JOURNAL OF

THE ROYAL SOCIETY

OF

WESTERN AUSTRALIA

INCORPORATED

VOLUME 45 (1962)
PART 2

PUBLISHED 31ST AUGUST, 1962

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

THE

President

Past President ....

Vice-Presidents ....

Joint Hon, Secretaries

Hon. Treasurer ....
Hon. Librarian ....
Hon. Editor

ROYAL SOCIETY

OP

WESTERN AUSTRALIA

INCORPORATED

COUNCIL 1961-1962

J. E. Glover, B.Sc., Ph.D.

N. H. Brittan, B.Sc., Ph.D.

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

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

R. W. George, B.Sc., Ph.D.

C. V. Malcolm, B.Sc. (Agric.).

L. J. Hollis, B.A., Dip. Com.

G. G. Smith, M.Sc.

J. E. Glover, B.Sc., Ph.D.

J. G. Kay, B.Sc.

R. J. Little.

D. Merrilees, B.Sc.

R. T. Prider, B.Sc., Ph.D., M.Aust.I.M.M., F.G.S.
R. D. Royce, B.Sc. (Agric.).

L. W. Samuel, B.Sc., Ph.D., F.R.I.C., F.R.A.C.I
D. W. Stewart, B.Sc. (For.).

W. R. Wallace, Dip. For.

Journal
of the

Royal Society of Western Australia
Vol. 45 Part 2

5. — Some Interesting Stomatapoda — mostly from Western Australia

By W. Stephenson*

Manuscript received — 20th February, 1962

Twenty-five species of stomatopods are dealt
with, stress feeing laid upon geographical dis-
tribution. Additional localities are given for
eight species previously known from Western
Australia, together with eight new records for
Western Australia, one new record for Lord
Howe Island, and three for Australia.

Three species are recorded from Christmas
Island (Indian Ocean) and one from the
Molluccas.

Squilla microphthalma H. Milne Edwards is
redescrlbed and Lysiosquilla brazieri Mlers
resurrected as a subspecies of L. latifrons (de
Haan).

Introduction

Since the publication of a check list of
the Australian stomatopods (Stephenson and
McNeill 1955) and a recent short note (Stephen-
son 1960) additional specimens from five main
sources have been examined: —

(a) extensive collections from the Western
Australian Museum, Perth;

(b) a few specimens from the Australian
Museum, Sydney, sorted from general
collections made in North Western
Australia ;

(c) specimens recently obtained in Eastern
and Northei*n Australia;

(d) a small collection from Christmas
Island in the Indian Ocean; and

(e) the small series of stomatopods in the
Macleay Museum (University of
Sydney).

In reporting upon these, specimens falling
within the known limits of distribution have
been omitted unless of special interest. The
following categories are considered separately
below: —

(a) species whose known distribution has
been extended;

(b) new Australian records; and

(c) redescribed species.

In this paper, lengths are measured in a mid-
dorsal line from the posterior end of the telson
(as near practicable, excluding spines) to the
anterior edge of the carapace excluding the
rostrum.

Extensions to Ranges

In general, only brief literature citations are
given, these being adequate for identification.

♦ Department of Zoology. University of Queensland. St.

Lucia. Queensland

Puller references and details of the known Aus-
tralian distributions are given in Stephenson
and McNeill (1955).

SQUilla fabricii Holthuis, 1941, pp. 249-53, text-
fig. 1.

One, Macleay Museum, labelled “Molluccas”.
The species is known from very few specimens
from the erstwhile Dutch East Indies.

Squilla granti Stephenson. 1953, pp. 201-8, Pig.
lA-D, Pig. 2A, B, D, F. G.

Two, Port Walcott, W. Australia, 8 fm, sand
with occasional sponges and corals, coll. R. D.
Royce (“Davena”), 3/vi/1960, W. Aust. Mus.
Reg. No. 20 (b)-61.

Two, 4-5 miles from Urquhart Point, 3-3i fm,
Westminster dredging, Weipa. Emberley River,
Gulf of Carpentaria, Queensland, coll. G.
Webster, l/vii/1961. Aust. Mus. Reg. No. P12543.

Previously known only from Queensland. The
present specimens differ from the original
material (Stephenson 1953) in certain minor
respects, e.g.* more pigmented. Black pigment
lines the gastric grooves of the carapace and
is present on the outer uropod. Here diffuse
patches occur on the distal portions of the
penultimate segments and on the proximal por-
tion of the ultimate segments.

Squilla incrnata Tate. Kemp, 1913, pp. 70-2, PI.
V, Figs. 57-9 (as S. oratoria var perpeitsa) .
Two, Entrance Point, Broome, W. Australia
between tides on rocky reef shore, Aug. 1929,
coll. A. A. Livingstone, Aust. Mus. Reg. No.
P13545.

New record for Western Australia.

Squilla laevis Hess. Kemp, 1913, pp. 49-50, PI.
Ill, Pigs. 35-7.

The Western Australian Museum collections
include: —

One, Exmouth Gulf, W. Australia, trawled,
coll. R. McKay. M.V. “Lancelin”, Oct. 1958, Reg.
No. 8-61.

Two, 40 miles S.W. of Carnarvon, Shark Bay,
W. Australia, trawled, coll. via. A. Snell. June,
1960, Reg. No. 42-61.

The Australian Museum collections include: —
One, between Broome and Wallal (Ninety
Mile Beach), W. Australia, dredged about 7 fm,
coll. R. Bourne, Reg. No. P10016.

33

The species is recorded from Queensland, New
South Wales, Victoria, South Australia and
Western Australia. The present records extend
the northern range in Western Australia, just
as recent collecting has extended the Queensland
range (Stephenson 1960).

Squilla mcneilli Stephenson, 1953, pp. 213-8,
Fig. 4A-F.

Dr. A. A. Racek reports (in correspondence): —
“This species has been found very commonly
between about 60 to 120 fathoms in an area
stretching from about due east of Barranjoey
lighthouse (Hawkesbury River' to due east of
Stockton, and in fact it has been so common
that I did not suspect it being an unusual

species at all Whatever the optimal depth

of this species might be, S. moieilli does not
seem to occur at all this side of 100 m or 50
fathoms.” Eleven species collected in the above
area, 3rd-9th July, 1959, were kindly forwarded
by Dr. Racek.

The majority of specimens had raptorial
dactyli with six teeth not five as in the type
material. Some specimens had six teeth on
one claw and five on the other.

The species had been recorded off the New
South Wales coast from Newcastle to Green
Cape in 25-90 fm.

Squilla miles Hess. Hale. 1924, pp. 492-5, PI.
XXXII, Pig. 1, text-fig. 381a-i.

One. near Albany. Cape Riche, W. Australia,
herring net. coll. A. Kalnins, March, 1954. W.
Aust. Mus. Reg. No. 38-61.

Juvenile (30mm), Shark Bay, W. Australia,
night drift. M.V. “Lancelin”. 31/iii/1957, W.
Aust. Mus. Reg. No. 9-61.

Two, 4 Mile Reef off Busselton, 60 ft, stomach
cf fiddle-ray. coll. B. Wilson. 28/xii/T958, W.
Aust. Mus. Reg. No. 46-61.

Common on the southern shores of Australia.
Previously recorded from Western Australia (but
from an unknown locality t by Alexander
(1916a). from Cottesloe by Hale (19296) and
from Albany by Stephenson and McNeill (1955).
The present records extend the northern range
of the species, particularly since Hess’ (1865)
locality of “Sydney” is suspect (see Stephenson
and McNeill 1955).

Squilla raphidea Fabricius. Kemp, 1913, pp. 88-
92, PI. Vn. Fig. 77.

One, Wyndham. W. Australia, coll. R. G. Pat-
terson, 1956. W. Aust. Mus. Reg. No. 170-56.

New record for Western Australia, previously
known from the Northern Territory, New South
Wales and Queensland.

Squilla terTareginensis Stephenson, 1953, pp.
208-13. Fig. 3A, B.

Male, Carnarvon, W. Australia, off reef, N.
Paul via P. Crackel, Feb. 1962, W. Aust. Mus.
Reg. No. 148-62.

Known previously only from the type material
from N. Queensland.

Lysiosquilla multifasciata Wood-Mason. Chopra,
1939, pp. 162-5, Pigs. 8, 9.

Two, Barred Creek, 40 miles N. of Broome,
W. Australia, mangrove fiats, low tide. coll. A.

Kalnins, 5-10/i/1960, W. Aust. Mus. Reg. No.
35-61.

One. Port Walcott. W. Australia. 8 fm sand
with occasional sponge and corals, coll. R. D.
Royce (“Davena”). 3/vi/1960. W. Aust. Mus.
Reg. No. 20 (d)-61.

New records for Western Australia. Only one
specimen was previously known from Australia,
from Dunk I.. Family Group. Queensland.

Lysiosquilla osculans Hale, 1924, pp. 501-2, PI.
XXXIII, Pig. 3, text-fig. 384 (as Lysio-
squilla vercoi var osculausi.

Two. Cottesloe. W. Australia, from gullet of
flounder, coll. D. Diamond. 1940, W. Aust. Mus.
Reg. No. 14/15-40.

New record for Western Australia, previously
known only from Victoria and South Australia.

Pseudosquilla ciliata (Fabricius). Kemp, 1913,
pp. 96-100. Bigelow. 1931. pp. 152-6, text-
figs. 3-6. Dollfus, 1938, pp. 198-200, Fig. 8
(with synonymy).

A single specimen is present in the Macleay
Museum labelled “Lord Howe Island.”

Within Australia previously recorded only
from Queensland.

P ;eudosquilla ornata Miers. Kemp, 1913, pp.
100-1. Komai, 1927. pp. 324-5. PI. XIV,
Pigs. 2-2b.

One, Christmas I., coll. E. Carr, July, 1961,
W. Aust. Mus. Reg. No. 110-61.

A widespread Indo-West Pacific species, ap-
parently never common, and not so far recorded
from Australia.

Gonodactylus chiragra (Fabricius). Kemp, 1913,
pp. 155-62. PI. ix. Fig. 107, Pig. 2 on p. 161.
Dollfus, 1938. pp. 205-13, text-figs. 14, 15
(with synonymy).

This widely distributed species is represented
in the Western Australian Museum by collec-
tions from: —

Yampi Sound; Point Gantheaume; Broome;
Port Hedland; Delambre I.. Enderby I. (Dampier
Archipelago); Point Cloates; Maud Landing;
Carnarvon; Cape St. Cricq; Cape Inscription;
North I.. Rat I. and Gun I. (Abrolhos Group);
Quobba; Dixon I., Nichol Bay; and also from
Christmas I.

Previously recorded in Western Australia by
Miers (18806) from Swan River and the
Abrolhos, by Pocock (1893) from Baudin I.,
Troughton I., Damma I., and Baleine Bank, by
Alexander (1916a, 6> from Port Hedland, and
by Balss (1921) from the Cape Jaubert vicinity.

Go^iodactylus falcatus (Forskal). Kemp, 1913,
pp. 167-9, PI. IX, Pig. 113, text-fig. 2 (as
G. glabrous Brooks). Dollfus. 1938, pp. 217-
222, figs. 18-20 (as G. glaber Brooks, with
synonymy).

This widely distributed species is represented
in the Western Australian Museum collections
by specimens from: —

Yampi Sound; Thevenard I.; Shark Bay; and
Cockburn Sound. It has been recorded from
Shark Bay by Alexander (1916a) and Dirk
Hartog I. by Hale (19296).

34

Cxonodactylus graphurus Miers. Kemp, 1913, pp.
169-71, text-fig. 1 on p. 170.

This widely distributed species is represented
in the Western Australian Museum collections
by specimens from: —

Broome: Port Hedland; Malus I., Gidley I.,
and Mermaid Straits (Dampier Archipelago) ;
Wreck Point. Abrolhos Is.; and Port Walcott.
Previously recorded in Western Australia from
Nichol Bay by Miers (1880b). from N.W. Aus-
tralia. Baudin 1., and Baleine Bank by Pocock
(1893) and from the Cape Jaubert vicinity by
Balss (1921),

Gonodaciylus pulchellus Miers. Kemp, 1913, pp.
177-9. PI. X. Pigs. 117-8. Dollfus, 1938, pp.
224-6, Fig. 22 (with synonymy).

Two off Gantheame Point. Broome. W. Aus-
tralia. dredged 4 fm. Aug. 1929, coll. A. A.
Livingstone, Aust. Mus. Reg. No. 13544.

One, 25 miles N.W. of Angel I.. Dampier
Archipelago. W. Australia, “Honolulu” dredge,
37 fm, sand. coll. R. D. Royce (“Davena”),
2/vi/1960. W. Aust. Mus. Reg. No. 18-61.

New record for Western Australia, with only
two previous Australian records, from Princess
Charlotte Bay and Hayman Island, both in
Queensland.

Gcnodactylus stoliurus ( Muller ). Kemp, 1913,
pp. 184-5.

One, Maud Landing. N.W. Australia, 10 fm.
weed and sand on rock, “Honolulu” dredge, coll.
R. D. Royce (“Davena”) 20/V/1960, W. Aust.
Mus. Reg. No. 10-61.

One, 5 Mile Fence, S. of N.W. Cape. W. Aus-
tralia, reef at low tide. coll. R. George and P,
Cawthorn, 3/iv/T961, W. Aust. Mus. Reg. No.
52-61.

According to Pocock (1893), Hansen con-
sidered that Miers’ (1880b. c) record of G.
irispmosns from Shark Bay refers to the present
species. G. stoliurus has been recorded from
Lancelin Is., Western Australia (Stephenson and
McNeill 1955).

Gonodaciylus trispmosus Dana. Borradaile, 1898,
p. 33, Pi. V. Figs. 1, la (as Protosquilla
trispmosa). Kemp. 1913, pp. 180-1.

One, with malformed telson, N.E. of Malus
I., Dampier Archipelago, W. Australia, 10 fm,
“Honolulu” dredge, coll. R. D. Royce
(“Davena”), 31/V/T960, W. Aust. Mus. Reg. No.
13 (b)-60.

The left half of the central portion of the
telson is deformed and bent ventrally, giving a
wider than normal separation between right and
left halves, with resultant difficulty in keying
the specimen.

Recorded previously from W. Australia as
follows: —

Swan River (Miers 1880b), Baleine Bank
(Pocock 1893) and Cape Jaubert vicinity (Balss
1921): also from Queensland (Hale 1929a,
Stephenson and McNeill 1955). Nevertheless
sufficiently uncommon to be worthy of mention.

The specimen from Swan River mentioned by
Miers (1880b, c) was evidently White's (1847)
type. Although G. trispinosus White 1847 is a
nomen nudum (see Kemp 1913, p. 180), the name
G. trispinosus was evidently first applied to this

Western Australian specimen (see Miers 1880b,
c). Two further Western Australian specimens
referred by Miers (1880b, c) to the present
species are now regarded as G. stoliurus.

Odontodactylus cultrifer (White). Kemp, 1913,
pp. 137-8, pp. 138-9 (as O. carnifer (Po-
cock)).

Two, W. side. Exmouth Gulf, W. Australia,
in trawl, coll. K, Godfrey, M.V. “Lancelin”,
26/ii/1956, W. Aust. Mus. Reg. No. 41-61.

New record for Western Australia, previously
recorded from Queensland.

Odontodactylus japonicus (de Haan). Komai,
1927, pp. 335-6, PI. XIII. Fig. 2.

One, Shark Bay or Exmouth Gulf, W. Austra-
lia, trawled, coll. R. McKay, M.V. “Lancelin”.
June-Oct. 1958, W. Aust. Mus. Reg. No. 14-61.

Alexander’s (1916b) record from Broome,
Western Australia was queried by Stephenson
and McNeill (1955), but since the species is now
known from Western Australia and from the
Capricorn Group, Queensland (Stephenson
1960), this querying appears unjustified.

Odontodactylus scyllarus (L.) Kemp, 1913, pp.
135-7. Komai, 1927. pp. 335-6, PI. XIII, Pig.
2 .

One, Christmas I., coll. E. Carr, July, 1961,
W. Aust. Mus. Reg. No. 111-61.

While widespread throughout the Indo-West
Pacific area, this species nowhere appears com-
mon. A specimen from Wewak, New CJuinea,
was recently received at the University of
Queensland.

Ne^v Australian Records
Squilla gcnypetes Kemp 1911

Squilla affiiiis Pocock. 1893, p. 474 (pariim fide
Kemp. 1913).

Squilla gorypetes Lloyd. 1908. p. 33 (sine desc.).
Squilla gonypetes Kemp, 1911. pp. 96-7: 1913, pp.
54-5, PL* IV, Figs. 42-44. Sunler, 1918, pp. 66-7.
Kemp end Chopra, 1921. pp. 300-1. Hansen.
1926, p. 10. Gravier. 1938. pp. 166-8, Fig. 1.
Stephenson and McNeill. 1955, p. 256 (In key).

9 (34 mm> approx. 10 miles N. of Long I.,
off Onslow, W. Australia, coll. B. R. Wilson
(“Davena”), 17/vi/T960. W. Aust. Mus. Reg. No.
29-61.

The pigmentation of the present specimen
follows exactly that mentioned by Pocock
(1893) and described by Kemp (1913). Pre-
viously the species had been recorded from
Burma, the Andaman Is., Arakan Coast, Per-
sian Gulf. Holothuria Bank (China Sea), Mergui
Archipelago, the Java Seas and the Gulf of
Suez, but not from Australia.

Squilla multicarinata White 1848

Squilla multicarinata White 1848. p. 144, Annulosa
PL VI. Fig. I, la; 1849. pp. 381-2. Miers
1880a. p. 20. Bigelow, 1895. p. 511 (In key).
Noblll, 1903, p. 38. Kemp. 1913, pp. 86-8, p. 196.
PL VI. Figs. 73-6. Sunler, 1918. p. 70. Kemp
and Chopra, 1921, p. 307. Paris!. 1922, pp. 102-3.
Komai, 1927, p. 322. Gravier. 1938, pp. 174-7.
Fig. 4.

Damaged 9 (estimated length c. 39 mm), 5
or 6 miles off Bezout I., Dampier Archipelago,
W. Australia, coll. B. R. Wilson (“Davena”),
5/vi/1960. W. Aust. Mus. Reg. No. 27-61.

35

PLATE I

Figs. A. B, C, D — Lysiosquilla latifrons brazieri.

Fig. A.— 35mm specimen (whole); Pig. B.— 55 mm specimen (carapace); Fig. C.— 80 mm specimen (telson); Fig.
D.— 75 mm specimen (telson). Figs. E, F, G—Squilla inicrophthalvia; Fig. E.— 32 mm specimen (carapace, etc.);'
Fig. P.— 39 mm specimen (telson); Fig. G.— 28 mm specimen (telson).

36

Fig.
of L.

— Lystosquilla latifrons brazieri — a. — Carapace, etc. of L. latifrons after de Haan (1844); b.— Carapace etc
latifrons after Komai (1927); c. — Telson of L. latifrons after de Haan (1844); d. — Telson of L. latifrons
after Komai (1927); e. — Telson of 80 mm specimen; f. — Telson of 75 mm specimen.

37

Previously known from Japan, the Philippines,
Singapore. Hongkong, Christmas I., South India
(Kemp 1913 >, Java Seas (Sunier 1918), Singa-
pore (Kemp and Chopra 1921). and Gulf of
Suez (Gravier 1938), but not from Australia.

In the key of Stephenson and McNeill (1955i.
the species comes out by the following route: —
1. 2, 4. 23 (first alternative). With 5. lirata
Kemp and Chopra (1921) it is distinguished by
the possession of a bilobed lateral process on
the sixth thoracic somite, and by the entire
surface of carapace and abdomen being multi-
carinate. It is separated from S. lirata by the
possession of a mandibular palp, by having five
teeth on the raptorial dactylus. and by other
features (see Kemp and Chopra 1921).

Redescribed Species

Lysiosquilla latifrons brazieri Miers 1880a

(Figs. le. If; Plate 1, Figs. A. B. C. D)

Lysiosquilla brazieri Miers, 1880a. p. 11. PI. I. Pigs.

3-6: 1880b. p. 125. Hiiswell. 1882, p. 206. Bigelow.

1893, p. 503 (in key). ChlltrOn, 1911, p. 139.

Lysiosquilla latifrc7is (dc Haan). Kemp. 1913. pp.

128-9 (partim). Stephenson and McNeill. 1955,

p. 248.

non Squilla latifrons de Haan. 1844, PI. LI, Pig. 3;

1849. p. 222.

non Lysiosquilla (Corcnis) latifrons (de Haan).

Miers 1880a. pp. 10-11.

non Lysiosquilla latifrons (de Haan). Rathbun.

1902. p. 54. Komai, 1927, pp. 333-5. PI. XIV.

Pigs. 3. 3a. 3b.

Discussion of Synonymy

Miers’ original description, based upon a
female spscimen from New South Wales, dis-
tinguished L. brazieri from L. latifrons as fol-
lows; — (a) having six teeth on each raptorial
dactylus instead of seven, (bi having a telson
armed with more numerous spinules (c. 14), (c)
not possessing a median sinus on the telson. and
(d) in having the short appendages of the last
pair of thoracic limbs almost linear.

These distinctions have been shown to have
no taxonomic value.

(a) Rathbun (1902) noted that de Haan's
original description of S. latifrons from Japan
did not conform with his figure, where six teeth
are shown on the right raptorial dactylus. Also
her own Japanese specimen possessed six teeth
on each dactylus. The same applies to the
Japanese specimen described by Komai (1927).

(b) Rathbun (1902) noted in her Japanese
specimen that the telson bore 12 small spines on
one side and 11 on the other, and later Chilton
(1911) noted upon New Zealand specimens of
L. brazieri that there were only 10 spinules upon
each side of the telscn.

(c The absence of a median sinus is of little
value, because Chilton’s material had ‘’a slight
indication of a sinus on the posterior margin of
the terminal segment”.

(d) Kemp (1913) inferred that a linear short
appendage on the last pair of thoracic limbs
was present in de Haan’s species, and this is
confirmed by Komai’s (1927) description.

The described distinctions between Miers* and
de Haan’s species are invalid and all recent
workers have relegated Miers’ name to the
synonymy (Chilton 1911, Kemp 1913, Komai
1927. and Stephenson and McNeill 1955).

However, certain other characteristics are of
possible taxonomic value. These are colour,
and ornamentation of the telson.

Colour, — Two freshly collected specimens from
the New South Wales/Queensland border waters
showed constant colour differences as compared
with the Japanese specimen described by Komai.
Certain cf these differences were visible even
in an old dried specimen from Port Jackson in
the Australian Museum.

Ornamentation of the telson, — There are con-
stant differences between the Australian
material and the Japanese.

To date only three Japanese specimens have
been reported in the literature (de Haan’s,
Rathbun 's. and Komai’s), plus two from New
Zealand (Chilton’s — omitting the numerous
washed up specimens which were not retained),
and two from Australia (Miers’, and that of
Stephenson and McNeill which is relisted be-
low). Attempts have been made to obtain
material from New Zealand (particularly Chil-
ton's specimens) but these have failed. Until
more material has been examined it must re-
main uncertain whether there are intergrades
in colouration and telson ornamentation be-
tween the Northern and Southern Hemisphere
forms. On the present inadec^uate data the two
forms appear to be separate both geographi-
cally and morphologically, although the struc-
tural differences are much less than usually
found between stomatopod species. For these
reasons Miers’ species is revived and given sub-
specific status.

Material Examined

9 (75 mm), Port Jackson. N,S. Wales, coll.
Mr. Tiley, dredge ‘'Samson”, Aust. Mus. Reg.
No. P5487. (Note: the abdominal segments are
telescoped in this dried specimen, which is con-
siderably the largest of the three examined.)

9 (80 mm), of Tweed Heads, Queensland/
N.S. Wales border, prawn trawled on clean sand
in 20 fm, ll/ni/1961, coll. W.S. (Squilla laevis
was the dominant stomatopod in the collection) .
(Note: left raptorial claw missing; specimen to
be deposited in the Australian Museum.)

9 (55 mm), off Kingscliff, Northern N.S.
Wales, in prawn trawl at 31 fm, June. 1961, coll.
L, Wale. (Note; both raptorial claws missing;
specimen to be deposited in the Queensland
Museum.)

Material Illustrated

Whole specimen and carapace — 55 mm speci-
men. Telson — 80 mm specimen (Fig. le; PI. 1,
Pig. O; 75 mm specimen (Pig. If; PI. 1. Fig. D).

Description

The present material agrees with Komai’s
(1927) description in all respects excepting the
following: —

Colour. — In L. latifrons Komai describes the
carapace as ‘‘marked with three dark bands
rather obviously defined”. In the present speci-
mens the posterolateral areas of the carapace
are conspicuously sooty black while the anterior
halves ai'e diffusely mottled and spotted with
black to a definite pattern (see PL I, Figs. A.
B). The central of Komai’s three dark bands

38

is represented only by two dark spots. Addi-
tional darker pigmented areas on the Australian
specimens comprise — the anterolateral portions
of the first abdominal somite and the endopo-
dite of the uropod. The latter carries a central
paler line.

Telson. — Komai describes this as “nearly as
long as broad” but in his figure, and also in de
Haan’s, and in the present specimens it is about
half as long as broad.

Komai described the ornamentation of the
dorsal surface of the telson as follows: —

“ set with a row of seven acute spines

situated at equal intervals near the posterior
margin: of these spines the median somewhat
surpasses the rest in thickness; the other three
pairs are gradually longer outwards; the sub-
median and intermediate spines as well as the
latter and the lateral spine, are interposed each
with a shallow longitudinal furrow, while such
a furrow does not exist between the submedian
spines and the median spine, so that the median
three spines make one group, and the lateral
four are separated from the other.” Komai ’s
description agrees closely with de Haan's
original figure of the telson (see Figs. Ic, Id).

In the present material Komai’s seven acute
spines are present, with a similar central group-
ing of three, and with four more lateral spines.
However, the intermediate spines are separated
from the submedians by a broad and fairly
deep excavation typically bearing additional
spines. The laterals are separated from the
intermediates by an almost equally broad, but
typically shallow excavation. These features
are clearly shown in Miers’ figure on L. hrazieri
(Miers 1880a, PI. I. Fig. 3t. On a lower level
than the above large spines, and slightly pos-
terior to them (but still well forward of the
terminal spinulation) further spinules or spines
are present, either as rows or reduced to
isolated spines (see Figs. le. If; PI. I. Figs. C.
D).

The following variations in spinulation of the
dorsal surface of the telson occur in the present
material: —

75 mm specimen. — Left space between sub-
median and intermediate dorsal spines with one
spine and a spinule, right space with a tubercle
and an elongated spine. Two confusedly ar-
ranged lower rows of spines are present, com-
prising 11 spines.

80 mm specimen. — Spaces between submedian
and intermediate dorsal spines each with a long
spine. On the left there are two smaller spines,
and on the right a single smaller spine, these
corresponding to the lower rows of the above
specimen.

55 mm specimen. — As 80 mm specimen except
single lower spine or spinule on each side.

The numbers of smaller spines on the posterior
margin of the telson between the mobile sub-
medians are as follows: — 75 mm specimen — 8
spines and one minute spinule on either side
of the mid line; 80 mm specimen — 9 spines on
each side with a minute spinule just left of
centre; 55 mm specimen — 9 spines on each side.
Miers’ figure of L. hrazieri (1880a, PI. I. Fig. 3)
shows about 7 spines on either side (total c. 14)

not 14 on either side as Chilton inferred. Chil-
ton’s own matei’ial had 10 spines on either side.
These numbers overlap with those of L. latifrons
from Japan, which are evidently at least equally
variable.

In dorsal view the lateral margins of the
carapace in L. latijrons have been figured as
rounded by de Haan (see Pig. la), and rela-
tively straight by Komai (see Fig. lb). In the
present specimens they are rounded (see PI. I,
Figs. A, B).

Sqiiilla microphthalma H. Milne-Edwards, 1837

(Figs. 2a. 2d; Plate I. Figs. E. F, G)

Squilla microphthalma. H. Milne-Edwards. 1837, p.
523. de Haan. 1849. p. 221. Bigelow. 1895, p.
509. Jurlch, 1904, pp. 368-9. PI. XXVI, Pig. 1.
Kemp. 1913. pp. 31-3, PI. I. Pigs. 17-20. Kemp
and Chopra, 1921, pp. 299-300. Serene. 1952,
pp. 5-11. text-figs. 10. 11. 16. 17, 20; PI. I. Figs.
2, 5; PI. II. Pigs. 2. 5 (under S. depressa).

Chlorida microphthalma (H. Milne Edwards).
Eydoux and Souleyet. 1841. pp. 264. 266 (fide
Kemp 1913 — not seen).

Chloridella microphthalma (H. Milne-Edwards).
Wood-Mason. 1895. p. 8. PI. IV. Figs. 1-5. de
Man. 1898. pp. 691-3, PI. 38. Figs. 76. 76a.

non Chloridella microphthalma (depressa) Miers
1880a, p. 14. PI, II, Figs. 1-4 ( — S. depressa
Miers) ) .

non Chloridella microphthalma Haswell, 1882. p.
207 ( r. s. depressa (Miers)).

Discussion oj Synonymy

Serene (1952) has vShown that previous re-
cords of S. ynicrophthalma from Australia refer
to 5. depressa Miers. In comparing these two
species he redescribed S. microphthalma on the
basis of two specimens from Indo China, and
illustrated the species extensively. The present
material differs from Serene’s redescription of
S. microphthalma in several particulars, includ-
ing; —

(a) carapace shape, with a broader front
and less inclined margins (compare
Fig. 2a and c);

(b) shape of rostrum, shorter (compare
Fig. 2a and c ) ;

(c) number of denticles between submedian
marginal teeth of telson. with 3-4 pairs
(typically 4) as against 2 pairs: and

(d) shape of bifurcate process of the
uropods.

In the present specimens the process
is relatively stouter, and with a more
clearly sinuous curve in front of the
external lobe of the longer spine (com-
pare Fig. 2d. and f).

Serene’s figured specimens closely resemble
Jurich’s from Zanzibar, w'hile the present speci-
mens closely resemble Kemp’s from Karachi and
Madras and de Man’s from Indonesia.

The differences between these two groups of
specimens lie within the range to be expected
within a stomatopod species, particularly since
Serene and Jurich had I’elatively large speci-
mens (lengths 85 mm and 67 mm respectively).
This conclusion has been reinforced by exami-
nation of a specimen from Zanzibar kindly
forwarded by Dr. A. J. Bruce. The specimen
( 3 , 4.2 mm) bsars the following habitat data —
“outflow of mangrove swamp, thick black mud,
trawled 5 fms, Mto Zingwe Zingwe, Zanzibar

39

Island, 4/vii/1961.” This specimen is inter-
mediate between Serene’s and the Australian
material as follows: — (a) resembles Serene’s in
having a long rostrum and having only two pairs
of denticles between the submedian teeth of
the telson and (b) resembles the Australian
material in the relatively broad and slightly in-
clined margins of the carapace, and in the shape
of the bifurcate process of the uropods (al-
though the external lobe is less strongly de-
veloped).

Material Exaviined

^ (20mm) Roebuck Bay, W. Australia, be-
tween tides on fiats, 8/viii/1929, coll. A. A.
Livingstone, Aust. Mus. Reg. No. P13540.

^ (32 mm), $ (39 mm) Roebuck Bay, W.
Australia, between tides sand flat, 8/viii/1929.
coll. A. A. Livingstone, Aust. Mus. Reg. No.
P13541.

^ (28 mm) W. side Fort Hill, Port Darwin,
N. Territory, between tides, coll. A. A. Living-
stone. Aust. Mus. Reg. No. P13542.

Material Illustrated

Carapace etc. — 32 mm specimen. Telson — 39
mm specimen (PI. I. Fig. P) and 28 mm speci-
men (PI. I. Pig. G). Front end of carapace and
rostrum — 39 mm specimen (Fig. 2a). Bifurcate
process of uropod — 32 mm specimen (Fig. 2d).

Description

Eyes. — Eyestalks short and broad, length ap-
proximately three-quarters that of the basal
segment of the antennular peduncle, maximum
breadth about 50 per cent, of or slightly more
than 50 per cent, of combined length of stalk
and eye. Cornea black, indistinctly bilobed,
with the maximum breadth of the cornea in a
dorsolateral direction being about one third
maximum length. Inner margins of the eye-
stalks in juxtaposition for about half their
length, thereafter diverging quite strongly (in
this last respect, they resemble those of 5.
latreilli (Eydoux and Souleyet) see Serene. 1952.
PI. I, Pig. 1 and 5. decorata (Wood-Mason) see
Kemp, 1913, PI. I. Pig. 13)*

Pig. 2 . — SQuilla microphthalma — a, b, c. — Rostrum and anterior part of carapace: a, 39 mm specimen; b, after
de Man (1898); c, after Serene (1952). d, e, f. — Bifurcate process of uropod; d, 32 mm specimen; e, after Kemp

(1913); f, after Serene (1952).

40

Antennules . — Peduncle approximately two-
thirds the combined lengths of the carapace and
rostrum.

Antennae . — Basal segment together with
proximal segment of exopodite slightly shorter
than the length of the distal segment (scale)
but not reaching to the middle of the eyes.
Peduncle of endopodite c. 1.4 times the length
of the antennal scale (excluding bristles) and
equal to or slightly longer than the anterior
breadth of the carapace.

Mandibular palp. — Present, three segmented
in all but the smallest specimen, here apparently
two segmented.

Rostrum. — iSemi-elliptical (length about 60
per cent, of breadth > to semi-circular.

Carapace . — Maximum breadth slightly less
than the length (excluding the rostrum) and
double the minimum breadth behind the antero-
lateral angles. Anterior borders between the
rostrum and the anterolateral angles distinctly
inclined posteriorly. Anterolateral angles each
with a well-developed spine. Lateral borders
straight in their anterior three-quarters, with
smoothly rounded posterior angles and smoothly
concave posterior border.

Cervical groove distinct, gastric groove dis-
tinct anterior to the cervical groove, but less
distinct posterior to it. No trace of median,
intermediate or lateral carinae, nor of marginal
carinae anterior to the cervical groove. Posterior
to cervical groove, marginal carinae distinct
and best developed in their reflected portions.
A pit present in the mid-line of the carapace
about one third of the way backwards from the
anterior margin. A mid-dorsal tubercle just in
front of the posterior margin of the carapace.

Raptorial claws . — Articulation of ischium and
merus not terminal, but slightly in advance of
the proximal end of the latter. Merus massive,
with concave inferior margin. Upper margin of
the carpus with a single distinct curved carina
running along three-quarters of its length and
terminating abruptly: near the termination
there is a tuft of setae. Propodus deep and
swollen, bulging particularly at its carpal arti-
culation, and with a smoothly curved lower
margin. Upper margin pectinate, with setae
amongst the teeth and there are the usual three
stout movable spines just inside the proximal
part of the upper margin, Dactylus with four
spines, the terminal one being much the longest.
The lower margin of the dactylus is sinuous.

Thoracic somites . — Fifth somite bearing on
each side a lobe (often blunted, but acute in
largest specimen) which continues ventrally as
a short ridge. Sixth, seventh and eight somites
bearing distinct but rounded intermediate
carinae: the lateral margins carinate. No
median or submedian carinae on the thoracic
somites.

Sixth and seventh somites with entire and
rounded borders. Lateral margin of eighth
somite partly covered by a bluntly rounded
anterior projection from the first abdominal
somite. Endopodites of the last three thoracic
appendages elongated ovoids.

Abdominal somites . — First five somites bear-
ing distinct intermediate, lateral and marginal
caiinae but median and submedian carinae not
px’esent. On the sixth somite, there are sub-
median. intermediate, lateral and marginal
carinae, the last two converging posteriorly. The
following carinae end in spines, brackets indi-
cating absence in some of the specimens: —
Carinae Abdominal somites

Submedian 6

Intermediate (5), 6

Lateral —

Marginal (4), 5, 6

Termination of marginal carina of fourth
abdominal somite rarely sharp enough to be
termed a “spine”: that of third somite even less
spiniform.

Telson . — Broader than long (c. 1.6 times) and
strongly convex. Median carina very distinct
and ending in a distinct spine. A row of tubercles
cn each side of the carina, these rows being re-
ferred to below as submedian tubercles. Up to
five tubercles present, being sometimes dis-
tinctly separate and sometimes more or less
fused. The most anterior on each side compi'ises
a broad elevated area and behind these the rows
converge, sometimes terminating in a median
tubercle behind the carina. Between these sub-
median tubercles and the marginal teeth of the
telson. there are 2-4 (typically 3) obliquely
pointing and elevated rounded ridges (or
elongated tubercles). Up to three pairs of ad-
ditional tubercles towards the margins of the
telson, representing basal inflations of the
carinae of submedian and intermediate teeth
respectively and lying near the base of the
lateral teeth. In the smallest specimen only the
submedian inflations are present.

Margins of the telson ornamented as fol-
lows: — a pair of blunt submedian teeth with
movable spines; between these there are 3-4
(typically 4) pairs of denticles: a pair of stout
but sharp intermediate teeth separated from
the submedians by 6-8 denticles on each side;
and a pair of acute sharp lateral teeth
separated from the intermediates by a single
denticle on each side. Anterior two-fifths of
margin of lateral tooth of telson wdth a stouter
carina than the remainder which, in the large
specimens, terminate in an inconspicuous
prelateral denticle. As stated above the carinae
of submedian and intermediate teeth are repre-
sented in the larger specimens by tuberculate
carinae. These sometimes show indications of
being composed of fused rows of smaller
tubercles.

Under surface of the telson microscopically
granular, and without a post-anal carina.

Uropods . — Bifurcated process of basal seg-
ment armed internally with 5-7 (typically 6)
sharp, elongate teeth. Longer process about
thrice the length of the shorter and with a
basal thickening on the margin facing the
shorter process. Thickened portion very con-
spicuous and sinuously curved.

External margin of the basal segment of the
uropod bearing 5-6 ( typically 6 ) articulating

spines, the terminal much the largest. The ter-
minal segment of the exopodite approximately
2i times as long as broad.

41

Colours after prolonged alcohol fixation : —
General colour of larger speciments biscuit, but
raptorial claws ivory coloured. Typically with
numerous black or sooty brown pigmented areas
of which the most conspicuous are: — (a) ros-
trum and anterior carapace (see PI. I, Fig. E);
(b) posterior carapace where a black line near
the border continues in the anterior reflections
of the marginal carinae; (c) on most of the
free thoracic somites and abdominal somites
where transverse lines occur near the posterior
margin of each. These lines are absent on the
fifth thoracic somite and typically on the sixth
abdominal somite and are feebly developed on
the fifth abdominal somite. There is a small
black spot in the posterior medial portion of
the penultimate segment of the exopodite of
the uropod.

In the smallest specimen the only pigmented
portion apart from the cornea consists of small
dark spots near the anterolateral margins of
the carapace, while in the 28 mm specimen, the
first five abdominal somites bear additional
diffuse spots of pigment running across each
segment, and the sixth abdominal somite is pig-
mented as in the previous ones.

Comments

In the four individuals differences in the den-
sities of pigmentation and the ornamentation
of the telson are within the expected range of
variation of a species. However the smallest
specimen differs in further particulars: — (a)
less curvature cn the inner margins of each
eyestalk; (b) in dorsal view eyestalks extending
relatively forw^ard in relation to the basal seg-
ment of the antennular peduncle. Only about
cne-eighth of the length of this segment projects
forward beyond the eye, and (c) mandibular
palp. This is proportionally smaller than in
the remaining specimens, and although ap-
parently undamaged, only two segments could
be distinguished instead of three.

The species was previously known from
Zanzibar, India. Indo-China and Indonesia but
not from Australia.

Acknowledgments

Thanks are due to (a) the Directors of the
following Museums for loan of specimens: —
Western Australian Museum (Dr. W. D. L.
Ride) Australian Museum (Dr. J. W. Evans),
Macleay Museum, University of Sydney (Miss E.
Hahn); (b) the Curators of the crustacean col-
lections at the first two Museums <Dr. R. George
and Mr. F. A. McNeill) who have been especially
helpful: (c) the following Directors or Curators
in New Zealand Museums for forwarding speci-
mens of hysiosguilla — Professors J. Morton, G.
Knox, L. R. Richardson; Drs. R. Falla, R. R.
Forster and Mr. R. W. Macdonald: and (d) to
Dr. A. A. Racek for specimens and information
on Squilla mcneilli and to Mr. L. Wale for a
specimen of Lysiosquilla latifrons brazieri: (e)
Miss May Rees for assistance throughout and
especially in preparing line drawings; and (f)
to the Research Committee of the University of
Queensland for financial support.

List of Species

fjfmtxiaclybis chiragra (Fabriciu.s) ....

Page

:u

0. Jalcatm (Forskal) , . ...

U

0. graphurus Miers

:h)

(I. pulehellus Miers

:h')

(i. •'itoUurm {Miillcr) ...

;ir)

(i, trispinosu^ Dana ..

Lysiosquilla mullijasciata W’ood-Mason

u

A. latifrons hrazieri Miers

88

A. oscalans Hale

84

Odoniodartylus ndtrifer (White)

85

O. jtipoiiirits (de Haan)

:h)

O. scyllaru.s (L.)

:^5

l*seudosq(iilla ciliuta (Fabricius)

84

P. ornatn Miers

84

Squilla fabricii Holthuis

:j8

*S'. gtmypHfis Kemp

.85

.S’, granfi Stephenson

:h8

iS'. itwrnala Tate

88

»S. laevis Hess ....

.83

S. mcneiUi Stephenson

:i4

*S. microphthnhnn H. M. Kdwards ....

;h)

.S’, miles He.ss ...

u

S. muUicarinnta White

.85

6', raphidea Fahrieius ....

u

S. terrareginemis Stephenson

u

References

Alexander. W. B. (1916a). — On a stomatopod new to
Australia, with a list of the Western Aus-
tralian species of the order. J. Roy. Soc. W.
Aust. 1: 8.

Alexander, W. B. (1916b). — rurther notes on Western
Australian stomatopods. J. Roy. Soc. W.
Aust. 1: 9-10.

Balss, H. (1921). — Results of Dr. E. MJbberg’s Swedish
Scientific Expeditions to Australia 1910-13.
XXIX. Stomatopoda. Macnira, Paguridea
und Galatheldea. K. Si^enska Veteyiskakad.
HandL 61; 1-24.

Borradalle, L. A. (1898). — On some crustaceans of the
South Pacific -Part 1. Stomatopoda. Proc.
Zool. Soc. Lond. 1898: 32-8.

Bigelow, R. P. (1895). — Scientific results of explorations
by the US. Fish Commission Steamer
“Albatross” No. XXXII. Report on the
Crustacea of the Order Stomatopoda col-
lected by the Steamer “Albatross” between
1885 and 1891 and other specimens In the
U.S. National Museum. Free. U.S. Nat. Mus.
17: 489-550.

Bigelow, R. p. (1931). — Stomatopoda of the southern and
eastern Pacific Ocean and the Hawaiian
Islands. B^tU. Mus. Comp. Zool. Harv. 72:
105-91.

Chilton. C. (1911). — Revision of the New Zealand
Stomatopoda. Trans. N.Z. Inst. 4.3: 134-9.
Chopra. B. (1939). — Stomatopoda. Sci. Rep. Murray
Exped. 6: 137-81.

Dollfus. R. P. (1938J. — Mission Robert Ph. Dollfus en
Egypte. Stomatopoda 11. M^vi. Inst. &gypt
37: 185-236.

Eydoux, A. M., and Souleyet, L. (1841), — "Voyage de la
‘Bonlte*." Zooiogie, Vol. 1. Crustacea, pp.
262-6. pi. V (Paris.)

Gravier. C. (1938).— IX. Stomatopoda (1). Les Stoma-
topodes du Golfe de Suez. M^m. Inst, tgypt.
37: 163-83.

Haan, W. de (1844).— Crustacea in Slebold’s “Fauna
Japoulca". Atlas. PI. LI (Leyden.)

42

Haan, W. de (1849). — Crustacea in Slebold’s “Fauna
Japonlca". Text pp. 220-225 (Leyden.)
Hale, H. M. (1924). — Notes on the Australian Crustacea.

No. 1. Jlec. S. Aust. Mus. 2: 491-502.

Hale, H. M. (1929a). — Crustacea from Princess Charlotte
Bay, North Queensland. The Isopoda and
Stomatopoda. Trans. iSoy. Soc. S. Aust. 53:
33-6.

Hale, H. M. (1929b). — Notes on the fauna of Dirk Hartog
Island. Western Australia. No. 4. Crustacea.
Trans. Roy. Soc. S. Aust. 53: 67-70.

Hansen. H. J. (1926).— The Stomatopoda of the Slboga
Expedition. Slboga Exped. Livr. 104, Monogr.
XXXV

Haswell, W. H. (1882). — "Catalogue of the Australian
stalk and sesslle-eyed Crustacea." (Austra-
lian Museum; Sydney.)

Hess, W. (1865). — Beitrtlge ziir Kenntnlss der Decapoden-
Krebse Ost-Australlens. Arch. Naturaesc?i.
31: 127-73.

Holthuls, C. B. (1941). — Biological results of the Snelllus
Expedition XII. The Stomatopoda of the
Snelllus Expedition. Te?n?ninc/aa 6: 241-294.
Jurich. B. (1964). — Die Stomatopodcn de deutschen
Tlefsee-Expedltlon. WIss. Ergebn. ‘’Valdivia”.
7: 361-408.

Kemp, S. (1911). — Preliminary descriptions of new
species and varieties of Crustacea Stoma-
topoda in the Indian Museum. Rec. Indian
Mus. 6: 93-100.

Kemp, S. (1913). — An account of the Crustacea
Stomatopoda of the Indo-Paclflc region
based on the collection in the Indian
Museum. Mem. Indian Mus. 4: 1-217.

Kemp. S., and Chopra, B. (1921). — Notes on Stomatopoda.

Rec. hidian AfiiS. 22: 297-311.

Komai. T. (1927). — Stomatopoda of Japan and adjacent
localities. Mem. Coll. Sei. Kyoto (B) 3: 307-
54.

Lloyd. R. E. (1908). — Remarkable cases of variation.
Rec. Indian Mus. 2: 30-5.

Miers, E. J. (1880a).— On the Squlllldae. Ann. Mag.
Nat. Hist. (5) 5: 1-30.

Miers, E. J. (1880b;. —On the Squillidae. Ann. Mag.
Nat. Hist. (5) 5: 108-27.

Miers, E. J. (1880c). — On a collection of Crustacea from
the Malaysian region. Part IV. Penaeidae,
Stomatopoda, Isopoda. Suctoria and
Xlphosura. Ann. Mag. Nat. Hist. 5: 457-72.

Mllne-Edwards, H. (1837).— “Hlstoire naturelle des
Crustaces. Vol. II. (Paris.)

Nobill. G. (1903). — Crostacei dt Singapore. Boll. Mus.
Torino 18 (455): 1-39.

Parisi, B. (1922). — Blenco degll Stomatopodl del Museo
de Milano. Atti Soc. Jtal. Milano. 61: 91-114.

Pocock, R. I. (1893). — Report upon the stomatopod
crustaceans obtained by R. W. Basset-
Smith Esq., Surgeon R.N., during the cruise
in the Australian and China Seas of H.M.S.
“Penguin". Commander W. U. Moore. Ann.
Mag. Nat. Hist. (6) 9: 473-9.

Rathbun, Mary J. (1902) .—Japanese stalk-eyed Crus-
tacea. Proc. U.S. Nat. Mus. 26: 54-5.

Serene. R. (1952). — Btude d'une collection de Stoma-
topodes de TAustralian Museum de Sydney.
Rec. Aust. Mus. 23: 1-24.

Stephenson, W. (1953),— Three new Stomatopoda (Crus-
tacea) from eastern Australia. Aust. J. Mar.
Frestiw. Re.s. 4: 201-18.

Stephenson, W. (1960). — Notes on Queensland Stomato-
poda (Mantis prawns). Qd. Nat. 16: 61.
nine species of Squillidae from the collec-

Stephenson. W., and McNeill. F. (1955).— The Australian
Stomatopoda (Crustacea) In the collection
of the Australian Museum, with a check
list and key to the known Australian species.
Rcc. Aust. Mus. 33; 239-65.

Sunler, A. L. J. (1918). — The Stomatopoda of the collec-
tion of the "visschery-statlon" at Batavia.
Contr. Faune Jnd. N4.erl. 1 : 62-75.

White, A. (1847). — "List of the Specimens of Crustacea
in the Collection of the British Museum."
pp. 83-85 (London.)

White. A. (1848). — Description of two new species of
Crustacea. Proc. Zool. Soc. Lond. 1848: 144.

White. A. (1849).— Description of two new species of
Crustacea. Ann. Mag. Nat. Hist. (2) 4: 381-2.

Wood-Mason, J. (1895). — Figures and descriptions of
nine species of Squillidae from the collec-
tion In the Indian Museum. (Calcutta).

43

6. — Larks, Mirafra javanica, of tropical Western Australia

By G. P. Mees*

Manuscript received — 20th February. 1962

On the basis of freshly collected material
the distribution and geographical variation of
Mira/ra javanica In Western Australia is dis-
cussed In greater detail than was possible by
Mayr & McEvey (1960), to whose paper the
present one may be regarded as complementary.

Introduction

In their important paper on distribution and
geographical variation of Mirafra javanica,
Mayr & McEvey (1960i suggested further field
studies. They also discussed the application of
ternary nomenclature in the species and some
other matters.

Prom May till August. 1960. my assistant. Mr.
A. M. Douglas, and I visited the northern part
of Western Australia. We paid special atten-
tion to Mirafra, and collected 39 specimens,
many of which are from localities whence the
species had not previously been recorded. This
resulted in a much improved knowledge of the
distribution of the various colour-types, includ-
ing the discovery of a very rufous population
in the Kimberley Division. I have thus been
able to extend the information given by Mayr
& McEvey and to discuss some of the questions
raised in their stimulating paper.

The first draft of this paper was sent to Dr.
Mayr and Mr. McEvey. to both of whom I am
indebted for useful comment and criticism, much
of which has been incorporated. I have to
thank Dr. D. Amadon (American Museum of
Natural History) for the loan of four topo-
typical specimens of Mirafra javaiiica vielvil-
lensis. Mr. R. Vollprecht. of the Perth Weather
Bureau provided me with the 1960 rainfall
figures for Wyndham. Pinally Mr. Douglas and
I want to express our sincere gratitude to the
many friends whose hospitality contributed so
much to the success of our stay in the Kimber-
ley Division.

Distribution, Habitat and Habitat Preference

Mirafra javanica is widely distributed in the
North-West Division from Minilya northwards,
and in the Kimberley Division (Fig. 1), and
there is no doubt that its distribution is con-
tinuous or nearly so. The occurrence seems to
be governed by the presence of seeding soft
grasses, the seeds of which are (at least in the
months May till August over which my ex-
perience extends* its main and perhaps its onlv
food. In stomachs I found exclusively seeds of
at least two different species of grasses, though
unfortunately these were not preserved foi
identification.

In the Kimberley Division, larks occur nox
only in the open fields and paddocks with rather
low and small soft grasses, but also in the very

♦ Western Australian Museum. Perth. Western Australia.

high Heteropogon contortus in open savannah;
usually, how'ever. as solitary specimens, never
in great numbers, and never in the denser types
of savannah-woodland.

The preceding notes on habitat do not con-
flict with my earlier statement that the distri-
bution of the species is continuous or nearly so,
for open savannah and open fields are so com-
mon in the Kimberley Division that for practical
purposes they may be regarded as a continuous
habitat. This is supported by the fact that at
every place visited by us in the Kimberley
Division we found larks. It may be noted that
Mayr & McEvey (p. 163, 164) do not seem to
be quite clear about the position. They loosely
use the word “isolates” for several of the
populations, but contradict this by saying

“ that suhrufescens is essentially

an inland race tending to intergrade with the
coastal isolates around it in the west”. Mr.
McEvey wrote me that the term “isolates” was
partly used as a substitute for “subspecies”,
mainly to avoid continuous repetition of the
words subspecies and race. Personally I would
prefer not to water down the term isolate as
this is likely to cause confusion.

Wherever I observed larks. I have made notes
on the colour of the soil, and feel justified in
stating — contrary to what is said of certain
species of larks in Arabia and the Kalahari
Desert — that they show no particular preference
for soils corresponding in colour with the colour
of their own plumage. When going to alight,
the birds are clearly guided exclusively by the
presence of seed-bearing grasses, quite irrespec-
tive of the colour of the soil.

This does not mean, however, that any idea
of correlation between colour of soil and colour
of larks should be abandoned. In much of the
Kimberley Division the rocks are of a reddish-
brown colour, consisting of reddish sandstone
and red weathered basalt. All true plains, on
the contrary, have a greyish-yellow colour. At
Beverley Springs and Kalumburu, though some
of the paddocks where one finds the greatest
concentration of larks are greyish-yellow, the
country is hilly so that the greater part of it
is reddish-brown in colour — and so are the larks
in these places. Going from Beverley Springs
over Mount House to Glenroy. the larks become
slightly paler in colour. But Glenroy is an
enormous plain. At Wyndham the hills consist
of red sandstone and their erosion products are
just as red as the reddish-brown sand found
in localities where i*ust-coloured larks occur.
However, the most conspicuous feature of the
country near Wyndham are the extensive grey
flats, the estuarine plains of several rivers, and
it is to that colour that the larks are adapted.
As a generalisation, therefore, I would say that

44

Fig. 1.— The distribution of Mirafra javanica and Its colour types in Western Australia. Localities: 1, Mlnllya
2, Maud’s Landing: 3, Point Cloates; 4, Minderoo; 5, Onslow; 6, Cane River; 7. Coolawanyah; 8, Mlllstream’
9. 14 miles east of Roebourne: 10. Sherlock; 11, Mundabullangana; 12, Port Hedland: 13, Tabba Tabba- 14*
Marble Bar; 15. De Grey; 16. Pardoo Sands. 116 mile peg; 17. Wallal. 30 miles beyond homestead; 18. Anna
Plains Outcamp; 19. Roebuck Bay Plains. Broome: 20 Broome-Derby Road, 85 and 88 mile pegs; 21, Point
Torment; 22, Meda; 23, Llverlnga; 24, Noonkantaah; 25. Beverley Springs: 26. Mount House; 27 Glenroy 28
Kalumburu; 29. Forrest River; 30, Wyndham; 31. Parry’s Creek; 32. Argyle (Ord River)

in hilly or mountainous areas rusty coloured
larks may be expected, in lowlands yellowish-
grey larks. The very rufous M, j. woodwardi
of the hilly North-West, and the pale M. j. halli
of the plains near Broome, further support this
assumption.

In view of the extreme paleness of the race
halli, special attention was paid to the colour
of the soil in its range. The coastal area along
the Eighty Mile Beach where we found and
collected larks consists of pale yellow dune sand.
Near Broome the coastal strip is also pale, but
it does not extend far inland and patches of
brownish and reddish soil occur very near the
town as I could ascertain when flying over the
area. This suggests that along the Eighty Mile
Beach the species does not range far inland,
but is confined to the pale coastal strip. The
coastal sands south-west of Port Hedland
(Mundabullangana) are just as pale as those
near Broome, but the larks there have behind
them the great reservoir of rufous larks from
the hilly Pilbara District so that any local selec-
tive influence would be undone.

Dr. Mayr has pointed out to me that the
situation in Australia is probably not directly
comparable with that in Africa, as there the
contrast between the substrates is exceedingly
great. Rich red rocks will meet along a geo-
logical fault line almost snow white limestones.
Elsewhere, black lava flows will meet whitish
sands. The differences in soil colour found in
the north of Western Australia are certainly far
less striking. Nevertheless it is worth pointing
out that in Australia no preference for soils
which correspond in colour with the plumage
of the birds seems to have developed, and in
fact, this is one of the questions posed by Mayr
& McEvey.

Mayr & McEvey (p. 190) enquire about the
possible significance of habitat colour (general
colour of a paddock of seeding grasses) as op-
posed to ground colour. As far as I have been
able to make out, this colour, which in the seed-
ing season is pale yellow, has had no influence
as a factor contributing to the colour of the
birds. This is not surprising when one con-
siders the fact that Mirafra javanica is essen-

45

tially a ground bird, so that cryptic coloration,
to be of any advantage, has to agree with the
colour of the ground, rather than with that of
its vegetation. Moreover, habitat colour is pre-
sumably identical in all parts of the range
(though seasonal variation from green to yellow
probably occurs), and if it has any effect, this
will be difficult to detect because it would work
in exactly the same way throughout the range.
Also there seems to be a preference for small
open places or places with scanty vegetation —
mere about this in the section on behaviour.

Miscellaneous Notes on Behaviour, etc.

Towards the end of July, 1958. I found the
larks at Millstream and Coolawanyah stations
in full song. Early that month there had been
heavy rainfall in the district and during the
last days of the month more rain fell; I men-
tion this because of the possibility that sexual
activity is stimulated by rainfall, as has been
shown in other Australian bird-species (cf.
Keast 1959). Some notes on the behaviour of
the birds during rain were published previously
(Mees 1961).

During our trip, we did not hear any song
until on my last day of field work in Wyndham,
July 31, 1960, when one or two birds were in
full song. This evidently was the first song of
the season because for a week I had been visit-
ing the same localities almost daily, and had
not heard any song. Rainfall in Wyndham
during 1960 was: Januai'y 897 pts., Februai’y
814, March 928, April 223, May 347, June till
October nil, November 250, December 460.

Apart from the singing birds in Wyndham,
the only noise I heard from larks in the period
May-July 1960 was a short double note uttered
in flight: “pitpit .... pitpit ....

When preparing specimens collected I have
taken notes cn size of gonads, state of ossifica-
tion of skulls, and condition of moult, the re-
sults of which are tabulated. In all specimens
of which the sex could be determined the gonads
are small to very small; this is particulaiiy
striking in the males, for in adult females the
normal size of a resting ovary is apparently
about 5 mm. which is fairly large; but none of
the females had enlarged oocytes. In the speci-
mens of uncertain sex, the gonads were pre-

TABLE I

Locality

Date

Sex

Gonads

Ossification skull

Moult

14 miles N. of

22.v.i!mo

small, 2 mm.

fully ossifiefi

body feathers

Roeboiirno

♦»

,,

9

fully os.sified

body feathers and tail

Mundabullangana

23.\M9(5(I

9 juv.?

9

partly ossified

no moult

Pardoo vSands ....

25. V. 1960

9

very small, mm.

}> f* —

,,

? imm.

partly ossified

Broome > Derby

l.VI.1960

fully ossified

moult

Road, 85-milo
peg

Broome - Derby

i.vi.umo

c?

tiny, less than 1 mm.

largely unossified

moult

Road, 88-milo
peg

Kalumburu

14. VI. 1960

? iram.

partly oSvSificfl

heavy moult

,, —

2. VI 1.1960

9 imm.

tiny, 3 mm.

>» —

1.VJI.1960

9 ad.

fairly large, 6 mm.

fully ossified

primaries, no moult in-

. , —

24.VI.1960

c?

small. 1^ mm.

fully ossified

body feathers
heavy moult primaries

,, —

19.V1,!960

9 ad.

fairly large, 6 mm.

fully ossified

wings

,, —

27. VI. 1960

small, I mm.

fully ossified

primaries and rectrices

,, —

12.VI.1960

fairly large, 1. testis

fully ossified

99 * • • •

14.VI.1960

? imm.

3J mm., r. testis
2J mm.

j)artlv ossified

heavy moult

99

24.V1.I960

small, 1 mm.

fully ossified

wings

Beverley Springs

19.VI1.1960

9

fairly small, 5 mm.

not completely ossified

99

3

tiny, less than 1 ram.

fully ossified

>> >>

11.VII.1960

9

fairlv small, 3J mm.

fully ossified

»> »>

19.V1I.1960

small, mm.

fully ossified

99 99

13.ViI.1960

9

fairlv small, 4J ram.

fully ossified

Wyndham

25.Vn.1960

9

fairly small, 4 j mm.

fully ossified

y > ....

31.V1I.1960

9

fairly small, 5 mm.

fully ossified

body feathers, heavy

,,

25.V1I.1960

small, 1 mm.

fully ossified

tail moult

»>

26. VII. 1960

<?

fairly small, 2 and

fully ossified

,,

28.Vir.1960

9

2i mm.

fully ossified

,,

25.VII.1960

9 imm.

.small. 4 mm.

partly ossified

»>

28.Vn.1960

9

fairly small, 5 mm

fully ossified

>>

26.Vn.1960

9

fairly small, 5 mm.

fully ossified

46

sumably so small and inconspicuous that they
could not be found. Unfortunately I failed to
obtain a singing bird at Wyndham; it would
have been interesting to compare the size of
its gonads with those of non-singing birds. My
observations fit in with McEvey’s (I960) state-
ment that the only months in which breeding
of Mirafra javanica in Australia has not been
recorded are May. June, July and August.

Many of the specimens show moult. Exami-
nation of the condition of ossification (or
pneuniatisation » of the skull enabled me to
check whether the plumage characters hesitat-
ingly claimed by Mayr & McEvey fp. 156-157)
to be diagnostic of young birds, do hold. I
found that the character of the pale edges to
the crown feathers, which gives the crown a
more or less scaly appearance, is apparently
a valid one. As regards the outer edges of the
primaries, allegedly wider in young birds, there
is certainly a tendency to this, but the differ-
ence is not very convincing.

The way to observe larks is to walk criss-
cross over likely-lcoking country (e.g.. places
covered with soft grasses) until one is flushed.
The flight is slow and with chai’acteristic fairly
short wing-beats. Usually they alight nearby.
The birds are to be found on small open or less
densely covered places, rather than in the
densest grassland. This probably is also the
reason that quite often we found them near
cattle— where the vegetation had been trampled
down.

Partly this preference for certain places may
be responsible for the concentration of birds one
often sees in a very limited area, but on the
other hand larks are definitely social. Not only
did I observe repeatedly how flying larks joined
each other, or would even be joined by those
that had been standing on the ground, but also
when a number were flying about at a time,
and one alighted, others would alter their course
and alight near the first individual.

There is no doubt that when not in song
the birds are rather incoirspicuoiis. When sing-
ing they inevitably make their presence known
widely. When not singing, however, they stick
to the ground, and will not fly up until dis-
turbed. Hence, the casual observer may easily
overlook them even in paddocks where they are
common.

Several authors, of which Bourke <1947) may
be mentioned in particular, have drawn atten-
tion to the amount of mimicry in the song
of Mirafra javanica: Bourke even states:

“ that mimicry forms the bird’s ‘normal’

song during the breeding season, and that from
February until September (approximately) the
song of the species consists of a short double
note— merely a call-note”. This call-note is
evidently identical with the one described by
me on a previous page — in my opinion it is in-
correct to call it song. As regards the song.
Bourke is certainly right that much of it con-
sists of imitation, but the statement that
mimicry forms the song definitely goes too far.

I have no notes on mimicry in Australia, but
did hear it from the nominate race, Mirafra
javanica javanica Horsfield. in Java. Fox-
example, on 29.VI.1947 near Buitenzorg. West

Java, I noted that there was apparently much
mimicry in the song, and that the ordinary call
of Pyc7io?iotus aurigaster (Vieillot) could easily
be recognised in it. Another time I recognised
the call of Caprimulgus affinis Horsfield and
the song of Prinia inornata blythi (Bonaparte),
both imitated almost perfectly (Tjibaroesa, West
Java. 2.VII.1949).

Geographical Variation

The colour vai’ies from rufous to white on
under-parts, and fiom black to grey, with the
edges of the feathei's varying from rufous to
white, on the dorsal surface. In the Northern
Territory there is a population isoderbergi)
with a very black back, and little rufous, but in
Western Australia the black-grey seems to be
moi*e or less correlated with the rufous-white
series, and does therefore net need a separate
discussion. I distinguish the following colour
types: —

( i ) No rufous at all except some pale brown
along the outer edges of the px'imaries;
under surface creamy white; back
grey. The white extreme.

(ii) As (i), but a very slight admixture
of brownish on the back; under sur-
face cream.

(iii) More bi*ownish than (ii) both above
and below.

(iv) All the feathex’s of the dorsal surface
with dark rufous borders, under sur-
face cinnamon. The rufous extreme.

In Fig. 1 I have tried to indicate the approxi-
mate ranges of these four colour types in
Western Australia. The figure also shows very
clearly the difficulties of practical classification
one encounters when trying to deal with these
poixulations trinomially. Going from Roebourne
along the Pardoo Sands to Anna Plains, one
covers the whole zone of intergradation be-
tween the I'ufous wo'odioardi and the pale halli:
from Broome going north-east via Derby to
Beveiiey Sprixxgs one encounters the same
gi’adual chaxige in opposite direction, fi*om halli
to melvillensis.

The three races mentioned, looodwardi, halli
and inelvillensis are extremes, hence there is
no problem in applying trinomials. We come
now to colour type Ui) however. This is at
Pardoo Sands and between Broome and Derby
merely an intermediate — or a product of inter-
gradation — between (i) and (iii) which, fox-
practical purposes, should not be named*. Near
Wyxidham, on the other hand, an indistinguish-
able population occurs which is not an inter-
mediate, but forms the end of a gradient of a
decreasing amount of I’ufous in the plumage.

•Admittedly the Pardoo Sands at the 116 mile peg.
where we collected larks of an intermediate colora-
tion , are a pinkish mixture of the red sand of the
interior wltli the whitish sand of the dunes; hence
one might argue that the larks at this locality are
intermediates because tltey live on a soil of inter-
mediate coloration, and not because of their Inter-
mediate geographical possition between woodwardi
and halli.

At the 85 and 88 mile pegs on the Broome-Derby
Road, where Intermediate larks were collected, the
soil Is pale yellowish with a pink tinge due to a
very slight admixture of red sand.

47

To colour type (iii) the same pertains as
to colour type Ui) ; it is another intermediate:
between Floebourne and Pardoo Sands and near
Derby a product of intergradation between
(iv) iwoodwardi) and (ii), but at Argyle
perhaps an extreme in the rufous direction as
no more rufous population is yet known to
exist near that place (it is very well possible that
the range of the melvillensis-hke population ex-
tends in an easterly direction to south of Wynd-
ham. but no material is yet available).

It must be realised that my division of the
larks in four colour types (as far as amount of
rufous in the plumage is concerned) is purely
arbitrary. With equal justice I might have
divided them in six or eight types. Also as far
as my experience goes, the gradients are per-
fect: the boundaries indicated between the
various colour types again are arbitrary which
mepis that specimens here ascribed to (ii)
which have been collected near the range of
(i> are pale, whereas specimens nearer to (iii)
are more rufous.

Pig. 2.— Grass plains at the 85 mile peg on the Broome-
Derby Road, habitat of Mira/ra javanica subsp. (colour
group Ilj. 1. VI. 1960.

Fig. 3.— Grass plains at Beverley Springs, habitat of
Mirafra javanica melvillensis. 19. VII. 1960.

Colour type (iii) corresponds to what
Mayr & McEvey call sitbrufesceiis, a race they
give a continuous range from the De Grey River

area in the west to the Gulf of Carpentaria in
the east. At De Grey it appears as an inter-
mediate between xooodwardi and halli, near
Derby as an intermediate between halli and
melvillensis, west of Forrest River it remains
to be discovered as an intermediate between
"fOTvestV^ and melvillexisis, south of Wyndham
(Argyle Downs) it may occur as a soil-adapted
population of its own accord, not as an inter-
mediate. It looks as if Mayr & McEvey’s
subrufescens is not the widely-ranging sub-
species these two authors thought it to be. but
consists of a series of geographically uncon-
nected populations, each of which has a dif-
ferent history.

Some Principles of Nomenclature

Mayr & McEvey, discussing the geographical
variation of Mirafra javanica, correctly stress
the fact: "that it is quite impossible to express
this complex variability adequately in terms of
orthodox subspecies". It is perhaps right to
state that I am a strong believer in the ternary
system of nomenclature, and the fact that it
is sometimes inadequate does not in my opinion
mean that the whole ternary system must be
rejected. I would even go farther and say that
Linnean nomenclature, with its present-day
implications of relationships, is only useful when
one clearly realises that classification, which
inevitably is limited to a small number of
systematic categories, is of necessity arbitrary,
even though we may call it a natural system.
Our whole knowledge of speciation and evolu-
tion confirms that there are all stages of inter-
mediates between our artificial classes, and that,
indeed, no two pairs of species or other system-
atic categories stand in exactly the same rela-
tion to each other. Once this is clearly
recognised. Linnean nomenclature can be used
with much advantage.

Mayr & McEvey have named populations, even
though very similar populations might occur
in other parts of the range of the species, a
method they defend as follows (p. 188): “Men-
tion has been made of the inadequacy of ortho-
dox nomenclature and it is merely added

here that their treatment will partly depend
upon whether one looks at them from the view-
point of the taxonomist or the student of evolu-
tion. To the taxonomist identical populations
(i.e.. poly topic subspecies) must be given the
same name. To the student such a terminology
would seem to obscure the fact of the indepen-
dent evolution obviously involved". However,
they add: "In the present cases the majority of
individuals in a given race are distinguishable
from the majority in the parallel form". As our
whole system of nomenclature is based on
mcrphology, it would, indeed, be undesirable to
abandon the generally accepted provision that
a form, to deserve a place in nomenclature, must
be morphologically recognisable.

It seems to me that nomenclature tends to
take too important a place in many systematic
publications. Systematic ornithology should be
concerned with the description of actual varia-
tion as found in nature, rather than with the

PLATE I (opposite)

Mirafra javanica halli, after a specimen from Anna Plains Outcamp. Lower figure, Mirafra

Figures of approximately natural size.

Upper figure

javanica melvillensis, after a specimen from Kalumburu

48

subsidiary problem of how to apply trinomials,
which is not a biological problem at all, but
largely a matter of tidyness of the human mind.

The geographical variation in most species of
birds, especially where island populations and
other discontinuities in distribution are con-
cerned. can be conveniently and profitably ex-
pressed in trinomials. In Mirafra javanica we
have a species that shows a strong geographical
variation (in the concluding section more evi-
dence for this variation being truly geographical
and genetic will be brought foi’ward), but the
peculiar nature of these variations causes the
application of trinomials to be of little use, if
not actually misleading. Nevertheless, to be
consistent with the treatment of other bird
species rather than for any other i*eason it seems
undesirable to reject all use of trinomials.

The tentative solution offered in the following
chapter is, as I realise full well, a half hearted
attempt at a compromise that may not satisfy
anybody but that should be acceptable until an
entirely new approach to the problem may be
made.

Nomenclature to be applied in Western
Australia

It is possible, and desirable, to name extremes,
in other words populations that cannot be con-
sidered to be intermediates or products of inter-
gradations. On the basis of this, the following
trinomials can be applied in Western Australia.

1. Mirafra javanica rooodwardi Milligan,
type locality Onslow. Synonym : Mirafra
javanica subrufescens Mathews. Very rufous,
and reduced spotting on the throat. Inhabits
the North-West Division of the State, and inter-
grades north-east of Port Hedland with the
following species,

Mayr & McEvey (p. 161' specify the type
locality of subrufescens as being “Tabba Tabba,
N.W. Aust. (south of Fitzroy River) which is
doubtless true, but rather misleading, for Tabba
Tabba is actually about 15 miles south of
Strelley and about 30 miles south-east of Port
Hedland. so that the type locality of subru-
fescens is not. as Mayr & McEvey suggest, in
the Kimberley Division, but in Pilbara, and four
hundred miles away from the nearest point of
the Fitzroy River, Birds from this locality are
only slightly paler than true woodwardi and
can be included in that race without distorting
the facts. Correspondence has revealed that
both Mayr and McEvey were under the misap-
prehension that Tabba Tabba was in the Kim-
berley Division, and that in applying the name
subrufescpis they have been partly misguided
by their incorrect idea of the geographical posi-
tion of its type locality. Certainly the birds of
niy group (iii> in the Kimberley Division,
where they range widely, have a better claim to
nomenclatorial recognition than the inter-
mediates of the Port Hedland area. If a
separate subspecific name is to be given to the
birds of this group, a new one would have to
be provided.

Specimens from Millstream and Coolawanyah
about which Mayr & McEvey received notes
fi'om me, were referred by these authors to
subrufescens. On the other hand I found that
these specimens agree very well with the type

of woodwardi which is in the collection of the
Western Australian Museum (Mees 1961).
Actually the type of woodivardi comes from very
near to what according to Mayr & McEvey is
an area of intergradation between woodwardi
and s7ibrufescens. If a type specimen is inter-
mediate between two races, the name will have
to be restricted to one of the components. Yet,
I cannot agree that specimens identical with
the type of woodwardi should be called sub-
rufescens (it is not a single individual I am dis-
cussing but a series of five, which is presumably
more or less representative of the populations
of Millstream and Coolawanyah). Actually this
supports my opinion that subrufescens must be
synonymised with woodivardi.

2. Mirafra javanica halli Bianchi, type
locality Roebuck Bay Plains near Broome (Plate
I. upper figure). The palest of all races, no
brown in plumage except on edges of primaries
and secondaries. Coastal plains from Broome
to Wallal. Intergrades at Pardoo Sands with
the preceding subspecies and near Derby with
the following.

3. Mirafra javanica melvillensis Mathews,
type locality Melville Island (Plate I. lower
figure). As dark rufous as woodwardi, but on
an average more blackish above, throat spotting
heavier. Distribution, Melville Island and the
north-western part of the Kimberley Division,
It is not suggested that there is any historical
or zoogeographical significance in the similarity
of the Melville Island and north-west Kimberley
populations, but I have been unable to find any
morphological character to distinguish them
from each other, except that the Kimberley birds
average very slightly paler below.

Also valid is the name soderbergi for speci-
mens from the northern part of the Northern
Territory, which have the unique combination
of very little brown with a strongly blackish
tinged upper surface. Contrary to Mayr &
McEvey I prefer not to apply a trinomial to the
population from near Wyndham (Mirafra
javanica forresti Mayr & McEvey) because
identical specimens occur as intergrades else-
where.

Incidentally. Mayr & McEvey (p. 167) de-
scribed ‘'forresW from Forrest River and Parry’s
Creek, but included a specimen from Wyndham
in '' subrufescens”. This is most unlikely be-
cause Wyndham is between Forrest River and
Parry's Creek, and the latter locality is only 15
miles south of Wyndham. and on the road to
Argyle Downs where ” subrufescens” occurs.
The specimens collected by me at Wyndham all
are much paler than material from Argyle
Downs, and evidently are typical “forresti”. The
most likely explanation seems to me that Mayr
& McEvey’s specimen labelled Wyndham was
actually collected somewhere inland from Wynd-
ham in a time that the need of exact labelling
was perhaps less strongly felt than nowadays.*

• I have now received this specimen, a flat skin, on
loan from Mr. McEvey. It was collected on July
30. 1959. In a grass and rock habitat on Wyndham
Airfield by Dr. G. Brown, hence there Is no doubt
about its locality of provenance. Curiously this
bird is, as Mayr & McEvey correctly state, far
browner than •'forresti’' and than any of the birds
I saw near Wyndham. It may be a straggler from
a different area, and it certainly shows that much
fieldwork and collecting remains to be done

49

Mayr & McEvey (p. 167) state that their race
"'forrestV' is confined, so far as is known, to the
Forrest River/Parry’s Creek area, but add a few
lines lower down that specimens from Point
Torment, Derby, Meda, and Lennard River are
referable to the same subspecies. Even so. as
their map shows, Mayr & McEvey were able
to give their new race a continuous range
throughout the northern Kimberley Division.
The discovery of vielvillensis in this area has
broken up the continuous range ascribed to
''forrestV\ which is one of my reasons for re-
jecting the name.

A special nomenclatural difficulty lies in the
fact that the oldest name given to an Australian
bird, horsfieldii, apparently applies to an inter-
mediate population. Fortunately this is outside
the area covered by this paper. It is likely that
no final classification can be arrived at without
a study of the genus Mirafra as a whole.

Historical Considerations and Evidence Against
Migration

Throughout this paper it has been taken for
granted that the variation in colour is genotypic
in character; this notwithstanding the fact that
Mayr & McEvey found one population (the
“Ather.on substrate type”) where the coloration
is apparently largely caused by stain. I have
found no evidence of this in Western Australia.
The po.ssibility that colour might be an ex-
tremely variable character, and that perhaps
a bird might be able to assume a plumage of
the colour of the soil it happens to be living
on at the time of moult can also be ruled out
for various reasons. One is that specimens kept
in captivity in Perth for some years did not
change colour. The most important reason is
that wherever I observed and collected larks
in any one locality <and at some places, like
Wyndham and Beverley Springs I observed
large numbers) all birds were remarkably uni-
form in colour. The individual variation in any
one locality is only very slight — which certainly
indicates that the colour is genetically fixed.
Contrary to Mayr & McEvey <p. 159) I failed
to find indications of increased variability where
strongly different forms approach each other
geographically.

Another point is that in Western Australia
there are nowhere abrupt changes from one
colour type into another. Everywhere, as far
as is known, smooth intergradation occurs. And
though for example along the Eighty Mile
Beach it is very striking how, over the com-
paratively small distance of two hundred miles,
the colour changes from the deep-rufous wood-
wardi to the pale halli, even here the change is
smooth and gradual.

A complicated pattern of distribution as the
present one, of populations that are uniform
in any one locality and are doubtless truly

genetical, can only be explained by assuming
that the birds are everywhere in their range
extremely sedentary. The geographic variations
being as they are. in many localities a bird
only a hundred miles away from its place of
birth would be conspicuous. No such aberrant
individuals have yet been found, though they
are almost certain to turn up when very large
series are collected. Mayr & McEvey (p. 178.
179) mention some aberrant birds from the
range of horsfieldii, but the position in south-
eastern Australia is probably different as set
forth below.

Another possibility, that the birds would
migrate in winter but return to their place of
birth in the breeding season, can be ruled out
for the same reason as given above: in the
period from May to August, the only months
in which no breeding is known to occur 'McEvey
1960), not a single aberrant bird that might
be regarded as a migrant of a population dif-
ferent from the local one was noted. It is also
difficult to see how selection for soil colour
could take place in such a case.

The preceding notes can be taken as evidence
against the occurrence of seasonal movements
in Western Australia, a point about which Mayr
& McEvey (p. 189) enquire. It is perhaps

significant that the race horsfieldii, the only one
in which substantiated evidence of seasonal
movements exists, has a wide distribution and
is according to Mayr & McEvey more or less
uniform over its whole range.

Finally. I want to express my agreement with
Mayr & McEvey’s thesis (p. 155): ‘‘This single
selection force, colour of the substrate, seems to
over-ride all others”. It is remarkable that this
should be so. particularly becau.se — as showm
on a previous page — no active preference seems
to have developed with it. Anyway, there is no
doubt that Mirafra javanica is a highly success-
ful species in Western Australia, that occurs
wherever suitable habitat is present. Contrary
to most other forms of wild life it may even
look forward to a rosy future, for man's activi-
ties, like stocking remote areas, are likely to
increase rather than decrease the extent of its
habitat.

References

Bourke, P. A. (1947). — Notes on the Horsfield Bush -Lark.
Em7i 47: 1-7.

Keast, J. A. (1959). Australian birds: their zoogeography
and adaptations to an arid continent, in
Bodenheimer: "Biography and Ecology in
Australia" (Junk: Deu Haag).

Mayr. E. & McEvey. A. R. (1960). — The distribution and
variation of Mirafra favanica in Australia.
Emu 60; 155-192.

McEvey. A. R. (1S60). — Oology of Mirafra javanica in
Australia. Emu 60: 249-261.

Mees, G. F. (1961). — An annotated catalogue of a col-
lection of blrd-sklns from West Pilbara,
Western Australia. J. Roy. Soc. W. Aust. 44:
97-143.

50

7. — Land Forms and Soils in the Avon Valley, near York,

Western Australia

By R. L. Wright*

Manuscript received — 19ih September, 1B$1

Four stages of downcutting and base-levelling
are recognised between a deeply weathered
surface and the main channel in part of the
Avon valley, Western Australia. Corresponding
to the first three stages are two rock-cut ter-
races and a double phase alluvial terrace. The
fourth stage Is represented by a later Incision
of the Avon and Its tributaries.

The form and distribution of remnants of
the deeply weathered surface and the terraces
are described in detail, and the relationships
and regional extent of the land forms recognised
in the field are further Illustrated by carto-
graphic analysis. This account of the land
forms Is followed by a discussion of their
relationships to the soils of the area which
have been previously described by Mulcahv
( 1959 . 1960 ).

The relationships between soils and eroslonal
and depositlonal surfaces are shown to be com-
plex. The main conclusions are that most of
the soils of the area are the result of geomor-
phologlcally relatively recent episodes; that
youthful soils can occur on geomorphologlcally
“old” sites: that sites inherited from the same
stage of landscape evolution may have differ-
ent soils; and that similar soils may occur on
sites relating to different geomcrphological
stages.

A denudation chronology concUides the paper.

Introduction

This paper describes cyclic land forms in the
Avon valley near York. Field work was con-
fined to that part of the valley between
Hamersley and Mount Hardey. however, detailed
mapping of land forms was carried out during
four weeks' field investigation of a small area
shown in Figure 1. The land forms recognised
in the field are supplemented by cartographic
analysis. Both field mapping and cartographic
analysis were concerned only with land forms
however. Mulcahy ( 1959. 1960) previously

mapped in detail the soils of the area with par-
ticular reference to their development in rela-
tion to stages in the erosion of a lateritised land
surface. Therefore, the account of the land
fornp in the present paper is followed by a dis-
cussion of their relationships to the soils as
mapped and described by Mulcahy.

The upper Avon follows a north- to north-
westerly course on the plateau surface of West-
ern Australia, approximately parallel with the
coast and about 60 miles inland. At Toodyay
it turns south-west through the escarpment
and reaches the coast near Perth. The plateau
surface in this area was named the Darling
Plateau or Peneplain by Jutson (1912) and it
is bounded on the west by the Darling Scarp.
It lies mainly between 800 and 1,000 feet above
sea-level, rising very gradually eastwards.

• Division of Land Research and Regional Survey.

C.S.I.R.O.. Canberra. A.C.T.

The Avon and its tributaries in the York area
follow two main directions: north-north- west,
corresponding to the strike of the underlying
rocks, and north-east or east-north-east, an
important jointing and fracture trend iPig. 2).
The Avon valley here forms a trench between
500 and 600 feet above sea-level and a mile or
so wide. Away from the river, the landscape is
that of a sub-maturely dissected plateau. To
the west, dominantly on granite, the plateau is
generally between 1.000 and 1.200 feet above
sea-level. Where gneiss crops out it is inter-
bedded with discontinuous quartzite bands and
other meta-sedimentary lenses (Johnston 1952),
These have been etched into bold steep-sided
hills such as Mt. Mackie (900 ft.) and Mt. Bake-
well (1,500 ft.). On the less resistant gneiss
east of the river the plateau is between 750 and
850 feet above sea-level. Rising above, however,
are the Needling Hills (1,200 ft.), formed on
quartzite, and Mt. Brown (1,000 ft.) and Mt.
Hardy (900 ft.) on granite. More resistant
lenses in the gneiss show themselves as local
slope breaks and minor valley-side benches.
Low narrow ridges are formed by granite and
dolerite dykes, the most prominent of which
have a north-north-west or north-east trend.

Field Methods

The benched valley sides of the Avon are
evidence that incision has taken place in stages,
with gently inclined surfaces or “flats” corre-
sponding to prolonged periods of base-levelling
and steeper linking slopes representing the
intervening periods of downcutting. A study
of the valley was made to see if stages of down-
cutting and base-levelling could be recognised.

Field work consisted of the recognition of
“fiats” and breaks of slope, and mapping tk em
using aerial photographs. In this context “flat”
does not imply lack of slope. It is an accepted
term as used by many geomorphologists (e.g.,
Linton 1948). The important considerations in
defining “flats” are the delimiting breaks of
slope between the “flats” and the steeper link-
ing slopes, and the break of slope at the lower
margin of a "flat” is particularly significant.
This break is commonly relatively well-defined
(see Figs. 3 and 4) and can be measured to
within a few yards and mapped with little dif-
ficulty. However, mapping is more difficult if
this lower break of slope is a convex “curved
break of slope” (Savigear 1956). In this case
the lower margin of the “flat” is mapped at the
measured upslope boundary of the convexity.
Slope breaks determined by outcrops of more
resistant rocks were distinguished from those

51

In FIG 3 PROFILES
(a), (b) FIG 4 SECTIONS

DEEPLY WEATHERED SURFACE
TERRACE I (OUTER MARGIN)
TERRACE Q (OUTER MARGIN;
INCISED DRAINAGE
^OOO" CONTOURS

3 Miles
J

Fig. 1. — Distribution of cyclic elements in the landscape.

which transgressed structure and which were
therefore interpreted as delimiting remnants of
the sides and floors of former valleys.

Although field mapping was not extended to
the Mortlock River, correlation of similarly
staged terraces within the Avon and Mortlock
catchments was established by their continuity
across the watershed at higher levels,

A barometric altimeter was used to supple-
ment map contour data fNo. 400 York 1-mile
series, Second Edition). Two sections were
levelled and other slope angles were measured
with an Abney level.

Cyclic Elements in the Landscape

Deeply Weathered Surface
Preserved on the higher parts of the crests
of the main interfluves are the remains of what
was formerly an extensive deeply weathered
erosion surface, remnants of which are readily
recognisable by their laterite cappings, delimit-
ing breakaways, and general accordance of level.
Although these remnants have been mapped by
Mulcahy they are vei*y restricted in area and.
therefore, consideration of them is important if
any attempt is to be made to reconstruct the
nature of the original surface. Thus, a brief

52

description is given of the altitudes and forms
of remnants in the area of detailed field work.

The largest remnant mapped <Fig. 1) is
situated one mile east of Ironbark farm
(862653)*, at an altitude of between 800 and
850 feet above sea-level, and forms part of the
Avon-Mortlock watershed. The western and
northern part of the residual consists of a broad
rounded rise between 30 and 40 feet above the
remainder, with a flat crest 60 yards wide.
On the north and south flanks of the rise, slopes
are relatively steep, attaining 4 per cent. Below
the rise the surface slopes at 1 per cent, in
the west, but elsewhere has no appreciable
gradient. The I’emnant is delimited by break-
aways between 10 and 30 feet high with slopes
up to 40 per cent., except to the south, where
the boundary is a long slope of between 3 and
4 per cent.

Immediately east of this residual, at slightly
more than 800 feet above sea-level, is a smaller
residual of the deeply weathered surface. Its
crest slopes are up to 4 per cent., as on the
Ironbark residual, and it is bounded by break-
aways only slightly less pronounced than those
of the larger residual.

♦ The figures in brackets are grid references on the
York 1-lnch map.

To the south of the Ironbark residual are
four small remnants of the surface extending
in a belt noith to south. The northernmost
one, east of Modedine (860632), is at a height
of 800 feet. The second remnant is at 850 feet
and is preserved on an outcrop of resistant rock
which may account for it being somewhat
higher than the other remnants. The third
remnant is slightly less than one mile north-
east of Red Knob (8706111, and the fourth,
worked for gravel, is crossed by the Goldfields
Road at Poison Hill <880601).

Three miles east, at Collins Hill, (920623). is
another small outlier of the surface at a little
above 800 feet. This residual contrasts with
other remnants of the old surface in that it
has no delimiting breakaways. It is bounded
by slopes less than 7 per cent.

East of Mt. Brown is another group of rem-
nants of the old surface. These are generally
at a height of 850 feet, but to the south an
altitude of 900 feet is reached.

In this area, therefore, remnants of this
erosion surface are found only on the watershed
between the Avon and Mortlock Rivers and are,
with one exception, very small.

These remnants east of the Avon are cut
in gneiss and are between 800 and 900 feet

Fig. 2.— Geology and drainage.

53

above sea-level. West of the river, however, on
the more resistant granite and meta-sedimen-
tary rocks, remains of a laterite-capped surface
which is correlated with that east of the Avon
are 200 to 300 feet higher. Numerous dis-
sected remnants, generally between 1,000 and
1,150 feet, extend from Mt. Ommanie (711657)
in the north to the "Stockdale” (708472) area
in the south.

Rising above the level of the deeply weathered
surface are Mt. Bakewell (1.500 ft.) and the
Needling Hills (1,200 -|- feet). Woolnough (1918)
suggested that hills standing above the plateau
level indicate an earlier surface which he named
the “Mount Dale Level”. Mt. Bakewell and the
Needling Hills are satisfactorily explained as the
topographic expression of resistant quartzites
on the old erosion surface. There is no evidence
of an earlier erosion surface in this area.

The Terraces

The deeply weathered surface has been
strongly dissected by the Avon and its tribu-
taries. Pour stages are recognised between the
old surface and the main channel. Correspond-
ing to the first three stages are two rock-cut
terraces and an alluvial terrace, each comprising
backing “slopes" at its inner margin, furthest
from the river, and lower, gently sloping “flats”
as its outer margin, nearest the river. The
fourth stage is represented by a later incision
of the Avon and its tributaries. The break of
slope between the “flat” of one terrace and the
backing “slope” of the terrace below is shown
on the map illustrating the distribution of
remnants of the ten*aces (Fig. 1). The terraces
can be traced almost continuously, rising up-
stream in the area mapped.

Terrace /. — The inner limit of the first stage
of entrenchment is the breakaway which
generally separates it from remnants of the
deeply weathered surface. A long concave slope
leads down to the terrace “flat” as much as
100 feet below. The terrace flat has been dis-
sected into spurs with gently sloping crests, and
the equivalent stage is continued into the ad-
jacent tributary valleys by slope breaks and
minor benches. The form of the Terrace I
spurs is illustrated in Figure 3. Except for
Spur IV, the crests are up to about 100 yards
wide and 400 yards long with slopes of about
1 per cent, towards the river. Spur IV. at the
head of a tributary valley, is gently rounded in
cross-section with marginal slopes of 3 per cent.
The margins of the crests generally slope at
between l and 2 per cent, towards the steeper
(up to 12 per cent.) flanks of the spurs. The
backing slopes, leading up to and including
the discontinuous breakaways of the deeply
weathered surface, attain 7 per cent.

The terrace was mapped on the east side of
the Avon valley, where it was seen to rise up-
stream and up both .sides of tributary valleys.
In the downstream part of the area it is
generally at an altitude of 700 feet. However,
below Mt. Bi'own it is at about 725 feet, and up-
stream it attains 800 feet near Quartz Hill
(848518). The terrace rises away from the Avon
to between 750 and 800 feet in the heads of
tributary valleys, e.g.. east of Ironbark farm
and near Poison Hill.

Although the outer margin of the terrace
occasionally coincides with outcrops of more
resistant rocks it does not follow any one band
for any distance and locally cuts across struc-
ture. Thus, in the tributary valleys west of
Poison Hill, the terrace transgresses minor
breaks of slope formed by thin band.s of more
resistant gneiss. Similarly, around Mt. Brown,
the terrace leaves the gneiss and transgresses
the granite.

The terrace was mapped in tributary valleys
as well as in the main valley, as “paired”
benches on both valley sides. What gradient
there may be on these valley-side "fiats” is to-
wards the Avon and the remnants can be traced
rising steadily up the main valley. In the tri-
butary valley west of Poir>on Hill the longitu-
dinal gradient of the old valley floor, measured
from the surviving remnants of the terrace on
the north side of the valley, is 2 per cent, to-
wards the Avon.

These characteristics all point to this land
form being the remains of a previous valley
floor, a cyclic feature, and a distinct stage in
the entrenchment of the Avon into the deeply
weathered surface.

Terrace II . — Thts terrace occurs primarily as
spur crests and also as restricted benches and
slope breaks which occur on valley sides below
the level of Terrace I. The spur crests are
usually less than 100 yards wide and gently
rounded in cross section (Hg, 3). They slope
at between 1 and 3 per cent, towards the spur
sides, which have gradients of up to 9 per
cent. In longitudinal profile the crests slope
at between 0.5 and 2 per cent, and are up to
300 yards long. The backing slopes are up to
7 per cent, but have local structural benches.

Terrace II is at an altitude of just above 600
feet in the downstream part of the area, rising
to 650 feet at the head of the tributary valley
in which Mobedine farm is situated. The ter-
race rises up the main valley and attains 700
feet south of Mt. Brown. It is also well-pre-
.'rerved near Hamersley (742699), extending from
a mile or so east of Rivoli (736694), following
the 600 feet contour westwards and rising up the
valley of Heal Brook to an altitude of 650 feet
west of Mt. Mackie.

Terrace II has been more closely dissected
than Terrace I. Thus, it is preserved as rounded
spur crest.s, with “flats” less typical than in
Terrace I. It does, however, have the same
fundamental characteristics as Terrace I. It
cannot be explained by more resistant outcrops
forming local base levels of denudation. Locally
it cuts across them, as at Mt. Brown where it
transgresses the granite boundary. In addi-
tion. the feature can be traced along both sides
of tributary valleys, rising upstream. It is, there-
fore, interpreted as indicating a further stage
of downcutting in the Avon valley.

Terrace III . — Unlike the older, rock-cut ter-
races, Terrace III is a depositional feature. It
comprises the present valley floor of the Avon
and the lower parts of tributary valley floors
and represents a period of aggradation follow-
ing the dissection of Terrace II. It is at an
altitude of about 550 feet flanking the Avon River
but rises to about 600 feet in the tributary

54

LONGITUDINAL PROFILES CROSS PROFILES

FEET

Pig 3 —Profiles illustrating form and height range of Terrace I and Terrace II remnants (Vertical exaggeration

X 10). Locations shown in Pig. 2.

valleys west of Poison Hill (880601). The range
in form of the terrace is illustrated in Figure
4. It extends between i and 1 mile downslope
and comprises outer “flats” between 100 yards
and h mile long with gradients of about 0.6 per
cent.> and backing “slopes” up to about 3 per
cent.

The deposits of the terrace embrace two sub-
stages — an older deposit consisting of lateritised
arkosic grits being overlain unconformably by
younger fine-textured alluvium (Mulcahy 1959,
1960). The maximum exposed thickness of the
grit.-: is about 6 feet but the base was not seen.
The younger alluvium is up to 15 feet thick in
the area of detailed field work.

In the Avon valley numei’ous exposures of the
grits in the floors of channels incised in Ter-
race III suggest that they are continuous be-
neath the alluvium of the terrace. These grits
are correlated with similar deposits, called the
Mortlock beds, which occur at a high level in
the upper reaches of the Mortlock River and
which occupy shallowly dissected valley floors
cut in Terrace I — i.e., Terrace II stage sites.
However, in contrast to the more or less com-
plete lateritic profiles in the Mortlock valley,
the deposits in the Avon valley have been trun-
cated and then reburied by the younger alluvium
(Mulcahy 1959, 1960).

Thus, Terrace III stage comprises two sub-
stages — the first consisting of downcutting into
Terrace II and the subsequent deposition of the
Mortlock beds; the second consisting of erosion
and truncation of the Mortlock beds followed
by the deposition of fine-textured alluvium.

The Channel and Lower Flood-plains . — The
youngest phase of downcutting has reached the
stage where the Avon and its tributaries are
incised up to 25 feet into Terrace III. and the
channel floors are cut into the weathered grits
(Fig. 4). Restricted lower flood-plains are
forming locally. These occur marginal to the
Avon and in the lower courses of tributary
streams and are typically less than 100 yards
wide.

Cartographic Analysis

Cartographic analysis of the York sheet (No.
400 1 mile-series, Second Edition) was applied
after field work had been completed and after
the cyclic features described above had been
recognised. The aim of this analysis was to
determine the regional significance of these
features and to illustrate their forms and their
relationships to each other.

Projected Profiles

Projected profiles (Barrel! 1920) were con-
structed for that part of the Avon valley between
Gwambygine (825453) and Hamersley (742699),
with parallel lines of section drawn normal to
the general trend of the valley (Fig. 5a). This
technique projects surface levels over a wide
area into one view to attempt to illustrate the
original forms of surfaces since dissected.
Similarly, by projecting features related to
stages of downcutting and base levelling — valley
side benches and “graded” valley floors — it could

55

a?.so be expected to show the relationships and
distribution of such intact forms in different
parts of the catchment.

The profiles were simplified and the various
levels emphasised by showing only interfluve
crests, valley side benches and valley bottoms,
and omitting connecting slopes.

The most striking feature on the profiles is
that the “reconstructed” deeply weathered sur-
face is shown to be considerably lower east of
the river that to the west. East of the Avon
it is generally 800-850 feet above sea-level, rising
away from the river to 900-950 feet. However,
west of the Avon this old surface is usually at
1,100-1,150 feet attaining 1,200 feet or so away
from the river and sloping gently to 950 feet
close to the Avon. Mt. Bakewell and the Need-
ling Hills rise well above the general level of
the surface.

The profiles illustrate Terrace I and Terrace
II stages of downcutting, shown as the valley
bottoms of entrenched tributary streams and
the valley side benches of the Avon.

West of the Avon, Terrace I stage is indicated
at about 700 feet close to the river, and by broad
valley bottoms rising away from the river to
850-900 feet and more in the extreme west. East
of the river this stage is reflected in benche.s
at 700-750 feet and in the level extending east-
wards at 750 feet.

Terrace II stage is seen in benches at 600-650
feet near the Avon and in broad valley bottoms
at this level east of the river, rising to 700 feet
in the extreme east. West of the Avon, Tei’race
II stage valley bottoms rise from 650 feet near
the main river, reaching 800 feet in the west.

Terrace III is illustrated by the broad “fiat”
at 550 feet west of the river.

River and Spur Profiles

River profiles were drawn because it might
be expected that, after rejuvenation, the surviv-
ing portion of an earlier profile would meet the
developing new profile in a nickpoint of charac-

teristic form. If such convexities or breaks in
the profile had cyclic significance they would be
expected to relate to remnants of the earlier
valley floors now preserved at terraces down-
stream from the nickpoints. Therefore, spur
profiles were drawn in conjunction with the
river profiles to verify whether these cyclic rela-
tions occurred. Figure 5 (b and c> shows the
profiles, and their location is shown on Figure
2 .

Structural differences may explain breaks in
river profiles. However, in view of the highly
foliated nature of the underlying rocks, it is un-
likely that structural differences could cause
corresponding breaks in the river profiles and
levels on the spur profiles. Neither could under-
lying structure result in the formation of the
distinct levels that characterise all the spurs.

In the western tributaries of the Avon, Ter-
race I stage nickpoints range in height from
750 to 950 feet, depending on how much the
lower profiles of this stage have been destroyed
by Terrace II rejuvenation. Nickpoints occur
at 650-800 feet where Terrace III stage re-
juvenation meets the Terrace H stage valley
floors.

Only two nickpoints are revealed in the river
profiles east of the Avon. These are at 750 feet
and may represent the junctions of Terrace II
stage and Terrace III stage profiles. The absence
of other breaks in the streams east of the Avon
may be due to more rapid regrading on the less
resistant rocks of that part of the area.

The spur profiles illustrate the associated
valley-side features. West of the Avon Terrace
I remnants are preserved at 750-800 feet and
Terrace II remnants at 650 feet on all the spurs.
East of the river these stages are indicated at
700-750 feet and at 600-650 feet on the spur
profiles.

Clinographic Curves (Fig. 6(

Areas between successive contours were
measured and clinographic curves (Hanson-
Lowe 1935) were drawn for part of the valley

Pig. 4._Terrace III sections showing (a) the “hat” where terrace is broad, and (b) the whole feature where
terrace is narrow. (Vertical exaggeration x 10). Locations shown in Fig. 2.

56

F««t above Sea Level Feet above Sea Level Feet above Sea Level

MILES

FIGURE 5. Profiles in the York area showing (a) projected profiles looking downvalley, (b) river and spur profiles west of the Avon, and
(c) river and spur profiles east of the Avon. (Vertical exaggeration x lO). Locations shown in Fig. 2.

1450 1350 1250 1150 1050 950 850 750 650 550

FEET ABOVE SEA LEVEL

Pig. 6.— Clinographic curves (a) west of the Avon, and (b) east of the Avon. Histograms showing
in successive radii, (c) west of the Avon, and (d) east of the Avon.

differences

57

west of the Avon and part to the east (area
ABCD on Fig. 2).* Histograms were drawn
showing the differences between succes.sive radii
used in constructing the clinographic curve.!
These have considerable usefulness in emphasis-
ing slope change."; the convex lower margins of
relatively gently sloping sectors of the curve
appear as maxima, whilst the backing concavi-
ties appear as minima.

West of the Avon . — The level of the deeply
weathered surface here is reflected by a gently
sloping section in the clinographic curve (Fig.
6a) at 1,000-1.150 feet and by a broad maximum
between these levels on the hi.stogram (Fig. 6c ).
Hills above are indicated at 1,150-1,250 feet in
the curve and by a slight maximum on the histo-
gram. The top of the curve rises sharply to Mt.
Bakewell. above 1.450 feet.

Terrace I stage incision into the upland sur-
face is represented by the steepening of the
curve at 1,000 feet and by decreasing column
heights on the hi.stogram. The gentler lower
slopes of this stage cause a break in the con-
tinuity of decreasing column height at 800-850
feet.

The further steepening of the curve below 750
feet, and the histop’am minimum at 700-750 feet
mark the upper limit of Terrace II stage inci-
sion. The convexities bounding the lower slopes
of this .stage are indicated by a .slight maximum
at 600-650 feet on the histogram.

Terrace III stage downcutting is not shown
by the curve or by the histogram.

East of the Avon . — The clinographic curve
(Fig. 6b » indicates that there is a greater area
of country at 800-850 feet than between succes-
sive contours at lower levels. The flattening of the
curve at 800-850 feet probably records the exist-
ence of both deeply weathered surface remnants
and Ten-ace I "flats". This may be due to the
fact that remnants of the deeply weathered
surface are small, or that the vertical interval
between them and Terrace I is less than 100
feet, or that the connecting slopes between the
deeply weathered remnants and Terrace I are
gentle.

The steepening of the curve at 650-700 feet
and the histogram minimum between these
levels indicate the striking slope change asso-
ciated with vigorous Terrace II stage incision.
The flattening of the curve at 600-650 feet and
the corresponding histogram maximum illustrate
the change to the graded lower .slopes of Terrace
II.

A very slight steepening below 600 feet in the
curve indicates Terrace III stage incision.

Land Forms and Soils

Mulcahy (1959. 1960) has correlated the soils
of the area with a number of “erosional and
depositional surfaces". His Quailing surface
consists of a laterite which occupies the highest
parts of the landscape. The Kauring surface is

* In the clinographic curve the radii of circles equal lii
area to that within each contour (x axis) are
plotted against height above sea-level (y axis) to
give mean gradients.

t The author is grateful to J. A. Mabbutt of the Divi-
sion of Land Research and Regional Survey for
suggesting the construction of these histograms.

a younger laterite which occurs at a slightly
lower level. The Quailing and Kauring surfaces
are regarded by Mulcahy as part of the "Old
Plateau" (Jutson 1934) and thus he deduces
that the laterite has formed in at least two
stages.

Sandy deposits derived from the Quailing and
Kauring surfaces and named the Quailing de-
positional and Monkopen surfaces form features
termed "spillways".

Mulcahy believes that rejuvenation of drain-
age in this area took place in response to post-
Tertiary uplift and that this resulted in a num-
ber of "erosion cycles". The oldest of these is
represented by the Belmunging and Mortlock
surfaces, "the remnants of the sides and floors
respectively, of valleys cut in the old plateau”
(Mulcahy 1960. p, 211).

His second cycle is represented by the Bal-
kuling surface which is a pediment cut in the
transitional and pallid zones. It is extending
by the retreat of "breakaways which bound the
old plateau and the Belmunging surface and
hence it must be younger than both of them"
(Mulcahy 1960. p. 213).

Mulcahy's youngest "cycle of erosion" is re-
presented by the York and Avon surfaces, the
upper limit of which is marked by "a slight in-
crease of slope" (Mulcahy 1960, p. 214) below
the Balkuling surface. Mulcahy and Kingston
(1961) mapped three soil types on the York
surface — shallow stony soils lYl in Fig. 7), red-
brown earths (Y2 and Y3 in Pig. 7), and poorly
drained soils (Y4 in Fig. 7).

The Mobedine surface of Mulcahy occurs as
a scree forming the noses of ridges at about the
600 feet contour in the Avon valley.

Summarising, therefore, Mulcahy recognises
three cycles of erosion resulting from the re-
juvenation of drainage whereas four stages in
the downcutting of the Avon into the deeply
weathered surface have been recognised in the
field by the author and described above.
Because of this, and in view of Mulcahy’s as.-ser-
tion that "the field evidence, then, shows that
a good correlation exists between the distribu-
tion of the soils and that of the major geo-
mcrphic surfaces, for which a relative age
sequence can be established" (Mulcahy 1960, p.
215), the distribution of soils is reviewed in the
light of the four stages recognised geomorpho-
logically.

The soils as mapped by Mulcahy and King-
ston (1961) and the cyclic elements recognised
here in an area exetending north-east from
Mackias Crossing are shown in Pig. 7. Because
of the complexity of this map. Pig. 8 shows
typical relationships between the cyclic land
forms, weathering zones, superficial deposits
and soils, generalised in three dimensions for
most of the area shown in Fig. 7.

The Soils of the Deeply Weathered Surface

The Quailing laterite is found on the higher
and lower parts of the largest remnant of the
deeply weathered surface east of Ironbark farm,
with the Quailing and Monkopen depcoitional
materials essentially on the slopes fringing the
higher part and extending down into the heads
of dissecting streams. The sites of the Kauring

58

DEEPLY WEATHERED SURFACE

59

laterite are at the margins of the Ironbark
residual, in areas of waxing slope development.
On the north and east they occur immediately
above the dissection heads of tributaries of the
Mortlock River. The Belmunging soil surface
also occurs on the lower parts of the Ironbark
residual but extends downslope transgressing the
boundai'y of the remnant. The other remnants
of the deeply weathered surface typically carry
only the Quailing laterite with some Quailing
depositional sands but the Balkuling and Bel-
munging soil surfaces, with exposures of
weathered rock, are found on the remnant im-
mediately north-west of Poison Hill.

Terrace I Soils

Terrace I stage sites carry a variety of soils.
In that part of the area drained by the Mort-
lock headwaters, the Belmunging soil surface
occurs extensively on the inner slopes below
remnants of the deeply weathered surface,
generally where the.se are not bounded by break-
aways. Good examples are found immediately
north and south of the Ironbark residual and
on slopes marginal to the Collins Hill residual.

Immediately south of the Ironbark residual of
the deeply weathered surface there is a small
outlying remnant of the Quailing laterite which
forms a bench on Terrace I. Here, therefore,
the terrace must be cut in the uppermost parts
of the old deep weathering profile.

The Balkuling soil surface is found down-
slope from remnants of the deeply weathered
surface but is usually separated from these by
breakaways. The soils have as their parent
material the pallid zone of the Quailing and
Kauring laterites or weathered rock. They occur
on pediments below the breakaways occasionally
extending downslope on to spur crests formed
by the remains of Terrace I “flats”, particularly
east and south-east of Mt. Brown. Here, at the
heads of tributary valleys, the base level for
Terrace I dowmcutting was within the pallid
and traditional zones of the weathering pro-
file. Elsewhere, how^ever. Terrace I downcutting
proceeded below the depth of weathering and
in these situations the York red-brown earths
generally occur on the lower parts of the ter-
race. where it is cut on unweathered rock. In ad-
dition, Terrace I has locally been closely
dissected into rounded spurs at the heads of
tributary valleys, and these remnant.s cari’y the
shallow stony variety of the York soils (e.g.,
profile IV. Fig. 3).

Terrace II Soils

Terrace II sites in the Avon valley are mainly
characterised by the shallow stony soils which
form part of the York soil surface, and these
occasionally extend up to the inner slopes. How-
ever, York red-brow'n earths also occur on the
terrace. Certain tributary valley heads of this
stage of downcutting are the principal sites of
the poorly drained variety of the soil comprising
the York soil surface. York red-brown earths
also locally occur in the floors of these tributary
valleys.

In addition, the Belmunging soil surface
occurs on certain Terrace II sites in the head-
waters of the Mortlock. mainly the inner slopes

of the terrace. Thi.s is well illustrated one and
a half miles north-east of Poison Hill on the
slopes below a large remnant of Terrace I
(903620). These sites are often less than 50 feet
above the main tributary channel and locally
the Belmunging soil surface extends down on
to gentle slopes only a few feet above, and 100
yards or less from the channel. In the more
dissected valley of the Avon, where post-Terrace
II stage downcutting and base levelling is much
further advanced, similar Terrace II sites (with-
out Belmunging soils) are about 100 feet above
the main channel.

The Mortlock surface occurs downslope from
the Belmunging surface in the floors of the
Mortlock and its tributaries, which have been
interpreted as shallowly dissected Terrace II
stage “flats”.

In this area, the deposits of the Monkopen
soil surface are associated with the Ironbark
reoidual of the deeply weathered surface. They
originate on the residual and “spill” out
through the breakaways to occupy pre-existing
valleys cut through Terrace I. To the south-
east of this residual, however, the deposits
spread out on to the low' interfluves between
the valleys and extend down on to the gentle
lower slopes, cut during Terrace II stage. They
also locally overlie the restricted slopes cut dur-
ing Terrace III stage incision.

All the spillways in the present area, with
one exception, are confined to the Mortlock
headw’aters. The exception is a small spillway
at the head of the Avon tributary which flows
pe.it Ironbark farm (862653).

Terrace III Soils

The most extensive soil type associated with
Terrace III is the alluvium of the terrace itself —
the solonized grey or browm soils of the Avon
surface. Exposures of underlying grits in tri-
butary channels incised in to the terrace form
part of the Mortlock surface.

The Mobedine surface occurs at the inner
margin of Terrace III in the Avon valley. It is
found on the slopes below the noses of spurs,
the crests of w'hich are Terrace II remnants.

Lastly, areas of the York .soil surface, mainly
the shallow stony soils but locally also the red-
brown earths and poorly drained soils, occur
in tributary valleys on the floors and adjacent
slopes cut during Terrace III stage downcutting.

Discussion

The limited distribution of the Quailing and
Kauring laterite residuals in this area makes
any ambitious interpretations impossible. From
the geomorphological evidence they could satis-
factorily be regarded as the remains of an un-
dulating surface on which the Quailing laterite
developed and subsequently underwent varying
degrees of stripping. In the area of detailed
field w'ork, the Kauring laterite is found above
Terrace I stage dissection heads or in other
areas of waxing slope development, that is. on
sites where the (Quailing laterite has been
truncated. These sites would appear, there-
fore. to be related to the advance of Terrace I
stage downcutting into the surviving remnants
of the deeply weathered surface. Alternatively.

60

2

UJ

X

u

u

u

O

ac

O

5

X

u

h-

<

u

2

O

>

CO O

X

K Q

< ii!

OD QC

2o

oi. a.
Of QC

Rg

>»

73

a>

Q.

ft

e 3

s

OQ

ctf

WJ

O

03

03

03 •

o c-
ft

■Si*

Sil

is

O'"*

«d ^

§•1

“■S

03* CS
03 03

§«

a>

1id5

•C ^

«.2

ta-o

03

03

a

io
B a

t: V

P wt

•o

s-S

bh

61

a pre-TeiTace I stage of downcutting could be
invoked to explain the truncation of the older
laterite. However, there is no geomorphological
evidence to suggest this and the first hypo-
thesis is favoured.

Remnants of the Quailing and Kauring
lat rites show distinct and restricted occurrences
on the watershed between the Avon and Mort-
lock Rivers. The geomorphic relationships of
the younger soil surfaces are more complex and
they are not necessarily restricted to any of the
cyclic surfaces recognised here. This is illus-
trated by the extension of the Belmunging and
Monkopen soil surfaces from the deeply
weathered surface across Terrace I and Terrace
II and on to younger sites.

The degree of truncation of the weathering
profile has influenced subsequent soil develop-
ment on Terrace I sites. Balkuling soils occur
where the terrace is cut in weathered rock;
downslope. where it is cut on unweathered rock,
York red-brown earths are found.

The formation of the spillways probably in-
dicates conditions of surface instability such as
would be induced by a change to drier conditions
resulting in a reduced vegetation cover. This
would lead to the movement of sandy material
from remnants of the deeply weathered surface
into the upper parts of tributary valleys and
to a reduction in the competence of streams to
carry away the derived material.

The occurrence of more than one band of
ferruginous concretions within the deposits pos-
sibly indicates a number of depositional
“periods” separated by soil forming “periods”
(Mulcahy 1959, p. 46).

The only geomorphological evidence for the
age of the .spillways is that the bulk of the de-
posits occur in valleys cut through Terrace I
but that the youngest sites on which they occur
are the slopes cut below Terrace II. Therefore,
they are mainly post-Terrace I in age. whilst
the youngest deposits at least formed during
Terrace III stage. On geomorphological
grounds, therefore, it is possible that the spill-
way deposits are, in part at least, linked with
the Mcrtlock deposits. Some support is given
to this by the contiguity of the spillway and
Mortlock deposits in many parts of the Mort-
iock valley as mapped by Mulcahy. The coarse
arkosic sediments of the Mortlock surface also
indicate an extensive aggradational phase and
may be due to the same change to increased
aridity postulated to have resulted in the de-
position of the spillways. In the Avon valley
this aggradational phase is dated as part of
Terrace III stage, although equivalent deposits
and the tributary spillways occur on Terrace
II stage sites in the upper reaches of the Mort-
lock. However, Mulcahy has emphasised the
multiple natui-e of the spillway deposits and it
is entirely feasible that the spillways may have
operated at all stages in the destruction of the
deeply weathered surface, from Terrace I stage
onwards. Therefore, the formation of the spill-
ways is not restricted to any geomorphological
stage in the concluding denudation chronology.

The Mortlock depasits have been subsequently
lateritised. The greater degree eof weathering
of these deposits compared with those of the

spillways could partly reflect site differences —
the Mortlock deposits occur on relatively poorly
drained valley floors whereas the spillways
occupy more freely drained tributary valleys. In
addition, the spillway sands being derived from
the old laterites might not be capable of further
weathering, whereas much of the Mortlock de-
posits were no doubt derived from valley sides
downslope from the old laterites and would,
therefore, contain material which would weather
readily.

The Belmunging surface could have developed
on the valley sides during this second period of
lateritisation. It is noteworthy that the Bel-
munging surface ts absent from the Avon valley.
Firstly, this may be because it was never well
developed on the steeper slopes of the more dis-
sected Avon valley, although in fact there are
numerous “flats” and gentle slopes where it
could have done so.

Secondly, subsequent rejuvenation and dis-
section. which have been very active in the Avon
valley, could have removed all trace of the Bel-
munging surface. This same rejuvenation has
not yet attacked the headwater areas of the
Mortlock to the same extent and Belmunging
remains are preserved there. The truncation of
the Mortlock layer beneath the alluvium in the
Avon valley testifies that such erosion has in
fact occurred. The pisolithic gravels of the
Mobedine surface and the scattered pisoliths
which occur upslope from it (Fig. 4b) may have
been derived from a soil which was related to
the Belmunging surface and which has been
removed from the Terrace II "flats” above.
However, the Mobedine gravel is different from
that of the Belmunging surface (Mulcahy, per-
sonal communication).

Lastly, the Belmunging surface in the Mort-
lock catchment may be a.ssociated with part of
the truncated deeply weathered profile — and the
greater part of the Avon valley is cut below the
level of that zone of weathering.

Thus, the pattern of soils in the York area
indicates that a phase of surface instability,
probably due to the change to more arid condi-
tions. occurred during a first Terrace III sub-
stage. This could have resulted in the accumu-
lation of the Mortlock materials and of much
of the spillway deposits. The Belmunging,
Mobedine and Mortlock surfaces, and possibly
the Kauring laterite. could indicate a younger
weathering phase. The erosion which trun-
cated the Mortlock profile in the Avon valley
during the second Terrace III sub-stage may
have removed the Belmunging surface and the
spillways from the valley. As a result the York
surface and the related Avon depc.sits formed
extensively in the Avon valley and locally in
the Mortlock valley.

Conclusions

In contrast to the three “cycles” of erosion
described by Mulcahy (I960, p. 211) and the
two “cycles” he writes of most recently (Mul-
cahy and Kingston 1961, p. 32), here, four
stages are recognised in the downcutting of
the Avon into the deeply weathered surface.
Obviously, therefore, soil surfaces cannot be
identified with cycles of erosion as implied

62

ty Mulcahy. Thus, in this section of the
Avon valley there is a very complex relation-
ship between the soils as mapped by Mul-
cahy and erosional and depositional sur-
faces recognised by detailed geomorphological
analysis. This complexity is to some extent re-
solved by the following conclusions: —

(1) In the York area all the soil surfaces
other than the Quailing and Kauring
laterites are seen to extend on to Ter-
race 11 and even younger Mtes and are,
therefore, the result of relatively i*ecent
episodes in the evolution of the physical
landscape. The Balkuling soil surface,
which alone is restricted to higher
levels, i.s associated with active pedi-
ments and, therefore, in part at least,
must be very “young” geomorpho-
logically.

(2) Youthful soils can occur on geomorpho-
logically “old” sites where these have
been stripped of an earlier soil. This
is expressive of the fact that land forms
.survive when their soil cover has been
removed. Thus the York shallow stony
soils are particularly associated with
Terrace II sites and are occasionally
developed on Terrace I sites.

(3) Sites inherited from the same stage of
landscape hi.*>toi‘y may have different
soils. For example. Terrace I has been
shown to truncate the deeply weathered
profile and. therefore, it offers differ-
ent sites for soil development based on
differences of parent material.

(4) Since the inception of geomorphological
stages occurs at different times in dif-
ferent parts of an area and .since one
stage continues its development after
the inception of later stages, stage
in the geomorphological sense does not
have a strict time connotation and need
not correspond with one period of soil
formation. Thus, similar soil.s may
occur on sites relating to different geo-
morphological stages. For example, the
deposition of the Mortlock layer oc-
curred on the Terrace II stage valley
floor of the Mortlock and the Terrace
III stage floor of the Avon. Similarly,
one of the youngest of Mulcahy’s sur-
faces. the York surface, occurs on sites
with similar physical characteristics
but which relate to different geomor-
phological stages.

These conclusions illustrate the complexity of
correlating soils and stages of geomorphological
evolution. Nevertheless, one example of the
value of analysins the relief of an area to see
how far stages of downcutting and base-levelling
could be recognised, could be in relating laterites
occurring at different levels in the landscape.
For instance Playford (1954) has suggested that
laterites occurring at differing levels in West-
ern Australia formed during one period of deep
weathering. If these laterites transgress land
forms relating to different geomorphological
stages, they could have been formed during one
period of deep weathering. If, on the other
hand, laterites at different levels are .separated

by one or more stages of downcutting which
truncate higher laterites then the lower
laterites would obviously be younger.

Denudation Chronology
The main stages of landscape development in
the York area may be summarised as follows: —

(1) The formation of the deeply weathered
surface. — This i.s the earliest recognis-
able stage in the history of the present
landscape and indicates a prolonged
period of base-levelling and the pro-
duction of a laterite profile.

(2) The Terrace I stage. — Consisting of
rejuvenation, downcutting and base-
levelling into and below the deeply
weathered surface, with the possible
initiation of spillway.s.

(3) The Terrace II stage. — Consisting of
rejuvenation, downcutting and base-
levelling into the Terrace I stage valley
floors, possibly with continuing spill-
way activity.

(4) The Terrace III stage. —

(i) Sub-stage 1. — This sub-stage
was initiated by rejuvenation
of the rivers and downcutting
into the Terrace II stage valley
floors. Broad valley bottoms
were eventually produced be-
low Terrace II.

A change to drier conditions
caused the deposition of the
Mortlock beds in valley bot-
toms and the filling in of tri-
butary valleys by “spillways”.

A subsequent weathering
phase resulted in lateritisation
of the Mortlock beds, the for-
mation of the Belmunging soil
on valley sides, and possibly
the development of the Kaur-
ing laterite on sites where the
Quailing laterite had been
truncated.

(ii> Sub-stage 2. — The rivers were
then rejuvenated and, in the
Avon valley, the Moitlock layer
was truncated, the “spillways”
were removed from tributary
valleys, and the Belmunging
surface was stripped from the
adjacent interfluves. The Mobe-
dine surface may have de-
veloped where lateritic gravel,
derived from the Belmunging
or a related profile, accumu-
lated on slopes cut in weathered
or relatively fresh rock mar-
ginal to the Avon valley bot-
tom.

This rejuvenation did not
invade the headwaters of the
Mortlock to the same extent,
and the spillways and the
Mortlock and Belmunging soils
are preserved there as a result.

The latter part of this stage
was marked by the deposition
of the Avon alluvium and the

formation of York soils where
older soils had been stripped
from the sides of the main
valley.

(5) The post-Terrace III stage. — A further
.stage of downcutting has resulted in the
incision of the rivers into the Terrace
in stage valley floor, and small flood-
plains are developing locally below Ter-
race in.

Acknowledgments

The author makes grateful acknowledgment
of the helpful criticism and advice received from
J. A. Mabbutt of the Division of Land Research
and Regional Survey in the preparation of this
paper. The Division’s drawing office is thanked
for preparing the figures. The author also
thanks M. J. Mulcahy of the Division of Soils
for introducing him to the area, and Professor
A. F. Wilson of the University of Queensland
for information concerning the geology of the
area.

References

Barren, J. (1920). — The piedmont terraces of the north-
ern Appalachians. Amer. J. Sci. 49: 227-258,
327-362, 407-428.

Hanson-Lowe, J. (1935). — ^The cUnographlc curve. Geol.
Mag. 72: 180-184.

Johnston. M. H. (1952). — The geology of the Hamersley
siding area. J. Roy. Soc. W. Aust. 36: 45-72.

Jutson, J. T. (1912). — Geological and physlographlcal
notes on a traverse over portions of the
Darling Plateau. Bull. Geol. Surv. W. Aust.
48.

Jutson. J. T. (1934). — The physiography (geomor-
phology) of Western Australia. Bull. Geol.
Surv. W. Aust. No. 95.

Linton, D. L. (1948). — The deliinltatlon of morphological
regions. London essays in Geography: 199-
217.

Mulcahy. M. J. (1959). — Topographic relationships of
laterlte near York. Western Australia. J. Roy.
Soc. W. Aust. 42: 44-48.

Mulcahy, M. J. (1960). — Laterltea and laterltlc soils in
South-Western Australia. J. Soil. Sci. 11:
206-225,

Mulcahy, M. J. and Hlngston. P. J. (1961). — ^The develop-
ment and distribution of the soils of the
York-Qualrading area. Western Australia, In
relation to landscape evolution. C.S.I.R.O.
Aust. Soil Publ. No. 17.

Playford, P. E. (1954). — Observations on laterlte in West-
ern Australia. Aust. J. Sci. 17: 11-14.

Savlgear. R. A. G. (1956). — Technique and terminology
In the investigation of slope forms. Premier
Rapport de la Commission pour I’Etude des
Versants, 1 : 66-75.

Woolnough, W. G. (1918). — The Darling peneplain of
Western Australia. J. Roy. Soc. N.S.W. 52:
385-396.

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, (h) Honorary Members, (c) Non-members resident outside
Western Australia, who must communicate the paper through an Ordinary or
Honorary Member. No paper shall be accepted from a Non-member resident in
Western Australia.

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

Journal

of the

Royal Society of Western Australia, Inc.

Volume 45 1962

Part 2
Contents

5. — Some Interesting Stomatopoda— mostly from Western Australia. By

W. Stephenson.

6. — Larks, Mirafra javanica, of tropical Western Australia. By G. F. Mees.

7. — Lands Forms azid Soils in the Avon Valley, near York, Western Australia.

By R. L. Wright.

Editor; J. E. Glover

Assistant Editors: G. F. Mees, R. D. Royce
Annual Subscription: Forty Shillings

The Royal Society of Western Australia, Inc., Western Australian Museum,

Perth

57799/4/62—535

ALEX. B. DAVIES, Government Printer, Western Australia