in
Figure 3. The coarse part of 4761 is poorly
sorted whereas the fine part is well sorted.
Sample 4763 is moderately sorted although the
acid- treated part (4763B) shows that the
removal cf organic calcite improves the sorting.
This effect is confirmed by qualitative examina-
tion of the larger foraminifera present in the
untreated part of the sample; they range in
random fashion from clearly immature small
specimens, to large mature specimens. The
results of these analyses are not very significant,
but they do support the idea, deduced from a
variety of other evidence, especially field evi-
dence, that the Charikange Beds were deposited
under swiftly varying conditions and that they
are turbidites, at least in part. A more formal
attempt to determine the processes of deposition
of these beds, using the CM patterns of Passega
(1957), had to be abandoned because most of
the samples could not be broken down without
the use of hydrochloric acid, and this destroyed
the large and vital fraction made up of trans-
ported coarse shell fragments and other organic
grains.

15

DIAMETER IN MM.

Fig. 3.

The larger foraminifera include both original
specimens and those derived from the underlying
Tina Calcarenite and the Betilonga Limestone.
The latter are easily distingushed. They always
have at least a partial coating of crystalline
limestone; they are worn, frequently broken, and
have a distinctive crystalline infilling. The
remaining specimens are considered to be in-
digenous to the sediment for the following-
reasons: they are often well preserved so that
those with flanges or delicate surface ornament

are entire, contrasting with the shattered shells
of most of the small molluscan species; they
show a spread in sizes; some are partially air-
filled and where infilling is present it is a
yellowish clay, the same as that found in many
of the smaller specimens, including the plank-
tonic ones; the adherent matrix is common to
the sediment. Finally, these indigenous larger
foraminifera belong to species of Lepidocyclina
(including “Multilepidina”) , Miogypsina, Cyc -
loclypeus (including subgenus Katacycloclypeus) ,

16

Planorbulinella, Gypsina and Operculma, respec-
tively, and so comprise what might be termed
a normal fauna. This fauna has not been found
in exposures of older sediments on Guadalcanal
even in the underlying older part of the Tanga-
raisu Shale; in any event this contains the same
fauna cf planktonic foraminifera as that of the
Charikange Beds.

The fauna cf smaller benthonic foraminifera
was studied to obtain additional evidence on the
nature of the original sedimentary environments.
Thirty-nine of the sixty-two species of smaller
foraminifera have been found also in modern
sediments. Thirty-three of these species are
benthcnic forms, which provide information on
the depth and possibly the bottom temperature.

It must be realised, however, that this kind
of evidence has inherent limitations. Firstly,
there are huge gaps in our knowledge of the
ecological and geographical ranges of the species
mentioned due to the inadequacy of the sample
coverage of the Indo-Pacific province. Secondly,
the concepts of some of the species are by no
means clear. Thirdly, in the records of Recent
species in the Indo-Pacific no distinction has
been made between the occurrences of living
and dead specimens. Hence, contamination of
live assemblages by displaced dead specimens,
which results in mixed death assemblages, can-
not easily be recognised. Finally, many of the
bathyal records might report displaced occur-
rences which cannot be confirmed and so the
shallower occurrences are considered more re-
liable indices of the living environment of a
species.

The following extant benthonic foraminifera
are found in the Charikange Beds:

Amphicoryna scalaris (Batsch) 1791.

Amphistegina lessonii d’Orbigny, 1826.

Anomalinella rostrata ( Brady)/ 1881.

Bathy siphon fusca Cushman, 1927.

Bolivina hantkeniana Brady, 1881.

Bolivina schwageriana Brady, 1881.

Bulimina inflata Seguenza, 1862.

Cassidulina subglobosa Brady, 1884.

Cassidulina pad flea Cushman, 1925.

Ceratobulimina pacifica Cushman and Harris. 1927.

Cibicides refulgens de Montfort. 1808.

Cymbaloporella tabellaeformis (Brady), 1884.

Dentalina communis (d'Orbigny), 1826,

Dentalina subemaciata Parr. 1950.

Elphidium craticulatum (Fichtel and Moll), 1798.

Eponides margaritiferus i Brady), 1881.

Eponides praecinctus (Karrer)', 1868.

Eponides procerus (Brady >, 1881,

Eponides subornatus (Cushman), 1921.

Gyroidinoides neosoldanii (Brotzen), 1936 ( — Gyroi-
dina soldanii (d'Orbigny) as used by Brady in
1884).

Hyalinea balthica (Gmelin), 1791.

Islandiella limbata (Cushman and Hughes), 1925.

Nodosaria lamulifera Boomgaart, 1949.

Nodosaria pauperata (d’Orbigny), 1846.

Nodosaria spirostriolata Cushman, 1917.

Nonion pompilioides (Fichtcl and Moll), 1798.

PlanuLina ivuellerstorfl (Schwager), 1866.

Rectobolivina bifrons striatula (Cushman). 1917.

Rectobolivina columellaris (Brady), 1881.

Robulus costatus (Fichtel and Moll), 1798.

Robulus orbicularis (d’Orbigny), 1826.

StUostomella abyssorum (Brady), 1881.

Stilostornella consobrina (d’Orbigny), 1846.

Uvigerina spp.

Although there are more planktonic than ben-
thcnic specimens of foraminifera in the Chari-
kange fauna, the fauna is mixed and so the
evidence given by ratios of plantonic: benthonic
foraminifera cannot be used to suggest the depth
of deposition of the Charikange Beds. Con-

sidered separately, the planktonic species in-
dicate open-sea conditions and possiblv deep
water with surface temperatures of* about
25 -30 C. It could have been that conditions
weie similar to those off Guadalcanal today.

The benthonic element comprises a mixed
fauna, in which species indicative of shallow-
water environments are present with species
suggestive of deep-water conditions. Amphis-
tegina lessonii, Ancmalinella rostrata. Elphidium
craticulatum, Lepidocyclinidae, and Miogypsini-
dae indicate a shallow, warm water population.
Fuithei . E. craticulatum and A. rostrata seem to
be common in warm, tropical waters not deeper
than 40 fathoms in the Indo-Pacific. Such an
assemblage probably represents an environment
of similar depth to Polski’s (1959) Central Shelf
fauna (150-285 feet), but the bottom tempera-
ture probably was 20°-25"C.

A deep-water environment is suggested by the
species Cassidulina pacifica, Planulina wueller-
storfi, Bulimina inflata, Gyroidinoides neosol-
dcuiii. Nonion pompiloides, Rcctobolovina bifrons
striatula, Nodosaria spirostriolata, Stilostomella
abyssorum and S. consobrina. There are only a
few depth records available for the last four of
these species, all of them for depths of the order
of 500 to 700 fathoms. The other five species
have been recorded from shallow waters of less
than 100 fathoms, but by far the great majority
of their records are for depths of 500 fathoms or
more. By and large this assemblage of species
suggests an original environment roughly corre-

^ 0 m,n in A J° , Polski ' s Middle Bathyal fauna
<2.000-4.500 feet and bottom temperatures of
” C.). As well, the species Hoglundina

elegans occurs higher in the Charikange Beds
and this species has been recorded many times
from deep water only.

*1 faunal evidence at least indicates that
the Charikange Beds were formed by mixing of
sediments. Marine sediments deposited in a
shallow-water tropical environment, along with
sediments from intermediate depths, were dis-
placed to a depth of about 500 fathoms by
slumping prior to consolidation of any of the
sediments. Transportation by turbidity currents
would thoroughly mix the sediments from the
shallower depths, and these in turn would mix
with the deep-water sediments accumulating at
their ultimate position of rest.

Conclusions on the Charikange Beds

The combination cf evidence afforded by the
regional field study and the sedimentary and
faunal character of the samples examined, leads
us to the following conclusions:

(1) The Charikange Beds were deposited
rapidly in a tectonically unstable
environment: the bulk of them are first-
cycle sediments.

<2) Slumping or other transport played a
part in the accumulation of the Chari-
kange Beds.

(3) The association of larger and planktonic
f oi am in if era was not due to the deriva-
tion of the former type from older
consolidated sediments; it is best ex-
plained by contemporaneous transport
by slumping or turbidity flow, which

17

carried natural life-assemblages of
warm fairly shallow water benthonic
foraminifera and their containing sedi-
ments into deeper water. This trans-
port vook place long before there was
any consolidation of the sediments.

(4) The association now 7 to be described
was essentially contemporaneous.

The Association

The following species of planktonic and larger
benthonic foraminifera are found in samples
4761, 4762 and 4763. Qualitative estimates of
frequency are given for each species. For the
larger forms, if more than 25 specimens could
be picked from about 100 gms of sediment the
species is said to be abundant (a): if 11-25
specimens, common (c) ; if 5-10 specimens, fre-
quent (f) : if less than 5 specimens, rare <r);
and if only one specimen could be picked, very
rare (v). For the planktonic forms the same
freouency range is based on about 5 gms of
sediment.

Larger Foraminifera

4761

+76:

Opcirnlina cfinnplanatu japonica
LepitlocffLiim (“ MitUilcpidfna ") xutnitinxtx
Op firm Una renma

a

c

c.

1 f

C priori i/pcus (C.) hidopar tj\cn *

e

f

Lepffiiiff/filinit (.V.) japonic^

f

f

f

Mrotji/pttina poh/morp/tu

r

Jdnnorbulfnelht <•('. larvaftt

r

r

Cue. ( Kulaeyfilorlypfiup martini

r

r

(,'t/pshm sqvumifoi'rnw
Lf pidor ad inn (.V.) marl ini

r

l*

Pin tun bid i tudlti sp.

Uelerostleyina sp. ... ■

Planktonic. Foraminifera

(Jlohitje.rinoiitex conqlobahtx
0 -aid fix obfiynux

(i niff fist quadi ibibatus tm mat a rax
(} oidfis quad fit obatm ttumiUffir
O-fjirffitt quadrilobcttax hilobux
GlabafpuidriM u/t.ixpira ultixpirn
<: in a tffihhcen.it ad vena
<•'- inti flfihixrenx (Ifihisdttia
Cl obor Ottilia xcUuln
<r'. si i. cf. (>'. hmitlfi
OtbuHnct nmvensa
Puilmialhta obUqnitor alula
Sphneroi dindlopshs kocltt
.S'. xnni unit ml ....

*s. it tfbdfihhccit.it ...

Notes on the Species
Larger Foraminifera

Lepidocyclina (“Multilevidina’') suv aensis
(Whipple) 1934 — Plate 1, Figs. 9-14.

This is the same species as that described by
Whipple as L. ( Cyclolepidina) suvae?isis from
Fiji. It is found in other parts of the Indo-

Pacific region and is usually considered to be of
Burdigalian age. A number of specimens are
figured to show variation in the nature of the
nucleoconch and the usual presence of rays
which can often be seen in equatorial section;
this is a feature distinguishing it from otherwise
similar species, e.g. L. C‘ M ”) luxarians (Tobler)
or L. (“M.”) irregularis (Hanzawa). It should
be stressed that the nucleoconch with two large
chambers (Fig. 9) is a rare variant. Only two
of about 100 specimens are like this; the remain-
der have from 3 to 7 chambers (the actual num-
ber may depend on the precise lccaticn of the
thin section with respect to the plane of equa-
torial chambers).

Following the work of Van der Vlerk (1961)
the internal features of the type of L. radiata
(Martin) are now known for the first time.
Their resemblance to those of the Guadalcanal
specimens is striking as may be seen by a com-
parison between Van der Vlerk’s description and
illustrations with those in Coleman 1963a
(Plates 3 and 4: note especially Figure 5 on
Plate 4), and also the illustrations in this paper.
It m^y well be that the Guadalcanal specimens
should be referred to L. radiata. This possibility
is left standing until more is known of the
variation in tcpotypic specimens of that species.
Van de Vlerk considers that the type of L.
radiata came from sediments younger in age
than Burdigalian.

The above remarks contribute to the problem
of the significance of multilocular embryonic
chambers in Lepidocyclina, and as a corollary to
this, the concept and validity of Multilevidina
and Pliclepidina, respectively (for references, see
Cole 1963a and 1963b). Discussion of this prob-
lem, however, lies outside the aim of this paper.

Levidocvclina ( Nephrolepidma ) japonica Yabe,
1906 — Plate 1, Figs. 7, 8.

On Guadalcanal this species is also found in
the Tina Calcarenite and the upper part of the
Betilcnga Limestone, but not in the lower part
(of Aquitanian age) as stated previously by Cole-
man (1963a). Specimens which are clearly de-
rived from these sediments are found in the
Charikange Beds together with those considered
to be indigenous.

Lepidocyclina (N.) martini (Schlumberger) ,
1900 — Plate 1, Fig. 1.

The lenticular shape without a marked central
boss and the marked elongation of the equatorial
chambers along the rays distinguish this species
frcm the otherwise similar L. ( N.) radiata (Mar-
tin) as it is generally conceived. According to
Eames et al (1962) L. (NJ martini is a Vindo-
bonian species but it has been found as part of

PLATE 1

i • t prtid ocuclina ( Nephrolepidina ) martini (Schlumberger) Equatorial section of UWA 50411, X20; 2 : Planorbuli-
L>)lal£cfP larvala ( Parker and Jones) Transverse section of UWA 50412, X32; 3 : ? Planorbulinella sp. Median
cfpVriminf TJW A 50413 X20' 4 5 : Miogvpsina polymorpha (Rutten) 4, transverse section of UWA 50414, X25; 5.
r>art-median section of UWA 50415. X20; 6 : Heterstegina sp. cf. H. suborbicularis d’Orbigny Partial median section
ttw a 50416 X30 7 8 • Lepidocyclina ( N ephrolepidina ) japonica Yabe 7. vertical section of UWA 50417, X13,

Q 1 Jn notorial section of UWA 50418. X13; 9-14 : Lepidocyclina (" Multilevidina ”) suvaensis (Whipple) 9. central
LrUon of eQua^rial section, UWA 50419, X25; 10, vertical section of UWA 50420 Xll; 11. central portion of

equatorial section, immature specimen, UWA 50421, X19; 12. equatorial section (slightly oblique ) of ™ A 50422,
eu lunuufti , . _ j; i. i oirtnfT i *? Id ovtoniQl onnPEirnnrp nf TTW A 50423. X 5.

X19 showing elongation of hexagonal chamberlets along rays; 13, 14 external appearance of UWA 50423 X5,

•vnd UWA 50424 X5‘ 15 : Cycloclypeus (Cycloclypeus) mdopacificus Tan Median section of UWA 50425, X13,

Cvcloclvieuti^ martini Van der Vlerk 16. median section of UWA 50426. X12; 17. external

inoearance of broken specimen. UWA 50427, X4.5; 18-20 : Operculina complanata (Def ranee ) japornca Knnz^

ttw a ^ndOR vs- io median section of UWA 50429. X12; 20. median section of UWA 50430.

18 external appearance of UWA 50428, X5: 19, median section of UWA 50429, X12; 20,
1 ' X12, a more tightly colled specimen

18

r T f

v v r

i* ... I e

r .... 1 <•

r v

r ! r f

v

f f a

v i v r

r r j r

r I r

a c- a

f r c

Plate 1

19

typically Burdigalian assemblages (Coleman
1963a) in the Tina Calcarenite, the Bonege
Limestone, the Lake Lee Calcarenite (Guadal-
canal) and the Anuha Calcarenite (Florida).

Miogypsina polymorpha (Rutten), 1911 — Plate
1, Figs. 4, 5.

Specimens of this species derived from the
upper part of the Betilonga Limestone are also
found together with indigenous specimens. They
cannot be distinguished morphologically.

Cycloclyveus ( Cycloclypeus ) indopacificus Tan,
1932 — Plate 1, Fig. 15.

This species is common in the Indo-Pacific
region in sediments of Vindobonian age. but it
occurs also in Burdigalian sediments on Guadal-
canal (Bonege Limestone. Lake Lee Calcarenite)
on Florida (Anuha Calcarenite) and also on Fiji
(in the “ Lepidocyclina Tuff”, see Eames et al.
1957).

C. ( Katacycloclypeus) martini Van der Vlerk.
1923 — Plate 1, Figs. 16, 17.

Species of this subgenus are characteristic of
the Burdigalian. It has not been recognised with
certainty from other Solomon Islands sediments.

Operculina complanata japonica Hanzawa, 1935
— Plate 1, Figs. 18-20.

Specimens of this subspecies show great varia-
tion. the extremes of which are connected by a
continuous series of intergrading specimens. It
is present in flood proportions in the Charikange
Beds but is rare in other Solomon Islands sedi-
ments of about the same age.

O. venosa (Fichtel and Moll), 1798.

Probably the most common species found in
Solomcn islands Tertiary and Quaternary sedi-
ments, O. venosa is also present in great num-
bers in the present-day beach sands. The oldest
specimens do not appear to be in any way
different from the modern ones.

Planorbulinella sp. cf. larvata (Parker and
Jones), 1865 — Plate 1. Fig. 2.

Although referred to P. larvata this is almost
certainly a new species having as its major dis-
tinguishing feature a laminated central portion;
the laminations are the result of lateral exten-
sions of the chamber walls (see Coleman, 1963a)
and do not appear to represent a gerontic fea-
ture. It is especially characteristic of Burdiga-
lian to Quaternary sediments in the Solomon
Islands.

Planorbulinella sp. — Plate 1, Fig. 3.

This rare species has been seen in thin sections
of the Charikange samples.

Gypsina squamiformis Chapman, 1900.

The Moluccan form described under this name
by Bursch (1947) also occurs in the Charikange
samples.

Heterostegina sp. — Plate 1, Fig. 6.

Only two specimens of this species have been
seen, both in thin section. Its features are not
sufficiently clear to permit a specific identifica-
tion but it could well be H . suborbicularis
d’Orbigny, a Pacific region Quaternary species.
Cole (1957) records it from Eccene sediments
from Eniwetok Atoll.

Plantonic Foraminijera

Globigerinoides conglobatus (Brady), 1879 —
Plate 2. Figs. 15, 16, 21.

One poorly preserved specimen seems to belong
here. Banner and Blow (I960) gave the range
of G. ccnglobatus as Pliocene to Recent, and
they pointed out that certain Eocene and Lower
Miocene records attributed to G. conglobatus are
of other species. Belford (1962) recorded this
species from the Upper Miocene of Papua-New
Guinea, and Umbgrove (1931, p. 63) listed it
from Miocene T to Recent of Indonesia.

This species is most abundant in the equatorial
part of the Pacific, in waters ranging from 16 °C.
to 33°C. (Bradshaw 1959). In the Indian Ocean
it is present in the central and equatorial water
masses (Beliaeva 1962), where the surface water
temperatures range from 10 °C. to 28 °C.

Globigerinoides obliquus Bolli, 1957 — Plate 2,
Figs. 10, 17.

This species is rare in the Charikange Beds.
The absolute range recorded for this species in
the Caribbean region is from the Globorotalia
kugleri zone (Upper Oligocene) to the Globige-
rina bulloides zone of Upper Miocene age (Bolli
1957; Blow 1959). Belford (1962) has rare
specimens from the Middle Miocene of Papua.
Its range in the Solomon Islands is from Upper
Oligocene to Upper Miocene-Pliocene.

Globigerinoides quadrilobatus im mat urns
(LeRoy), 1939 — Plate 2, Figs. 11, 18.

This subspecies can be distinguished from G.
quadrilobatus trilobus (Reuss) by its more highly
arched umbilical aperture and the less embrac-
ing ultimate chamber although transitional
forms are difficult to place with certainty. In

PLATE 2

1 : Sphaeroidinellopsis subdchiscens (Blow) Umbilical view, UWA 49146, X75; 2-4 : ? Pulleniatina obliquiloculata
< Parker and Jones) 2, lateral view; 3. umbilical view; 4, dorsal view; UWA 49149, X75; 5, 8 : Sphaeroidinellopsis
seminuLina (Schwager) 5. umbilical view of multiloba- type, UWA 49159, X60; 8, umbilical view of typical form,
UWA 49147, X60; 6. 13. 14 : Globoquadrina dehiscens advena Bermudez 6, lateral view; 13, dorsal view; 14, umbilical
view; UWA 49143. X60; 7, 9 : Globigerinoides quadrilobatus trilobus (Reuss) 7. dorsal view; 9. umbilical view;
UWA 49139, X75; 10, 17 : Globigerinoides obliquus Bolli 10. umbilical view; 17, dorsal view; UWA 49141, X100:
11-18 : Globigerinoides quadrilobatus innnaturus (LeRoy) 11, umbilical view; 18, dorsal view; UWA 49137, X75;
12, 19 ; Globoquadrina dehiscens dehiscens (Chapman, Parr and Collins) 12, umbilical view; 19, dorsal view;
UWA 49144, X60; 15, 16, 21 ; Globigerinoides conglobatus (Brady) 15, umbilical view; 16, dorsal view; 21, lateral
view, UWA 49140, X60; 20, 23, 24 : Globoquadrina altispira altispira (Cushman and Jarvis) 20, umbilical view; 23.
lateral view; 24, dorsal view; UWA 49142, X75; 22 : Orbulina universa d’Orbigny UWA 49145, X50; 25, 26, 30 :
Sphaeroidinellopsis kochi (Caudri) 25. dorsal view; 26, umbilical view; 30. lateral view', UWA 49148, X60; 27. 31-33,
37, 38 : Globorotalia cf. G. tumida (Brady) 27, lateral view; 31. umbilical view; 32, dorsal view; UWA 49136b, X75.
33! lateral view; 37. dorsal view; 38, umbilical view; UWA 49136a, X100; 28. 28 : Globigerinoides quadrilobatus
sacculifer fBradv) 28, umbilical view; 29. dorsal view; UWA 49138. X75; 34-36 : Globorotalia scitula (Brady) 34,

umbilical view; 35, dorsal view; 36, lateral view; UWA 49094, X60.

20

Plate 2

21

the Solomons this subspecies ranges from the
Aquitanian to the Upper Miocene-Pliocene, but
previous records from the Indo-Pacific region
extend its range into the Upper Oligocene.

Globigerinoides quadrilobatus sacculifer (Brady),
1877 — Plate 2, Figs. 28, 29.

The specimens from the Charikange fauna are
most like the lectotype of G. quadrilobatus sac -
culifer chosen by Banner and Blow (I960).
Although a transitional series between G. quad-
rilobatus irregularis LeRoy and this subspecies
could be recognised in the Charikange fauna, no
attempt was made to separate these subspecies
in this work.

This cosmopolitan species is well represented
in Noegene to Recent sediments of the Indo-
Pacific region (Umbgrove 1931: LeRcy 1941;
Crespin 1963: Chang 1959: Belford 1962). Brad-
shaw (1959) found it in Pacific waters ranging
in surface temperature from 15 C. to 33 C., but
it was most common between 25 C. and 29 C.
In the Indian Ocean it is present in the tropical
and central parts with a temperature range of
IO C. — 28C., according to Beliaeva (1962).

Globigerinoides quadrilobatus trilobus (Reuss),
1850 — Plate 2, Figs. 7, 9.

This subspecies of G. quadrilobatus has been
separated from G. quadrilobatus immaturus by
characters of its final chamber, which is more
inflated and more embracing, and by the more
elongate, narrower primary aperture. A cos-
mopolitan subspecies, G. quadrilobatus trilobus
first appeared in the Lower Miocene and has
persisted to the present. It has been widely
recorded frcm the Indo-Pacific region.

Globcquadrina altispira altispira (Cushman and
Jarvis), 1936 — Plate 2, Figs. 20, 23, 24.

Although the poor preservation of specimens
assigned to Glcbcquadrina has destroyed or
obscured details of the umbilical teeth which
especially distinguish this genus, the specimens
possess ether features of the test which chaiac-
terise known specimens of Globoquadrina. in
particular G. altispira altispira.

This species is distinguished from G. dehiscens
dehiscens by its high trochoid spire and the
axially elongate chambers; an otherwise similar
subspecies. G. altispira globcsa. has globular and
less elongate chambers in the last whorl.
Although previous records indicate that G. alti-
spira altispira became extinct late in the Bur-
digalian or early in the Helvetian. Chang s
1 1959 ) Upper Miocene record from Taiwan and
Belford’s (1962) observation of the species in
strata dated as Pliocene from Papua-New
Guinea, might be taken as evidence for a
markedly longer range for G. altispira altispira
in the equatorial part of the Indo-Pacific than
elsewhere. However, Chang 1 is doubtful,

because two of his figured specimens (Chang
1959, Plate 2. Figs. 8a-9c) belong to other forms
of Globoquadrina. Belford’s identifications, on
the other hand, seem correct but the age deter-
minations of some of his younger faunas may
be in error. Although this evidence is equivocal,
additional records from Wreck Island (Lloyd
1961) and the Solomons (McTavish 1963)
showed that G. altispira altispira probably be-

came extinct late in the Tertiary T. It seems
likely that the range for this species is Aqui-
tanian to Tortonian.

Globoquadrina dehiscens advena Bermudez,
1949 — Plate 2, Figs. 6, 13, 14.

The range for this subspecies seems to be
closely similar to that of G. altispira altispira
except that it probably became extinct earlier in
the Tortonian, as it does in the Solomons.

Globoquadrina dehiscens dehiscens (Chapman,
Parr and Collins), 1934 — Plate 2, Figs. 12, 19.

The Charikange specimens generally agree
with those figured by Blow (1959) and Jenkins
(I960), in which the apertual face is neither as
prominent nor as smooth as that of the holo-
type. The first chamber of one specimen seems
to be divided into two, thus giving the appear-
ance of five chambers in the last whorl.

This species is moderately long-ranging with
a wide distribution. It ranges from Middle or
Upper Oligocene to Upper Miocene, but it is most
ccmmon in the Low : er and Middle Miocene. In
the Solomons G. dehiscens became extinct in the
Sphaeroidinellopsis seminulina fauna or Tertiary
Upper T age.

Glcborotalia scitula (Brady), 1882 — Plate 2,
Figs. 34-36.

A single, typical specimen of G. scitula has
been found in the Charikange fauna. Although
the range of G. scitula < sensu lato f is Aquitanian
to Recent according to Blow (1959), the species
is only present in Vindobonian and younger sedi-
ments of the British Solomon Islands. It has
been found in small numbers over a wide range
of temperatures <12°C.— 29 c C.) in the Pacific
(Bradshaw 1959).

Globorotalia sp. cf. G. tuinida (Brady), 1877 —
Plate 2, Figs. 27, 31-33, 37, 38.

This is an extremely variable species, and its
variation is much greater than that found in
Upper Miocene populations of G. tumida from
Guadalcanal. The more obvious variation is in
the shape of the chambers and the outline of
the periphery. However, there is variation in
the numbers of chambers in the last whorl,
commonly five to seven with eight in rare speci-
mens. Further, there is a range in the convexity
of the spiral, but spiro-convex specimens appear
to be rare.

Seme specimens with only a slightly convex
spiral side are strongly reminiscent of G. cul-
trata fijiensis Cushman, G. fohsi robusta Bolii,
and G. cultrata multi-earner ata Cushman and
Jarvis. In this respect, it is interesting to note
that Banner and Blow: (I960, p 27) have sug-
gested that G. fohsi robusta is pseudomorphous
after G. tumida. Still other specimens which
appear to be spire-convex are not unlike G. cul-
trata panda Jenkins. However, all these speci-
mens can be related to G. cf. G. tumida because
of their tumid tests and because there are speci-
mens gradational between them and forms more
like typical G. tumida. There are other speci-
mens that should perhaps be referred to G. cul-
trata menardii . They have raised sutures and
are less tumid than most other specimens from
the Charikange Beds. However, they are con-

nected to specimens like typical G. turnida by
transitional forms, so they have been included
in G. cf. G. tumida too.

Banner and Blow (1960, p. 27) suggested that
typical G. tumida did not become distinct from
G. cultrata menardii until Upper Miocene time
although representatives of it first appeared in
the Upper Tortonian. Evidence from Guadal-
canal supports this belief (McTavish 1963).
Indeed, G. cf. G. tumida might be nothing more
than a primitive population of G. tumida.

Globorotalia tumida is a tropical species. In
the Pacific it is most common between tempera-
tures of 29 °C. and 3l L C. and is not present in
water at surface temperatures less than 19 c C.
(Bradshaw 1959). Beliaeva <1962, p. 10) re-
corded it from the Arabian and equatorial waters
of the Indian Ocean where the surface-water
temperature ranged from 23 °C. to 28“ C.

Orbulina universa d’Orbigny, 1839 — Plate 2,
Fig. 22.

The specimens of O. universa are poorly pre-
served. The ultimate chamber of these speci-
mens appears to embrace the preceding cham-
bers completely and the test bears pore-like
apertures and finer pores over its surface.

Since LeRoy’s first important paper (LeRoy
1948) cn the OrbttZma-surfaCe a large literature
on the occurrence of this species has grown, and
it. seems widely accepted that pre-Miocene
records cf O, universa are based on misidentifi-
caticns or incorrect stratigraphical information.
Recent evidence (Carter 1958: Cita and Elter
1960: and Glaessner 1959, 1960) from widely
separated areas of the Indo-Pacific region and
Europe indicates that Orbulina definitely made
its entry in post-Aquitanian times and suggests
that it first appeared early in the Helvetian.
Pulleniatina obliquilcculata (Parker and Jones),
1862 — Plate 2, Figs. 2-4.

Specimens earlier referred doubtfully to
Globorotalia opima continuosa Blow (McTavish
1963, p. 294) now seem more likely to be primi-
tive forms of P. obliquiloculata rather reminis-
cent of Globigerina nipponica Asano. Possibly
this latter species is the juvenile stage of P.
obliquiloculata for it is markedly smaller than
adult specimens of this species and is not unlike
the assumed primitive forms of P. obliquiloculata
found in the Charikange Beds. Banner and
Blow (I960) believed that typical P. obliquilocu-
lata ranged from the Pliocene to the Recent.
However, it appears to range from late in the
Tortonian < Tertiary Upper T) to the Recent-
in the Indo-Pacific. region where it has been
widely recorded. Its range in the Solomons is
consistent with the latter view.

Sphaeroidinellopsis kcchi (Caudri), 1934 — Plate
2. Figs. 25, 26, 30.

This species is rare in the Charikange fauna.
It differs from S', seminulina in having five or
six, occasionally seven, chambers in the last
whorl, a more open umbilicus, arched aperture,
and radially elongate chambers. Glaessner
1 1943, p. 69) listed S. kochi as a Miocene guide-
fossil in the Indo-Pacific. Evidence from Vene-
zuela (Blow 1959) and the Solomons (McTavish
1963) suggests that this species existed only
during the Middle Miocene.

Sphaeroidinellopsis seminulina (Schwager), 1866
— Plate 2, Figs. 5, 8.

Sphaeroidinellopsis seminulina is the most
common planktonic species in the Charikange
fauna. Specimens of S. multiloba (LeRoy) have
been included in S. seminulina in the belief that
they represent a mature stage of this species.
Accordingly, S. seminulina has been recognised
by the three or four chambers in the last whorl
with the last chamber being considerably smaller
than the combined size of those preceding it.
Glaessner (1943, p. 69) listed this species as a
Miocene guide-fossil for the Indo-Pacific region,
where it has been recorded from such widely
separated areas as Borneo, Taiwan, New Zealand
and Australia. In the Solomons it is present in
sediments ranging in age from Burdigalian to
Tortonian. Rare specimens in Upper Miocene
regiments are probably reworked.

Sphaeroidinellopsis subdeliiscens (Blow), 1959 —
Plate 2, Fig. 1.

Banner and Blow (I960) chose S. subdeliiscens
as the type species cf their new genus Sphaeroi-
dinellopsis , which is distinguished from Sphae-
roidinella by its lack of supplementary apertures.
The last chamber of S, subdeliiscens is more or
less equal to the preceding two chambers and
this is the features which best distinguishes it
from 5. seminulina . However, transitional forms
between these species are present in the Chari-
kange Beds. S. subdeliiscens is confined to the
Middle Miocene and the early Upper Miocene.
In the Indo-Pacific region it has been found in
Taiwan. Sylvania Guyot and Papua-New Guinea.
In the Solomons it is present in the Middle Mio-
cene Globigerina nepenthes and Sphaeroidinel-
lopsis seminulina faunas and small numbers
have been recognised in the basal part of the
Upper Miocene Globigerina dutertrei fauna.

Age Significance

If considered alone, the larger foraminifera
from the Charikange Beds samples would be
thought to comprise a Burdigalian or an early
Vindobonian assemblage. The following species
are regarded as essentially Burdigalian ( Ter-
tiary f,.,) : Lepidocy cline (“Multilepidina” )

suvaensis , Miogypsina polymorpha, Cyclccly-
pens < Katacycloclypeus) martini, Lepidocy cliria
(Nephrolepidinaj japonica. A Vindobonian age
is suggested by the presence of L. (N.) martini
and C. (C. > in dopaci ficus; both these species may
be found in older sediments of Burdigalian age.
The suggestion of a late Burdigalian to early
Vindobonian age is a compromise and not a
studied calculation; it probably exceeds the
limits of refinement possible with larger for-
aminifera during this part of the Tertiary. So
far as the planktonic foraminifera are concerned
the abundance of species of Sphaeroidinellopsis,
especially S. seminulina, with species of Globo-
quadrina, indicate that this association should
be correlated with the Sphaeroidinellopsis semi-
nulina fauna as it is expressed in the Malaita
Group in the Solomon Islands. This Malaitan
fauna was correlated in turn (McTavish 1963,
pp. 68-69) with the Sphaeroidinella seminulina
zone of Venezuela (see Blow 1959) which is
shown by Blow and Banner (1962, o. 137) as
spanning the Tortonian-Sarmatian boundary.

23

The older age limit of this zone has not yet been
established and so it would be too specific at
this stage to restrict the Malaitan Sphaeroidinel-
lopsis seminulina fauna to the Tortonian, a
younger age than is indicated by the larger for-
aminifera, although the stratigraphic position
of the Charikange samples would support this
assignation. We have been content therefore to
describe this association as simply Vindobonian.
Nevertheless, if it is indeed true that the forami-
nifera preserved in the samples were contem-
poraneous, then the possibility remains that the
following species of larger foraminifera survived
into Tortonian time: Lepidocyclina (“Multile-
leyidina”) suvaensis (Whipple), Lepidocyclina
(NephrolepidinaJ japonica Yabe, Miogypsina
polymorpha (Rutten), Cycloclypeus < Katacyc-
loclypeus) martini Van der Vlerk, and Opercu-
lum complanata japonica Hanzawa.

Acknowledgments

We are grateful to Dr. M. F. Glaessner, Uni-
versity of Adelaide, for reading the manuscript
of this paper and for offering a number of con-
structive criticisms and helpful comments.

References

Banner, F. T. ( and Blow. W. H. (I960).— Some primary
types of species belonging to the superfamily
Globigerinaceae. Contr. Cushman Fdn .

Foram. Res. 11: 1-41.

Beliaeva, N. B. (1962). — Raspredelenie planktonnik fora-
minifer v vodakh i osadkakh Indiiskogo
Okeana. Dissertation, Geol. Instit. Akad.
Nauk SSSR.

Belford. D. J. (1962). — Miocene and Pliocene planktonic
foraminifera, Papua-New Guinea. Bull. Bur.
Min. Resour. Aust. 62-1: 1-51.

Blow, W. H. (1959).— Age, correlation and biostratigraphy
of the Upper Tocuyo (San Lorenzo) and
Pozon formations. Eastern Falcon, Venezuela.
Bull. Amer. Paleont. 39 No. 178: 67-251.

Blow, W. H., and Banner, F. T. ( 1962) .—“Fundamentals
of Mid-Tertiary stratigraphical correlation;
pt. 2. The Mid-Tertiary (Upper Eocene to
Aquitanian) Globigerinaceae.” (University
Press: Cambridge.)

Bolli, H. M. (1957). — Planktonic foraminifera from the
Oligocene-Miocene Cipero and Lengua for-
mations of Trinidad, B.W.l. Bull. U.S. Nat.
Mus. 215: 97-123.

Bradshaw. J. S. ( 1959).— Ecology of living planktonic
foraminifera in the North and Equatorial
Pacific Ocean. Contr. Cushman Fdn. Foram.
Res 10- 25-64.

Bursch, J. G. (1947). — Mikropalaeontologische Unter-
suchungen des Tertiars von Gross Kei
(Molukken). Abh. schweiz. palaont. Ges. 65:
1-69.

Carter, A. N. ( 1958 ) .—Pelagic foraminifera in the Ter-
tiary of Victoria. Geol. Mag. 95: 297-304.
Chang, L. S. (1959).— A biostratigraphic study of the
Miocene in Western Taiwan h ased on smaller
foraminifera (planktonic). Proc. Geol. Soc.
China 2: 47-72.

Cita. M. B. and Eher, G. (I960). — La posizione strati-
grafica delie marne a Pteropodi delle Langhe
e della Collina di Torino ed il significato
cronologico del Langhiano. R.C. Accad.
Lined (8) 29: 360-369.

Cole, W. S. (1957). — Larger foraminifera from Eniwetok
Atoll drill holes. U.S. Geol. Surv., Prof. Paper
260 -V : 743-783.

(1963a). — Illustrations of conflicting inter-
pretations of the biology T and classification
of certain larger foraminifera. Bull. Amer.
Paleont.. 46 No. 205: 1-63.

(1963b). — Analysis of Lepidocyclina radiata

(Martin). Bull. Amer. Paleont., 46 No. 208:
153-185.

Cole, W. S., Todd, R., and Johnson, C. G. (I960). —
Conflicting age determinations suggested by
foraminifera on Yap, Caroline Islands. Bull.
Amer. Paleont. 41 No. 186: 77-112.

Coleman, P. J. (I960). — The geology of Central- North
Guadalcanal. Rec. Geol. Surv. Brit. Solomon
Is. 1957/58 1: 4-13.

(1963a). — Tertiary larger foraminifera of the

British Solomon Islands, south-west Pacific.
Micropaleontology 9: 1-38.

(1963b). — Stratigraphic and structural notes

on the British Solomon Islands — with refer-
ence to a first geological map. Rec. Geol.
Surv. Brit. Solomon Is. 1959/62 2 (in press).

Coleman, P. J., Grover, J. C., Stanton, R. L. t and Thomp-
son, R. B. (1963). — A first geological map of
the British Solomon Islands Protectorate.
Rec. Geol. Surv. Brit. Solomon Is. 1959/62
2: (in press).

Crespin, I. (1953). Micropalaeontology. In: Condon, M.

A., Johnstone, D., Perry. W. J.. and Crespin.
I. The Cape Range Structure Western Aus-
tralia. Bull. Bur. Min. Resour. Aust. 21:
43-75.

Eames, F. E., Banner, F. T., and Blow, W. H. (1957). —
London Palaeontological note No. 171 (Bri-
tish Petroleum Company Ltd.) on fifteen
samples from the Fiji Islands. Rep. Geol.
Siirv. Fiji 42.

Eames, F. E., Banner. F. T., Blow, W. H.. and Clarke.

W. J. (1962). — “Fundamentals of Mid-

Tertiary Stratigraphical Correlation.” (Uni-
versity Press: Cambridge.)

Glaessner, M. F. ( 1943 ) .—Problems of stratigraphic cor-
relation in the Indo-Pacific region. Proc.
Roy. Soc. Viet. 55 (n.s.): 41-80.

(1959). — Tertiary stratigraphic correlation in

the Indo-Parific region and Australia. J
Geol. Soc. India 1: 53-67.

(I960).— West -Pacific stratigraphic correla-
tion. Nature, Lond. 86: No. 4730, 1039-1040.

Jenkins, D. G. ( I960).— Planktonic foraminifera from the
Lakes Entrance oil shaft. Victoria, Australia.
Micropaleontology 6: 345-371.

LeRoy, L. W. ( 1941 ) .—Small foraminifera from the late
Tertiary of the Netherlands East Indies.
Colo. S'ch. Min. Quart. 36: (1).

(1948). — The foraminifer Orbulina universe

d'Orbigny, a suggested Middle Tertiary’ Time
Indicator. J, Paleont. 22: 500-508.

Lloyd, A. R. (1961). — Foraminifera from the sub-
surface Miocene of Wreck Island, Queens-
land. Thesis, Univ. Adelaide.

McTavish, R. A. (1963). — Smaller Foraminifera from the
British Solomon Islands. Ph.D. thesis, Univ.
W. Aust.

Passega, R. (1957). — Texture as characteristic of classic
deposition. Bull. Amer. Ass, Petrol. Geol. 41:
1952-1984.

Polski, W. (1959). — Foraminiferal biofacies off the North
Asiatic coast. J. Paleont. 33: 569-587.

Umbgrove, J. H. F. (1931).— Tertiary foraminifera. Leid.
Geol. Meded. 5: 35-91.

Van der Vlerk, I. M. (1961). — Lepidocyclina radiata (K,
Martin), 1880. Proc. Acad. Sci. Amst. Ser. B..
64: 620-626.

24

5.— Discontinuous and Presumed Vicarious Plant Species in Southern

Australia

By J. W. Green*

Manuscript received — 18th June, 1963

The details of distribution of discontinuous
and presumed vicarious species pairs in south-
western and south-eastern Australia are pre-
sented. Selected discontinuous species are
listed and their distributions shown on outline
maps, and these species are discussed in rela-
tion to the geological and climatic history of
the Australian continent. It is suggested that
some disjunctions may have their explana-
tion in long-distance wind dispersal. Presumed
vicarious species known to occur in the two
regions are listed, and are discussed in relation
to discontinuous species and geological and
climatic history.

Introduction

Comparison of the vascular floras of south-
western and south-eastern Australia has re-
vealed the existence of several hundred species
in common, of which some 35 have been
selected for study according to the principles
outlined in the section below entitled “Discon-
tinuities between South-western and South-
eastern Australia”.

All are autochthonous species whose disjunc-
tion is well marked and which have no obvious
mechanism whereby they may be dispersed over
long distances by water or by animals. All but
a few are restricted to the temperate area of
southern Australia shown in Figure 1.

In addition, about 50 pairs of presumed vica-
rious species have so far been noted in the two
regions.

The only previous comparison of the vascular
plants of the two regions at the specific level
appears to be that of Hooker (I860), who
stated that 83 species were common to south-
western and south-eastern Australia. He gave
the number of such species for each genus but
mentioned no specific names.

Discontinuous Species

The area occupied by a species is usually said
to be discontinuous (or disjunct) if it is broken
into two or more portions which are separated
by a distance exceeding the “normal dispersal
capacity” (Cain 1944) of propagules of the
species. The determination of dispersal capacity
must depend on experimental data, but an esti-
mate of dispersal capacity may be obtained
from the size and morphology of the propagules.
Discontinuity may be assumed if the actual dis-
tance separating the populations exceeds this
estimate.

Minor Discontinuities in Eastern Australia

In central and south-eastern Australia several
examples of discontinuities are well known.
Eucalyptus cladocalyx is of particular interest

* Department of Botany, University of New England,

Armidale, New South Wales. Present Address: Botany

Department, School of General Studies, A.N.U., Can-
berra.

because of its discontinuous occurrences on
Eyre Peninsula, Kangaroo Island and in the
Flinders Range, even though the tree has been
planted successfully in intervening areas, prov-
ing the suitability, at least for growth, of such
habitats.

Other examples are Acacia pence, occurring
in several separated localities in central Aus-
tralia and south-western Queensland, Eucalyp-
tus globulus and E. regnans, both occurring in
south-eastern Australia and Tasmania, Dillwy-
nia orecdoxa, restricted to the Victorian Gram-
pians and the Braid wood -Clyde Mountains area
of New South Wales, Schoenus turbinatus,
Lasiopetalum ferrugineum and Phebalium den-
tatum, all of which are discontinuous between
the Sydney district and the Gibraltar Range
in northern New South Wales, and Eucalyptus
nitens, which is discontinuous between the
northern and southern tablelands of New South
Wales.

Specht et al. (1961) have described the dis-
junct distribution of Eucalyptus elaeophora
(now to be known as E. goniocalyx, according
to Johnson 1962) in South Australia, Victoria
and New South Wales, while Willis (1962)
mentions several examples of Tasmanian plants
which are found only in the Grampians in main-
land Australia <e.g. Leptospermum nitidum
and Pomaderris apetala), a number of disjunc-
tions between the Grampians and the area from
East Gippsland to central coastal New South
Wales (Psilotum nudum, Davallia pyxidata,
Howittia trilocularis , Dodonaea truncatiales , and
Westringia glabra J and one species which is
unknown between the Grampians and north-
eastern New South Wales ( Swainsona brachy-
carpa ) .

The distances between the disjunct areas of
the above species are mostly smaller than those
of the species mentioned below\ and some may
well prove to be continuous w r hen more infor-
mation on dispersal capacity becomes avail-
able.

Discontinuities between South-western and
South-eastern Australia

The examples listed below were obtained by
a study of the published literature, combined
with the examination of specimens from several
Australian herbaria. No critical taxonomic
work has yet been done, but in most cases
specimens from the south-western and south-
eastern populations of each species have been
compared in their gross morphology.

On the basis of distance, these may be classed
as major discontinuities. Most are separated
by about 750 miles, and a few by a greater
distance. It is not considered likely that

25

Fig. 1

additional collections would extend the ranges
much further into the intervening area; the
Nullarbor Plain is noted for its paucity of
species, and the sclerophyll communities in
which the majority of the species occur are
certainly absent. In any case the existing
habitat conditions suggest that their survival
would be impossible.

It is considered that the distances involved
here are probably greater than the normal
dispersal capacity of propagules of these species.

In selecting a short list for study, the follow-
ing classes of plants were eliminated from the
several hundred apparently discontinuous
species mentioned in the Introduction: (i)

species capable of growing in semi-arid regions,
because of their potential capacity of migration
between the two regions by marginal spread;
(ii) littoral species, because they are presum-
ably capable of marginal spread along the
coast, or dispersal by sea water or animals of
the shore, without any great problem in estab-
lishment; and (iii) aquatic species of rivers and
lakes, again without any apparent problems in
dispersal or establishment. For the sake of
simplicity, species having any occurrence out-
side temperate Australia and Tasmania have
also been eliminated. The remainder, then, are
those species whose disjunctions are most diffi-
cult to explain.

The species are listed below and their ranges
are shown on the maps in Figures 2-3.

LILIACEAE

Borya nitida Labill.

Calectasia cyanea R.Br.

Lomandra micrantha (Endl.) Ewart
Thysanotus tenellus Endl.

T. sp. (undescribed)

IRIDACEAE

Orthrosanthus multiflorus Sweet
ORCHIDACEAE

Caladenia latifolia R.Br.

C. mensiesii R.Br,

Corybas dilatatus Rupp et Nicholls
Leptoceras fnnbriatum Lindl.

Microtis atrata Lindl.

M. orbicularis Rogers
Pterostylis robust a (Ewart) Rogers
P. vittata Lindl.

Thelymitra antennifera Hook.f.

T. flexuosa Endl.

T. fusco-lutea R.Br.

T. macmillanii F.Muell.

T. rubra Fitzg.

PAPILIONACEAE

Daviesia brevifolia Lindl.

Dillwynia cinerascens R.Br.

D. uncinata (Turcz.) J.M.Black
Sphaerolobium daviesiodes Turcz.

S. vimineum Sm.

RUTACEAE

Microcybe pauciflora Turcz.
POLYGALACEAE

Comesperma polygaloides F.Muell.
RHAMNACEAE

Cryptandra leucophracta Schlecht.
STERCULIACEAE

Thomasia petalocalyx F.Muell.
EPACRIDACEAE

Acrotriche cordata (Labill.) R.Br.
Leucopogon hirsutus Sond.
LOGANIACEAE

Logania vaginalis (Labill.) F.Muell.
STYLIDIACEAE

Levenhookia pusilla R.Br.

Stylidium perpusillum Hook.f.
COMPOSITAE

Lagenophora huegelii Benth.

26

27

28

There is no marked tendency for the majority
of the species studied to be more widely dis-
tributed in either south-western or south-
eastern Australia. Borya nitida, Calectasia
cyanea and Orthrosanthus multiflorus are prin-
cipally western with a limited occurrence in the
east, while the reverse is true of Sphaerolobium
vimineum, Comesperma polygaloides, and many
of the orchids. Leucopogon hirsutus is very
restricted in both regions. Three species,
Lomandra micrantha, Dillwynia cinerascens and
Sphaerolobium daviesioides have noticeably
different forms in the two areas, the second
being discussed in more detail below. Thely-
mitara fLexuosa occurs in three separated areas,
the eastern disjunction being over 750 miles.
The widest disjunction is that of Stjlidium per-
pusillum, which is not recorded between Bussel-
ton, Western Australia, and western Victoria, a
distance of 1,560 miles.

It has been possible to pay special attention
to the two species of Dillwynia which are dis-
continuous between the south-western and
south-eastern regions. A large range of speci-
mens has been examined from both regions
showing that D. uncinata is markedly uniform
in gross morphology over its entire area, while
the disjunct populations of D. cinerascens seem
to be morphologically distinct. All western
specimens of the latter have spinescent branch-
endings while all eastern specimens have none,
and it appears that the two populations repre-
sent a pair of vicarious subspecies.

Presumed Vicarious Species

Closely allied to the concept of discontinuous
species distributions is that of vicarious pairs or
sets. The history of the latter has been sum-
marized by Cain (1944) who defines vicarious
species as “closely related allopatric species
w T hich have descended from a common ancestral
population and attained at least spatial isola-
tion.” They are thus the equivalent of the
Artenkreise of Rensch. As pointed out by Vier-
happer (1919) (quoted by Love 1954), it is
possible to visualise two ways in which corre-
sponding taxa in different areas may have
arisen: (i) as true vicariads , which have pene-
trated into a new 7 area and later become
differentiated, and < ii ) as false vicariads or sub-
stitution taxa, which have differentiated prior
to their occupation of new 7 areas. Love is of
the opinion that the terms vicariism , vicariad
and vicarious should be confined to the second
type, and suggests that where lack of infor-
mation permits no such subdivision, only the
collective term corresponding taxa should be
used. This would apply to the Australian
examples quoted below 7 .

The concept of vicariism has been applied to
taxonomic ranks other than species but, as
mentioned by Turrill (1959), the phytogeo-
graphical significance of vicarious families,
tribes or genera is often obscure. The concept
has also been applied to communities.

The accompanying list (Table I) show 7 s pairs
of species in which the members of each pair
appear to be closely related systematically and
far removed geographically. In most cases
Bentham’s Flora Australiensis has been the
primary source of information, and it has been
verified that Bentham considered the species

of each pair closely related. In some cases
(marked with an asterisk) Bentham actually
commented on the closeness of the relationship.
Wherever appropriate more recent monographs
and revisions have been consulted.

South-Western species

South-eastern species

CY PER ACE A E

Lepidosperma angustutum R.Br.

L. a men rum R.Br.

/,. granite R.Br.

L semiterex E. Mueil. ex B< ieek

L . leptophyll urn Be n 1 1 i .

L. lorluomm F. Mueil.

LI El ACE A E

Lomandra caeupitosa (Benth.)

L. sororin (F. Mueil. ex Benth.)

Ewart

Ewart

Xanthorfhoea preissii Endl.

X. quadrangulata F. Mueil.

CAS CAR IN' AC EAE

*Casuarina decussala Benth.

C. torulosa

('. trichodon Mit|.

G. strieta Ait.

PROTE.

U EA E

A denantbos flavidiflora F. Mueil.

A. terminals R.Br.

Lam bert in m alt (flora Lindi.

L. formosa Sm.

DROSKKACKAE

* Drosera rosmlatn Echm.

1). whittakeri Planeh.

1). memiesii R.Br.

I). plaiichonii Hook.f.

MIMOSACEAE

Acacia r.nch/irarva Meissn.

A. daUaehiana K Mueil.

A . direr gem Benth.

A. vomeriformix A. Cunn.

•.!. leptoneura Benth.

A. ripens A. Cunn.

A . pentadenia Lindi.

A. miteheUii Benth.

PA PI I.IONACKAE

Bossiaea spinoxa (Turcz.) Domin
H. pedunculari s Turcz.

B. tufa R.Br.

B. biloba Benth.

Chori?ema genistioides (Meissn.)

B. foliom A. ('mm.

B. microphyllu Sm.

II. helerophylla Vent.
B. cinema R. Br.

C purviflorum Benth.

C. A. Gardner

Daviesia anceps Turcz,

1). alata Sm.

Jacksonia sternbergiana Hucg.

./. darkii F Mueil.

Mirbelia orata Meissn.

M. oridobioides F. Mueil.

Oxylobium microphyllum Benth.

O. cord i folium Andr.

0. trieuspidatim Meissn.

(). procumbent F. Mueil.

Pultenaea spinulosa (Turcz.)

P. tend la Benth.

Benth.

RUTACKAE

Boronia alata Sm.

II. algida F. Mueil.

B. erenulata Sm.

II. serrnlata Sm.

B. peniciUata. Benth.

H. falcifolia A. Cunn.

B. riminea Lindi.

B. parii flora Sm.

STACK HOI

SI ACE A E

*Stackhousia pubescent A. Rich.

•S’, monvgyna Labill.

RHAMXACKAE

Spyridium complication E. Mueil.

-S', coadilfolium Reiss.

STKKCI LI ACE A E

Lasiopetalum acutiflorum Turcz.

L. ferrugineutn Sm.

D1XEKN IACEAE

nibbertia grand ipe.s Benth.

II. billardieri F. Mueil.

II. ere nut a Andr.

II. dentatu R.Br.

II. inelusa Benth.

II. rirgata R.Br.

II. inner attain (Turcz.) Benth.

//. aeicualrix F. Mueil.

THYMJBL ABAC EAE

Pimelea preissii Meissn.

P. striela Meissn.

M YRTACEAE

l)a ruin ia xanguinea (Meissn.)

D. mirropelala ( F. Mueil.)

Benth.

Benth.

Eucalyptus ealopliylla Lindi. \

E. Jicifolia F. Much. /

/ E. qummifera (Gaertn.) Ifoehr.
\E. intermedia R. T. Baker

Leptospermum erubexeens Sehau.

L. atlemmlum Sm.

M (cromyrtus drummondii Benth.

M. mmutiflora Benth.

EPACK1DACEAE

• Leucopogon ronoste phio ides DC.
* L. conostephioides DC.

L. rufus Lindi.

L. de for mis R.Br.

L. gilbertii Stschegl.

L. concur vus F. Mueil.

SOLAN ACE A E

Anthocercis microphgUa F. Mueil.

A. myosolidea F. Mueil.

GOODEN IACEAE

Dampiera altissima F. Mueil. ex

D. putt (folia Benth.

Benth.

29

This is believed to be the first published list
of presumed vicarious species pairs between
south-western and south-eastern Australia.
Wood and co-workers ‘Wood and Baas-Bscking
1937, Wood 1949 and Wood and Williams I960'
have compared the species occurring in dry
sclerophyll forests of South Australia. New
South Wales and the Australian Capital Terri-
tory, but their examples of supposed vicarious
species pairs appear to be little more than rep-
resentative species of the various genera,
chosen without regard to systematic affinity or
even total geographic range. Three species of
Acacia which are quoted are A. myrtifolia (rep-
resenting South Australia, even though it also
occurs in the other regions), A. discolor (New
South Wales) and A. falciformis <A.C.T. >. One
of these is bipinnate and the two phyllodineous
species are rot closely related systematically.

Discussion

The present study lias revealed the occur-
rence of geographically separated but system-
atically related populations in south-western
and south-eastern Australia which show many
different degrees of relationship. In some cases
it has not been possible to differentiate the
populations from the two areas, in others mor-
phological differentiation has been at an infra -
specific level, while further examples are given
in which full specific distinctions are recog-
nized, even though the populations are suffi-
ciently similar to be considered true vicariads
(i.e., presumably derived from an immediate
common ancestor).

The existence of differentiated populations is
most easily explained in terms of the Tertiary
and Quaternary history of southern Australia,
but in the case of discontinuous species in which
the populations are morphologically indis-
tinguishable it is felt that comparatively recent
long-distance dispersal should be considered as
a possibility.

It is likely that many of the species referred
to above occupied former continuous areas which
became broken up by the onset of unfavourable
climatic conditions. Unfortunately, knowledge
of the Australian flora of the upper Tertiary.
Pleistocene and early Recent is scanty compared
with that of the lower Tertiary. It is thought
that the mid-Tertiary flora was a mesic one,
that peneplanation was widespread and that the
climate was humid and warm. Peneplanation
was modified by vertical land movements begin-
ning in the Miocene and culminating in the
Upper Pliocene or Pleistocene, but the conse-
quent climatic and edaphic changes are con-
sidered inadequate to explain the discrepancy
between the distributions of Tertiary and
present floras (Crocker 1959).

Crocker and Wood (1947) have suggested the
existence of a Recent arid period, sudden and
drastic enough to have had profound effects on
a pan-Australian flora in the southern part of
the continent. Their picture of the retreat of
the pre-arid flora to refuge areas such as the
Stirling Range, Mount Lofty Range. Grampians
and Flinders Range is borne out by the present
study of discontinuous species, many of which
have been collected from two or more of these
areas. Willis (1962) mentions several Gram-

pians endemics “having undoubted Western
Australian affinities." Crocker (1959) quotes
the existence of “disjunct vicarious pairs, . . .
major species disjunctions and the occurrence
of relic species’’ as good evidence for the thesis
of retraction and expansion.

Herbert (1929) quotes several species of
Eucalyptus (e.g. E. diversifolia and E. flock-
toniaeJ which are discontinuous between west
and east, and suggests that their distribution is
best explained by the onset of arid conditions
in a previously well-watered central zone. He
does not suggest a time for the climatic change.

Smith-White (1954) suggests that the areas
of many species may have been simply bisected
by the Miocene inundation of the Nullarbor
Gulf, aided by a tract of arid country to the
north, and that the bisection has been main-
tained to the present “in turn by physiographic,
edaphic and climatic barriers".

Burbidge (1960) favours rather the late
Pleistocene as the time of separation, on the
basis of discontinuities between the two areas at
the specific level.

The many degrees of morphological diver-
gence between western and eastern populations
may suggest that not all species were separated
simultaneously. Stebbins (1950) points out that
explanations in terms other than formerly
continuous areas are possible, and mentions the
possibilities of populations having always been
separated although not to such a degree as at
present and of the former existence of “stepping
stones". The two extreme hypotheses (geo-
logical history and more recent long-distance
dispersal) are not mutually exclusive, and com-
binations of the two are possible.

Overseas work on discontinuous species dis-
tributions has been summarised by Cain (1944).
Most explanations have been based on past
geological history, although often without direct
fossil evidence. In some cases the discovery of
fossil records from areas outside the living
range of a species has given irrefutable evidence
of contracting areas, but the necessary specific
identification of the fossil specimens presents a
formidable problem. In Australia we have no
such records at the specific level which are
likely to help elucidate west-east discontinuities.
At the level of the subgenus, the occurrence of
fossil leaves of corymbose eucalypts in Tas-
mania may point to the contraction of a former
area occupied by a group of species, but gives
no direct evidence at the specific level.

While geological history must be considered
in explaining most discontinuities between
south-western and south-eastern Australia, the
possibility of propagules of certain species
travelling long distances cannot be ruled out.
Most objections to the general hypothesis of
distance dispersal concern the establishment of
propagules in a foreign environment, in com-
petition with local species. Turrill (1959) has
mentioned the ability of polyploids to extend into
habitats unfavourable to their diploid progeni-
tors but cautions against the drawing of broad
conclusions. The role of vegetative reproduc-
tion as an aid to establishment is discussed by
Baker (1953), and facultative apomixis could

30

well be important. Taylor (1955) has discussed
the establishment of alien species under natural
conditions on Macquarie Island.

The instances of Eucalyptus cladocalyx and
E . ficifoiia , whose natural ranges are much
smaller than in cultivation, may indicate a
low capacity for dispersal or establishment,
although these distributions may be of “young"
species which have not had time to occupy all
available habitats.

It will be noticed that a high proportion of
orchids occurs in the list of discontinuous
species set out above. Seeds of Microtis atrata
were measured and found to be about 200 x 65 g
while those of Thelymitra flexuosa were about
180 x 100 fi. The smallest non-orcidaceous
seeds were those of L evenhookia pusilla (about
500 x 220 u). There seems no reason why seeds
of this magnitude should not be carried long
distances in the atmosphere. Ridley (1930)
quotes examples of mineral particles 1 200th in.

< 1,270 diameter being identified 970 and 600
miles from their respective sources and con-
cludes by saying that “dust seed (Orchidaceae)

, . . may travel a distance of as much as 700
miles*’ in one flight. Taylor (1955) suggests
that species with very small seeds may have
arrived on Macquarie Island as a result of
wind transport, quoting as evidence the dis-
covery of pollen grains of Podocarpus 600 miles
from their nearest source of supply. Small
(1921) quotes an experiment in which it was
found that a light breeze of about two miles per
hour is sufficient to support a dandelion fruit
in the air indefinitely.

Accepting for the moment the possibility of
wind dispersal having operated between south-
western and south-eastern Australia, there is
no climatological information to suggest the
more likely direction. According to Kendrew
(1937) and Gentilli (undated), southern Aus-
tralia is characterised by prevailing winds in a
general westerly direction in summer and
easterly in winter, but there is much day to day
variation. It is often assumed that the south-
western region served as a centre of origin of
many autochthonous species, on the basis of the
high proportion and number of endemic genera
found there.

There do not appear to be any well defined
animal migration routes which could have
carried propagules between the two areas in
recent times. Aborigines are not likely to have
carried seed deliberately for any distance
although it is suggested that accidental carriage
by aborigines may have occurred during their
8,000 year occupancy of Australia. Alter-
natively, early human occupants of this country
may have disturbed the habitat sufficiently to
create favourable conditions for the establish-
ment of alien propagules carried by long dis-
tance wind dispersal.

With regard to the vicarious species, the main
problems are the verification of true systematic
affinity between the members of each pair, and
the determination of whether the species are
true or false vicariads. Love (1954) has applied
cytotaxonomic studies to corresponding types
from North America and Europe and has shown
92 pairs of truly vicarious taxa and 41 pairs of
substitution species, having different chromo-

some numbers. He has found that both vicarious
and substitution types are to be found at
different stages of separation, from habitat
separation in the same region to physiographic
or historic separation in different regions.

It is envisaged that future investigations will
be undertaken in the following stages, with a
view to obtaining further information relative
to the foregoing observations:

<i> Field collection of material for detailed
comparison of morphology, anatomy, cytology,
breeding systems and habitat preferences (the
information so far collected from herbarium
labels has proved too scanty to provide useful
comparisons of habitats).

(ii> Investigation of survival of small seeds
in the laboratory under conditions of tempera-
ture and humidity likely to be encountered on
a transcontinental journey.

(iii) Cultivation and crossing of western and
eastern individuals, in order to obtain some
measure of genetic divergence.

Acknowledgments

I wish to thank the curators of the Sydney,
Melbourne, Adelaide and Perth Herbaria for
permission to consult their collections and for
providing further information on distributions
in many cases. In particular the assistance of
the following persons is gratefully acknow-
ledged: Mr. J. H. Willis of the Royal Botanic
Gardens and National Herbarium, Melbourne,
and Professor N. C. W. Beadle, Associate Pro-
fessor G. L, Davis, Mr. J. B. Williams, and Mr.
R, A. Boyd, all of the Department of Botany,
University of New England.

References

Baker, H. G. (1953). — Race formation ancl reproductive
method in flowering plants. S.E.B. Sym-
posiu n "Evolution.” (C.U.P.)

Bentham, G. (1363-78). — "Flora Australiensis.” (Reeve:
London. )

Burbidge, N. T. (1960). — The phytogeography of the
Australian region. Aust. J. Bot. 8: 75-212.
Cain, S. A. (1944). — “Foundations of Plant Geography.”
(Harper: New York.)

Crocker. R. L. (1959). — Past climatic fluctuations and
their influence upon Australian vegetation.
In “Biogeography and Ecology in Australia .”
(Junk: Den Haag.)

Crocker. R. L., and Wood. J. G. (1947).— Some historical
influences on the development of the South
Australian vegetation communities and
their bearing on concepts and classification
in ecology. Trans. Roy. Soc. S. Aust. 71:
91-136.

Gentilli. J. (undated). — "Australian Climates and Re-
sources." (Whitcombe and Tombs: Perth.)
Herbert, D. A. (1929). — The major factors in the present
distribution of the genus Eucalyptus. Proc.
Roy. Soc. Qd 10: 165-93.

Hooker. J. D. 1 1860 Botany of the Antarctic Expedi-
tion III. Flora Tasmaniae 1. Introductory
Essay.

Johnson, L. A. S. (1962). — Studies in the taxonomy of
Eucalyptus. Contrib. N.S.W. Nat , Herb. 3:
103-26.

Kendrew, W. G. (1937).— “The Climates of the Con-
tinents.” (Oxford University Press: London.)
Love. A. (1954). — Cytotaxonomical evaluation of corres-
ponding taxa. Vegetatio 5-6: 212-24.

Ridley. H. N. ( 1930).— "The Dispersal of Plants through-
out the World.” ( Reeve: Kent.)

Small, J. (1921). "A Textbook of Botany.” (Churchill:
London. )

Smith-White, S. (1954). — Chromosome numbers in the
Boronieae (Rutaceae) and their bearing on
the evolutionary development of the tribe
In the Australian flora. Aust. J. Bot. 2:
237-303.

31

6pecht, R. L. Brownell, P. F., and Hewitt, P. N. (1961).—
The plant ecology of the Mount Lofty
Ranges, South Australia. 2. The distribution
of Eucalyptus elaeophora. Trans. Roy. Soc.
S. Aust. 85: 155-76.

Stebbins, G. L. (1950). — “Variation and Evolution in
Plants.’’ (Columbia University Press: New
York.)

Taylor, B. W. (1955). — The flora, vegetation and soils
of Macquarie Island. Aust. Nat. Antarct.
Res. Exped. Rep. Ser. B., vol. 2.

Vurrill, W. B. (1959). — Plant Geography. In “Vistas in
Botany.” (Pergamon: London.)

Vierhapper, F. (1919). — “Uber echten und falschen
Vikarismus. Oest Bot. Zeitschr. 68: 1-22.

Willis, J .H. (1962). — Land flora of Victoria. From Viet.
Year Bk. No. 76.

Wood, J. G. (1949). — Vegetation of Australia. In “The
Australian Environment.” (C.S.I.R.O. and
Melbourne Univ. Press: Melbourne.)

Wood, J. G., and Baas Becking, L. G. M. (1937). — Notes
on convergence and identity in relation to
environment. Blumea 2 : 329-38.

Wood, J. G., and Williams, R. J. (1960). — Vegetation. In
“The Australian Environment.” (C.S.I.R.O.
and Melbourne Univ. Press: Melbourne.)

32

INSTRUCTIONS TO AUTHORS

Papers may be submitted to the Society in accordance with Rules and
Regulations 38 to 41 inclusive (see below). They should be addressed to The
Honorary Secretary, Royal Society of Western Australia, Western Australian
Museum, Perth.

Authors are solely responsible for the factual accuracy and for any opinion
expressed in their papers. They are particularly requested to verify references.
Alterations to MSS. submitted to the printer will be allowed only under excep-
tional circumstances, and no changes will be permitted after galley-proof stage.

In the preparation of MSS. authors are required to follow the C.S.I.R.O.,
Guide to Authors (C.S.I.R.O., Melbourne, 1953), except that papers longer than
10,000 words (30 foolscap pages of pica type, with 6 inch lines 40 to the page) will
not normally be accepted.

Authors may be required to meet half the cost of preparation of the blocks
of diagrams and illustrations.

Authors shall receive a total of 30 reprints free of charge. Further reprints
may be ordered at cost, provided that such orders are submitted with the MS.

RULES AND REGULATIONS

38. Every paper intended to be read before the Society or to be published
in the Society’s Journal must be sent to the Secretaries at least seven days
before the date of the next ensuing Council meeting, to be laid before the
Council. It will be the duty of the Council to decide whether such contribution
shall be accepted, and if so, whether it shall be read in full, in abstract, or
taken as read. All papers accepted for publication must be read or otherwise
communicated at an ordinary meeting prior to publication.

39. A Publications Committee, appointed by the Council, shall recommend
to the Council whether a paper presented to the Society shall be published in
the Society’s Journal. The Publications Committee may obtain an opinion from
any person it may select on the suitability of any paper for publication.

40. Publication in the Society’s Journal shall only be available to (a)
Ordinary Members, (b) Honorary Members, (c) Non-members resident outside
Western Australia, who must communicate the paper through an Ordinary or
Honorary Member. No paper shall be accepted from a Non-member resident in
Western Australia.

41. The original copy of every paper accepted for publication by the
Society, with its illustrations, shall become the property of the Society, unless
stipulation is made to the contrary, and authors shall not be at liberty to
publish their communicated papers elsewhere prior to their appearance in the
publications of the Society unless permission for so doing is given by the
Society, or unless the Society fails to publish the paper in the Journal of the year
in which it is read or otherwise communicated, or of the succeeding year.

Journal

of the

Royal Society of Western Australia

Volume 47

1964

Part 1
Contents

1. — The Environment of the Quokka (Setonix brachyurus) in the Darling

Range, Western Australia. By G. M. Storr.

2. — An Ordovician Cystoid (Pelmatozoa, Echinodermata) from Western Australia.

By Ida A. Brown.

3. — Description of a new Stonefish of the family Synaceidae from Western

Australia. By R. J. McKay.

4. — Association of Larger and Planktonic Foraminifera in single samples from

Middle Miocene sediments, Guadalcanal, Solomon Islands, south-west
Pacific. By P. J. Coleman and R. A. McTavish.

5. — Discontinuous and Presumed Vicarious Plant Species in Southern Australia.

By J. W. Green.

Editor: R. W. George
Assistant Editor: R. D. Royce

Annual Subscription: Forty-five Shillings

The Royal Society of Western Australia, Western Australian Museum, Perth

79342/1/64—570

ALEX. B. DAVIES, Government Printer, Western Australia

JOURNAL OF

THE ROYAL SOCIETY

OF

WESTERN AUSTRALIA

VOLUME 47
PART 2

PUBLISHED 13TH JULY, 1964

REGISTERED AT THE G.P.O., PERTH FOR TRANSMISSION BY POST AS A PERIODICAL

THE

ROYAL SOCIETY

OP

WESTERN AUSTRALIA

COUNCIL 1963-1964

President
Past President
Vice-Presidents

Joint Hon. Secretaries

lion. Treasurer
Hon. Librarian
Hon. Editors

C. F. H. Jenkins, M.A.

W. D. L. Ride, M.A., D.Phil.

W. R. Wallace, Dip. For.

J. H. Lord, B.Sc.

Margaret E. Redman, B.Sc.

J. G. Kay, B.Sc.

R. D. Royce, B.Sc. (Agric.).

Ariadna Neumann, B.A.

J. E. Glover, B.Sc., Ph.D. (1963).

R. W. George, B.Sc., Ph.D. (1964).

A. S. George, B.A.

A. B. Hatch, B.Sc., Dip. For.

L. E. Koch, M.Sc.

R. J. Little.

D. Merrilees, B.Sc.

W. C. Packer, Ph.D.

P. E. Playford, B.Sc., Ph.D.

L. W. Samuel, B.Sc., Ph.D., F.R.A.C.I., F.R.I.C.

Journal
of the

Royal Society of Western Australia
Vol. 47 Part 2

6. — The Woolgorong Stony Meteorite

By G. J. H. McCall* and P. M. Jefferyf

Manuscript received — 17th September , 1963

Numerous fragments of a stony meteorite
were recovered from Woolgorong Station, north-
north-east of Mullewa, Western Australia and
approximately 300 miles north of Perth in July,
1961, and later in the same year. This find
almost certainly represents a possible fall noted
in this locality in December, 1960. A double
report was heard at this time, but the authors
consider that this does not necessarily imply
a multiple fall since phenomena of this kind
are generally attributed to compression wave
effects high in the atmosphere. The meteorite
is a veined and brecciated, grey hypersthene-
olivine chondrite. Prior’s class III, and shows
evidence of pronounced recrystallisation. Re-
covery of fragments from shallow burial in the
soil and from the surface, has allowed a con-
fident reconstruction to be made. There is evi-
dence that it had an aerodynamically stable
form under conditions of atmospheric entry,
since surface patterns and the distribution of
thickened and thinned fusion -crust reveal a dis-
tinct orientation. Such orientation supports
the idea that this was a single fall, complicated
only by fragmentation at or near the point of
impact.

Microscopic study has revealed considerable
structural variation, some areas revealing well-
formed, though recrystallised, chondrules.
Others show almost complete recrystallisation to
a granular aggregate of polysomatic and larger
fractured grains. In these areas of recrystalli-
sation the chondrular structure is only vaguely
defined.

Chemical tests, etch -tests, X-ray diffraction
studies and two full chemical analyses have
been carried out, and the results are given
here. The minerals determined include kama-
cite, iron sulphide (troilite?) olivine (chrysolite-
hyalosiderite > , orthopyroxene, oligoclase, and
possibly a calcium silicate (rankinite? ).

Introduction

In August, 1961, the report of a possible
meteorite discovery reached the Geological Sur-
vey of Western Australia (Lord, unpublished
data). The report came from Mr. F. R. Wick-
man, Manager of Woolgorong Station (Fig. 1),
the site of the discovery (latitude 27 : 43' S,
longitude 115° 50' E).

This discovery followed seven months after
the experience of unusual phenomena including
sounds normally associated with explosions,
heard by people at or near the Station. Even at
that time the possibility of a meteorite fall was
discussed, though the possibility that an air-
craft in distress was involved was also suggested,
and indeed an air search is reported to have
been initiated. The meteorite was recovered as
a direct result of attempts to locate the object
which had caused this disturbance.

* University of Western Australia, Geology Department.
i University of Western Australia, Physics Department.

The detailed investigation of this fall was
passed over to the Director of the West Austra-
lian Museum. Dr. W. D. L. Ride, and he visited
the site of the find with Mr. D. Merrilees later
that year, recording details of personal impres-
sions of the 1960 phenomena and examining the
traces of the fall. They collected further frag-
ments of the stone, in addition to those origi-
nally collected by Mr. A. J. Noldart of the Geo-
logical Survey of Western Australia (Lord,
unpublished data).

At the request of Dr. Ride, the investigation
has been supervised in its later stages by honor-
ary associates of the Museum, members of the
Meteorite Advisory Committee which acts in an
advisory capacity to the Trustees of the
Museum; Dr. McCall, who has supervised petro-
graphic and chemical investigations, and com-
piled the information here given concerning
details of the fall and external features of the
meteorite; and Dr. Jeffery who has carried out
X-ray diffraction studies. The actual chemical
analysis was carried out at the British Museum
(Natural History), London: an additional X-ray
diffraction determination was also carried out
at the British Museum (Natural History).

This meteorite was initially supposed to be
achondritic — if this was correct it would have
been an exceedingly rare occurrence; however,
access to the whole collection of recovered frag-
ments showed that initial microscopic studies
had been carried out on a very strongly recry-
stallised area within the stone, an area showing
little trace of rounded chondrules, and that the
bulk of the fragments show distinct, rounded
chondrular insets. While the find is not as
rare as was at first suspected, the amount of
material recovered, the diverse internal texture
and structure, and the excellent preservation
of external features characteristic of oriented
meteorites, make this one of the most interest-
ing meteorite finds in this State.

History of Fall and Reconstruction of Meteorite

Discovery of the Meteorite

In July 1961, Mr. W. Hamlet and Mr. C.
Monger both employed at Woolgorong Station,
unearthed what they called a “sky stone” at a
point situated a few hundred yards north-west
of the Station homestead. A shallow crater
some twelve inches deep had been formed in the

33

topsoil, and the mass had apparently disinte-
grated into angular fragments on impact. The
manner in which the crater was gouged into
the soil suggested (to those who have visited the
site) impact at a reasonably shallow angle of
incidence, and no abnormal velocity, on a flight
bearing of 320° — that is, close to north-west.
The preservation of a trail of small meteorite
fragments and dust along the same trend line
supported this conclusion. Some larger frag-
ments were collected up to three feet from the
small crater, perhaps due to rebound on impact.
The geologist of the Geological Survey of West-
ern Australia who first visited the site reported
that the site of the fall was a broad, flat plain
of red soil, which overlies a thick, indurated
crust of the type locally known as the “Murchi-
son Cement/’

The Reported Fall

It must be borne in mind that all the reports
bearing on the phenomena observed in Decem-
ber, 1960, were made to Dr. Ride and Mr. Merri-
lees a year later. However, they correspond so
closely with other descriptions of meteorite falls,
contained in the literature, that there seems no
reasonable doubt that the meteorite fell at this
time.

As is not uncommon, auditory effects pre-
dominated — the only visual observation recorded
being a slight haze or dust cloud noted by Mr.

Tig. 1.— Sketch map showing the position of the Wool
gorong fall and meteorite recovery.

F. R. Wickman, who was at the homestead on
the afternoon shortly before Christmas, 1960
(Dec. 20th?) and at about 1400 hours heard a
swishing noise followed by a dull bang “like a
couple of plugs of gelignite going off forty feet
underground.” The suggestion of two distinct
reports, one at Woolgorong and one to the
north-north-west, towards Yallalong (Fig. 1) is
present in several recorded descriptions of this
event. At Yallalong a noise was heard which
suggested that an aircraft was in distress, and
it is said that a light aircraft actually took off
to search for it. Though this record of a double
report could mean a double fall — meteorites do,
in fact, usually fall in multiple associations
rather than as single bodies — such sound effects
are more commonly attributed to compression
waves high in the atmosphere, as the still speed-
ing meteorite enters the atmosphere and com-
mences to decelerate (an effect akin to the
supersonic bangs of jet aircraft). The evidence
of orientation <p. 38) suggests that this was in
fact, a single fall up to the point of fragmenta-
tion at or near contact with the ground. If a
second stone does exist, the nature of the ter-
rain and the rapid deterioration of stony
meteorites are factors likely to preclude further
recovery.

The reports were heard up to thirty miles from
the homestead at Woolgorong, and this in itself
supports the idea that the sound effects were due
to shock- waves — it is unlikely that the report
produced on impact with the earth of a mass
of about one hundred pounds weight could ever
be heard so far from the point of impact. Stones
of this size reach the surface of the earth at a
velocity in no way comparable with their
velocity through space because of the effect of
atmospheric braking, and the shallow and partial
burial of this meteorite testifies to impact at a
velocity probably not far off, or at, the speed of
free fall: the close association at the site of the
find of the fragments which were later pieced
together with perfect interlocking junctions
testifies to very late fragmentation, probably on
impact, though the cracks had clearly started
to form before this time.

Mr. Hamlet recalls a loud report and a noise
like a thunderstorm, but saw no flash; Mrs.
Wickman likened the noise, again, to an aircraft
in difficulty, and believes that a short (but ap-
preciable) pause separated the two reports.

The Meteorite Fragments

A total weight of about eighty pounds of frag-
ments has been recovered. Of these the bulk,
comprising five large and numerous smaller frag-
ments are held at the West Australian Museum
(W.A.M. No. 12113 a, b and c).* Another large,
fusion-crust coated fragment, weighing lb lb.,
was retained at. Woolgorong Station.

Two small fusion-crust coated fragments are
held in the collections of the Geological Survey
of Western Australia and the Government
Chemical Laboratories. Another such fragment
has been supplied to the Keeper of the Minera-
logical Collections at the British Museum
(Natural History).

■ Some of the material collected was temporarily stored
in contact with galena samples. Any geochemical
work on trace element content should therefore
be restricted to specimens marked 12113b which are
not so contaminated.

34

Plate 1.

A. Some larger fragments fitted together to reassemble the entire termination of the original mass (eight
fragments can be fitted together exactly, but an assemblage of only four Is shown here). The thin,
brown fusion-crust, which covers the surface contrasts with a grey, freshly-broken surface revealed
on the left hand side of the photograph. This surface is believed to be the anterior surface. Radiat-
ing striae and regmaglypts show clearly on the surface of fusion-crust (coin indicates scale).

B. End-on view of the same reconstructed termination showing contrasting fusion-crust surfaces, the
posterior surface in atmospheric flight being rough, blackened and thickened (scale as above)

C. Facetted fragment from the opposing termination of the original mass (scale as above).

D. Fragment showing elongated regmaglypts.

E. Fragment showing the thickened fusion-crust of the posterior surface, contrasting with that showm
in Plate 1 A above (X 3).

35

The fragments show mostly freshly- fractured
surfaces, appearing grey with patches and
streaks of iron oxide staining, and are of moder-
ately coarse grain. Specks of metal punctuate
these broken surfaces, appealing glistening and
of grey to brown colour on faces freshly cut with
the diamond saw, but rapidly tarnishing to the
brown colour seen on all the faces opened up at
the actual time of fragmenation. Thin, thread-
like veinlets traverse the stone; most are infilled
with glassy material, projections of the fusion-
crust inwards at the time of incipient fragmen-
tation (Plate IE). These are apparently what
Krinov <1960. p. 190) calls “surfaces of the
second kind." which develop only during the
final stage of progress of the meteorite through
the atmosphere. There are, however, some
veinlets infilled with sulphide (p. 38) and these
veinings must be due to some process which oc-
curred long before the brief transit of the stone
through the terrestrial atmosphere — a process
which presumbly occurred within the parent
planetary body.

Rounded chondrules show clearly on some
broken surfaces: some are up to half a centi-
metre in diameter, and most of them are com-
posed of ferromagnesian silicates showing bluish -
grey within the lighter grey granular base, and
not apparently aggregated with any metal or
felspar. Metallic chondrules of similar rounded
shape are not numerous, but can be recognised;
some form solid metallic pellets up to half a
centimetre in diameter but most are of consi-
derably smaller size. In the granular base, w r hich
forms considerable chondrule-free areas and
encloses the chondrules w^here these are mani-
fest, creamy-white specks of felspar can be re-
cognised.

Some fracture surfaces show a brecciated ap-
pearance, due to the presence of insets of irre-
gular form, appearing as angular fragments
(Plate 2B) — these are mostly broken or partly
recrystallised chondrules, but some seem to be
simply fragments of varying texture — it is how>-
ever, essentially a monomictous breccia in that
all the fragments stem from a single type of
stony meteorite.

The fusion-crust is evident on many frag-
ments; it shows a pronounced variation in thick-
ness and character, the thickness ranging from
less than 0.5 mm. to more than 0.75 mm., and
the areas of thickening (Plate 1E> being
characterised by a noticeably roughened sur-
face, black in colour and traversed by a fine
network pattern of polygonal shrinkage cracks,
the diameter of the polygons being in the order
of half a centimetre. In contrast, the fusion-
crust shows in the areas of thinning (Plate 1),
a smoother surface and a brownish colouration;
striations are here more in evidence, but poly-
gonal shrinkage cracks less so. Perfect thumb-
print regmaglypts are sparsely indented in the
fusion-crust (Plate 20; the regmaglypts at the
border area of the thinned crust, near the sharp
coigns in the facetted meteorite (Fig. 2), are
however, almond-shaped, and elongated parallel
to the finer striations on the glass surface (Plate
1A, D>. According to the accepted theory, the
circular thumb-prints appear where there is
little ablation while the elongated regmaglypts
appear where ablation results in strongly linear
flow effects. Striae are present on both thickened

Plate 2. A. Fragment showing a glass-veneered surface
due to the fact that the meteorite has
fractured along a plane of fracture of the
“second kind 1 ' (p. 36) (X £).

B. Cut face of a fragment showing circular
and angular chondrular insets of ferro-
magnesian minerals, giving to the meteo-
rite a brecciated character. The arrange-
ment of the nickel -iron and sulphide along
stringers is evident, and there is one large,
wholiy-metallic chondrular inset (scale,
width of the specimen is inches).

C. Thumb-print regmaglypts (scale X 1.25).

and thinned areas of fusion -crust, but are far
less evident on the rough, black, thickened areas
of fusion-crust (Plate 1C) as are shallow 7 reg-
maglypts.

Reconstruction and Orientation
It has not been possible to fit together all
the available fragments so as to reconstruct the
entire mass, enclosed by fusion-crust, as it was
before fragmentation. How’ever. it is probable
that at least tw r o -thirds of the mass is repre-
sented by fragments held in known collections.
The difficulty arises due to the small size of
many of the fragments and the fact that a
portion from the middle of the mass seems to

36

be not represented by any large fragments. A
single termination comprising about one-third
of the original mass was successfully recon-
structed, the various pieces, numbering eight in
all, being joined on tightly-interlocking faces.
The reconstruction of four of the largest frag-
ments into this termination is shown in Plate
1A. This reconstruction of one termination of
the mass allowed a confident estimation of the
length and cross-section shape of the stone, by
means of extrapolating the very regular curva-
ture of the faces of this part of the mass. The
total length of the mass was probably not more
than three feet, the maximum width just over
twelve inches, and the maximum depth six to
seven inches. The probable shape of the original
mass is shown in Fig. 2, together with the
fusion-crust patterns.

A fortunate chance in the correspondence of
a set of striations and the nature of the crust
on the surface of one large fragment with those
on the thickened surface of the reconstructed
termination, allowed this fragment to be iden-
tified and orientated as the opposing termina-
tion of the boat-shaped mass. This fragment,
shown in Plate 1 C, is drawn in its probable
position in Fig. 2. It seems that one termina-
tion was spatulate while the other was facetted,
and the lateral margins of the flattened boat-
shaped mass were also different, one being con-
tinuously curved while the other had a steeply-
facetted form and was bounded by abrupt coigns.

The pattern of striations, regmaglypts and
fusion-crust thickening suggests that this mass
had an aerodynamically-stable character while
travelling through the atmosphere — that is, it

Regmaglypts

Assemblage of five
interlocking fragments
assembled together
as a termination of
the flattened elon-
gated original mass

Striae

Single termination
fragment oriented by
means of roughened
posterior and longi -
tudinal basal striae.

TOTAL LENGTH
c. 3 feet

Regmaglypt

Thin brown
anterior fusion
crust. *•

Roughened black
posterior surface
thickened crust,
shrinkage cracks.

UrTrr-bF FUGH T

L\N£

Striae normal
to this section

anterior

\ posterior

\

\

Fig. 2. — Diagrammatic reconstruction of the probable form of the original mass before fragmentation, showing
.surface patterns and their probable relationship to the attitude adopted by the mass while in atmospheric
flight. The cross-hatched plane represents a cross-section of the meteorite.

37

did not rotate to any extent. This means that
it is an orientated meteorite; such a character
is somewhat unusual, and is supposed to be
evinced by meteorites which have entered the
atmosphere as a single mass, and not suffered
fragmentation until impact or the very last
stage of their transit through the atmosphere.
The probable flight attitude is shown in Fig.
2, and the extreme anterior point was probably
situated near to the coin, shown to give the
scale in Plate 1 A. In spite of the evidence
of orientation the mass does not seem to have
been symmetrical.

Petrography

The Metallic Fraction — Megascopic Details

The metallic fraction occurs as grey to brown-
ish specks, stringers and pellets within the
stone, and is mostly fringed by hazy, brown
iron oxide discolouration of the surrounding
silicate minerals. It forms between 10 and 15
per cent, by volume. One large pellet, half a
centimeter in diameter was removed bodily,
etched with 8 per cent. HNO.,, and tested for
nickel and sulphide. Etching revealed a coarse
irregular pattern (Fig. 3>, clearly not an orga-
nised octahedral Widmanstatter pattern. The
possibility that this pattern could stem from
aggregation of the sulphide within the metallic
areas is suggested. That considerable sulphide
is present was shown by dissolving some of the
metal in concentrated HC1, H 2 S fumes being
evolved ( troilite dissolves in this way); and the
reaction with dimethylglyoxime gave a strong
pink colouration indicative of nickel.

X-ray diffration studies <p. 42) provided
confirmation of the tentative recognition of the
principal metal component as kamacite, and the
absence of taenite revealed by these studies
confirmed that the etch-pattern could not be
an octahedral Widmanstatter etch-pattern.

The Silicate Fraction — Megascopic Details.

The silicate minerals of meteorites tend to
present unfamiliar surface appearances to the
petrologist familiar with terrestrial olivine,
pyroxenes and felspars. In this meteorite the
felspars have the usual creamy -white colour, but
the light grey colour of the olivine, and the
bluish colour of the ferromagnesian minerals of
the chondrules are atypical of terrestrial olivine
and orthopyroxenes. The finely-shattered or
flnely-granular nature of the crystals which is
apparent on microscopic examination, probably
accounts for this unusual appearance in the
hand-specimen. The metal-free and felspar-

, 05cm i

Fig. 3. — Irregular etch pattern developed after etching
the cut and polished surface of a large metallic pellet
(Plate 2 B) with 8’Z HNO.i. The cross-hatched area is
kamacite, the black enclaves are etch-pits, probably
reflecting sulphide inclusions within the nickel -iron.
The central area is a deep cavity produced on etching —
its origin is uncertain.

free character of some chondrules is very notice-
able, as is the fact that micro-faulting within
the stone has displaced portions of some chon-
drules relative to the remainder. This, together
with the angular and fragmental nature of
some chondrules accounts for the brecciated
appearance seen in certain areas of the meteor-
ite (Plate 2 B>. The light grey colour of the
bulk of the silicate fraction, together with the
more bluish-grey colour of the chondrules leave
no doubt that this should be classified as a grey
chondrite (in the older terminology of Brezina),
and the structure noted above requires the ad-
dition of the term brecciated.

The Metallic and Silicate Fractions —
Microscopic Examination
Under the microscope, the meteorite reveals
a predominant content of transparent silicate
minerals and subordinate opaque nickel-iron.
It is almost free from glass, except for fusion-
crust and extensions of it into cracks. The
metal percentage is below the upper limit of
metal content of aerolites and thus no other
appellation can reasonably be given. The sili-
cate minerals are; —

Olivine (var. chrysolite — hyalosiderite)
Orthopyroxene ( enstatite-hypersthene )
Plagioclase felspar (var. oligoclase)

Plate 3.

A.

B.

C.

D.

E.

F.

i granular chondrule composed largely of orthopyroxene (dark-grey). Felspar
a (black) fringe the chondrule and stringers of this material tend to terminate at
n * margins. Vague fan structure is recognisable (X 63. Plane polarised light).

A ? otl ? e , r such chondrule. but mainly composed of olivine. It shows similar fring-
H^kel-iron and sulphide (black) and felspar (white), but here the minerals
choi ™L ru i e at its margins, probably in the course of the partially com-
pleted process of recrystalMsation. Troilite (black) occurs in a veinlet on the right hand side of the
pnotograpn, a veinlet of the type that occasionally transects the chondrules (X 63. Plane polarised
right ) -

Photomicrograph. A grated chondrule apparently entirely composed of olivine lamellae without
associated glass; clearly strongly recrystallised (X 125, Plane polarised light).

Photomicrograph. An exocentric fan chondrule. apparently entirely composed of olivine (X 63,
Plane polarised light).

Photomicrograph, a microporphyritic chondrule consisting of sulphide (black), felspar (white) and
dark, dusty aggregates of cryptocrystalline material or glass; this chondrule appears as if partial
assimilation has occurred during the recrystallisation process (X 100. Plane polarised light).
Photomicrograph. A fan chondrule of cryptocrystalline material or glass (cloudy black), sulphide
1 black ) and felspar (white). Olivine shows as grey granules iX 100, Plane polarised light).

38

PLATE 3

39

No other silicate minerals could be optically de-
termined.

Olivine . — Entirely fresh, it shows some evi-
dence of subhedral form in “gabled’- termina-
tions of some larger grains, but is mostly
anhedral. It may be of somewhat fibrous ap-
pearance in the chondrules. Irregular granules
tend to show a polysomatic character, being
granulated into innumerable minute granules
bounded by planes of cleavage and parting
within the larger crystal (Plate 4 A, B). The
granulation has locally proceeded so far that
the granulated aggregates of minute grains bear
no trace of the larger grains from which they
have been derived. As the dominant silicate
mineral, olivine occurs in both chondrules
(Plate 3 C. D) and areas of almost complete re-
crystallisation devoid of apparent chondrules
(Plate 4 A > : it is evident in the lamellae within
grated chondrules (Plate 3 0.

The diagnostic optical characteristics are an
interference figure indicating a biaxial mineral
with a very large optic angle <c.8(F) and a
negative sign, and refractive index values rang-
ing from 1.68 to 1.72. These properties fit in
with a slightly ferroan variety, compositionally

slightly towards the hyalosiderite range from
chrysolite (Mg 9 oFe 2 o) . This agrees with the
chemical determination as probably c.Mg 7 r,Fe_. r ,
(p. 41). This is the typical olivine of Prior’s
Class III, hypersthene-olivine-chondrite.

Orthopyroxene . — This can be recognised in
thin sections on account of better cleavage and
lower birefringence than the olivine. It is, how-
ever, often difficult to differentiate in fine grains.
Present in both chondrules and areas of almost
complete recrystallisation, it is of a non-
pleochroic variety. The optical sign seems to
be negative (though few good interference-
figures were obtained) and the most probable
identity is hypersthene lacking pleochroism.
Deer. Howie and Zussman (1963. p. 30) state
that in orthopyroxenes the pleochroism is not
a factor of increasing iron content but is in-
dependent of the MgO/FeO ratio, and thus non-
pleochroic, ferroan orthopyroxenes are not un-
known. It must, however, be noted that while
enstatite is ruled out as a modal component on
account of the chemical evidence tp. 41), modal
bronzite is possible since Prior's classification
(1922) is chemically based and involves assess-
ment of the MgO/FeO molecular ratio in all the

A B

Plate 4. A. Photomicrograph. A strongly recrystallised area of the meteorite. The finely cracked, polysomatic
nature of the olivine grains (grey) is apparent, as is the manner in which the felspars (white) are
aggregated interst.itially and along stringers with nickel iron and troilite (black) iX 163. Plane polar-
ised light).

B Photomicrograph A similar recrystallised area under higher magnification. Kamacite shows black,
olivine grey, and oltgoclase felspar white. The texture of this area of the meteorite appears not
dissimilar from that of many terrestrial igneous rocks (X 320. Plane polarised light).

40

silicate fraction, while in fact olivine tends to
take up rather more iron in proportion to mag-
nesium than orthopyroxene (Prior and Hey,
1953, p. x).

Flagioclase . — Present as colourless interstitial
granules, translucent and showing distinct
cleavage but only seldom showing lamellar
twinning, the felspars are quite fresh. Refrac-
tive index determinations and measurements of
the extinction angles on a few recognisable
lamellar-twinned grains indicated that it has the
composition of oligoclase, somewhere between
AnaoAbso and An 28 Ab 72 , slightly more sodic than
the theoretical deduction of An K . (andesine)
from the normative calculation.

Modal analysis is not entirely satisfactory
because of the difficulty of differentiating be-
tween fine grains of pyroxene and olivine, and
the variability of the material. The following
would, however, seem to be a fair approximate
assessment of the mode: —

Metal. Nickel-iron (kamacite) + sulphide
(troilite) 10-15

Silicates. Ferromagnesian (olivine -P hy-
persthene) 80 dr

Oligoclase 3-10

Microscopic study using reflected light allowed
a clear distinction between kamacite and troilite.
The later developed veinlets are entirely of
sulphide. Bright reflectant specks aggregated
with the kamacite may be schreibersite.

Texture and Microscopic Structure
The texture and structure are very variable.
Chondrules provide the most striking structural
feature; these mostly show well-rounded shapes
(Plates 3 A, B) but some are broken, and some
appear to have been partly absorbed in the re-
crystallised containing mass, again losing their
regular shape. The grated chondrules show no
evidence of ever having possessed any regular
shape (Plate 30. In the strongly recrystallised
areas the chondrular structure is often not ap-
parent. The range of internal textures is that
normally associated with chondritic aerolites;
the granular type called micro porphyri tic, the
fan forms showing an exocentric focus (Plate
3 D > , and grated chondrules are all abundant,
but monosomatic (single crystal > chondrules
have not so far been recognised. Felspar and
metal are absent from most of the chondrules.
There are sparse metallic chondrules, either
composed entirely of metal or of sulphide and
cryptic isotropic material (glass?) associated
with felspar (Plates 2 B, 3 E, F) ; the latter in-
clude both microporphyritic and fan forms.

The relationship of the felspar and nickel-iron
(native metal and sulphide) to the ferro-
magnesians and to the chondrular insets is one
of the most interesting features of this meteorite.
In recrystallised areas both nickel-iron and
felspar are interstitial to the ferromagnesian
grains (Plates 4 A, B); often this interstitial
growth is arranged in narrow stringeis and
there are very w'ell defined veinlets of the same
material (Plate 3 B). These stringers tend to
terminate in peripheral fringes around the chon-
drules (Plates 3 A, B), but rare examples, such
as that shown in Plate 3 B. cut through them.
The relationship suggests a late mobilisation and
introduction of nickel-iron (-{-sulphide) and

felspar into a chondrite showing only very limited
development of these minerals in the form of
sparse chondrules. There seems to be some
evidence of the same fringing relationship of
sulphide and felspar to sulphide/felspar/glass
chondrules as seen in the silicate chondrules,
and these too seem to have been early structures
which have been later invaded by material
identical with that which composes them. The
alternative interpretation — that the sulphide/
felspar chondrules represent crystallisation in
spheroids from these stringers seems unlikely.
Lovering (Moore, 1962, p. 195) has suggested a
sequence of genetic phases in meteorites, and
this felspar-metal invasion which seems to ac-
company recrystallisation seems to fit in well
with this concept.

The classificatory terms of the older Brezina
system, veined and recrystallised (crystalline),
seem entirely justified by the structural and
textural evidence.

The texture and mineralogy of the recrystal-
lised areas (Plates 4 A, B) of this meteorite
would be difficult to differentiate from that of
some ultrabasic igneous rocks were it not for
the presence of nickel /iron with subordinate
sulphide rather than entirely combined in ore
minerals — oxides and sulphides characteristic
of such rocks. This comparison could well have
a petrogenetic significance, in respect of the
derivation of crustal igneous rocks from the
mantle.

Chemical Analysis

Two full chemical analyses have been carried
out on specimens chosen at random from the
collection at the West Australian Museum. The
only selective factor was that specimens of
fresh appearance were chosen.

Before sending the specimens for analysis a
set of specific gravity determinations was made
on five of the fragments. The results obtained
were 3.45, 3.46. 3.53, 3.54, 3.58; Average 3.51.

Results of analysis by Dr. A. A. Moss, British
Museum (Natural History) are as follows: —

Analysis by the Method of Prior (1913).

Lab. No. 2943.

Percentage Composition

Fe

6.31

TiOa

0.10

MnO

0.29

Na jO

1.04

Ni

0.99

ALO*

2.98

FeO

14.07

K.O

0.14

Co

0.05

Cr 3 Os

0.54

MgO

24.88

HjO+ 0.04

FeS

6.43

P-O:.

0.09

CaO

1.89

H 2 0 —

009

SiO 4

39.95

Total 99.88%

Normative mineral composition
Olivine (close to Mg 1 .. 1 Fe«>..,SiO l > 46.94

Bronzite* 25.19

Felspar t 12.69

* The conversion of a meteorite analysis by Prior (1913)
to normative values involved the allocation of excess
CaSiO to bronzite, a mineral that does not con-
tain any calcium silicate. This seems without any
real basis and it would perhaps be better to express
the CaSiOu excess simply as CaSiO.j. However, in
conformity with the accepted practice, the figure
for bronzite given here includes CaSiO:, excess: it.
must be noted that this will make the bronzite
value somewhat higher than the expected modal
orthopyroxene content.

t The felspar normative values are Or .84. plagioclase
remainder (c.An^oAbeo) .

41

Ilmenite

0.18

Chromite

0.81

Apatite

0.21

Troilite

6.43

Nickel Iren (Ni/Fe

1/6.3)

7.35

Molecular ratio MgO

/FeO in

Silicate Fraction

3.14

A second analysis by the Government Chemi-
cal Laboratories, Perth, shows no significant
divergence.

X-Ray Diffraction Studies

The following minerals were recognised in the
course of X-ray diffraction studies carried out
at the Physics Department, University of West-
ern Australia.

Metallic Fraction
Kamacite

The sulphide could not be conclusively identified.
Silicate Fraction
Olivine (“forsterite”)

Enstatite (60% certain identification)

Rankinite (only very doubtful identification
since no calcium silicate has ever been re-
cognised in meteorites).

The recognition of orthopyroxene of the
enstatite-hypersthene series was confirmed by
further X-ray diffraction studies of pyroxene
separated from the meteorite by Mineralogists
of the British Museum (Natural History).

In the course of his investigation of olivines
from stony meteorites Dr. B. Mason has carried
out optical and X-ray diffraction studies of
olivine from this meteorite, and confirms that
it is Fo 7 .-, Fa_>5 (written communication).

Acknowledgments

The fact that this occurrence has been re-
corded at all is entirely due to the interest
shown by Messrs. F. R. Wickman, W. Hamlet,
and C. Monger of Woolgorong Station. Dr. G. F.

Claringbull and Dr. A. A. Moss of the British
Museum (Natural History) assisted the writers;
by carrying out chemical analysis and an addi-
tional X-ray diffraction study, and Mr. J. H.
Lord, Director of the Geological Survey of
Western Australia, Dr. W. D. L. Ride and Mr.
D. Merrilees of the Western Australian Museum
contributed towards the interpretation and de-
scription of this fall. Mr. K. Bauer produced
the photographic illustrations (except Plate 1 E),
and Miss R. Hunt drafted the line illustrations.

Appendix

List of institutions holding specimens of the
Woolgorong Meteorite —

West Australian Museum (12113a, b and C„
and thin sections 12283-12286)*.

University of Western Australia (48334).

Geological Survey, Western Australia
(1/4878).

Government Chemical Laboratory, Perth
(N DC 3205).

Kalgoorlie School of Mines.

British Museum (Natural History).

American Museum of Natural History, New
York.

Smithsonian Institute, Washington.

References

Deer, W. A., Howie, R. A., and Zussman, J., (1963). —
“Rock Forming Minerals.” (Longmans:
London.) Vol. 2.

Krinov, E. L., (I960) — “Principles of Meteoritics.” (Per-
gamon Press: Oxford, London, New York
and Paris.) Translation.

Moore, C. B., (editor) (1962) — “Research on Meteorites."

(John Wiley and Sons: New York and

London.)

Prior, G. T., (1913) — The meteoritic stones of Baroti,
Punjab, India, and Wittekranz, South
Africa. Miner. Mag. 17, 22-32.

(1922) — The classification of meteorites.

Miner. Mag. 19, 51-68.

Prior, G. T., and Hey, M. H.. (1953) — “Catalogue of
Meteorites.” (British Museum (Nat. Hist.). I
2nd Ed.

* The photomicrographs used as illustrations 3a-3f, and
4a-4b were prepared from these thin sections.

42

7. — The Agamid Lizards of the Genus Tympanoeryptis in

Western Australia

By G. M. Storr*

Manuscript received — 17th March, 1964

Seven species and subspecies inhabiting West-
ern Australia (including parviceps sp. nov.) are
described and their distribution outlined. The
relationships between centralis, tetraporophora
and lineata are discussed. Doubt is raised
whether the genus is natural. The key includes
all known species and subspecies of Tympano-
cryptis.

Introduction

The genus Tympanoeryptis , like Nephrurus
<Storr 1963), is most diversified in the more
arid parts of Australia; and being equally dif-
ficult to collect, the discovery ox its various
species has been similarly protracted. Three
species were described in the second half of
the nineteenth century, but no more were
added until Mitchell < 1948) revised the genus
and proposed several new taxa. The purpose
of this paper is to describe yet another species,
to amend the current concept of certain taxa.
and to bring up to date our knowledge of the
variation and distribution of all Western Aus-
tralian members of the genus.

In the descriptions of species and subspecies
diagnostic characters have been printed in bold-
face type. Unless stated to the contrary, all
material examined is lodged in the Western
Australian Museum, and all localities are in this
State. Since two species ( tetraporophora and
uniformis ) are herein recorded from Western
Australia for the first time, it is quite possible
that other eastern taxa will be found here; ac-
cordingly maculosa, intima, and pinguicolla (all
of Mitchell 1948) have been included in the
key.

Genus Tympanoeryptis Peters

Type (by monotypy): T. lineata Peters, 1864,
Mber. Akad. Wiss. Berlin 1863: 230.

Diagnosis. Distinguishable from all other
Australian Agamidae by the absence of an ex-
ternal ear opening.

Description. Small, terrestrial, cryptozoic
and somewhat sluggish lizards with a greatly
or moderately depressed head, body, and tail.
The limbs, head, and tail are relatively small.
The dorsal scales are heterogeneous, consisting
mostly of small, smooth or keeled scales, mixed
with tubercles, i.e., large scales, more or less
raised and usually keeled, spinose, or mucro-
nate. A gular fold always and a dorso-lateral
fold sometimes present. Most species lack
femoral pores. Pre-anal pores usually two but
absent from females of most taxa.

Distribution. Endemic to Australia, where it
is absent cnly from the east coast and south-
west corner of the continent. There are seven
known species, five of them occurring in West-
ern Australia with a joint range from the Kim-

* Western Australian Museum, Perth, Western Australia.

berley Division south to the northern Wheat-
Belt, the Eastern Goldfields, and the Great Aus-
tralian Bight.

Key to Species and Subspecies

1. Dorsal tubercles conspicuous, more cr less
raised, with or without keels, spinose or

mucronate 2

Dorsal tubercles inconspicuous, fiat and
smooth maculosa

2. Rows of scales between nasal and labials.

4-8 3

Rows of scales between nasal and labials,

2-3 ... .... .... parviceps

3. Dorsal tubercles clumped in transverse and

oblique ridges 4

Dorsal tubercles isolated 5

4. Tail bands sharply oblique, usually narrower

than pale interspaces cephala gigas

Tail bands almost transverse, much wider
than pale interspaces cephala cephala

5. Conspicuous dorsal pattern: neck almost as

wide as or wider than head 6

Moderate or obscure pattern; neck much
narrower than head 7

6. Neck slightly narrower than head; South

and Western Australia lineata lineata

Neck slightly wider than head: southern

Victoria lineata pinguicolla

7. Dorsal tubercles irregularly distributed 8

Tubercles tend to be in four longitudinal

rows intima

8. Hind -leg about as long as trunk tetraporophora

Hind-leg considerably shorter than trunk 9

9. Eye almost completely hidden from above by

bowed-out supraciliary ridge lineata centralis

Eye protrudes from below straight supra-
ciliary ridge uniformis

Tympanoeryptis lineata lineata Peters

Tympanoeryptis lineata Peters. 1864. supra cit.
Near Adelaide, South Australia.

Material examined. R 14184 (10 miles NE of
Fraser Range HS.), R 17418 (12 miles SW of
Balls donia), R 12222 (12 miles W of Naretha),
R 19101-4 (Naretha), R19105-10 (Seemore
Downs), R 15209 (Rawlinna), R 16502 (Forrest).

Description. Moderately stout and depressed
Sides of neck puffed out almost to width of head.
Supraciliary ridge moderately acute and
(viewed from above) only slightly bowed out-
wards and not concealing the somewhat protru-
sive eyes. Facial trough not deep or angular.
Snout-vent length of largest specimen (R 19103,
a gravid 9 > 65 mm.

Rostral 2-3 times as wide as first labial.
Mental truncate polygonal or elliptical, deeper
than wide, and usually narrower than rostral.
Upper labials 12-15. Nasal approximately hep-
tagcnal, flat or slightly convex, the moderately
large eval nostril piercing it towards the top
and entering forwards and downwards, and
separated from the labials by 4-5 rows of tecti-
form scales.

43

▲ U NEAT A UNEAT A
A UN EAT A CENTRALIS

X TETRA POROPHOR/
■ PARVICEPS
• UNIFORM IS
© CEP H ALA CEP Ha
O CEPHALA G/GAS

4

\LA

f \x*

r •

_ 1 1

1

1 1
1 1
1
1
1
1
1
1
1
1

20

) °

■

0 u\

o

o

° 1

> n

C

5/n q

o

o

o

o °

. °

D o

0

o

1 ]

A A A a *-■

1

1

1

1

1

1

1

j 1

30

<#° aa-

A

A

i aTt

35°

11 5°

120° 12 i

r~

5° 730°

Fig. 1.— Map of Western Australia, showing location of specimens of Tympanocryptis.

44

Scales on top of head strongly keeled, usually
unicarinate but occasionally multicarinate or
rugose, smallest above the temples and supra-
ciliary ridge. Scales on nape small and bluntly
keeled, becoming larger and more strongly
keeled on back, where they are mixed with
scattered tubercles, i.e. scales that are enlarged,
raised, sharply keeled and mucronate. Tubercles
are absent from the longitudinal white lines and
tend to be sparse in the space between the
dark transverse bands. There are some tubercles
on the base of the tail, but distally the scales
are more uniform, their keels aligned longi-
tudinally. Scales on upper surface of limbs
(like tail) large and strongly keeled; on the
under surface smaller and feebly keeled.

Gulars and ventrals smooth. Palms and soles
covered with small, imbricate, bluntly keeled
scales. Lamellae under fourth toe 18-22, spinose
and sharply bicarinate. Claws relatively short.
Pores in both sexes, 2 pre-anal and occasionally
a small one on thigh.

Dorsal ground colour fawn. Five prominent
silvery grey or white longitudinal lines running
back from neck; of these the vertebral is much
the widest (4 times as wide as the dorso-
laterals) and, like the laterals, ends at the level
of the pelvis; the dorso-laterals reach the
proximal third of tail. These lines dilate as
they cross over dark chocolate-brown bands (a
broad one across neck, 4 narrower ones across
body, and 12-17 indistinct ones on tail ) and may
become brownish where they cross the pale
interspaces. The transverse bands are widest
where they contact the vertebral line. The
squares bounded by the transverse bands and
the vertebral and dorso-lateral lines are some-
what paler and less reddish than the corre-
sponding spaces between the dorso-lateral and
lateral lines. Upper surface of limbs, like tail,
vaguely banded with chocolate-brown. A short
dark chocolate transverse bar in front and a
short longitudinal bar to each side of inter-
parietal. A pale transverse bar from orbit to
orbit, with an ill -defined dark bar immediately
in front of it. Under surface whitish except for
grey or black flecking on chin, chest or abdomen
(these dark scales, especially on chin, may be
arranged in broken, more or less longitudinal
lines).

Distribution. South-eastern Western Austra-
lia west to Fraser Range and east through
southern South Australia to north-western Vic-
toria where it begins to intergrade with the
south Victorian race pinguicolla (Mitchell 1948).

Tympanocryptis lineata centralis Sternfeld

Tympanocryptis lineata centralis Sternfeld, 1924. Abh.
Senckb. Naturf. Ges. 38: 234. Hermannsburg Northern
Territory (M. von Leonhardi).

Material examined. R 21861-4 (Balgo Hills
Mission), R 15141-3 (Warburton Range Mis-
sion), R 15696 (20 miles E of Warburton Range
Mission), R 20988 (western pass, Blackstone
Range), R 20746 (Blackstone Mining Camp),
R 20757-8 <Pass of the Abencerrages, Rawlinson
Range), R 21454-6 (Tennant Creek, Northern
Territory), R 20925 (Victory Downs, Northern
Territory ) .

Description. More slender than nominate
race, with narrower neck and shorter tail and
limbs. When viewed from above, supraciliary
ridge more strongly bowed outwards and con-

cealing more but not all of eye. Snout-vent
length of largest specimen (R 21456, a gravid
9 ) 61 mm.

Rostral less than twice as wide as first labial.
Mental usually deeper than wide and narrower
than rostral. Upper labials 10-14.

Head scales more uniform in size and less
strongly keeled than in nominate race, especially
in vicinity of interparietal where they may be
quite smooth. Supratemporals not markedly
smaller than neighbouring scales. Dorsals more
heterogeneous than in nominate race, with a
greater contrast between the smooth or slightly
keeled, weakly imbricate small scales and the
sharply keeled, spinose tubercles.

Gulars and ventrals weakly keeled. Lamellae
under fourth toe 15-21, mostly unicarinate.
becoming bicarinate distally. Claws, especially
those of fingers, much longer than in nominate
race, and the longest in the genus. Two pre-
anal and occasionally a small femoral pore
(pores are seldom discernible in females).

Coloration darker (reddish or purplish
brown) and less strongly patterned than in
nominate race. The longitudinal lines,,
especially the laterals, may be obscure or ab-
sent. When present, the vertebral line is only
twice (not four times) as wide as the dorso-
laterals.

Distribution. Hills of the far eastern interior
of Western Australia, east into the Northern
Territory and probably north-western South
Australia <R 20925 was taken about a mile north
of the South Australian border).

Comments. Loveridge (1934) wrote, “the
grounds on which Sternfeld based this race
are somewhat slender”. They certainly were,
for one of his two characters was invalid, viz.
that centralis has a much longer tail than
typical lineata. Sternfeld had no specimens of
the latter and relied on the imperfect descrip-
tion in Boulenger (1885). Mitchell f 1948 » went
so far as to relegate centralis to the synonomy
of lineata lineata. Yet centralis, as the fore-
going description shows, differs consistently and
markedly from lineata, which occurs only 300
niiles to the south and is quite uniform through-
out its extensive range in the south-east of
this State and, judging from Mitchell’s photo-
graphs and descriptions, in southern South
Australia.

I have hesitated to include in centralis a
specimen (R 13222) from Giles. More material
is required before deciding whether the discre-
pancies between it and centralis are due to its
youth (snout-vent length 25 mm.) or are
characteristic of another but undescribed taxon.
There is no indication of longitudinal white
lines; and the dorsal tubercles are large, ob-
tusely keeled, not spinose or even mucronate.
and tend to be aligned longitudinally. It is
thus similar in some respects to the two speci-
mens from the northern MacDonnell Ranges,
described briefly by Mitchell (1948, p. 70).

Tympanocryptis tetraporophora Lucas & Frost

Tympanocryptis tetraporophora Lucas & Frost. 1895^
Proc. Roy. Soc. Vic. 7 : 265. Admlnga, South Australia
{W. Baldwin Spencer).

Material examined. R 11752 (Kimberley Re-
search Station, Ord River >, Macleay Mus. R 930
(King Sound), National Mus. D 7701 (holotype).

45

Description. Relatively slender with a narrow
head and long legs. Tail long, tapering gradu-
ally. Supraciliary ridge acute and (viewed from
above) fairly straight in outline and almost
wholly concealing eye. Rostral ridge acute.
Racial trough shallow and rounded (R 930) or
deep and angular <R 11752), Snout-vent length
of largest specimen (R 930, a presumed 9L 63.5
mm.

Rostral 3-4 times as wide as high, wider than
polygonal mental, about equal to combined
width of first two labials. Nasal situated on or
below rostral ridge, elliptical or rounded poly-
gonal, convex, pierced a little above its centre
by the large e’iiptical nostril which enters for-
\ arc? , and sepaiated from labials by 4-5 rows of
facials. Upper labials 11-15, increasing in size
backwards, tectiform or keeled, and as large
as or larger than adjacent facials.

Scales on top of head mostly with a strong
keel, sometimes tricarinate or rugose, smallest
above orbits, and largest on occiput and temples.
Scales on nape small, strongly imbricate, keeled
or spinose. becoming larger on back where they
are irregularly mixed with tubercles, i.e. larger,
raised scales which are strongly keeled and
spinose. Extreme base of tail has scattered
tubercles, not much larger than the sharply
keeled, mucronate ordinary scales. Scales on
upper surface of limbs large, sharply keeled, and
mucronate or spinose.

Scales on chin keeled, becoming larger,
mucronate and more strongly imbricate on
throat. Ventrals large, imbricate, keeled and
mucronate. Subcaudals similar in size to
supracaudals but less strongly keeled, not
spinose and only slightly mucronate, the keels
aligned longitudinally. Scales on palms and
soles strongly imbricate, raised, and trispinose
(the central spine much the highest). Lamellae
under fourth toe 19-21, bicarinate and trispi-
nose (the outer spine much the smallest and not
arising from a keel). Claws moderately long.
Pores in males only, 2 pre-anal.

Head and back brownish grey, becoming paler
on distal half of tail. Indistinct brown bands
across back, narrowly broken by the vertebral
and dorso-lateral lines which, being little paler
than the ground colour, are scarcely discernible.
Several caudal bands, of which the first 2-4
are darkest and least obscure. Limbs paler than
back, not or vaguely banded. Under surface
whitish, with or without brown flecking.

The above description is based on the two
Kimberley specimens, which differ as much
among themselves as from the type, which I
have been able to examine through the courtesy
of Mr. J. McNally, Director of the National
Museum, Melbourne. The rostral is higher in
the type, so that it is only twice as wide as high
and no wider than the mental. The obtuse
rostral ridge is swollen arcund the nasal, which
is separated from the labials by 6-7 rows of
facials. Most of the head scales in the type
have one or more small pits near their posterior
edge.

Distribution. Restricted in Western Australia
to the Kimberley Division. (Owing to past con-
fusion of this species with lineata, it is not pos-
sible to ascertain from the literature its distri-
bution beyond this State. It probably extends

from the Kimberleys south and east through
the Northern Territory to far northern South
Australia. )

Comments. Lucas and Frost founded this
species mainly on its possession of femoral pores.
A year later (1896) they virtually retracted it
after discovering, as I have done, that lineata
too may have femoral pores. Nevertheless.
tetraporophora is a recognisable taxon, differing
from linea J a in having little colour pattern, a
more slender habitus, and considerably longer
limbs. The only question is whether it or
centralis represents nneata in central Australia.
Only one cf them can be regarded as a race of
lineata, for they are almost certainly sympatric
over a large part of the Northern Territory
(they may however be ecologically separated, fer
centralis has only been found among rocks). I
have retained centralis as a race of lineata, be-
cause of its greater similarity in colouring. It
is interesting that centralis and tetraporophora
diverge from lineata in opposite directions with
respect to length of limbs.

Tympanccryptis parviceps sp, nov.

Hclotype. R 16984 (in Western Australian
Museum), an adult 6 collected by G. M. Storr
and B. T. Clay on October 18, 1962.

Type Locality. Eleven miles south-east of
Ningalco, Western Australia, in 22° 48' S. 113
49' E.

Paratypes. R 19095-6 (2 miles N of Ninga-
loo ) ; R 8833, R 13221, R 13483, R 19094 (Point
Cloates); R 19C97-8 (Quobba); R 10654, R
11265-70, R 13164, R 20494-5 (Bernier Island*.

Description. Body relatively slender, moder-
ately depressed. Head small, not greatly wider
than neck. Fore-legs short, hind-legs long.
Supraciliary ridge acute and i viewed from
above) bowed outward and concealing most of
eye. Rostral ridge continuous with supraciliary
ridge and similarly acute. Facial trough deep
and angular. Chin terraced, i.e. the plane of
the chin, instead of curving gradually round the
ventro-lateral corner of jaw changes abruptly
along a straight line running obliquely back
from each side of mental (see Figure 2). A
small but distinct dorso-lateral fold. Snout-
vent length of largest specimen <R 19094, a 9)
46.5 mm.

Fig. 2. — Ventral surface of head of T . parviceps illustrat-
ing the "terracing” cf chin. Drawn by Miss R. Hunt.

46

Rostral much wider than high, about equal
to combined width of first two labials, and
wider than the hexagonal or truncate polygonal
or elliptical mental, which is nearly as wide as
deep. Upper labials 8-12, keeled, much smaller
than the facials immediately above them.
Nasal large, situated well below rostral ridge
and extending to bottom of facial trough,
rounded polygonal, flat or slightly concave,
pierced towards its top by an elliptical nostril
which enters downwards and forwards, and
separated frcm labials by 2 (sometimes 3) rows
of facials, the upper row small and smooth, the
lower large and keeled.

Scales on top of head strongly keeled, largest
in the frontal and interpariental regions,
smallest above the supraciliary ridge and on the
occiput where they become rugosely conical.
Scales on nape small but strongly keeled. Scales
in wide strip down middle of back uniform,
slightly raised, strongly keeled and weakly
spinose. Dorso-laterals vary greatly in size, the
smallest being little larger than the nuchals.
Caudals similar to but larger than median dor-
sals, Scales on upper surface of limbs similar
to median dorsals.

Scales on chin weakly keeled, becoming higher
and subspinose on throat. Ventrals large and
mucronate on chest, becoming smaller and
weaker on abdomen. Scales on under surface
of limbs similar to those on chest. Scales on
palms and soles flat and weakly keeled. Lamellae
under fourth toe 18-23, bicarinate and bispinose
( the outer spines much the smaller ) . Claws
slender and moderately short. Pores in <? 6
only, 26-34 (8-10 pre-anal, 16-25 femcral).

Dorsal ground colour pale brownish grey,,
palest and tending to fawn on the median dor-
sal strip (where all the scales are uniformly
large). The dark brown on sides of body and
tail extends on to dorsum in the form of nar-
row hourglass-shaped bars, broken by the pale
median strip. Limbs not or vaguely banded
with brown. Under surface whitish except for
grey or black marbling on chin and throat.

Distribution. Mid-west coast of Western Aus-
tralia from Ningaloo south to Quobba and
Bernier Island.

Comments. This is the most distinctive
species in the genus, and its belated description
has been due to its confusion with Ampliibolurus
adelaidensis rather than to its restricted distri-
bution. Within its range it is apparently plenti-
ful, especially at Point Cloates where several
specialists in other animal groups have
incidentally collected parviceps .

Lacking an external ear opening and a
lateral line of tubercles on the base of tail,
parviceps should never have been mistaken for
adelaidensis. Nevertheless their similarity is
more than superficial; for these two species,
alone among the Agamidae I have examined,
possess a “terraced” chin. In view of this and
of similarities in colour pattern and dorsal scala-
tion, they are believed to be more closely related
to each other than to any species in the genus
to which they are at present assigned.

TvmnanocrvDtis uniformis Mitchell

Tympanocryptis uniformis Mitchell 1948, Rec. S.
Aust. Mus. 9: 76. Near Darwin. Northern Territory

(P. Wesselmann).

Fig. 3. — Left (top to bottom): Tympanocryptis tetraporophora , T. 1. centralis , and T. L lineata . Right (top to

bottom): T. cephala gigas, T. uniformis, and T. parviceps.

47

Material examined. R 13638 (20 miles SE of
lAiluigui, West Kimberley > .

Description. Body moderately depressed.
Head large. Snout short, sloping steeply in
profile. Limbs small. Tail tapering rapidly <as
in T. cephalaj. Supraciliary ridge acute and
(viewed from above) straight in outline and not
concealing the very protrusive eyes. Rostral
ridge obtuse. Facial trough shallow but angu-
lar.

Rostral a little more than twice as wide as
high, slightly wider than combined width of
first two labials, and a little wider than the
truncate elliptical mental. Upper labials 15,
increasing in size backwards, and considerably
smaller than the facials immediately above
them. Nasal situated on swollen rostral ridge,
rounded polygonal, convex, pierced towards its
top by a small circular nostril which enters
downwards and forwards, and separated from
labials by 5-6 rows of facials.

Scales on top of head strongly keeled an-
teriorly, becoming rugose posteriorly, smallest
above supraciliary ridge, largest prefrontally
and around interparietal. Dorsals weakly im-
bricate, weakly keeled, mucronate, and sparsely
mixed with isolated tubercles (enlarged,
slightly raised, unicarinate scales, their keel
terminating in a short spine or mucro), which
are highest on sacrum. The largest tubercles
tend to be aligned in four longitudinal rows,
two on each side of dark vertebral stripe, which
consists of two rows of uniform black scales,
slightly larger and more strongly imbricate and
keeled than other dorsals. Tubercles extend to
hase of tail, which is otherwise covered above
with keeled mucronate scales, and below with
non-imbricate scales whose keels are aligned
more regularly. Scales on upper surface of
limbs more strongly keeled. Two highly spinose
scales towards posterior edge of thigh.

Scales under head mostly smooth and weakly
imbricate. Ventrals largest on chest and tend-
ing to be striate or weakly keeled. Scales on
lower surface of limbs much smaller and less
strongly keeled than those above. Scales on
palms and soles still smaller but more sharply
keeled, distally becoming spinose, especially on
palm. Lamellae under fourth toe 16-18, strongly
bispinose, except proximally where the outer
spines are scarcely discernible. Claws short and
stout. No pores.

Dorsal ground colour grey. Body and tail
crossed by darker grey bands (of uniform width
and as wide as interspaces), one across neck
and four across body, all of them only slightly
darker than ground coloration and thus barely
discernible. Caudal bands darker and more
prominent, at least on proximal half of tail.
More conspicuous than the body bands is a
dark grey or black vertebral strip from occiput
to lump. Much fainter are a dark discontinuous
dorso-lateral and lateral line, discernible only
where they cross the transverse bands. Limbs
pale greyish brown, obscurely banded with
darker brown. Entire under surface white ex-
cept for pale greyish brown flecking round jaws.

Distribution. From the Kimberley Division of
Western Australia east into the Northern Terri-
tory.

Comments. The above description is of speci-
men R 13638, collected by Prof. C. L. Camp on
June 23, 1960, which constitutes the first record
for this State and the only known specimen
apart from the type.

Tympa nocry ptis cephala gigas Mitchell

Tympanocryptis cephalus gigas Mitchell, 1948, Rec. S.
Aust. Mus. 9: 65. Between Ashburton and Gascoyne
Rivers, Western Australia.

Material examined. R 3990 (between Wells
40 and 43. Canning Stock Route) ; R 13363
(Jigalong); R 13640, R 17689 (Turee Creek);
R 14C60 (Wandagee); R 19111 (“Bernier
Island'’); R 13476 » 20 miles S of Dongara); R
19099 (11 miles N of Mt. Magnet); R 13639
(Yalgoo>; R 13991 (Paynes Find); R 19649
(Mt. Magnet); R 15802 (Belele); R 2006 (Mee-
katharra) ; R 7302 (Wiluna); R 1606 (Lake
Wells); R 19770 (Albion Downs); R 1240, R
1242. R 1305, R 1360, R 1757-8, R 13323 (Laver-
ton); R 20-1 (“Goldfields”); R 7067 (Grants
Patch); R 3748 (Hampton Hill); R 4329. R
6542-4 (Kalgoorlie) ; R 6434-5, R 6754-5 iKur-
rawang); R 2346 (Londonderry); R 19100 (10
miles W of Coolgardie).

Description. Body much depressed. Neck
much narrower than head, which is usually a
little wider than long. Tail short, tapering
rapidly. Supraciliary ridge moderately acute
and (viewed from above) bowed out but not
concealing eye. Rostral ridge obtuse, swollen
in front of nasal. Facial trough not deep or
ang^ar. Snout -vent length of largest specimen
(R 6755) 57 mm.

Rostral about twice as wide as high, slightly
narrower than combined width of first two
labials, and considerably wider than the trun-
cate elliptical mental which is usually deeper
than wide. Upper labials 12-16, smaller than
the facials immediately above them. Nasal
situated immediately below rostral ridge,
rounded polygonal, slightly convex, pierced a
little above its centre by the small elliptical
nostril which enters downwards and slightly
forwards, and separated from labials by 5-6 rows
of facials.

Scales on top of head rugosely tectiform or
conical; largest and highest prefrontally, on
temples, and around interparietal; smallest and
lowest above supraciliary ridge and on mid-line
between orbits. Scales on nape small and
strongly imbricate, becoming larger and weakly
imbricate or juxtaposed on back. Dorsal
tubercles large, raised, unkeeled, trispinose (the
central spine much the highest and sharpest)
and arranged side by side to form short trans-
verse or oblique ridges; the two kinds of tuber-
cular ridge generally alternate down the back,
the transverse ridges tending to be distant from
the mid-line, and the oblique ridges close to it.
The tubercles extend to base of tail, the re-
mainder of its upper surface, like its entire
under surface, being covered with uniform
scales, their keels aligned longitudinally.

Scales on chin imbricate and -weakly keeled.
Ventrals weakly imbricate and keeled on chest,
becoming smooth and juxtaposed on abdomen
Except at elbow and knee, scales on upper sur-
face of limbs large, strongly keeled and mucro-
nate, and mixed on fore -arm and thigh with

48

tubercles like those on back. Scales on lower
surface of limbs similar to subcaudals. Scales
on palms and soles small, strongly keeled, and
spinose. Lamellae under fourth toe 16-18, bi-
carinate and bispinose, the outer keels and
spines much the smaller and sometimes barely
discernible. Claws strong and short. Two pre-
anal pores in most specimens.

Dorsal ground colour brownish red. A dark
reddish brown band across narrowest part of
neck, sometimes broken by 3 short white longi-
tudinal lines (one vertebral, two dorso-lateral).
8-11 dark bands across tail. The first caudal
band, situated only a little behind level of vent,
is transverse and has a sharply defined darker
posterior edge. The next two bands are in-
creasingly oblique (from above they appear V-
shaped with the angle pointing backwards),
similarly with a dark posterior edge, and
separated from each other (as are the first and
second bands) by a large whitish or pale grey
space. The remaining caudal bands are more
obscure and uniform in colour (i.e. not darkly
margined) and tend to be broken and to ex-
tend to lower surface of tail. Limbs vaguely
banded. Entire under surface whitish.

Distribution. Western Australia, from the
Ashburton River drainage and the Great Sandy
Desert south to Dongara, Paynes Find, and the
vicinity of Kalgoorlie.

Comments. The above description is based on
specimens from the Eastern Goldfields, the only
region from which there is adequate material.
Northwards, the following changes occur:

(1) the head scales become lower and less
rugose and tend to be simply unicari-
nate or smooth;

(2) the head becomes narrower than long;

(3) the tubercles become smaller, lower, less
numerous and not so inclined to be ar-
ranged in definite and symmetrical
patterns ;

(4) the nasal tends to move up on to ros-
tral ridge.

North to at least the Ashburton, there is no
geographical variation in colouring, apart from
a tendency for the flanks to become dark grey,
which may diffuse on to back broadly or in the
form of vaguely transverse bands. The head
may be similarly diffused with grey above orbits
and temples.

Though the extreme populations are fairly
distinct, they are connected by a chain of inter-
mediates. Moreover, the various north-south
dines seem to operate independently of each
other and any division into two races would
necessarily be arbitrary. Mitchell’s concept of
gigas is therefore broadened so as to include all
populations south of the Ashburton.

Tympanocryptis cephala cephala GUnther

Tympanocryptis cephalus Gunther 1867. Ann. Mag.
Nat. Hist. (3) 20; 52. Nickol Bay, Western Australia
(Duboulay).

Material examined. R 12495 (Mardie).

Distribution. A small sector of the north-
west coast of Western Australia, west of Roe-
bom* ne.

Comments. As nominate cephala is only
known from the two syntypes in the British
Museum and from our R 12495, it is not possible

to give a full description of it. All that can be
done at present is to show briefly how r cephala
differs from the much better knovm gigas.

According to the original description, Boulen-
ger (1885), and the photograph of a syntype
(Mitchell 1948), nominate cephala differs from
gigas as follows:

(1) the body is not so depressed, and the
head is narrower

(2) the dorsal tubercles are not (or not so
markedly ) arranged into welts

(3) there are darker bands (‘'more or less
indistinct”) across the body

(4) the caudal bands are separated from
each other by much narrower inter-
spaces.

The specimen from Mardie (only 60 miles
south-west of the type locality) is very young
(snout-vent length 26.5 mm) but seems essen-
tially similar to the syntypes. It further differs
from our series of gigas in having

(1) more strongly keeled and imbricate
ventrals (this is especially noticeable on
the abdomen, where the scales show
no tendency to become smooth and
juxtaposed)

(2) fewer lamellae (13) under the fourth
toe. compared with 14-19, seldom fewer
than 16, in gigas

(3) much weaker spines on palms, soles and
subdigital lamellae (indeed the last 2-3
lamellae are quite smooth).

In some characters, e.g., the relative smooth-
ness of head scales and the position of the
nostril with respect to the rostral ridge, typical
cephala is nearer to northern gigas than the
latter is to southern gigas. In other charac-
ters, especially coloration, there is sharp
enough break to warrant at least tentatively the
recognition of the tw r o races.

Measurements

The following measurements were made on
all specimens: (1) total length, (2) length of
tail, (3) snout-vent length, (4) length of hind-
leg (including longest toe, without claw) — all
to the nearest 0.5 mm; and (5) length of head
(measured axially from snout to end of low r er jaw)
and (6) maximum width of head — both to near-
est 0.1 mm. All these measurements were ex-
pressed as per cent, length of trunk (snout- vent
length minus length of head), and the mean and
standard deviation from the mean were calcu-
lated for each sample; with the exception of
head width they are given in Table I. For the
number of specimens examined in each taxon

TABLE I

Mean length, of head, tail , fore-leg. and hind-leg, all
expressed as per cent, length of trunk ( with standard
deviation in brackets ).

Head

Tail

Fore-leg

Hind-lag

l. 1 meat a

40 2(4 -2)

220-1(22 -4)

59-3(4-5)

93-5(9-1)

f. rent mu'!

37 7(2

205*5(10*3)

55-8(3-4)

82-1(5-2)

tetraporophorn

37 ’5

210

03-3

101-7

pnrticep*
uniform is

36 ‘2(2 -8)
42

190-2(21 - ft)

54-0(4-8)

58

99-4(7-5)

85

c. cephala

43

141

00

92

e. tjvjos

40*6(4*6)

175*0(18-9)

65-7(4-8)

92-3(6-9)

49

TABLE II

Number of specimens examined , mean length of
trunk, ratio fore-leg to hind-leg. and ratio width to
length of head ( with standard deviation in brackets).

Number

of

Specimens

Length

of

Trunk (mm)

Fore-leg

as % ,,f
Hind-leg

Width of Head

as % of its
Length

l. lineata

Id

36 • 4

63-6(3*3)

83-4(4*6)

l. centralis ... 1

18

34 9

67-1(2-7)

87-6(3-4)

tetraporophora

3

42 0

62-0

82-0

parviceps

19

28-8

54-3(3-9)

87*0(2-5)

uniformis

1

24

68

91

c. cephala ....

l

18-5

65

1 00

c. giyas

36

29-3

70-6(3-3) |

99-6(5-8)

see Table II. which also gives mean length of
trunk and mean ratio length of fore-leg to hind-
leg and width to length of head.

Tables I and II show that the lineata group
1 including tetraporophora ) comprises larger and
longer-tailed lizards than in the rest of the
genus.

Despite its broken distribution (in isolated
ranges surrounded by sandy deserts or otherwise
unsuitable lowlands), lineata centralis exhibits
remarkably little variability in the relative
length of its appendages. In contrast, the more
continuously distributed lineata lineata is highly
variable in structure (probably exceeding in this
respect cephala gigas, which undergoes marked
clinal variation in several characters). How-
ever. 1 . lineata is much more uniform in colora-
tion than l. centralis , which tends to lose its
dorsal pattern from east to west.

Discussion

As illustrated by the map in Mitchell (1948,
p. 58), the genus Tympanocryptis attains its
greatest diversity in the Lake Eyre basin. In
Western Australia, however, the several taxa
tend even more than in Nephrurus (Storr 1963 )
to be allopatric. There is probably widespread
sympatry only in the Kimberley Division (be-
tween uniformis and tetraporophora ) . Further
collecting in the far east of the State may re-
veal the presence of tetraporophora within the
range of lineata. Elsewhere the probability of
sympatry is small. As parviceps is restricted
to coastal dunes, it will at most only approach

the range of cephala , which alone represents the
genus on the great Precambrian block from the
Pilbara south to the Goldfields.

Since this study was begun it has become in-
creasingly doubtful whether the genus Tym-
panocryptis is tenable as it is at present under-
stood. As mentioned earlier, parviceps is in
some important respects more like Amphibolurus
adelaidensis than any Tympanocryptis . On the
other hand, the 2-pored species of Tympano-
cryptis, especially the lineata group, are remi-
niscent of various species of Diporiphora.

The closure of the ear aperture has, perhaps,
no great phylogenetic significance. At any rate*
if Tympanocryptis is to stand, other characters
should be found. The relatively small head
cannot be reckoned an independent character*
for it is almost certainly related to the atrophy
of the ear, which in turn is probably connected
with their small size and sedentary and cryptic
habits. There thus remains at present only
their tendency to spininess, which however is
shared with such species of Amphibolurus as
adelaidensis and barbatus. Indeed, the latter,
with its extremely depressed body, differs from
Tympanocryptis only in its larger size and the
presence of an ear aperture.

The generic classification cf Australian
Agamidae has remained exactly as Boulenger
left it in 1885, and a thorough and simultaneous
revision of all Australian genera is long over-
due. Since such a study has been planned by
another worker, I have not pursued the matter
further and have retained for the present
Tympanocryptis in its traditional sense.

References

Boulenger. G. A. (1885). — “Catalogue of the lizards in
the British Museum (Natural History).”
I. 2nd ed. (Brit. Mus. (Nat. Hist.):
London.)

Loveridge, A. (1934). — Australian reptiles in the Museum
of Comparative Zoology, Cambridge, Mas-
sachusetts. Bull. Mus. Comp. Zool. 77 (6).
Lucas. A. H. S., and Frost, C. (1896). — Reptilia in
“Report on the work of the Horn Scientific
Expedition to Central Australia. II.”
(Dulati: London.)

Mitchell, F. J. (1948). — A revision of the lacertilian
genus Tympanocryptis. Rec. S. Aust. Mus.
9: 57-86.

Storr. G. M. (1963). — The gekkonid genus Nephrurus in
Western Australia, including a new species
and three new subspecies. J. Roy. See. W.
Aust, 46: 85-90.

50

8. — On Beryl from Western Australia

By P. L. C. Grubb*

Manuscript received— 19th November . 1963

Varieties of pink, light blue, and white beryl
are described from two localities in Western
Australia. Their contrasting features are
examined by optical, physical, x-ray diffraction,
chemical, and spectrographic means. The slight
differences in colour are considered to be caused
by a variety of factors including trace element
distribution, minute partially crystallised fluid
inclusions, and the degree of incipient kaoli-
nisation. The effect of alkali ion content on
unit cell parameters is examined but fails to
reveal any consistent relationship. Finally, the
nature and composition of former liquid inclu-
sions is briefly analysed and comparisons with
other reported samples drawn.

Introduction

Of the three varieties of beryl discussed here,
two (the light blue and white) were collected
from the undulating Mt. Marion pegmatite
sheet (Lat. 31° 5' S„ Long. 121° 30' E.), where
they form part of a single large zoned crystal.
The third (a reportedly caesium-bearing type)
was obtained from a pegmatite body at Wodgina
(Lat. 21° 25' S., Long. 118° 30' E.).

Optical and X-Ray Diffraction Properties

Zoning . — Although isomorphous replacements
of several elements within the beryl crystal lat-
tice are common, macro- and microscopic zon-
ing of single crystals appears to be quite rare.
The total diameter of the zoned Mt. Marion
crystal is of the order of 3 cm and the outer
cloudy white zone is between 4 and 8 mm thick.
The junction of these two zoned portions is
quite sharp, but in one section it consists of a
gradational fine-scale repetition of these.

Refractive indices . — Slight variations in re-
fractive indices between the three main beryls
exist ( Table 1), and it is noteworthy that in the
pink Wodgina beryl the incipiently kaolinised
portions have refractive indices up to 0.002
higher than the fresh grains.

TABLE 1

Refractive indices of Western Australian
beryls

1

2

3

4

Xe

No

1 • 577
1 • 586

1 * 57'i
1 • 588

1*581
1 • 591

1 *576
1 • 583

1. Light blue core 1 . , . ,

2. Cloudy white margin I 10 ' MaHon h, ' r > 1 cr - vs,al

3. Pink beryl from Wodgina

4. Clear green beryl from Yinniet hurra (Nourish 1950)

Inclusions . — In all three beryls microscopic
rod-like inclusions occur, these being most con-
spicuous in the clear light blue and pink
varieties (Fig. 1 and 2). Despite their super -

* C.S.I.R.O., Mineragraphic Investigations, c/o University
cf Melbourne, Parkville, N.2, Vic. Australia.

ficial resemblance to those figured by Little
(1960) however, these differ in being partially
or sometimes wholly crystalline so that on fine
pulverising — particularly in the Wodgina beryl —
they can be isolated from larger grains and their
refractive indices compared with these of im-
mersion oils.

Fig. 1. — Rod -like quartz inclusions in Mt. Marion beryl
Plain polarised green light. X 250 magnifications*

Fig. 2.— Fluorite inclusions in pink beryl from Wodgina
Western Australia. Plain polarised green light. X 480
magnifications.

Identification of these solid inclusions was
carried out by x-ray diffraction with the result
that traces of quartz were discerned in the light
blue variety and of fluorite in that of Wodgina.

51

It is also of interest to note that whereas Nor-
rish ( 1950 > observed minute tubular inclusions
of “water” parallel to the “c” crystallographic
axis in beryl from Yinnietharra, he found no
evidence of this water in a complete chemical
analysis of the same specimen.

Alteration . — Although no traces of secondary
alteration products were detectable by x-ray
diffraction means, close scrutiny under high
magnification revealed a slight dusting effect of
some grains in the cloudy white variety and to
a lesser extent in the pink variety. This, as at
Mt. Marion, is coupled with a slightly higher
loss-on-ignition of the outer white crystal mar-
gin compared with the clear light blue core. It
seems logical to conclude therefore, that the
conspicuous white appearance of some beryl
from Mt. Marion is caused by incipient kaolini-
sation.

Unit cell parameters . — Calculation of the unit
cell parameters for the three beryl types from
Mt. Marion and Wodgina show slight divergen-
cies but this appears to bear no evident rela-
tionship to their chemical composition (Table
2). No support therefore can be found here for
the hypothesis of Sosedko (1957) and Frank-
Kamenetsky & Sosedko (1958) that the “c” unit
cell parameter increases stoichiometrically with
the alkali content due to the crystal lattice dis-
tortion caused by the substitution of small beryl-
lium ions by larger alkali ions. Neither does there
appear to be a very constant relationship be-
tween the “a” unit cell parameter and total
alkali ion content as suggested by Schaller,
Stevens and Jahns (1962).

Specific gravities . — Specific gravities deter-
mined using the Berman precision balance show
the highest values for the pink Wodgina beryl,

whereas the cloudy white Mt. Marion beryl, be-
ing slightly kaolinised, is less than its light
blue counterpart (Table 2).

Chemical analyses . — Partial chemical analyses
of the three beryl types from Wodgina and Mt.
Marion are listed in Table 2. In addition Quali-
tative and semi-quantitative speetrographic.
anaylses of each showed slightly higher contents
of cobalt and manganese in the pink Wodgina
sample, slightly more titanium and zinc in the
blue Mt. Marion type, and rather more alumi-
nium, magnesium, calcium, iron, chromium,
sodium, and lithium in the white Mt. Marion
beryl. It is of interest, however, that no caesium
could be detected in any of the samples studied.

Conclusions

Inclusions in beryl, although rare, have re-
cently been described from several world
localities, yet for the most part these are be-
lieved to be fluid bodies consisting essentially
of water, carbon dioxide, and sulphur (Little
1960 and Zwaan 1958). Eppler (1958), on the
other hand, considers that the chatoyancy ex-
hibited in some beryl crystals is due to minute
oriented ilmenite needles. Subsequently (1960)
in an almost opaque beryl sample showing
asterism, he recorded thin tabular pyrrhotite
crystals, crystal groups of quartz and epidote,
rod-like apatite, and clusters of pyrite. Most
recently Schaller, Stevens, and Jahns (1962)
have recorded oriented inclusions of quartz,
tourmaline and fluorite in an unusual beryl from
Arizona.

The inclusions in both the Mt. Marion and
Wodgina beryls, although now largely crystal-
line, appear to have consisted of the late resi-
dual fluid existing while the beryl was in the

TABLE 2

Chemical composition and calculated unit cell constants for beryl

1

2

3

4

5

(5

7

Si O n

64-85

05-32

04-17

01-88

Bed

1315

12-00

11-82

10-54

alo..

17-52

17-77

17-42

17-10

0-37

0-13

0-12

0*08

JVIgO

0 14

0-25

0*21

0*22

CaO

trace

trace

0-07

nil

0-20

0-44

0-44

Li .( )

0-34

0-60

0-04

0 • 52

0-30

1-23

0-00

MiiO

0-01

l race

nil

trace

-Na.,Q

0-04

1-66

1 • 00

0 • 94

1-27

1-39

2 - 50

K ..6

0-10

0-10

0 16

0-16

nil

nil

nil

(Rto C*),o

nil

nil

nil

0-08

0-27

0-07

4-13

ICO -f 105 r 'C ....

2-10

2 55

1-84

2*19

1-76

1 -88

2-26

11*0— 105 t: C

0-70

0 • 00

0-00

nil

0 • OS

0-00

0-16

Tib*

trace

trace

nil

0*01

0-01

0-0L

p..o;

nil

-r-

<V> a

F

trace

trace

nil

nil

l - lL

• M-

Cl

nil

....

99-93

100-02

99*96

99-92

Unit Oil Constants

e/a

Specific Gravity

9- 15
9-21
1 ■ 00(1
2' 71

9- 19
917
0-998
2-70

9-20

9*21

1001

2-84

9-188
9 • 1 8 9
1 • 000
2-71

1. Clear light blue core zone of Ml. Marion beryl. Western Australia.

2. Cloudy white margin of Mt. Marion beryl. Western Australia.

3. Light, pink beryl from Wodgina, Western Australia.

4. Clear light green beryl from Yinnietharra, Western Australia. (Xornsh 1950.)

5. Green beryl from pegmatite, C.S.S.R, "I

White beryl from pegmatite. C.S.S.ll. ^.Sosedko, T.A. (1957).
tik beryl from pegmatite, T'.s.s.K. j

9 • 202
9 183
0*998
2-72

9-202 9-200

9-209 9 • 227

1- 0008 1003

2 - 75 2-78

52

process of crystallising. So that while at Wod-
gina the residuum was essentially fluorine-rich,
at Mt. Marion it was correspondingly siliceous.

Although the cloudy white nature of some
Mt. Marion beryl is readily explained through
late stage kaolinisation, the pink and blue
colours of the remaining two varieties is not so
readily understood. Nevertheless, it is considered
that these characteristics may arise through the
contrasting distribution of trace elements; the
pinkish hue being due to the presence of cobalt
and manganese, and the blue probably to
chromium. An additional factor, particularly
in the pink Wodgina beryl may be the anoma-
lous refraction effects produced by its minute
fluorite inclusions.

Acknowledgments

For critical reading and comments of this
note the writer is indebted to Mr. A. J. Gaskin,
Chief of Division of Applied Mineralogy,
C.S.I.R.O.

The partial chemical analyses of the new
varieties of beryl were carried out by Mr. T. H.
Donnelly, to whom much thanks is due.

References

Eppler, W. F. (1958). — Notes on asterism in spinel and
chatoyancy in chrysoberyl, quartz, tour-
maline, zircon, and scapolite. J. Gemm. 6-
251-263.

(1960). — An unusual star-beryl. J. Gemm.

7: 183-191.

Frank-Kamenetsky, V. A., & Sosedko, T. A. (1958).— On
the character of isomorphism in alkaline
beryls. C. R. Acad. Sci. U.R.S.S. 118: 815-
817.

Little, W. M. (1960). — Inclusions in cassiterite and
associated minerals. Econ. Geol. 55 (3)-
485-509.

Norrish, K. (1950). — An x-ray study of West Australian
beryl. J. Roy. Soc. W. Aust. 34: 1-16.

Schaller, W. T., Stevens, R. E., and Jahns, R. H. (1962).

— An unusual beryl from Arizona. Amer.
Min. 47: 672-699.

Simpson, E. S. ( 1948) .—“Minerals in Western Australia.”
Vol. 1 (Government Printer: Perth.)

Sosedko, T. A. (1957). — The change of structure and
properties of beryls with increasing
amounts of alkalies. Zap. Vsesoyuz. Miner.
Obshch. 86: 495-499.

Zwaan, P. C. (1958). — Remarks on the inclusions in an
aquamarine. Proc. Acad. Sci. Amst. (B) 61:
260-264.

53

9. — Contributiones Florae Australiae Oecidentalis XII 1

By C. A. Gardner*

Manuscript received — 16th July, 1963

Descriptions are given of 35 species previously
undescribed. They have been selected from a
large number of plants awaiting description be-
cause of their more than usual interest. The
Actinostrobus has for long been confused with
A. pyramid alls, and is by far the most wide-
spread species of this endemic genus; the
Casuarina. a plant of unusual habit within
the genus is of very localised occurrence. The
seventeen Proteaceae have been included since
they will be incoi pcrated in that portion of
the “Flora of Western Australia” nearing com-
pletion. During the preparation of this work
it was found that considerable confusion existed
in the genus Dryandra, of which five new species
are now described. Of the species of Banksia,
one tB. pilostylis) has been confused with B.
media, while the species of Grevillea are frem
localities previously little known botanically.

Two species of the Papilionaceae are included
because they belong to the toxic sections of
Oxylcbium and Gastrolobium the latter being
already a confirmed toxic species, and the former
will prove to be so.

In the Myrtaceae two species of Eucalyptus
are described; one from Kimberley apparently
has a restricted range while the other species has
been confused with a form of Eucalyptus
eudesmioides, but with which it has no true
affinity. One new Regel ia is described, and the
names of two well-known species have been
changed in accordance with International Rules.
The two species of Wehlia are perhaps the most
attractive of the genus, and both are from the
Eremaean Province. The species of Darwinia
bears a strong affinity to a South Western
species, but grows in the Eremaea.

In the Chloanthaceae are three spectacular
plants, a Lachnostachys ; a Newcastlia with
extraordinary foliage, and also of tall erectly
branched habit, its golden-tomentose leaf under-
surfaces making it at once a plant of singular
attractiveness; and thirdly the most showy of
the species of Pityrcdia, named P. spectabi’is.
a species in danger of extinction owing to agri-
cultural extension in the area in which it grows.

In the Labiatae is a small compact rare
Prostanthera with pale blue flowers.

In the Goodeniaceae, a cushion-like Leschen-
aultia with pale blue corollas and a white
throat is described together with a magnificent
biennial Goodenia with long spike-like in-
florescences, the flow'ers almost hidden in the
dense sPk of the spikes.

Leptospermum sericeum, for many years re-
garded as synonymous with Kunzea sericea.
has been restored as a valid species; it has been
included for its undoubted horticultural value,
and is probably the largest -flowered and most
attractive species of this genus in Australia.

CUPRESSACEAE

Actincstrcbus arenarius C. A. Gardn. sp. nov.

Frutex 2-5 m. altus, ramis plusminu c ve divari-
cates, rigidiusculis, ramulis triquetris; folia
ternata. squamiformi-triquetra, ad basin adnata,
sursum patula. parte libera late ovata, acuta.
Amenta mascula ignota. Strobili foeminei mani-
feste pedicellari, pedicelli squamis abbreviate
appressis obtecti; strebilus ovoideo-globosus,
breviacuminatus, squamis exterioribus vacuis,
tenuibus, parum auctis, arete appressis, in-
teriores gradatim majoribus, ovatis, marginibus

* Herbarium, Western Australia.

tenuis; squamis fertilibus induratis, ovato-
lanceolatis, in apicem attenuatae erectis; semina
magna, pauca, 2- vel 3- aiata. luteo-brunnea.

Hab. in regionibus interioribus arenosis; stationes
maxime australes a fiumine Murchiscn usque oppidem
Bruce Rock valde insignis, occidentalem versus Darling
Range non transgrediens; frequenta cccurrens in distr.
Irwin et Avon in arenosis apertis. Typus prope Tam-
min, Gardner n. 610. Novem. 1920.

Formerly included with A. pyramidalis Miq.
this species remains distinct in the more glau-
cous foliage and cones, the fertile scales of
which have erect and somewhat acute, not in-
curved apices, they are also larger, and glau-
cous, the leaves cf A. cirenarius are ovoid-
deltoid in the free portion, whereas in A. pyra-
midalis they are ovate and more spreading.

Leaves mostly 2 mm. long, the free part
slightly less than 1 mm.; cones 1.5- 1.6 cm.
long, 1.6 cm. broad; seeds 1.1 cm. long.

CASUARINACEAE

Casuarina ramosissima C. A. Gardn. sp. nov.

Frutex probabiliter dioicus, usque 1 m. altus,
ramis ramulisque subconforibus, numerosis.
verticillato-ramosis, ramulis 2-3- articulatis,
cinereo-tomentcsis, fere rigidis, patentibus,
glaucis, prof unde sulcatis demum glabris; inter -
nodia (basalia excepta) 1-2 cm. longa, dentes
4, erecti, deltcidei, obtusi, ciliati, apice nigri.
Amenta mascula ignota. Strobilis sessilis, verti-
cillatis vel oppositis, breviter cylindricis vel
ovoideo-cylindricis, basi truncatis, circ. 1 5 cm.
longis et 1 cm. latis; bractea magna late ovata.
dorso rufo-tomentosa, apicem versus in appren-
dicem contracta, apice late ovata, atro-fusca
marginibus fimbriolata; bracteolae basi con-
natae, vix exsertae, dorso lato verrucosa
areolatae et dense fibreso-piloso; achaenia atro-
fusca, crassa vel turgida et breviter pilo:a, ala
obliqua, vix scariosa.

Hab. in distr. Avon prope Dandarragan, in arenosis
glareosis in fruticulosis, Gardner 9013, (TYPUS).

In its compound branchlets this species is
related to C. microstachya Miq. and C. Drum-
mondiana Miq., all the remaining species hav-
ing undivided branchlets. From C. microstachya
it differs in its taller stature, longer ashy-
tomentose branchlets, very different bracts, and
sulcate. not tetragonal internodes. From C.
Drummcndiana it can be distinguished by the
much longer, not glabrous internodes, larger
fibrous- hairy bracteoles and darker coloured,
less hairy and shorter winged achenes.

PROTEACEAE

Concspermum sericium C. A. Gardn. sp. nov.

§Euconosperuum Benth.

Fruticulus vel suffrutex habit u simile C.
coeruleum, cauli scapiformi, basi folicso sericei,
superne aphylli, dichotomcso-ramoso vel scapi-

54

formi; folia erecta, lingulata elongata, usque 12
cm. longa, uninervia, nervoso-marginata, glabra,
infra medium attenuata in petiolum basi dila-
tata, acuta apicibus ustulata.

Spicae breves, fere in capitulum, contractae,
multiflorae, longe pedunculatae folia longe
superantes et quasi corymbum floriferum pul-
chrum constituentes ; bracteae florigerae lanceo-
latae basin versus attenuatae, in apicem
acuminatae productae cymbiformae, subtus et
marginibus sericeis, pallide coeruleae, perianthio
subaequales.

Perianthium 4 mm. longum, pulchre coeru-
leum, tubum glabrum, limbi labium superius
integrum, leviter concavum dorso breviter
birsutum, acutum. marginibus elegante ciliatum;
labium inferior prof unde trilobum. lobi anguste
lanceolati, acuti, marginibus longe ciliati,
dorsale pilosi.

Hab. In distr. Irwin prope flumen Hill juxta Badgin-
garra in arenosis apertis. flor. m. Decembri. F. Lullfitz
n. 2156 (TYPUS).

The affinity of this species is with C. coerulewn
Meissn. and C. debile Kipp, differing from the
former in the narrow cymbiform not woolly
bracts and the densely ciliate lobes of the lower
lip, and from the latter in much longer obtuse
leaves, and the indumentum of the flowers.

Lambertia orbifolia C. A. Gardn. sp. nov.

Frutex 2-2.5 m. altus, ramis ramulisque erectis,
villoso-tomentosis pilisque longis intermixtis,
pilis patentibus. Folia opposita, erecto-patula,
sessiles, orbiculares, integra, basi cordato-
orbiculares, subcucullata, conspicue penninervosa
et reticulata, 2 cm. longa, 1.8 cm. lata. Involucri
4-6 flori, axillares vel ramulis terminati; bracteae
exteriores oblcngo-ovatae valde obtusae vel
tiTincatae, dorso dense tomentosae, interiores
multo longiores sed similes.

Perianthium rubrum, 4.7 cm. longum; seg-
menta. basin attenuata, superne dilatata et
leviter curvata, sursum contracta et spiraliter
revoluta (obtorta) et hirsuta, limbum angustum,
erectum, breviter hirsutum, apicibus breviter
barbatis; squamae hypogynae lineari-lanceolatae,
acuminatae, liberae. Ovarium longe et dense
pilosum; stylus filiformis.

Hab. in distr. Stirling baud procul King George’s
Sound, juxta flumen King, in arenosis glareosis in
fruticetis, fl. m. Junio, leg. A. J. Gray (TYPUS).

Differs from all other species in the orbicular
cordate leaves, but closest to L. inermis R. Bt\,
from which it differs in the larger hairy
perianth, the narrower truncate bracts, and
much broader sessile leaves.

Grevillea globosa C. A. Gardn. sp. nov.

§Plagiopoda.

Frutex erectus usque ad 2 m. altus, ramis
ramulisque erectis, ramulis juvenilibus sericeo-
pilosus. Folia erecta. pinnatisecta, 7-9-fida,
lobis tenuis sed rigid is, anguste linearea vel
subteretia, leviter pubescentibus, subtus bisul-
catis, demum glabrescentibus, glaucis.

Racemi terminali, breviter pedunculati foliis
multo breviores; peduncuh rufo-tomentosi;
pedicelli 4 mm. longi: sericei. Flores zygomor-
phicum racemo terminali. Perianthium rubrum,
basi dilatatum, segmenta perangusta, lineari-
oblonga, extus sparse sericeo-villosa, sursum

attenuata, sub limbo globosa revoluta, intus
glabra, extus in partis inferioribus sparse
sericeo-villosa, limbo albo-villoso. Torus
obliquus; glandula hippocrepiformis, plana;
ovarium dense villosum, stipite ovario longiore;
stylus 17 mm, longus, parte inferiore sparse
hirsutus, ceterum glaber, curvatus, disco stig-
matico orbiculari, laterali terminatus. Fructus
ignotus.

Hab. in distr. Austin, Pindar 32km. septentrionalem
versus juxta confragosos lateritos, fl. m. Januario. F.
Lullfitz n. 2241 (TYPUS).

The affinity of this species is with G. tenuiloba
C. A. Gardn., from which is differs in the much
longer leaves with more numerous segments, the
short =t globose racemes and conspicuously
silky perianth-limb.

Grevillea ninghanensis C. A. Gardn. sp. nov.

Frutex nanus, usque ad 20 cm. altus, crebre
divaricate -ramosa, ramulis appresse-pubescenti-
bus. Folia fasciculata, subsessilia, in ramulis
patulis disposita, coriacea, viridia, linearia,

1- 1.6 cm. longa, mucronato-pungentia, basi
sensim attenuata vel subsessilia, marginibus
valde revolutis, glabra, supra convexa et laevia,
subtus bisulcata.

Racemi axillari, laxissimi, secundi, 2-4 cm.
longi, rhachis sparse pubescentibus, pedicelli

2- 3 mm. longi, appresse-sericei, perianthii tubus
10-11 mm. longus, sub limbo parvo globoso
revolutus, segmenta perangusta, basi dilatata,
extus sparse appresse-sericea, intus parte media
sparse appresse-sericea, superius glabra; torus
obliquus, fere orbiculatus; ovarium dense seri-
ceum, stipitatum; stylus circ. 12 mm. longus;
basin versus crassus, glaber, apice recurvatus,
disco stigmatico orbiculari, laterali terminatus;
glandula semilunaria, parvum prominens.

Hab. in distr. Austin, prope Ninghan, in acacietis in
argillaceis fl. m. Aug. Gardner n. 12502 (TYPUS).

This species is related to those of the Section
Plagioyoda which have entire leaves with closely
re volute margins, and loosely racemose flowers.
It differs from G. deflexa in having narrower,
shorter and more closely revolute leaves, erect
racemes, and shorter styles. It is also a much
smaller shrub with the branches close to the
ground.

Grevillea fulgens C. A. Gardn. sp. nov.

SPlagicyoda Eenth.

Frutex 90 cm. altus, ramis divaricatis vel
erectis, subflexuosis, ramulis laepe erectis,
juvenilia pubescentia.

Folia erecto-patentia, linearis, basi attenuata,
breviter petiolata, obtusa, rigide mucronulata,
7-8 cm. longa. integra (vel rarius 2-3 loba
lateralia nascentes) marginibus arete revolutis,
supra glabra, unisulcata, subtus bisulcata.

Flores zygomorphi, in axillis foliorum singuli
vel bini, et ramulis axillaribus terminantae;
pedunculi appresse villosiusculi circ. 5-7 mm.
longi. bracteati. bracteae exteriores lanceolatae.
brevae, interiores magnae, 2, obovato-cuneatae
truncatae, vel breviter dentatae, mucronatae,
dorso ferruginoso-tomentosae. sursum glabrae,
alabastri includens; pedicelli rufo-hirsuti, 5-6
mm. longi.

Perianthium scarletinum lilacino-suffusum,
sub limbo globoso recurvum, torus obliquus;
segmenta basi dilatata, extus ( limbo excepta*

55

glaber, tubus parte basali extus sparse hirsuta,
cetera glabrus; limbus apicem versus sparse
hirsutus; glandula hippocrepiformis; ovarium
stipitatum, dense albo-pilosum, stipite ovario
paulo breviore; stylus crassus, fere rectus, circ.

2 cm. longus, laxe et appresse pilosus, disco
stigmatico orbiculare, laterali.

Fructus rectus, ovoideo-oblongus, glabrus, 1-2
cm. longus styli longe persistens basi coronatus.

Hab- in distr. Eyre montem Desmond ad collis latera
petraea. in fruticulosus densus, ti. m. Aug-Septem.
Gardner 14070 (TYPUS) etlam 13718, 12883.

Although the position of the torus is that of
the species of the Section Plagiopoda, it has no
close affinities within this Section, nor with the
species of the Sect. Hebegyne. It has the bracts,
leaves and fruit of G. bracteosa Meissn. but is
otherwise unrelated.

Grevillea candelabroides C. A. Gardn. sp. nov.

§Cyclcidenia

Frutex vel arbor usque 5 met. alta; rami
patenti vel erecti, primum pubescentes demum
glabrescentes. Folia alterna, pinnatisecta, seg-
menta 8-9, lineari-filiformia, compressa, acuta,
glauca, subtus bisulcata, 8-13 cm. longa, in toto
usque 20 cm. longa.

Racemi numerosi, stricte erecti, in paniculam
dispositae, breviter pedunculati, folia excedenti;
pedunculi atro-purpurascenti, usque 5-6 cm.
longi; pedicelli 4 mm. longo, glabri.

Perianthium extus glabrum, intus sub medio
barbatum ceterum glabrum, tubum sub limbum
rovolutum; torus rectus; glandula hypogyna
annularia, subcupulari, ovarium glabrum, longe
stipitatum; stylus 8-9 mm. longus, glaber, disco
stigmatico obliquo in truncatum coronatum
attenuatum.

Fructus oblique ovoideus, compressus, sub-
laterali affixus, brunneus vel pallidus, saepe
apice subulato curvatus (basi styli persistente)
ad 1.3 cm. longus; semina subovalia, compressa,
circ. 1 cm. longa, griseo-viridia, undique latius
alata.

Hab. in distr. Irwin interiores, prope Mingenew et
Indarra septentrionalem versus extendit ad super
flumine Murchison, distr. Austin regionem invadit. in
fruticetis fl. m. Decembri et Januario, Gardner 12312;
in arenosis apertis prope Ajana Gardner 12062 (TYPUS).

The species is related to G. leucopteris Meissn.,
differing in the spruce-like compact habit, the
strictly erect candelabra-like panicles of racemes
which are both erect, the glabrous branchlets,
almost sessile racemes which scarcely exceed the
leaves, the glabrous and somewhat viscid fruit
and the presence of an oblique truncated stig-
matic cone. It flowers in summer, whereas G.
leucopteris blossoms in September and October.

Grevillea Gordoniana C. A. Gardn. sp. nov.

SCycloptera

Frutex elatus subarborescens usque 7 m. altus,
ramis ramulisque erectis, ramulis appresse-
pubescentibus.

Folia erecta, densa, sparsa, integra, rigida,
teretibus, juvenilis rufo-tomentulosa, exsulca,
breviter acuta vel mucronata, apicem ustulatum.

Panicula ampla, terminalis, speciosa, e racemis
numerosis, breviter pedunculatis, densifloris,
subglobosis, composita; pedunculi 3-6 mm. longi:
pedicelli gracili, glabri, 6-7 mm. longi: bracteis

56

aurantiaco-tomentosis, obtusis, diu persistent-
ibus, late ovatis, vel orbicularibusve, unguicula-
tis, erectis, intus glabris.

Perianthium luteum, 7-8 mm. longum, sub
limbo globoso revolutum; torus rectus, cupulatus,
segenta (valde inaequalibus) basi dilatata, extus
glabra, intus parte basali dense barbata, superius
glabra; glandula hypogyna cupulata; toro
adnata; ovarium glabrum, stipitatum, stipite
ovario longiore. glandula marginibus inserta;
stylus 7-8 mm. longus, glaber, apice breviter
recurvatis et disco stigmatico magno orbiculari
terminatus, in centro umbonato.

Fructus elongatus, anguste-oblongatus, sinu-
atus, 2 cm. longus, obliquus, verrucosus et
viscidus; semina lineari-sigmcidea, undique
latius alata.

Crescit in district Austin flumen Murchison 40km.
septentrionalem versus in lutoso arenosis in fruticetis,
fl. m. Decern . Gardner 14273 (TYPUS).

It is difficult to place this anomalous species
in any given Section of the genus. The lateral
umbonate stigmatic disc is not in conformity
with the Section Cyclcptera, but the inflores-
cence being paniculate, and the cupular torus
lined by the disc appear to warrant its retention
with the Section. The peculiar fruits which
do not freely emit the even more peculiarly
shaped seeds are without parallel in the genus.
The very short rhachis and subumbellate in-
florescence, the dehiscence of the perianth seg-
ments above the actual base of the perianth,
and the orbicular tomentose bracts are certainly
without parallel in the section Cycloptera, but
the attachment of the ovary stipes to the disc
margin make it impossible of insertion in the
Section Cycladenia.

This remarkable plant commemorates the
name of Mr. D. N. Gordon of Myall Park, Glen-
morgan, Queensland, who has cultivated
numerous Western Australian plants in his
garden, and who was instrumental in providing
the first specimens of this species through one
of his collectors (Mr. A. J. Gray).

Grevillea Rogersoniana C. A. Gardn. sp. nov.

§Anadenia

Arbor vel frutex usque 8 m. alta, ramis
erectis vel erecto-patentibus, crebo foliatis,
ramulis teretibus.

Folia petiolata, oblongo-spathulata, basin
versus sensim attenuata, integra vel cuneata
et bi- vel triloba, lobis ovato-triangularibus,
obtusis vel calloso-mucronatis, plana, laevia,
glauco-viridia, trinervis vel quinquinervis,
omnino glabra.

Racemi numerosi, densifiori et multiflori, in
paniculis terminalibus dispositi, 7 cm. longi;
pedunculi sparse appresso-sericei. Flores regu-
lares. breviter pedicellati, pedicelli 1 mm. longi.

Perianthium roseum, 8.10 mm. longum, rectum,
limbo anguste-elliptico, segmentis filiformis, basi
haud dilatatis, extus et intus glabris; torus
rectus; glandula hypogyna parva, semiannularia,
parum prominens. Ovarium glabrum, stipita-
tum, stipite 2 mm. longum: stylus circ. 1 cm.
longus conico-stigmatico recto, basi marginato.

Fructus ovoideo-globosus, dense resinoso-
tuberculatus, mucronatus, leviter compressus,
valvulis crustaceis. Semina 2, elliptico-
hemispherica alba, laevia, exalata, 1.5 cm
longa, 11-12 mm. lata.

Hab. in distr. Irwin septentrionalem versus ad Shark
Bay, in arenosis fruticetis, fl. m. Aug. Septem. Gardner
13517 (TYPUS).

Affinity with G. petrophiloides Meissn., from
which it differs principally in the greater size,
more robust habit, the flat entire or simply
lobed leaves, the lobes being broad and flat, also
in the larger more verrucose fruits which, al-
though viscid have much smaller resinous cavi-
ties and smaller tubercles than is the case with
G. petrophiloides. G. petrophiloides sometimes
has flat-leaf-segments, but these are always
narrow and proportionately very much longer
and always linear.

The species commemorates Mrs. W. Rogerson
of Gosnells who first directed my attention to
this handsome species.

Hakea tamminensis C. A. Gardn. sp. nov.

§Euhakea Benth.

Frutex horridus, 2 m. altus, crebo ramosus,
ramis divaricato-ramosis et intricatis, ramulis
tomentoso-villosis.

Folia alterna, lineari-teretia, divaricato-
patentia, vel inferiores erecta, rigida, coriacea,
stricta, glabrescentia, 5-6.5 cm. longa, basi non
contracta, apice rigide pungentia.

Racemi ad umbellas redacti, axillari. multi-
flori; breviter pedunculati, foliis multo breviori;
bracteae ovatae, glabrae, stipitatae; pedicelli 4
mm. longi, dense hirsuti. Floribus purpureis
vel rubris.

Perianthium sub limbo parvo globoso revo-
lution, 4-5 mm. longum, segmenta lineari-
oblonga, basi haud dilatata, extus sparse
hirsuta, intus glabra. Torus parvus, rectus;
glandula hypogyna erecta, hemisphaerico-trun-
cata, parum prominens; stylus breviter exsertus,
glaber, apicem versus incrassatus, disco stig-
matico laterali terminatus.

Fructus valde obliquus, crassus, ovoideus,
brevi crassique pedicellatus, 3.5 cm. longus, 2.5
cm. latus. gibbosus, prominenter et longitu-
dinaliter cristatus et supra prof unde sulcatus;
semina obovata ala ovato-oblonga, secus
marginem basique decurrente, hinc paulo valde
atra rugulosa.

Hab. In distr. Avon prope Tammin in arenoso
giareosis, fl. m. Julio-Aug. Gardner n. 11997 (TYPUS).

Affinity with H. circumalata Meissn., from
w'hich it differs principally in the densely
hirsute pedicels and the cristate, deeply and
broadly sulcate fruit.

Banksia laricina C. A. Gardn. sp. nov.

^Oncostylis

Frutex circ. 1 m. altus, ramis erecto-
patentibus, ramulis cano-puberulis. Folia
linearia, circ. 1.5 cm. longa, distincte petiolata,
mucronata. supra viridia et laevia, marginibus
arete revolutis, subtus unisulcatis vel canali-
culatis.

Spicae pedunculatae, ovoideo-globosae vel
breviter cylindricae, 2.5-3 cm. longae; bracteae
crassae, cuneatae. rufo-villosae, truncatae, apici-
bus albo-tomentosae.

Perianthium luteum, appresso-sericeum. sub
limbo ellipsoideo reflexum, 1.3 cm. longum, limbo
3 mm. longo; stylus glabrus tenuis sed rigidus
sursum hamosus, perianthio multo longior

arcuatim exsertus, conico-stigmatico atro, brevi,
exsulcato. Conus fructifer, 5 cm. longus, 8.5
cm. latus; folliculi lignei, glabri et prominentes,
valde exserti, 3. 3-3. 5 cm. lati. marginibus tenuis
et aliquando undulatis.

Hab. in distr. Darling prope fiumen Moore et Beer-
mullah. in arenosis depressis subhumidis. fl. rn. Jun-
Julio, Gardner, 12840 (TYPUS).

The affinity is with B. nutans R. Br., but the
bracts are white-tomentose at the truncate tips,
the cones erect, and the follicles are entirely
different, being much exserted without any per-
sistent perianths surrounding them, are much
broader and with thin margins, entirely unlike
those of any other species of the genus.

Banksia pilostylis C. A. Gardn. sp. nov.

SCyrtostylis Benth.

Frutex 2-3 m. altus, ramis ramulisque erecto-
patentibus, ramulis cano-tomentosis.

Folia sparsa, petiolata, patentia, linearia,
truncata. remote serrato-dentata, basi attenuata,
9-12 cm. longa, 12-15 mm. lata, dentibus
rigidis, acutis marginibus recurvis, supra laevia
glabra, nitentibus, subtus costato-reticulata
lacunis albo-tomentosis, tantum conspicuis
punctata, venulis glabratis.

Spicae ramis terminatae. cylindricae, 5-10
cm. longae, erectae, densiflorae,

Perianthium pallide luteum in vetulis demum
luteo-brunneum, dense villosum; limbo elliptico,
acuto, densissime villoso, limbo obtuso, 4 mm.
longo, apicibus ipso glabro; stylus arcuatus,
rigidus, laxe pilosus, perianthium paulo ex-
cedens, cylindrico-stigmatico vix latiore, brevo,
sulcato terminatus.

Folliculi transversi, glabri, obtusi, rufo-
brunnei, cum pallide luteo maculati, 2 cm. lati,
dorso convexi.

Hab. in distr. Eyre in arenosis inter flumina Oldfield
et Young septentrionalem versus ad collibus “Fitz-
gerald Peaks'’ in arenosis dunosis fruticetis. etiam
prope Starvation Boat Harbour, fl. m. Oct. Gardner
12149 (TYPUS).

The fact that this species is so common
within the area given above, suggests that it has
not been overlooked, and must have been seen
and collected by Baxter and Maxwell. It has a
certain resemblance to B. media R, Br., but the
leaves are much longer and more white under-
neath, and the spikes are also narrower and
longer. The hirsute style is quite distinct from
B. media. It is possible that Bentham in-
cluded this species with B. media notwithstand-
ing the difference in shape, size and indumen-
tum of the lower leaf surface (B. media having
ahvays much shorter and relatively broader
leaves almost glabrous underneath, a perianth
which has a much shorter indumentum, and the
limb of w r hich is soon glabrous). The style is
glabrous, and the spike much shorter' and
thicker. The two species are entirely distinct.

Banksia Bcnthamiana C. A. Gardn. sp. nov.

SCyrtostylis Benth.

Frutex erectus, usque 4 m. altus, ramis
erectis, ramulis brevibus, patentibus, cano-
tomentosis.

Folia erecta, rigida, elongato-linearia, usque
19 cm. longa et 6-7 mm. lata, utrinque glabra,
subconcolorata, apice breviter acuta, sursum
ultra medium remote dentata, dentibus acutis,
deltoideis, sinubus costa media parallelis, in

57

sicco utrinque luteo-viridia. pagina superiore
enevvia, laevia, subnitidula, subtus simulate
glabra et haud nitidula et lacunls albido-
tomentosis praeditis, tantum sub lente conspicue
parce punctata, marginibus crassa in petiole
brevi sensim attenuata, costa media supra
prominula immersa, subtus crassa.

Spicae densae ramulorum terminantibus
erectae, cylindricae, 7-9 cm. longae, usque 6 cm.
latae, folia pauca cbvallata: bracteae pubes-
centes, patulae, rufo-barbatae.

Perianthium aurantiacum, 2 cm. longum,
adpresse flavido-sericeum, laminae lineari-
lanceolatae. 5 mm. longae. Stylus rigidus,
rectus, demum sursum leviter curvatus perian-
th io leviter longior; stigma 4 mm. longa, stylo
glabro haud crassior sed basi constricta ab eo
distinct um continuum attenuatum, laeve.

Folliculi parvi, transversi 12 mm. lati
depressi tomentosi, perianthiis marcescentibus
obtengentibus.

Hab. in distr. Austin ab Dal wall inu ad lacum Monger,
attingit et prope oppidulum Wilroy, in arenoso
glareosis, fl. m. Decembri-Januario. prope Wubin Gard-
ner 12097; Wilroy Gardner 12075 t TYPUS).

The species is related to B. Eldei'iana F. Muell.
et Tate, but can be distinguished by the vesti-
ture of the perianth, as well as by the longer
erect flowering spikes, and the narrower sulcate
stigmatic cone. It is also a larger shrub.

Dryandra cynaroides C. A. Gardn. sp. nov.
iEudryandra Meissn.

Frutex erectus, 1-1.3 m. altus, sparse ramosus,
rami erecti, dense villosi; rami juveniles inter
folia squamis lanceolatis tectis. Folia subver-
ticillata, lineari-lanceolata in circuitu, in
petiolum longum angustata, apice pungentia,
subpinnatifida; lobis remotis et lanceolato-
triangularibus, erecto-patentibus, pungentibus,
marginibus recurvis, sinubus latis et ad costam
mediam parallelis; supra glabra laeviaque,
nervis prominentibus, subtus albo-tomentosis,
costa media prominenti.

Capitula sessilia, terminalia. solitaria, saepe
ramulis lateralis abbreviata terminantia, foliis
floralibus obvallatis; involucra permagna
usque 3 cm. expansa; bracteae numero-
sissimae, exteriores lineari-filiformes, interiores
setaceae, plumosae, ciliatae, exterioribus
longiores, omnes apicibus plumosae.

Perianthium supra basin glabrum, turgidum
et dense sericeo-tomentosum, segmentis fere
medium cohaerentia, sursum breviter tomen-
tosa, deinde glabra, laminae prof unde sulcatae,
glabrae, angustae; stylus perianthio vix
excedens, rectus, glaber; cylindrico-stigmatico
vix latiore terminatus.

Hab. in distr. Avon Pingelly orientalem versus, in
arenosis lateritls. fl. m. Novem. Gardner 14255 (TYPUS).

The affinity is with D. erythrocephala C. A.
Gardn., from which it differs in the longer
perianth 5.5 cm long, the much shorter broadly
lobed leaves with recurved (not revolute) mar-
gins, and the elegantly ciliate bracts and glab-
rous style, etc. The flowers are yellow.

Dryandra subpinnatifida C. A. Gardn. sp. nov.
§Eudryandra Meissn.

Fruticosa 1.5 m. alta, crebre virgato-ramosa,
ramis erectis, glabra, insuper pluriramosa,
ramulis brevibus, patentibus, conferte foliosis;

folia linearia, usque 23 cm. longa, infra media
spinoso-pinnatifida, rhachi vix 1 mm. lata
lobis patentibus, anguste-triangularibus vel
linearibus, spinescentibus, sinubus rectangu-
lariter dispositis, sursum linearia, integra
(inusitate sparse lobata) lobis acutis, 4 mm.
lata, supra glabra et laevia, costa media im-
pressa, subtus niveo-albo tomentosa, marginibus
recurvis, costa media prominentibus.

Capitula terminantia et praecipue copios
ramulos laterales florentibus foliis numerosis
obvallatis capitula manifeste excedentibus;
involucrum turbinatum cum bracteis numero-
sissimis quorum exteriores foliolatis, interiores
lanceolato-linearibus, acuminatis, omnes apici-
bus patulae et glabrae.

Perianthium circlter 3 cm. longum, infer ius
tomentosum ipso basin costatmn glabrum.
sursum sericeo-villosum. limbus obtusus, plu-
mosus; stylus glaber, perianthium excedens.
rigidus, cylindrico-stigmatico anguste conico.

Crescit in distr. Stirling, Popanyinning occidentalem
versus in collibus glareosis, £1. m. Septem. F. Lullfitz
(Typus).

The affinity of this species is with D. squarrosa
R. Br. from which is differs in the much larger
perianth and entirely different foliage which
has no similarity to any other species of the
genus. The flower-heads of D. squarrosa are
described as being axillary; those of D. sub-
pinnatifida are terminal.

Dryandra foliosissima C. A. Gardn. sp. nov.

SEudryandra Meissn.

Frutex erectus. usque 2 m. altus, dense ramo-
sissimus, ramis ramulisque brevibus, patentibus.
Folia densa, patentia vel recurva, usque 23 cm.
longa, linearia, marginibus recurvis, pinnati-
fida, acuta, lobis brevibus, remotis, deltoideis.
patentibus, mucronulatis, breve-decurrentibus.
2-3 mm. longa, vix 3 mm. lata, sinubus latis.
costa media parallelis, supra viridia et glabra,
costa media prominula immersa. subtus costato-
reticulata. sordide albo-tomentosis.

Capitula ramos et ramulos laterales copiosos
terminantes expansa ad 4.5 cm. diam., foliis
numerosissimis capitula longe superantibus;
involucri turbinati, bracteae exteriores linear i -
subulatae, appressae, interiores latiores et rufo-
tomentosae. omnes acuminatae et erectae.

Perianthium circ. 2.5 cm. longum, ad basin
costatum sericeo-hirsutum. supra dense
tomentoso-lanosum, sursum (unguibus) sparse-
pilosum. limbo angustissimo extus apicem
versus sericco-barbato. Stylus glaber, perian-
thium paulum excedens. cylindrico-stigmatico
non latiore terminatus. Folliculis obovoideis,
dense rufo-tomentosis vix compressis, 1.4 cm.
longis 7-8 mm. latis.

Hab. in distr. Eyre in saxosis in fruticetis, sumno
montem Desmond prope Ravcnsthorpe, in fruticetis
densis, fl. m. Junto et Julio, Gardner 12886 (TYPUS)

This species has a close affinity with D.
mucronulata, R. Br.. with which is has been
confused. It agrees in many respects with the
description given by Bentham in the Flora
Australiensis, but there are several small, but
important differences.

D. mucronulata has flat leaves which are
divided almost to the midrib, the lobes being
triangular and acute or finely mucronate, and
the sinuses are acute. The inner bracts are

58

obtuse and mucronate. In D. foliosissima , which
apparently has the same habit as D. mucronu-
lata, the sinuses are very broad and parallel to
the midrib, the leaves very much longer, the
margins strongly recurved, and the teeth much
smaller. The bracts are all acute or acuminate.
It appears to be restricted to the vinicity of
Mount Desmond, whereas D. mucronulata would
appear to be restricted to the vicinity of King
George’s Sound and the Stirling Range. D.
foliosissima appears to be still more closely re-
lated to D. squarrosa R, Br., of which I have
not seen specimens. The fragment in the Mel-
bourne Herbarium doubtfully referred to as D.
squarrosa by Bentham is D. polycephala. From
the description given of D. squarrosa, the in-
volucres would appear to be very different from
those of D. foliosissima.

Dryandra subulata, C. A. Gardn. sp nov.

§Acrodontae D.C.

Fruticulus humilis 30-35 cm. altus, ramis
stricte ascendentibus, juvenilibus tomentosis
vetulis glabratis; folia erecta, anguste-linearia,
plerumque 25-30 cm. longa, 1.5-2 mm. lata,
integra, margine crasso arete revoluta, in
petiolum breviter attenuata (ipso basin sub-
dilatatum) longiuscule mucronato-spinescentia
juvenilia sparse sericeo-puberula. adulta
demum glabra. subtus bisulcata, sericeo-
tomentosa.

Capitula subsessilis, ramos et praecipue
ramulos laterales copioso terminantia. expansa
usque 2.5 cm., foliis mumerosis capitula multo
superantibus obvallata; involucri late turbinati;
bracteae numerosissimae, exteriores lineari
subulatae, usque 5-6 cm. longae, basi dilatatae,
viscideae, interiores breviorae, omnes rectis et
apicibus subulatae.

Perianthium ca. 2 cm. longum. basin brevem
glabrum, sursum tomentoso-villosum, limbum
glabrum, tubum sursum attenuatum et glabrum;
stylus rectus glaber, perianthium non excedens;
cylindrico-stigmatico atro-fusco, leviter sulcato.

Hab. in distr. Irwin prope flumen Hill, in fruticetis
apertis in arenoso-glareosis, fl. m. Septem. Gardner
12175 (TYPUS) etiam locis eisdem, A. J. Gray.

This species differs from the others of this
Section in habit, in the exceptionally long en-
tire linear leaves none of which closely surround
the flower-heads, the heads being surrounded
by external subulate bracts which are almost
three times longer than the involucre proper.

Dryandra arborea C. A. Gardn. sp. nov.
s iEudryandra Meissn.

Arbor usque 5 m. alta, ramis patentibus,
trunco usque 30 cm. diametri, cortice nigra;
ramulis brevibus ad apicem solum foliolatis;
folia petiolata, oblanceolata, plana, subpinna -
tifida. basi cuneata, lobis patentibus, triangu-
laribus. decurrentibus, pungentibus, subtus
reticulatis, glaucis, subalveolatis, sinubus
acutis; capitulis terminalibus, involucro floribus
multo breviore, squamis < bracteae) glabriusculis
ciliatis, lanceolatis, acutis, demum glabrescent-
ibus.

Perianthium stylum subaequantum. unguibus
dense tomentosum; stylus breviter exsertus, basi
lanuginosus; stigmatico cylindrico obtuso.

Folliculi oblique ovati, biconvexi, glabriusculi
demum glabri, reticulato-striati, margine acuto
compressi.

Hab. in distr. Coolgardie in rupestribus “Iron Knob”,
collibus Koolyanobbing, fl. m, Oct. Gardner 12217 Oct.
1959 (TYPUS).

This species was previously collected by Young
between Ularring and Mount Churchman, and also
by John Forrest, and w T as named D. armata by
Mueller, although it has very little in common
with the latter, apart from the woolly base of
the style. The involucral bracts are quite dis-
tinct from those of any other related species,
and the floral leaves are not obvallate. It is
also the only arborescent species of the genus
known to me. D. armata R. Br. fsensu strictu)
is confined to the south coastal areas, and mav
be easi y distinguished by its hirsute branchlets,
long hairy and long-ciliate bracts, and in addi-
tion these latter always have nigrescent-tomen-
tose apices, a characteristic confined to this and
the closely related D. Purdieana Diels and D.
Gilbertii S. Moore.

P APILJON ACE AE

Oxylobium rigidum C. A. Gardn. sp. nov.

§Podolobium

Frutex humilis, usque 30 cm. altus, glaber;
ramis erectis: ramulis angulosis. Folia opposita,
breviter petiolata. oblongo-lanceolata. plana,
rigida, costa media prominens, basi leviter sed
conspicue cordata, apice acuta, subpungentia.
glauca; stipulae setaceae, nigrae, petiolis
longiores. Flores pauci ad ramulorum apices
in racemes pedunculati, verticillato-dispositae.
pedicelli circ. 15 mm. longi; bracteolae deciduae.
Calyx coriaceus, circ. 6 mm. longus, campanu-
lato-cupuliformis, glaberrimus, glaucis, lobi 3
inferiores ovato-lanceolati, circ. 2.5 mm. longi,
tubo breviores, lobis superioribus approximate
altius connati, truncati, marginibus ciliolati;
vexillum calyce haud duplo longior cordato-
emarginatum, alae vexillo paulo breviores,
obtusissimae; Carina aliis multo longior;
ovarium longe stipitatum; villosum 4-ovulaturn.

Hab. in distr. Eyre montem Madden septentrionalem
versus, in glareosis fruticetis, fl. m. Oct. Gardner 13635
I TYPUS ) .

The affinity of this species is with O. race -
mosum (Turcz.) C. A. Gardn., from which it
differs in the usually acute leaves, the shorter
pedicels, upper ciliolate calyx-lobes silky ovary,
and number of ovules, and the much more rigid
glaucous foliage with scarcely evident venation
Oxylcbium racemosum is a tall shrub of 1-2
metres in height, with longer racemes.

Gastrolobium appressum C. A. Gardn. sp. nov.

Series Racemosae Benth.

Fruticulus 20-30 cm. altus, ramis subflexuosis
glabris, ramulis brevibus, dense foliolatis
appresse-pubescentibus, demum glabratis. Folia
ternata, brevissime petiolata vel subsessilia.
rigida, erecta, arete appressa, coriacea, ovato-
lanceolata. acuta, leviter concava, basi leviter
sed conspicue rotundata, 6-7 mm. longa, mani-
feste reticulata, glabra, supra reticulata, subtus
pallida; stipulae nullae. Flores ad ramulorum
apices in racemulos pauciflorae, positae, pedicelli
graciles, subpatentim sericeo-pubescenti, pler-
umque 2 mm. longi; calyx glaberrimus. 6 mm
longus, lobi 3 inferiores lanceolati acuminati

tubuin adaequantes; corolla calyce fere duplo
longior, praeter carina purpurescenti, auran-
tiaco-lutea, alae carinaque vexillo breviores,
aequilongae; ovarium villosum distincte stipi-
tatum; legumen ignotum.

Hab. in distr. Irwin prope Gunyidi. in arenosis glare-
osis in fruticetis. fl. m. Septem Gardner 12745 (TYPUS);
prope Watheroo in locos slmilibus; in distr. Avon, Dal-
walllnu 25 km. occidentalem versus in arenosis saxosis.

The affinity is with G. hamulosum Meissn., but
the branches are more divaricate, and the
leaves appressed to the branchlets. The ap-
parent complete absence of stipules is another
feature distinguishing this species.

MYRTACEAE

Eucalyptus cupularis C. A. Gardn. sp. nov.

Arbor usque 10 m. alta, cortice alba, basin
versus rimosa, sursum leviter obtecta; ramulis
leviter angulatis. Folia alterna, petiolata,
rigida, crassiuscula, concoloria, opaca, in
petiolum 2-2.5 cm. longum attenuata; lamina
anguste lanceolata vel lanceolata, viridia, apici
acuta vel acuminata, falcata, 16-18 cm. longa,
costa media supra elevata et canaliculata, nervis
lateralibus patentibus. anastomosis, nervus
intramarginalibus ab margine distinctis.

Pedunculi axillari, erecti, validi, subtereti,
sursum incrassati, 8-10 mm. longi, umbellam

5- 7 floram gerentes; flores sessiles; alabastri
juvenilibus bracteati.

Hypanthium cylindro-cupulatmn, 5-6 mm.
longum, laeve, basi contractum, 2 -costa turn vel
-anguiatum; 6 mm. longum; operculum late
ovoideo-hemisphaericum, umbonatum, hypan-
thio brevior (3.5 mm. longum). Stamina adulta
non visa, filamenta in alabastris flexuosis et
incurva.

Fructus cylindrico-cupulatus, lignosus, dia-
metro transversali 7-8 mm., sed tantum circa
8 mm. longus, margine crasso, discus elevatus,
annulatus; capsula 4-5-locularis, valvis del-
toideis, conspicue et rigide exsertis.

Hab. in distr. Ord. prope oppidum Hall’s Creek, in
colliculis lapidosis schistosis occidentalem versus Gard-
ner 10217 (TYPUS) et in locis slmilibus R. A. Perry
3180; in montibus lapidosis in Lat. 18 45' S. Boundary
Survey, S. J. Stokes sine n.

In the absence of flowers it is difficult to as-
sign the systematic position of this attractive
tree, and its affinities must remain obscure. It
certainly has not been previously described.

Euclayptus jucunda C. A. Gardn. sp. nov.

Frutex elatus vel arbor parva, usque 7 m.
alta, ramulis juvenilibus angulosis, brunneo-
eorticatis; folia alterna, distantia, viridi-
glaucescentia, oblongo vel anguste-lanceolata,
falcata, in petiolum conspicuum circ. 1.2 cm.
longum attenuatum, apice acuminata vel acuta,
sine petiolo 6- usque 9 cm. longa, costa media
prominula, nervis lateralibus supra vix con-
spicuis, subtus elevatis et conspicue obliqo
divergentibus, laxe anastomosis, vena intra-
marginalia ab margine remota.

Pedunculi erecti, subtereti vel tantum obscure
vel leviter angulosi. nunquam complanati,
axillares sed plerique e ramulorum parti
inferiorum foliorum jam destitutam enascentes,
1-1.2 cm. longi. Flores plerumque sat numerosos

6- 12. manifeste pedicellatos gerentes; pedicelli
subteretes, circ. 6 mm. longi, hypanthio vix
breviores.

Hypanthium anguste-campanulatum, sub flore
expanso circ. 6 mm. longum, basi attenuatum;
stamina flavida, filamentis in alabastris flexuosa.
inflexa, antherae oblongae, loculis parallelis;
operculum late conicum vel ovoideum, acutum
vel obtusum, 4 mm. longum.

Fructus ovoideo-globosus (diametro trans-
versali 1.3 cm. metiens) sed tantum circ. 1.5-1. 7
cm. longus, fere laevis ad marginem orifici 3.5
mm. diametro; capsula trilocularis, valvis pro-
funde inclusis et apparenter tribus.

Hab. in distr. Irwin septentrionalem valde divulgatus
in arenosis crescens; prope flumen Murchison in
arenosis lutosis in fruticetis, Gardner; prope flumen
Greenough juxta pontem mullewensem in fruticetis.
fi. m. Febr. R. D. Royce 1122; prope Tenindewa, G. H.
Burvill; baud procul a Yuna. G. E. Brockway ; inter
Mullewa et flumen Greenough. J. Reeves 70; in arenosis
fruticetis prope flumen Greenough. fl. 6. Januario 1959.
Gardner 12066 (TYPUS).

Belonging to the Macr antherae, this species
does not appear to have any close affinities.

Blakely has named a variety globosa of
Eucalyptus eudesmioides F. Muell. which I have
not seen, but the species named here has no
affinities with the Eudesmieae.

Regel ia megacephala C. A. Gardn. sp. nov.

Fruticosum glabrum usque 3.5 m. altum, rami
erecti, ramuli abbreviate dense foliati. Folia
opposita, quadrifariam decussata, erecta, vel
leviter patentia. obovato-orbicularia, plana vel
concava. sessilis, basi subcordata, glauco-
viridia, subtus prominenter 5-nervia, breviter
acuta vel mucronata.

Capitula globosa, ad ramorum valde abbre-
viatum terminantia, magna, rhachis albo
sericeo-villosa; flores purpurae, pulchrae.
Calycis tubus turbinatus, 2.5-3 mm. longus,
albo-pubescens; sepala lanceolata, acuta, tri-
nervia, calycis tubo longiores, villosa; petala
alba subscariosa, fere oblonga, truncata erecta
glabrique, marginibus hyalina, ciliata, circa 3.5
mm. longa; phalangium circ. 2-andrum, circ.
7 mm. longum, glabrum, antherae luteae.

Capitulae fructiferae ovoideo-globosae vel
ovoideae, usque 3 cm. longae et 2.3 cm. latae.
Calyces fructiferae plusminusve concretae,
truncatae, late hemisphaerico-urceolatae, 1-1.2
cm. latae; calyx fructifer immersus late
obovoideus, capsulam valde superans, valvis
obtusis.

Hab. in distr. Avon prope oppidum Coomberdale, iu
collis saxosis quartzosis, fl. m. Decem. Gardner 12167
(TYPUS).

Affinity with R. ciliata Schau., differing prin-
cipally in the glabrous branches, the erect and
appressed obovate-orbicular larger leaves, the
relatively longer calyx-lobes, and the much
larger fruiting heads. The shrub is also much
larger and erectly branched.

R. velutina (Turcz.) C. A. Gardn. comb. nov.

Beaufortia velutina, Turcz. in Bull. Phys
Math. Acad. Petersb, x. 346 (1852).

R. cymbifolia (Diels) C. A. Gardn. comb, nov,

Beaufortia ? cymbifolia Diels in Engler’s Bot.
Jahrb. xxxv. 431 (1904).

Diels did not see normal 9 flowers of this
plant. It proves to be a Regelia . and I have col-
lected a topotype from the original locality: —
prope Tambellup in fruticetis arenosis, fl. m.
Oct. Gardner n. 13885.

60

Eremaea purpurea C. A. Gardn. sp. nov.

Frutex rigidus circiter 30 cm. altus, ramis
crassis, erectis densissime foliolatis, ramulis
ultimis exceptis glaberrimus. Folia patentia,
oblonga, crassa, supra concava, 2-2.5 mm. longa,
obtusa, glaberrima, viridia, tantum leviter
gladuloso-punctata. Flores purpurae, pulchrae,
solitarii ad ramulorum apices sessiles, bracteis
parvis, ovatis, caducis. Calycis tubus campan-
ulato-turbinatus, argenteo-villosus, vix 3.5 mm.
longus, lobi deltoidei subacuti, trinerves, tubo
breviores glabri; petala purpurea, obovato-
orbicularia, concava. obtusissima, erecta, glabra,
circ. 4 mm. longa, marginibus subscariosa, basin
versus attenuata; phalagium 15-andrum, fila-
menta purpurea, antherae flavae.

Calyx fructifer late hemisphaerico-cupulatus,
capsula non superans, circ. 6 mm. longus, 9 mm.
latus; semina erecta, inaequaliter tetragono-
clavata, angulis membrana hyalina frimbriata
ala t is.

Hab. in distr. Darling prope Culalla, in arenosis
humidis, fl. m. Decem. Gardner (TYPUS); prope Muchea
in arenosis depressis, H. Steedman.

The affinity of this species is with E.
ebracteata F. Muell., from which it differs in
the much smaller and broader glabrous leaves,
the small purple flowers and the relatively
smaller calyx-lobes. The style is short and
glabrous.

Melaleuca conothamnoides C. A. Gardn. sp. nov.
§Capitatae Benth.

Frutex multiramosus circ. 1 m, altus, ramis
erecto-patentibus, ramulis erectis, omnino gla-
berrimus; folia alterna vel sparsa, breviter
petiolata, arcuato-patentia, oblongo-lanceolata
vel oblonga, obtusa vel apicibus rigide mucro-
nata, basi contracta vel breviter attenuata,
glauca, rigida, plana vel leviter concava, 3-5-
nervia. Capitula globosa vel ovoidea, ad
ramorum partem superiorem terminantia,
breviter pedunculata, rhachis breviter sericeo-
tomentosa; flores mediocres, purpurae, calycis
tubus late ovoideus, puberulus; calycis lobis
breviter semiorbicularis, marginibus tenuiter
hyalinis; petala ovata, purpurea, subscariosa,
concava, obtusissima, patentia, glabra, margini-
bus integra; phalangium glabrum, purpureum
circ. 13-andrum, antherae brevae, luteae.
Ovarium triloculatum. Capitulae fructiferae
globosae vel brevi-cylindricae vel ovoideae, 7-12
mm. longae et 9 mm. latae. Calyces fructiferae
ovoideae, lobis deciduis; capsulam valde
superans, cii\ 4 mm. longae et 3 mm. latae.

dlst r. Irwin inter Caron et Latham, Blackall
Gardner 2690; inter Morawa et Koorda. Blackall
3469 distr Avon: Wongan Hills, R. D. Royce 2192; Bal-
d, 7 Gardner; Prope Wyalkatchem, J. W. Green 835;
tuackall; prope Tammin 5 km. occidentalem versus in

1990°fTYPUs!° SlS in planietis ’ Ga rdner 1122, Novem.

Austin distr. a form from Boolardy, collected
by A. B. Cashmore has larger flower-heads and
somewhat longer leaves.

The affinity of this species is with M. cordata
Turcz., differing mainly in the shape of the
leaves which are never cordate at the base, the
almost petaloid calyx -lobes, and the more
numerous filaments. Both species are often
found in the same area under the same ecological
conditions.

61

Leptospermum sericeum Labill.

Leptospermum sericeum Labill. Nov. Holl. PI.
Sp. XI p. 9, pi. 147 (1806).

In September 1962, I collected at Cape Le
Grand specimens of a Leptospermum in flower
(Gardner 14117) which have been matched with
the type in the Florence Herbarium.

Since 1867, when Bentham confused this
species with Kunzea sericeci, the species has been
deleted from the Flora of Western Australia.
Brown and Baxter both collected Leptospermum
sericeum in fruit, and, since the specimens were
not in flower, they were placed by Bentham
under Kunzea sericea.

The recent collection of this attractive species
proves that L. sericeum is the common plant
of the islands of the Recherche Archipelago,
since Kunzea sericea does not occur within 160
km. of the coast. Labillardiere’s designation
“Capite van Diemen” for the plant is due to an
error.

Wehlia aurea C. A. Gardn. sp. nov.

Frutex densus multiramosus, usque 40 cm.
altus, omnino glaber; ramis erectis, virgatis.
ramulis ultimis brevibus, denissime foliatis;
folia confertissima, sparse, arete imbricata,
erecto-appressa, subsessilia, crassiuscula, ellip-
tico-obovata, laevia, supra leviter concava,
subtus convexa, crebre nigro-punctata, valde
obtusa, 2 mm. longa.

Flores in ramorum apicibus denissime spicatae,
brevissime pedicellatae, aureo-luteae, circiter 5-6
mm. diam.; bracteae scariosae, calycis tubus
involventes, caducae, orbiculares, medio uni-
costatae, marginibus hyalinae. Calycis tubus
hemisphaericus, vix 2 mm. longus, leviter com-
pressus, lobi rotundati, tubo subaequalibus,
glabri, marginibus scariosis, integri; petala
orbicularia, aureo-lutea, marginibus integra.
Stamina circ. 35-40; ovarium globosum; ovula
2, ad apicem placenta a basin collateraliter
affixa; stylus elongatus.

Hab. in distr. Austin prope Pindar, in arenosis lutosis,
fl. m. Oct. Gardner 7778 (TYPUS), etiam inter Peren-
jon et Dalwallinu Blackall, ad lacum Monger extendit;
prope Canna in fruticetis arenosis, Gardner 2688.

The species is distinct in its minute appressed
broadly obovate smooth leaves, in the whole
plant quite glabrous, and in the leafy spikes of
yellow flowers.

Wehlia grandiflora C. A. Gardn, sp. nov.

Frutex 1.2 m. altus, ramis patentibus, omnino
glabra. Folia petiolata. patentia, 5-6 mm.
longa, cunea to -oblonga, subtriquetra, valde
carinata, supra convexa, in apicem mucronatum
recurva terminantia; flores purpurae, axillari-
bus solitarii vel bini, brevipedunculata; bracteae
parvae; bracteolae scariosae, cordatae, arete
imbricatae; longitudinaliter striatae, purpureae,
marginibus albis scariosis. Calycis tubus cam-
panulatus, vel hemisphaericus, minute foveolato-
punctatus; sepalis suborbicularibus, roseis,
trinervis, lobis, hyalinis, petalis subdiniidio.
Petala roseo-purpurea, late obovata (apicibus
subacuta), purpureo-striata ; stamina numerosa
(circi. 35-40) petala excedentia; filamentis pur-
pureis ; antherae oblongo-ellipticae, recurvae,
rimis longitudinalier dehiscentibus, glandula
cornuta, dorsalia, prominens, incurva; stylus

flexuosus, compressus, staminibus longioribus,
in depressis ovarii insertus; ovulis 2, collat-
eraliter affixa

Hab. in distr. Austin prope Menzies 25 km. meri-
dionalem versus, in arenosis dunosis, fl. m. Novem-
Decem. Gardner 13879 (TYPUS).

Distinguished from all other species of the
genus in the larger widely spreading longer
leaves, the larger flowers and distinctly cordate
bracteoles and long exserted style. Its closest
affinity is W. staminosa F. Muell., which occurs
on stony hills between Mount Hale, Meekatharra
and Lake Carnegie.

Darwinia Masonii C. A. Gardn. sp. nov.
iGenetyilis (D.C.) Benth.

Frutex erectus, 1.5-2 m. altus vel ultra, ramis
erecto-patentibus, cortice albo-griseo. Folia
dense conferta, basin versus attenuata et brevi-
ter petiolata. subtriquetra, pallide viridia, saepe
1 cm. longa. apice breviter acuta. Capitulis
subsessilibus, ramulis terminantibus; involucrum
hemisphaericum; bracteae numerossisimae.
pluriseriatae, exteriores oblongo-lanceolatae
interiores ovato-contractae, rubrae, usque 15
mm. longae; bracteolis oblongo-lanceolatis,
valde concavis vel cymbiformis, submembrana-
ceis, mucronulatis. 5 mm. longis. Calycis tubus
circ. 4 mm. longus, durus, manifeste 10-
costatus, inferiori turbinatus, lobis minutis;
petala ovato-lanceolata, alba, circ. 3 mm. longa,
obtusa, glabra; stamina 10. brevia, glabra;
staminodia filamentis similia sed angustiora,
aequilonga; stylus 10-11 mm. longus, sursum
sparse et breviter pilosus, pilis patentibus.

Hab. in distr. Austin prope montem Gibson, in
fruticetis in glareosis. fl. m. Aug. D. Mason (TYPUS).

This species is closely related to D. leiostyla
(Turcz.) Domin, notwithstanding the totally
different climatic and edaphic requirements. The
involucral bracts are much like those of D.
leiostyla, and in both species the calyx lobes
are minute. The leaves however, are narrower
and subtriquetrous, and the flower-heads
smaller.

VERBENACEAE

Lachnostachys bracteosa C. A. Gardn. sp. nov.

Frutex nanus, circ. 20-30 cm. altus, ramis
numerosis. erecto-patentibus. ramulis erectis.
Folia opposita. decussata, linearia, sessilia et
valde aecurrentia, obtusa. supra glauca et ±
transverse bullata. olivacea et sparse tomentosa,
demurn glabrescentia, usque 1 cm. longa, 1.5
mm. lata, marginibus arete revoluta, subtus
i costa media) albo-tomentosa.

Spicae terminates, cylindricae, dense albo-
tomentosae, sessiles vel subsessiles; bracteae
rigidae, lanceolatae, cymbiformae, petiolatae,
solitariae vel tripliciter cum uno petiolo,
marginibus longe tomentosae, dorso dense et
longe brunneo-tomentosae, acutae, petiolae per-
sistentibus. Calyx obovoideus, breviter pedicil-
latus, extra dense lanato-plumosus, sordide
albus, lobis lanceolatis, acutis, tubus et lobis
intus glabris. Corollae tubus turbinatus, calyce
subaequilongus, tubo truncato pilis longibus
intus prope basin ortis, exceptis glabro. Stamina
6-7 (rarius ad 2 reducta > filaments elongata ad
margine tubum affixa, dilute-violacea, basin

versus dilatata; antheris dithecis, thecis rimis
longitudinalibus dehiscentibus. Ovarium ovoid-
eum vel globosum, dimidio superiore sericeo-
pilosum; stigmato capitato.

Hab. in distr. Eyre prope lacum King occidentalem
versus in fruticetis in arenoso-lutosis, fl. m. Oct. Gard-
ner 13636 (TYrUS).

What appears to be a peculiarity in this
species is to be seen in the cymbiform bracts
which have a slender but rigid petiole. These
are either single, or three -branched, each bract
having a flower in its axil. These persistent
petioles, simple or trifid, persist after the flowers
have fallen, giving to the inflorescence a distinc-
tive appearance not seen in any other species
of Lachnostachys. The presence of the petioles
of these bracts indicate that the inflorescence
is not a spike, but a spike-like panicle. Whether
or not this condition is found in other species
(except L. coolgardiensis S. Moore, its closest
affinity) remains to be seen. This, and the
much-branched stems which are comparatively
slender and more woody than in other species
are another distinctive feature of the plant.

Newcastlia chrysophylla C. A. Gardn. sp. nov.

Frutex circ. 4 m. altus, ramis erectis, dense
tomentosis, ramulis erectis, aureo-tomentosis.
Folia alterna, spars a vel rare opposita, obovata
usque late-obovata, petiolata, basin attenuata,
cum petiola 14 mm. -3.7 cm. longa, 5 mm. ad
1.0 cm. lata, crassa, plana vel marginibus
leviter recurva, supra medium crasse crenulato-
dentata, obtusa, supra olivacea et scabrida,
subtus dense aureo lanato-plumoso-tomentosa.

Spicae terminales 4-5 cm. longae paniculae
dispositae, breviter pedunculatae, bracteae ig-
notae. Calyx globosus. 3-4 mm. longus, extus
dense aureo-plumosus, lobis deltoideus, acutius-
culis, intus glabris. Corollae tubus obovoideus,
calyci subaequilongis, lobis late-oblongis, obtusis,
albis, glabris; filamenta breviora, infra medium
tubum adnata, superne libera, quam antherae
multo longiora. Ovarium globosum, supra
medium dense pilosum, 2-loculatum, loculis ?
1-cvulatis, ovulo unico tantum maturascenti;
stylus brevissimus (2 mm. longus), sparse
pilosus, breviter bilobus,

Crescit in distr. Austin flumen Murchison 40km.
septentrionalem versus, in arenosis lutosis in fruticetis.
fl. m. Decem. Gardner 12054 (TYPUS).

Only known from this locality, and apparently
rare, this remarkable shrub is particularly at-
tractive by reason of the dense golden tomentum
of the undersurface of the leaves which con-
trasts with the dark olivaceous upper surface.
The stern is often 7 cm. in diameter, and it is
by far the largest species of this genus. The
genus is found almost entirely on the desert
regions of the continent, and its occurrence al-
most on the boundary of the South-West Pro-
vince is remarkable. Its affinities are with N.
Dixonii F. Muell. et Tate and N. clirysotricha
F. Muell. differing from both in the leaves which
are petiolate, and of different size and shape,
having the golden calyces of N. chrysotricha
which has linear leaves with revolute margins,
and from N . Dixonii in the shape of the leaves
and the spicate inflorescence.

The species is worthy of cultivation.

62

Pityrodia spectabilis C. A. Gardn. sp. nov.

§Eupityrodia E. Pritzel

Suffrutex diffusus vel erectus, usque 15 cm.
altus, sparse ramosus omnino dense lanatus,
ramulis crassus, internodiis brevibus. Folia
opposita, oblonga vel oblongo-obovata, sessiles,
sed non decurrentes, plana, crassa sed mollis,
marginibus integra, novellis dense lanata,
vetulis sparsim floccosa, usque 5.3 cm. longa,
obtusa, basi attenuata.

Flores numerosae in axillis superioribus dis-
positae, pedicellatae, magnae, pedicelli 3-4 mm.
longi. Calyx fere ad basin divisus, 1.5 cm.
longus, lobis lanceolato-obovatis, obtusis, dorso
plumosis. Corolla coccinea vel aurantiaca, 3
cm. longa, ad basin tubi angusta, subito cam-
panulato-dilatata, extus glabra, intus supra
basin dense tomentoso-sericea, supra medium
sparse et breviter pilosa; lobis brevibus, patentis,
subaequalibus, semiorbicularis, obtusis, margini-
bus denticulatis. Stamina exserta, filamentis
infra medium tubi insertis, antherae brevissime
appendiculatae. Ovarium dense sericeo-
pilosum ; stylus glabrus, corolla excedens.

Hab. in distr. Irwin prope Buntine in arenosis lutosis,
fl. m. Decern. Gardner 12023 (TYPUS); prope Canna

D. H. Perry, fl. m. Oct.

The species belongs to the Section Eupityrodia

E. Pritzel with affinity to P. axillaris (Endl.)
Druce differing in the much larger calyx and
corolla — almost twice as long, the calyx -lobes
being also much broader and thicker, the al-
most equal corolla-lobes which are denticulate
and not entire, the much exserted stamens, and
the dark or violet prominent striations of the
corolla. It is by far the largest flowered mem-
ber of the genus.

LABIATAE

Prostanthera scutata C. A. Gardn. sp. nov.

§ Euprostan thera

Frutex nanus usque 20 cm. altus, ramosissimus,
ramis erecto-patentibus, vel diffusis, crebro
foliatus. Folia opposita, in petiolis abbreviate
attenuata, apice cbtusa, basin versus attenuatus,
mollis, plana, integra undique cano-tomentosa,
uninervia, 8-15 mm. longa.

Flores in axillis superioribus dispositae, soli-
tariae, breviter pedicellatae; bracteolae 2,
lineari-spathulatae, 2.5 mm. longae. Calyx 1.2
cm. longus, venis reticulatis, labio supero aucto,
late ovatus, breviter 3-lobatus; labio infero
integro, superiore duplo breviore. Corolla
pallide coerulea, extus undique breviter pilosa,
intus glabra, circ. 15 mm. longa. tubo cylindrico
supra ovarium leviter constricta, deinde
ampliato, lobis suborbicularibus, obtusis. 2
superioribus obovatis alte connatis, inferioribus
obovatis, lebo mediano eniarginato, lateralibus
ovato-oblongis, retusis, stamina inclusa,
antheris omnibus appendiculatis. Calyx fruc-
tifei amplus, obtusus; labio supero magno 1.5-
1.7 cm. longo, 1.5 cm. lato, rotundo, eleganter
nervoso, breviter et obtuse 3-lobat.o; labio infero
ovato, superius duplo breviore, obtuso.

Hab. in distr. Irwin prope flumen Hutt, in arenosis
lutosis, fl. rn. Decembri, Gardner 14266, (TYPUS) etiam
prope oppidum Wilroy, in fruticetis in arenosis fl. m.
Novem Gardner s.n.

GOODENIACEAE

Leschenaultia pulvinarls C. A. Gardn. sp. nov.
§Euleschenaiiltia Benth.

Suffrutex humilis, dense ramosus ramis
ramulisque prostratis, dense pulvinaris, usque
30 cm. diam.

Folia dense-conferta, rigida, lineari-subulata,
apice longiuscule mucronata, dense and breviter
hirsuta, 6-7 mm. longa et 0.4 mm, lata.

Flores in axillis superioribus solitarii; ovarium
lineare strictum vel leviter curvatum, 5.5 mm.
longum, glabrum, calycis-lobi subulati acuminati,
3.5-4 mm. longi, scabride-hispidissimi. Corolla
coerulea, tubus circ. 6 mm. longus, extus glab-
rus, intus tota extensione sparse et breviter
pilosus, lobi tubo subaequilongi. 2.3-3 mm. longi
acuminati, omnes subaequaliter latiuscule alati.
in flore patentes, alis latis, rotundis, obtusis.
obsolete venosis; stamina 3.2 mm. longa, fila-
menta tenuia. Capsula ignota.

Hab. in distr. Stirling, a Corrigin circ 40km. meri-
dionalem versus, in arenosis, fl. m. Oct. Gardner 13620
( TYPUS ) .

This species belongs to the Section Euleschen-
aultia, with affinity to L. expansa R, Br., from
which is differs in the dense pulvinate habit,
like that of L. iubiflora R. Br., and in the subu-
late hispid foliage and calyx lobes of which there
is no counterpart within the genus.

It is known only from the type locality, and
it may be in danger of extinction, the total
area being less than half a mile in extent, and
most of it farming land.

Goodenia sericostachya C. A. Gardn. sp. nov.
§Monochila G. Don.

Herba biennis vel forsan perennis usque 35
cm. alta, dense albido-sericeo-tomentosa, cauli-
bus teretibus, simplicibus vel plerumque erecto-
ramosis, validis. Folia inferiora subrosulata
numerosa, obovato-oblanceolata, basin versus
sensim angustata, ima basis paulum dilatata,
integra, usque 5 cm. longa et 12 mm. lata,
obtusa, plana crassa sed molliter tomentosa,
folia caulina minora, circ. 2 cm. lenga, obtusa,
basi sessilia, integra.

Scapi elongati, robusti, paullum ramosi, dense
villosi; flores sessiles dense conferti in spicam
densam, inferne saepe interruptum, superhe
densiorum efformantes. Calycis tubus cbovoid-
eus, denissime sericeo-pilosus, calyci lobi
oblongo-lanceolati, acuti, ovario subaequilongi.
Corolla intense violacei, extus praeter alas
glabra dense sericeo-villosa, lebi omnes subae-
qualiter alati: omnes subaequales digitatim

patentes, acuti; stamina libera; stylus elongatus,
breviter et sparse pilosus; indusium dorso
breviter pilosum, margin© breviter et dense
ciliatum. Capsula ellipsoidea, 6 mm. longa,
longe lanato-villosa, calycis lobis persistentibus
coronata; semina nigra, subrotundata, tenuiter
et anguste marginata.

Hab. in distr. Irwin circ. flumen Murchison 45 km.
septentrionalem versus in arenoso-lutosis, fl. m. Decern
Gardner 12430. 5. Jan. 1960 -TYPUS).

This species is remarkable in that it com-
bines the floral structure of Goodenia with the
habit and inflorescence of Verreauxia . It must
be included in the Section Monochila, with G.

63

scapigera R. Br. and G. Watsonii F. Muell. et
Tate, but differs from these in the vestiture of
the plant, and from all the species of Goodenia
in the spicate inflorescence. It is a remarkably
attractive plant. In the season of its discovery
it was fairly common in the restricted area of
its habitat, visits to the same area during the
last two or three years have failed to disclose

its presence, even when not in flower. It is prob-
able, that like some other plants, especially
Velleia discophora, it will not again appear until
the country is burned.

The holotypes of the species described in this
paper will be deposited in the Western Austra-
lian Herbarium.

I

64

INSTRUCTIONS TO AUTHORS

Papers may be submitted to the Society in accordance with Rules and
Regulations 38 to 41 inclusive (see below). They should be addressed to The
Honorary Secretary, Royal Society of Western Australia, Western Australian
Museum , Perth.

Authors are solely responsible for the factual accuracy and for any opinion
expressed in their papers. They are particularly requested to verify references.
Alterations to MSS. submitted to the printer will be allowed only under excep-
tional circumstances, and no changes will be permitted after galley-proof stage.

In the preparation of MSS. authors are required to follow the C.S.I.R.O.,
Guide to Authors (C.S.I.R.O., Melbourne, 1953), except that papers longer than
10,000 words (30 foolscap pages of pica type, with 6 inch lines 40 to the page) will
not normally be accepted.

Authors may be required to meet half the cost of preparation of the blocks
of diagrams and illustrations.

Authors shall receive a total of 30 reprints free of charge. Further reprints
may be ordered at cost, provided that such orders are submitted with the MS.

RULES AND REGULATIONS

38. Every paper intended to be read before the Society or to be published
in the Society's Journal must be sent to the Secretaries at least seven days
before the date of the next ensuing Council meeting, to be laid before the
Council. It will be the duty of the Council to decide whether such contribution
shall be accepted, and if so, whether it shall be read in full, in abstract, or
taken as read. All papers accepted for publication must be read or otherwise
communicated at an ordinary meeting prior to publication.

39. A Publications Committee, appointed by the Council, shall recommend
to the Council whether a paper presented to the Society shall be published in
the Society’s Journal. The Publications Committee may obtain an opinion from
any person it may select on the suitability of any paper for publication.

40. Publication in the Society’s Journal shall only be available to (a)
Ordinary Members, (b) Honorary Members, (c) Non-members resident outside
Western Australia, who must communicate the paper through an Ordinary or
Honorary Member. No paper shall be accepted from a Non-member resident in
Western Australia.

41. The original copy of every paper accepted for publication by the
Society, with its illustrations, shall become the property of the Society, unless
stipulation is made to the contrary, and authors shall not be at liberty to
publish their communicated papers elsewhere prior to their appearance in the
publications of the Society unless permission for so doing is given by the
Society, or unless the Society fails to publish the paper in the Journal of the year
in which it is read or otherwise communicated, or of the succeeding year.

Journal

of the

Royal Society of Western Australia

Volume 47 1964

Part 2
Contents

6. — The Woolgorong Stony Meteorite. By G. J. H. McCall and P. M. Jeffery.

7. — The Agamid Lizards of the genus Tympanocryptis in Western Australia.

By G. M. Storr.

8. — On Beryl from Western Australia. By P. L. C. Grubb.

9. — Contributiones Florae Australiae Occidentalis XIII. By C. A. Gardner.

Editor: R. W. George
Assistant Editor: R. D. Royce

Annual Subscription: Forty-five Shillings

The Royal Society of Western Australia, Western Australian Museum, Perth

82169/4/64—570

ALEX. B. DAVIES, Government Printer, Western Australia

JOURNAL OF

THE ROYAL SOCIETY

OF

WESTERN AUSTRALIA

VOLUME 47

PART 3

PUBLISHED 30TH OCTOBER, 1964

REGISTERED AT THE G.P.O , PERTH FOR TRANSMISSION BY POST AS A PERIODICAL

THE

ROYAL SOCIETY

OF

WESTERN AUSTRALIA

COUNCIL 1964-1965

President
Past President
Vice-Presidents

Joint Hon. Secretaries

Hon. Treasurer
Hon. Librarian
Hon. Editors

W. R. Wallace, Dip.For.

C. F. H. Jenkins, M.A.

J. H. Lord, B.Sc.

D. Merrilees, B.Sc.

Margaret E. Redman, B.Sc.

A. B. Hatch, M.Sc., Dip.For.

R. D. Royce, B.Sc. (Agric.).

Ariadna Neumann, B.A.

R. W. George, B.Sc., Ph.D. (1964).

J. E. Glover, B.Sc., Ph.D. (1965).

B. E. Balme, B.Sc. (Hons.).

J. S. Beard, M.A., B.Sc., Ph.D.

A. S. George, B.A.

J. G. Kay, B.Sc.

P. E. Playford, B.Sc., Ph.D.

R. T. Prider, B.Sc., Ph.D., M.Aust.I.M.M., F.G.S.
W. D. L. Ride, M.A., D.Phil.

D. L. Serventy, B.Sc. (Hons.), Ph.D.

Journal
of the

Royal Society of Western Australia
Vol. 47 Part 3

10. — Notes on the Skulls of two Western Australian Rodents with a key to
the Skulls of the Rodents of Southwestern Australia

By Ernest Lundelius, Jr.*

Manuscript received — 18th February. 1964

Remains of two species of Western Australian
rodents. Pseudomyi >• (Gyomys) Occident alls and
Pseudomys ( Pseudomys ) rawLinnae, hitherto
known from small samples arid few localities
have been found in several caves along the west
coast of Australia and along the south edge of
the Nullarbor Plain. Pseudomys (Gyomys) occi-
dentalis is reported for the first time from the
southern Nullarbor Plain. Minor morphological
differences can be seen in samples from different
areas.

Pseudomys ( Pseudomys ) rawlinnae is shown to
vary in the presence of an accessory cusp on
the anterior edge of M<, the degree of inter-
orbital constriction and the degree of develop-
ment of the spine on the zygomatic plate. This
variation makes it difficult to distinguish this
species from Pseudomys (Thetomys) nanus on
the basis of fragmentary material.

Statistical tables on each species and a key to
the skulls of the rodents of southwestern Aus-
tralia are given.

Introduction

Investigations of cave sediments from Western
Australia have resulted in the collection of
abundant material of two species of rodents,
Pseudomys (Gyomys) occidentals and Pseu-
domys ( Pseudomys ) rawlinnae , which have pre-
viously been known only from sparse material.

The author (Lundelius 1957, 1960, 1963) has
previously reported the presence of these two
species in cave deposits in Western Australia
but no details concerning their morphology
were given. The samples which form the basis
of this report are large enough to give reliable
estimates of the variability of some cranial and
dental characters. This should facilitate com-
parisons with samples from other areas as they
are found.

The present samples are obtained from the
following caves: Murraelellevan, 32 2'S. 126

4'E.; Madura, 32° 0'S. 127 O'E.; Webb and

Snake Pit, 31° 46S. 127° 51'E.: and Abrakurrie,
31° 39'S. 128° 26'E.

Pseudomys (Pseudomys) rawlinnae

This species was named by Troughton (1932 '
on the basis of material from the vicinity of
Rawlinna, Western Australia. Finlayson (1939)
described a series from the vicinity of Ooldea,
South Australia and gave the range and mean

* Department of Geology. University of Texas, Austin,
Texas.

of the measurements of four males and four
females. All of the material considered here
comes from caves along the Eyre Highway
between Balladonia and Eucla.

The largest sample and the one which forms
the basis of the statistical data, comes from
Murraelellevan Cave. Smaller samples are
available from Madura. Webb, Snake Pit and
Abrakurrie caves. Except for the material from
Madura and Abrakurrie caves which contain
some fill, all the specimens were collected from
the surface where they were associated with
remains of the rabbit Oryctolagus cuniculus and
the house mouse Mus musculus. These two
species have been introduced by the European
colonists within the last century. While there
is always the possibility of the remains of the
introduced species being mixed with older
material, the condition and the thorough
mixing of the bones indicates that all the sur-
face material is essentially the same age. The
presence of well-preserved bones of only small
animals indicates that owls were responsible
for the bone deposits.

The skull of this species closely resembles
that of Pseudomys shortridgei but is smaller
and more lightly built. The anterior edge of
the zygomatic plate is gently concave as in
Pseudomys shortridgei, but there is more of a
tendency to develop a dorsal spine. Some speci-
mens have a well developed spine, others have
the dorsal border of the zygomatic plate angled
but with no spine. Although members of the
genus Notomys also have the spine they are
easily distinguished by the broadening of the
upper surface of the anterior part of the zygoma.
There is no broadening in Pseudomys rawlinnae.

The skull shows considerable variation in the
width and shape of the interorbital area (Fig.
1). The width of the zygomatic arches also
appears to be variable but an objective mea-
surement is difficult on the present material.
Some of this variation in the skull is probably
attributable to age differences. Two skulls
( CNHM PM-4294, PM4352) which show narrow
zygomatic width, shortened rostra and rounded
skulls have unworn teeth indicating that they are
from relatively young individuals. Sexual dif-
ferences may also be involved but no data are
available on sexual dimorphism.

65

Fig. 1. — Dorsal views of
variation in interorbital

eight skulls of Pseudomys (Pseudomys) rawlinnae from Murraelellevan Cave showing
region. x2. A — PM 4294; B — PM 4357; C — PM 4295; D — PM 4296; E — PM 4339; F — PM 4344-
G— PM 4349; H— PM 4347.

TABLE 1

Statistical Data for a Sample of Pseudomys (Pseudomys) rawlinnae from Murraelellevan Cave and a Sample

from Ooldea, South Australia.

Sample from Murraelellevan Cave

No.

Mean

Standard

Deviation

Total length upper molars

38

4-95± 03

•23

Length M 1

38

2-42±-0l

•097

Length M-

37

1 -43 ±-01

•11

Length M 3

38

1 • 15± *01

•084

Width M 1

38

1*66 ±-01

•079

Interorbital width
Nasal length

2*2

3

3 * 88 dr * 04
10-82

•208

Palatal length

29

1 3 • 28 zh * 1 l

•608

Length anterior palatal foramina

29

IH7± -09

•504

Alveolar length upper molars ....

39

5 • 40 ± *04

•287

* Data from Finlayson (1939.)

Sample from Ooldea*

□efficient

of

ariation

Observed

4

Males

4 Females

Range

Mean

Observed

Range

Mean

Observed

Range

4-6

4-64-5-48

5-1

1!^

I

o

o

5-2

50-5-3

4-0

7-4

7-4

2-23-2-65
1 • 30-1 • 76
•97-1-29

....

....

4-7

1 • 53-1 • 79

5-36

3-49-4-19

4-0

3-9-41

4-1

3 -9-4 -5

10-43-11-45

11-5

11-3-11-7

11-2

11-0-11-6

4-58

12-21-14-55

15*6

15-1-16-3

16-0

15-6-16-6

8-17

5-05-7-03

6 • 6

6 • 6-6 • 6

6-8

6 -7-6 -9

5-31

5-02-6-17

.... j

....

The teeth are similar to those of Pseudomys
shortridgei but, as in the case of the skull, they
are smaller (Tables 1 and 2). The laminae of
the upper teeth do not slope as much as in
Pseudomys shortridgei. The teeth show some
variation in the degree of development of the
external row of cusps, presence of an accessory
cusp on the anterior edge of M 1 and the number
of posterior cusps of M :i and their configuration.

Of 89 specimens from Murraelellevan Cave, 18
(or 20%) possess a well defined accessory cus-
pule on the anterior edge of M l . A statistical
comparison of samples with the cuspule and
without it, shows no difference in the length of
M'. Of 79 right M 3 ’s from Murraelellevan Cave,
75 have a single posterior cusp free of the
midloph, 2 have it connected to the midloph
and 2 have two free posterior cusps.

TABLE 2

Statistical Data on a Sample of Pseudomys ( Pseudo -
my s) shortridgei from the Top One Foot of Sediments ,
Hasting's Cave , Western Australia.

• PSEUDOMYS NANUS WEDGE'S CAVE
A PSEUDOMYS NANUS TYPE
x PSEUDOMYS NANUS RECENT, W.A.

O PSEUDOMYS NANUS HASTING'S CAVE
a PSEUDOMYS RAWLINNAE MURRAELELLEVAN CAVE

Kg.

Mean

Stand-
dard
I )evia-
tion

Co-

efficient

Ya na-
tion

Observed

Range

Total Length
Upper Molars

25

5*7(5 4. *054

•27

4*70

i 5-23-6*11

Length M l ....

34

2-68- -030

•IS I

6 • 54

2*32-2*9!

Length M 2 ....

26

1*70- *023

*12

(5 • 99

1 - 4.i

i 1*91

Length M 8 . .

25

1 *35 • 021

*11

7*93

1-la

i -l *5i

Width M l ....

34

1*1)1 - • 018

*11

f> • 58

1 • 70-2 01

An examination of the coefficients of variation
given in Table 1 reveals that they are all within
the range to be expected in a population of
mammals. This implies that either this popula-
tion does not exhibit the high variability said
to be characteristic of some Australian desert
species (Jones 1923) or the sample from
Murraelellevan Cave was accumulated in a very
short period of time.

A comparison of the Murraelellevan Cave
material with Finlayson’s figures shows the two
samples to be similar. However, a comparison of
measurements of five characters of Finlayson’s
Ooldea sample with the Murraelellevan Cave
sample shows that the former measurements
always average larger. In two characters, nasal
length and palatal length, no specimen from
Murraelellevan Cave is as large as the largest
Ooldea specimen (Table 1). These minor differ-
ences probably do not have much significance.
Both samples are small so far as some characters
are concerned and may not be representative of
either population. It is to be expected that
different populations of a species will show
small differences.

The mandibles and lower dentitions are so
similar to those of Notomys mitchelli, which
occurs in the same area, that differentiation is
extremely difficult and no attempt is made to
describe them here.

The skull of Pseudomys rawlinnae , despite its
general resemblance to that of Pseudomys
shortridgei , is readily distinguished by its gener-
ally smaller size, lighter build and better
developed anterior spine on the zygomatic plate.
There seems to be no reason for not regarding
them as good species until such time as mor-
phologic intergrades are found.

Pseudomys rawlinnae is far more likely to be
confused with Pseudotnys nanus. Both species
have a spine on the zygomatic plate, no broaden-
ing of the upper anterior surface of the zygoma
and, as noted above, some specimens of Pseu-
domys rawlinnae have an accessory caspule on
the anterior edge of M 1 as in Pseudomys nanus.

Figures 2, 3 are scatter diagrams of width of
M l against length of M' and length of M 1
against length of M ;! showing the relationships
between a Recent sample of Pseudomys nanus,
2 subfossil samples of Pseudomys nanus and a
sample of Pseudomys rawlinnae . It is clear that
the principal difference between Pseudomys
nanus and Pseudomys rawlinnae , is size. This
is of little help in deciding whether they might
be conspecific since size is frequently different

1.5 -

□

a

oo

a

a

□

x

D

□

X

D

0 °o

O □

A

—I l I l I l 1 1 1—

2 2 2.3 24 25 2.6 2.7 2B 2.9 3.0

LENGTH M 1

Fig. 2.— Scatter diagram showing relationship between
length of M 1 and width of M l in several samples of
Pseudomys ( Thetomys ) nanus and Pseudomys (Pseu-
domys rawlirtnae from Western Australia,

16

1.5

ml

2

14

1.3

1.2

O O

(9

□ □
a

x

n □ Bp'Ll

o o
□

l.l -

B B

1.0 -

2.2 2 3 2 4 2 5 2.6 2 7 2.8 2 9 3.0

LENGTH m-

Fig. 3. — Scatter diagram showing relationship of length
of M 1 and length of M :i in several samples of Pseudomys
( Thetomys ) nanus and Pseudomys (Pseudomys) raw-
linnae from Western Australia. See figure 2 for explan-
ation.

in different populations of the same species.
It is also interesting that the subfossil samples
of Pseudomys nanus seem to show a slightly
larger size than the Recent sample.

Although the type locality of Pseudomys nanus
is along the west coast of Western Australia, a
large part of the intervening area is essentially
the same as the known area of occurrence of
Pseudomys rawlinnae and it is quite possible
that it actually occurs over a wide area. Finlay-
son ( 1941 ) has recorded a rodent which he refers
to Pseudomys nanus from 10 miles south of
Koonapandi in Central Australia. The presence
of very similar rodents in widely scattered
localities in central and western Australia sug-
gests that they might be local populations of
one widely distributed species.

Pseudomys (Gyomys) occidentalis

Pseudomys < Gyomys ) occidentalis was des-
cribed by Tate (1951) from two specimens from
Tambellup. southwestern Australia. Until the

67

material from the caves in western Australia
was discovered these were the only known
specimens.

The material described here comes from caves
along the west coast of Australia and from the
Nullarbor Plain (Lundelius 1957, 1960, 1963).
This new material not only adds to the known
geographic range of the species but also shows
that the species varies morphologically over its
range.

B

Fig- 5.— Lingual view of mandible of Pseudomys
(Gyomys) occidentalis from Murraelellevan Cave, x 4.

A

B

Fig. 4.— Ventral view of maxillary (A) and palate (B)
of Pseudomys (Gyomys) occidentalis from Murraelel-
levan Cave, x 5.

The presence of Pseudomys (Gyomys) occi-
dental's in the Recent deposits at Jurien Bay
reported by the author (Lundelius 1957)
extended its known range 300 miles northward
and indicated a wide distribution in the semi-
arid portions of western Australia. The dis-
covery of remains of this species in surficial
deposits of Murraelellevan Cave on the Nullarbor
Plains extends its range 500 miles eastward into
a very dry area and indicates that the species
had in the very recent past a very wide distri-
bution in the southern part of western Australia.

The skull of this species is characterized by a
broad zygomatic plate with a straight anterior
edge and no spine at the top. The anterior
part of the zygomatic arch is not broadened.
The molars, both upper and lower, are very
small in proportion to the size of the skull, very
short crowned and have almost no inclination
to the laminae. No specimens have been seen
with an accessory cuspule on the anterior edge
of M 1 .

Table 3 gives measurements for the type and
paratype, several specimens from Murraelellevan
Cave and statistical data on a sample from
Hasting’s Cave near Jurien Bay. It can be seen
that there are few significant differences between
these three widely separated samples. The
greatest difference is the generally smaller size
of the Nullarbor sample. Although the sample
is small it consistently falls slightly below or in
the lower end of the observed range of the
Hasting’s Cave sample.

The differences between the Hasting’s Cave
sample and the type and paratype from Tam-
bellup are much smaller. The length of the
lower molars, M,, M 2 , width of M,, length of
M 1 , M J and width of M 1 of the Tambellup speci-
mens are either slightly above or at the upper
end of the observed range of the Hasting’s Cave
sample.

68

TABLE 3

Statistical Data on a Sample of Pseudomys (Gyomys) occidentalis from the Surface Layer of Hasting’s Cave and
Murraelellevan Cave, Western Australia, compared with the Type and Paratype

Hasting’

S C

ave >

sample

Recent West
Australian

•rn

Muri

aelelhn

Sam|

an

le

Cave

No.

Mean

standard

Deviation

Coefficient

Variation

i )bsen
Range

id

CN11M

34725

Type

CN

34

HM

726

No.

M.c

an

Observed

Range

Length

M 1

10

1-

96 +

038

*12

6-

28

l

78-2

14

2 • 09

2-

11

Length

M-

15

1-

36 +

01

•04

2

85

i

27 1

41

1 • 30

1 •

39

Length

M 3

9

1

02 +

03

•09

8-

70

88 1

15

• 95

98

Width

M l

10

1

34 -

009

•03

2

24

l-

28 1

44

1-37

1

47

Width

across M‘‘s (alveolar)

12

5

20 r

052

• 18

3

42

4

84-5

50

Width

between M ’’s (alveolar) ....

15

2

78

049

• 1 9

6

82

2

•>•>.•■>

99

Width

across M ;, 's (alveolar) ...

5

5

11

•21

4

15

4

84 5

32

Width

between M 3 ’s (alveolar) ....

11

3

34 +

081

•27

8

02

2

92 3

so

Width

lerosfe ant. end premaxillu

16

o

75 +

065

•26

9

43

2

42 3

30

2*34

2

56

Lpper

alveolar length

26

4

71 +

056

- 29

6

19

4

26 5

82

4-44

4

46

2

4-

30

4

26 4-

34

Length

post incisor diastema ....

21

7

66

091

•42

5

44

7

00 s

44

7-62

7

91

Interorbital width

8

4

13 •

17

•50

12

1 9

3

68-5

23

4 *04

4

32

Depth

>f rostrum at tubercle

15

6

99 i

098

•38

5

43

6

08-7

45

6-18

6

57

Length

incisive foramina

16

4

06,

•065

•26

5

26

4

61-5

47

5-1*

5

6*

Length

lower molars

13

4

53 +

022

•08

1

78

4

38 4

68

4 • 58

4

67

Length

M,

20

»>

01 +

•013

•06

3

13

1

85 2

12

2*13

2

10

3

1

92

1

•83 1

99

Length

M,

14

T

35 i

•008

•03

2

•52

1

30 1

41

1-37

T

46

Length

m 3

14

i

10 -

•018

•07

6

72

97 1

23

• 99

1

07

19

Width

M

20

i

i9 ;

•009

•04

3

•30

1

13 1

29

1-23

1

30

3

1

17

l

-16-1

Length

lower post incisor diastema

19

4

30 •

•066

•29

6

65

3

•73 4

•91

3-82

4

10

4

3

93

3

•56-4

49

Alveola

r length lower molars ...

20

4

60 -

•029

•13

2

•80

4

39-4

86

4 • 62

4

58

4

4

54

4

•15-5

42

Depth

of jaw at center M j

18

4

12 +

• 059

. 25

5

•98

3

•55-4

50

3*78

4

00

4

3

29

2

• 98 -3

61

* Measurements from Tate (1951) ; C'NHM, Chicago National History Museum.

Fig. 7. — Lingual view of mandible of Pseudomys
( Gyomys ) occidentalis from Hasting’s Cave, x 3.5.

These differences between samples would
probably prove to be statistically significant if
larger samples were available. The differences
are of the kind and degree usually seen between
subspecies and probably represent subspecies of
Pseudomys occidentalis.

Key to the Skulls of Rodents of Southwestern

Australia

The identification of skeletal material of
rodents is usually considered to be a task for a
specialist with access to large collections of
comparative material. This is frequently the
case and will probably remain so in many
instances. It is possible to find in the rodents
of southwestern Australia consistent and easily
recognized characters which, with few excep-
tions will allow an identification to species. The
key presented here is based on the skull and
upper teeth. The identification of the mandibles
and lower teeth is very difficult even with access
to comparative material.

The geographic area over which this key is
expected to apply is western Australia from
Geraldton to the south coast and eastward along
the Nullarbor Plain to Eucla.

The identification of species of Rattus and
Mus may give trouble. The recognition of the
four species of Rattus mentioned in the key

69

should give no trouble but additional species of
Rattus may have been recently introduced from
Asia and will not be identifiable. It is also
probable that Asiatic species of Mus have been
introduced recently. The author has seen
specimens trapped in Western Australia which
are not referable to any of the species of Mus
or Rattus previously recognized in Western
Australia.

A number of species of Notomys have been
named from Western Australia, four from the
western coastal area. Although Mack (1961)
has synonomized many of these there is still
no reliable way to differentiate the species on
the basis of fragmentary material. Conse-
quently this key does not go beyond the generic
level.

Some characters, such as the presence of an
accessory cusp on the anterior edge of M 1 , the
presence of a spine on the zygomatic plate and
the number of roots on M 1 . which are widely
used in the literature should be used with some
caution. The accessory cusp on M 1 is supposed
to be present in Pseudomys < Thetomys ) nanus ,
Leggadina, and present or absent in Notomys
and absent in Pseudomys (Pseudomys) short-
ridgei. Rattus is supposed to have five roots
on M 1 ; Pseudomys , Leggadina, Notomys are
supposed to have three. The zygomatic plate
is supposed to have a spine in Pseudomys
(Thetomys) nanus and Notomys.

3. Snout short; incisive foramina

slit-like, occupying a large
part of the post incisive dia-
stema and extending well back
of the front edge of M>; inter-
orbital constriction at middle
of frontal .. . .....

Snout long; incisive foramina
elliptical with rounded ends,
occupying comparatively small
part of the post incisive dia-
stema and extending only to
anterior edge of M 1 ; inter-
orbital constriction well an-
terior to middle of frontal

4. Size large, molar row greater

than 7 mm

Size medium, molar row less
than 7 mm

5. Zygomatic plate close to ver-

tical; ridges on sides of brain-
case parallel

Zygomatic plate inclined out-
ward from skull; ridges on
side of braincase rounded

6. M 1 longer than M 2 plus M ;1

notch present on back side of
upper incisors

M 1 shorter than M 2 plus M :<
notch absent on back side of
upper incisors

7. Anterior edge of zygomatic

plate concave with anterior
projecting spine

Anterior edge of zygomatic
plate straight or convex

In fact, as noted above, a significant per-
centage of specimens of Pseudomys (Pseudomys)
rawlinnae and a few specimens of Pseudomys
( Pseudomys ) sliortridgei possess the anterior
cuspule on M 1 . The spine on the zygomatic
plate is apparently always present in Pseudomys
(Thetomys) nanus and Notomys and is variable
in Pseudomys (Pseudomys) rawlinnae. A few
specimens of Pseudomys ( Pseudomys ) short -
ridgei have four or five roots on M 1 .

The result of these variations is the virtual
impossibility of distinguishing positively Pseu-
domys (Thetomys) nanus an d Pseudomys (Pseu-
domys) rawlinnae on the basis of cranial
material. For this reason geographic distribu-
tion has been used to separate them until satis-
factory morphological criteria are found.

A few specimens of Pseudomys (Pseudomys)
shortridgei which have the anterior accessory
cuspule on M 1 are unlikely to be confused with
any other species because of size and the lack
of a spine on the zygomatic plate. Those with
more than three roots on M 1 are easily dis-
tinguished from Rattus by the poor development
of the outer row of cusps on M l ' z .

8. Anterior portion of zygomatic
arch broadened (2 to 3 times
wider than remainder of arch)

Anterior portion of zygomatic
arch not broadened

9. Found in Nullarbor Plain

Found along west coast of
Western Australia

LO. Size very small (length M 1 -**
less than 4 mm)

Size medium to large (length
more than 4 mm) ...

11. Accessory cusp on anterior end
of Mi

Accessory cusp not present on
anterior end of M 1

12. Molars small in proportion to
size of skull with weak lami-
nae

Molars robust with well deve-
loped laminae and cusps

a i

1. M- absent; M- greatly en-
a i

larged (length greater than 6
mm), wears to form 3 large
lakes; root pattern of M 1 com-
plex ... .... .... Hydromys

cl try sog aster

M- present; M- not greatly en-

3 1

larged 2

2. M 1 5-rooted; 3 well developed
cusps on 2 anterior lophs of

M 1 .. 3

Mi 3-rooted (or occasionally
4), only 2 well developed cusps
on 2 anterior lophs 6

13. Size large (length M'- ;i more
than 6.5 mm) with two rows of
heavy cusps on upper molars

Size medium (length Mi-^ less
than 6.5 mm) with two rows of
well developed cusps and one
of poorly developed cusps on
upper molars

14. Upper molar row more than
7.5 mm; a small oval foramen
immediately behind upper in-
cisors; bullae large

Upper molar row less than 7.5
mm; no small oval foramen
immediately behind upper in-
cisors; bullae small

5

Rattus fuscipes
Rattus sp.

Rattus norvegicus

Rattus rattus

Mus musculus

7

8
10

Notomys

9

Pseudomys (Pseu-
domys) rau-linnae

Pseudomys ( The-
tomys) nanus

11

12

Leggadina

hermannsbergensis

Pseudomys
(Gyomys) albo -
cinereous

Pseudomys
(Gyomys) occid cn-
talis

13

14

Pseudomys (Pseu-
domys) shortridgei

Leporillus conditor

Leporillus apicalis

70

Acknowledgements

The author is indebted to Mr. L. Glauert of
the Western Australian Museum and the
authorities of the British Museum of Natural
History for permission to study material in their
care. Prof. R. T. Prider and Prof. H. Waring
made available the facilities of the Geology and
Zoology departments of the University of
Western Australia. The author’s wife, Judith
Lundelius, assisted with the field work and read
the manuscript. The field work was done
during the author’s tenure as a Fulbright Fellow
at the University of Western Australia.

References

Finlayson, H. H. (1939). — Records and descriptions of
Muridae from Ooldea, South Australia.
Trans . Roy. Soc. S. Aust. 63: 354-364.

(1941). — On Central Australian mammals.

Part II the Muridae. Trans. Roy. Soc. S.
Aust. 65: 215-231.

Jones, Frederick Wood (1923). — “The mammals of South.

Australia.” (Govt. Printer: Adelaide.)
Lundelius, Ernest (1957). — Additions to knowledge of
ranges of Western Australian mammals. W.
Aust. Nat. 5: 173-182.

(1960). — Post Pleistocene faunal succession

in Western Australia and its climatic inter-
pretation. Proc. XXI Int. Geol. Congr.:
142-153.

1963. — Vertebrate remains from the Nullar-

bor Caves. J. Roy Soc. W. Aust. 46: 75-80.
Mack, George (1961). — Mammals from south-western
Queensland. Mem. Qd. Mus. 13: 213-229.
Tate, G. H. H. (1951). — The rodents of Australia and New
Guinea. Results of the Archbold Expe-
ditions. No. 65. Bull. Amer. Mus. Nat. Hist.
97: 187-430.

Troughton, Ellis LeG. (1932). — On five new rats of the
genus Pseudomys. Rec. Aust. Mus. 18:
237-294.

71

11. — Effect of Salt, Temperature and Seed Scarification on Germination of
Two Varieties of Arthrocnemum halocnemoides

By C. V. Malcolm*

Manuscript received — 19th May, 1964

Green, but ripe seed heads of Arthrocnemum
halocnemoides varieties pterygosperma and per -
granulatum were collected and dried and the
resulting seeds germinated well.

Germination of both varieties was tested in
solutions containing nil. 0.8, 1.6 and 2.4 per cent
sodium chloride. Tests were carried out in a
factorial experiment at eight temperature re-
gimes. Salinity caused reduction and delay in
germination of both varieties. 50 per cent re-
duction in germination of var. ptergosperma
occurred at about 8 g/1 NaCl and of var. per-
granulatum at about 20 g/1 NaCl. Salinity and
temperature interacted in their effects on ger-
mination.

Var. pterygosperma gave significant germina-
tion only at 5-35 C the temperature range at
which var. pergranulatum gave best germina-
tion.

Scarified and unscarified seed of var. per-
granulatum. was tested at six temperature re-
gimes. Scarification caused an increase and an
acceleration of germination. There was an in-
teraction between scarification and tempera-
ture.

ol: summer. Green inflorescences were collected,
air dried in the laboratory, and seed obtained by
threshing and cleaning. The resulting seed
germinated well and was used in the trials
reported in this paper.

In some preliminary attempts at germinating
seed of Arthrocnemum spp. in the laboratory
it became clear that several factors were opera-
tive. Scarification appeared to assist germina-
tion of seed of A. halocnemoides var. pergranu-
latum and temperature seemed to influence
germination of var. pterygosperma. The
germination trials discussed in this paper were
designed to elucidate these effects and to study
the effects of salt, a factor of considerable
importance in the natural habitat of the species.

Method

Introduction

Extensive areas of land in the agricultural
areas of Western Australia are too highly saline
for the growth of normal crops. As reported
by Burvill (1956), in the results of a statistical
survey, farmers in Western Australia estimated
that in 1955 there were 933,000 acres of salty
Samphire ( Arthrocnemum spp.) flats or Ti-tree
( Melaleuca spp.) flats on their farms in the
agricultural areas. This land is described by
Smith (1961, 1961a) as “valley water-logged”
and has an extremely salty watertable of 1,000
to 4,000 grains per gallon total soluble salts,
closer to the surface than 4 ft.

Smith and Malcolm (1959) discussed the use
of Kochia brevifolia and Atriplex spp. for grow-
ing on salt affected soils but the plants were
not recommended for waterlogged areas. In
seeking plants for waterlogged salty areas, the
naturally occurring Arthrocnemum spp. have
been investigated. The germination studies
reported here are part of a programme of
research into the use of Arthrocnemum spp.
for sowing as fodder plants on highly saline
winter waterlogged soils.

Seedheads of most plants dry off or ripen at
some stage in their life cycle. The seed may
then be harvested. In A. bidens, the seed heads
are retained in a dried and woody condition.
However, in A. halocnemoides varieties per-
granulation and pterygosperma the seed heads,
or inflorescences, may remain green and fleshy
for several months after the seed has apparently
ripened. As a result the opportunity for the
inflorescences to dry off passes with the passing

‘ Soil Research and Survey Section. W.A. Department
of Agriculture, Jarrah Rd., South Perth.

(i) Effect of temperature and salt on the
germination of A. halocnemoides varieties
pergranulatum and pterygosperma
The salinity treatments, nil, 0.8. 1.6 and 2.4
per cent, sodium chloride, were imposed by
placing the seeds in 15 mis of the appropriate
solution in a petri dish. The solutions were
pipetted off and replaced at regular intervals to
avoid concentration effects due to evaporation.

The temperature regimes were as follows:

20°C. constant 5-35°C. fluctuating

30°C. „ 15-30°C.

35°C. „ 15-45°C

60°C. „ 15-60°C.

The fluctuating temperatures were thermo-
statically controlled to give gradual changes
simulating normal day-night fluctuations as
described by Quinlivan (1962).

The temperature and salinity treatments were
combined factorially and three 50 seed replica-
tions were tested at each combination.

Germination was counted at 2-4 day intervals
over the duration of the trial which lasted for
33 days. Germination was taken to have com-
menced when the radical emerged from the
testa.

(ii) Effect of Temperature and Scarification
on Germination of A. halocnemoides var.
pergranulatum

The following temperature regimes were used:

35 C. constant

5-35°C. fluctuating

60 °C.

15-30°C.

15-45°C.

15-60°C.

72

For each temperature three replications each
of scarified and unscarified seed were tested
The scarification was applied at random to
seedlots, after they had been counted, by lightly
rubbing with fine emery paper.

Germination was counted at regular intervals
over the duration of the trial which lasted for
28 days.

Results

(i) Effect of temperature and salt
Both temperature and salinity caused marked
effects on germination (Table 1). Var. ptery-
gosperma gave significant germination only in
the 5-35°C. temperature treatment. Var.
pergranulatum also showed a preference for the
5-35°C. range, but gave some germination at
other temperatures.

A constant temperature of 60 °C. gave rapid
germination with var. pergranulatum, and seven
days from the commencement of the trial the
germination figures shown had been reached.
No germination occurred thereafter. It appeared
that the high temperature had broken the
dormancy of the seed but caused a high
mortality rate. By contrast seed in the 5-35°C.
range did not commence to germinate until
seven days had passed but reached a final
percentage of 75 as against 36 per cent, for the
60 C. treatment.

There was an interaction between temperature
and salinity effects similar to that shown by
Malcolm (1963) for Kochia brevifolia. While
0.8 per cent. NaCl caused no reduction in
germination of var. pergranulatum at 5-35 C.,
at 60 C. it caused a reduction from 36 to 25
per cent, germination. Similarly, 1.6 per cent.
NaCl caused a reduction from 75 to 52 per cent.
(31 per cent reduction) in germination of var.
pergranulatum at 5-35°C., but at 60' C. a reduc-
tion of from 36 to 13 per cent. (64 per cent,
reduction) was caused. Thus, in each case
salinity and a temperature of 60 C. interacted
to give greater reductions in germination than
was shown for either factor operating separ-
ately.

TABLE 1

Effect of temperature and salt on germination of two
varieties of Arthrocnemum lialocnemoides
(trwvm i"*r cent, germination from 3 replication* of 50 seeds)

Variety

pterygonperim

pergranulatumi

Sftitrifty

0

o-s

10

2-4

0

(l-K

1

1*0

2* 4

Germination

14 da;

ya iiftij

■r commencement •

experiment

Temperature .

20

1 1

1

0

0

0*3

0

o l

0

30

1 |

0

1)

0

0

0

o j

0

35

1 i

l

0 ,

1

7

3

3 |

1

00

0 1

l

0

0

30

IS

1 1

0

5-35

7

4

o !

1

28

17

3

0

15 30

O |

1

o '

0

1

3

1

0

15* 45

1 '

1.

ii

0

0

1

1

0

15-00

l 1

1

0

0

10

0

1

1

FT uni Germination 33

(Lay* ;

iR.er (*t

tnimem-einent of experimi

eat

Temperature C. :

20

l '

3

1)

0

1

1

0

0

30

l 1

0

0

0

0

1

0

0

35

1

1

0

1

s

4

3

1

00

0

1

0 1

0

30

25

13

3

5—35

72 |

37

15

•>

75

75

52

0

15-30

4

1

0

0

4

5

5

0

15-45

J

l

1

0

10

3

1

0

15-00

1

1

0

0

20

13

1

1

Three temperature factors appear to influence
germination. As discussed, high temperature
accelerates germination of var. pergranulatum,
though similar effects were not noted for var.
pterygosperma. The effect of high temperature
is shown in the germination of var. pergranu-
latum in the progressive increase in germina-
tion through the 15-30°C., 15-45*C. t and 15-
GO J C. ranges. High temperature failed to affect
germination of var. pterygosperma.

A second factor operating is temperature
fluctuation, the effect of which is most clearly
seen in the results for var. pterygosperma.
While constant temperatures of 20, 30 and 35°C.
gave almost no germination, a fluctuating tem-
perature of 5-35 C. gave 72 per cent, germina-
tion.

Finally a relatively cold temperature is
required by both varieties as the minimum
temperature in the fluctuating range. Ranges
of 15-30. 15-45, and 15-60 °C. gave almost no
germination with var. pterygosperma but a
range of 5-35 °C. gave high germination. The
range 15-45 C. is of the same magnitude, 30
degrees, as the 5-35°C. range but does not
include a sufficiently cold temperature. Var
pergranulatum gave similar results but effects
of high temperature caused a variation. It is
of interest to note that germination of both
varieties occurred in the glasshouse where the
temperature range was approximately 10-31 C.

The specific temperature requirements indi-
cate dormancy problems. The response to the
60 °C. treatment suggests hard-coatedness (see
below). The response to fluctuating tempera-
tures and to a relatively cold temperature e.g.
5 or 10°C. may be due in part to hard-
coatedness and partly to physiological dormancy.

(ii) Effect of Temperature and Scarification

In this trial, the effects of temperature were
similar to those discussed under Section <i > , but
varied in the effect of high temperature. 60 C.
constant and 15-60 °C. fluctuating temperatures
failed to give comparable figures. The reason
for this is not known but may have been due
to the use of a different batch of seed.

Scarification caused a marked increase and
acceleration in germination. For the 5-35 C.
range unscarified seed gave a germination of
43 per cent., but scarified seed gave 64 per cent,
germination. In the 15-45 C. range the differ-
ence was even greater, 11 per cent, germination
being obtained without scarification and 46 per
cent, with scarified seed.

Scarification resulted in reasonable germina-
tion being obtained at temperatures which
normally gave very poor results. For example
35 C. germination without scarification was 9
per cent, but with scarification was 36 per cent.

Seed which had been scarified gave 34 per
cent, germination by the end of 13 days testing
at 15-45 C. but unscarified seed had only given
5 per cent, germination over this period. The
large difference was due in part, to the acceler-
ating effect of scarification.

It is likely the scarification treatment was not
even on all seeds in view of the difficulty of
scarifying such tiny seeds in a controlled man-
ner. The seeds are encased in a hard black
testa and scarification was continued until the

73

TABLE 2

Effect of temperature and scarification on germination of A. halocnemoides var. pergranulatum (mean per cent.

germination from 3 replications of 50 seeds each)

( commenced 1 1 .x .63 )

Scarified i'nscarlfied

Date

Temp. (•.

Tom

i. a

35

60

5-35

15-30

15-45

15-60

35

60

5-35

15-30

1 5-45

15-60

1 4.x. (13

0

0

0

1

3

0

0

0

0

0

0

0

1

16.X. 63

s

3

0

•>

10

9

0

l

0

1

1

•>

1 7.X.63

7

3

1

•>

17

10

0

l

8

3

1

2 1.x. 63

<)

3

16

7

25

1 1

•;

1

3

l

3

-j

24.x. 63

12

4

23

9

34

12

3

1

4

3

l

3

28.X.63

33

5

32

11

39

13

5

1

13

4

3 1.x. 63

36

II

55

16

45

15

9

4

39

10

6.xi 63

36

1 1

64

1 i

46

17

9

5

43

1 1

6

6

8.xi.03

36

11

64

17

40

17

9

5

43

7

11

majority of seeds in the sample bore light
coloured patches. The variations are accounted
for in the design of the experiment.

Discussion

(i) Effect of temperature and salt

Novikoff (1946) reported 50 per cent, reduc-
tion in germination for several vegetable crops
to occur at 18-22 gm/1 NaCl. Malcolm (1963)
found that for Kochia brevifolia 50 per cent,
reduction in germination occurred at 15-20 gm/1
NaCl. In the present study, var. pterygo -
sperma gave 50 per cent, reduction in germina-
tion at about 8 gm/1 NaCl, and var. pergranu-
latum probably at about 20 gm/1 NaCl. It may
be concluded that from the point of view of salt
tolerance at germination, var, pergranulatum
is similar to many vegetable crops and var
pterygosperma is less tolerant.

The results indicate that temperature is a
vital factor determining when A. halocnemoides
will germinate. Moreover, germination will not
occur when the salinity of the soil solution is
greater than 2.4 per cent. NaCl. Smith (1961a)
reported seasonal changes in salinity of a bare
soil at Quairading. Arthrocnemum spp. are
growing nearby. The lowest value obtained for
the 0-j inch samples was 0.80 per cent. NaCl on
the oven dry basis in July and the average for
the July sampling was 1.11 per cent. NaCl. The
soil in question was a fine sandy clay loam.
The combination of low winter temperatures
and high salinity would provide a severe
limitation to the germination of the varieties
under study. It is apparent that the plants
have developed a means of avoiding germina-
tion into too harsh an environment. A great
deal of rain would be required to leach salts
from the topsoil sufficiently to allow germina-
tion to occur.

The results indicate that field planting of the
species could be undertaken in the autumn
before rain makes the soil unworkable, and the
seeds would germinate when conditions become
favourable. Salt tolerance of the seedlings must
increase extremely rapidly and early in life
surpass that of most other plants.

(ii) Effect of Temperature and Scarification

The scarification treatment was more effec-
tive in breaking dormancy than the best
temperature regime under test, but still did

not ensure good germination at unfavourable
temperatures. On the other hand the seed coat
appears to be slowly permeable on at least a
proportion of seeds — those which germinate at
5-35 c C. without scarification. Scarification
allows the slowly permeable seeds to germinate
more rapidly and allows further impermeable
seeds to germinate.

Assuming the main effect of scarification is
to allow water to gain ready access to the seed,
an interaction can be shown between scarifica-
tion and temperature. Germination of scarified
seed can be considered as the potential germi-
nation for each temperature range. Leaving
seed unscarified would be expected to cause the
same proportional decrease in germination at
all temperatures. However, at 15-45°C. the
germination of unscarified seed amounted to
11 per cent, about one quarter that for scarified
seed. Whereas at 5-35°C. unscarified seed gave
43 per cent, germination, about two thirds that
for scarified seed. The 5-35°C. temperature
range appears to influence the permeability of
the seed coat in some way and make up for
non-scarification.

Acknowledgements

The assistance of Mr. T. E. H. Aplin, of the
Botany Branch of the Western Australian
Department of Agriculture, who identified the
species, is gratefully acknowledged.

The assistance of Mr. B. Goldspink and Mr.
B. Marshall in the laboratory was greatly
appreciated.

ncici cuucs

Burvill, G. H. (19561. — Salt Land Survey. 1955. J. Agric
W. Aust. 5 (3rd series) : 113-119.

Malcolm, C. V. (1963). — M.Sc.( Agric. ) Thesis. An Agro-
nomic Study of Kochia brevifolia. W A
University.

Novikoff, V. (1946).— Notes sur l’utilisation des auxes
salees. Ann. Serv. Bot. & Agron. Tunisie
19 : 139-162.

Quinlivan, B. J. ( 1961 ).— M.Sc.( Agric.). Thesis. Some
Aspects of Seed Production and Hard Seeded-
ness in the West Australian Blue Lupin
Vv .A. University.

Smith. S. T. (1961). — Soil Salinity in Western Australia.

J. Agric. W. Aust. 2 (4th series) : 757-760.

Smith. S. T. (1961a). — M. Sc. (Agric.) Thesis. Some
Aspects of Soil Salinity in Western Australia.
W.A. University.

Smith. S. T and Malcolm. C. V. (1959).— Bringing
Wheatbelt Saltland back into Production.
J. Agric. W. Aust. 8 (3rd series) : 263-267.

74

12. — Ashbed and Nutrients in the Growth of Seedlings of Karri (Eucalyptus

diversicolor F.v.M.)

O. W. Loneragan* and J. F. Loneragant

Manuscript received — 19th May, 1964

When karri seedlings were grown in pots of
soils from karri forest areas they showed re-
markable responses to nitrogen and phosphorus
fertilizers. The addition of nitrogen or phos-
phorous fertilizers separately gave only small
increases in yield of seedlings, while the simul-
taneous addition of both fertilizers gave very
large responses (up to three-fold in height and
twenty-fold in dry weight).

Responses to ash and to heat treatment of the
soil were also obtained. The major part of the
ash response was shown to be due to the supply
of phosphorus in the ash. The heat response is
also believed to be largely due to increased
supply of nutrients.

It is suggested that part of the well known
ashbed response of karri forests is due to
nutrients released in the ashing process.

Introduction

Growth of karri (Eucalyptus diversicolor
F.v.M.) seedlings in the south-west of Western
Australia is stimulated considerably in ashbeds
where heaps of forest slash have been burned
(Loneragan, 1961). Growth of young karri
seedlings in non -ashbed areas is generally so
poor that deliberate creation of ashbeds has
become part of karri forest regeneration practice.

Karri ash contains plant nutrients including
phosphorus, calcium, magnesium, and potassium
(Stoate, 1950: Hatch, 1960) which may be

important in the ashbed response. However,
other marked chemical and physical changes
are associated with ashbed treatment (Hatch,
1960) and no nutrient responses of karri have
been recorded.

The work described in this paper was under-
taken to assess nutrient deficiencies of a number
of karri soils for optimum growth of karri
seedlings, and to see to what extent nutrients
might account for some of the stimulating
effects of ashbeds.

Methods

Experiments 1, 2, 3.

General. — A red-brown sandy loam (karri-
type) soil was taken from an area which showed
typical ashbed response in regenerating karri
seedlings at Crowea, 20 miles south of Manjimup.
Soil from the top six inches was sieved, mixed,
and 3,000 g placed in plastic bags in unglazed
earthenware containers of 8 -inch diameter at
the surface.

All fertilizer additions are expressed as weight/
acre calculated from a surface area basis < 1
cwt/acre 0.405g/pot). The plastic bags were

* Forests Department, Manjimup, Western Australia,
j- Department of Soil Science and Plant Nutrition. Insti-
tute of Agriculture. University of Western Australia.

perforated so that the soil was freely drained.
The pots of soil were kept in the open at the
Forests Department, Manjimup.

Karri (Eucalyptus diversicolor) seeds were
sown in all pots on August 11 and 12, 1961.
Natural rain was supplemented by de-ionised
water as required. Plants were thinned to 5 per
pot on October 2, to 3 per pot on October 30,
and to 1 per pot on December 6.

The heights of plants from cotyledonary node
to growing point was measured at frequent inter-
vals throughout the experiment: data in tables
refer to heights on April 26. about two weeks
prior to harvest. At this time the thickness of
the stem just above the cotyledonary node was
also measured with calipers.

Shoots and roots of plants of experiments 1
and 3 were separated at the cotyledonary nodes
and harvested on May 4 and 3, 1962, respectively:
plants of experiment 2 were not harvested.

Harvested material was dried in an oven at
105 C, cooled in a desiccator, and weighed.

Detail . — Experiment 1 — To test the effects of
nutrients on the ashbed response of karri seed-
lings on karri-type soil.

No basal fertilizers were applied. Treatments
were applied in quadruplicate, in a 2 x 2 x 2
factorial design as follows:

Heat — nil or 12 hours at about 150 C in an
oven: the soil was placed directly in
the earthenware pot for this purpose:
the plastic bag was replaced after treat-
ment.

Ash — nil or 3 tons of ash/acre. Ash was
collected from tree bark which had been
burned in a hot fire; it was mixed
throughout the top inch of soil.

Nutrients — nil or N, P. K, Ca, Mg, Mn, Cu,
Cl, Zn, B, Mo, Co. The nutrients were
added as the same salts, at the same
rates, and in the same manner as were
used for basal and for treatment dress-
ings in experiment 2, with the omission
of NaHiPOi.

Experiment 2 — to test the response of karri
seedlings on karri-type soil to applications of
K, Ca, Mg, S, Mn, Cu, Cl. Zn, B, Mo, and Co
fertilizers.

Basal dressings of nitrogen and phosphorus
were applied. NH, NO< was applied at 14 cwt/
acre in aliquots of 1 cwt/acre to the soil surface
at regular intervals. Phosphate was applied as
either the sodium or the potassium salt at levels
equivalent in phosphate to 4 cwt of superphos-
phate/acre.

75

Treatments were applied in an unreplicated 2'
factorial design as follows:

KHv-POi at 1.5 cwt/acre or NaH-POi at 1.3
cwt/acre

CaCOi — nil or 2 cwt/acre
MgSCh — nil or 56 lb/acre
MnS0i.4H.-0 + CuS0,.5H>0 + ZnS0,.7H 2 0
— nil or 14 -f- 7 4- 7 lb/acre
Na*Bf 07 . 10 H 2 0 4- (NH, ) (; .Mo 7 0- 1 .4H,0 +
C 0 CL. 6 H-O — nil or 34 + 1 + 1 lb /acre.
All salts except CaCOs, which was applied in
solid form, were applied in solution to the sur-
face of the soil. When all salts had been applied
they were mixed through the surface inch of
soil.

Experiment 3 — To test the response of karri
seedlings on karri-type soil to applications of
N, P. and K fertilizers.

Basal dressings of salts of Ca, Mg, S, Mn, Cu.
Cl, Zn. B, Mo. and Co were applied. The nut-
rients were applied as salts in the same form,
rate, and manner as the nutrient treatment in
experiment 2, with the omission of KH PO, and
NaHsPOi.

Two treatments were applied in octuplicate in
a 2 x 2 factorial design as follows:

KH 2 POi — nil or 1.5 cwt/acre applied in
solution to the soil surface.

NH, NO y — nil or 14 cwt/acre applied as in
experiment 2.

Experiment 4.

Loamy sand was taken from the surface of
soils in two adjacent areas on either side of a
narrow jarrah-marri (Eucalyptus marginata-E.
ccdophylla) and karri-marri ecotone at Snake
Gully near Manjimup (see Plate 1. Loneragan.
1961).

The soils were prepared at the same time and
in the same way as the karri soil used in experi-
ments 1, 2, and 3, with the exception that six-
inch diameter pots were used. <0.23 g of ferti-
lizer/pot 1 cwt/acre).

A simple experiment was designed to test the
effects of nitrogen and phosphorus fertilizers on
the growth of karri seedlings on the karri-marri
and the jarrah-marri soils.

The basal and treatment dressings of fertilizers
were identical to those of experiment 3. The
treatments were imposed in duplicate.

Experiment 5.

Loam was taken from ashbed and non-ashbed
areas of the surface soil of a regenerating karri
forest at Easter Brook, 20 miles west of Man-
jimup. Pots of soil were prepared as in experi-
ment 4.

Treatments consisted of either no fertilizers
or dressings equivalent to 10 cwt of commercial
N. P. K mixed fertilizer (“Nitrophoska red”)
per acre. These were imposed in duplicate.

Seed was sown on 11 July, 1962, thinned to
1 plant per pot, and harvested on 15 may. 1963.

Experiment 6.

Loam was taken from ashbed and non-ashbed
areas of the surface soil of a regenerating karri
forest at Mattaband, 35 miles south-east of
Manjimup. Pots of soil were prepared as in
experiment 4.

Treatments were applied in duplicate as
follows:

N — nil or 10 cwt of calcium ammonium
nitrate/acre.

P — nil or 10 cwt of lime -super /acre.

Karri seeds were sown on 1 June, 1962,
thinned to 1 plant per pot, and harvested on 15
May, 1963.

Experiments 7, 8.

General. — Soil was taken from the same area
as in experiments 1, 2, and 3. Soil from the
top six inches was mixed and 1,800 g placed in
plastic bags in polystyrene containers of 6 -inch
diameter at the surface (0.23g fertilizer /pot
1 cwt/acre). The containers were not drained
and the water content of the soil was main-
tained around field capacity by adding deionised
water to weight. The plants were grown in a
glasshouse at Perth: the glasshouse was shaded
to about half natural daylight from December 1.

Karri seeds were sown in all pots on August 8,
1962. Plants were thinned to 5 per pot on
September 11. and to 1 per pot on November 16.
Tops of plants were harvested on December 19,
1962 and treated as in experiments 1, 2, and 3.

Detail .—Experiment 7— To test if some of the
response of karri seedlings to ash is due to
supply of phosphate.

Basal dressings were applied of salts of N, K.
Mg, S, Mn, Cu, Zn, B. Mo. and Co. N was
applied in solution as NE,NO ;i in aliquots of 1
cwt/acre given at intervals throughout the
experiment to give a total of 4 cwt/acre. All
other basal nutrients were applied in the same
form and at the same rate as in experiment 3:
in addition, K^SO, was applied in solution at 2
cwt/acre. The basal dressings were pipetted on
the soil surface, allowed to dry, and mixed
thoroughly with all the soil in the pot.

Treatments were applied in quadruplicate in a
4 x 4 factorial design as follows:

P — phosphate equivalent to superphos-
phate at O. 4, 10. or 25 cwt/acre: the
replicates were split into pairs which
received either KH.PO, or NaH-PO, in
solution at appropriate rates.

Ash — O, 1, 3, or 9 tons/acre; ash was
prepared as in experiment 1.

After application, the treatments were allowed
to dry and were mixed thoroughly through the
volume of the soil.

Experiment 8 — To test if the response of karri
seedling to ash is due in any way to an effect
on nitrogen supply.

Basal dressings were applied of salts of P, K,
Mg, S, Mn, Cu. Cl, Zn, B, Mo, and Co. The salts
were applied in the same forms and at the same
rates as in experiment 2, with the omission of
NaH-POi and CaCO s , and the inclusion of
K 2 SO 1 at 2 cwt/acre. The basal dressings were
pipetted on the soil surface, allowed to dry, and
mixed thoroughly with all the soil in the pot.

Treatments were applied in quadruplicate in
a 2 x 2 factorial design as follows:

NHjNO.j — nil or 4 cwt/acre applied in
solution in aliquots of 1 cwt/acre.

Ash nil or 3 tons/acre: ash was prepared
and mixed as in experiment 4.

76

■rf /

■

Fig. 1. Effects of ash. heat, and nutrients on the growth and form of karri seedlings grown for eight months
in pots of karri topsoil (Experiment 1). Top— no heat. Bottom— heat treatment of soil before planting.
Left to right— no ash. no nutrients: no ash, 4- nutrients: 4- ash. no nutrients: 4- ash, 4- nutrients.

77

Results

Ash, heat , and nutrient responses
Ashbed treatments markedly stimulated the
growth of karri seedlings grown in pots of karri
soil (Table 1). Application of ash, in the absence
of nutrients and of heat treatments, increased
the height of the seedlings one-and-a-half times,
doubled the stem diameter, and trebled the dry
matter of tops and roots. Heat treatment of the
soil prior to sowing also gave, in the absence of
ash and of nutrients, similar large increases in
seedling growth.

TABLE 1

Effects of ash. heat, and nutrients on the growth of
karri seedlings for eight months in pots of karri top-
soil (Experiment 1 ).

No Heat

Heat

- A sh

| | Ash

— Ash

1 H- Asli

Dry weight of shoot

7

26

35

79

(g x 10 1 per plant)

16 2

271

259

353

Height uf shoot/ — cm.

- —

27

41

45

59

66

S3

7«

1 S7

Diameter of stem— '

—

14

26

24

37

cm. x 10 •

T

52 j

57

60

65

Drv weight of shoot

2

7

10

23

percentage maximum!

-1-

46

V 7

73

100

Height x girth— per-

—

1

7

7

2* ?

rentage maximum

48

73

74

100

All main treatment effects, significant at. P< 0*001 : all iuter-

artions, not significant. Dry weight data were transformed to
square roots "before analysis.

Nutrients applied to the same soil stimulated
karri seedling growth to an even greater extent
than ashbed treatments (Table 1). In the
absence of ash and heat treatments, nutrients
nearly trebled height growth and stem diameter
and increased dry matter production twenty-
fold. The more striking stimulation of dry
matter production, compared with height and
stem diameter, was partly due to the effects of
nutrients in increasing branching and leaf pro-
duction (Fig. 1) and partly due to the fact that
the weight of the main stem increases with a
function of the product of its height and square
of its diameter. Despite the striking effects of
treatments on seedling form, the product of
height x girth reflected relative effects of treat-
ment on dry weight production very closely
(Table 1).

In the presence of ash and heat treatments,
nutrients again increased growth of karri seed-
lings. Interactions between treatments were in
all cases not significant.

The nature of the nutrient response

The striking effect of the nutrient mixture in
stimulating karri seedling growth was due almost
entirely to the supply of nitrogen and phosphate
in the mixture. None of the nutrients K, Ca,
Cu, Mn, Zn, Co, Cl, B, or Mo, was deficient in
the soil since, when nitrogen and phosphate
were applied, the omission of any of them had
no effect at all on the appearance, shoot height,
or stem diameter of karri seedlings. The omis-
sion of Mg SO 4 from the nutrient mixture had

no effect on shoot height, but depressed stem
diameter slightly (by 13%: significant at 5%)
No dry weight data were collected on plants of
this experiment. On the other hand, the failure
to supply either nitrogen or phosphorus to the
seedlings, even in the presence of all other
nutrient salts, led to severe restriction of growth
(Fig. 2).

For optimum seedling growth, nitrogen and
phosphorus appear to be deficient in many karri
forest soils. Large responses of karri seedlings
were obtained when nitrogen and phosphorus
fertilizers were applied to pots of all soils
examined (Table 2).

TABLE Z

Effects of nitrogen (N) and phosphorus ( P ) fertilizers
on the growth of karri seedlings grown in pots of var-
ious soils from karri forest areas (Experiments 4. 5, 6)
Dry weight of shoots (g x 10 1 per plant)

— N —I 1 — N - F

X p

| 4N -|-P

Snake Gully (karri-marri) .. .

8 16

17

96

Snake (iull.v ( jnrrah-marri)

2 4

4

109

Has ter Brook (ashbed)

20

77

Easter Brook (non-ash bed) |

21

rn*

76

-Mattaband (ashbed)

11

86

Mattaband (non-ashbecl) ....

4 14

21

91

The nature of the ash response
Phosphorus. — A major part of the effect of ash
in stimulating karri growth on soil in pots was
due to an increase in phosphorus supply to the
plant (Table 3). As in previous experiments,
karri seedlings grew poorly when phosphorus
was omitted from the soil to which all other
known essential nutrients were added. In the
absence of phosphorus, addition of ash up to
9 tons per acre increased dry matter production
of the shoots six times. But in the presence of
phosphorus the response to ash disappeared
(interaction significant at P<0.001).

TABLE 3.

Effects and interactions of phosphate fertilizers and
ash on the growth of karri seedlings grown for four
months in pots of karri topsoil given adequate dress-
ings of other nutrients (Experiment 7).

Dry weight of shoots (g x 10 1 per plant).

Ash

Tons/'aere

Form of
Phosphate

Phosphate ( = 1* in ewt.s snper/acre)

0

4

10

25

0 /

Nu

4

67

69

76

K

4

66

70

50

i /

Na

8

69

66

74

\

K

7

71

77

77

3 /

Na

12

72

72

81

K

13

7 /

79

84

q s

Na

18

73

82

77

• \

k

2S

70

.84

75

The response of karri seedlings to ash in this
experiment is thus due to its effect on phos-
phorus supply. This effect is explained by the
phosphorus actually added to the severely phos-
phorus-deficient soil in the ash (0.3%). A dress-
ing of 9 tons of ash per acre would contain

78

Fig. 2.-— Effect of ammonium nitrate (N) and potassium
phosphate (P) on the growth of karri seedlings grown
in pots of karri topsoil given adequate levels of other
nutrients (Experiment 3). Interaction significant at

£< 0 . 001 .

phosphorus equivalent to 6 cwt of superphos-
phate per acre. The failure of this treatment
to stimulate growth of karri to the same extent
as equivalent dressings of phosphate salts was
probably due to poorer availability of the phos-
phorus from the ash.

Nitrogen. — In addition to increasing phos-
phorus supply to karri seedlings, ash treatment
increased nitrogen supply to an appreciable
extent (Table 4). In the presence of all known
nutrients except nitrogen, karri seedling growth
was again poor. In the absence of nitrogen,
addition of ash doubled the yield of shoots. In
the presence of nitrogen there was no response
at all to ash (interaction significant at P<0.001).
Thus, in this experiment the response to ash
must have been due entirely to an effect in
increasing nitrogen supply.

TABLE 4

Effects and interaction of ammonium nitrate and ash
on the growth of karri seedlings grown for four months
in pots of karri topsoil given adequate dressings of
other nutrients. (Experiment 8).

Dry weight of shoots (g x 10- 1

1 per plant)

— Ash

'i* Ash

Xo ammonium nitrate

i

S '

10

+ ammonium nitrate

S(>

86

The ash contained no measurable nitrogen,
so that its effect on nitrogen supply must have
been indirect.

Other factors. — Although in experiments 7 and
8 all of the ash response was replaced by
nutrients, in experiment 1 there was some
response to ash even in the presence of nutrients.
The additional response given by ash in the
presence of nutrients was probably partly due

to the supply of sub-optimal amounts of nitro-
gen in the nutrient treatment, combined with
heavy leaching in the early stages of growth.
However, it is still possible that in experiment 1
a small part of the ash response in the presence
of nutrients was due to factors other than nitro-
gen and phosphorus supply. Because of small
differences in experimental technique (e.g. size
and drainage of the pots, depth of mixing of
treatments, water supply), it is not possible to
make an absolute comparison between experi-
ment 1 and experiments 7 and 8.

However, it is clear that under all conditions,
nitrogen and phosphorus supply are the major
limiting factors to the growth of karri seedlings
on the soils examined. It is also clear that the
prime effect of ash is to increase the supply of
phosphorus.

Discussion

A number of soils from karri forests have been
shown to be extremely deficient in both nitrogen
and phosphorus for the growth of karri seed-
lings. Soils from regenerating Eucalyptus
obliqua forests have previously been shown to
be deficient in phosphorus for growth of E.
obliqua seedlings (Attiwill, 1962). It seems,
then, that the growth of seedlings and perhaps
also of trees regenerating in eucalypt forests
may be limited by the same widespread defi-
ciences of nitrogen and phosphorus which limit
the growth of agronomic crops on so many
Australian soils.

Beadle (1954, 1962) has in fact suggested that
soil phosphorus status is an important deter-
minant of plant communities in Eucalypt forest
areas of eastern Australia. In addition, he has
shown a marked response of the seedlings of
three Eucalypt species to soils with increasing
phosphorus status. His photographs show
changes in leaf size and seedling form with
increasing soil phosphorus similar to those pro-
duced by nutrients in karri seedlings.

The striking nutrient responses of karri seed-
lings reported in this paper should not be extra-
polated directly to the forest. The results were
obtained with seedlings grown for a very short
period of time, in the absence of competition,
and with adequate water. All of these con-
ditions are modified in the forest. Moreover,
nutrient treatments in pots caused dramatic
changes towards a different form of the karri
seedlings. However, the results do indicate the
potential which exists for stimulation of tree
growth by adding fertilizers to karri forest soils.

Heat treatment of soils and dressings of wood
ash also stimulated growffh of karri seedlings.
The effects of ash were shown to be largely
due to increased phsophorus supply to the seed-
lings. Heat treatment of forest soil has also
been shown to liberate phosporus for the growth
of Eucalyptus obliqua seedlings (Attiwill, 1962).
In addition, heat treatment of the same soil
appears to have released some nitrogen to the
seedlings. There are reports of increased supply
of nitrogen from organic matter after heating
(Vlamis et al., 1955), although there may, in
fact, be a loss of total nitrogen from the soil
through burning (Barnette and Hester, 1930).
It is suggested that in the present experiment
the main effects of heat may also have been

79

through the release of nutrients for the growth
of karri seedlings. No attempt was made in this
work to resolve the effects of heat.

Even when the soil had been heat treated and
large quantities of ash applied, nutrients stimu-
lated karri seedling growth. This suggests that
ashbed responses in the forest could be greatly
augmented by fertilizers. Such fertilizers would
probably be best applied at the earliest oppor-
tunity to encourage trees to dominate the site.

In some pot experiments, nutrients replaced
ash. However, the effects of fire which produce
ashbeds in the forests are complex (see review
by Ahlgren and Ahlgren, 1960). Therefore,
extrapolation to the forest situation of the
effects of heat and ash treatments on the growth
of karri seedlings in pots has even more limita-
tions than the extrapolation of nutrient
responses. Marked effects of burning forest
slash on the physical and chemical properties
of the soils in karri forests have been recorded
by Hatch (I960). In addition to these effects,
burning of forest slash may modify the viability
of seed and therefore the regeneration of plants
in the vicinity. Heat from the fire may destroy
surface roots and seeds of competitive species,
and so permits unhindered germination and
growth of seed freshly introduced from above
by seed trees in the area. This competition-free
seed bed may be an important part of the ash-
bed response in karri regeneration (Loneragan,
1961). At the same time, the ability of seeds
of some Acacia species to germinate is increased
by mild heat conditions on the fringe of the ash-
bed, so that leguminous species may be favoured
to colonize the fringe areas of the ashbed. In
the long term, these species could be expected
to improve the nitrogen status of the soil. Such
changes in the botanical composition of the
competitive species produced by ashbeds could
produce marked changes in the competition
which the karri suffers for light, nutrients, and
water.

Whatever the complexity of the long term
ashbed response, the results presented in this
paper clearly indicate that nutrients supplied by
ashbeds could make an important contribution
to the accelerated growth of young karri seed-
lings in the ashbeds of regenerating forests.

Acknowledgements

The authors are grateful to Dr. D. W. Goodall,
C.S.I.R.O. Division of Mathematical Statistics,
for the statistical analyses and to Mr. A. J. Hart
for considerable help in Experiments 5 and 6.
They also wish to thank Mrs. D. Collins and
Miss C. M. Armstrong for technical assistance.
The W.A. Department of Agriculture kindly
helped with glasshouse facilities. The authors
would also like to thank Mr. A. C. Harris, Con-
servator of Forests, and Mr. G. E. Brockway for
encouragement and stimulating discussion.

References

Ahlgren, 1. F. and Ahlgren. C. E. (1960). — Ecological
effects of forest fires. Bot. Rev. 26 : 483-533.

Attiwill, P. M. (1962). — The effect of heat pre-treatment
of soil on the growth of Eucalyptus obliqua
seedling's. 3rd Gen. Conf. Inst. For. Austral.

Barnette, E. M. and Hester. J. B. (1930). — Effect of burn-
ing upon the accumulation of organic matter
in forest soils. Soil Sci. 29 : 281-284.

Beadle. N. C. W. (1954). — Soil phosphate and the de-
limitation of plant communities in Eastern
Australia. Ecology 35 : 370-375.

(1962).— ibid. II. Ecology 43 : 281-288.

Hatch, A. B. ( 1960) .—Ashbed effects in Western Austra-
lian forest soils. Bull. For. Dep. W. Aust.
No. 64.

Loneragan, O. W. (1961). — Jarrah (Eucalyptus viargin-
ata Sm.) and karri (Eucalyptus diversicolor
F.v.M.) Regeneration in South-West West-
ern Australia. M. Sc. Thesis, University of

Western Australia.

Stoate, T N. (1950). — Nutrition of the pine. Bull , For
Bur. Aust. No. 30.

Vlamis, J. Schultz, A. M. and Biswell, H. H. (1955).—
Burning and soil fertility. Calif. Agric. 9 :
7-10.

80

13. — Foraminiferal Evidence for the Paleocene Age of the King’s Park

Shale (Perth Basin, Western Australia)

By Brian McGowran*

Manuscript received — 16tii June, 1964

Several species of fora min if era in the King’s
Park Shale are found also in other Australian
faunas dated firmly as Paleocene. Some, includ-
ing Globorotalia chapmani Parr, Globorotalia
pseudomenardii Bolli, Lamarckian ruguLosa
Plummer, are Paleocene index fossils. There is
no significant evidence for dating the King’s
Park Shale as Eocene.

Introduction

In 1938 the late W. J. Parr described a fora-
miniferal fauna found in samples from two bores
in King’s Park, Perth, Western Australia. Parr
concluded that the fauna was Eocene, and prob-
ably Upper Eocene, in age. Later he prepared
several more samples, mainly from the nearby
Langley Park bore, which were richer in both
numbers and age-diagnostic species than the
original material. However, nothing further had
been published on the fauna when Parr died in
1949. The fauna has been studied subsequently
but no worker has made a serious attempt to
revise Parr’s conclusions in the light of new,
richer material or the recent major advances
in our knowledge of Lower Tertiary biostrati-
graphy. Nevertheless this assemblage and its
dating have been relied upon quite heavily,
particularly by Crespin (see refs.), in studies of
other Lower Tertiary faunas in Australia.

The material prepared by Pari- has been used
by the writer for comparative purposes in his
studies on the Australian Paleocene (unpubl.
thesis, 1962). This paper presents evidence for
redating the King’s Park Shale as Paleocene.
Reference is made continually to the Paleocene
faunas of the Carnarvon Basin. Western Aus-
tralia. and the Otway Basin, Victoria. Papers in
preparation on these faunas will discuss and
figure most of the species invoked here, as well
as some of the repercussions in species identi-
fication arising from the age revision of the
King’s Park Shale. However, Globorotalia
chapmani Parr is discussed and figured below.

Material. — King’s Park Bore No. 1: 10 samples
between 780 ft. and 120 ft.

King’s Park Bore No. 2: 8 samples between
728 ft. and 150 ft.

Langley Park Bore: 7 samples between 975 ft.

and 208 ft.

Claremont Bore: 1 sample, 300-350 ft.

Types of 17 species erected by Parr and other
specimens.

Specimens separated by M. F. Glaessner from
core at 1505 ft.. South Perth Bore (see Glaessner
1956),

For localities, see Coleman, 1952, text-fig. 1.

* University of Adelaide, South Australia.

Acknowledgements

The Parr material was made available and
permission to write this paper was granted by
Prof. R. T. Prider and staff, Dept, of Geology,
University of Western Australia. This coopera-
tion is acknowledged most gratefully. Thanks
are due also to Prof. M. F. Glaessner (Dept, of
Geology, University of Adelaide), who read and
criticized the manuscript, and to Mr. B. E. Balme
(Dept, of Geology, Univ. of Western Australia)
for his comments and for presenting the paper.

Stratigraphic Relationships

The following notes are based on McWhae
et al. (1958) and references given by these
authors.

The sedimentary sequence under consideration,
part of the thick infill of the Perth Basin, has
been named the King’s Park Shale by Fair bridge
(in Coleman, 1952). The formation is known
only from bores in the Perth area, where it
varies, in drilled thickness, from 660 feet in the
King's Park no. 1 bore to more than 1000 feet in
the South Perth bore, less than two miles away.
The sequence consists of grey calcareous shales
and mudstones, pyritic in part. The King’s Park
Shale overlies the South Perth Formation
(uppermost Jurassic to Lower Aptian) and the
Osborne Formation ( ?Aptian, Albian to Ceno-
manian); ana it is succeeded by Quaternary
limestones and clays.

Previous Studies

In arriving at his conclusions, Parr (1938,
p. 69, 90 > relied largely on Cushman's studies on
the North American Eocene and Oligocene
faunas. Cushman published on the Paleocene of
North America largely between 1940 and 1949:
also, important studies on Paleocene faunas by
Brotzen (1948) and others were net available to
Parr. Added to these points in Parr’s defence
are several others: the Midway of Texas (Plum-
mer 1927) was thought of as Lower Eocene
instead of Dano-Paleocene; the biostratigraphic
value of planktonics was becoming appreciated
only then (e.g., Glaessner 1937a); and the
Upper Eocene microfaunas of southern Aus-
tralia were not yet known. Indeed, very little
has been published on the Paleocene even to the
present day. Consequently it is hardly surpris-
ing that Parr's conclusions can be revised by
new approaches to the problem.

Concerning the King’s Park fauna, Crespin
(1950, p. 424) stated that “ . . . there is no indi-
cation with what part of the Eocene it can be
correlated”. Subsequently, however (Crespin

81

1954, 1956a.b, 1958, Raggatt and Crespin 1955 >.
Crespin has emphasized the similarity of various
faunas mostly in South Australia and Victoria to
the fauna described by Parr; and she has
accepted the Upper Eocene age of the latter.
On the other hand, Crespin (1956a > has noted
the similarity in some respects of the King’s
Park fauna to the Paleocene faunas of the Car-
narvon Basin (see also below).

Coleman (1952) studied the foraminifera from
six bores in the Perth metropolitan area. He
noted (p. 37) that “the faunal assemblage as a
whole is unlike that of any other region in Aus-
tralia”; but he did not discuss its age, concen-
trating instead on bore-to-bore correlation. In a
footnote <p. 42). however, Coleman records a
suggestion by Crespin that the King’s Park
assemblages are closely similar to the faunas in
the Swedish Paleocene (Brotzen 1948), and that
“it is probable then that at least part of the
fauna of the Perth Basin ... is of Paleocene
age”. Glaessner (1956) and Belford (unpub-
lished report to Bureau Mineral Resources,
Canberra; referred to by Cookson and Eisenack,
1961) found King’s Park Shale species in samples
from the South Perth and Rottnest Island bores
respectively. Both workers dated their samples
on the basis of Parr’s conclusions.

Churchill (1960) refers in passing to the
Paleocene age of the King’s Park Shale at a
relatively high level in the subsurface (Narrows
bore 164 ft.). Churchill’s evidence presumably
is palynological but to the writer’s knowledge
this evidence has not been published.

The most recent discussion of the age of these
rocks is provided by Cookson and Eisenack
<1961). In a study of micrcplankton and pollen
from the Rottnest Island bore to the west of
Perth, they state (p. 47): “Although the

evidence ... is inconclusive, it shows that the
Rottnest deposits are Eocene and clearly younger
than the Paleocene to Lower Eocene Pebble
Point Formation of Victoria . . . The Upper
Eocene age suggested by Belford for the Rottnest
Bore deposits on the basis of foraminifera is not
incompatible with their pollen and microplank-
ton content, but it is equally possible that they
may have been older than this”. The material
studied by Cookson and Eisenack came from
between 1480 and 1595 feet: there is no reason
to doubt the continuity of this deposit with the
King’s Park Shale on the nearby mainland.

It is clear that several workers regard Parr’s
original age determination as being too young.
Mr. B. E. Balme (Dept. Geol., Univ. W. Aust.) has
kindly informed the writer (pers. comm.) that
a Paleocene age makes interpretation of the
palynological evidence very much easier.

Correlation and Age of the King’s Park Shale

Note on comparison with Upper Eocene
faunas. — Several species occurring in the fauna
are found also in various firmly dated Upper
Eocene sediments However, an inspection of
several Upper Eocene faunas, particularly those
from the Brown’s Creek Clays in western Vic-
toria, has shown that the resemblances are less
striking than Crespin and others have implied.
In the absence of monographic studies, the true

faunal relationships cannot be fully demon-
strated. But we may note a few examples of
alleged occurrences of King’s Park species in the
Upper Eocene (e.g. Carter 1958) which are based
on misidentification (the Eocene species will be
discussed elsewhere) :

Bolivinopsis crespinae Parr.

Alabamina westraliensis (Parr).

C ivllpguem belina rugosa (Parr).

More significantly, several important Upper
Eocene species (see Parr 1947, Carter 1958.
Glaessner and Wade 1959, Ludbrook 1963) are
not found in the King’s Park Shale. Among
others in this category, there are:

Globigerapsis index (Finlay).

Pseudohastigerina micra (Cole).

Hantkenina alabamensis compressa Parr.

Asterigerina adelaidensis (Howchin).

Carpentaria hamiltotiensis Glaessner & Wade.

Maslinella c/iapmani Glaessner & Wade.

Lamarckina airensis Carter.

Comparison of King's Park Shale with .4us-
tralian Paleocene faunas. — The King’s Park
Shale shows considerable overall faunal resem-
blance to the Carnarvon Basin Paleocene. Parr
(1938) identified nearly 70 species in the King’s
Park Shale: Coleman (1952) has found more
than 90. Of about 100 species found so far by
the writer (unpubl. thesis) in the Boongerooda
Greensand (McWhae ct al. 1958, and refs.), more
than 50 have been identified in the King’s Park
Shale material at hand.

About 100 species have been identified also in
two Paleocene faunas from the Wangerrip
Group, Otway Basin, western Victoria (Baker
1953, and refs.). Again, more than 50 occur
also in the King’s Park Shale. This similarity
is greater than with the Boongerooda Greensand
because the two Victorian faunas, from the
Pebble Point Formation and the Rivernook
Member of the Dilwyn Clay respectively (Baker,
1953), are mutually quite distinct.

Correlation of the King’s Park Shale with the
sequence in western Victoria is based largely on
the mutual occurrence of the following ben-
thonic species. That some of them are very
rare is of little importance; it is their presence
that matters, because not one is known authen-
tically from the Australian Upper Eocene.

Vag inn Una longiforma (Plummer).

Citharina subplumoides (Pam.

A n gulag erina sp.

"Bolivinopsis crespinae Parr" (believed to be Spiro -
bolivina. em mendorferi ( Jennings ) ) .

Alabamina westraliensis (Parr) (A, wilcoxensis Toul-
min is considered a junior synonym).

Ceratobulimina westraliensis Parr.

Lamarckina rugulosa Plummer.

Lamarckina aft. L. naheolensis Cushman & Todd

Epistominoides aft. E. midwayensis Plummer.

Anomalinoides westraliensis (Parr) (Cibicides danica
Brotzen is considered a junior synonym).

As well as providing a firm basis for correla-
tion within the Australian region, this list of
species suggests most strongly that the King's
Park Shale is Paleocene in age. With the
exception of Angulogerina sp.. all these species
or closely related forms are found in the Paleo-
cene of the North American Gulf Coast (Olsson
1960, Cushman 1951 and refs.). Similarly,
several occur in the Paleocene of Scandinavia
i Brotzen 1946) and Poland (Brotzen and Poza-
ryska 1957, 1961). Lamarckina rugulosa perhaps

is the most significant Paleocene index fossil in
the above list (see also Pozaryski and Pozaryska
I960). Only a few specimens have been found
in the King’s Park Shale, but L. rugulosa is com-
mon in the Rivernook Member of the Dilwyn
Clay.

Planktonic species and their significance . —
The King’s Park Shale is less rich in plankonics
than the Boongerooda Greensand, which is
•‘Tethyan 1 ’ in its species diversity and large
numbers of specimens. Nevertheless the plank-
tonics are the most conclusive evidence for dat-
ing the formation. The following elements have
been identified:

Globorotalia chapmani Parr <60 specimens).

Globorotalia pseudomenardii Bolll (2 specimens).

Acarinina mckannai (White) and related forms
(relatively common).

Globigerina Hnaperta Finlay species group.

Chiloguembelina crinita (Glaessner) <20 specimens i.

Chiloguembelina trinitatensis (Cushman & Renz)
(2 specimens).

Zcauvigerina aegyptiaca Said & Kenawy (3 speci-
mens).

In 1957 Bolli published a detailed zonation
of the Low r er Tertiary Lizard Springs Forma-
tion of Trinidad, with five zones spanning the
Dano-Paleocene interval. This zone sequence
has been extended to other sections in Europe
and the Americas, unaltered or in somewhat
modified form, and the writer has used it in the
Carnarvon Basin Paleocene. The zone sequence
for the Dano-Paleocene, as modified by Berg-
gren <1964), is as follows:

I Globorolalia velas-
I coensis subzone.

Globorotalia velascoensis Zone | Globorotalia pseu-

I domenardii sub-
1 zone.

Globorotalia pusilla pusilla— Globorotalia angulata
Zone.

Globorotalia uncinata Zone.

Globigerina daubjergensis — Globorotalia trinidaden-
sis Zone.

For the purpose of this discussion, the G.
velascoensis Zone is equated with Upper Paleo-
cene.

It has become clear over the past few years
that Globorotalia chapmani, masquerading as
“Globorotalia elongata Glaessner”, etc. (see p.
85), is restricted in its total range to the upper
part of the Paleocene. The only significant
occurrence of G. chapmani in apparently
younger sediments was in the Nanafalia Forma-
tion at the base of the Wilcox in Alabama,
dated as basal Eocene by Loeblich and Tappan
(1957). The Nanafalia, however, is Paleocene.
in the G. Pseudomenardii subzone (Bramlette
and Sullivan 1961. Gartner and Hay 1962, Berg-
gren 1964). Globorotalia chapmani first appears
in the G. pusilla pusilla-G. angulata Zone, and
occurs most commonly in association with G.
pseudomenardii, as in the Boongerooda Green-
sand. There are records from the G. velascoen-
sis subzone (e.g. Hay I960) though the species
is not found in the highest known Paleocene in
the Carnarvon Basin. Thus its appearance at
depth in the King’s Park Shale (South Perth
1505 ft.. Langley Park 951 ft.) and its range up
to and including the highest samples available
(King’s Park 120 ft., King’s Park No. 2 150 ft ,
Langley Park 208-224 ft.) is decisive for age and
correlation.

The occurrence even of two specimens of
Globorotalia pseudomenardii (Langley Park 400
ft., 208-224 ft.) also is significant. Gartner and
Hay (1962) have emphasized the wide distribu-
tion and restricted range of G. pseudomendardii.
Whether the species always is restricted to the
G. pseudomenardii zone as defined by Bolli
<1957, see also Bolli and Cita 1960, Hay I960) is
not quite clear. It seems probable that like G
chapmani, it persisted with restricted numbers
and distribution into the G. velascoensis sub-
zone. While common in the Boongerooda
Greensand and its deeper water equivalents in
the Carnarvon Basin, G. pseudomenardii is
represented in the highest Paleocene only by a
few small specimens.

A Paleocene age is supported by the remain-
ing planktonic elements of the King’s Park
Shale fauna. These species occur in the Boon-
gerooda Greensand together with additional
species not found in the King’s Park Shale. The
variable species group referred tentatively to
Acarinina mckannai < White) (see von Hille-
brandt, 1962, Gohrbrandt, 1963, and references
therein) includes also morphotypes which have
been identified by some workers under other
names; but all are characterized by reniform
chambers in a low to (particularly) high spired
arrangement. Large samples of well preserved
material are needed to study this variation
adequately. The presence of this group sug-
gests the G. velascoensis Zone ( s . I .) equivalents
without definitely excluding Lower Eocene.
More locally, however, the group is restricted in
the Carnarvon Basin like G. chapmani, being
abundant in the G. pseudomenardii subzone but
not occurring higher. The Middle Eocene index
species Globigerina orbiformis Cole < Porti-
culosphaera mexicana (Cushman) ), as identified
and figured by Parr (1938), belongs to A.
mckannai.

Giimbelina venezuelana Nuttall var. rugosa
Parr 1938 is a junior synonym of Giimbelina
crinita Glaessner 1937. The occurrence of
Chiloguembelina crinita , Ch. trinitatensis Cush-
man & Renz and Zeauvigerina aegyptiaca Said
& Kenawy support the evidence cited above still
further. Beckman (1957) has studied the dis-
tribution of Chiloguembelina etc. in the Trini-
dad Lower Tertiary, with reference to Bolli’s
zone sequence. He has found that the three
species listed are restricted to the Upper Paleo-
cene. Other records (Loeblich and Tappan
1957, Said and Kerdany 1961 > support this.

General Discussion and Conclusions

An analysis of the planktonic foraminiferal
fauna, particularly the occurrence throughout
of Globorotalia chapmani, has shown that the
King’s Park Shale is Paleocene in age. The
same evidence, however, is not good enough to
indicate positively the interval spanned in terms
of the biostratigraphic sequence currently used.
But it is likely that most of the formation was
restricted in time of deposition to the GZoboro-
talia pseudomenardii subzone. The fauna,
particularly the presence of the Acarinina
mckannai species group throughout, suggests
the absence of the G. pusilla pusilla-G. angu-
lata Zone. And apart from the long-ranging
Globigerina Hnaperta species group, the King’s

83

G/oboro£a/ta chapmani Parr. I 4. 8, 9: spiral views; 3, 6, 7; edge views; 2, 5: umbilical view. Adherent matrix

tralia. All specimens from Parr Collection therein. All figs. X132.

84

Park species are not found in G. velascoensis
subzone equivalents in the Carnarvon Basin, but
they are all present ( except perhaps Ch. trinita-
tensis ) in the Boongerooda Greensand and other
G. pseudomenardii subzone equivalents.

The King’s Park Shale can be correlated with
the Otway Basin Paleocene in Victoria on the
basis of several mutually occurring benthonic
species which have been found also in the
Paleocene of Europe and North America. These
species do not have the biostratigraphic pre-
cision of the planktonics at present, but the
fauna as a whole decidedly is Paleocene in
general aspect. Of course, some species extend
into the Eocene. The only definite Lower
Eocene foraminiferal faunas known at present
in Australia occur in the Carnarvon Basin,
where planktonic index species have been found
by Glaessner (McWhae et al. 1958) and the
writer. The King’s Park Shale contains no
species which could suggest a definite correla-
tion with these faunas or with the more wide-
spread and sometimes rich Middle to Upper
Eocene faunas. All the species in common have
unknown but long stratigraphic ranges. Hence
all the negative evidence available, that is, the
absence of Eocene planktonic and significant
benthonic species, supports the positive evidence
embodied in the presence of Paleocene plank-
tonic and benthonic species in the King’s Park
Shale.

We have discussed the Paleocene only with
reference to its zonal subdivision. The use of
stage names has been avoided; this subject is
in a state of flux and is outside the scope of
the present paper. It is sufficient to date the
King’s Park Shale as Upper Paleocene, definitely
younger than Danian and Lower Paleocene.

These conclusions are more definite than those
reached by Cookson and Eisenack (1961).
Cookson and Eisenack compared the Rottnest
bore microplankton and microflora with Cook-
son’s “Microflora C” of the Victorian Lower
Tertiary (Cookson 1954), particularly by the
presence of the pollen species Protecicidites
pachypolus Cookson and Pike. “Microflora C”
is of indefinite Lower Tertiary age. though
definitely younger than the Pebble Point Forma-
tion in Victoria. The Pebble Point Formation
has been regarded as Lower Eocene with
Paleocene affinities by Baker (1953 and refs.)
who relied on somewhat inconclusive studies of
its molluscan and foraminiferal content by
Teichert, Singleton, Glaessner and Parr; and
Cookson and Eisenack have assumed a Paleo-
cene to Lower Eocene age on this basis, not on
the basis of its microflora or microplankton
content. A study of the Pebble Point foramini-
fera by the present writer has suggested a
correlation with the Globorotalia pusilla pusilla-
G. angulata Zone, slightly older than the
Globorotalia pseudomenardii subzone. A Middle
Paleocene age is supported also by the correlat-
ing of the Rivernook Member of the overlying
Dilwyn Clay with the Globorotalia velascoensis
subzone or uppermost G. pseudomenardii sub-
zone, that is, Upper Paleocene. Thus, it appears
on biostratigraphic evidence that “Microflora
C” in part is of Upper Paleocene age. On the
other hand, the interval under consideration in
the Rottnest Is. bore (1480 to 1595 feet) may

be younger than the King’s Park Shale as
developed beneath Perth, but this is most
unlikely unless the beds dip steeply to the west
(indeed, Coleman (1952) states that an easterly
dip is possible) or are faulted.

Notes on Species

Globorotalia chapmani Parr
PL 1, figs. 1-9.

Globorotalia chapmani Parr, 1938, p. 87. pi. 3, fig. 8. 9.

Globorotalia membranacea ( Ehrtnberg ) ; Glaessner
1937b. p. 385, pi. 4. fig. 38; Subbotina, 1953. p. 205. pi.
16. fig. 7-10.

Globorotalia elongata Glaessner; BollI, 1957, p. 77. pi.
20 . fig. 11-13; Loebllch and Tap pan, 1957. p. 189. many
figs.

Globorotalia pseudomenardii Bolll; Nakkadv, 1959 n
462; pi. 4. fig. 3.

Globorotalia troelscni Loebllch & Tappan; Gohrbancii,
1963, p. 51. pi. 6. fig. 13-15 (synonymy)

Parr’s species has been overlooked consist-
ently by recent students of Lower Tertiary
faunas, presumably because of the misleading
title of his paper. Parr’s description of Globoro-
talia chapmani is given in full;

“Test biconvex, oval, the dorsal surface more
convex than the ventral, which is umbilicate;
periphery lobulated, peripheral margin rounded;
chambers comparatively few, not more than
five in the last-formed whorl, each much larger
than its predecessor; sutures depressed, not
limbate, gently recurved on both sides of the
test; wall smooth and punctate; with a silvery
lustre; aperture an elongate slit with a slight
lip, opening at the base of the last-formed
chamber into the umbilical depression. Length
up to 0.65 mm”.

Notes on Morphology . — Globorotalia chap-
vnani is characterized by an imperforate mar-
ginal band which gives the impression of being-
a keel, but oriented thin sections show that the
wall is not thickened in any way in this region.

The test varies somewhat in plan, from almost
equilateral to strongly elongate due to variable
increase in chamber size. The chambers of G.
chapmani vary in inflation and the sharpness
of the periphery, but typically they have an
equally biconvex arrowhead shape in profile,
and the early chambers usually are depressed
below the outer whorl. In some individuals this
is not so; the inner whorls are raised giving a
broadly convex spiral surface. On both sides
the sutures are distinctly depressed.

Remarks. — A metatype of Globorotalia pseu-
doscitula elongata Glaessner (collection of M. F
Glaessner) is very similar to juveniles of
Globorotalia pseudomenardii Bolli. Russian
workers (e.g. Shutskaya 1956) no longer dis-
tinguish the variety elongata formally from
pseudoscitula. This species is quite distinct
from “ Globorotalia elongata ’’ of recent workers
(see also Gohrbandt, Z. c.).

“ Globorotalia elongata ”, as figured by Loeblich
and Tappan, Bolli, and others, is quite variable
in profile due to variation in compression of
chambers and angularity of margin. Specimens
from the Salt Mountain Limestone, Alabama-
examined through the kindness of Prof. L. D.
Toulmin (Tallahassee, Florida) show clearly
that this form is G. chapmani.

85

Specimens of Globigerina compressa Plummer
(Upper Midway, Lockhart, Texas, coll, H. J.
Plummer, in collection of M. F. Glaessner)
have been examined. They are smaller than
Globorotalia chapmani . The chambers become
distinctly compressed in the adult stage but they
are more rounded in profile than in chapmani,
and the margin lacks the imperforate zone
(see also Hofker. Nat. Maandblad, 47:3-4, 1958).

Crespin <1954; Raggatt and Crespin 1955)
has recorded Globorotalia chapmani from the
Upper Eocene of Southern Australia. In the
absence of description or figures, it is not known
to which species she was referring. G. chapmani
has not been seen in any of the material at
hand from the same formations, and Carter
(1958) did not mention its presence in Victoria.
Possibly Crespin was referring to Pseudohasti-
gerina.

Occurrence. — In several samples; 60 specimens
examined.

References

Baker. G. (1953). — The relationship of Cyclammina-
bearing sediments to the older Tertiary de-
posits south-east of Princetown, Victoria.
Mem. Nat. Mus. Mclb. 18 : 125-134.

Beckmann. J. P. (1957). — Chiloguembelina Loeblich and
Tappan and related foraminifera from the
Lower Tertiary of Trinidad, B.W.I. Bull.
U.S. Nat. Mus. 215 : 83-96.

Berggren, W. A. (1964). — The recognition of the Globo-
rotalia uncinata Zone (Lower Paleocene.) in
the Gulf Coast. Science (In Press).

Bolli. H. M. (1957). — The genera Globigerina and Globo-
rotalia in the Paleocene-Lower Eocene Lizard
Springs Formation of Trinidad, B.W.I. Bull.
U.S. Nat. Mus. 215 : 61-82.

Bolli, H. M. and Cita, M. B. (1960). — Upper Cretaceous
and Lower Tertiary planktonic foraminifera
from the Paderno d’Adda section, northern
Italy Rep. XXI Int. Geol. Congr. 5 : 150-161.

Bramlette, M. N. and Sullivan. F. R. (1961). — Coccoli-
thophorids and related Nannoplankton ol
the early Tertiary in California. Micropa-
leont. 7 : 129-188.

Brotzen, F, (1948). — The Swedish Paleocene and its
foraminiferal fauna. Sverig. Geol. Unders.
Afh. 493 : 1-140.

Brotzen. F. and Pozaryska, K. (1957)— The Paleocene in
Central Poland. Acta Geol. Polon. 7 : 273-277.

(1961). — Foraminiferes du Paleocene et de

T Eocene inferieur en Pologne septentrionale.
Remarques paleogeographiques. Rev. Micro-
paleont. 4 : 155-166.

Carter. A. N. (1958). — Tertiary foraminifera from the
Aire District, Victoria. Bull. Geol. Surv. Viet.
55 : 1-76.

Churchill, D. M. (1960). — Living and fossil unicellular
Algae and Aplanospores. Nature, Lond. 186 :
493-494.

Coleman, P. J. (1952). — Foraminiferal investigations in
the Perth Basin, Western Australia. J. Roy.
Soc. W. Aust. 36 : 31-43.

Cookson, I. C. and Eisenack, A. ( 1961 ) —Tertiary micro-
plankton from the Rottnest Island bore,
Western Australia. J. Roy. Soc. W. Aust.
44 : 39-47.

Crespin, I. (1950). — Australian Tertiary microfaunas and
their relationships to assemblages elsewhere
in the Pacific region. J. Paleont. 24 : 421-429.

(1954). — Stratigraphy and micropalaeon-
tology of the marine Tertiary rocks between
Adelaide and Aldinga, South Australia. Rep.
Bur. Miner. Resour. Aust. 12 : 1-65.

f 1956a).— Migration of foraminifera in Ter-
tiary times in Australia. Ibid. 25 : 1-15.

(1956b). — Changes in ideas of age of certain

beds in the Australian Tertiaries. Ibid. 25 ■
16-25.

(1958).— The occurrence of Hantkenina in

Western Australia. Micropaleont, 4 : 317-319,

Cushman, J. A. (1951). — Paleocene foraminifera of the
Gulf Coastal region of the United States and
adjacent areas. Prof. Pap. U.S. Geol. Surv
232 : 1-75.

Gartner, S. and Hay, W. W. (1962).— Planktonic fora-
minifera from the type Ilerdian. Eel. Geol
Helv. 55 : 553-572.

Glaessner, M. F. (1937a). — Planktonforaminiferen ausder
Kreide und dem Eozan unci ihre strati-
graphische Bedeutung. Stud. Micropaleont.
Moscow 1 : 27-52.

(1937).— Studien fiber Foraminiferen aus der

Kreide und dem Tertiar des Kaukasus. I.
Die Foraminiferen der altesten Tertiarschi-
chten des Nordwestkaukasus. Probl. Paleont
Moscoiv 2-3 : 349-410.

(1956). — Crustacea from the Cretaceous and

Eocene of Western Australia. J. Roy. Soc. W
Aust. 40 : 33-35.

Glaessner, M. F. and Wade, M. (1959).— Revision of the
foraminiferal family Victoriellidae. Micro-
paleont. 5 : 193-212

Gohrbandt, K. (1963). — Zur Gliederung des Palaogen im
Helvetikum nordlich Salzburg nach plank-
tonischen Foraminiferen. Mitt. Geol. Ges.
Wien 56 : 1-116.

Hay. W. W. (1960). — The Cretaceous-Tertiary boundary-
in the Tampico Embayment. Mexico. Rep.
XXI Int. Geol. Congr. 5 : 70-77.

Hillebrandt, A. von (1962).- — Das Paleozan und seine
Foraminiferenfauna im Becken von Reichen-
hall und Salzburg. Ach. Bayer. Akad. Wiss.
Math.— Nat. Kl. (N.F.) 108 : 1-182.

Loeblich, A. R. and Tappan, H. (1957 (.—Planktonic
Foraminifera of Paleocene and early Eocene
age from the Gulf and Atlantic Coastal
Plains. Bull. U.S. Nat. Mus. 215 : 173-198.

Ludbrook, N. H. (1963). — Correlation of the Tertiary
rocks of South Australia. Trans. Roy. Soc.
S. Aust. 87 : 5-15.

McGowran, B. (1962). — Foraminifera and biostratigraphy
of the Australian Maastrichtian and Paleo-
cene. Unpubl. Thesis Univ. Adelaide: 1-391.

McWhae, J. R. H., Playford, P. E., Lindner, A. W., Glen-
ister. B. F. and Balme, B. E. (19581.— The
stratigraphy of Western Australia. J. Geol.
Soc. Aust. 4 (2) ; 1-161.

Nakkady, S. E. (1959).— Biostratigraphy of the Urn Elg-
hanayem section, Egypt. Micropaleont. 5 :
453 -4 (2 .

Olsson, R. K. (I960). — Foraminifera of latest Cretaceous
and earliest Tertiary age in the New Jersey
coastal plain. J. Paleont. 34 : 1-58.

Parr, W. J. (1938). — Upper Eocene foraminifera from
deep borings in King’s Park, Perth. 'Western
Australia. J. Roy. Soc. IV. Aust. 24 : 69-101.

(1947). — An Australian record of the fora-
miniferal genus Hantkenina. Proc. Roy. Soc.
Viet, (n.s.) 58 : 45-47.

Plummer. H. J. (1927). — Foraminifera of the Midway
Formation in Texas. Bull. Univ. Tex. Bur.
Econ. Geol. Tech. 2644 : 1-206.

Pozaryski, W. and Pozaryska, K. (I960).— On the Danian
and Lower Paleocene sediments in Poland.
Rep. XXI Int. Geol. Congr. 5 : 170-180.

Raggatt. H. G. and Crespin, I. (1955).— Stratigraphy of
Tertiary rocks between Torquay and Eastern
Views. Victoria. Proc. Roy. Soc. Viet, (n.s.)
67 : 75-142.

Said, R. and Kerdany, M. T. (1961).— The geology and
micropaleontology of the Farafra ~ Oasis,
Egypt. Micropaleont. 7 : 317-336.

Shutskaya, E. K. (1956). — Stratigraphy of the lower
horizons of the Paleogene of the central pre-
Caucasus. based on foraminifera). Trav.
Inst. Sci. gdol. Akad. Nauk S.S.S.R. 164 :
1-119 (in Russian).

Subbotina, N, N. (1953). — (Fossil foraminifera of the
U.S.S.R. : Globigerinidae, Hantkeninidae and
Globorotaliidae i. Vses. Nejt. Nauk-Issl
Geol. Trudy 76 : 3-29 (in Russian).

86

14. — Projasus — a New Generic Name for Parker’s Crayfish, Jasus parkeri

Stebbing (Palinuridae : “Silentes”)

By R. W. George* and J. R. Grindleyt
Manuscript received — 19th May, 1964

The family Palinuridae is divisible into the
Stridentes (producing sound by antennal move-
ment) and the Silentes.

Jasus parkeri Stebbing, belongs to the Silentes
and because of its uniqueness warrants the
erection of a new genus. Projasus gen. nov. is
intermediate between the other two genera of
Silentes, Jasus and Palinurellus.

A key to the genera of Silentes is included.

Introduction

Stebbing (1902) originally placed his new
South African species, parkeri , in the genus
Jasus but commented “there is no resemblance
so far as I know to any other living Palinurid
hitherto described”. On the basis of Stebbing’s
description and figure alone. Holthuis < 1946 :
148) considered the species “no Jasus at all”
and allocated it to the genus Puerulus com-
menting . . it has for instance, a distinct
stridulating organ, which is entirely absent in
the genus Jasus”. Barnard 1950 : 541) replaced
parkeri in the genus Jasus stating that the
stridulating organ is not present, but also
recognised that ”... there is a close resemblance
to Panulirus t Puerulus] angulatus Bate . .

Parker, in his excellent papers (1883 and
1884) on the classification of palinurids, divided
them into two categories the “Silentes” and the
“Stridentes”. The ability to produce sound by
rubbing a process of the antenna over a raised
ridge on the antennular plate has been used
extensively in keys and descriptions, yet few
workers have adopted Parker’s terminology for
the two distinct groups. These terms, “Silentes”
for the non-stridulating, and “Stridentes” for
the stridulating group are very apt and their
use is to be encouraged.

This paper presents the results of re-examin-
ation of Stebbing’s and other specimens of
parkeri from the Souih African Museum and
concludes that the species parkeri does not
belong to either Jasus or Puerulus but to a new
genus of the Silentes, Projasus gen. nov.

We are very appreciative of Dr. Holthuis’
assistance and comments in the final stages of
preparation of the paper.

Projasus gen. nov.

Type. Jasus parkeri Stebbing, 1902.

Diagnosis . — This genus consists of one species
parkeri and it differs from all other Palinuridae
in possessing four longitudinal rows of spines,
separated by smooth regions on the carapace. It
differs from the other genera of Silentes in com-
binations of characters as discussed below.

* Western Australian Museum, Beaufort St , Perth, W.A.
t South African Museum, Cape Town, South Africa.

Description . —

i Antenna articulated simply; stridulating
organ absent.

ii Carapace with angular sides and with
longitudinal spiny ridges, otherwise
smooth.

iii Abdomen smooth, first five segments with
median carina.

iv Supraorbital processes pointed, unarmed
and obliquely erect.

v Eyes produced laterally on eyestalks, not
recessed.

vi Median rostrum and clasping processes
very small.

vii First abdominal segment of female with-
out pleopod.

Remarks. — Parker (1883 and 1884) recognised
only one species group in the Silentes and for
this group he proposed the sub-genus Jasus.
Although he did not tabulate Palinurellus Von
Martens 1878 in his Silentes group (1884, p. 304),
he clearly recognised its lack of a stridulating
organ and suggested that it was close to the
“parent species” of the family Palinuridae ( 1884,
p. 303). Palinurellus is now recognised (see
Holthuis 1946) as an actual member of the
Palinuridae.

Parker’s original concept of Jasus was based
on the species lalandii H. Milne Edwards, 1837;
edwardsii Hutton, 1875 and hugelii Heller, 1862
(this species is now regarded as a synonym of
verreauxii H. Milne Edwards, 1851; see Holthuis
1946). These non-stridulating species form a
natural group having in common a cylindrical,
uniformly spiny carapace and abdominal terga
without median carina, either smooth f ver-
reauxii J or with squamiform sculpture (lalandii
group). Projasus gen. nov. is therefore clearly
outside the boundaries of the Jasus species group
as envisaged by Parker (cf. ii, iii. above).

Palinurellus bears a general resemblance to
Projasus in the possession of median, but less
pronounced, carinae on the abdominal terga.
However, Palinurellus has no spines on the cara-
pace, and no supraorbital processes, the eyes
are simply recessed at the sides of a large
rostrum which lacks clasping processes; the first
abdominal segment of the female possesses a
pleopod (Holthuis pers. comm., 1964) (cf, iv, v,
vi, vii above).

Both Holthuis (1946) and Barnard (1950)
noted the close general resemblance of parkeri
to Puerulus (one of the Stridentes), presumably
based on the angular shape of the carapace.
The absence of a stridulating organ however
easily separates Projasus parkeri from all
Stridentes and places it in the Silentes group.

87

Fig. 1. — A, B, C. Projasus parkeri; A, female pleopod of second abdominal segment B, female pleopod of third
abdominal segment C, anterior region of carapace. D, Jasus lalandii anterior region of carapace. E. F. Palinurus
gilchristi; E, anterior region of carapace F, underside of stridulating pad. X-Y indicates axis of antennal

articulation.

88

We regard Projasus as a genus of Silentes
intermediate between Palinurellus and Jasus.
It differs as much from the other genera of
Silentes, as Puerulus does from the other genera
of Stridentes.

Key to the Genera of Silentes Palinurids

( l > Supraorbital processes present ;
median rostrum and clasping pro-
cesses small; carapace with spines 2
Supraorbital processes absent; median
rostrum large; clasping processes
absent; carapace with low tubercles

(2) Carapace with rounded sides and
covered with many subequal spines;
abdomen smooth or squamiform
Carapace with angular sides and with
longitudinal spiny ridges; abdomen
with median carina

Palinurellus

Jasus

Projasus

This key is based on our examination
of all taxa of Silentes recognised by Hol-
thuis (1946 and 1963) except Palinurellus
ffundlachi and Palinurellus gundlachi var.
xcienecki.

Fig. 2 . — Projasus parkeri female, carapace length 58 mm,
off Buffalo River. South African Museum No. A995, now
housed in Western Australian Museum.

Projasus parkeri (Stebbing.)

(Figs. 1 A, B, C, and 2.)

Jasus parkeri Stebbing 1902 p. 39, pi. 7 and 1910 p.
375; Barnard 1950 p. 540.

Puerulus parkeri ; Holthuis 1946 pp. 110 and 148.

Type Locality . — . . shrimp trawl, Buffalo
River north 15 miles. Depth, 310 fathoms.
Bottom, coral and mud”. Hat. 33°S, long. 28"E.J

Description . — The original description and
figure given by Stebbing (1902) require little
amplification. Only two features are redes-
cribed here — the female pleopods and various
structures in the anterior region of the carapace.

In his key to the genera of Palinuridae, Hol-
thuis (1946 p. 113) indicated the value of the
shape of the various female pleopods as taxono-
mic characters and illustrated fifteen pleopods
of several species but parkeri was not among
these. Stebbing (1902) and Barnard (1950)
gave only brief descriptions of the female
pleopods but did not illustrate them.

Because of the lack of detail in the anterior
region of the carapace in Stebbing’s (1902)
original illustration of parkeri , Holthuis (1946)
interpreted the exaggerated ridges of the anten-
nular plate as part of a stridulating mechanism.
As a result he incorrectly allocated it to the
genus Puerulus .

Female pleopods (fig. 1A and B». — There are
no appendages on the first abdominal segment.
The endopod of the pleopod of the second seg-
ment is almost as large as the exopod and bears
a reduced stylamblys (fig. 1A>. On the third,
as on the fourth and fifth abdominal segments,
the endopod of the pleopod is not foliate and
the stylamblys is present (fig. IB).

Anterior region of carapace, fig. 1C. — The
rostrum is small, acute, upturned and embraced
by two small, pointed clasping processes. The
narrow antennular plate runs forward and
downward and its exposed sides are longitudin-
ally convex. The articulating process of each
antenna abutts the latero-distal region of the
antennular plate and merely pivots on it when
the antenna is raised.

Comparison of Silentes and Stridentes

In the Silentes, the sides of the antennular
plate are not well developed or polished and
the axis of articulation of the antenna passes
through the point of articulation of the anten-
nal process with the antennular plate. Due to-
the simple rotation of the antennal process, no
sound is produced when the antenna is raised.
The axis of articulation of the antenna is indi-
cated by the oblique broken lines X-Y for P.
parkeri in fig. 1C and for Jasus lalandii in fig.
ID.

On the other hand, the Stridentes have highly
polished areas on the sides of the antennular
plate and well developed lateral processes on
the antennae. These processes have a ribbed
pad on their under surface and the axis of
articulation is nearly transverse. When the
antenna is raised, the ribbed stridulating pad
moves over the polished area on the side of
the antennular plate producing the characteris-
tic stridulation of the Stridentes group. In

89

figs. IE and F, the anterior region and the
underside of the stridulating pad of Palinurus
gilchristi is illustrated. The genus Panulirus
with its very broad antennular plate probably
represents the greatest development of the
stridulating apparatus.

The Silentes however, are not without means
of sound production. A. Heydorn of the Divi-
sion of Sea Fisheries, South Africa, reports that
Jasns lalandii makes loud “clicking” noises
under water without movement of the antennae
and believes that the sound is probably pro-
duced by the mandibles.

References

Barnard, K. H. (1950). — Descriptive catalogue of South
African Decapod Crustacea. Ann. S. Air.
Mus. 38 : 1-864.

Holthuis, L. B. (1946). — The Decapoda Macrura of the
Snellius Expedition 1. Temminckia 1 : 1-178.

(1963). — Preliminary descriptions of some

new species of Palinuridea. Proc. Acad. Sci.
Amst. (C) 66 : 54-60.

Parker. T. J. (1883). — On the structure of the head in
“Palinurus" with special reference to the
classification of the genus. Nature. Lond.
29 : 189-190.

(1884). — On the structure of the head in

Palinurus with special reference to the
classification of the genus. Trans. Proc. N.Z
Inst. 16 : 297-307.

Stebbing, T. R. R. (1902). — South African Crustacea.
Mar. Invest. S. Afr. 2 : 1-92.

(1910). — General Catalogue of South African

Crustacea. Ann. S. Afri. Mus. 6 : 281-599.

Memorial

Douglas Elwood White

Douglas Elwood White — M.Sc., Ph.D.,
F.R.A.C.I,. died as the result of a car accident
on the evening of 4th April, 1964, at the early
age of 53 years.

His long association with the University of
'Western Australia where he was Professor of
Organic Chemistry at the time of his death,
was largely responsible for the expansion of
organic chemistry in this State during the post-
war years.

He graduated from the University of Sydney
with first class honours in 1930 and completed
his M.Sc. in 1932. He later went on to Oxford
where he worked in the Dyson Perrins Labora-
tories and was awarded the degree of Doctor of
Philosophy in 1937.

White first lectured in Organic Chemistry at
the University of Sydney and was then appointed
biochemist to its School of Public Health and
Tropical Medicine. In 1941 he became lecturer
in Organic Chemistry at Melbourne University
and in 1943 joined the University of Western
Australia as Lecturer in Charge of the same
discipline. He was appointed Reader and Head

of this Department in 1945 and in 1958 became
the University’s first Professor of Organic
Chemistry.

His early research on essential oils developed
into the broader field of plant chemistry parti-
cularly in relation to the flora of this State.
His varied contributions to science covered work
on tannins, coumarins, alkaloids, terpenoids and
the oestrogenic isoflavones of sub-terranean
clover.

A period of study leave in 1950-51 was spent
at the Technical High School, Zurich, where
White developed a specialised interest in poly-
terpenes on which he later became an authority.

Despite his great interest in research, he was
an able administrator and served as the Dean
of the Faculty of Science from 1957 to 1959.
He gave freely of his time to a number of
scientific societies and was chairman of the
organising committee of ANZAAS for the Perth
Congress in 1959. In his long association with
the Royal Society of Western Australia he acted
on the Editorial Committee from 1954 to 1957,
was a Council member in 1953-54, Vice President
in 1954-55 and President in 1955-56.

90

15. — Geographical Variation and Distribution of some Birds from Western

Australia

By G. F. Mees*

Manuscript received — 17th March, 1964

Revisional and distributional data are given
on ten species of passerine birds from Western
Australia.

Introduction

Over the past few years I have collected
evidence on the validity of certain subspecies of
birds described from Western Australia, together
with other information. Some of these findings
alter or are complementary to recent revisions
(Mayr 1948, 1954, Mees 1961), and therefore
warrant publication. All material discussed is in
the collection of the Western Australian
Museum.

Coraeina novaehollandiae (Gmelin )
Earlier I stated that no evidence of migratory
movements in Western Australia existed (Mees
1961, p. Ill), and that no individuals of Cm.
melanops were known from within the range of
Cm. subpallida.

However, in July 1959 I was surprised to find
flocks comprising dozens of birds in the scrub-
country south of Carnarvon; there were indi-
viduals with dark and with pale mantles, clearly
both melanops and subpallida. In the first
fortnight of August 1959, the species was also
common at North West Cape, though not in such
great numbers. The concentration of hundreds
of birds near Carnarvon was far above anything
one would expect of a local breeding population
or even of local movements, and I am convinced
that true migration was involved. On 14 May.
1960 I especially noted that in the same area
not a single individual was observed. In
December 1962 I saw only an occasional pair of
what seemed to be subpallida , According to
Peters Sz Mayr (in Mayr & Greenway 1962, p.
171), subpallida winters in the Lesser Sunda
Islands and Kei Islands. It may be that the
subspecies does partially migrate to these
islands, but I have not seen specimens, and the
majority of individuals seems to stay in Aus-
tralia.

Material collected since my previous paper
was written confirms the occurrence in winter
of both subspecies in the north-west of the
State. I give here a full list of specimens and
localities.

Cm. subpallida. 9 , 5-VIII-1959, Milyering Well,
Yardie Creek Station, North West Cape (A 8636) ;
9, 20-V-1960, Cossack (A 8526); & imm., 20-V-
1960, Cossack (A 8524); $ 20-V-1960, Cossack
(A 8525); 9 8-IX-1959, Edmund Station, Barlee
Range (A 8649).

* Formerly Western Australian Museum, Perth, Western
Australia. Present address, Rijksmuseum van Natuur-
lijke Historie, Leiden, Holland.

C.n. melanops. £ imm., 5- VIII- 1959, Milyering
Well. Yardie Creek Station, North West Cape
< A 8924); 9 imm., 7-VIII-1959, Milyering Well,
Yardie Creek Station, North West Cape < A 8925 ) ;
9 11- VII- 1959, Murchison River Bridge, north
of Geraldton (A 8953); 6 , ll-VI-1962. Legendre
Island, Dampier Archipelago (A 8835); & 22-1-
1962, Beagle Bay Road, west Kimberley Division
(A 895D; £ 21-VI-1960, Kalumburu, north

Kimberley Division (A 8622).

Note especially the specimens of melanops
from Yardie Creek and Legendre Island, in the
range of subpallida. The specimen from the
Murchison River is rather pale for melanops
and may be considered somewhat intermediate;
this is what one would expect from a breeding
bird of that locality. Keast (1961. p. 413) refers
to the form subpallida as an isolate, but I
regard it is unlikely that it is.

Previously I noted that the testes of C. novae-
hollandiae are dark slate in colour. I have since
collected Coraeina papuensis hypoleuca in the
Kimberley Division (Kalumburu and Beverley
Springs) in which I found the same feature.

Acrocephalus stentoreus gouldi Dubois

Mayr (1948), in his revision of the Australian
reed -warbler, accepted as valid the name car ter ae
Mathews, though with grave misgivings: “ . . .
the type of carterae has a relatively and actually
much larger bill than gouldi. I would refrain
from naming a new form on such slight basis
and with only a single specimen available, but
now that the name is in the literature, it must
be recognized”. A specimen collected by me
near Derby on 6 June, 1960, almost topotypical
of carterae , has a bill of average length for
gouldi , from individuals of which race I am
unable to distinguish it. Moreover, in June 1962,
during a visit to the American Museum of
Natural History, I compared the type of carter ce
with specimens from south-west Australia, and
found a south-west bird with a bill as long as
the bill of carterae. Therefore, carterae is doubt-
less a synonym of gouldi. At the time of my
visit to Derby, the small pool where I collected
a specimen harboured several individuals, but
my friend Mr. P. Slater, at the time a resident
of Derby, who accompanied me, mentioned that
he had not previously observed reed-warblers in
that locality. I regard it as possible that the
birds observed were migrants from the south.
As yet there is no proof of the existence of an
indigenous breeding population in the Kimberley
Division.

The validity of gouldi itself is in need of con-
firmation, as I have pointed out in a previous
paper (Mees 1961, p. 113).

91

Pachycephala pectoralis 'Latham)

According to the latest published information,
in south-western Australia this species extends
from the mouth of the Murchison River to the
Esperance area, and inland to Bunketch
(observed once), Kellerberrin (rare), Lake Grace
and Mt. Ridley (Reid 1951, Ford & Stone 1957,
Serventy & Whittell 1962).

In the first half of 1963 I found it in several
localities well inland from the range indicated,
and it appears that the species is much more
widely distributed than was hitherto known.
The observations are as follows (Fig. 1): Mt.

Hampton, 10 March, one singing male; Mt. Hol-
land, 15 April, two singing males; Parker Range,
15 April, fairly common; Moorine Rock, 16 April,
several individuals of both sexes; Carrabin, 16
April, a pair; Great Eastern Highway near the
301 mile post (47 miles west of Coolgardie), 12
June, at least two singing males; Mordetta, 10
July, several individuals; eleven miles north of
Pingaring, 10 July, one singing male; Jitarning,
21 and 22 July, several individuals.

While none of these observations were made
in the breeding season, nothing is known of
migration, and I regard it as likely that the

species is a resident in the localities listed. It
is surprising that this common and familiar
species has been overlooked in such an extensive
part of its range.

Pachycephala lanioides Gould

In his review of the Australian members of the
genus Pachycephala, Mayr (1954) recognized
four subspecies of P. lanioides: bulleri, carnar-
voni (consistently misspelt carnavoni by Mayr),
lanioides and fretorum.

Mayr did not have much material at his dis-
posal; he was also much impressed by the patchy
occurrence of mangroves along the Australian
coast. Earlier Mack (1933) had been even more
handicapped by lack of material. Over the past
few years many additional specimens have been
collected by myself and other staff members of
the Western Australian Museum.

Before discussing the variation of the species.
I will say something about mangrove birds in
general. It is true, as Mayr states, that the
occurrence of mangroves is patchy. But does
this necessarily mean that Mayr (p. 10) is also
right in calling the populations inhabiting such
mangrove patches “true isolates”? In my
opinion the problem of the mangrove birds, and
of other birds inhabiting a specialised habitat
which is patchy in occurrence, is to find and
colonize all the available habitat, because per-
haps no single patch is large enough to ensure
the continued existence of a population for any
great length of time. In other words, birds of
this group, to be successful, must have fairly
good dispersal faculties. This is in conflict with
a natural tendency to become sedentary in any
good patch of habitat. Probably both these
opposite selective factors are at work, resulting
in a species with moderate mobility. On the
basis of this the geographical variation of P.
lanioides can be explained satisfactorily. On

OCoolgardie
• 301 mile post

Southern Crosso
Carrabin, •Moorine Rock

°Merredin • Parkers Range
Kellerberrin# #Mt. Hampton

Mt Holland •
O Hyden

O Norseman

Mt. Ridley

TPjg. i. — The inland distribution of Pachycephala pectoralis (dots) in Western Australia. Towns indicated

open circles are put in for orientation.

92

by

the map (Fig. 2) I have indicated the distribu-
tion of P. lanioides as ascertained from material
examined and from Mayr’s records, and also the
two major gaps in the mangrove, as ascertained
during visits to the north and north-west of the
state in 1959, 1960 and 1962: a gap of about
220 miles between Carnarvon and Tantabiddy
(North West Cape), and one of about 140 miles
between Pardoo and La Grange.

The new material confirms the conclusions as
regards geographic variation drawn by Mayr,
except that I recognise not four, but three races,
which from south-west to north-east are:

1. Pachycephala lanioides carnarvoni
(Mathews 1913).

Pachycephala lanioides bulleri Mayr 1954.

Differs strikingly from the other races in that
the female and the male in immature plumage
are essentially brown in colour, not grey. I fail
to find any differences between birds from
Carnarvon, Exmouth Gulf and as far up the
coast as Port Hedland (included in bulleri by
Mayr ) . Most interesting is a single adult female
from Pardoo, which is greyish brown in colour,
almost exactly intermediate between carnarvoni
and the nominate race. This is what one would
expect from its intermediate geographical posi-
tion. I have not examined topotypical material

of “ bulleri ” from De Grey, but the mangroves
between there and Exmouth Gulf are more or
less continuous, show at least no major gaps,
and in Mayr’s description there is nothing in-
consistent with the view that this population
differs from carnarvoni only in a slight trend
towards the nominate race, a trend more pro-
nounced in the Pardoo bird. As in my opinion
there is no point in subspecifically naming inter-
mediates where smooth gradients exist, and as
naming is in such cases actually misleading.
bulleri has to be synonymized with carnarvoni,
to which it is closest. I may add that I do not
quite understand why under the subspecies
bulleri, Mayr remarks: “This ecologically

specialized bird is undoubtedly confined to
isolated pockets of mangrove”. The remark is
very much to the point, but not especially for
this race, but for the whole species.

Though, as Mayr pointed out, Mathews (1930,
1931) had synonymized his name carnarvoni, it
should be noted that A. J. Campbell (1918),
who was an infinitely better ornithologist than
Mathews, had previously mentioned its diag-
nostic characters.

Measurements. — Males: 97, 97.5, 100, 100, 101,
101, 101, 102, 102, 103. Males in female plumage:
96, 96, 99, 100. Females: 93. 95, 96, 97, 98, 99.
101 .

93

Distribution. — The coast of Western Australia
from Carnarvon to De Grey; intermediates
towards the following race at Pardoo.

2. Pachycephala lanioides lanioides Gould
1840. Males are similar to those of the preced-
ing subspecies, females differ by being grey, not
brown, above, and much paler with hardly any
buff below.

Measurements. — Male, 102. Male in female
plumage; 96. Female: 97.

Distribution. — Only known from the West-
Kirn berley Division, from Broome to Point Tor-
ment, but almost certainly ranging farther north.

3. Pachycephala lanioides jretorum De Vis
1889. Similar to the nominate race but slightly
smaller.

The four specimens taken by me at Wyndham
confirm Mayr’s conclusion as regards the applic-
ability of the name fretorum to the slightly
smaller eastern birds. At the same time it must
be remarked that my material is in rather worn
plumage and that the difference in size is slight.
Specimens from Wyndham agree in wing size
with material from the Roper River, the eastern-
most locality whence the species is certainly
known. They also agree with the measurements
published by De Vis ( 4 : 93. 96; $: 92 >.

Measurements. — Males; 95, 99. Females: 93.
94. The following measurements of material
from Roper River in the H. L. White Collection
were supplied by Mr. McEvey. Males: 94, 95. 97.
Immature male: 96. Females: 92, 94.

Distribution. — Northern Australia from the
Cambridge Gulf to the Roper River; apparently
Kimberley, Gulf of Carpentaria.

Discussion. — Both Mack (1933) and Mayr
(1954) have commented on the inadequacy and
ambiguity of De Vis’ <1889) description, I have
studied it carefully, and in my opinion it makes
sense only if in its first paragraph the word
“ 1101 ” is added, so that it comes to read: “Among
the birds brought from Cambridge Gulf is a
young female Pachycephala which is NOT
identical with an adult of the same sex pro-
cured at Kimberley on the Gulf of Carpentaria
in company with two males. These which have
hitherto been supposed to be P. lanoides, Gld.,
must now be considered to constitute a distinct
species. The writer proposes for it the name
P. fretorum

When this is done the description becomes
understandable; the receipt of the specimen
from Wyndliam, which De Vis considered as
typical P. lanioides (misspelt lanoides by him),
was the first of the species from a western
locality available to him, and enabled him to
differentiate his three eastern birds from true
lanioides. For this reason also, descriptions are
given of the three Gulf of Carpentaria birds:
adult male, immature male, adult female, but
not of the juvenile female from Cambridge Gulf.
Similarly, in the list of measurements these two
males and one female are listed as P. fretorum.
while for comparison measurements of one
specimen of P. lanioides are given; the sex of
this specimen has not been recorded again, as
it had already been said to be a young female.
The type locality, southern shores of Torres

Straits, which would be more or less correct for
the birds from Kimberley, Gulf of Carpentaria,
but not for Cambridge Gulf, supports my views.
So does the fact that Saville-Kent (1889, p, 220)
lists his specimen from Cambridge Gulf as
Pachycephala lanoides. which name (with the
incorrect spelling!) must have been given to
him by De Vis.

On the other hand I must admit that in the
list of birds collected at Cambridge Gulf (p. 236)
De Vis lists this same individual as Pachycephala
fretorum. This, however, can be ascribed to his
well-known carelessness.

From the preceding discussion it will be clear
that I do not regard the specimen from Cam-
bridge Gulf as a cotype of P. fretorum. Kim-
berley. Gulf of Carpentaria, will have to be
regarded as the type-locality, unless it is
definitely proven that the species does not occur
there. As mentioned by Mayr. the only specimen
of the four discussed by De Vis now present in
the Queensland Museum is the one from Cam-
bridge Gulf. This was confirmed once more by
Mr. Mack (in litt., 5-VI-1963); ‘T have checked
the old bird collection here in connection with
De Vis' species Pachycephala fretorum , and the
only specimen present is what appears to be a
young bird, possibly a female, with a culmen
20 mm. long. Originally, De Vis has written on
the small label Pachycephala “lanoides”. The
name “ lanoides ” has been marked out and fre-
torum put in its place. There are no other
specimens of P. lanioides in the collections’ \
It is unfortunate that none of the three indi-
viduals of fretorum described by De Vis can be
traced; it should also be noted that he says
nowhere that they were in the Queensland
Museum, and I am inclined to believe that they
were not. as it is unlikely that all three of them
would have disappeared without leaving a trace.

How De Vis could find a culmen length of
17.0 to 19.0 mm in fretorum and of 25.0 mm in
the individual from Cambridge Gulf is obscure.
Possibly he measured the exposed culmen of
fretorum and the entire culmen of lanioides .
which would make it about right.

Colluricincla harmonica julietae Mathews
Mathews (1942)* based the name Colluricincla
brunnea julietae on a single individual from
Sturt Creek, which was described as differing
from brunnea in its smaller size. The type in
the collection of the Western Australian
Museum, regd. no. A 4004, has a wing length of
117 mm (Mathews gave it as 116), and though
its plumage is abraded, even freshly moulted
the wing would not have been more than one
or two mm longer.

Serventy (1946) concluded that Mathews's paper with
the description of this and other new forms was pub-
lished in 1943. and at the time he was right. The
International Commission on Zoological Nomenclature
has since ruled, however, that circulation of author’s
copies and other preprints constitutes valid publica-
tion (Stoll & al. 1961), and certainly in a case like the
present one. in which the exact date of publication is
printed on the paper, there is no reason not to accept
this date, 12th February. 1942. An interesting point
that appears is that now Cervinipitta kimbleyensis
was published earlier that the emendation Cervinipitta
kimberleyensis, which appears on a slip circulated with
the completed volume in 1943, but not with the
original preprints.

94

Measurements of brunnea in our collection
(specimens from Kalumburu, Wyndham. Nichol-
son Station, South Alligator River and Eureka >
are: <$ , 127, 129, 132, 134, 134; 9, 129, 130, 131.

134. It is clear from these measurements that
Mathews was right and though personally 1
would not base a subspecies on a single worn
individual and on geographical grounds strongly
doubt its validity, now that it has been named,
it cannot be rejected until additional material
may necessitate a reconsideration of its validity.
Perhaps it is significant that the male from
Light House Rock Pool, Nicholson River, with a
wing length of 127 mm, is slightly smaller than
other specimens of brunnea examined. Nichol-
son River is about a hundred miles north-east
of Sturt Creek, the type locality of julietae.

The northern populations of C. h. rufiventris
are of about the same size as julietae, but they
differ in being grey-backed rather than brown-
backed, and though I have no hesitation in
making both rufiventris and brunnea subspecies
of harmonica the relationship of julietae is with
the latter rather than with the former, as,
indeed, one would expect on geographical
grounds.

Melithreptus lunatus chloropsis Gould
Gould (1848) described this subspecies from
Swan River as having the unfeathered skin
round the eye green, for which reason he named
it chloropsis. Mathews (1909) named Meli-
threptus whitlocki from Wilson’s Inlet as having
the skin round the eye white all year. It may
be pointed out that the names given by Mathews
before his notorious contribution of January-
1912, are usually valid, and the existence of two
races, one with white and one with green over
the eye, in the south-w T est of Western Australia
has been accepted by several subsequent authors
(Whittell & Serventy 1948, Serventy & Whittell
1951, but corrected by Serventy & Whittell
1962), I have observed and collected specimens
in various parts of Western Australia, both near
Perth and along the south coast, and found that
in adult birds of both sexes the lunate patch
over the eye is always white, w T hile the lower
eyelid is greyish blue in colour. In juvenile
birds, on the contrary, which are easily distin-
guished by having a brownish tinge on the
mantle and by the brown, not black, crown, the
skin above the eye is pale blue. Gould’s state-
ment to the contrary, the skin is never green.
It is clear, therefore, that Western Australia is
inhabited by one race only, which retains the
misnomer chloropsis. The nominate race differs
by having the skin above the eye orange, and
by its very slightly shorter bill.

Meliphaga virescens (Vieillot)

The geographic variation of this species,
w T hich has a continuous range over the whole of
Australia, is so gradual that it is impossible to
express it satisfactorily in nomenclature. The
main trend is one of larger size and darker
colour towards the south and the coastal areas,
and of smaller size and paler colour inland and
north. In a case like this, where no sharp
boundaries exist anywhere, it is wise to refrain
from giving too many names. In a previous
paper (Mees 1961) I commented on the large

size of the type of lewisi from Lewis Island,
Dampier Archipelago. The bird, sexed as a
female, has a wing of 96 mm.

In 1962 a collecting party from the Western
Australian Museum visited the Dampier Archi-
pelago, and obtained a number of specimens of
M. virescens , though unfortunately none on
Lewis Island. Wing-lengths of this material
are: Legendre Island, 6 : 92.5. 96; 9 : 88.

88.5. Dolphin Island, £ : 93. 93; 9 : 87.

Though these birds are very slightly darker in
colour than inland specimens (from the Can-
ning Stock Route) of forresti , and perhaps
slightly larger on an average, they can best be
included in that race. The fresh material con-
firms my earlier suspicion that the type of
lewisi was wrongly sexed and is a male, which
w T ould account for its large size.

Meliphaga leucotis novaenorciae (Milligan)

The name novaenorciae has not generally been
accepted and Milligan’s (1904a) description to
the contrary, I cannot detect any difference in
colour between specimens from the south-west
and the east of Australia, but there is a dif-
ference in size, as was correctly pointed out by
Milligan (1904b). Material in the Western
Australian Museum measures: New South

Wales: 6 , 94 (juv.), 96,101, 102; 9 . 87, 87, 98.
Victoria: 9 , 88, 92, 95; sex unknown, 92. West-
ern Australia: $, 89, 91, 91, 91. 92, 94, 96; 9,
81. 81, 82.5, 84, 85.

It is likely that the “9” from New South
Wales with a wing of 98 mm is incorrectly sexed,
so that we get for the eastern states: <3, (94),
96-102, 9 , 87-95; and for Western Australia:

6 , 89-96; 9 , 81-85. On the basis of these

figures novaenorciae seems tenable as a form
slightly smaller than the nominate race.

Notwithstanding its name, this bird has never
been found at New Norcia; the Wongan Hills
evidently form its western limit of distribution.
Eastwards it extends at least to Zanthus (Whit-
lock 1922, p. 175).

Artamus cyanopterus (Latham)

Keast (1958), followed by Mayr (in Mayr &
Greenway 1962), recognised perthi Mathews,
though as a poorly differentiated race, on the
basis of having only the 2nd and 3rd primaries
edged with white, while in birds from eastern
Australia the 2nd, 3rd and 4th primaries show
white edges. The material of this species in the
Western Australian Museum shows that not only
is white usually present on the 4th primary of
Western Australian birds, as already indicated
by Keast, but also that in eastern birds it is
sometimes feebly indicated or practically absent
(see table). This further weakens the case for
retention of perthi and in my opinion the name
should be synonymized. Artamus cyanopterus
cannot be divided into subspecies.

4th 4th

primary primary

broadly narrowly

with with

white white

4th

primary
aiiioM:
w itluntt
white

New Smith Wales (7 specimens) 4

Western Australia <15 specimen*) 5

95

Cracticus nigrogularis nigrogularis (Gould)

In a previous paper (Mees 1961, p. 126 foot-
note) I mentioned that birds from the extreme
south-west of the Kimberley Division belong to
the nominate race. The available material is
the following:

No.

Sex

Locality

Date

Wing

Bill

Weight

(grams)

A 841 2

•

o im.

La Grange

28. v. 1900

109

47

132

A 84 13

$ ad.

»»

i9

171

45

126

A 8415

o ad.

Leanwarringah
Well (between
Broome and
Derby)

30.V.1 900

178

46

124

AS414

$ ad.

Derby ....

3. vi. I960

176

47-5

120

Two recently collected specimens
north Kimberley Division measure:

from the

No.

Sex

Locality

Date

Wing

Hill

Weight

(grams)

AS418

<$ inn

Kahnnburu . .

20. vi. 1900

159

43

AS419

ad.

99

24.vi.1960

154

38

....

Field observations at Pardoo and Anna Plains
show that the range of the nominate race is
continuous along the Eighty Mile Beach. The
boundary between the two races in the Kimberley
Division remains to be found, but as birds from
Wotjulum clearly are pure picata , the nominate
race cannot range far into the Kimberley Divi-
sion; also the boundary between the races must
be rather abrupt, which suggests secondary
contact. It is perhaps reasonable to assume
that the nominate race has only comparatively
recently worked its way up along the coastal
strip of the Eighty Mile Beach. Perhaps the
Kimberley specimens of the nominate race
average slightly smaller than more southern
birds, but they are much too large to be regarded
as intermediates.

References

Campbell, A. J. (1918). — Notes on some additions to the
H. L. White collection. Emu 18 : 2-8.

De Vis, C. W. (1889). — List of birds, lizards, and snakes
collected at Cambridge Gulf. Proc. Roy. Soc .
Qd. 6 : 236-239.

Ford, J. R. & Stone, P. S. (1957).— Birds of the Keller-
berrin-Kwolyin District, Western Australia.
Emu 57 : 9-21.

Gould, J. (1848).— ‘The Birds of Australia" IV. part 30
( London ) .

Keast, J. A. (1958). — Seasonal movements and geo-
graphic variation in the Australian wood-
swallows (Artamidae). Emu 58 : 207-218.

(1961). — Bird speciation on the Australian

continent. Bull. Mus. Comp. Zool. Harv
123 : 303-494.

Mack, G. (1933). — The White-breasted Whistler. Emu
33 : 1-3.

Mathews, G. M. (1909). — Descriptions of two new sub-
species of Australian birds. Bull. Brit. Orn .
Cl. 25 : 24.

(1930). — "Systema Avium Australasianarum”

(Taylor & Francis : London.)

(1931). — "A List of the Birds of Australasia”

(Taylor & Francis : London.)

(1942). — New forms of Australian birds J.

Roy. Soc. W. Aust. 27 : 77-78.

Mayr, E. (1948). — Geographic variation in the reed-
warbler. Emu 47 : 205-210.

(1954). — Notes on Australian whistlers (Aves

Pachycephala) . Amer. Mus. Novit. 1653 : 22 pp’.

Mayr, E. & Greenway, J. C. ( 1962 ) .-‘‘Check-List of Birds
of the World" XV (Museum of Comparative
Zoology : Cambridge, Mass.)

Mees, G. F. (1961). — An annotated catalogue of a collec-
tion of bird-skins from West Pilbara, West-
ern Australia. J. Roy. Soc. W. Aust. 44 :
97-143.

Milligan, A. W. (1904a). — Notes on a trip to the Wongan
Hills, Western Australia, with a description
of a new Ptilotis. Part I. Emu 3 : 217-226.

(1904b). — Plumage phases of Ptilotis leucotis .

Emu 3 : 234.

Reid, D. (1951). — A Bunketch bird list. W. Aust. Nat.
3 : 68-71.

Saville-Kent. W. (18891. — Preliminary observations on a
natural history collection made in connec-
tion with the surveying cruise of H.M.S.
"Myrmidon" at Port Darwin and Cambridge-
Gulf — September to November, 1888. Proc.
Roy. Soc. Qd 6 : 219-236.

Serventy, D. L. (1946). — Publication dates on some
recent names. Emu 46 : 236.

Serventy, D. L. & Whittell, H. M. (1951). — ‘‘A Handbook
of the Birds of Western Australia", 2nd ed.
(Paterson Brokensha : Perth.)

(1962). — "Birds of Western Australia", 3rd ed.

(Paterson Brokensha : Perth.)

Stoll, N. R. & al. (1961). — "Code International de
Nomenclature Zoologique adopte par le XVe
Congres International de Zoologie/Interna-
tional Code of Zoological Nomenclature
adopted by the XV International Congress
of Zoology" (Int. Trust Zool. Nomencl. :
London.)

Whitlock. F. L. (1922). — Notes from the Nullarbor Plain.
Emu 21 : 170-187.

Whittell. H. M. & Serventy, D. L. (1948). — "A Systematic
List of the Birds of Western Australia”
(Government Printer : Perth.)

96

INSTRUCTIONS TO AUTHORS

Papers may be submitted to the Society in accordance with Rules and
Regulations 38 to 41 inclusive (see below). They should be addressed to The
Honorary Secretary, Royal Society of Western Australia, Western Australian
Museum, Perth.

Authors are solely responsible for the factual accuracy and for any opinion
expressed in their papers. They are particularly requested to verify references.
Alterations to MSS. submitted to the printer will be allowed only under excep-
tional circumstances, and no changes will be permitted after galley-proof stage.

In the preparation of MSS. authors are required to follow the C.S.I.R.O.,
Guide to Authors (C.S.I.R.O., Melbourne, 1953), except that papers longer than
10,000 words (30 foolscap pages of pica type, with 6 inch lines 40 to the page) will
not normally be accepted.

Authors may be required to meet half the cost of preparation of the blocks
of diagrams and illustrations.

Authors shall receive a total of 30 reprints free of charge. Further reprints
may be ordered at cost, provided that such orders are submitted with the MS.

RULES AND REGULATIONS

38. Every paper intended to be read before the Society or to be published
in the Society’s Journal must be sent to the Secretaries at least seven days
before the date of the next ensuing Council meeting, to be laid before the
Council. It will be the duty of the Council to decide whether such contribution
shall be accepted, and if so, whether it shall be read in full, in abstract, or
taken as read. All papers accepted for publication must be read or otherwise
communicated at an ordinary meeting prior to publication.

39. A Publications Committee, appointed by the Council, shall recommend
to the Council whether a paper presented to the Society shall be published in
the Society’s Journal. The Publications Committee may obtain an opinion from
any person it may select on the suitability of any paper for publication.

40. Publication in the Society’s Journal shall only be available to (a)
Ordinary Members, (b) Honorary Members, (c) Non-members resident outside
Western Australia, who must communicate the paper through an Ordinary or
Honorary Member. No paper shall be accepted from a Non-member resident in
Western Australia.

41. The original copy of every paper accepted for publication by the
Society, with its illustrations, shall become the property of the Society, unless
stipulation is made to the contrary, and authors shall not be at liberty to
publish their communicated papers elsewhere prior to their appearance in the
publications of the Society unless permission for so doing is given by the
Society, or unless the Society fails to publish the paper in the Journal of the year
in which it is read or otherwise communicated, or of the succeeding year.

Journal

of the

Royal Society of Western Australia

Volume 47

1964

Part 3
Contents

10 —Notes on the Skulls of Two Western Australian Rodents with a key to the
Skulls of the Rodents of Southwestern Australia. By Ernest Lundelius, Jr.

11. — Effects of Salt, Temperature and Seed Scarification on Germination of Two

Varieties of Arthrocnemum halocnemoides. By C. V. Malcolm.

12. — Ashbed and Nutrients in the Growth of Seedlings of Karri (Eucalyptus

diversicolor F. v. M.). By O. W. Loneragan and J. F. Loneragan.

13. — Foraminiferal Evidence for the Paleocene Age of the King’s Park Shale

(Perth Basin, Western Australia;. By Brian McGowran.

14. — Projasus — a New Generic Name for Parker’s Crayfish, Jasus parkeri Steb-

bing (Palinuridae: “Silentes”). By R. W. George and J. R. Grindley.

Memorial — Douglas Elwood White.

15. — Geographical Variation and Distribution of some Birds from Western Aus-

tralia. By G. F. Mees.

Editor: R. W. George
Assistant Editor: R. D. Royce

Annual Subscription: Forty-five Shillings

The Royal Society of Western Australia, Western Australian Museum, Perth

85491/7/64—570

ALEX. B. DAVIES, Government Printer, Western Australia

JOURNAL OF

THE ROYAL SOCIETY

OF

WESTERN AUSTRALIA

VOLUME 47 (1964]

85977 / 7 / 64—570

Annual Report of the Council of The Royal Society of Western Australia

for the Year ending 30th June, 1963

Membership

There are 246 members on the register of the
Society. They are as follows: —

Honorary 1

Life 3

Corresponding 2

Ordinary 212

Associate 14

Student 14

In the past year 14 Ordinary and 1 Student
Members were elected. Four Ordinary Members
resigned.

It is with regret that we record the deaths of
two Honorary Life Members, Mr. Ludwig Glauert
and Miss Enid Allum. Mr. Glauert, a foundation
member, served on Council for many years and
was President for two years. He received the
Kelvin Medal from the Society in 1945. Miss
Allum joined the Society in 1915, and served as
Treasurer for many years.

Council

Eleven Council meetings were held at the
Zoology Department of the University in the
past financial year. Attendances were as fol-
lows : —

Dr. W. D. L. Ride 8, Professor R. T.

Prider 10. Dr. N. H. Brittan 7. Dr. R.

W. George 5, Dr. J. E. Glover 10, Dr.

L. W. Samuel 11, Mr. C. F. H. Jenkins

7, Mr. J. G. Kay 7, Mr. R. J. Little 9,

Mr. C. V. Malcolm 6, Mr. D. Merrilees 7,

Mrs. A. Neumann 9, Mr. P. Pillow 11,

Miss M. E. Redman 11, Mr. R. D. Royce

8, Mr. W. R. Wallace 8.

Publications

The Editor, Dr. Glover, maintained the high
standard of the Journal in the publication of
volume 45, parts 2, 3 and 4 and volume 46,
part 1. Mr. Royce and Dr. Mees assisted with
the distribution and editing.

T e Proceedings, compiled by Mr. Malcolm,
appeared regularly with notices of General
Meetings, summaries of previous Meetings, and
items of general interest to members.

Library

The Librarian, Mrs. Neumann, has catalogued
all holdings, revised the entries for the C.S.I.R.O.
List of Scientific Serials and reviewed the ex-
change list. Although £60 was spent on binding,
there remains a considerable amount to be done.

New exchanges were arranged with six insti-
tutions: Institute of Caribbean Studies of the
University of Puerto Rico; University College
of Addis Ababa, Ethiopia; Academy of Science
of the Kazakh S.S.R., Alma Ata; Deutsche
Akademie der Wissenschaften zu Berlin, Ger-
many; Institute of Geology, University of
Parana, Brazil; Naturwissenschaftlicher Verein
fur Steiermark, Graz, Austria. The total number
of exchanges is now 217.

General

The Standing Committee on Conservation,
which was appointed by Council to advise on
matters involving the preservation of Western
Australian flora, fauna and geological features,
has beeen active. The members are- Mr. C. F.
Jenkins (President), Professor B. J. Grieve, Drs.
A. R. Main, P. E. Playford, W. D. L. Ride and
D. Serventy, and Messrs. G. E. Brockway, T.
Cleave, A. J. Fraser, J. F. Morgan and R. D.
Royce.

During the year, two films were shown and
seventeen papers and talks were given. They
were entitled as follows: —

“Symposium on flies”, by C. F. H.
Jenkins, Dr. D. R. R. Snow and
Dr. A. Zorbas.

“The pebble crabs of W.A.”, by Dr.
R. W. George.

“The life cycle of the plant pathogen
Scleratina sclero+orium in W.A.” by

R. M. Henderson, presented by Mrs.
E. R. L. Johnson.

“The foraminifera of Oyster Harbour,
W.A.”, by K. McKenzie.

“Some aspects of forestry in Africa”,
by W. R. Wallace.

“Cave paintings of the West Kimber-
ley”, by Dr. P. E. Playford.

“Botanic gardening”, by Dr. M. Phillips.

“Seismicity of Western Australia”, by
I. B. Everingham.

“Fossil mammals of the Nullabor caves”,
by Dr. E. Lundelius, presented by
Dr. W. D. L. Ride.

“Evolution in Isotoma”, by S. James.

“Development of the South Perth Zoo”,
by C. F. H. Jenkins.

“Fossil coastline of Geographe Bay, near
Capel, W.A.”, by J. G. Kay.

“Behaviour and physiological studies of
lizards”, by W. R. Dawson, P. Licht,

S. D. Bradshaw and V. Shoemaker.

“Symposium on conservation”, by Drs.

W. D. L. Ride, A. R. Main and P. E.
Playford and Mr. R. D. Royce.

“Honey and beeswax production in
Tanganyika”, by Dr. F. G. Smith.

“The gekkonid genus Nephrurus in
Western Australia”, by Dr. G.
Storr.

“Agricultural prospects of the For-
restania area”, by G. H. Burvill.

In conclusion, the Society would like to record
its thanks to the State Treasury for its con-
tinued support of the publication of the Journal,
to the Government Printer for his co-operation
in the Journal publication, and to the Museum
Board for the housing of the library and for
providing premises in which to hold the regular
meetings.

W. D. L. RIDE,

President.

M. E. REDMAN,

Joint lion. Secretary.

Journal
of the

Royal Society of Western Australia
Vol. 47 Part 4

16. — A Review of Australian Fossil Marsupials

Presidential Address, 1963
By W. D. L. Ride, M.A., D.Phil.*

Delivered — 15th July, 1963

Contents

Classification ....

Phylogeny
Marsupi carnivora

A ustralian Marsupi carnivora

Thylacinidae

Dasyuridae

Glaucodon ballaratensis
Dasyurus affirm and D. bowlingi
Sarcophilus lama-rims and S. prior
Myrmecobiinae ....

Peramelina

Jschnodon australis
Diprotodonta ....

Wynvardiidae : Wynyardia bassiana
Phalangeridae

Perikoala palankarinnica

Burramys parvus ....

Macropodidae

Generic Classification
Structure and Terminology
Phyletic lines
The Grange Burn potoroo
Biprotodontidae
Vombatidae

incertae sedis : Notoryctidae
The Australian Radiation : Origin, Isolation and

Drift ....

Appendix: Statistical Determination of Subspecies
References

Page

98

100

101

103

105

109

110
111
111
111
112
112
112

114

115
118
118
119
119
121
123
123
125
125

125

126
128
129

Introduction

In recent years there has been a great resur-
gence of interest in Australian marsupials. In
the case of the modern mammals this has pro-
duced a vast body of information through the
application of modern physiological and ecologi-
cal techniques, knowledge which is now replac-
ing the natural history anecdotes and zoological
inferences of past generations. In a similar
fashion, the advent of the four-wheel drive
vehicle, of plastics, and the discovery of new
techniques for handling data, have produced
encouraging signs that knowledge of fossil
marsupials is also likely to increase. Today,
palaeontological field-work aimed at the discovery
of new fossil marsupials in Australia is going on
from most of the Australian museums and also,

with spectacular success, from several American
universities. Unfortunately, as yet, only a small
fraction of the results of this new work has
been published.

Despite our present-day activity and the
publications of our predecessors, our knowledge
of pre-Pleistocene marsupials remains small
because descriptions are, as yet, available of
only very few Tertiary fossils and of these only
one (Wynyardia) is certainly older than Miocene.
Preliminary descriptions have been published of
some of the recently discovered middle and
upper Tertiary forms from Central Australia
(Stirton 1955, 1957a), while stratigraphic infor-
mation from a number of sources is also being
made available (Stirton, Tedford & Miller 1961,
Ludbrook 1963, 1963a, Balme 1963). When the
full descriptions, proper comparisons, and the
ages of these fossils are known they will provide
the first real framework for palaeontological
studies on the Age of Marsupials in Australia.

Although there are great numbers of Pleisto-
cene fossil marsupials in the Australian Museums,
little has been done with them since they were
originally described. Thus, advances in the
knowledge of the anatomy and classification of
recent forms, and concepts of population biology
and of faunistics, have played little part in their
interpretation. In addition, most of the early
workers did not recognise the need for accurate
stratigraphic localization. Thus, it is scarcely
surprising that one of the most pressing tasks
in Australian Palaeontology today is that of
rediscovering the classic localities, and localizing
the old material through comparison with new
It is only in this way that the species names
in use can be stratigraphically allocated, and it
is only through these new collections that ranges
of morphological variation of the various species
will be determined. At present “species” are
often represented by samples which may not be
stratigraphically homogeneous and population
studies based upon them may be quite erroneous.
An outstanding example of such work has been
the studies of Woods (1960a), and more recently
of Bartholomai (1962, 1963), on the Darling
Downs faunas. Woods has shown that this
material, which was formerly regarded as a
single rather mixed unit, belongs to two distinct

* Western Australian Museum, Beaufort Street, Perth.
85977 — ( 2 )

97

successional faunas. Although physical super-
position has not yet been demonstrated, these
are probably Upper Pliocene and Lower (or Mid)
Pleistocene.

Such work, if it is to be done to modern
standards, is very slow and it is fortunate that
(even omitting the numbers of overseas workers)
there are in Australia today almost as many
active researchers in this field as there have
been in the whole of our history.

The review which I present here is a general
one of the outlines of our knowledge of Aus-
tralian fossil marsupials to date. However, in
it I also take the opportunity to give fresh data
and conclusions relative to a number of fossils
which I have studied but have not yet published
upon. These are Glaucodon ballaratensis, the
Thylacinus complex, the Grange Burn “cuscus”
(which is not a cuscus but a potoroo). and
Wynyardia bassicina. I also propose a new
ordinal classification of the Marsupialia to bring
their arrangement more into line with that
accepted for the Eutheria. I also consider the
current status of the Continental Drift hypo-
thesis and conclude that it provides an adequate
explanation for anomalies in the composition of
the Australian mammalian fauna.

Classification

Today, marsupials occurring in Australia are
usually classified into three major groups. These,
following Simpson’s (1930) example, are con-
ventionally given the rank of Superfamilies, as
are the three other major taxa of marsupials
which occur in Palsearctic and Neotropical
regions (see also Simpson 1945). This rather
low rank for each of these groups fits well with
the traditional practice of regarding the Mar-
supialia as a single Order of Mammalia, but it
also implies an unwarranted degree of homo-
geneity in a group of animals which have evolved
for just as long as the Eutheria have; by con-
trast these are currently subdivided into twenty -
six Orders.

Fossil marsupials, particularly those of the
Australian Pleistocene, clearly demonstrate a
wide variety of forms which, if they had been
eutherian, would have been distributed among
several orders. Thus, if the kangaroos and
wallabies are antelope- and deer-like, the
sthenurines bovid-like and the diprotodontids
hippo-like then the native cats can only repre-
sent some such unspecialized Carnivora as the
civets and mongooses. The bandicoots are very
different again from both of these groups and
clearly invite comparison with yet another order.

This idea is not anything original for taxono-
mists have long recognized this inconsistency in
classification. For example, Cain (1959, p. 214)
has said in explanation: "Because of their

peculiar features Tmarsupialsl are always ranked
as a single order of mammals within a separate
class, although the briefest inspection is enough
to show that there is at least as much difference
between a kangaroo and a dasyure ( for example)
as between an insectivore and a rodent, let alone
a rodent and a lagomorph. Because eutherian
mammals were the first to become familiar to
the anatomists of the western world, and even
more because our species belongs to them, they
have been taken as normal and the extraordinary

‘abnormality’ of the marsupial urinogenital
system has sufficed to keep marsupials as a single
order. Had we known the marsupials first and
especially if we had belonged to them, they
would have been classified in several orders, and
no doubt the ‘abnormality’ of the eutherian
genitalia would have sufficed to keep the number
of eutherian orders much lower than at present.’’

In the classification which follows I do not
discard the six superfamilies but, in order to
emphasize the breadth of the whole radiation,
and since I believe that a synthetic view of
palaeontology and neontology requires it, I
group these further.

In an earlier review of the beginnings of the
marsupials and of the main features of their
phylogeny (Ride 1962) I had pointed out that
marsupials had formerly been grouped in two
higher taxa (suborders) and the abandonment
of these by Simpson in 1930 was one of the most
important advances in marsupial classification
of our time. Through discarding them Simpson
promoted a freer mental approach to problems
of interrelationship within the marsupials as a
whole. The problem which now arises is whether
or not a new ordinal classification would have
the same restrictive effect as did the subordinal
classifications attributed to Owen and Bensley
(or de Blainville). I think that this is unlikely
because it cuts across no phylogenetic divisions
It also raises no problems of intercontinental
migration.

The earlier classifications raised problems in
both of these directions.

Although Owen’s classification into Polyproto-
dontia and Diprotodontia was not intended to be
phylogenetic it became so in the minds of
students and, when the caenolestoids were added
to the Diprotodontia, argument inevitably
became confused with the issue of Continental
Drift. By constrast, Bensley’s classification was
phylogenetic but was complicated by his mis-
interpretation of the structure of the didelphoid
foot ( see Goodrich 1935), but better understand-
ing today makes it possible that his Syndactyla
will eventually be used as a “cohort” in future
marsupial classifications. Both systems (as
phylogenetic arrangements) had their proponents
and the problems in the minds of students as
to which was more likely to be correct were
only disposed of by Simpson’s action in abandon-
ing them altogether. He rejected these “key-
classifications” which were based on single
characters and subdivided the Marsupialia into
five or six groups of very different importance
and variety but of equal or nearly equal anti-
quity.

In 1930 there seemed to be no good reason for
grouping the superfamilies into taxa at higher
level but, since then, fossils of about the right
age (Palaeocene and Lower Eocene) and with
suitable characters to provide a transitional
stage between the non-Australian superfamilies
Didelphoidea and Borhyaenoidea have been dis-
covered (i.e. Eobrasilia Simpson 1947 and Patene -
like forms Simpson 1948, p. 49). If the Dasyu-
roidea be added to these — and there is little to
argue against this course in spite of the absence
of transitional fossils between them and the
South American superfamilies — then the argu-

98

ment that Dasyuroidea, Didelphoidea and Bor-
hyaenoidea form a single unit comparable with
the eutherian Carnivora becomes even stronger.
Most authors regard the modern Dasyuroidea
as not very greatly modified descendants of the
original didelphoid invaders of Australia and.
since the Phalangeroidea at least (and also the
Perameloidea > warrant ordinal status by com-
parison with Eutheria, I take the formal step of
proposing four orders of Marsupialia.

The classification which I adopt is as follows:
Class MAMMALIA Linnaeus, 1758.
Infraclass METATHERIA Huxley, 1880.
Superorder MARSUPIALIA Illiger, 1811.
Order 1. MARSUPICARNIVORA nov.
Superfamilies

1. Didelphoidea : American Opossums
etc., U. Cret — L. Mioc, Pleist — R,
N. Amer.; Eoc-Mioc, Western
Europe; Palaeoc — R, S. Amer.

2. tBorhyaenoidea : South American

marsupial carnivores. Palaeoc-
Plioc, S. Amer.

3. Dasyuroidea : Australian Native-
cats etc. ?Mioc — R, Aust.; R, N.
Guinea.

Families

1 Dasyuridae : ?Mioc — R,
Aust.; R, N. Guinea.

2 Thylacinidae : U. Plioc — R.
Aust.; ?Pleist, N. Guinea.

Order 2. P AUCITUBERCULATA * Ameghino,
1894.

Families

1. Caenolestidae : Oppossum rats etc..
Palaeoc — R, S. Amer.

2. iPolydolopidae : Palaeoc — Eoc, S.
Amer.

Order 3. PERAMELINA Gray, 1825.

Family

1. Peramelidae : Bandicoots, Plioc —
R, Australia; R, N. Guinea.

Order 4. DIPROTODONTA Owen, 1866.
Families

1. Phalangeridae : Possums, cuscuses,
marsupial lions, etc.. ?Mioc — R,
Aust.; R f N. Guinea.

2. $Wynyardiidae : Oligocene* Tas-

mania.

3. Vombatidae : Wombats. Pleist —
R, Aust.

4. tDiprotodontidae : Mioc — R. Aust.;

?Plioc, N. Guinea.

5. Macropodidae : ?Mioc — R, Aust,;
?Plioc — R, N. Guinea.

Marsupialia incertae sedis

Family Notoryctidae : Marsupial Moles. R,
Aust.

* I follow Cabrera 1919 in adopting Ameghino’s name
for this taxon and like him exclude Microlestes and
Hypsiprynopsis from the concept as well as the
“Hypsiprymnoidea”— see Ameghino 1903 (Vol. 13, p.
962 of Obras completas) .
v taxa comprising extinct forms only.

Numbers of genera in orders . — If the fossil and
modern genera of these orders are counted and
the generic concepts of a single conservative
author are employed (G. G. Simpson 1945), they
fall well within the range of magnitude of orders
of Eutheria. Thus:

Marsupialia — Marsupicarnivora 64. Pauci-
tuberculata 21, Peramelina 5, Diproto-
donta 48. Incertae sedis 1.

Eutheria (some orders only) — Embri-
thopoda 1, Tubulidentata 1 (possibly
2§ ) , Dermoptera 3, Pholidota 1 (possibly
4§), Pyrotheria 4 (possibly 6§), Dino-
cerata 8, Astrapotheria 9. Hyracoidea 13.
Sirenia 16 , Proboscidea 24, Lagomorpha
33, Perissodactyla 158, Carnivora 377,
Artiodactyla 419.

In making these comparisons, however, we
must remember that we know little of the fossils
of some of these marsupial orders and the values
shown here for the numbers of genera included
in each are not really comparable with those
given for the eutherian orders because they are
too low. This is made clear by comparing the
approximate percentages of genera of marsupial
and eutherian orders which are known only as
fossils. Since a number of the smaller orders
of Eutheria are known only as fossils, compari-
son has been limited here to those which, like
the orders of Marsupialia, contain some living
representatives. The percentages of genera
known from fossils only in each order are:

Marsupialia — Marsupicarnivora 68%, Pera-
melina 0% Paucituberculata 86%, Dip-
rotodonta 35%.

Eutheria — Tubulidentata possibly 50% §,
Dermoptera 67%, Pholidota possibly
75% §, Hyracoidea 77%, Sirenia 88%,
Proboscidea 92%, Lagomorpha 70%,
Perissodactyla 96%, Carnivora 70%,
Artiodactyla 80%.

The same data reveal that our knowledge of
non -Australian fossil marsupials is comparable
with that of eutherians but knowledge of the
Australian forms lags far behind that of euther-
ian orders of comparable size. This is empha-
sized by the removal of the Australian compon-
ent from Marsupicarnivora; this causes the
value to rise to 80%, with 0% of purely fossil
genera known within the Dasyuroidea||.

The marsupial orders do not match the
eutherian orders only in numbers of genera but
also in the breadth of their adaptive radiations.

I have already made brief mention of similari-
ties between Diprotodonta and Artiodactyla and.
in general outline, a similar comparison can be
drawn between Marsupicarnivora and Carnivora.
In making such a comparison it must be first
recognized that the total adaptive scope of the
two orders is not directly overlapping. Thus,
some ecological branches (like the Pinnipedia
and some wholly vegetarian Procyonidae) of the

§ Figures qualified thus are obtained by including those
genera which Simpson includes with reservation.

In these figures Glaucodon (Dasyuroidea) and Iach-
nodon (Peramelina) are not included because they
were published after the date of Simpson's 1945
monograph from which the data are drawn. To
insert these without adding newly described genera
of Eutheria would be pointless.

85977 — ( 3 )

99

radiation of Carnivora are missing from Mar-
supicarnivora, but the Marsupicarnivora is
widened in other directions through including
many of the equivalents of the eutherian order
Insectivora within it; these are excluded from
Carnivora. Otherwise, remarkable similarities
exist between the taxa. Relatively unspecialized
carnivores approaching Mustelidae and Viver-
ridae are found in Borhyaenoidae ( Amphipro-
viverrci) and Dasyuroidea (Dasyurus), and
omnivorous animals like the marsupials Didel-
phis and Philander can be equated with members
of the Raccoon family (Procyonidae, e.g. Bas-
sariscus). Sarcophilus, the Tasmanian Devil, is
like a Wolverine ( Gulo ) or a Hyaena in its car-
nivorous specializations and Thylacinus is like
a wolf. Chironectes (the Water Opossum) with
its slightly flattened tail and webbed toes might
even be regarded as an early approach to otter-
like form. One of the most remarkable ecologi-
cal and functional similarities is that between
Thylacosrnilus the Marsupial Sabre-tooth
(Pliocene, South America) and Smilodon and
the other true sabre-tooth tigers (see Simpson
1941a). As far as I know there is no direct
equivalent between marsupials and such highly
specialized herbivorous carnivores as the pro-
cyonids Potos (the Kinkajou) and Ailuropoda
(the Giant Panda) but, as already mentioned,
this is only to be expected when broad com-
parisons are being drawn between orders.

The insectivore-like forms among the Mar-
supicarnivora include Didelphoidea: Perather-
ium, Eocene-Miocene Europe, North America;
Marmosa*, mouse-opossums, Recent South
America; and Dasyuroidea: Sminthopsis, Plani-
gale and Antechinus , Marsupial-mice, Recent

* Information on the dietary preferences of South
American mammals referred to in this review is
from Hall & Dal quest 1963.

and Pleistocene, Australia. Some of these are
arboreal, e.g. Marmosa and Caluromys of South
America, and Phascogale (sens, strict.) of Aus-
tralia. This latter genus should possibly be
equated with the tree-shrews Tupaia. If Noto-
ryctes (the Marsupial Mole) should prove to be
a dasyuroid its only equivalent is also in the
Insectivora (i.e. Talpa or Chrysochloris) ,

Finally, the taxa have comparable ranges in
time: Carnivora occurred from the L. Palaeocene
to Recent, and Marsupicarnivora from the U.
Cretaceous to Recent.

PHYLOGENY

Lacking fossils we can only derive the probable
relationships of the Australian marsupial orders
by inference from the characters of their modern
representatives. Such a phylogenetic tree is
shown in fig. 1. This is essentially the same as
that shown in my 1959 review of some problems
of marsupial phylogeny (Ride 1962) but incorp-
orates the ordinal classification given here. It
illustrates the conclusion that the major groups
of marsupials may be arranged as a series of
structural grades. Recently Masai (I960) and
Dillon (1963) have published results which
threw some doubt on the position at which
duplicicommissuraly (i.e. the possession of a
fasciculus aberrans in the forebrain) should
first appear in this dendrogram. Their results
are so at variance with those of earlier workers,
such as Elliot Smith and A. A. Abbie, that I
have not taken them into account and it is
quite clear that further examination of speci-
mens of the various species must be made in
case the apparent disagreement is produced by
differences in technique, in interpretation, or
even through faulty material.

\

PAUCITUBERCULATA

Posterior incisors reduced
in umi groups (but not oil)

( pseudodiprotodonty)

MARSUPICARNIVORA
(Didelphoidea , Bor hyae noidea,
Dasyuroidea )

I

PERAMELINA
Incisors 5 or 4
3

1+

Incisor number variously
reduced in specialized groups
remains unreduced beyond
i in others

♦ + 4- +

2nd A 3rd pedal digits
conjoined (syndactyly)
Incisors possibly reduced
from 5 to 5
4 3

DIPROTODONTA
i*l

J+l
1+1
l+i
l+i

Fasciculus aberrans developed
in forebrain ( duplicommissuraDi
__ Incisors reduced by loss of
♦ anterior elements to 3-1
3-1

(diprotodonty )

ANCESTRAL MARSUPIALIA
Incisors reduced from -j- to A- (polyprotodonty)
Ureters rotated to lie mesial to mullerian ducts
Epitympanic recess beginning to be floored by alisphenoid

Fasciculus aberrans not developed in fore brain

(simpticicommissural )

2nd ft. 3rd pedal digits not cor\joined ( dladoctyl )

Sjndictyl lints shown thus
Ouplicommissunl lints thus

THERAPSIOA

I

Fi g x — The sequence of morphological differentiation in the evolution of marsupial orders (modified from Ride

1962 ).

100

Fig. 2. A family tree of marsupials Horizontal distances between unbroken lines limiting phyla represent known
numbers of genera. Stipple represents aquatic barriers (modified from Ride 1962).

If the results of Masai and Dillon were to be
confirmed, they would produce no change in the
general phyletic outline but they would have
the result of pushing the evolutionary origin of
the fasciculus aberrans into the various lines
within the already evolved Diprotodonta, since
some of these will be like all other mar-
supials in lacking it while most will possess it.
However, Masai states that some Diprotodonta
posses a corpus callosum. If this is so, they
differ from all other marsupials and resemble
Eutheria — a somewhat remarkable state of
affairs.

The widely held view of the fundamental
dichotomy between marsupial and eutherian
female urinogenital systems has been challenged
by Kean (1961). He holds that the median
(eutherian) vagina has become reduced in the
marsupial line, following the reduction in size
of the neonatus, and lateral vaginae have evolved
as neomorph seminal ducts. It is Kean‘s view
that embryological evidence is equivocal.

Figure 2 is a graphic summary of geographical
and temporal distributions combined with the
dendrogam of Fig. 1. Horizontal breadths of
the ascending phyla represent the known num-
bers of genera at each horizon, and the diagram
gives some indication of marine barriers to
dispersal. However, it takes no account of the
possible role of the Tethys sea as a barrier to
dispersal from northern to southern Eurasia and
leaves the question unanswered as to whether

marsupials entered Australia by means of a
southern intercontinental connection or across
Wallace’s Line.

MARSUPICARNIVORA

The fossil evidence of relationship between
the two main groups of American Marsupicarni-
vora (the Didelphoidea and Borhyaenoidea) lies
in the fossils Eobrasilia, Patene and a problema-
tical group of five isolated molars of which
Simpson (1948, p. 49) says: “These various teeth
appear to represent a morphological group of
fairly unified nature and intermediate in its (ad-
mittedly too few) known characters between
borhyaenids and didelphids. The isolated teeth
could be referred to either family. Among the
Borhyaenidae they would be the most primitive
known forms, among the Didelphidae the most
advanced in the borhyaenid < or predaceous )
direction. On the whole, I think them some-
what closer to borhyaenids, but they are almost
perfectly intermediate as far as they go. In
conjunction with the almost equally primitive
Patene , they give a clue, slender but real, to the
ancestry of the Borhyaenidae.”

Simpson sums up his attitude (1948, p. 40) to
the relationship of didelphoids, dasyuroids and
borhyaenids by saying: “Traced as far as possible
to its fundamentals, the structure of borhyaenids
in general seems clearly derivable from a primi-
tive stock certainly pre-thylacinid and compar-
able only to the didelphids and the most primi-

101

tive dasyurids* It resembles both these stocks
(which are very similar in essentials), and there
is no conclusive evidence of relations to one or
the other. Adaptively the group parallels the
dasyuroids, but its inferred structural ancestry
appears rather more didelphoid. The most
probable conclusion is that in the Upper Creta-
ceous the didelphoids and dasyuroids had a com-
mon ancestry sharing the primitive characters
of both groups, somewhat more like the Didel-
phidae than the Dasyuridae as we know them
because the former are more conservative. From
this stock the borhyaenids arose. They may very
well have arisen from a carnivorous line pro-
gressing in the direction of the Dasyuridae, but
not very far along this line, and surely before it
had acquired its most characteristic specializa-
tions.”

Of course, there are no fossil intermediates be-
tween dasyuroids and the American forms but
comparison of the characters of even modern
Dasyuroidea with Didelphoidea and Borhyae-
noidea reveal no differences which separate them
unequivocally.

Modern dasyuroids differ principally from
didelphoids in that they are specialized in a
carnivorous (or insectivorous) predaceous direc-
tion and are primarily cursorial. Thus Dasyu-
ridae seldom possess the opposable hallux of
the scansorial didelphoids and in this they re-
semble some Borhyaenidae (see Prothylacinus
in Sinclair 1906, p. 371 and pi. 54) but not all.
Whether the dasyuroid foot with its small and
non-fully-opposable (and often very reduced)
hallux and its frequently high degree of ter-
restrial specialization is more primitive than
the highly specialized arboreal form of the
didelphoid foot is not known. I know of no
description of the hallux of a mesozoic mam-
mal and I consider that Bensley’s (1903, p. 163)
view that the terrestrial foot of dasyuroid mar-
supials is secondly derived from the arboreal
one is no more than an assumption. Bensley’s
belief is derived from Dollo’s statements (1899,
1900) and from Bensley’s own observation (p.
191) that the didelphoid Marmosa, which has an
opposable hallux, meets the requirements of a
prototype for the entire Australia marsupial
radiation because it possesses “an indication of
the syndactylous condition of the Phalangeri-
dae’\ His conviction that this approach to
Marmosa was reasonable was strengthened for
him by the presence of the “primitive” five
external stylar cusps on the upper molars of
the same genus. Goodrich (1935) has shown
that his statement regarding syndactyly in
Marmosa is erroneous. Bensley also pointed out
that the dasyuroids differ from didelphoids in
incisor number; dasyuroids never possessing
more than 4/3. However, this statement is
only true in relation to modern didelphoids
since some (e.g. Eodelphis cutleri, Cretaceous,
Belly River Pm„ Canada. Simpson 1929, p. 128)
have also reduced their lower incisors to three.
Borhyaenidae reduce their lower incisors even
further; thus Borhyaena excavata and B.
tuberculata have an incisor formula of 3/2 while
Thylacosmilus possesses none at all.

The molars of most modern dasyuroids differ
from those of most didelphoids in that they
lack the full complement of external stylar

Didelphoidea Dasyuroidea

Fig. 3.— The cusps (black circles) and stylar cusps
(stipple) of the upper molars of Didelphoidea and
Dasyuroidea. (Didelphoids after Simpson 1929. p. 119;
dasyuroid second upper molars in collection W. Aust.

Mus.)

cusps which Simpson has shown (from a study
of very numerous Cretaceous didelphid upper
molars as well as that of a large series of later
didelphids from all known horizons; Simpson
1929, p. 119) is ancestral in the Didelphoidea —
see fig. 3. These cusps vary considerably but
Simpson says that, in teeth of the didelphid
type, stylar cusp A is always present and is only
slightly external to the paracone save on M\
B is never well developed, while C and D are about
equally developed and usually form ridge-like
crests parallel to the outer border. E is always
present except on M 4 . In modern Dasyuroidea
there are seldom more than three stylar cusps
present. However, in Borhyaenoidea although
the earliest forms possess the full complement
of stylar cusps (e.g. the five Patene-\\ke molars
illustrated in Simpson 1948, p. 49) more spe-
cialized forms have progressively fewer, Patene
and Procladosictis (Palaeocene and Eocene)
having only twin styles opposite the paracone
and metacone and Amphiproviverra , Prothy-
lacinus and Borhyaena (Miocene) having only
a single anteroexternal stylar cusp (Sinclair
1906). In Thylacosmilus (Pliocene) they are

102

absent (Simpson 1941). The lack of a full
complement of stylar cusps can thus scarcely
be said to be a dasyurid character.

It is unfortunate that very little detailed
descriptive comparative work has been pub-
lished on the epitympanic region of marsupials
since it seems likely that this structure may
prove to be a better indication of separate phyle-
tic lines in the Marsupicarnivora than the den-
tition. In Dasyuridae the epitympanic sinus
anterior to the epitympanic recess is always
ventrally enclosed in an alisphenoid bulla which
grips the tympanic ring laterally (fig. 4). It
is widely open into the epitympanic recess pos-
teriorly. The dorsal and anterolateral part of

Dasyurus geoffroyi

ring

epitympanic sinus
bulla

epitympanic

recess

Dasycercus cristicauda

epitympanic sinus
lanterior)

epitympanic^ sinus
periotic

hypotympamc
sinus in
exoccipital

tymponic ring

alisphenoid bulla
(cut away)

Fig. 4. — The epitympanic region of Marsupicarnivora.
In both dasyuroids the alisphenoid bulla is cut away
to reveal the epitympanic sinus. In Dasycercus cristi-
cauda the squamosal and exoccipital are also cut away.
The tympanic ring of the specimen of Didelphis has
been lost. (Specimens in W. Aust. Mus. )

the epitympanic sinus is also greatly excavated
into the substance of the alisphenoid and the
squamosal where it may even penetrate into
the root of the zygomatic arch dorsal to the
glenoid. While the ear of Thylacinus is super-
ficially rather different in appearance with its
small flattened, and posteriorly incomplete,
bulla and dorsally situated epitympanic sinuses,
it is clearly derivable from that of the more
typical dasyurid condition. In Didelphis (and
in Wynyardia see p. 115) there is no dorsal
o: anterolateral “excavation” and the cavity
(which probably represents the epitympanic re-
cess alone) is only partly floored by an incom-
plete tympanic (or bullar) wing of the alis-
phenoid. Since the marsupials alone floor the
epitympanic recess with the alisphenoid (Ride
1962 ) the uncomplicated condition seen in Didel-
phis is more likely to be primitive than the
fully enclosed and “excavated” condition of
Dasyurinae. Simpson’s (1929, p. 129 > illustra-
tion of the glenoid region of the Cretaceous
didelphoid Eodelphis, while very incomplete, sug-
gests a condition similar to that of Didelphis:
the illustrations of Borhyaenoidea by Sinclair
(1906) show yet another condition which may
indicate that the structure will be useful in
phylogenetic studies.

It seems that proper study of the Marsupi-
carnivora will reveal distinct lineages coming
from a single basal stock but because Dasyu-
roidea on the one hand and Didelphoidea and
Borhyaenoidea on the other are not far separate
and subject to very similar pressures, and have
been given equal ecological opportunities, a
commonly -held mosaic of characters will per-
sist giving rise to many examples of pairs of
species in these different phyletic lines: pairs
which more closely resemble each other than
they do their less distantly separated relatives
within their own lines. An example of this sort
may well be the very strong resemblance between
Thylacinus and Miocene Borhyaenidae, and it
is probably also the reason for the result ob-
tained by Wood (1924) who, when he compared
forty-nine characters of Miocene Borhyaenidae,
modern Dasyuroidea (including Thylacinus )
and modern Didelphoidea, only obtained one
“character” which unequivocally separated
Dasyuroidea from the rest combined. This
“character” was the location of dasyuroids in
Australia while the other two superfamilies
were confined to the Old World and the Americas
(seventeen characters separated Didelphoidea
from the other two, while five separated
Borhyaenidae) .

Australian Marsupicarnivora

Superfamily Dasyuroidea

The native cats and their allies are a com-
plex group. They are entirely Australo -Papuan
and exhibit all grades of carnivorous specializa-
tion from those of the eaters of very small prey
(e.g., insects and other arthropods, small mice
and birds) like Planigale and Sminthopsis to
those of highly specialized mammal killers and
carrion feeders like Sarcophilus and Thylacinus.
The transformation through the degrees of
carnivorous specialization in normal members
of the family takes place without reference to

103

generic distinction and is closely connected with
increase in the size of the body (Bensley 1903.
p. 91). Among Dasyuroidea there are highly
specialized forms which have become so modi-
fied that their immediate relationships are un-
known. These are Thyladnus, the Tasmanian
Wolf and Myrmecobius the Banded-anteater
(termite-eater). These two genera are usually
placed in separate major taxa from the others
in a non-committal fashion; I employ Myrme-

cobiinae and Thylacinidae here in this man-
ner. Their relative status as subfamily and
family is entirely artificial and is adopted be-
cause, subjectively, the specializations of Myr-
mecobius for termite eating are more under-
standable as a development of typical dasyuroid
structure than the obviously widely divergent
yet parallel carnivorous specializations of
Thyladnus which contrast strongly with those of
Sarcophilus and other Dasyuridae. These paral-

Fig. 5. — Type specimens of fossil TilylaciJius . (a) Syntypes of T. spelaeus lOwen); Brit. Mus. iN.H.j Nos. Geol
Ml 0800, M10801. (b) T. major Owen; specimen unknown fig. from Owen. (O T. rostralis De Vis. Qd. Mus No
F 730 (photograph by courtesy Qd Mus.). (a) and (c) to same scale; scale of <b) unknown.

104

lei specializations probably reveal an early
divergence of two predaceous phyla within
Dasyuroidea.

It might even be reasonable to give the Thyia-
cinidae superfamily rank and regard their rela-
tionship to Dasyuroidea as being the same as
that between Didelphoidea and Borhyaenoidea.
However, to date, only one genus and probably
only one species is known and it seems best to
leave things as they are until further material
shows that the lineage of Thylacinus is as old
and as diverse as that of the Borhyaenoidea.
Stirton et al (1961) suggest that a specimen from
the Etadunna Formation of central Australia
( ?Mioc, Ludbrook 1963a) may possibly be a
dasyurid ancestor of Thylacinidae and, if this
should prove to be correct, family rank will prob-
ably be sufficient (but see p. 108).

Thylacinidae Bonaparte, 1838

Only one genus of Thylacinidae is known and
the earliest specimen yet discovered came from
the upper Pliocene Chinchilla Sand of the Darl-
ing Downs of Queensland. Today it is confined
to Tasmania where the only living species (T.
cynoceplialus ) is rare and possibly verges on ex-
tinction.

During the Pleistocene Thylacinus was wide-
spread, occurring in Tasmania, continental Aus-
tralia (where its remains are known from the
fluviatile Diprotodon Beds of the Darling Downs,

various cave deposits in Victoria, New South
Wales, South Australia and in south-western
Australia as far north as the Namban River),
and even in New Guinea where a single mandible
was found recently by Dr and Mrs R. Bulmer
among pebble tcols and flakes in an archaeolo-
gical excavation near Kiowa in the Eastern
Highlands (Van Deusen 1964).

Size, or proportional difference, has been the
main criterion used by authors to differentiate
the various described species of Thylacinus .
Thus, T. breviceps (which is almost certainly a
female and undoubtedly a juvenile specimen:
Aust. Mus Sydney No. 774) was described by
Krefft (1871) as a small species of modern
Thylacine; and T. spelaeus* of the Pleistocene
was described by Owen (1845) as a larger species
than T. cynocephalus, and T. rostralis likewise
by De Vis (1894). De Vis claimed for his species
that its proportions differed so much from those

* The status of the name T. major Owen 1877 is un-
certain. The species was figured (Foss. Mamm. Aust.
PI. V.) from the hinder part of a left mandibular
ramus with M 2 , 3 , 4 . M ;{ and M 4 are drawn as having
very reduced talonids and are quite unlike those of
Thylacinus while M 2 is an ordinary Thylacinus tooth.
Lydekker (1887, p. 264) says that T. major was ap-
parently given inadvertently for T. spelaeus and that
the specimen illustrated by Owen was apparently
drawn from the last three true molars of Sarcophilus
laniarius added to the hinder part of a mandible of
T . cynocephalus. I found no specimen of Thylacinus

corresponding with Owen’s figure in the collection
of the British Museum (Nat. Hist.).

TABLE 1.

Thylacinus: Comparison of adult modern Tasmanian population with Western and Eastern (including specimens
from New South Wales, Victoria and South Australia east of the Nullarbor) Cave-fossils

x Block 1 Block 2 Block 3

Character

M odern Thylacinm cynocephalus

Western cave-fossil Thylacinus

Eastern cave-fossil Tlu/lacinus
( — spelaeus)

Ref.

No.

Morphological

Name

X

mm

Observed

range.

mm

n

s.

MIII1

V

X

mm

Observed

range

mm

n

s

mm

V

X

mm

Observed

range

mm

n

s

mm

V

1

M 2 pr.-me....

1 5 • 3

130 10-0

52

0-9

5 • 76

13-0

11 4 160

12

M

8 • 38

15-6

14-3-17 • 1

4

1-8

11-28

»>

M 3 pr.-me.

17-8

100 19-7

40

1 -2

6-50

15-4

13-2 170

9

1 - 1

7-41

17-6

15-8-20-2

3

2-3

13-20

3

M 1 4 length ....

45 • 5

41 * 7 50 1

39

2-8

6 17

40-0

36-8-45-9

7

3-1

7-78

44-9

41 -4 50-6

3

5-0

11-16

4

P 4 length

1 0 • S

9-2 11-9

50

0-6

5 • 27

10-7

9-3 12- 1

8

1 • 1

9 • 79

11-2

10-1 12-3

4

l-l

9-81

5

P, (.’length...

49 '4

41 1-59-4

47

4-3

8-76

49-9

38-0 57-9

4

9-1

18-20

52 • 6

50 • 9 54 - 3

2

2-4

4-58

6

M 4 length

15-7

13-917- 1

46

1 -0

6-20

14-9

12-8 16-9

8

1-4

9-46

10-5

14-4 18-6

4

2-0

12-42

7

Mand. Length

1 75

142-206

47

16

9 • 30

163

124 189

5

25

15-17

159

153-165

2

9-0

5-37

8

Ht. at P 4

30*8

25 ■ 8-37 • 4

49

2-8

9 * 12

29 ■ 9

24-4 34-8

7

3-2

10-64

29-7

26-6-33-7

3

3-7

12-3

F t Test

Character

Variance Ratio

Comparison

of Means

Ref.

No.

Morphological

Name

Between blocks
1 and 2 above

Between blocks
1 and 3 above

Between blocks 1
t,

and 2 above
P

Between blocks
t

1 and 3 above
P

1

M- pr.-me.

1 -79

4-73**

8*1119

<0-01

0-6467

>0-5

2

M 3 pr.-me.

1 -03

3 • 91 *

5-94

<0-01

0*273

>0-5

3

M 1 - 4 length

1-23

3- 18*

4-69

<0*01

0-338

>0-5

4

P 4 length

3-41**

3-70**

0-2471

>0-5

1 • 272

0*5 01

5

P 4 -C length

4-41**

3 *23

0-1866

>0-5

0 • 9029

0-5 (.)■ I

6

M i length

2-12

4-38**

1 - 9636

01 0 • 05

1 - 429

0*5-0- 1

i

Maud, length

2-32

3 62

1 • 455

0-5-01

1-356

0* 5-0*1

a

Ht. at P 4

1 -28

1-69

0-7802

0 '5-0*1

0-6482

>0*5

**

Significant at 5% level.

105

Significant at 1% level.

TABLE 2.

Thylacinus : Sexual dimorphism in modem adult Tasmanian population

X

Block 1

Block 2

F

t Test

diameter

modern Thylacinus ?

modern Thylacinus 3

Ref.

No.

Morphological

Name

X

mm

Observed

Range

mm

n

s

mm

v

X

mm

Observed

Range

nun

li

s

nun

V

t

P

1

M- pr.-me.

14 4

13 (5-15 *3

12

0-5

3-80

10 -0

15-0-10-0

10

0 3

1-85

3-409*

8*092

<0-01

2

M 3 pr.-me.

Mid

10-1 17 -7

11

0 - 5

3-18

1 9 • 0

18-3 19-7

9

0-0

2-95

1-110

9 • 542

<0-0L

3

M‘-‘ length

42 0

41 -7-44-1

9

0 • s

1 -82

48 - L

46*3 50 1

9

1-3

2-71

0 • 283

10-878

<0-01

4

P, length

10 2

9-2-10-7

13

0-f>

4-48

11-2

10-9 11*7

10

0*3

2-03

3-349*

(5 • 407

- ; 0-01

5

Id • f>

41-5 50-9

11

2 • <»

5 - 53

53 ■ d

51 -7-550

8

1 -3

2-5

3-074*

5-151

< 0-01

6

Al 4 length

14-8

13- 9-10-2

9

0-7

4-59

10-4

151 17-0

9

0*0

3-67

1 • 270

7-825

0*01

7

Marnl. length

1(M

154-173

10

0

3-81

189

175 198

9

8

1 - 33

1-778

8 • 407

<0-01

8

Hi. at. I\

28 • 7

20 • 7-29 - 9

11

1 -0

3 • 45

33-2

31 1-30-0

9

1-0

4-77

2-501

7-107

<0-01

* Significant at 5% level.

of T. cynocephalus that it probably also differed
from T. spelaeus since Owen merely regarded
this as a larger Thylacine than T. cynocephalus.
De Vis said that both T. rostralis and T. cyno-
cephalus occurred in the Chinchilla Sand.

The phenomenon of “gigantism” is as marked
among the Australian Pleistocene forms as it
is in other continental faunas and it has gen-
erally been believed that Thylacinus of the
Pleistocene is an example of it. However, records
of the presence of T. cynocephalus throughout
these deposits is a complicating factor; fur-
thermore, in recent years, collections of thy la -
cines from the caves of south-western Australia
have been made by the Speleological Group of
the Western Australian Naturalists’ Club (and
m particular by D. L. Cook and R. Howlett.
see Cook 1963, 1963a) and these have resulted
in the discovery of further material of a very
small thylacine of which fragments were first
collected in the Mammoth Cave of Western
Australia by L. Glauert in about 1909.

It is possible that these reputedly large and
small thylacines are merely large and small
individuals of a single species and in order to
examine this I have, over a considerable period,
collected data on thylacines from the British
Museum (Nat. Hist.), the Oxford University
Museum, the Australian Museum, the National
Museum of Victoria, the South Australian
Museum, the Queensland Museum, the Queen
Victoria Museum Launceston, and the Western
Australian Museum. In all, some sixty modern
specimens which were collected in Tasmania
have been measured in an attempt to gain an
indication of the range of variation which might
be expected in a fossil population of this genus.
It is not possible to give more than a prelimi-
nary account of this study here, a study which
is complicated by the fact that many specimens
are mutilated (particularly the fossils) so that
all characters measured are present in only a
few individuals and, in the case of the fossils,
very few even possess comparable characters.
So far the work has been confined to the analy-
sis of single characters but it is hoped ulti-
mately to subject the data to multivariate
methods. Here, in order to illustrate the con-
clusions to date, a selected group of eight sepa-
rate characters is discussed (see Tables 1 and 2).

In Thylacinus cynocephalus variances are
high. Coefficients of Variation (V) range from
about 5-10 but there can be no doubt that only
one species is involved in this single living Tas-
manian population. The effects of growth have
been eliminated as far as possible by selecting
characters 1, 2, 3, 4, and 6 because these are
dental characters taken on the enamel caps of
permanent teeth. In the case of characters 5,
7. and 8 these are certainly likely to be effected
by growth since they all include bone, but even
these only include measurements from indivi-
duals in which dental development is complete.

Excluding the material from the deposits of
the Darling Downs, and also the New Guinea
specimen, it seems reasonable to divide the
remaining fossil specimens into two samples.
One of these (Block 3, Table 1), that from the
caves of New 7 South Wales, Victoria and eastern
South Australia includes the type specimens
of T. spelaeus from the Wellington Caves,
N.S.W. (Brit. Mus. (N.H.) Geol. Nos. M/10800
and M/10801; Fig. 5) while the other (Block 2)
is from the caves of the south-west.

The problem of analysis of these samples is
complicated by the fact that Thylacinus cyno-
cephalus is strongly sexually dimorphic (Fig. 6a
and Table 2) and the data have had to be
examined to see whether the samples depart
sufficiently from normality to render the use of
statistical tests invalid when these are con-
structed on the assumption that they are drawn
from populations with normal distributions. It
is found that only in the case of mandibular
length w : ould there be any doubt about the
validity of applying normal statistical pro-
cedures.

Tests of variance ratio between modern and
fossil samples (Table 1) show 7 significantly dif-
ferent variances. Since the modern population
is undoubtedly of a single species, it is tempting
to conclude that more than one taxon is in-
cluded in each fossil sample and that this is
responsible for the high variance. However, in
the situation wffiere strong sexual dimorphism
occurs (and in particular where the sex ratio
can be assumed to be 1:1, or close to it) the
variance cannot be expected to decrease with
sample size because both sexes are likely to be
represented. Dr. R, D. Hughes, of the Australian

106

n

12 -

modcrn

Th ylacinus

Wastarn
cava fossil
Th vlacinus

9 -

6

5

ino'flP'flpiflOip'Piflo

T i • i • i 1 • 1 1 •

,L,_C\iC\Jpf2^tjjdt£(£?0
M 2 Class intervals (mm)

50

40

35

M

1-4

M 4 Class intervals (mm)

NOT SEXED

Fig. 6a. — Histograms illustrating the distribution of
classes of two characters of modern Thvlacinus and
Western Cave-fossil Thylacinus. Where sexes are known
these are indicated by different shading.

Fig. 6b. — Ranges of values of four characters of three
populations of Thylacinus. Tas., Tasmanian modern;
W.A., Western Cave-fossils; E.A., Eastern Cave-fossils.
Vertical lines represent observed range. Horizontal lines
represent sample means, stippled blocks superimposed
upon Tas. and W.A. samples represent one standard
deviation outside the confidence limits of the means.
M 2 , are diagonal measurements across metacone and
protocone, and M 1 - 4 are measurements of total

antero-posterior length.

National University, and I are examining this
problem in greater detail and the results will be
reported fully elsewhere, but, at present, we are
of the opinion that the size of the variances in
each of these small samples of fossils are those
which might be expected in samples drawn from
single species as exemplified by the much larger
Tasmanian sample. We conclude that there is no
longer any need to postulate sympatry between
T. cynocephalus and T. spelaeus in the Pleisto-
cene of eastern Australia.

Since individuals of T . spelaeus are supposed
to be larger than those of T. cynocephalus their
means should differ. Comparison of means by
Student’s t test (Blocks 1 and 3) shows no sig-
nificant differences and it can only be concluded
that, so far, this work does not support the
separation of these two species.

Comparison between the means of T. cyno-
cephalus and Western Cave thylacines reveals
a different situation. Here t tests (between
Blocks 1 and 2) are significant in characters
1, 2, 3 at the 1% level and approaches signi-
ficance at the 5% level in character 6. All that
now remains is to decide whether the difference
is sufficient to warrant the use of a specific or
subspecific name. Various approaches to this

107

kind of problem are possible (see Appendix)
and it is sufficient to say that using a test
which is compounded of the statistical discus-
sions of the 75% Rule of Amadon and the 90%
Rule of Mayr, Linsley and Usinger, the charac-
ters examined do not support separation even
at a subspecific level at the present time < see
Fig. 6b for a graphic representation of this).
However, although recognition by name is not
justified there is no doubt that in the Pleisto-
cene of Western Australia there existed a popu-
lation of Thylacinus cynoceyhalus which on an
average contained smaller individuals than the
modern form (and by inference its eastern
Pleistocene representative).

The status of T. rostralis and the thylacines
from the two Darling Downs faunas remains un-
solved. The Holotype of T. rostralis (Queens-
land Museum F.730, Fig. 5) from Ellangowan
nr. Cambooya in south-eastern Queensland is
from the Pleistocene fluviatile deposits and lies
at the upper end of the range of variation of
T cynoceyhalus (Character 1 of my Table
1 15.5, 2 17.9, 3 47.4, 4 12.0,

5 58.5, 6 17.1, 7 203, 8 38.9) and

the only other specimens from the same beds
(Queensland Museum F.3742 from King’s Creek,
Clifton, 4 14.1. 5 67.5, 8 approximately

43- Brit. Mus. (N.H.) 35973 from Gowrie Creek,

1 17 . 1 , 2 20.2) also have values that are

very high. It is thus possible that T. rostralis
is a distinct form. A value of 18.6 for a single
isolated last lower molar (character 6) from
the Pliocene Chinchilla Sand also suggests that
this form may possibly be distinct as well.

The New Guinea thylacine falls outside the
known range of measurements of modern
thylacines in at least one character (length of
p 3 ). Hobart Van Deusen has a full description
of the specimen in preparation for the Novi-
cates series of The American Museum of Natural
History. It is hoped that a C 14 date will also
be available.

Origins. — The origin of the Thylacinidae is
not yet known although Stirton et al (1961,
p. 35) have suggested that a dasyurid fossil
(comparable in size with Dasyurus viverrinus )
found in Lake Napakaldi in Central Australia
may possibly be ancestral because it possess
three premolars, graded from front to rear, and
lacks a metaconid on Mi. But other Dasyuridae
(see below) possess a very reduced metaconid
on M, and others have three premolars graded
from front to rear and it seems far more likely
that a more significant feature of the thylacinid
lower dentition, and the one in which it differs
from that of advanced Dasyuridae like Sarco -
pliilus, lies in the development in thylacine lower
molars of a main posterior shearing crest along
the ridge from protoconid to hypoconid; this
bypasses the metaconid. In the dasyurid line

Pig 7.— Cusps of the upper and lower maxillary teeth
of Dasyuroidea. Right upper tooth rows and left lower
rows illustrated. Note main differences between Thy-
lacinus and others are (a) uppers: Thylacinus has poorly
developed stylar cusps: (b) lowers: Thylacinus has main
shearing crest to hypoconid from protoconid; Dasyurus
and Sarcophilus have this to metaconid (pa paracone,
pr protocone, me metacone, prd protoconid, pad para-
conid, med metaconid, end entoconid. hyd hypoconid,
hyld hypoconulid). Ridges are shown in continuous
line, valleys in broken line.

108

Fig. 8. — Holotype of Glauccdon ballaratensis Nat. Mus. Viet. No. P16136 Magnification X2. (Stereo-pair by courtesy

Nat. Mus.)

the shearing ridge which is developed is that
which runs from the apex of the protoconid;
through the tip of the metaconid to the ento-
conid; in its fully developed form the metaconid
is absorbed without trace into this ridge. In both
kinds of specialization the metaconid is reduced
as its end product, but in the thylacinid line
the shear is developed by accentuating the
talonid and the hypoconid crest, in the dasyurid
line it is developed by specializing the meta-
conid crest and reducing the talonid (see fig. 7).

Dasyuridae with three premolars which in-
crease in size from front to rear include Smin -
thopsis rufigenis, S. ferruginif rents, and Murexia
maxima (Tate 1947, p. 151, table 5). The meta-
conid of Mi is so reduced as almost to be ab-
sent in the newly described Antechinus rosa-
mondae Ride, 1964a (specimen W.A.M. No.
M 3421).

Dasyuridae Goldfuss, 1820
Dasyurinae

The subfamily Dasyurinae includes a wide
range of native cats and their allies including
the so-called marsupial mice. It excludes the

Myrmecobiinae (the marsupial Banded-anteat-
er). From morphologically greatest to the least,
species range from the massive Tasmanian Devil,
Sarcophilus, through the more typical Das -
yurus *, the arboreal treeshrew-like Phascogale ,
their terrestrial relatives Dasycercus, the smaller
Antechinus and Sminthopsis , to the minute
Planigale.

Few fossil dasyurines are known. The oldest
is probably the as yet undescribed form men-
tioned by Stirton et al (1961, p. 35) from the

* I follow Simpson 1945 in rejecting Pocock's (1926)
concept of four generic names for five species of
very similar spotted native cats. These additional
‘•genera” Dasyurops , Dasyurinus, Satanellus, are
based in the main upon slight variations in the
development of the great toe and in the nature of
the footpads. If similar reasoning was to be applied
to the related genus Sminthopsis (which mercifully
has so far evaded the attentions of the splitter) we
could have in Western Australia alone up to six
“genera” of these. Pocock is also well known as a
splitter of eutherian cats ( see Simpson 1945, pp.
231, 2 for remarks on this and a rather diverting
comment on the philosophy of splitting as applied
to cats) but it is not generally realized that the
present situation in marsupial “cats" is also due to
Pocock’s attentions.

109

Etadunna (?Mioc.) formation. This unnamed
species has been discussed above under Thyla-
cinidae.

The most important fossil dasyurid yet
described is Glaucodon ballaratensis Stirton.
1957; its age is not known but is possibly Plio-
cene (see below). Several Pleistocene species
of modern genera have been described. These
give little information of phylogenetic import-
ance except perhaps of gigantism.

Glaucodon ballaratensis Stirton, 1957

Glaucodon is known from a single almost
complete right mandibular ramus which lacks all
teeth but the first and fourth molar (fig. 8).
It was found in a well at a depth of fifty feet
near Ballarat and reached the National Museum
of Victoria in 1914. Gill (1957, p. 191), in
describing the site, commented that an intensive
study is needed before the fossil can be dated
with accuracy but the materials in which the
well was sunk appear to have been laid down
subsequent to the eruptions of the volcanoes
which are represented today by Mount Mooroo-
kyle and McRorie’s Hill. He considers that the
deposit is either Pleistocene or Pliocene.

Irrespective of its as yet underter mined age,
the fossil is of great interest because it is mor-
phologically intermediate in its few characters
between the “normal” large Dasyuridae (e.g.,
Dasyurus maculatus) and the more specialized
short-faced Tasmanian Devil Sarcophilus. In
the skull, these two sorts of Dasyurinae differ
principally in the characters associated with
the greatly shortened face of Sarcophilus
coupled with its more extreme development of
“carnassial” dentition of the pseudocreodan
type ( see Butler 1946); in Sarcophilus great
emphasis has been placed on the paraconid
crest (particularly between paraconid and proto-
conid) and the talonid and metaconid are re-
duced as compared with the ancestral didel-
phid or even dasyurid molar type (fig. 7). Short-
ening of the face is a trend which is often
repeated among Dasyurinae and it commonly
finds expression in reduction and crowding of
premolars; examples of it are Antechinus rosa-
mondae which may be compared with A.
flavipes, and Dasycercus cristicauda which may
be compared with Phascogale. Even Dasyurus
has lost one of the original number of three
premolars so that, even in this long-faced genus,
shortening has occurred to some extent.

The molar teeth of Glaucodon are shown in
stereoscopic pair in fig. 8.

In Mi, the protoconid is large, there is no
paraconid, the metaconid is a small but distinct
cone adpressed to the protoconid. The talonid
is about one-third of the total length of the
tooth. The hypoconid is well developed, the
entoconid is small, and there is possibly a small
hypoconulid between them. Since the metaconid
of M, is barely present in Sarcophilus and it is
often almost completely reduced in D. macula-
tus , this tooth may be somewhat more primi-
tive in this respect than either; however the
metaconid is closely adpressed to the protoconid
and the tooth is particularly reminiscent of that
of Sarcophilus in its massiveness and lack of
lateral compression.

In M ( the protoconid is large and tall and
the paraconid is also well developed. The meta-
conid is somewhat smaller. The main shearing
crest of the tooth comprises the ridges from
the apices of the paraconid and protoconid.
and that between the protoconid and the meta-
conid (which is set lower than the paraconid).
The talonid is much reduced and consists of
little more than a hypoconid. The disparity
between the heights of the paraconid and meta-
conid (compare 3 and 4 of Table 3) is a Sarco-
philus- like character not seen in D. maculatus
where these two lingual cusps are more or less
equal. Although, as mentioned below, the
talonid of the M., of D. maculatus may be re-
duced as much as in Glaucodon, in D. macula-
tus the hypoconid and entoconid are separate
and even a hypoconulid may be present. In
Sarcophilus the talonid is usually represented
by little more than an isolated cuspule (possibly
the hypoconid) posterobuccal to the metaconid.
a condition morphologically more like that of
Glaucodon although advanced on it.

There are few characters in Glaucodon upon
which to base detailed comparison. However,
these are compared with Sarcophilus and Dasyu-
rus maculatus in Table 3 (and in figs. 7 and 8).
In these few characters the more highly speci-
alized Sarcophilus differs from the less speci-
alized D. maculatus in having:

(a) An enlarged molar trigonid in compari-
son with its talonid; in particular it
has almost totally reduced the talonid
in M, (2 of Table 3). The trigonid of
M» is also compresed bucco-lingually.
(b» A paraconid which is increased in
height relative to the protoconid (3 of
Table 3).

TABLE 3.

Comparison of characters of Dasyurus maculatus ,
Glaucodon and Sarcophilus.

/>.

mttrulflfus

(IhiH-

oodoft

Sarr.ophUm

n.

range %

%

range

n.

l. Total length of M, i
a* % of length of
mandible from an-
terior til* of posterior
edge of masseteric
shell' ..

3

29 -8-^*2

39-6

37 *8- -42 -7

7

2. length of M , exclud-
ing talonid as % of
total length of M, .

5

(>7-7-77- 1

77-0

92 • 0-97 • 4

>

3. .M , height of paraconid
as % of protoconid

5

52 0 63*3

50- 1

03-4-72-2

0

4. M , height of meta-
conid ns % of proto-
eonid

5

52* 0-00 - 7

42-7

31 -3-38-0

0

0. Depth of mandible
expressed as % of
length of mandible
from anterior tip to
posterior edge of
masseteric shelf

4

15 7-16 -7

20 • 5

217-23*0

7

0. Depth of mandible

expressed as % of
molar tooth row

3

49-8-50-2

51 9

51 *5-68*9

7

110

(c) A metaconid which is reduced in height.
This is almost a vestige in some molars
(4 of Table 3>.

(d) Incisor alveoli which are much crowded
so that the alveolus of the second inci-
sor is raised above the plane of the
other two ( this condition occurs in
D. rnaculatus but it is not as well
marked) .

<e> Premolar alveoli which are crowded,
each pair being set at an angle to the
long axis of the ramus. There are no
diastemata between them.

(f) The depth of the mandible is propor-
tionally greater in comparison with its
length (5 of Table 3 ).

(g) The molar tooth row is proportionally
longer in comparison with the length
of the mandible (1 of Table 3).

Glaucodon compares with these in the follow-
ing way:

(a) The trigonid of Mi is compressed bucco-
lingually and the talonid is more re-
duced than in most D. rnaculatus (al-
though 1 have measured a specimen of
D. rnaculatus which had a talonid
which was reduced to a similar degree
— see Table 3).

(b) The height of the paraconid of M.. as
compared with that of the protoconid
is within the range of D. rnaculatus and
not of Sarcophilus.

< c ) The metaconid is greatly reduced to-
wards the condition in Sarcophilus.

(d) The incisor alveoli are as in Sarcophilus.

(e) The premolar alveoli are as in Sarco-
philus.

(f) The depth of the mandible, expressed
as a percentage of its length, is inter-
mediate between that of D. rnaculatus
and Sarcophilus.

(g) The length of the molar tooth row, as
a percentage of the length of the
mandible, is within the range of Sarco-
philus and not of D. rnaculatus.

In all, as far as its known characters go, there
is little doubt that Glaucodon can be regarded
as structurally ancestral to Sarcophilus. In some
characters it still lingers on the Dasyurus side
and in others it is clearly Sarcophilus- like. In
one character (the well developed conical but
adpressed metaconid on Mi) it is like neither.

Dasyurus affinis and D. bowlingi

D. affinis McCoy (?1862) and D. bowlingi
Spencer &; Kershaw (1910) are the only two
species of fossil Dasyurus which have been de-
scribed to date.

The name D. affinis was published in Quarter
Sheet N.W. of the Geological Survey of Victoria
in a note which merely states u Dasyurus affinis
(McCoy) New Species nearly as large as D.
rnaculatus but differing in proportions”. The
material comprises two syntypical left mandibu-
lar rami (Nat. Mus. Viet. P 7426, P 1510D which
have since been figured by Gill (1953b). Gill
advises me that the precise locality of the cave
from which the specimens come is in basaltic

tuff, under a flow of Newer Basalt on a small
tributary near the head of Toolern Creek, 4$
miles S. cf Gisborne and 11 miles S.W. of Couan-
galt Post Office The material is Holocene and
no revision of the status of the species has been
published to date.

D. bowlingi is from the dune sands of King
Island and Deal Island in Bass Strait and the
measurements and other data presented by
Spencer & Kershaw of a good series of speci-
mens suggest that the species was much larger
and also morphologically distinct from D. macu-
latus with which it was sympatric. The geo-
logical age of the material is not known but
Spencer & Kershaw infer that the species per-
sisted into modern times and was seen alive by
Peron during the Baudin expedition of 1801.

Scarcophilus laniarius and S. prior

S. laniarius (Owen, 1838) and S. prior De Vis
(1883) are the only two species of fossil Sarco-
philus which have been described to date al-
though Stirton (1957, p. 131) records an unde-
scribed species from the Pliocene at Kalamurina.
on the Warburton River, South Australia, which
he says is closer to S. laniarius than to S. har-
risi Boitard (the modern species).

It is widely accepted that S. laniarius is a
gigantic Pleistocene form of the modern species
and Gill's (1953, p. 87) statement of mandibular
measurements 4 ‘from 15 to 50 per cent, larger
than the average extant Sarcophilus ” suggests
that, unlike the position in Thylacinus spelaeus
and cynocephalus, the distinction between the
two should be maintained.

Sarcophilus prior De Vis, 1883 comprises only
the proximal articular surface of a right tibia
with a little of the shaft. It is from ‘‘Darling
Downs” and is somewhat larger than the cor-
responding bone in the modern specimen of
Sarcophilus with which De Vis compared it.

Myrmecobiinae

There are no fossils and nothing significant
has been added to the problems of the affinities
of Myrmecobius since the beginning of the cen-
tury. During the 19th century Myrmecobius was
believed to be an unmodified survivor from the
Jurassic with affinities with Mesozoic mammals,
but the modern view which can be directly at-
tributed to Bensley is that its unusual dental
characters (and among these its high dental
formula) have been derived from those of
normal Dasyuridae. Bensley (p. 100) said that
many of the characters of the incisors, canines,
and premolars “which appear at first sight to
be primitive, are repeated in the Peramelidae,
where they are undoubtedly the result of retro-
gression. The patterns of the lower molars are
directly derivable from those of the smaller Das-
yurinae”, and of the upper molars Bensley is also
of the opinion that the basic pattern which can
be recognized running through these very vari-
able teeth is approximately that characteristic
of normal Dasyurinae as well. Great dental
aberration is characteristic of myrmecophagous
mammals (e.g. Proteles , the Aard-wolf).

For a while, the phylogenetic picture was
somewhat confused by the description of the
fossil Myrmecoboides (Gidley, 1915) from the

Mid Palaeocene of North America. This now
turns out to be a eutherian insectivore and may
be disregarded in this context (Simpson 1945, p,
172).

PERAMELINA

Even less is known of the Peramelina as fossils
than the Australian Marsupicarnivora with the
exception of enormous numbers of specimens of
modern species in the Pleistocene cave deposits
of various parts of Australia.

The modern bandicoots appear to fall into at
least two distinct groups, one of these includes
only the Rabbit Bandicoots ( Macrotis ) while
the others are obviously fairly closely related,
with the Pig-footed Bandicoot ( Chaeropus > as a
highly specialized derivative. Peravneles tenui -
rostris Owen. 1877 and Perameles wovibeyensis
Broom, 1896b have both been described from
the Pleistocene of New South Wales.

There has been no revision of the status of
P. wovibeyensis but, from what I know of the
fauna from the Wombeyan Caves, N.S.W. which
contains it (Ride 1960 > I would not be surprised
to find that it is identical with a modern form.

Perameles tenuirostris (Owen 1877, PL V., figs.
10, 11) is from the Wellington Caves, N.S.W.
and is stated by Lydekker (1887. p. 255) to be
identical with the modern P. nasuta.

Ischnodon australis Stirton (1955) is the only
Tertiary species which has been described.

Isclincdon australis Stirton. 1955

This fossil bandicoot which comprises a single
fragmentary mandibular ramus with two pre-
molars and two molars in position was described
from the late Tertiary Palankarinna Fauna from
the Mampuwordu Sands at the Woodard Quarry
near Etadunna Station, South Australia. Stirton
was unable to determine the exact relationship
of the bandicoot (see Fig. 9> but he suggested
that it was nearer to the rabbit bandicoots
• Macrotis) than to other Peramelidae. In par-
ticular, he considered that the presence of a
somewhat reduced paraconid and reduced hypo-
conulid on the molars indicated relationship to
Isoodon and Perameles but the fact that they
were reduced although present suggested Macro -
tis relationships where (he implies) the para-
conid is lost.

Comparison of Stirton’s figure and measure-
ments with Macrotis in the collection of the
Western Australian Museum reveal that his
generic diagnosis includes some specimens of
Macrotis lagotis because some of the details of
dentition which he regarded as being of generic-
value, i.e. the presence of a much reduced para-
conid and hypoconulid, and the presence of a
small stylar cusp antero-buccal to the hypo-
conid, are very variable in that species. In fact
all these characters of Ischnodon are present in
specimens WAM No. M 1399 from Bridgetown
in south-western Australia, and WAM No. M 632

from Laverton. W.A., yet, as Stirton has im-
plied, other specimens lack the paraconid com-
pletely (e.g, WAM M 898 from Wiluna).

The only feature of Ischnodon at present de-
scribed which I consider to be of significance
are its low crowned molars w r hich are lower than
those of any rabbit bandicoot known to me. For
the present, the generic distinction can only be
maintained in this respect.

DIPROTODONTA

The Diprotodonta is the most diverse, and to
me most structurally interesting, of all the mar-
supial orders. It is generally thought of as a
herbivorous order but it actually contains a com-
plex series of radiations within itself. Primitive
modern members of it (Bensley 1903, p. 124) are
insectivores (e.g. the small modern phalangers
such as Distoechurus and Cercartetus, and in
addition it contains at least one rapacious car-
nivore {Thylacoleo) , and several smaller modern
forms with strong carnivorous tendencies {Bet-
tongia). There are many herbivores of brows-
ing and grazing habit and there is even a fos-
sorial bear-like animal (the Wombat). Among
the browsers and the grazers were the heavy
’‘pachydermatous" Diprotodontidae, the gazelle
and cervid-like Macropodinae, and their close
relatives the browning- adapted bovid-like Sthe-
nurinae {see Ride 1959 discussion of Procopto-
don ). Some of these phyla have even become
convergent upon each other as have the diproto-
dontid palorchestines upon kangaroos (Woods
1958, Bartholomai 1962).

The known Pleistocene fossils, of which there
are many, are mainly of large and moderately
large animals and help us to fill out the details
of this great radiation; the great numbers of
them and their diversity gives to Diprotodonta
a much better appearance of balance of know-
ledge between fossils and living genera. How-
ever these appearances are still misleading be-
cause we have no lineages and thus can talk
only of evolutionary lines by arranging Recent
or Pleistocene animals in order of primitiveness
and arguing from them.

The sub-taxa of Diprotodonta which may be
distinguished by this means comprise five and
possibly six distinct phyletic lines. These are
the Phalangeridae as represented by the modern
Australian Possums, Cuscuses. Flying Possums,
Pigmy Possums, etc. and Pleistocene forms in-
cluding Thylacoleo : the Macropodidae as repre-
sented by the modern Kangaroos. Wallabies and
Rat Kangaroos (these latter may be widely sep-
arate from true Macropodidae and, in modern
taxonomic practice it is usual to separate them
off as a subfamily Potoroinae) ; the Diprotodon-
tidae as represented by the giant extinct forms
Nototheriuviy Diprotodon, Palorchestes, etc.; the
Vombatidae which comprises the Wombats and
their extinct giant relatives; and the Wynyardii-
dae which contains only Wynyardia bassiana of

Fig. 9 (opposite).

Fig. 9. — Upper figure: Holotype of Ischnodon australis after Stirton 1955. Stereo-pairs of mandibles of Macrotis
lagotis for comparison of paraconid. hypoconulid and cuspule between protoconid and hypoconid. (Upper pair
WAM No. M632, Laverton. W.A. Lower pair WAM No. M1399, Bridgetown, W.A.) Upper pair show dP* in position.
In lower pair P* is erupting. Stereo-pairs and line drawing not to same scale.

112

hyld

ant. cingulum

113

the Oligocene. In addition there is possibly a
separate line containing only the Koala Phas-
colarctos (see Troughton 1959, pp. 75 and 84 for
a discussion of the problems of the relationships
of Phascolarctos ) .

Wynyardiidac Osgood. 1921
Wynyardia bassiana

The oldest diprotodont known to us is Wyn-
yardia bassiana, a remarkably complete and
semi-articulated skeleton from the Oligocene
which was found prior to 1876 at Fossil Bluff
near Wynyard in Northern Tasmania. It re-
ceived numerous mentions in literature (Gill
1957) before being described by Sir Baldwin
Spencer in 1900 (Spencer 1901). Unfortunately,
Spencer was so impressed by its great age (at
that time thought to be basal Eocene) that he
made rather extravagant claims for it on the
basis of a detailed description and discussion of
its characters — a discussion based upon inade-
quate* material of other marsupials.

Spencer concluded that Wynyardia was inter-
mediate between the “Pclyprotodonts" (Marsu-
picarnivora and Peramelina) and Diprotodonts
and moreover that it possessed a number of
characters unknown in any other marsupial.
Subsequently, Wood Jones became interested in
this fossil and ultimately obtained the loan of
Its principal parts for redescription. Unlike
Spencer, he concluded that the fossil had no
non-marsupial characters, that it had no speci-
fically non-diprotodont characters, and that it
must be considered an ally of the phalangers
but probably differed from the modern lightly
built animals in both gait and bodily habit.

Unfortunately, Wood Jones, always a direct
man, demonstrated his disagreement with Bald-
win Spencer’s conclusions in a manner which
has allowed subsequent workers to misinterpret
his results and moreover to misread his conclu-
sions. Thus Wynyardia has come to be regarded
as an animal very much like the modern pos-
sum Trichosurus < see Gill 1957). Admittedly,
Wcod Jones claimed that Wynyardia had many
resemblances to Trichosurus but this claim was
in the main because his results were based
almost entirely upon comparison between the
fossil and a similarly mutilated specimen of
Trichosurus ( see Wood Jones 1931).

Thus a suspicion arose that Wynyardia might
even have been a modern possum which had
fallen into a fissure in the limestone and had
become incorporated into the Fossil Bluff se-
quence (Gill 1957 gives an excellent account of
this ) .

The most important recent work on Wynyardia
has been done by E. D. Gill, Curator of Fossils
at the National Museum of Victoria. By using
physical and stratigraphical methods he re-
examined the provenance of the fossil and con-
cluded that it was undoubtedly contemporary

* Comparison by Spencer with an inadequate series and
insufficient information regarding variation is the
only way in which I can interpret the mis-state-
ments of marsupial anatomical detail which he
makes. The mis-statements do not relate so much
to the fossil (of which his description is excellent)
but in regard to the distribution of various charac-
ters among living Marsupialia.

with the Fossil Bluff fauna and is not a subse-
quent inclusion. Today, this fauna is considered
to be Oligocene! (Gill 1962, p. 249).

The establishment that the fossil was genuine
together with the misinterpretation of Wood
Jones' results has resulted in claims being made
that the Possum (i.e. Trichosurus > has an an-
tiquity dating back to the Oligocene ( see Gill.
1955, p 92). Results of examination by myself (the
full details of which will be published elsewhere)
do not support this view and reveal that, con-
trary to current opinion (which has grown like
a snowball around a slender nucleus of misin-
terpreted results). Wynyardia is a very re-
markable animal quite unlike any known dipro-
todont combining as it does a number of very
primitive marsupial characters with robustly ex-
pressed specializations peculiar to itself. Any
resemblance to Trichosurus which the fossil
possesses is in those characters which represent
relatively unspecialized diprotodonty.

Wynyardia comprises a broken skull (fig. 10),
the inferior border of the left mandibular ramus,
a coronoid process (which has become lost since
Spencer’s examination), the axis vertebra, a
series of nine articulated presacral vertebrae in-
cluding the last thoracic, portions of fused sacral
vertebrae, a large portion of pelvis comprising the
dorsal ramus of the right ischium including the
acetabulum, and a more or less complete right
ileum, one epipubic bone, a more or less com-
plete left femur and a fragmentary right femur,
the left tibia and fibula, together with some
fragments of ribs. At the time of Baldwin Spen-
cer's and Wood Jones' examinations, the skull,
mandibular ramus, long bones of the hind limb,
sacrum, and pelvic fragments had been extracted
from the matrix. After Wood Jones examined
the material, the vertebral column had also been
extracted and was available for examination.
The otic region was still filled with matrix and
glue and I removed these.

That Wynyardia is a diprotodont is established
by the presence, at the anterior end of the man-
dible, of the unmistakable root of a single large
incisor and as Wood Jones pointed out, the
upper incisor alveolus which remains in the
premaxilla is consistent with its having an en-
larged first upper incisor of phalangerid type.
Along the broken dorsal margin of the mandibu-
lar ramus there are four circular structures
which are almost certainly the broken tips of
the roots of cheek teeth, these are followed by
a pathological cavity and what is almost cer-
tainly the alveolus of the last molar. If this
mandible and these “teeth" are then positioned
on the skull it will be seen that the anterior
end of the zygomatic arch lies midway along
the toothrow; this is another diprotodont char-
acter not found in Dasyuridae or Peramelidae.
In these it lies at the posterior end of the tooth-
row. The position of the posterior end of the
palate is indicated by the pterygoids and con-
firms this. The pterygoid fossa in the mandible
can also only be matched in Diprotodonta among
Australian forms.

j Spencer's geological contemporaries regarded it as

“basal Eocene” while Wood Jones followed Chap-
man’s “no older than the Miocene” and Howchln’s

“no older than the Pliocene” (Jones 1931, p. 97).

114

Primitive characters found in no diprotodont
are best seen in the region of the middle ear
(see stero pairs in fig. 10 a, b). Unfortunately,
the bullar wing of the alisphenoid has been
broken away and the tympanic is missing but
sufficient completely undamaged surface on the
post-glencid process remains to show that the
tympanic was not tubular and fused to the post-
glenoid process as in Diprotodonta in spite
of Wood Jones’ statement that the post-glenoid
process and the auditory region do not differ
from that of Trichosurus. He argued that the
smooth surface of the post-glenoid process might
be due to erosion but this is not so; except for
about 3 mm of its postero-mesial end (length
12 mm), the structure is covered by a finished
surface which is quite unlike the ragged can-
cellar appearance of an eroded contact area
irom which a fused tympanic tube has been
torn. On the other hand, I suspect that the
innermost 3 mm which is eroded does represent
the point of contact with the tympanic ring (the
differences between the post-glenoid and tym-
panic region of Wynyardia and Trichosurus can
be se:n in fig. 10 b, d) and if this is so, it could
represent an intermediate stage between the free
ring of Didelphidae and the fused tube of the
Phalangeridae.

The most important primitive character of
the middle ear of Wynyardia is the complete
absence of epitympanic sinuses either anterior
or posterior to the epitympanic recess. Van der
Klaauw (1931. p. 82) says that these sinuses are
absent in mcnotremes and didelphids, while
other marsupials show well-developed sinuses in
the squamosal. The very small epitympanic re-
cess between the periotic and the portion which
remains of the tympanic wing of the alisphenoid
suggests that the alisphenoid bulla was at most
very short and merely shielded the epitympanic
lecess. In fact, in general appearance the whole
region is almost identical with that of modern
Didelyliis virginiana (fig 10 c) except that the
anterior face of the periotic is slightly more
flattened in the fossil which suggests that the
tympanic wing might have contacted it here.
The morphology of the squamosal behind the
eai and its entire relations in the nuchal and
post-glenoid regions are also comoletely un-
phalangerid and can be matched only in Mar-
supicarnivora.

Other characters of a marsupicarnivoran
aspect, although probably of less importance
being inherited by other Diprotodonta, are the
the nasals which are precisely those of
Dtaelphis, the lachrymo-nasal contact, ‘'the great
sagittal crest, the very short post-tympanic
legion, and the position of the sacral articula-
• 101 ^ 1 i 1 This is very far anterior as

m Didelphis, not mid-way to the acetabulum as

i? a ^ an ^ er *^ ae ‘ The triangular cross section
of the ilium is also typically Didelphis- like.

Some of the specializations of Wynyardia
seem to be related to its extremely robust build
and, probably, erect carriage of the body which
must have been somewhat Koala-like. This ro-
bustness is represented in the skull by the depth
of the mandible (only the tips of the roots of
the cheek teeth are present along the broken
dorsal surface so that it seems certain that much

115

is missing of the ramus which is already not
much more slender than the mandible of Tricho-
surus) and by the large areas of attachment of
the pterygoid and masseter-temporal muscle
groups. It is in the post-cranial skeleton how-
ever that adaptations to an erect or semi-erect
posure are most marked. Lumbar vertebrae,
which are typically phalagerid in possessing well
developed slender anapophyses ( hyposphenes > ,
have transverse processes which are horizontal
and form an acute angle with the plane of the
anterior zygopophyses. This acute angle is only
exceeded by the Koala (and approached by the
cuscus ) . The koala and the cuscus are both
animals which habitually adopt an erect posi-
tion. The pelvic girdle suggests the same con-
clusion. The epipubes are very large; in modern
Diprotodonta they are largest in the Koala (ap-
proximately half the length of the femur)* in
Wynyardia even the incomplete epipubis is
greater than half the length of the femur (64.7
mm: 106.4 mm). On the ilium, which is trian-
gular in section, the area of origin of the gluteus
medius is large ( see Elftman 1929 for termi-
nology) which agrees with the observation that
the greater trochanter of the femur (which car-
nes its insertion) is also large. On the other
hand the lesser trochanter is smaller than that
of Trichosurus or Phalanger and this agrees
wefl with the observation that the area of origin
of the iliacus on the ilium is smaller in Wyn-
yardia than in these genera. The acetabulum
is strongly buttressed anteriorly. These fea-
tures would argue (see Elftman 1929, p. 213)
that the animal was able to remain for consider-
able periods in the upright position where the
gluteus medius would be one of the muscles
concerned in its support.

The u bones of the hindlimb are robust and
the tibia is short as compared with the femur
The fibula is also much closer to the tibia in
size than is that of modern phalangerids (e.g.
Trichosurus). However, it is of phalangerid
loim as distinct from that of Macrcpodidae or
Peramelidae (see Barnett <fc Napier 1953. p. 209).
In all, it would seem that the specializations of
the postcranial skeleton and in particular the
robustness of the bones of the hindlimb the
proportions of tibia and fibula, and the short-
ness of the distal part cf the limb in comparison
with the femur, suggest that Wynyardia was a
slow, deliberately moving animal without salta-
tory or cursorial specializations.

Since Wynyardia is so very different from all
known phalangerids— an Oligocene diprotodont
with clear evidence of marsupicarnivoran origins
ol a didelphid kind, but yet highly specialized
in its own peculiar way — I follow Osgood (1921)
m including it in a separate family which will
include those primitive diprotodonts which com-
bine diprotodonty with a primitive middle ear
which dike that of Didelphis ) lacks epitympanic
sinuses.

Phalangeridae Thomas, 1888

The Phalangeridae is the stem family of the
Diprotodonta with the Wynyardiidae (as de-
fined here) at its base; although the Wynyardia
itself is probably too specialized to be directly

116

< a) Centra view of sku11 * <b) ° tic region. (O Otic region of Didelphis virginiana
otiC f egi ° n oi Tnchosurus vulpecula for comparison. In Tricliosurus the tubular tympanic
meatus has been broken away from the post-glenoid process and from the squamosal. The floor of the alisphenoid
bulla has also been removed to reveal the epitympanic sinuses for comparison.

117

ancestral. Offshoots which are probably pre-
Miocene in time of separation are the Vomba-
tidae, the Diprotodontidae. and the Macro-
podidae.

The main stock is the Phalangerinae, a sub-
family which contains all of those forms which
Bensley regards as basal in dentition to the
phalangerid radiation (including Acrobates, Dis -
toechurus, Eudromicia , Cercartetus) , as well as
certain larger forms which have progressed be-
yond the insectivorous habit but which, never-
theless, have remained rather generalized in
molar form <Le. Petaurus, Trichosurus , Wynlda
and Phalanger).

The Tarsipedinae, which is probably a rela-
tively modern offshoot, contains only the highly
specialized nectar-feeding Tarsives which thus
occupies a position analogous with that of Myr-
mecobius in the Dasyuridae: and, as is the case
there, fossils are unknown.

The Phascolarctinae, as it is at present re-
cognized, includes the Koala Phascolarctos, the
ringtails Pseudocheirus and the Greater Glider
Schoinobates. These are selenodont in denti-
tion but there is very great suspicion that too
much phylogenetic emphasis has been placed
on this character ( see below and Troughton
1959, p. 84).

The Thylacoleoninae comprises the marsupial
lions. Obviously derived from large phalangers,
they represent the supreme diprotodont adapta-
tion to the rapacious habit, a habit which is
not strange to some cf their smaller relatives.
Two species are known. T. carnifex Owen is the
typical Pleistocene form and T. crassidentata
Bartholomai (1962) is a slightly less highly
evolved form from the Pliocene Chinchilla Sand.
Woods (1956) has revised Thylacoleo. Gill ( 1954 >
discusses habits and distribution and provides a
full bibliography.

Most of these forms of Phalangeridae are fair-
ly small animals and accordingly it is not sur-
prising that there are only a few fossils beyond
the owl pellet deposits of the Upper Pleistocene
and Recent. Of the phalangerid fossils the earli-
est and most important is Perikoata palanka-
rinnica Stirton, 1957a from the Miocene of the
Etadunna Formation; a form which Stirton
places in the Phascolarctinae.

Of the Pleistocene species of Phalangeridae
only one has been sufficiently spectacular to have
played an important part in phylogenetic specu-
lation. This is Burramys parvus from the Wom-
beyan Caves of New South Wales. A brief dis-
cussion of Perikoala and Burraviys is included
here.

Perikoala palankarinnica Stirton, 1957

Perikoala was originally described as part of
the Palankarinna ?Pliocene fauna but Stirton
et al (1961, p. 36) now refer it to the underlying
Miocene Etadunna Formation.

Unfortunately, this remarkable fossil is only
known from very fragmentary material: an

edentulous maxillary fragment and a broken
mandibular ramus with a broken last lower pre-
molar and the first two molars which are almost
complete. I have not seen the material but
Stirton (1957a) says that it resembles both

Trichosurus vulpecula and Phascolarctos and he
concluded that Perikoala was clearly phascolarc-
tine, presumably because of the crenulated
enamel in the basins of the teeth and because of
the wide separation of hypoconid and entoconid
in the posterior premolar which (although only
a fragment of it remains) was obviously un-
usually molarized for a phalangerid. In large
Phalangerinae (e.g. Trichosurus , Phalanger, and
Wyulda ) this tooth is specialized as a sectorial
and not broadened as in Phascolarctos . As com-
pared with Pseudocheirus , Schoinobates and
Phascolarctos, Perikoala is not selenodont (or
subselenodont) but quadriscuspid or even bilo-
phodont yet it has the large number of small
folds in the crenulated enamel so typical of the
Koala.

The modern practice of separating the Koala,
the Ringtails, and the Greater Glider from the
Phalangerinae because of their selenodonty
(subselenodonty) follows such authorities on
dental morphology as Bensley (1903, p. 135);
he and various earlier workers were so impressed
by the appearance in some marsupials of seleno-
dont dentition, that they could not avoid draw-
ing parallels between them and the selenodont
and bunodont sections of the eutherian ungu-
lates and hence placed great phylogenetic
weight on this character. Accordingly, Phasco-
larctos was placed with Pseudocheirus and
Schoinobates in a separate subfamily. The
molars of Perikoala and Phascolarctos are un-
doubtedly similar in their crenulations but it
is timely to remember Thomas’ statement (1888.
p. 167) in relation to Pseudocheirus that
“this genus, with its close ally Petauroides
I Schoinobates I , by the complicated subseleno-
dont character of its molars, stands somewhat
apart from most of the other Phalangers, and
approaches Phascolarctos, in which a rather
simpler form of the same modification is ob-
servable. In young Cuscuses, however, a tend-
ency towards the same structure is also visible,
but the crests on the molars soon wear off, and
then there appears but little difference between
their molars and the simple smoothly quadri-
cuspid ones of Petaurus , Dactylopsila, and their
allies/'

In fact, one cannot help wondering whether
the relationships of Phascolarctos and Perikoala
might not lie with Phalanger , rather than with
Pseudocheirus with which Phascolarctos bears
scarcely any resemblance except in the ques-
tionable selenodonty of its molars. When more
is known of Perikoala, it may provide a useful
indication here.

Burramys parvus Broom, 1896

Burramys parvus Broom is a very small
phalangerine with very large sectorial premolars
(fig. 11) which closely resemble those of Hypsi -
prymnodon, (which is in some respects the most
primitive of the Macropodidae, and of Propleo-
pus the giant Pleistocene rat-kangaroo. Until
recently, Burramys was only known from a single
deposit, the ? Upper Pleistocene Broom fauna of
the Wombeyan Caves, New South Wales (Ride
I960). Wakefield (I960) has since obtained it
from Buchan in Victoria.

118

Because of its remarkable premolar. Broom
described Burramys as a link between the
phalangers and kangaroos, and later (Broom
1898) even concluded that it was “probably very
closely allied to the small Phalanger from which
Thylacoleo was descended”. Ultimately Tate
(1948), without further examination, removed
Burramys into the Macropodidae from the
Phalangeridae.

In 1956 I revised the status of Burramys fol-
lowing re-examination of some of Broom’s ori-
ginal material and some additional specimens
which were prepared from breccia collected by
Broom at the type locality. Burramys was
shown to be a phalangerid whose supposed
macropod affinities rested solely upon the struc-
ture of ihe premolar, while other characters
all argued against such relationships. In parti-
cular, these are the structure of the third pre-
molars, the nature of replacement of the milk

Fig. 11 . — Burramys parvus. Occlusal views of restored
cranium and mandible (from Ride 1956). x 2.5.

premolar (the peculiar “double” replacement of
Macropodidae is absent and specialization of
P-i has proceeded in the opposite direction) and
the structure of the masseteric fossa together
with the absence of a masseteric canal. The
case for special thylacoleonine affinities is
equally slender and rests upon no more than
the enlarged sectorial, a character which is
likely to be convergent since the teeth of the
“plagiaulacoid” type have been developed inde-
pendently by at least phalangers, macropods,
caenolestoids, multituberculates, and tarsioids
'Simpson 1933).

It is very probable that Burramys is a mem-
ber of an aberrant line of Phalangerinae pos-
sessing some relatively unspecialized charac-
ters, e.g., a Distoechurus- like palate and quadri-
cuspid bunodont molars, but also some very
specialized ones such as large grooved sectorials,
reduced fourth molars, and elongate incisors,
characters which probably point to some pecu-
liar insectivorous (or microcarnivorous) adapta-
tion.

Macropodidae Gray, 1821

As many fossil species of the family of kan-
garoos and wallabies have been described as all
other Australian fossil marsupials put together.
To anyone who knows our present marsupial
fauna this is scarcely surprising because, quite
apart from their commonness, macropods are
generally fairly large animals, are often gregari-
ous and are the dominant herbivores (as they
probably were during the Pleistocene as well);
it is thus likely that they provided the main
source of prey for marsupial lions, thylacines,
and other large carnivores (or scavengers) of
their time — all are features which are likely to
lead to good representation in the fossil record.

In spite of their commonness, or perhaps
partly because of it, our knowledge of their
taxonomy is more confused than that of other
groups and it is clear that Simpson’s < 1930, p.
69) hope that this state of affairs would soon
be corrected has not yet been fulfilled. The
confusion is the result of factors, usual in palae-
ontology, which are

(a) the uncertainty of generic limits

(b) poorly studied subjective synonymy be-
tween named species and poorly stated
species limits, and

(c) generally unknown temporal relations
between the named species.

Generic classification

Since almost all fossil Macropodidae are Pleis-
tocene, it is not unreasonable to suppose that
the generic limits of the modern Macropodidae
will be applicable except in a few cases where
whole genera have become extinct as has hap-
pened in the Sthenurinae where both Sthenurus
and Procovtodon are no longer extant. However,
not only has the generic classification of fossil
kangaroos always lagged behind that of modern
kangaroos and wallabies, but generic limits
among modern Macropididae are still undecided
< see Ride 1962a for a statement of the positions
taken by various authorities in modern times).
Fortunately, taxonomic activity is such that in-
vestigations which are currently in progress of

119

behaviour, chromosomes, blood sera, soft part
anatomy, bone, and dentition, will greatly ad-
vance our interpretation during the next decade,
and then the taxonomist of fossil kangaroos will
be in a much better position to produce a com-
parable arrangement. At present the taxonomist
of fossil kangaroos can only place his generic
limits by interpreting a limited and imperfect
number of characters and without much help
from neontological studies which can often lead
to much more meaningful interpretation.

Today, the area of most confusion among
modern animals is that surrounding the limits of
Macropus, Megaleia , Osphranter, Wallabia and
the other middle-sized wallabies. This confu-
sion is made worse for the taxonomist of modern
Macropodidae by the uncertain relationship be-
tween the type species of Wallabia ( W . bicolor)
and the group of large fossil wallabies called
Protemnodon ; Stirton (1963) has shown that
these large fossil wallabies are distinctive, but
it is still not clear whether the characters in
which they differ from some of the modern wal-
labies are merely expressions of size increase
which alone is scarcely a character upon which
to separate genera. The position is made even
more unstable by the knowledge that Wallabia
bicolor differs very greatly in chromosome mor-
phology from other modern middle-sized walla-
bies and is scarcely likely to be congeneric with
them; these may require yet another generic
name. At present we can only await the result
of further studies to clarify the position; in the
meantime it would seem very reasonable for
the neontologist to do as most authors do and
deliberately ignore the fossils, and also the dif-
ferences between Wallabia bicolor and other
wallabies, and use the name Wallabia for this
group.

It is probable that this deliberate cleavage
between neontological classification and palae-
ontology is desirable for stability in the names
of fossils as well. At present the incomplete
nature of most material of fossil species pre-
cludes its identification with genera which are
established by the neontologist upon the char-
acters provided by more complete material.
Thus, while taxonomists of modern animals find
little difficulty in deciding whether or not a
species, or population, belongs with one or other
of the hypsodont kangaroos Macropus or Me-
galeia , some fossil species can scarcely be so as-
signed; therefore the taxonomist of fossil kan-
garoos is forced to adopt wider limits and in-
clude both genera together. Since these limits
imply equivalence with modern genera to zoo-
geographers and others, much might be said for
the abandonment by palaeontologists of generic
names of Macropodidae which possess living ani-
mals for type species and fall back on the use
of purely fossil species concepts which will be
dictated by the nature of the material. Thus,
kangaroos with high crowned molars w f ith rapid
dental progression and premolars which are
early shed (see below), would be distinguished
from forms with relatively low crowned molars
and with tenacious premolars; these latter forms
could be further subdivided into those very large
forms with very great premolars which are never
shed, and those smaller forms which shed even
their premolars in extreme old age.

The most recent classification of the species
of fossil kangaroos (Simpson 1930 » was regarded
by its author as unsatisfactory. In it some
species are obviously greatly lumped while others
are in monotypic genera which are clearly not
comparable. This is a legacy which is the re-
sult of the alternating actions of describers (who
split) and synthesizers <w T ho lumped). We are
now in a describing phase with descriptions
being added to a lumped classification, a classi-
fication which is due in greatest part to Owen.
De Vis, and Lydekker.

Tire history of this sequence commenced in
1838. Prior to 1863, the date of completion of
Gculd’s Mammals of Australia, the taxonomy of
modern Macropodidae as generally accepted in
Britain ( see Owen 1840. and Waterhouse 1846)
was uncomplicated containing only three genera
Macropus, Hypsiprymnus ( Potorous ) and
Dendrolagus. The five species described by Owen
prior to 1873 (i.e. 1838-1859) were all included
in the first two of these. Of the remaining six-
teen species which he described (i.e. 1873-1877),
cne was a potoroine which he included in Bet-
tongia Gray 1837, two were placed in Osphranter
which had been introduced by Gould, one (M.
ferragus ) was placed in Macropus, and the re-
mainder were all placed in six new genera (as
well as these, Owen even placed in these new
genera two of the five species of his early period
and the solitary Macropus , i.e. M. ferragus, of his
later period ) . Owen thus left twenty-one species
distributed between ten genera.

Owen was followed by Lydekker who, in his
standard work, the Catalogue of the Fossil Mam-
malia in the British Museum (Natural History)

( 1887 ) , placed all of these in the four genera
Macropus , Stlienurus , Procoptodon, and Aepy-
prymnus (ignoring the status of Hypsiprymnus

Potorcusi . Since we would remove Aepyprym-
nus today to the Potoroinae and Stlienurus and
Procoptodon to the Sthenurinae, Lydekker actu-
ally placed all Macropcdinae in the single genus
Macropus.

De Vis followed him, describing fifteen species
of Macropodidae and three new genera between
1883 and 1895. These three new genera had one
species each and of them one, Triclis < Prov-
leopus) . is a potoroine while the other two are
fragments generically not determinable. The
remainder of his species were described as be-
longing to Stlienurus (in which he included
Procoptodon), Macropus which he confined to
the h 5 r psodont kangaroos, and Halmaturus a
name which he applied (actually, misapplied,
following others) to all the remainder.

The most recent complete revision is Simp-
son’s (1930) Post-Mesozoic Marsupialia, and
again the synthesis produced a lumped tax-
onomy. In effect Simpson returned to the con-
cepts of Lydekker. except that he recognized
Protemnodon Owen, and placed all the Halma-
turus fossils of De Vis into Macropus.

Today we have thus a classification of fossil
kangaroos which is excessively lumped when
compared with the generic classification of Re-
cent forms. The only Macropodinae outside
Macropus are Brachalletes De Vis and Synap-
todon De Vis which are probably not determin-
able, and Protemnodon which is confined to an

120

uncertain number of very large, large-premo-
lared fossil wallabies; and the newly described
Prionotevinus Stirton (1955) which has not yet
been satisfactorily compared with any other
genera.

Structure and Terminology in Macropodidae

Descriptions of fossil macropod species are par-
ticularly difficult to evaluate and this leads to
uncertainty in subjective synonymy. In the
main these descriptive complications are due to
nature of the macropod skull and dentition and
in particular to its ontogenetic processes, but
they are also unduly confused at present by the
existence of more than one set of dental
terminology.

Like the Phalangeridae from which they were
clearly derived, the Macropodidae possess quad-
ricuspid molars but, unlike most of them, these
are often further complicated by the develop-
ment of lophs, links, and folds of enamel which
increase their efficiency as grinders. Canines
are generally absent and, as in Phalangeridae,
incisors are reduced to three above and a single
pair below. However, the acquisition of a graz-
ing habit and the development of elongate sec-
torials has resulted in the elaboration of the
masseteric musculature and in particular a mas-
seteric foramen is developed in the mandible
(see Ride 1959). Premolars have been reduced
to two only in the permanent series and with
this has come a unique sequence of replacement
in which the most posterior permanent premolar
supplants not only its milk predecessor but also
the preceding permanent tooth. Since both per-
manent teeth are elaborate sectorials (or are
derived from them) while the milk tooth is a
molariform tooth, the macropod toothrow al-
ways comprises a sectorial followed by a row of
molariform teeth. In Hypsiprymnodon alone of
the Macropodidae both permanent premolars
have a brief coexistence in fully erupted state
(Woods 1960, Ride 1961). Complexity does not
end here, however, because in many Macropodi-
dae the molar tocth-row moves bodily forward
during the life of the animal (fig. 12) with the
result that the position of various teeth in rela-
tion to various cranial features are progressively
altered, and even the angles of the various teeth
and their longitudinal dimensions change. Fur-
ther, the permanent premolar (and even suc-
ceeding molars) may be shed from the front
of the jaws as in the great kangaroos, or molars

may be impacted and shed laterally from be-
hind the permanent premolar as in some wal-
labies. Since characters such as the relative
positions of teeth to bony processes, and even
the degrees of development of these in Macro-
podidae, are so very dependent upon onto-
genetic stages, they are very difficult to interpret
when they are used in descriptions which in-
volve insufficient material to establish the vari-
ous developmental sequences.

Differences in the terminology of premolars
applied by two schools of dental nomenclature
are nowhere more obvious than in the taxonomy
of Macropodidae since the characters of the tw r o
permanent premolars are often diagnostic. One
school holds that since more than three perma-
nent premolars have not been discovered in
known Marsupialia these should be numbered
1, 2, 3. The other school (of which I am one,
and I have consistently applied the 1, 3, 4
nomenclature in this review) notes that since
Marsupialia and Eutheria are probably mono-
phyletic and that Eutheria possess up to four
premolars, known Marsupialia have probably
lost one. The last tooth with a predecessor
(i.e., the most posterior premolar) is always in
contact with the first molar (unless, as in some
Dasyuridae, it is absent), accordingly this is
called 4 and is regarded as homologous with the
fourth premolar of Eutheria and Triconodonta.
This conclusion was adopted by Thomas whose
1888 Catalogue forms the foundation of all
modern taxonomic work on marsupials; by
Lydekker whose Catalogue (1887) summarizes
and synthesizes Richard Owen’s work; by H.
H. Finlayson, E. Le G. Troughton and G. H. H.
Tate whose monographs on Australian Marsu-
pialia together make up the greater part of all
modern taxonomic work on this group; and by
C. W. De Vis who, second only to Richard Owen,
was responsible for the greater part of the
descriptive literature on fossil Macropodidae.
The other two premolars are called by this
school P3 and PI in accordance with the observa-
tion that there is most usually a gap between
the most anterior and the intermediate pre-
molars.

While two schools of premolar terminology
have been in existence for some time (the
literature of American marsupials generally ap-
plies the PI. 2, 3 terminology) a second school
of incisor terminology has only lately arisen.
It was introduced by Stirton (1955) who used

n-p 3

Ly (sectorial)

A— p 3

/ -4- dP 4 (molariform)

A p 4 (sectorial) j'*

/4— m 1

1 — M'

M 4

A— M* ^

|— M 4

A.M.N.H 66164 O.U.M. 8027 U.S.N.M. 38551 AMNH 882 Leiden 1888b Holotype.

12, — The succession and forward movement of cheek-teeth of Macropodidae illustrated by the modern wallaby
Protemnodon parma. The transverse line indicates the position of the tips of the descending zygoma tic processes
of the maxillae. The youngest individual is on the left, and the oldest is on the right (after Ride 1957).

121

122

the names I 2 ^ ‘ for the upper incisors and I*
for the lower, in place of the usual I 1 2 s and I,.
stating at the time that this was based upon
the assumption tnat the primitive incisor

formula in marsupials is | ^ - — f- and

1 • 2 ■ 3 • 4 5

that the remaining incisors of the Macro-

^ 0 • 2 • 3 • 4 0

podidae are - — - — - — - — — . Stirton has since

(1963) abandoned this terminology, but Marcus
(1962) has adopted it. I have reviewed (1962,
pp. 297. 8) the embryological evidence for the
primitive incisor formula in marsupials and
there is little doubt that evidence exists which
suggests that the formula is at least six in the
upper jaw. It is also likely that the diprotodont
lower incisors are the third or fourth teeth of
the series. In spite of this, I would not at pres-
ent advocate the introduction of a further in-
cisor terminology to upset that which is in al-
most universal use until the fossils are found
which demand it.

Phyletic lines in Macropodidae

Today, there appear to be three phyletic
lines of Macropodidae. One of these, the Poto-
roinae (Hypsiprymnodontinae and Potoroinae
of earlier authors — see Woods 1960) appears to
be an early offshoot combining a number of
primitive characters (such as a hallux and
simple alimentary canal in Hyprispymnodon,
and relatively simple molars in most of the
group), with a number of extreme specializa-
tions (like anterior vaginal expansions, plagi-
aulacoid premolars, and enormous masseteric
canals. See Pearson 1949, Woods 1960, Ride
1961).

The most modern of the offshoots from the
main stem is the Sthenurinae. These are sec-
ondarily adapted for browsing from the main
grazing line < see Ride 1959) and culminate in
Procoptodon with deep, heavy mandibular rami,
broad crushing premolars, wide highly “orna-
mented” molars, erect lower incisors, reduced
upper incisors, elongate descending zygomatic
processes of the premaxillae, and a coossified
mandibular symphysis.

The Macropodinae are the stem forms of the
radiation. At their most highly evolved they
are supremely modified for grazing and swift
cursorial (saltatory) existence. However, with-
in the Macropodinae numbers of different kinds
of modification are visible; some of them are
small animals with scarcely any of the grazing
modifications we associate with the group (e.g.,
Lagostrophtis, see Thomas 1887) while others
are large grazers which place almost no depend-
ence upon their sectorial premolars and shed
them early, meeting all their masticatory re-
quirements with a rapidly progressing molar row.
In these “end forms” molars are hypsodont to
meet the abrasive effects of silica -containing
grasses and, like those of their eutherian
counterparts, lophodont molars have links which
increase the efficiency of their grinding surfaces.
Some Macropodinae have enlarged sectorials al-
most to the extent of the Potoroinae. In the
past some of these forms with large sectorials
were very large (e.g., Protemnodon) but today

they are generally of moderate size and include
the forest and thicket dwelling wallabies and
tree-kangaroos ( Wallabia bicolor, Dorcopsis .
Dorcopsulus and Dendrolagus ) ; these Macropo-
dinae possess simpler molars than their grazing
adapted relatives.

Until recently, Palorchestes was believed to be
a giant kangaroo, however revision by Woods
(1958) has shown unequivocally that this genus
represents a slenderly built line of Diproto-
dontidae which possibly bears to the heavily
built Diprotodontinae the same relationship as
the eutherian tapirs do to the rhinoceroses.

At present it is not possible to provide a de-
tailed arrangement of fossil kangaroos because
proper comparison has yet to be made between
the confused samples described from the various
strata of the Darling Downs and the various
species described by Owen. Revisions of fossil
taxa of Macropodidae, or parts of them which
have been completed in recent years are of
Sthenurus (Bartholomai 1963, Marcus 1962),
Protemnodon (Stirton 1963 >, Propleopus (Woods
1960). However, one form does require special
comment because it has to date been referred
to the Phalangeridae and (as a cuscus) has even
taken its place in palaeoclimatic speculation
(Gill 1957, p. 159: 1961, p. 336). This is the fos-
sil tooth from Grange Burn.

The Grange Burn Potoroo

The Grange Burn Potoroo is one of the most
tantalizing cf fossils. It is possibly Pliocene and
comprises a solitary enamel cap of a macropod
second or third right lower molar (Nat. Mus.
Viet. No. P 15777). It was collected by Edmund
D. Gill, of the National Museum, in the bank of
Grange Burn, one mile upstream from Forsyth’s
Bank, near Hamilton, Victoria. It came from
a fossil podsol 6 ins. below T basalt, where the
Burn flow T s off the basalt on to Tertiary rocks.
Gill (1953a, 1957) regarded the tooth as being
cuscus-like and Stirton (1957) agreed, compar-
ing it with the second and third right upper
molars of Phalanger maculatus and Ph. orien-
talis . He noted that the main difference between
the fossil and the cuscuses was that the fossil
was larger, more elongate, and possessed a
“prominent median lingual inflection”; he re-
garded the tooth as being much too worn to
show any indication of the crenulated surface
typical of the Phalanger tooth.

Comparison with a wider range of diproto-
donts reveals that the tooth is undoubtedly that
of a potoroine. It is inseparable in general
morphology, positions of wear facets, and pat-
terns of exposed dentine from modern Potorous
but it is very much larger having about twice
the linear dimensions of equivalent teeth of
P. gilberti. It is as distinct from the other
potoroines Bettongia, Caloprymnus , and Aepy-
prymnus in size as it is from Potorous, but de-
tailed comparison of cusp patterns with these
has not been made. It is distinguishable from
Propleopus in its smaller size, more acute lingual
cusps, complete absence of a median longitudinal
valley, and in its inequality between anterior
and posterior transverse dimensions.

123

The fossil tooth (fig. 13) has four cusps; two
of these, the protoconid and hypoconid, are low
in relief and rounded (protocone and hypocone
of Stirton 1957, p. 122) while the other two. the
metaconid and entoconid, are high and angular
(paracone and metacone of Stirton). As meas-
ured from the base of the enamel to the tip of
the cusp, only the protoconid is really lower
than the others (prd 3.3 mm. med 3.9 mm. hyd
3.8 mm. end 3.9 mm), but the relief of the
occlusal surface is such that the two sharply
angled lingual cusps appear to stand above their
rounded buccal counterparts. There are two
clearly-defined flat areas at opposite ends of the
periphery of the tooth which are, without doubt,
areas of contact with adjacent teeth in the
tooth -row. Of these, the anterior is broadest
and from its buccal edge a broad, shallow
channel rises onto the occlusal surface of the
tooth to separate off a small '‘antero basal
cingulum” (more properly this channel is the
remnant of the entrance to the trigonid basin
between protoconid and the almost completely
reduced paraconid, see Ride 1961). Dentine is
exposed in small sub-circular areas at the tips
of the buccal cusps and there are also three
smaller dark patches in other places on the
occlusal surface which almost certainly mark
the presence of fissures or pits in the enamel;
the most prominent of these lies postero-buccal
to the entoconid and represents the posterior
part of the talonid basin ( see Ride 1961 no
56. 7).

The occlusal surface of the tooth shows only
moderate wear; the posterior surfaces of the
semi-lophids (i.e. the areas which lie posterior
to the transverse ridges which run from the
acute metaconid to the rounded protoconid, and
from the acute entoconid to the hypoconid) are
covered with enamel which has a faintly pitted
surface which is obviously quite unworn since
it lacks facets, or the microscopic scratches
which are typical of dentally abraded surfaces.
However, in contrast, the anterobuccal faces of
the metaconid and entoconid have well marked
wear facets with pronounced microgrooves; these
wear facets have not passed through the enamel
into the dentine beneath. Wear through to the
dentine is only present on the tips of the two
buccal cusps; here, more or less circular areas
of dentine are exposed. The rounded buccal
shoulders of these cusps are also well-marked
with microgrooves and in particular, these are
concentrated in slight facets anterobuccal and
posterobuccal to each cusp.

The length of the tooth from anterior to
posterior contact areas is 8.3 mm, breadth at
level of anterior cusps 7.8 mm, and breadth at
level of posterior cusps 6.4 mm.

Since other authors have regarded this tooth
as being most like a right upper molar of
Phalanger while I regard it as being a right
low T er molar of Potorous , differences between
these teeth are rather important (fig. 13a, b)
and are tabulated here.

Kt M 2 Phalanger ( nudicaudatus and maculatus)

1. All cusps sharply angular.

2. Wear facets on posterior faces of semi-lophids.

3. When wear is at the stage where dentine shows
through the tips of lingual cusps only and not
through the buccal cusps, the anterior faces of the
semi-lophids are deeply ridged with crenulated
enamel.

4. As the tooth wears, the area of exposed dentine
spreads anteriorly and posteriorly from the tips of
the lingual cusps to form two Vs along the antero-
postero ridges. After this stage it begins to spread
buccalwards along the transverse ridges of the semi-
lophids.

5. The antero lingual face of the protocone is flattened
and facet-like. There is no antero basal cingulum.

6. There is no fissure posterior to the transverse ridge
between hypocone and metacone.

Rt Mo Potorous (gilberti)

1. Lingual cusps angular, buccal cusps rounded.

2. Wear facets on anterior faces of semi-lophids.

3. When only two of the cusps l i.e. the low rounded
buccal cusps) show exposed dentine there is no
crenulated enamel on any faces of the tooth.

4 As the tooth wears the area of exposed dentine at
the tips of the two rounded buccal cusps spreads
transversely across the tooth, not anteroposteriorly
until after the pairs of transversely placed cusps
become joined.

5. The antero buccal face of the protoconid carries a
broad shallow channel which opens out onto the
occlusal face of the tooth, separating a small
"antero basal cingulum".

6. The posterior part of the basin of the talonid is
represented by a fissure in the enamel, posterior to
the transverse ridge between the posterior cusps
(end & hyd).

The Grange Burn fossil possesses all the char-
acters of the Potorous molar and none of those
of Phalanger .

The only fossil potoroine which approaches the
Grange Burn fossil in size is Propleopus oscillans.
Woods (1960, p. 201) gives dimensions of the
lower molars of two specimens of P. oscillaris as
being M, 9.5 x 8.7 and 9.3 x 9.2, M 2 10.8 x 9.8 and
10.1 x 10.0, M 3 11.2 x 10.3 and 10.7 x 10.2 mm;
the Grange Burn fossil measures 8.3 x 7.8 mm.
There are also considerable resemblances in
general morphology. Comparison with stereo-
scopic pairs provided through the kindness of
Mr, J, T. Woods (fig. 13c) reveals that the Pro-
pleopus lower molar differs in having more
rounded lingual cusps, a more pronounced
median longitudinal valley (defined Ride 1961,
p. 56, under M>) which is completely lacking
in the Grange Burn fossil but visible in all
molars of the holotype of P. oscillans. The
molars of P. oscillans also seem much more
square (except Mi and M 4 which the Grange
Burn fcssil cannot be; it is not Mi because the
anterior moiety is broader than the posterior
moiety and is not M 4 because there are contact
surfaces at both ends of the tooth).

Since the differences between the Grange
Burn fossil and P. oscillans are slight and are
certainly no greater than those between related
species in other genera of Macropodidae it is
not possible to decide with any degree of cer-
tainty whether this single tooth should be placed
in Propleopus on grounds of size, or whether it
should be regarded as a gigantic Potorous on
grounds of morphological similarity. It seems
safest to conclude, for the present, that it is a
potoroine and leave naming it until further ma-
terial is discovered. The present tooth contains
sufficient features to be certain that the species
is new. but it will require the discovery of a
fourth molar or a premolar to provide sufficient
diagnostic characters to name the genus.

124

Diprotodontidae Gill, 1872

The Diprotodontidae are the giants of the
Australian fossil record. These great quadrupe-
dal herbivores are well represented in the fossil
record as far back as the Miocene. Unfortun-
ately, there has been no modern synthesis of
the group (which is now known to contain two
divergent phyla, the lightly built Palorchestes -
like forms and the more typical diprotodontids)
and the status of a number of the species, e.g.
D. minor, D. bennetti, D. longiceps, is not clear
from the literature. De Vis '1891b) provides a
key to the genera; he also considers that Seep-
arnodon probably belongs here (see under Vom-
batidae) .*

The most valuable statements cf the temporal
distribution of diprotodontids are in surveys by
Woods (1960a, 1962) and data selected from
these are tabulated below. In this table many
formations and localities are omitted to avoid
duplication but sufficient are given to indicate
the dipretodontid faunal associations, and the
distributions of various species. As Woods sug-
gests, it seems likely that Diprotodontidae will
prove to be stratigraphically important.

Age

Formation, etc.

Species

Recent. ...

Tartangan culture

BP 0.570

/ )i prof od on opt at a m

Pleistocene

Mowbray Swamp, Xas.

Sotothcrium tanman train
-Y. mitchelli

lHproUxhm Beds, Darl-
ing Downs, Qd.

Pi protoilon opt alma
Sot other i urn mitchelli
Pa/orchcstcs azael
E uowen ia robusla

3*13 or die

Otibanda Lake Beds,
X.G.

y ototfi er i u m tva Intense.
(7 M finixfiolophus)

Chinchilla Sand. Darl-
ing Downs. Qd.

Eu omnia, i /rata,

Euryztp/omo dunense
Palorchestes parvus
Pi protoilon sjj. (Woods 1902.
p. 40)

M am pu w< ud u Sands.

S,A.

M etiiscoloplms mawsoui
E uowen ia-Iike form

Miocene

< 'heltenhammii Stage,
Vic.

Xototheri.um (premolar figd
Stirton 1957 see Woods
1962, p. 45)

Diprotodontidae. Maxillary

and. mandibular fragments

Etadunna Km.. S.A.

Small diprotodontid with
7 1’alorchestine affinities

(Stirton fit al 1901, pp. 30.
7)

Eote : The sequence of the formations within each ago does not
necessarily indicate temporal succession.

The Otibamla Lake Beds were formerly thought to be
Pleistocene bill parts of them arc probably Upper Pliocene.
Stirton (19(53, footnote to p. 1-14) gives a K, date of
4 5 million years for pyroelastios associated with them.

Vombatidae Iredale & Troughton, 1934.

The origin of wombats is unknown although
their diprotodonty and syndactyly clearly indi-
cate that they are Diprotodonta. Dental
characters are aberrant in them and give little
real clue to their relationships with other
families in the order. Various attempts have

♦Since this address was given Stephenson (1963) has
endorsed this conclusion and has. in addition,
described a new species and genus of diprotodontid.
Diarcodon parvus.

been made to elicit affinities from these and
other anatomical features, and Tate (1951, p. 3)
has reviewed this work.

No Tertiary wombats have yet been described
and no fossils are reported as having been col-
lected in the various deposits of this age investi-
gated by Stirton in Central Australia or Victoria.
Woods lists none from the Chinchilla Sand
(Woods 1960a, 1962) nor does De Vis (1891)
although the general term “Darling Downs”
which he uses for Queensland wombats may
include this Pliocene formation.

The most recent revision of the wombats is
by Tate (1951) who dealt with both fossil and
recent forms. This work, when taken together
with Lydekker’s (1887) catalogue (to which
Tate does not refer), provides a very adequate
summary of present knowledge of the fossils.
However these two works differ slightly in
arrangement. Lydekker’s list is most useful in
providing concise descriptions and a synoptic
view of Owen’s fossil species, but it should be
read in the light of De Vis* further revision of
the status of the species P. mitchelli and
P. thomsoni (De Vis 1891). In his arrangement,
Lydekker adopts a classification which is that
used for modern wombats today, i.e. he recognizes
two main kinds of normal-sized wombat (the
plains-dwelling Lasiorhinus wombats and the
Vombatus species of the more densely timbered
country) although of course he uses only the
name Phascolomys which was then in general
use for them all. He employs Phciscolonus for
the Pleistocene Giant Wombat.

Of the fossil wombats known today, Lydekker’s
levision omits only P. hacketti Glauert, 1910,
P. angustidens De Vis, 1891 (which were
described since) and P. pliocenus McCoy, 1874
(which was not noticed by him). Tate places
P. hacketti and P. pliocenus with Vombatus, and
P. angustidens with Lasiorhinus. Both Tate and
Lydekker exclude P. curvirostris from these two
groups; Lydekker regarding it as incertae s edis.
while Tate creates a new genus, Ramsay ia, for
it. Tate regards the position of P. medius Owen
and P. magnus Owen as uncertain, Lydekker
includes them in the Vombatus group.

Phascolonus , the genus of Giant Wombats, is
universally accepted and is now known from
very much better material than Lydekker or
Owen possessed; this is from Lake Calabonna
(Stirling 1913). The genus Sceparnodon is con-
ventionally placed in synonymy with Phasco-
lonus , but with some reserve (Tate 1951, p. 12)
because of De Vis’ (1891a) strong arguments
that these were not synonymous and moreover
were not even closely related. He regarded
Sceparnodon as being diprotodontid (see above
above under Diprotodontidae).

MARSUPIALIA incertae sedis
Notoryctidae Ogilby, 1892.

Notoryctes, the marsupial mole, is unknown
as a fossil and it is so highly specialized in
dentition, skull structure, and limb structure
that arguments as to its affinities which are
based upon these features can only produce
tentative results. It is usual to follow Stirling's
(1891) original conclusions and relate it to the

125

Dasyuridae ( see Wood Jones 1923, and Simpson
1945). However. Bensley (1903, p. 173) regarded
the foot structure of Notoryct.es as being derived
from “one of ordinary phalangerine type” and
its dentition as having been “so modified away
from the usual dasyurid and peramelid type that
it does not present any of the special characters
which distinguish the dentition of these two
families” (Bensley 1903, p. 117). Accordingly he
placed Notoryctes (p. 210) in the “polyprotodont
syndactyla” with the Peramelidae. Cabrera
il919, p. 90) followed Bensley. Van der Klaauw
(1931, p. 254) has pointed out that the structure
of the tympanic in Notoi-yctes agrees more with
Diprotodonta than with other marsupials. Some
authors (e.g. Trough ton 1959, p. 75) have pre-
ferred to remain uncommitted and, in view of
the introduction of ordinal classification here,
it seems wisest to follow them.

The Australian Radiation — Origin, Isolation
and Drift.

The Australian radiation, by comparison with
mammal radiations of other inhabitable land
masses, is unbalanced in the degree in which
various major taxa of terrestrial mammals (and
in particular Cohorts) are represented in it.
Thus, with the exception of the dingo (an
obviously Pleistocene, or Recent, introduction).
Ferungulata are absent but represented in all
others: Unguiculata (except for flying forms
such as bats) are absent but present in all others;
and Glires, while present, are confined to repre-
sentation from a single suborder of Rodentia
(and a single family of these) when other con-
tinents have representatives of both orders and
all suborders.

The explanation of this can only be that the
Australian Continent has been through a very
long period of isolation which has lasted through
much of the Age of Mammals. However, in
this isolation, Australia has developed a wide
and comprehensively diversified mammalian
fauna, almost entirely marsupial, which has
paralleled the mixed marsupial and eutherian
radiation of South America, and pure eutherian
radiations of other continents, in almost all of
the kinds of niche available to mammals. Thus,
while in South America the marsupials provided
the carnivorous predators and smaller insec-
tivores and carnivores, the archaic ungulates of
that continent provided the herbivores ( see
Simpson 1950), in Australia all three main
adaptive avenues were filled with marsupials.
Moreover, they were not interrupted by invasions
of similarly adapted mammals from outside
during this period (or if they were the invaders
were repelled leaving no trace).

During the early part of this century many
responsible zoologists believed that the long
isolation of Australia, taken together with the
fact that the greater component of the Austra-
lian fauna was marsupial and similar to that
of South America, could best be explained by
the theory of Continental Drift. Further,
palaeontologists working on South American
marsupials argued for the existence of special
relationships between separate parts of the
South American Marsupial fauna and parts of

the Australian fauna; in particular between
Thylacinus and Borhyaenidae, and between
Diprotodonta and Paucituberculata. I have
reviewed this work, and that of others, and
conclude that these relationships cannot be
supported 'Ride 1962).

Most zoologists today believe that an alterna-
tive explanation, due principally to W. D.
Matthew and G. G. Simpson, is more reason-
able; (his states that Australia has been no more
isolated during the whole of the Age of Mam-
mals than it is today and during this period
entry to it has been by a combination of chance
and rafting ability across the water gaps of the
Malaysian archipelago. Thus, it is the effect
of chance which has led to the imbalance of
the Australian fauna. In contrast, many botan-
ists ( see Good 1963, but also see Burbidge 1960
p. 156 for expressions of reserve) would not
agree with this conclusion and still hold to the
theory of Continental Drift, observing that the
short distance across Torres Straits marks a
greater floral discontinuity than any similar
distance on the earth’s surface and moreover
floral continuity with the archipelago occurs
through New Britain, the New Hebrides and
New Caledonia and, allowing for differences of
latitude, even to New Zealand; they conclude
therefore that the floral discontinuity can best
be explained by the hypothesis that the gap be-
tween Australia and New Guinea has only be-
come narrowed by the northward drift of the
Australian continent.

In recent years geophysical work on palaeo-
magnetism supported by data of palaeoclima-
tology ( see Runcorn 1962, Nairn 1961) have
made it reasonably certain that Continental
Drift has occurred. The direction of the mag-
netic field of rocks of known age on a single
continent can be plotted to give a polar wander-
ing curve for that continent which probably
indicates the positions of the magnetic poles at
various ages relative to it. It is now known
that polar wandering curves of the various
continents do not agree and that the pole posi-
tions given by various continents for any parti-
cular age can only be made to agree by displac-
ing the continents relative to each other, i.e,,
by “drifting” them. Until very recently, it
was not known whether Continental Drift as
postulated from these data occurred late enough
in time for it to effect the composition of the
Australian Mammal fauna. However, the work
of Irving et al (1963) have made it abundantly
clear that Australia only achieved its present
latitude during the late Tertiary and that since
the Mesozoic it has been moving slowly north-
wards across what is now the Southern Ocean.

It seems that as biologists we are now in the
fortunate position of having a very reasonable
working hypothesis which has been erected solely
on non-biological data and which we can now
test by biological observation. Fossil marsu-
pials do not yet help us; such marsupials as
have yet been found in New Guinea are Pleisto-
cene and upper Pliocene (Stir ton 1963, Woods
1962) while the distribution of modern marsu-
pials in New Guinea and the adjacent islands
is entirely consistent with its being the pro-
duct of alterations in Pleistocene sea level, and

126

even through chance distribution across very
short water barriers over a very short space of
time.

On the other hand a very wide water gap
between Australia and Asia during much of
the Tertiary is supported by the apparent lack
of success enjoyed by non -marsupial invaders
before at least the Miocene (the Muridae first
appear in the fossil record in the Pliocene of
Europe and Asia). During the early Tertiary,
Asia certainly possessed such highly efficient
short water-barrier crossers as Primates, In-
sect] vor a. Scuriomorpha, Hystricomorpha. etc.
After all. even elephants succeeded in crossing
into Celebes and across Wallace's Line into
Flores during the Pleistocene (Hooijer 1962).

Added in Proof

Since this address was submitted for publica-
tion. a number of articles with important bear-
ing on it have come to my notice. These are by
Chow (1963). Hofer (1952), Hooijer (1964),
Kean, Marry a tt & Carroll (1964), Macintosh &
Mahoney ( 1964), Mahoney (1964), Mills (1964),
and Stephenson (1964).

The nature and character of ancestral mar-
supials (see Fig. 1 and under Phytogeny) are
clarified by J. R. E. Mills whose work (1964,
Proc. Linn. Soc. Lond.. Vol. 175, pp. 117-133)
would indicate common ancestry of Eutheria
and Marsupialia among the Amphitheriidae of
the Upper Jurassic. Formerly, inadequate ma-
terial of Pantotheria led Simpson ( 1945, pp. 169,
70 ) to state that “Despite a general resemblance,
the dental differences in detail l between panto-
theres, marsupials and eutherians I are great,
and various attempts to homologize molar cusps
or to specify closer affinities are so contradic-
tory and hypothetical as to provide no worth
while data for taxonomy.” Now, however, newly
prepared specimens of Peramus and Amphi-
theriurn from the Upper Jurassic suggest that
the relationship is closer than hitherto sus-
pected and that the molar pattern of Amphi-
theriidae (in which Mills includes Peramus ) is
well on the way to the tribosphenic condition
characteristic of both marsupial and placental
fines in the Upper Cretaceous. Mills is of the
opinion that, on purely dental evidence, the divi-
sion between marsupial and placental could date
from this stage. In connection with marsupial
dental homologies (see under Macropodidae —
Structure and Terminology) it is of interest to
note that specimens of Peramus examined by
Mills possess a canine, four premolars, and four
molars; specimens of Amphitheriwni also possess
four premolars and five molars, although Mills
notes that up to seven molars are known in
specimens not seen by him.

R. I. Kean, R. G. Marry att & A. L. K. Carroll
(1964, Aust. J. Zool., Vol. 12, pp. 18-41) have
also added data relevant to Kean’s earlier argu-
ment that the marsupial female urinogenital
system is derived from the eutherian condition.
They report the presence of epithelial cells lin-
ing the median “pseudovagina” of a large series
of Trichosurus vulpecula and conclude that this
epithelium acts as a retarding influence in the
closure of the canal after parturition; they are
of the opinion that this lining is vestigial and

not rudimentary, hence marsupials with a per-
manently lined, and open, median canal are, in
their opinion, primitive.

The taxonomic status of Dasyurus affinis (sec
under Dasyuridae) has been settled by J. A.
Mahoney (1964, Proc. Roy. Soc. Viet. Vol. 77, pp.
525-533). Both syntypes, in mensuration and
morphology, fall within the range of variation
of Dasyurus maculatus, a modern species.

N. W. G. Macintosh and J. A. Mahoney (1964,
Proc. Roy. Soc. Viet. Vol. 77 pp. 507-514) have
also added to the complexity of the Thylacinus
problem by recording a small Thylacine molar
from Fromm’s Landing in the Murray Valley,
South Australia. This tooth comes from a hori-
zon dated at between 1806 ± 85 B.C. and 1931 ±
85 B.C. It is the first small Thylacinus tooth
(the measurement of M 1 pr.-me. which these
authors give is less than any of the south-west-
ern cave Thylacines measured by me) to be re-
corded to the east of the Nullarbor in an area
which is otherwise occupied by the larger form*
The tooth is also morphologically unusual in
that the protocone is divided.

The identity of Sceparnodon (see under Vom-
batidae and Diprotodontidae) has been clarified
by N. G. Stephenson (1964, Proc*. zool. Soc. Loud.
Vol. 142, pp. 537-546). He concludes that it is
a member of the Diprotodontidae. Stephenson
also expresses the opinion < p. 540) that the only
specimen known of Phascolomys curvirosMs (see
under Vombatidae) is possibly a badly eroded
rostrum of P. magnus “or at least of a member
of the wombat series of comparable size to this
species”. Stephenson has also (1963, Palaeon-
tology Vol. 6, pp. 615-624) commented on the
distinctness of Sarcophilus laniarius and S.
harrisi (see under S. laniarius — Dasyuridae);
his measurements of two specimens of laniarius
and cf one of harrisi (p. 618) do not support
differences in size claimed for them by other
authors.

Earlier (1962), I had drawn attention to the
fact that, in the absence of Asiatic marsupials,
the presence of the notoungulate Palaeo sty lops
in the Palaeocene of Mongolia was evidence that
a migration route between South America and
Asia was open in the early Tertiary. Chow Min-
Chen (1963, Scientia Sinica. Vol. 12, pp. 1889-
1893) has now described a possible megalony-
choid xenarthran ( Chung chienia sichuanica )
from the Eocene of Honan. Megalonychoids are
unknown outside South America until the Plio-
cene (Simpson 1945, p. 69). D. A. Hooijer (1964,
Zool. Meded. Vol. 40, pp. 37-44) has published
further information on the pygmy stegodont
from Flores. He says that it is similar in size
to the species known from Celebes but regards
it as unlikely that both are conspccific. The
material from Flores comprises two milk molars.

Finally, I am indebted to Dr W. D. Turnbull of
the Chicago Natural History Museum for bring-
ing to my notice Helmut Hofer’s remarkable
“tiber das gegenwartige Bild der Evolution der
Beuteltiere” (1952, Zool. Jb . Abt. 2. Vol. 72, pp.
365-437) and to Mrs. A. Neumann for translat-
ing it. While he reviews no original work on
Australian fossil marsupials more recent than

that of Wood Jones on Wynyardia , Hofer gives
much interesting speculation on the origin of
the various marsupial phyla and t in particular,
he summarizes the work of various German
anatomists and embryologists which is not gen-
erally known to readers of the more usual litera-
ture on Australian marsupials.

Hofer bases most of his speculation and re-
view of American marsupials on the work of
Simpson, and of Australian marsupials on Bens-
ley and Tate. He concludes that, despite lack
of fossils, dasyurids are certainly derived from
didelphoids, and that borhyaenoids are distinct
from Thylacinus. He regards the perameloids
as a very early lateral branch of the south-
east Asian or Australian dasyuroid- or didel-
phoid-like marsupials, paralleling the South
American Caroloameghinia among didelphoids.
The origin of diprotodonty receives particular
attention because of Hofer’s own interest in
masticatory musculature. He regards shorten-
ing of the mandible by mutation as a necessary
step and cites modern experimental work with
dogs to show that such a step is not only pos-
sible, but can also be functional. He is of the
opinion that this mutation could have been of
frequent occurrence among didelphoids and
that this has led to the parallel development of
diprotodonty in Caenolestoidea and Phalange-
roidea. He supports this view with the observa-
tion that the sectorial tooth in these groups is
not homologous, but this has now lost much of
its force as the result of de Paula Couto’s work
of 1952 (discussed in Ride 1962, p. 295) which
has shown that this view of the sectorials is no
longer tenable.

Acknowledgments

1 am indebted to the Directors and Curators
of collections of fossil marsupials in Australia
and in particular to Mr. J. T. Woods, Director of
the Queensland Museum, for information and
photographs of fossils, for hospitality and for the
loan of specimens; to Mr. E. D. Gill of the
National Museum of Victoria for access to Glan-
codon and for the lean of the Grange Burn
tooth; to Mr. Basil Marlow of the Australian
Museum for the loan of Thylacines; to Dr. W.
Bryden, Director of the Tasmanian Museum, for
hospitality and permission to work on Wyn-
yardia ; to Mr. D. L. Cook for the loan of west-
ern cave-fossil Thylacines; to Professor L. H.
Wells, formerly of the University of Edinburgh,
for placing Robert Broom’s collection in my
hands and giving me permission to prepare it.
I am also grateful to Mr. B. K. Bowen of the
State Fisheries Department, Western Australia,
for discussions on the statistics involved, and
to Mr. Duncan Merrilees who criticized this
manuscript. Finally, the whole work owes much
to my former supervisor and colleague Profes-
sor A. J. Cain who taught me that there was
much more to taxonomy than classifying ani-
mals; and to Dr. Arthur Tindell Hopwood, form-
erly of the British Museum (Nat. Hist.), who
spent very many valuable hours giving me the
benefit of his unrivalled experience of Sir
Richard Owen’s collections of fossil mammals
from Australia.

The line drawings in the text are by Miss
Rosemary Hunt except where otherwise acknow-
ledged. Photography (other than that by my-
self) is acknowledged in the text. I am indebted
to Miss Helen Williams for her technical assist-
ance.

APPENDIX

Statistical determination of Subspecies

The granting of subspecific status (and hence
trinomials) to populations is arbitrary and de-
pends upon the morphological distinctness of
populations as a whole; subspecies may not be
wholly definable in absolute terms by the char-
acters of each member. Various authors have
discussed the statistics cf this ( see Amadon 1949,
Simpson 1961, p. 174 and Mayr, Linsley &
Usinger 1953, p. 143) giving definitions as leni-
ent as that which requires only 75% of indi-
viduals of a subspecies to be determinable from
75% of another, to more rigid requirements such
as that of Mayr et al. who require 75% of the
individuals of one population to differ from
97% of the other (or 90% from 90%) and
Amadon who requires 75% to be separable from
99.9% (or 97% from 97%).

Mayr et al. propose the use of the statistic
Coefficient of Difference (of samples)

X;, - X,.

(CD) Where X a

S a + Si,

is the higher of the two sample means and s a its
standard deviation.

In this the numerator expresses the distance
between the observed means and where this is
less than the combined standard deviations the
value of CD will be less than 1.0. Using this
statistic, Mayr et al. require a CD value of 1,28
for separation of subspecies and Amadon would
require a CD of 1.9.

As pointed out by Amadon, the use of such
a calculation takes no account of the probable
discrepancy between observed and actual ( i.e
population) means and standard deviations.
This matters little where samples are large, but
in fossil samples where values are frequently
very small, account must be taken of it; or at
least it should be made more difficult to achieve
significance with small samples than with large
ones. Therefore 1 include confidence intervals
in the test which I use here in order to relate
these sample values to populations. Since the
introduction of confidence intervals imposes up-
per and lower limits, there are two (population'
Coefficients of Difference for each set of observa-
tions. These are designated CD>, and CD*.

-

1—

r-

Co- 1 )

/ (“* - 1 ) <

\

n a ^.025

V % ^.025

128

CD,

OS

1 ) -s

AJ

*.975

where: X a is the higher value of the two sample
means,

S ;l its standard deviation,

N a the number of observations.

^ the confidence limit for

the population mean with 95% proba-
bility (tn a the value for t with n a de-
grees of freedom),

/N — 1\ s 2 , /N — n s 2
V a /a ana \ a /a

% 2 v2

n A - -975 n *“*025

a a

the lesser and greater values respectively
of the 95% confidence limits of the vari-
ance (n a degrees of freedom). See Simp-
son, Roe and Leewontin 1960, p. 161.

Similarly Xb is the lower value of the two
sample means, etc.

As mentioned above, taxonomists differ in the
value of CD which they would require before
recognizing distinct subspecies. In this con-
nection, it would seem to be logical that the area
below the curve of a population which the
taxonomist should require to be free of over-
lap from another population should include the
whole of the values which occur at greatest
frequency, i.e., it should include that part of
the curve within the range of 1 standard devia-
tion on both sides of the mean; thus 84% of
the population should be quite distinct. Ideally
this should be distinct from the whole of any
adjacent population but in practice it is prob-
ably sufficient that it should be overlap free
from 97% of it (i.e., 2 standard deviations from
its mean).

Thus, I would regard a population CD value
of 1.5 as adequate to establish subspecies.

Procedure. — In order to test two samples for
subspecific distinctness :

(a) Compare means of samples (t test) in
all variates likely to give significance.

(b) For variates giving significant t values
calculate CDi and CD 2 .

To interpret Results:

(a) If both CD, and CD 2 are 1.5 or greater
the two populations warrant subspeci-
fic separation.

(b) If neither reaches 1.5 it is probable
that the populations are not subspeci-
fically distinct.

(c) If CD i is less than 1.5 and CD> is 1.5
or more then it is not possible with
the material at hand to give subspecific
recognition although further material
might require it.

Comparison of Thylacinus . — In comparison
between samples of modern T. cynocephalus and
Western Cave-fossil Thylacinus, the observed
means of the characters M 2 , M 3 , M'* 4 were sig-
nificantly different (P < 0.01). CD, CD,, and
CD* were calculated for these and are given
here.

CD

CD,

CD a

M*

1*21

0*48

2*23

M*

104

0*32

2*10

M 1 4

0-93

0 16

2*15

In the character M,, where the differences be-
tween the means approach significance at the
5% level, CD 0.336, CD, is a negative quan-
tity, and CD 2 1.28.

References.

Amadon, D. (1949). — The seventy-five percent rule for
subspecies. Condor 51: 250-258.

Ameghino, F. (1903). — Los diprotodontes del orden de los
plagiaulacoideos y el origen de los rcedores
y de los polimastodontes. An. Mus. Nac. B.
Aires (3a) 2: 81-192. Republished (1932) in
Obras completas y correspondencia cientifica
de Florentino Ameghino 13: 889-993.

Balme. B. E. (1963). — Palynological Report No. 98. Lake
Eyre No. 20 Bore, South Australia. Appendix
to Investigation of Lake Eyre. Rep. Invest.
S. Aust. Geol. Surv. no. 24: 89-102.

Barnett. C. H.. and Napier, J. R. (1953).— The form and
mobility of the fibula in metatherian mam-
mals. J. Anat. 87 : 207-213.

Bartholomai, A. (1962).— A new species of Thylacoleo
and notes on some caudal vertebrae of
Palorchestes azael. Mem . Qd Mus. 14: 33-40.

— (1963). — Revision of the extinct macropodid

genus Sthenurus Owen in Queensland. Mem.
Qd Mus. 14: 51-76.

Bensley, B. A. (1903). — On the evolution of the Aus-
tralian Marsupialia; with remarks on the
relationships of the marsupials in general.
Trans. Linn. Soc. Lond. Zool. (2) 9: 83-217.

Broom, R. (1896).— On a small fossil marsupial with
large grooved premolars. Proc. Linn. Soc.
N.S.W . (2) 10: 563-567.

(1896a).— On a small fossil marsupial allied

to Petaurus. Proc. Linn. Soc. N.S.W. (2) 10:
568-570.

(1896b). — Report on a bone breccia deposit

near the Wombeyan Caves, N.S.W., with
descriptions of some new species of mar-
supials. Proc. Linn. Soc. N.S.W. 21: 48-61.

(1898).— On the affinities and habits of

Thylacoleo. Proc. Linn. Soc. N.S.W. 23:
57-74.

Burbidge, N. T. (I960).— The phytogeography of the
Australian Region. Aust. J. Bot. 8: 75-211.

Butler, P. M. (1946). — The evolution of carnassial denti-
tions in the Mammalia. Proc. Zool. Soc.
Lond. 116: 198-220.

Cabrera A. (1919). — “Genera Mammalium. Monotre-
mata, Marsupialia. “ (Museo Nacional de
Ciencias Naturales: Madrid.)

Cain A. J. ( 1959).— Deductive and inductive methods in
post-Linnaean taxonomy. Proc. Linn. Soc.
Lond. 170: 185-217.

Cook D L. (1963). — The fossil vertebrate fauna of
Strong’s Cave, Boranup. Western Australia.
W. Aust. Nat. 8: 153-162.

-(1963a). — Thylacinus and Sarcophilus from

the Nullarbor Plain. W. Aust. Nat. 9: 47-48.

129

De Vis, C. W. (1883). On tooth-marked bones of extinct
marsupials. Proc. Linn. Soc. N.S.W. 8: 187-
190.

(1889).— On the Phalangistidae of the post-

Tertiary period in Queensland. Proc. Roy.
Soc. Qd 6: 105-114.

(1891).— Remarks on post-Tertiary Phascolo-

myidae. Proc. Linn. Soc. N.S.W. (2) 6: 235-
246.

(1891a).— The incisors of Sceparnodon. Proc.

Linn. Soc. N.S.W. (2) 6: 258-262.

(1891b). — In confirmation of the genus

Owenia so-called. Proc. Linn. Soc. N.S.W.
(2) 6: 159-165.

(1894). — A thylacine of the earlier nototherian

period in Queensland. Proc. Linn. Soc.
N.S.W. (2) 8: 443-447.

Dillon, L. (1963). — Comparative studies of the brain in
the Macropodidae. Contribution to the
phylogeny of the mammalian brain II. J
Comp. Neurol. 120: 43-51.

Dollo, L. (1899).— Les ancfctres des marsupiaux, etaient-
ils arboricoles? Trav. Sta. Zool. Wimereux
7: 188-203.

(1900).— Le pied du Diprotodon et l'orlgine

arboricole des marsupiaux. Bull, Sci. Fr.
Belg.: 275-280.

Eiftnmn. H. O. ( 1929).— Functional adaptations of the
pelvis in marsupials. Bull. Amer. Mus. Nat.
Hist. 58: 189-232.

Gidley, J. W. (1915).— An extinct marsupial from the
Fort Union with notes on the Myrmecobidae
and other families of this group. Proc. U.S.
Nat. Mus. 48: 395-402.

Gill, E. D. (1953).— Distribution of the Tasmanian Devil,
the Tasmanian Wolf and the Dingo in S.E.
Australia in Quaternary time. Viet. Nat..
Melb. 70: 86-90.

( 1953a ) —Australian Tertiary marsupials

Aust. J. Sci. 16. 106-108.

(1953b).— Catalogue of Quaternary types and

figured specimens in the National Museum.
Melbourne. Mem.. Nat. Mus., Melb. no. 18:
157-168.

(1954). — Ecology and distribution of the ex-
tinct giant marsupial Thylacoleo. Viet. Nat.
Melb. 71: 18-35.

(1955). — The problem of extinction with

special reference to Australian marsupials.
Evolution 9: 87-92.

(1957). — The stratigraphical occurrence and

palaeoecology of some Australian Tertiary
marsupials. Mem. Nat. Mus., Melb. no. 21:
135-203.

(1961).— The climates of Gondwanaland in

Kainozoic times. Chapter 14, pp. 332-353 in
“Descriptive palaeoclimatology” ed. A. E.
Nairn. (Interscience Publishers: New York.)

( 1962).— Cainozoic. Pp. 233-253 in The

Geology of Tasmania. Ed. by A. Spry and
M. R. Banks. J. Geol. Soc. Aust. 9 (2): x.
107-362.

Glauert, L. (1910) —The Mammoth Cave. Rec. W. Aust.
Mus. 1: 11-36.

Good, R. (1963). — On the biological and physical re-
lationships between New Guinea and Aus-
tralia, in Pacific Basin Biogeography Sym-
posium. 10th Pacific Science Congress 1961:
301-309.

Goodrich. E. S. (1935). — Syndactyly in marsupials. Proc.
Zool. Soc. Lond. 1935: 175-178.

Hall, E. R., and Dalquest, W. W. (1963).— The mammals
of Veracruz. Publ. Mus. Nat. Hist. Univ.
Kans. 14: 165-362.

Hooijer, D. A. (1962). — Palaeontology of hominid de-
posits in Asia. Advanc. Sci.. Lond. 18: 485-
489.

Irving, E., Robertson, W. A., and Stott, P. M. (1963).—
The significance of the paleomagnetic results
from Mesozoic rocks of Eastern Australia.
J. Geophys. Res. 68: 2313-2317.

Jones. F. Wood (1923).—' “The Mammals of South Aus-
tralia’ * Pt. 1. Containing the monotremes
and the carnivorous marsupials. (Govt.
Print.; Adelaide.)

(1931). — A re-examination of the skeletal

characters of Wynyardia bassiana, an extinct
Tasmanian marsupial. Pap. Roy. Soc. Tasm.
1930: 96-115.

Kean. R. I. (1961).— The evolution of marsupial repro-
duction. Tech. Pap. For. Res. Inst. N.Z. No.
35.

Klaauw, C. J. van der (1931).— The auditory bulla in
some fossil mammals with a general intro-
duction to this region of the skull. Bull.
Amer. Mus. Nat. Hist. 62: 1-352.

Krefft, G. (1871). — Description of a new species of
thvlacine (Thylacinus breviceps). Ann. Mag.
Nat. Hist. (4) 2: 296-7.

Ludbrook. N. H. (1963). — Correlations of the Tertiary
rocks of South Australia. Trans. Roy. Soc
S. Aust. 87: 5-15.

(1963a). — Subsurface stratigraphy. Part II of

Investigation of Lake Eyre. Rep. Invest. S.
Aust. Geol. Surv. no. 24: 71-88.

Lydekker. R. (1887). — “Catalogue of the fossil Mammalia
in the British Museum (Natural History).’’
Pt. V I Marsupials 146-2891. (British Museum
(Natural History): London.)

McCoy, F. (?1862). — Geological Survey of Victoria Quarter
Sheet 7.N.W.

(1874). — Tertiary Mammalia. Phascolomys

pliocenus (McCoy). Decade 1, pp. 21-2, pis.
3-5 in “Prodromus of the palaeontology of
Victoria.” (Govt. Print.: Melbourne.)

Marcus, L. F. (1962).— A new species of Sthenurus
(Marsupialia, Macropodidae) from the Pleis-
tocene of New South Wales. Rec. Aust. Mus.
25: 299-304.

Masai, H. (I960).— Several observations on comparative
neuroanatomy. Med. J. Osaka Univ. 10:
303-320.

Mayr. E., Linsley, E. G.. and Usinger, R. L. (1953).-
Methods and principles of systematic zoology.
New York, McGraw Hill, x, 336 p.

Nairn, A. E. M. ed. (1961). — Descriptive palaeoclimat-
ology. New York. Interscience Publishers.
380 p.

Osgood, W. H. (1921).— A monographic study of the
American marsupial, Caenolestes. Publ. Field
Mus. (Zool. Ser .) 14 (1): 1-156.

Owen, R. (1838). — [Report on fossils from Wellington
Caves] in Mitchell. T. L. “Three expeditions
into the interior of Eastern Australia with
descriptions of the recently explored region
of Australia Felix, and of the present colony
of New South Wales.” Vol. 2: 359-366. (T
& W. Boone: London.)

(1840).— Outlines of a classification of the

Marsupialia. Trans. Zool. Soc. Lond. 2: 315-
333.

(1845). — Descriptive and illustrated catalogue

of the fossil organic remains of Mammalia
and Aves contained in the Museum of the
Roval College of Surgeons of England.”
[Marsupialia: 291-336.] (Taylor: London.)

-(1859).— On the fossil mammals of Australia.

Part 1. Description of a mutilated skull of
a large marsupial carnivore (Thylacoleo car-
nifex > Owen) from a calcareous conglomerate
stratum, eighty miles S.W. of Melbourne.
Victoria. Phil. Trans. 1859: 309-322.

(1877).— “Researches on the fossil remains of

the extinct mammals of Australia, with a
notice of the extinct marsupials of England. ’
2 vols. (J. Erxleben : London.)

Pearson. J. (1950). — The relationships of Potoroidae to
the Macropodidae (Marsupialia). Pap. Roy.
Soc. Tasm. 1949: 211-229.

Poccck, R. I. (1926).— The external characters of Thy-
lacinus. Sarcophilus, and some related mar-
supials. Proc. Zool. Soc. Lond. 1926: 1037-
1084

Ride, W. D. L. (1956). — The affinities of Burramys parvus
Broom a fossil phalangeroid marsupial. Proc
Zool. Soc. Lond. 127: 413-429.

— ( 1957 ) . — Protemnodon parma (Waterhouse >

and the classification of related wallabies
(Protemnodon, Thylogale, and Setonix)
Proc. Zool. Soc. Lond. 128: 327-346.

(1959). -Mastication and taxonomy in the

macropodine skull. Pp. 33-59 in “Function
and taxonomic importance”; ed. by A. J.
Cain. (Systematics Assn.: London.)

(1960). — A fossil mammalian fauna of the

Burramvs parvus breccia from the Wom-
beyan Caves, New South Wales. J. Roy. Soc.
W. Aust. 43; 74-80.

(1961) .—The cheek-teeth of Hypsiprymnodon

moschatus Ramsay 1876 (Macropodidae; Mar-
supialia). J. Roy. Soc. W. Aust. 44: 53-60.

(1962) .— On the evolution of Australian mar-
supials. Pp. 281-306 in “The Evolution of
Living Organisms." Ed. G. W. Leeper. (Melb.
Univ. Pr. : Melbourne.)

130

(1962a). — On the use of generic names for

kangaroos and wallabies (subfamily Macro-
podinae). Aust. J. Sci. 24: 367-372.

(1934). — A list of mammals described from

Australia between the years 1933 and 1963
(comprising newly proposed names and ad-
ditions to the Australian faunal list). Bull.
Aust. Mamm. Soc. no. 7: supplement: 1-15.

(1964a). — Antechinus rosamondae, a new spe-
cies of dasyurid marsupial from the Pilbara
District of Western Australia; with remarks
on the classification of Antechinus. W . Aust.
Nat. 9: 58-65.

Runcorn, S. K. ed. (1962).— “Continental drift.” (Aca-
demic Pr.: New York and London.)

Simpson, G. G. (1929). — American Mesozoic Mammalia.
Mem. Peabody Mus. Yale 3: 1-171.

(1930). — Post-Mesozoic Marsupialia. “Fos-

silium Catalogus” 1: Animalia. Pars 47. Ed.
J. F. Pompeckj. (W. Junk: Berlin.)

(1933). — The “plagiaulacoid” type of mam-
malian dentition. J. Mammal . 14: 97-107.

(1941). — The affinities of the Borhyaenidae.

Amer. Mus. Novit. no. 1118: 1-6.

(1941a). — The function of saber-like canines

in carnivorous mammals. Amer. Mus. Novit.
no. 1130: 1-12.

(1945). — The principles of classification and

a classification of mammals. Bull. Amer.
Mus. Nat. Hist. 85: i-xvi, 1-350.

(1947). — A new Eocene marsupial from Brazil.

Amer. Mus. Novit. no. 1357: 1-7.

( 1948) . — The beginning of the age of mam-
mals in South America. Bull. Amer. Mus.
Nat. Hist. 91: 1-232.

(1950).— History of the fauna of Latin

America. Amer. Scient. 38: 361-389.

(1961). — “Principles of animal taxonomy.”

(Columbia Univ. Pr.: New York.)

Simpson, G. G., Roe, A., and Lewontin, R. C. (1960). —
“Quantitative zoology.” Rev. ed. (Harcourt:
New York.)

Sinclair, W. J. (1906). — Mammalia of the Santa Cruz
beds. Marsupialia. Rep. Princeton Exped.
Patagonia 4: 333-460.

Spencer, B. (1901). — A description of Wynyardia bassiana ,
a fossil marsupial from the Tertiary beds of
Table Cape, Tasmania. Proc. Zool. Soc. Lond.
1900: 776-794.

Spencer, B., and Kershaw, J. A. (1910). — A collection of
sub -fossil bird and marsupial remains from
King Island, Bass Strait. Mem. Nat. Mus.,
Melb. 3: 5-35.

Stephenson, N. G. (1963). — Growth gradients among
fossil monotremes and marsupials. Palaeon-
tology 6: 615-624.

Stirling, E. C. (1891).— Description of a new genus and
species of Marsupialia, “Notoryctes typlilops" .
Trans. Roy. Soc. S. Aust. 14: 154-187.

(1913).— On the identity of Phascolomys

(Phascolonus) gigas , Owen, and Sceparnodon
ramsayi, Owen; with a description of some
parts of its skeleton. Mem. Roy. Soc. S.
Aust. 1: 127-178.

Stirton, R. A. (1955). — Late Tertiary marsupials from
South Australia. Rec. S. Aust. Mus . 11: 247-
268

(1957). — Tertiary marsupials from Victoria,

Australia. Mem. Nat. Mus., Melb. no. 21:
121-134.

(1957a). — A new koala from the Pliocene

Palankarinna fauna of South Australia. Rec.
S. Aust. Mus. 13: 71-81.

(1963). — A review of the macropodid genus

Protemnodon. Univ. Calif. Publ. Geol. Sci.
44: 97-161.

Stirton, R. A., Tedford. R. H., and Miller, A. H. (1961). —
Cenozoic stratigraphy and vertebrate paleon-
tology of the Tirari Desert, South Australia.
Rec. S. Aust. Mus. 14: 19-61.

Tate, G. H. H. (1947). — On the anatomy and classifica-
tion of the Dasyuridae (Marsupialia). Bull.
Amer. Mus. Nat. Hist. 88: 101-155.

(1948). — Studies on the anatomy and phylo-

geny of the Macropodidae (Marsupialia).
Bull. Amer. Mus. Nat. Hist. 91: 237-351,

(1951). The wombats (Marsupialia, Phas-

colomyidae). Amer. Mus. Novit. no. 1525:
1-18.

Thomas, O. (1887). — On the wallaby commonly known
as Lagorchestes fasciatus. Proc. Zool. Soc.
Lond. 1886: 544-547.

(1888). — “Catalogue of the Marsupialia and

Mcnotremata in the collection of the
British Museum (Natural History).” (British
Museum (Natural History): London.)

Troughton, E. Le G. (1959). — The marsupial fauna: its
origin and radiation. Pp. 69-88 in “Bio-
geography and ecology in Australia.” Ed. by
A, Keast and others. Junk: The Hague.)

Van Deusen, H. M. (1964). — First New Guinea record of
Thylacinus. J. Mammal, (in press).

Wakefield, N. H. (I960). — Recent mammal bones in the
Buchan District. Viet. Nat., Melb. 77: 164-178,
227-240.

Waterhouse, G. R. (1846). — “Natural history of the
Mammalia.” Vol. 1. Marsupiata. (Bailliere:
London.)

Wood, H. E. (1924). — The position of the “sparassodonts"
with notes on the relationships and history
of the Marsupialia. Bull. Amer. Mus. Nat.
Hist. 51: 77-101.

Woods, J. T. (1956). The skull of Thylacoleo carnifex.
Mem. Qd Mus. 13: 125-140.

(1958). — The extinct marsupial genus Palor-

chestes Owen. Mem. Qd Mus. 13: 177-193.

(1960). — The genera Propleopus and Hypsi-

prymnodon and their position in the Macro-
podidae. Mem. Qd Mus. 13: 199-212.

(1960a). — Fossiliferous fluviatile and cave de-
posits. Pp. 393-403 in “The geology of
Queensland”; ed. by D. Hill and A. K.
Denmead. (Melb. Univ. Pr.: Melbourne.)

[Originally published as vol. 7 of the Journal
of the Geological Society of Australia.]

(1962). — Fossil marsupials and Cainozoic con-
tinental stratigraphy in Australia: a review.
Mem. Qd Mus. 14: 41-49.

131

CONTENTS OF VOLUME 47

Part 1 < published 24th April, 1964 )

1. The Environment of the Quokka (Setonix brachyurus) in the Darling

Range, Western Australia. By G. M. Storr 1

2. An Ordovician Cystoid (Pelmatozoa, Echinodermata) from Western Aus-
tralia. By Ida A. Brown 3

3. Description of a new Stonefish of the family Synanceidae from Western

Australia. By R. J. McKay. 8

4. Association of Larger and Planktonic Foraminifera in single samples from

Middle Miocene sediments, Guadalcanal, Solomon Islands, south-west
Pacific. By P. J. Coleman and R. A. McTavish 13

5. Discontinuous and Presumed Vicarious Plant Species in Southern Aus-
tralia. By J. W. Green. 25

Part 2 (published 13th July, 1964)

6. The Woolgorong Stony Meteorite. By G. J. H. McCall and P. M. Jeffery. 33

7. The agamid Lizards of the genus Tympanocryptis in Western Australia.

By G. M. Storr. 43

8. On Beryl from Western Australia. By P. L. C. Grubb 51

9. Contributions Florae Australiae Occidentals XIII. By C. A. Gardner. 54

Part 3 (published 30th October, 1964 )

10. Notes on the Skulls of Two Western Australian Rodents with a key to the
Skulls of the Rodents of Southwestern Australia. By Ernest Lundelius, Jr. 65

11. Effects of Salt, Temperature and Seed Scarification on Germination of

Two Varieties of Arthrocnemum halocnemoides. By C. V. Malcolm. 72

12. Ashbed and Nutrients in the Growth of Seedlings of Karri (Eucalyptus

diversicolor F. v. M.). By O. W. Loneragan and J. F. Loneragan 75

13. Foraminiferal Evidence for the Paleocene Age of the King’s Park Shale

(Perth Basin, Western Australia). By Brian McGowran 81

14. Projasus — a New Generic Name for Parker’s Crayfish, Jasus parkeri Steb-
bing (Palinuridae: “Silentes”). By R. W. George and J. R. Grindley. 87

15. Geographical Variation and Distribution of some Birds from Western Aus-
tralia. By G. F. Mees 91

Part 4 (published 29th December, 1964 )

16. A Review of Australian Fossil Marsupials. Presidential Address, 1963. By
W. D. L. Ride.

97

The Royal Society of Western Australia

Council 1963-1864

President
Past President
Vice-Presidents

Joint Hon. Secretaries

Hen. Treasurer
Hon. Librarian
Hon. Editors

A. S. George, B.A.
A. B. Hatch, B.Sc.
L. E. Koch, M.Sc.
R. J. Little.

D. Merrilees, B.Sc.
W. C. Packer, Ph.I
P. E. Playford, B.JE
L. W. Samuel, B.S

C. F. H. Jenkins, M.A.

W. D. L. Ride, M.A., D.Phil.

W. R. Wallace, Dip. For.

J. H. Lord, B.Sc.

Margaret E. Redman, B.Sc.

J. G. Kay, B.Sc.

R. D. Royce, B.Sc. (Agric.).

Ariadna Neumann, B.A.

J. E. Glover, B.Sc., Ph.D. (1963).
R. W. George, B.Sc., Ph.D. (1964).

Dip.For.

Ph.D.

., Ph.D., F.R.A.C.I., F.R.I.C.

Council 1964-1965

President
Past President
Vice-Presidents

Joint Hon. Secretaries

Hon. Treasurer
Hon. Librarian
Hon. Editors

W. R. Wallace, Dip.For.

C. F. H. Jenkins, M.A.

J. H. Lord, B.Sc.

D. Merrilees, B.Sc.

Margaret E. Redman, B.Sc.

A. B. Hatch, M.Sc., Dip.For.

R. D. Royce, B.Sc. (Agric.).

Ariadna Neumann, B.A.

R. W. George, B.Sc., Ph.D. (1964).
J. E. Glover, B.Sc., Ph.D. (1965).

B. E. Balme, B.Sc. (Hons.).

J. S. Beard, M.A., B.Sc., Ph.D.

A. S. George, B.A.

J. G. Kay, B.Sc.

P. E. Playford, B.Sc., Ph.D.

R. T. Prider, B.Sc., Ph.D., M.Aust.I.M.M., F.G.S.
W. D. L. Ride, M.A., D.Phil.

D. L. Serventy, B.Sc. (Hons.). Ph.D.