ay cee
PU ee ee eee

Saeee as
3, y:
CE tettares re
oie

Yes

ets :
Hy Ton A Bae,

yy

NOY « a9 AAD Se

i,

v ee) o
4 BRITISH : MUSEUM (NATURAL HISTORY) ood. 1g 1G.1V. /.192

——

INSECTS OF SAMOA
AND OTHER SAMOAN TERRESTRIAL
“ARTHROPODA

PART II LEPIDOPTERA :
FASC. 3. Pp. 117-168 7 |

GEOMETRIDAE
By LOUIS B. PROUT, FES.

- WITH TWO TEXT-FICURES AND ONE PLATE

’ LONDON — :
PRINTED BY ORDER OF THE TRUSTEES OF THE BRITISH MUSEUM
SOLD AT {

THE British Musnum (Naturan History), Cromwern Roan, 8,W.7

AND BY

B, Quarrron, Lrp.; Dutav & Co., Lrp.; Tax Oxvrorp Universiry Press; anp
WHatpon & Veet Lrp., LONDON; ALSO BY OLIVER & ae HDINBURGH

1928

“Isoued 24th March, 1928.) . [Price Two Shillings and Sixpence.

INSECTS OF SAMOA

AND OTHER SAMOAN Tee ®
ARTHROPODA

Although a monograph, or series of papers, dealing comprehensively with

the land arthropod fauna of any group of islands in the South Pacific may be _ |

expected to yield valuable results, in connection with distribution, modification
due to isolation, and other problems, no such work is at present in existence.
In order in some measure to remedy this deficiency, and in view of benefits
directly accruing to the National Collections, the Trustees of the British
Museum have undertaken the publication of an account of the Insects and other
Terrestrial Arthropoda collected in the Samoan Islands, in 1924-1925, by
Messrs. P. A. Buxton and G. H. E. Hopkins, during the Expedition of the
London School of Hygiene and Tropical Medicine to the South Pacific.
Advantage has been taken of the opportunity thus afforded, to make the studies —
as complete as possible by including in them all Samoan material of the groups —
concerned in both the British Museum (Natural History) and (by courtesy of
the authorities of that institution) the Bishop Museum, Honolulu.

It is not intended that contributors to the text shall be confined to the
Museum Staff or to any one nation, but, so far as possible, the assistance of the —
leading authorities on all groups to be dealt with has been obtained.

The work will be divided into eight ‘“‘ Parts ’’ (see p. 3 of wrapper), which
will be subdivided into “ Fascicles.”’ Each of the latter, which will appear as
ready in any order, will consist of one or more contributions. On the
completion of the work it is intended to issue a general survey, summarising
the whole and drawing from it such conclusions as may be warranted.

E. E. AUSTEN,

Keeper of Entomology.

British Museum (Natura. History),
CroMWELL Roan, S.W.7,

& 2 Re
R > nd re
% Q 3 Ss) ®
<= uA) x e) a
o % ie Ets
R Q aS) Ned
#4 3

wie Dew

f Samoc

s
&s

teglits

Inliegend erhalten Sie Besprechungsbelege aus:

Anatomischer Anzeiger .
Anatomischer Bericht

Archiv fiir Protistenkunde
Botanisches Centralblatt (N. F.)
Centralblatt f. Bakteriologio, |. Abt. (Ref.)
Centralblatt fiir Bakteriologie, |!. Abt.
Centralblatt fiir Pathologie
Jahrhiicher f. Nationalékonomie (Ill. F.)
Soziale Praxis . i
Weltwirtschaftliches Archiv
Zeitschrift fiir Botanik .

Zeitschrift fiir arztliche Fortbildung .
Zeitschrift fiir Immunitatsforschung .

Zoologischer Bericht .
Allgemeines Statistisches Archiv

TOR Bde 8 Nriie ae)
POMS Bde ar ce Heft
aS) ae BY tie aes INS
NQF esi cies IN eee
Oe Se eens Nr...

HOie ee Ba :
Qe BOs eee: Heft
NOs Bites Nps tees
MO Wee denies Heft

NG ES Bdts eee Heft
TORY cB deans: Nr
LOR eNO B di owen: Heft
LOS Bay eis Heft
He Bol Heft

; is totally
ollected by
1897) from
mination—
Kl., Ixxxv,
vo Memoirs
iographical

ost easterly
, excepting
inds at the
made some:
in ancestral
as: but 1t
OMETRIDAE
-more than
; there are
2zreat Indo-
lements) is
some Cases,
species (e.g.

a, Gymnoscelis

Hist,
schen

Lesse7
ernalt

ddd,
Fateu
der
mie Deu
f Same

teglit

sects of Samoa and other Samoan terrestrial Ar-
hropoda. Part III, fase. 3: Geometridae by. l;).B.-Prout.
ondon, British Museum (Nat. Hist.) 1928. 51 S., 2 Textabb., 1 Taf.
‘reis 2 Sh. 6 Pence.
Die Spanner der Polynesischen Region und der éstlichen melanesischen
In sind sehr unvollstandig bekannt. Nur Neu-Seeland und Hawaii sind
ser bekannt. Samoa und Tonga haben ihre Geometriden von Westen her
alten. Die ganze indo-australische Spannerfauna ist sehr homogen. Die
‘tungen und zum Teil selbst die Arten sind sehr weit verbreitet. Von
18 auf Samoa gefundenen Gattungen ist keine endemisch oder von be-
rankter Verbreitung. Von Arten und Subspecies sind nur 10 auf Samoa
chrankt, von denen 9 als neu beschrieben werden.

Zacher (Berlin-Steglitz).

|
1
:
|
|
|
|
|
!
1

sects of Samoa and other Samoan terrestrial Ar-
thropoda. Part III, fasc. 3: Geometridae by L. B. Prout.
London, British Museum (Nat. Hist.) 1928. 51 S., 2 Textabb., 1 Taf.
Preis 2 Sh. 6 Pence.

Die Spanner der Polynesischen Region und der éstlichen melanesischen
iseln sind sehr unvollsténdig bekannt. Nur Neu-Seeland und Hawaii sind
esser bekannt. Samoa und Tonga haben ihre Geometriden von Westen her
rhalten. Die ganze indo-australische Spannerfauna ist sehr homogen. Die
rattungen und zum Teil selbst die Arten sind sehr weit verbreitet. Von
en 18 auf Samoa gefundenen Gattungen ist keine endemisch oder von be-
chrankter Verbreitung. Von Arten und Subspecies sind nur 10 auf Samoa
leschrankt, von denen 9 als neu beschrieben werden.

Zacher (Berlin-Steglitz).

; is totally
ollected by
1897) from
mination—
Kl., lxxxv,
vo Memoirs

iographical

ost easterly
, excepting
inds at the
made some
in ancestral
a3) bub vit
OMETRIDAE.
-more than
; there are
2reat _Indo-
lements) is
some Cases,
species (e.g.

a, Gymnoscelis

i “Dee eb oO Os : 1d
EEElE ES ePeegePhPeege oA
oS oe oS oe Op 6 a. a ee ee
cs oO ga Par akzeE YH Sm
SRoetse Ee 8 oS Fee B's s

5) 5 ay o no & Ss 28 0
228-35 6 2 gsypst..g 3 3 q
~_om Soe Sn SS ee ae
fe PA~2)-y- we 2, ye rsD. A POD yA > 3 = 3 ;
Pay Arwavr yr ee a)
: z 4
~ \ ) 3
(222 [0a7g~ut TT 42g} 4! ew
heal = 4
ADYIDS — a
"UODLON WEPSILYISaS Nau ETH ~ UaUEH Waa “gyUDLYOSEG
pownge {no OT AM PUTyS saetuadegne pun ua r UOA SBUMGIALCLHs AOZYUNLYOS

Zo
“2G WK UO KPO YOSTWEPUe auULaY Foy Usbunzcgog wauaepun/el vowDs 72) OT
eer UOQ FEFIPAGHMIA FPSO KUIS PUIG Uagey ary yeqyes ThA une pun uab
“UNZGUN OFC cuelaurou uctyos 287 DUNO seUuUDds ayoerrpuzenvmoput azupe a
“UBZTVYULE AOU UBBEAM UO UaPEsZewdsH aay? Usgny PoUaT Pun —NEUDe “quupy ag
4OSSA9 PUTS PITMaALY prim puvpeec-nay,l anni Oppurs LS] POQU?) Iwas 922 upasur
USYOSPSSUD]L au USYOITESO sap pun iiseeu wayne saul [ Og! “ep sou vd sy 1d
ene 8 un ef 81 pdad "TafO5 1 « ‘sqqnaras a fee rc "B26)
(°R8IH 30M) UNOS WS? et As ‘uagwir "INO dot “ST °F eq PUPT ALY AMO 8 2) re *sen/

‘TIT 2407 "opadoay24y LEMAR AES 3 UD OUT £ A2uze pun Downe so szoasur

some Cases,
species (e.g.

Anisodes obliviaria, Eupithecia ewpitheciata, Micrulia tenuilinea, Gymnoscelis

refusaria, Orsonoba clelia),

ao

i

2
ci

1 (3)

<a

P3505 oes WASH Lat

so) MAGS deb stne a w SHON ohh adh 4 snbdndenoad 2% aden,
= | eHOMS% G WE S Bev sAsto? ehdntust Si we ke B22
‘i Tesnolsm sgsiot ls is6 Tab Havas rotgeh secs ssiganylo’ seh auiienec 9G
SG bNs {Poon bear banlesG-well “UW .Qbbe sel Lous “Hoe Jet shoawh
“OS RHISgy HOW. As sbtsSamos OV Siat cecal sapiet Hoes weiss ss ssaoinhed
wtG oR UOMO A fies Sel onwot wennnge asiootlowwen-ohst egang tO
SON .takh eNEXD Ateoi “sise tate SOSH 8 sth sedhee List mus bas x59
HH mas eho os imsbes silos ie | syouityny MeKnshews sy Somos Yin BL

_benge Asso of WM Hale sstoasqsdue Hrs sesth ao’

ree

eNSh¥aa Aedes) rdoaged ser elu.

“shoo

YSSNNsd moony G sate oO Hiei vome’ ces escent

eURU Ss EOMGWAY Ree

@ sensb sag “Khas é a6

[NSEOmS OF SAMOA

Part III. Fasc. 3

GEOMETRIDAE
By Lovis B. Prout, F.E.S.
(With 2 Text-figures, and | Plate.)

Existinc knowledge of the GEoMETRIDAE of the Samoan Islands is totally
inadequate, and indeed scarcely extends beyond the few species collected by
Woodford at Apia, Upolu, and described by Warren (Nov. Zool., iv, 1897) from
types in the Tring Museum, and the few—partly of uncertain determination—
recorded by Rebel (Denkschr. K. Akad. Wiss. Wien, Math.-Naturw. Kl., Ixxxv,
1910; Jahrb. Hamb. Wires. Anstalt, xxxu, Betheft 2, 1915). Rebel’s two memoirs
are further of value as containing some topographical and_ bibliographical
notes, but these do not concern the GEOMETRIDAE in particular.

As to the Geometrid fauna of Polynesia in general and of the most easterly
islands of Melanesia, our present information is equally fragmentary, excepting
only as regards New Zealand at one extreme and the Hawaiian Islands at the
other. Mr. Meyrick (Fauna Hawanensis, 1, (2), 124 seg. 1899) has made some
very interesting and thought-provoking comparisons, suggesting certain ancestral
connections between the New Zealand and Hawaiian Lepidoptera; but it
seems clear that Samoa (together with Tonga) has received its GEOMETRIDAE
from the west, and there is no need, on the present occasion, to make more than
a passing reference to those connections. With the Fijian species there are
some definite affinities ; but indeed the general homogeneity of the great Indo-
Australian fauna (omitting the Subantarctic and North Pacific elements) is
strikingly shown by the very close generic agreements, and even in some cases,
so far as can at present be determined, by the actual identity of species (e.g.
Anisodes obliviaria, Eupithecia ewpitheciata, Micrulia tenwilinea, Gymnoscelis
refusaria, Orsonoba clelia),

II (3) 1

118 INSECTS OF SAMOA.

The thirty species now known to occur in the Samoan Islands belong to
eighteen genera, none of which are endemic or even restricted in their geographical
range. A brief note on each genus from this point of view is given in its place,
although it must be admitted that, notwithstanding the work of the last forty
years, the taxonomy of the family does not yet rest, in all respects, on a very
stable basis. As regards the species or subspecies themselves, ten at least are
confined to Samoa, nine of these being here described as new ; coloured figures
will be given in Part III, Fascicle 4.

HEMITHEINAE.

In the collection before me, this subfamily is represented by only two
genera and, with the exception of a solitary specimen of Pyrrhorachis, by
Thalassodes alone. Agathia, Anisozyga(?) and Conmabaena, which reach Fiji
(Agathia dimota Prout, Anisozyga pacifica Felder,* Comibaena cheramota Mey-
tick), and Mesurodes, only known to occur in Fiji, would be not unhkely additions,
but most of the widely distributed genera seem to extend no further east than
the Solomon Is.

Thalassodes Guenée.
Spec. Gén. Lép., ix, 359, 1858. ee: ck, Trans. Ent. Soc. Lond., 1886, p. 204, 1886.—Prout,

Gen. Ins. 129, p. 151, 1912; Lep. Cat. 14, pp. 95-97, 1913.

Distributed throughout India, Malaya, Melanesia and Polynesia, with a
few wide-ranging species even in Africa. The Pacific Islands have long been
known as the habitat of three of the species, T. pilaria Guenée, T. chloropis
Meyrick, and T. tumoclea Druce, all of which, indeed, have already been recorded
from Fiji, so that their occurrence in Samoa is not surprising. Together with
many other members of this genus, they are almost identical in markings, so
that their discrimination one from another, to say nothing of possibilities of
racial variation and of the differentiation of further species within the same
group, presents problems of considerable intricacy, especially since a large part
of the available material is in a more or less defective condition. I believe I can
distinguish four Samoan species, apart from the striking novelty about to be
described and a few indeterminate specimens.

* As Felder’s unique type is exceedingly like the common A. TERRES Walker, of Australia,
there may be some error regarding its locality.

GEOMETRIDAE. 119

1. Thalassodes charops, sp. n.

$2, 538-54 mm. Face green. Palpus in 9 14, in Q over 14, terminal seg-
ment in both sexes moderately elongate, about 3 of second segment; green,
beneath white. Vertex and base of antenna white; occiput green. Thorax
and abdomen above green, with a conspicuous whitish longitudinal lme ; beneath
and at extremity white. Legs white, the anterior brown on inner side; hind
tibia in ¢ dilated, with hair-pencil, terminal process reaching middle of first
tarsal segment.

Fore wing with apex rather acute, termen oblique, almost straight, very
faintly sinuous, tornus pronounced ; M! separate ; American green, with short
and not very copious white strigulae ; costal margin white, mostly as far as C,
at base only very narrowly; lines white, very slender; antemedian weak,
very oblique outwards, anteriorly obsolete; postmedian straightish ; fringe
white. Hind wing with termen only very weakly bent in middle; M! stalked ;
concolorous with fore wing; a white cell-spot, slightly angular on base of R?,
slenderly outlined with brown-red ; postmedian line of the normal Thalassodes
form, but very weak posteriorly ; fringe white.

Under side much paler, not strigulated; the white markings showing
faintly ; fringes white.

Upolu: Malololelei, 2,000 feet, type 3, 30.vi.1924 ; allotype 9, 21.xi.1924.

Savai: Fagamalo, 1 9, 1925.

A fine species, not manifestly closely allied to any other. Thalassodes
leucoceraea Prout, 1925 (Malay Peninsula), and 7. floccosa Prout, 1917 (Malay
Peninsula and Java), share with it the white cell-spot of the hind wing, but
otherwise their shape, the non-separation of M! of the fore wing and the different
scheme of markings suggest that the present species has been independently
evolved. The whole genus, even including its few African representatives, is
very homogeneous.

This species will be figured, in colour, on a plate which will appear in
Part ITI, Fascicle 4.

2. Thalassodes chloropis Meyrick.

Thalassodes chloropis Meyrick, Trans. Ent. Soc. Lond., 1886, p. 204, 1886 (Fiji).
2“ Thalassodes veraria Guenée ”’ (Bethune-Baker, Proc. Zool. Soc. Lond., 1905, p. 94, 1905 (Fiji)).

Upolu: Malololelei, 1 ¢, 22.11.1924, 1 9, 22.11.1924, 3 O9, 24.11.1924, 3 Q9,
21.1v.1924, 2,000 feet; Apia, Vailima, 1 9, 14.viil.1922 (Armstrong); Apia,

120 INSECTS OF SAMOA.

1 9, 30.vi.1924, 1 9, 2.ix.1924, 1 2, 30.x.1924, 1 Q, 18.vill.1925, 1 9, x.1925,
1,000 feet, 1 9, undated (Woodford), in coll. Mus. Tring.

Tutuila: Pago Pago, 3 99, 1.1924, 2 29, x.1923 (Steffany).

Apart from Fiji, 7. chloropis was previously known to have been found in
Tonga (Friendly Islands). The 7’. “ chloropis”’ recorded from the Jaintia Hills
by Hampson has of course nothing to do with the present species, but is that
which Swinhoe later (1902) named 7’. curiosa. The Samoan specimens, or at
least those from Tutuila, may differ racially from the name-typical in having
the postmedian line rather broader ; but few good specimens (and none such
from Fiji) are yet known to me. A @ from the Vavau Group, Tonga, in Mus.
Tring, more nearly resembles Th. timoclea Druce (vide infra) in colour and macu-
lation. The genus Thalassodes in general, and the present section in particular,
will repay very careful study and analysis when sufficient material can be brought
together.

3. Thalassodes timoclea Druce.
Proc. Zool. Soc. Lond., 1888, p. 227, t. xiii, f. 6, 7, 1888 (Fiji).

Upolu: Apia, 1 9, 27.vi.1924.

The distinctive characters of this species are somewhat problematical, since
the band on the 2 described by Druce, upon which he largely relied, was evi-
dently the result of some injury, possibly in the pupal state. The 3, however,
which was first described and must certainly be treated as the holotype, differs
essentially from the other and larger green-faced Fijian Thalassodes (T. chloropis
Meyrick) in having the third segment of the palpus shorter. Its hind legs are _
lost, but there is a small abdominal spine. I refer provisionally to 7. tumoclea,
the present 2, which is not in very good condition but apparently agrees closely
with Druce’s types. Apart from the palpus, it differs from the Samoan forms
of T. chloropis in the slightly brighter (less blue) green colouring, and the
extremely fine postmedian line, which moreover is perhaps rather more proximally
placed than is usual in 7. chloropis ; the white dorsal line on the abdomen is
distinct.

ie Thalassodes, sp.

Upolu: Malololelei, 1 3, 1.vii.1924 (Armstrong).
The specimen is hopelessly discoloured: the terminal segment of the
palpus certainly shorter than in 7. chloropis and apparently even than in 7.

GEOMETRIDAE. 121

tumoclea, the hind tibial process vestigial (in 7. chloropis well developed), M}
of the fore wing just stalked (in 7’. timoclea and generally in T. chloropis separate).
The face appears to have been green. The wing-length slightly exceeds that of
T. chloropis, the apex of the fore wing is minutely produced, its termen almost
straight, moderately oblique, and the bend in the mid-termen of the hind wing
is weak.
5. Thalassodes pilaria Guenée.
Lhalassodes pilaria Guenée, Spec. Gén. Lép., ix, 361, t. xv, f. 2, 1858 (Tahiti).
“ Thalassodes quadraria Guenée,” ‘Rebel, Denkschr. K. Akad. Wiss. Wien, Math.-Naturw. Kl.,
Ixxxv, 428, 1910 (Samoa).

Upolu: Apia, 1 9, 13.1x.1923 (Swezey and Wilder) ; 1 9, 14.1x.1923 (Swezey
and Wilder); 2 33, 2 29, undated (Woodford) ; 1 3, 1 9 in Mus. Tring, 2.x1.1924,
1 9, vi.1925 ; 1 9 (very worn).

Tutuila: Pago Pago, 1 9, 1.1924 (Steffany), 1 ¢.

Manua: Tau, 3 gd, 20, 23.11.1926 (Judd); Ofu, 2 gg, 2 29, 27.11.1926
(Judd).

This species differs from 7. chloropis in the face being red-brown (the face
in both species is unfortunately very liable to injury), the third segment of the
palpus and the terminal process of ¢ hind tibia rather shorter (the process in
the g of T. chloropis being nearly half as long as the first tarsal segment) and
in having M! of the fore wing stalked with R?. Under the name 7. quadraria,
Rebel writes: “ Upolu, 3, of which 1 g¢ Vaimea 7 June and 1 9 Malifa 10 June
quite agree ; face red-brown, quite as in Indian.” It is perhaps legitimate to
conjecture that the third example, not mentioned in greater detail, belonged
to T. chloropis.

T. pilaria is now known to occur in Fiji, Samoa, the Society Islands, and
Pitcairn Island. On the first-named islands, according to Veitch and Green-
wood (Proc. Linn. Soc. N. S. Wales, xlvi, 507, 1921, xlix, 154, 1924), the larva
has been found on Rosa, sp., Ricinus communis, Inocarpus edulis, Mangifera
mdica, and Eugona indularis.

Pyrrhorachis Warren.

Nov. Zool., iti, 292, 1896.—Prout, Gen. Ins., 129, p. 238, 1912; Lep. Cat., 14, p. 162, 1913.—
Turner, Proc. Linn. Soc. N.S. Wales, xxxv, 569, 1910 (Pyrrhorhachis, ex err.).

This compact little genus comprises only a few very closely related species ;
certain others of doubtful affinity, provisionally referred to it, would be better

122 INSECTS OF SAMOA.

excluded. Thus P. albifimbria Warren (1896) and P. marginata Warren (1899)
should undoubtedly be assigned to Comostolopsis; P. rubripunctata Warren
(1909) would be better placed in Chloéres, while the actual affinities of P. caerulea
Warren (1893) and P. cosmetocraspeda Prout (1916) are not yet quite clear.
There remain the following species: P. deliciosa Warren (1896), founded on a
single example from the Natuna Islands and probably a race of the species
next in order, ¢.e. P. pyrrhogona Walker (1866), of 8. India and Ceylon, with
subspecies turgescens Prout (1917.—Khasi Hills); P. marginata T. P. Lucas
(1888), of Queensland and the Tenimber Is.; P. augustata Prout (1917.—
Loyalty Is.) ; P. cornuta Warren (1896), distributed from the Malay Peninsula
and the Philippine Is. to the Louisiade Archipelago and the Solomon Is.; P.
viridula Warren (1903.—British and Dutch New Guinea); P. rhodometopa
Prout (1913.—Mt. Goliath, Central Dutch New Guinea), possibly a form of the
following; P. ruficeps Warren (1906.—British and Dutch New Guinea).

The species now described extends the range of the genus eastward into
Polynesia. It belongs to the group containing P. ruficeps and P. rhodometopa
—green, with brighter red borders and crown—which is connected by P. viridula
with the two bluer species, P. pyrrhogona and P. cornuta.

6. Pyrrhorachis rhodoselas, sp. n.

dg, 19 mm. Near P. ruficeps Warren (1906) and P. rhodometopa Prout
(1913). Body, as in the former, with the red dorsal streak not interrupted with
ereen at base of abdomen; its colour still brighter, without pale central dash
on thorax posteriorly, but with an imdistinct dark dot in the centre of each
segment.

Fore wing with costal margin brighter red than in the allied species, and
less densely dark-irrorated ; red distal border intermediate in width, but
widening appreciably at tornus, and near its proximal boundary marked through-
out with small, deep violet-grey interneural spots ; a crescentic yellow cell-mark,
scarcely noticeable with the naked eye. Hind wing similar, except costally.

Fore wing beneath with costal margin pale pink to whitish, at extreme
edge proximally bright rosy, wanting the black admixture seen in the allied
species.

Upolu: Apia, type g, xii.1924.

This specimen will be figured, in colour, on a plate which will be published
with Part III, Fascicle 4.

GEOMETRIDAE. 123
STERRHINAE.

This subfamily is xepresented by four genera comprising seven species,
six of which are included in the three largest and most nearly cosmopolitan
genera—Amnisodes, Scopula, and Sterrha, while the remaining species is a very
interesting Symmacra. These are the only four Sterrhine genera yet known to
extend into the Pacific eastward of the Solomons ; but it is not impossible that
representatives of Organopoda, Ptochophyle, Chrysocraspeda, Problepsis, or
Bytharia, all of which reach that group or the Bismarck or Louisiade Archi-
pelagos, may yet be discovered in Polynesia.

Symmacra Warren.
Nov. Zool., iii, 116, 1896.

I refer to this genus only four forms, and even so it embraces two distinct
sections.

A. (Sect. typ.). Hind tibia in g with strong hair-pencil and two very
unequal spurs ; facies of Discoglypha: S. reqularis Warren, 1896, Khasi Hills ;
S. mobtrusa (Warren), 1897, S. Celebes (pr. subsp. 2).

B. Hind femur of ¢ also hairy, a strong tuft at its base, the tibia usually
with only a single spur; facies of Dithecodes: S. solidaria Guenée, Spec. Gén.
Lép., ix, 348, 1858, Ceylon (nec auctt. al.) = 8. quadraequata Walker, Last Lep.
Ins., xxii, 762, 1861, Borneo = S. validaria Walker, List Lep. Ins., xxxv,
1607, 1866, Celebes; S. solidaria baptata Warren, Nov. Zool., iv, 224, 1897,
Samoa.

‘ Symmacra”’ inconspicua Warren (1899), is not strictly congeneric, but
belongs rather to Dithecodes.

7. Symmacra solidaria baptata (Warren).
Sterrha (?) baptata Warren, Nov. Zool., iv, 224, 1897 (Apia).

Upolu: Malololelei, 2,000 feet, 1 9, 24.11.1924 ; 1 9, 21.iv.1925.

Tutuila: Pago Pago, 3 99, 1.1924; 1 gf, 3 99, 11.1924 (Steffany).

The author of this species, of which the only specimen hitherto known was
the type Q, failed to detect its affinities. Except in the glistening white, not
green, colouring, this form differs little from the very widely distributed S.

124 INSECTS OF SAMOA.

solidaria Guenée, and now that the 3 has been met with it seems safe to treat
it as a race. Both are flushed with reddish on the under side, but in typical
S. solidaria this suffusion is widely spread and becomes very strong at the
costa of the fore wing, while in the race baptata, especially in the Q, it is slighter,
chiefly developed in and distally to the cell of the fore wing. On the hind
wing the clear white, slightly raised cell-mark extends for the length of DC?:3 ;
in typical S. solidaria it is reduced and more punctiform.

S. s. solidaria is known to me as occurring in Ceylon, Assam, Burma, the
Malay Peninsula, W. China, Borneo, Celebes, Java to Lombok, Timor, Wetter L.,
Buru, Kei Is., New Guinea, Queensland, the D’Entrecasteaux Is., Woodlark LI.,
the Louisiades and Bismarck Archipelagos and Guadalcanar I. It was long
known as S. validaria Walker, Walker’s older name quadraequata having been
overlooked, and Guenée’s S. solidaria entirely misidentified. Oberthiir’s figure
of Guenée’s type (Ht. Lép. Comp., xii, fig. 3215) leaves no doubt about the
determination, notwithstanding the poor condition of the specimen and its
entirely erroneous generic location as a “ Nemoria.”

Anisodes Guenée.
Spec. Gén. Lép., ix, 415, 1858.—Hampson, Faun. Brit. Ind., Moths, iii, 446, 1895.

Except in the Holarctic Region, where it is replaced by the closely allied
Cosymbia Hiibner, a few islands such as Hawaii and New Zealand, and the
southern part of South America, this genus is universally distributed, though
less prevalent in Africa than in the Indo-Australian and Neotropical Regions.
It is extraordinarily interesting on account of the very great diversity of 3
secondary sexual characters, on which have been based numerous “ genera,”
better treated for the present as subgenera, though some remarkable divergences
in the genitalia suggest that there may be some heterogeneous elements which
it will ultimately be necessary to exclude. The 99 are on the whole very stable
in structure, and the principal venational variations—presence or absence of
areole and approximation or wide separation of M! of the hind wing—do not
appear to be of generic value. Two species have already been recorded from the
Samoan Islands and a third is represented in the present collection, though
unfortunately only by a single 9. A very probable addition is A. (Brachycola)
decolorata Warren (Nov. Zool., iv, 215, 1897), described from a specimen from
Lifu, Loyalty Islands, though a closely similar form is known to occur in Tahiti.

GEOMETRIDAE. 125

The subgenus Brachycola Warren (= “‘ Pertzera Meyrick”? Hampson, Faun.
Brit. Ind., Moths, 11, 446, 1895, ex err.), typified by A. absconditaria Walker,
of India and Malaya, consists of some six or seven Indo-Australian species, and
is characterised by having the g hind tibia extremely short, with a tuft of red
hair and 3 long, curved spurs. The subgenus Pisoraca Walker (distribution
nearly coterminous with that of the genus) has the 3 hind tibia of normal length
but with 3 spurs, and is represented in Fiji by one recently discovered species
near—perhaps a race of—A. compacta Warren, 1898, a species of which the type
was obtained in the Kei Is., though it has since been found to occur throughout
a wide area in New Guinea, as also in the Louisiade and Bismarck Archipelagos.
Another member of the genus, A. (Perizera) prionodes Meyrick (Tr. Ent. Soc.
Lond., 1886, p. 209, 1886), was described from a single g from Fiji in the col-
lection of Dr. T. P. Lucas and may be a form of the variable A. pauper Butler
(Ann. Mag. Nat. Hist., (5), xx, 245, 1887) (Solomon Is.) = A. syntona Meyrick
(Tr. Ent. Soc. Lond., 1889, p. 487, 1889) (Port Moresby). In the subgenus Perixera
Meyrick (= Phrissosceles Warren) it is the hind femur of the g which bears a
red tuft, while the tibia is normally formed, with a pair of terminal spurs. A
typical (Fijian) Anisodes, without appreciable secondary sexual characters
and with only terminal spurs on the hind tibia, is likewise a solitary 3 of a race
or close ally of a widely distributed species (New Guinea and the islands east-
ward, as far as the Solomons) which I believe to be A. praetermissa Bastelberger
(Int. Ent. Zeit. Guben, ti, 38, 1908). I have not yet been able to examine the
type of the latter—a single specimen, apparently in poor condition. In
addition, some fifteen other Anisodes are known to me as occurring in the
Solomon Is.: one in the subgenus Zeugma Walker = Dizuga Warren (3 hind
tibia with 4 spurs), one in Brachycola, 7 in typical Anisodes, one in Crypsiplocia
Warren (jf with costal fold beneath), two in Plochucha Warren (¢ with costal
fold above), one in Stibarostoma (vide infra), and two in Perizera. The genus
probably numbers nearly three hundred species, of which slightly over one-half
are Neotropical.

8. Anisodes (Xenoprora) samoana (Warren).
Brachycola samoana Warren, Nov. Zool., iv, 216, 1897 (Apia, Upolu).

“Samoa,” 1 g, iil.—vili.1921 (O’Connor) ; “ Pago,” 1 9, v.1896 (P. de la
Garde) ; both in Mus. Brit. ? Upolu: Malololelei, 2 99, 24.ii., 6.vii.1924.

126 INSECTS OF SAMOA.

Warren founded this species on a single 2 in Mus. Tring, collected by
Woodford. In my manuscripts I had, with a considerable degree of confidence,
treated as identical with it A. parallela Warren, described ten pages later from
a specimen from Lifu, Loyalty Is., and made the type.of the genus (subgenus,
vide supra) Xenoprora. This is characterised by the peculiar ¢ palpus, which
is short, thick, upcurved and heavily rough-scaled above. In doing as I did
I was not far wrong, seeing that the two are unmistakably subspecies or repre-
sentative species, with identical palpi, even if they do not ultimately prove to
be absolute synonyms. But my experience in the matter of Cleora samoana and
its group (vide infra) has made me hesitant with regard to premature unions ;
and I notice that the only known ¢ of A. samoana, besides appearing slightly
narrower-winged and more yellowish than the rest of the Xenoprora 33 thus
far known, shows on the fore wing beneath, at and in front of the end of the
cell, a roundish patch of dense, somewhat specialised scaling, of which I cannot
discover a trace in any other example. None of the material, however, is in
really fresh condition and I would not unduly stress the distinction. The 2 9°
from Malololelei are of a larger, more yellowish-tinged, superficially very
different-looking form from Warren’s type 9, bat I am inclined to think that
they belong to the same species.

Besides the Samoan material detailed above, and the 4 gg and 3 99 of A.
parallela in the Tring Museum, examples of the same species or superspecies are
known from Fiji and the Society Islands. From the former group, the Joicey
collection contains a ¢ from Viti Levu (Woodford), ex. coll. Druce, of fairly
large size, fleshy-tinged rather than yellowish, with the median shade of the
fore wing rather strong, and in general agreeing with A. parallela rather than
with A. samoana. <A second 3, on which I made no special notes, has passed
through my hands, having been submitted to the Imperial Bureau of Entomology
for determination. From the latter group (Tahiti), the British Museum has
2 33, extremely worn, which perhaps represent another race, since both have
a more consolidated and darkened (black-brown) cell-mark on the hind
wing.

No other species of Xenoprora is known, but the subgenus is clearly the
progenitor of Hmmesura Warren (India to Celebes), which has the same ¢ palpus
but contorted 3 venation. Stibarostoma Warren (India to the Solomon Is. and.
(Jueensland) has also its 3 specialisation on the palpus, but in that group the
second segment is long. In all three subgenera the areole is obsolete.

GEOMETRIDAE. 127

9. Anisodes (Perixera) obliviaria Walker.

Anisodes obliviaria Walker, List Lep. Ins., xxi, 645, 1861 (Ceylon).

Anisodes suspicaria Snellen, Tijd. Ent., xxiv, 80, t. vii, f. 6-6 c, 1881 (Celebes).

“ Anisodes obrinaria Guenée,”’ Moore, Lep. Ceyl., iii, 446, t. cic, f. 4 a, 1887 (ez err.).
Perixera rufidorsata Warren, Nov. Zool., iii, 312, 1896 ; viii, 24, 1901 (N. India) (ab.).
Perizera rufannularia Warren, Nov. Zool., iv, 221, 1897 (Upolu: Apia).

Tutuila: Pago Pago, 1 gy, 1.1924, 1 9, x.1923 (Steffany).

I cannot yet see that this very widely distributed species even forms
definable races. The extreme ab. rufidorsata (with enlarged, broadly dark-
ringed cell-spot on the hind wing) is only known to me as occurring in N. India
(not infrequently), Java and New Guinea, but nowhere constant ; intermediates
are pretty general. I have seen specimens of A. obliviaria from Ceylon, India,
the Malay Peninsula, Java, Timor, Babar, the Obi Is., Amboina, New Guinea,
the Trobriand Is., Woodlark I., the Louisiade Archipelago, Queensland, the
Bismarck Archipelago, the Solomon Is. and Samoa. Tring Museum has 1 2
from Apia as well as the type 9 P. rufannularia.

10. Anisodes hypocris, sp. n.

°, 33 mm. Head and body concolorous with wings; the face, upper edge
of palpus and part of the fore leg (especially of the coxa) redder. Palpus long
(about 24), the third segment longer than the second.

Fore wing with termen crenulate ; areole wanting, R? arising before middle
of DC, DC? incurved anteriorly ; light cinnamon, weakly irrorated with greyish,
leaving a clearer postmedian band about 2 mm. in width; markings rather
blurred, greyish; cell-spot ocellate, though rather small; antemedian line
very obscure, apparently thick or double ; median rather thick, considerably
beyond ceil-spot, outbent about R*—M!, slightly incurved before and behind ;
postmedian 2-3 mm. from termen, accentuated by minute darker vein-dots,
incurved between the radials and again posteriorly ; subterminal indistinct,
between vague, somewhat macular dark shades; terminal dots not strong.
Hind wing with termen strongly and irregularly crenulate, the longest tooth
at R°, that at R! also rather pronounced ; cell-spot large, white ; median shade
rather close beyond it ; distal markings nearly as on fore wing.

Under side considerably paler; cell-marks indicated, on hind wing white,
though less sharply than above; postmedian dots also indicated, the other
markings faint or obsolete.

128 INSECTS OF SAMOA.

Upolu: Malololelei, 1 9, 21.vi.1924.

Without reference to the 3, it is not possible to give the subgeneric position,
but I conjecture that the present species will prove to be a Perizera. The
palpus and scheme of markings agree with those of the two most widely distri-
buted members of that group: P. flavispila Warren, 1896 (N. India, Burma,
Hainan and Hong Kong, with races (?) in Sambawa and Sumba and in Queens-
land, New Guinea and Fergusson Island, the latter being P. lophosceles Turner,
1908) ; and P. argyromma Warren, 1896 (distributed from Ceylon to New Britain).
It may even prove a close relative of the latter, notwithstanding its more
crenulate margins.

This species will be figured, in colour, on a plate which will appear in
Part IIT, Fascicle 4.

Scopula Schrank.

Fauna Boica, ii. (2), 162, 1802.—Acidalia Treitschke, 1825 (nom. praeoce.).—Leptomeris Hiibner,
1826.—<Arrhostia Hubner, 1826.—Craspedia Hiibner, 1826.—Dosithea Duponchel, 1829.—
Pylarge Herrich-Schafier, 1856.—Phyletis Guenée, 1858.—Lycauges Butler, 1879.—Tricho-
clada Meyrick, 1886.—Runeca Moore, 1888.—Longula Staudinger, 1892.—Cnidia Gumppen-
berg, 1892.—Synelys Hulst, 1896.—Induna Warren, 1897.—Acidalina Staudinger, 1897.—
Pleionocentra Warren, 1898.—Chlorocraspedia Warren, 1899.—Lipocentris Warren, 1905.

An enormous genus, showing remarkably little structural variation except
in secondary sexual characters in the g. The 3 genitalia, whether in the
Palaearctic or Nearctic Region, South America, South Africa, Malaya or
Tasmania, are quite homogeneous in their essential formation, and all that is
yet known of the exotic larvae fully bears out the naturalness of the grouping.
Wanting in Chili, Patagonia, New Zealand (excepting one obviously introduced
species (vide Meyrick, Tr. N. Z. Inst., xvi, 57, 1883) and the Hawaiian Islands,
the genus is otherwise more or less strongly represented everywhere, perhaps
least prominent in the Neotropical Region. Including a few which are still in
manuscript, | know about 470 named species, besides some 50 others which have
not yet been worked out.

The species upon which Meyrick founded his genus Trichoclada, and which
perhaps in some measure connects Scopula with the allied genus Problepsis
Lederer, is only known as occurring in Fiji: epigypsa Meyrick, Tr. Ent. Soc.
Lond., 1886, p. 208 (Trichoclada) = nivipennis Butler, Tr. Ent. Soc. Lond., 1886,
p. 436 (I[daea) = cernea Druce, Proc. Zool. Soc. Lond., 1888, p. 227, t. xiii, f. 8
(Acidalia). Two other species, belonging to the central group of Scopula,

GEOMETRIDAE. 129

inhabit the Loyalty Is., namely: S. wndilinea Warren, 1900, only known from
a specimen from Lifu, though Swinhoe (Tr. Ent. Soc. Lond., 1902, p. 657) quite
erroneously treated it as identical with S. cretata Warren (7.e. S. destituta Walker,
1866, Sula Is.); and S. oppilata Walker, 1861 = S. crossophragma Meyrick,
1886, which is known also to occur in Toekan Besi, Celebes, from Bali to the
Tenimber Is., in Queensland, New South Wales, Kei Is., Amboina, New Guinea,
the Bismarck Archipelago and the Solomon Is. The species are mostly small
and not particularly attractive, and in all probability several additional ones
will be found in the Pacific.

11. Scopula homodoxa (Meyrick).

Acidalia homodoxa Meyrick, Tr. Ent. Soc. Lond., 1886, p. 208, 1886 (Tonga).
% “ Craspedia remotota [sic] Guenée,”’ Rebel, Denkschr. K. Akad. Wiss. Wien, Math.-Naturw. K1.,

Ixxxv, 429, 1910 (Samoa).

“Samoa,” 1 9, li.—vii., 1921 (O'Connor); Upolu: Apia, 1 9, 29.v.1924,
2 99, 8.vi.1924, 1 9, 24.x.1924.

Also 2 99 from Tonga (Vavau, Neiafu), 111.1925.

Meyrick apparently confused two or three species, and, since he did not,
in his earlier work, select a holotype, there has been some difficulty in deciding
upon the application of his name. The author gives Port Moresby, Tonga, and
Fiji as localities, but, as his New Guinea ¢ has the hind tarsus } or less and
deviates in some other details, I have made the Tonga J the holotype and believe
that the species does not extend westward of the Loyalty Is.; see my notes
under Scopula pseudodoxa Prout, Nov. Zool., xxvii, 297, 1920. The three
specimens in the British Museum from the Solomon Is. (Alu), referred by Mr.
Meyrick (Tr. Ent. Soc. Lond., 1889, p. 485, 1889) to S. homodoxa, are rubbed
examples of S. oppilata Walker, with the very characteristic fringes worn off.

12. Scopula, sp.
Craspedia, sp. Rebel, Denkschr. K. Akad. Wiss. Wien, Math.-Naturw. K1., |xxxv, 429, 1910.

Upolu: Malifa, vi, two specimens, very worn, recorded by Rebel. <‘ Glossy
silky-white, one (a ¢) 11 mm., showing traces of red scaling before the
vertex.”

130 INSECTS OF SAMOA.

Sterrha Hiibuer.

Verz. bek. Schmett., 309, 1826.—Hyria Stephens, 1831 (nom. praeocc.).—Ptychopoda Stephens,
1831.—Ania Stephens, 1831.—Pythodora Meyrick, 1886.—Janarda Moore, 1888.—Mnesterodes
Meyrick, 1889.—Xenocentris Meyrick, 1889.—Argia Gumppenberg, 1890.—Andragrupos
Hampson, 1891.—Lophophleps Hampson, 1891.—Carphozera Riley, 1891.—Pelagia Gumppen-
berg, 1892.—Aphrogeneca Gumppenberg, 1892.—Synomila Hulst, 1896.—Ptenopoda Hulst,
1896.—Lophosis Hulst, 1896.—Hemipogon Warren, 1897.—Leptacme Warren, 1897.—
Strophoptila Warren, 1897.—Brachyprota Warren, 1897.—Polygraphodes Warren, 1897.—
Cacorista Warren, 1899.—Anteots Warren, 1900.—Hyriogona Warren, 1900.—Neochrysa
Warren, 1900.—Pogonogya Warren, 1900.—Prospasta Warren, 1900.—Thysanotricha Warren,
1903.—Deinopygia Warren, 1904.—Lobura Warren, 1906.—Omopera Warren, 1906.—Argy-
roscelia Warren, 1907.—Pareupithex Warren, 1907.—Hurthestes Dognin, 1914.

Another enormous genus, with a similar distribution to that of Scopula.
Less strictly homogeneous, showing still wider divergences in shape—sometimes
in the § only, sometimes in both sexes—and even in the details of the venation,
while the wealth of secondary sexual characters, chiefly in the development and
position of hair-tufts or specialised scaling on the wings or legs of the 3, has
given authors an opportunity for the erection of a large number of the “ genera ”’
cited in the synonymy above. It is noteworthy that the 22 of some of the
very divergent forms are so closely similar inter se, not only in structure but in
marking, as to be extremely difficult to distinguish.

Speaking generally, the simpler forms belong chiefly to the Palaearctic
Region, in particular to the Mediterranean countries. Of the 515 named species
known to me (to which must be added, as in the case of Scopula, some
50 unworked), 42 belong to the typical, almost exclusively Palaearctic, subgenus
Sterrha, with a pair of spurs on the ¢ hind tibia ; about 240, largely Palaearctic,
to the subgenus Ptychopoda, with the hind tibial spurs of the § wanting; while
the rest have been formed into a number of small and partly somewhat artificial
groups, which are almost exclusively exotic. Andragrupos, Hemipogon, Stro-
phoptila, Thysanotricha, Lophophleps, and Mnesterodes belong to the Indo-
Australian Region; Cacorista to Africa; Lophosis and Ptenopoda to North
America; Hyriogona, Deimopygia, Prospasta, Pareupithex, Lobura, Neochrysa,
Brachyprota, Hurthestes, Omopera, and Pogonogya account for most (about 67)
of the Neotropical species ; Xenocentris is distributed, except in Africa, but is
predominantly Indo-Australian.

The section Ptychopoda, not hitherto known eastward of the Solomon Is.,
is now represented in Samoa by S. rufula Warren. I know of no particularly

GEOMETRIDAE. 131

likely addition, since the rest of the Indo-Australian representatives of it are
mostly Indian or Australian; a very few species are Malayan, and only two
or three are known to occur in New Guinea and its satellite islands. Specimens
of a member of the section Hemipogon—H. sumplex Warren, 1899 (Louisiade
Archipelago)—have recently been received from Dampier I. and New Hanover ;
this species has very close relatives in H. liwidula Warren, 1903, of New Guinea,
and H. liparota Turner, 1908, of Queensland. The type of Warren’s Thysano-
tricha—T. ziczacata Warren, 1903—is from the Solomon Is. (Isabel I.). That
of Meyrick’s Mnesterodes—M. trypheropa Meyrick (New Guinea), = P. angusti-
perms Warren, 1897 (N. Queensland), = A. ? floccosa Pagenstecher, 1900 (New
Britain)—reaches in one direction to Timor, and I am under the impression that
I have seen specimens of it from Fiji; it would not be surprising if it proved
to occur in the Samoan group. The somewhat isolated species S. rhupistis, on
which Meyrick founded his genus Pythodora, is at present only known to occur
in Fiji.
13. Sterrha rufula (Warren).

Eois (%) rufula Warren, Nov. Zool., vi, 334, 1899 (Solomon Is.: “ Tugela”’ [recte Tulagi, off

Isabel I.]).

Upolu: Malololelei, 1 9, vii.1925 (Wilder); 1 9, 2,000 feet, 24.11.1924
(P. A. Buxton) ; 1 9, 23.ix.1924; 1 9, 21.v.1925; Apia, 1 9, 13.1x.1923 (Swezey
and Wilder).

Tutuila: Pago Pago, 1 9, 1., 2 29, 11.1924 (Steffany).

Also 1 2 from 8. New Hebrides (Tanna), 1x.1925,.

Range: % Moluccas (Buru); Kei Is.; British New Guinea; Louisiade
Archipelago ; St. Matthias; Bismarck Archipelago (New Hanover) ; Solomon
Is. Probably S. (L.) untformis Warren, 1896 (Queensland), and S. (L.)
perspersata Warren, 1897 (Timor) will also prove to be forms of this insect, in
which case the oldest name for the collective species will be S. unzformas.

Apparently rather variable. The Samoan form seems rather larger and
paler than the Solomons type, and more yellowish than rufous. The New
Hebrides example is larger still and sharply marked, but has more typical
colouring.

Since this species (besides S. uniformis and S. perspersata) was described
from the 9, which seems much the commoner sex, it is desirable to add a note
on the ¢ structure. Antennal ciliation of ¢ scarcely 1, hind tibia rather long,

132 INSECTS OF SAMOA.

strongly thickened distally, tarsus extremely short. No doubt related to the
croup of S. (A.) actiosaria Walker (Ceylon, also widely distributed in India,
Malaya, etc.), or may be likened to the larger S. (S.) carnearia Warren (Nov. Zool.,
iv, 63, 1897) (2, Labuan) = S. (P.) carneola Warren (Nov. Zool., iv, 223, 1897)
(g and 2 Labuan [loc. typ.] and Penang), which, however, further differs in
having the termen of the 3 hind wing ventricose.

LARENTINAE.

A very large subfamily, but chiefly characteristic of temperate and even
subarctic climates, the few tropical members of its more typical genera (Larentia,
Xanthorhoé, Huphyia, etc.) occurring almost exclusively at high altitudes. The
nine genera here recorded from the Samoan Is. are for the most part such as
would be expected—Colliz, Eupithecia (in a specialised form), Chloroclystis
and Gymnoscelis with their offshoots, and Sauris. The most doubtful as regards
its exact generic location is the new species which I assign to Asthena.

“ Larentia” finitima Walker (List Lep. Ins., xxiv, 1203, 1862), founded on
4 2° from the Isle of Pines (Kuni) and provisionally referable to Xanthorhoé,
sharing to some extent the shape and coloration of the larger Hawaiian X.
leucoxyla Meyrick ; Scotocyma albinotata legalis Warren (Nov. Zool., ui, 384,
1896), Loyalty Is.; Sterrhochaeta chlorodesma (Meyrick), (Tr. Ent. Soc. Lond.,
1886, p. 194)=Coremia picta Butler (Tr. Ent. Soc. Lond., 1886, p. 439), Fiji;
Colliz subligata Warren (Nov. Zool., iii, 385, 1896) (? C. ghosha Walker, subsp.),
Loyalty Is.; the unnamed species treated by Meyrick as the 2 of Gymnoscelis
erymna (vide imfra); Hybridoneura cristata (Warren) (vide infra, p. 145) ;
Sauris elaica Meyrick (Tr. Ent. Soc. Lond., 1886, p. 193), Fiji; a related
species found in the Loyalty Is., recently described as S. curvicosta Prout,
1928; Hots cymatodes (Meyrick) (Tr. Ent. Soc. Lond., 1886, p. 192), New
Hebrides—the foregoing appear to be the only Larentiine species yet known to
occur in the Pacific islands, apart from those which are here recorded from
Samoa.

Collix Guenée.
Spec. Gén. Lép., x, 357, 1858.

A very natural genus, comprising about twenty known species. Of these,

one belongs to Africa with Madagascar, and another to Réunion and Mauritius ;

GEOMETRIDAE. 153

the rest are entirely Indo-Australian, mostly inhabiting the islands. C. ghosha
Walker has a race (C. sticticata Warren, Nov. Zool., ix, 361, 1902) in the Solomon
Is.; C. elongata Warren (Nov. Zool., ix, 360, 1902), and C. purpurilita Prout
(Nov. Zool., xxxti, 41, 1925) were described from specimens from that group ;
the Bismarck and Louisiade Archipelagos also provide forms.

14. Collix lasiospila (Meyrick).
Cidaria lasiospila Meyrick, Tr. Ent. Soc. Lond., 1886, p. 194 (Fiji).

Tutuila: Pago Pago, 1 9, 12.viii.1925.

Rather larger than the only Fiji specimen I have seen (a 9 in the British
Museum) otherwise identical ; not so large as the ¢ type in coll. T. P. Lucas,
according to the measurements given by Meyrick. The determination is only
conjectural, since Meyrick does not describe the very distinctive under side,
on which the veins are broadly yellow; but I know no other similar Fijian
Larentid to which his description could be applied. Meyrick adds, however,
that the same species occurs in Australia, and that there may be some error
as regards the locality ; but, since there is nothing in his Australian “ Revision ”
(Proc. Linn. Soc. N. S. Wales, Series 2, Vol. V, pp. 791-879, 1891) to throw light
upon it I cannot pursue the matter further. Similar, though not identical,
species do occur in Australia.

Eupithecia Curtis.

Brit. Ent., 1, 64, 1825.—Tephroclystia Hiibner, Verz. bek. Schmett., 323, 1826; Meyrick, 7’r.
Ent. Soc. Lond., 1892, p. 65 (Lephroclystis, ex err.) ; Hulst, Tr. Amer. Ent. Soc., xxiii, 265,
1896 (Lephroclystis, ex err.); Staudinger and Rebel, Cat. Lep. Palaearct. (i), 308, 1901 ;
Warren et al., passim.

In its more typical forms Hupithecia, perhaps the most extensive Geometrid
genus in the world, does not concern us here. It occurs almost everywhere
m the Palaearctic, Nearctic, Neotropical and Ethiopian Regions, and has a
good many representatives in India and perhaps parts of Malaya. But in the
Moluccas and New Guinea it seems to be very sparsely represented or almost
absent, except for a few specialised offshoots which may claim generic separa-
tion. Among the latter is Mnesiloba Warren, with double areole, as in the
section Hucymatoge Hiibner, but having in addition a ¢ hind wing specialisation,

111 (3) 2

134 INSECTS OF SAMOA.

in that the tornal region is separated from the rest of the wing and more or less
distorted. Besides EL. ewpitheciata, there are two unnamed species, virtually
indistinguishable in the 2 but with different g modifications, both inhabiting
New Guinea and the D’Entrecasteaux Is.

15. Eupithecia (Mnesiloba) eupitheciata (Walker).

Phibalapteryx eupitheciata Walker, List Lep. Ins., xxvi, 1720, 1862 (Australia).

Bupithecia intentata Walker, List. Lep. Ins., xxxv, 1676, 1866 (Sarawak).

Cephalissa delogramma Meyrick, Tr. Ent. Soc. Lond., 1886, p. 195 (Fiji and Tonga).

Hupithecia dentifascia Hampson, Ill. Het., viti, 117, t. 152, f. 12, 1891 (Nilgiris)

Chloroclystis intentata (part.) Hampson, Faun. Brit. Ind., Moths, iii, 391, 1895 (Ceylon).

Eupithecia eupithectata Hampson, Faun. Brit. Ind., Moths, iii, 398, 1895 (Khasi Hills, Nilgiris,
Ceylon, Shan States, Australia, Fiji, Tonga); Prout, Journ. Bomb. Nat. Hist. Soc., xxxi,
319, 1926 (Upper Burma, Java, Sangir, Celebes, New Guinea).

Mnesiloba eupitheciata Warren, Nov. Zool., viii, 196, 1901; Turner, Proc. Roy. Soc. Vict., xvi,
237, 1904 (Brisbane).

Mnesiloba seminigra Bastelberger, Ent. Zeit., xix, 76, 1905 (Assam).

Upolu: Malololelei, 2 99, 22, 24.11.1924; 1 9, 2,000 feet, 24.vi.1924 ;
3 29, 2,000 feet, 21.iv.1925.

With the exception of the 9 taken 24.vi.1924, which is fully normal in
size, the Upolu examples are rather small. All have the hind wing more or less
pale and the subapical shading of the fore wing moderately bright, so that we
perhaps have to deal with a local race. But the species is everywhere extremely
variable, and so precisely similar is the range of forms exhibited by it that I
should hesitate to make any geographical separations in default of very plentiful
and very perfect material. The median area of the fore wing is particularly
meonstant, and may be darkened either: throughout; only in its posterior part
(ab. semmgra Bastelberger) ; in its anterior and distal part, forming a some-
what triangular shade, with its base on the costa and its apex in the angle of
the postmedian line about M!; at its proximal and distal edges; or it
may remain wholly whitish. Two of the Samoan specimens belong to the last-
named, the others approximating to the triangularly-marked form, though in
varying degree, in no case with the triangle well-defined proximally and
posteriorly, nor very completely dark-filled.

I have given the hitherto recorded range under the synonymy, but have
also received specimens from Peninsular Siam, Pahang and Sumatra.

GEOMETRIDAE. 135

Chloroclystis Hiibner.

Verz. bek. Schmett., 323, 1826.—Meyrick, Tr. Ent. Soc. Lond., 1892, p. 65.—Hampson, Faun.
Brit. Ind., Moths, iii, 390, 1895.—Turner, Proc. Roy. Soc. Vict., xvi, 228, 1904.—Dietze, Biol.
Eupith., 23, 1913.—Simotricha Warren, Nov. Zool., iv, 395, 1897.—Dyserga Petersen, Iris
xxii, 281, 1909.

Distinguishable from Hwpithecua by the subcostal venation of the fore
wing, SC! running into, or at least anastomosing with C. A large genus, though
not nearly so large as Eupithecia and characteristic chiefly of the Indo-Australian
Region. In the secondary sexual developments it almost rivals Anisodes.
Assuming that these should be treated as merely subgenera, there are perhaps
about two hundred known species ; of these only six are Palaearctic and about
forty Ethiopian. The few New-World species which have been placed here
do not belong to the genus with the exception of C. elacachroma Bastelberger
(Jahrb. Nass. Ver. Nat., lxi, 83, 1908 = EF. aquanivaria KE. D. Jones, Proc. Zool.
Soc. Lond., 1921, p. 343, t. i, f. 3), of Bolivia and Brazil, which forms a special
section of it.

The name-typical section, with but slightly modified scalmg on the 3
hind wing, contains only the European C. coronata Hiibner, with races in Japan
(C. luconda Butler) and N. W. India (unnamed); its immediate offshoots
(Sesquiptera Warren, Gymnopera Warren = ? Oligoclystia Bastelberger, Chloro-
plintha Warren and Aétheolepis Warren) with increased specialisation of that
wing, are Indo-Australian, but are not yet known to extend beyond New Guinea.
The group with simple ¢ wing-structure (Calliclystis Dietze) is distributed
throughout the greater part of the range of the genus, and is strongly repre-
sented in New Zealand by Pasiphila Meyrick (= Helastiodes Warren), which is
not always structurally distinguishable from it, though very commonly the 3
antenna bears long fascicles of cilia. To this group (Callclystis) belong, with
the possible exception of C. bosora, the few Polynesian Chloroclystis, all of which
are now known to occur in the Samoan Is. One of the groups with tufts on the
costal margin of the g (Bosara Walker) extends to the Solomon Is. in a form
which I suppose to be a race of C. dilatata Walker (Borneo) and of which C.
pelopsaria Walker (Sula) is probably another race. Thamnocausta Warren,
with specialised oval patches in the 3 on the hind wing above and on the fore
wing beneath, is now known as occurring in the Bismarck Archipelago, speci-
mens of C. malachitis Warren (1903) having recently been received at the Tring
Museum from New Britain. The rest of the sections need not be noticed here.

156 INSECTS OF SAMOA.

16. Chloroclystis bosora (Druce).
Larentia bosora Druce, Proc. Zool. Soc. Lond., 1888, p. 228, t. xiii, f. 10, 1888 (Fiji).

Upolu: Malololelei, 3 92, 24.11, 21, 24.v1.1924; Apia, 1 9, 20.v.1922
(Armstrong).

Druce founded his “ Larentia”’ bosora on a single @ and the only other
specimen that | have hitherto seen is another 9 from the same islands. The
Samoan examples are rather larger than the type, with rather less white above.
and beneath with the reddish flush rather less deep. The character of the
markings of the fore wing above, and especially the form of the palpus, which
has the second segment smooth-scaled and slightly down-curved, suggest that
this is an overgrown relative of C. adnuataria Walker, which is known in three
races or very close allies: C. admiataria Walker, 1862 (Ceylon; generally
distributed in India, the Malay Peninsula, N. Pagi Is., Celebes, and Sambawa) ;
C. fragiis Warren, 1899 (Louisiade Archipelago ; also in the Kei Is., British
New Guinea and perhaps on St. Matthias Is.); C. bryodes Turner, 1907 (N.
Queensland). The postmedian line of the hind wing, which is weaker and but
slightly angled, and the subterminal maculation of both wings are, however,
very distinct from the markings of C. admixtaria, and the 3, when discovered,
may show higher specialisation.

17. Chloroclystis mempta, sp. n.

32, 183-15 mm. Head, body, and wings predominantly drab, variable in
tone, the pale ground-colour being more or less tinged with olive, the dark
markings more cimnamon-brownish or fuscescent. Face darker above and
below than in middle ; no definite cone. Palpus 13 to 2, second segment rough-
scaled above and beneath; first and second segments beneath sprinkled with
fuscous. Vertex rather pale. Antennal scales at ends of segments rather dark,
shghtly projecting ; ciliation in 9 scarcely $. Abdomen, except third and fourth
somites, with unequal dark belts; body beneath pale, in places almost white :
2 robust. Hind tibia with the inner spurs long, especially the proximal one ;
the outer short.

Fore wing rather short ; apex moderately rounded ; SC! running into C :
markings rather variable ; basal and subbasal tending to form ill-defined narrow
bands; antemedian line at or slightly before 4, a little out-bent in middle;
postmedian from a rather strong dot at 3% costa, excurved in anterior part, then

GEOMETRIDAE. 137

at least as oblique as termen, scarcely incurved; median area traversed by
ill-defined lines or—in the J and 1 9—with more or less dark shading proximally
and distally, converging into a more definite dark spot from M? to hind margin ;
lines beyond more or less parallel, with the proximal subterminal fairly strong,
band-like, an indistinct pale longitudinal mark behind R3; a slight cloudy sub-
macular terminal band; terminal line little darker, interrupted by pale dots
at the veins. Hind wing with termen faintly sinuous between the radials, and
more markedly between M! and SM? ; concolorous with fore wing, but looking
slightly paler on account of a widening or accentuation of the pale band outside
the postmedian ; postmedian a little angled about R*; longitudinal pale spot
of outer area not or scarcely developed.

Under side paler, more glossy; the only noticeable markings a cloudy
postmedian, arising on the fore wing from a costal spot, and on the fore wing
an antemedian costal spot (generally weaker).

Upolu: Malololelei, type ¢ and the most similar 9 (allotype), 2,000 feet,
21.iv.1925; a more cinnamon-tinged 9, 1.v1.1924; a greyer, more regularly
Imed 9, slightly reminiscent of Gymnoscelis fasciata Hampson, 25.11.1924. This
species may be placed next to C. sordida Warren (1903), which is distributed
in New Guinea and its islands and I believe reaches the Solomon and
Bismarck groups, while it is probably only a subspecies of (identical with ?)
C. inops Warren (1898, erroneously as Gymnoscelis), of the Kei Is. and Amboina.
These latter, however, have both the proximal spurs of the hind tibia short,
while C. pauaxillula Turner (1907), of N. Queensland, likewise closely similar
in shape and facies, has the ¢ hind wing hairy beneath (sect. Dasymatia Warren).

A figure of this species, in colour, will appear in Part III, Fascicle 4.

18. Chloroclystis lepta (Meyrick).

Pasiphila lepta Meyrick, Tr. Ent. Soc. Lond., 1886, p. 191, 1886 (Tonga and Marshall Islands).
¢ Pasiphilodes lepta Warren, Nov. Zool., 11, 110, 1895.

Tutuila: Pago Pago, 1 9, 1.1924 (Steffany).

Apparently variable, but only four specimens, all from different islands,
are yet known to me, so that it is not at present possible to form an opinion
on the nature of the variation. A @ in the British Museum, from Moorea,
Society Is., is strongly marked above and agrees pretty closely with the present
example. Meyrick’s original from the Marshall Is. is a little paler, that from

138 INSECTS OF SAMOA.

Tonga paler still. The under side is always weakly marked. The originals
were both marked by their author “cotype,” but Sir George Hampson has
made the ¢, from the Marshall Is., the holotype.

Micrulia Warren.
Nov. Zool., iti, 396, 1896.

This genus differs from the two previous genera in that the hind tibia in
both sexes has only three spurs, the proximal one very long. The characters
upon which Warren bases his genus are: hind wing of the ¢ triangular, with
termen nearly straight, abdominal margin puckered above and bearing beneath,
from anal angle to M, tufts of thick hair. He adds that it is otherwise as
Chloroclystis ; but, besides the tibial character, it differs in having SC! of the
fore wing parallel with C and closely approximated, but not anastomosing.
Only the type species entirely agrees with this characterisation, but I consider
Opistheploce Warren, with termen of hind wing in $ “ convolute and folded over
above,” to be merely a section of Micrulia, and would refer to Opistheploce the
following: M. cinerea Warren (1896) (Batjan, only the type known); M. rufula
Warren (1899) (praec. form. ?) (Milne Bay ; also found from Dutch New Guinea
to Choiseul) ; AZ. ewrotosoma Hampson (1903) (sequ. 2 2) (Ceylon) ; M. medio-
plaga Swinhoe (1902) (Borneo ; also occurs in Ceylon).

19. Micrulia tenuilinea Warren.

Micrulia tenuilinea Warren, Nov. Zool., iii, 391, 1896 (Khasi Hills).
Megatheca dentosa Warren, Nov. Zool., viii, 31, 1901 (Queensland).
Gymnoscelis lobata 9 Hampson, Journ. Bomb. Nat. Hist. Soc., xxi, 1246, 1912 (ex err., nec 3}

(Ceylon).

Upolu: Malololelei, 2,000 feet, 1 3, 25.11.1924.

Like several of the small and obscure Eupitheciids, this species must have
an enormously wide distribution, but has been much overlooked. I had seen
specimens of it from Ceylon, Assam, the Malay Peninsula and Pulo Laut and
had confidently labelled the ‘ Queensland” example in Tring Museum, with
the facies and setting of the specimens in the same collection from the Khasi
Hills, as “err. loc. (probably Khasis).” In view, however, of the astonishing
discovery of it on Upolu, I must revise that judgment. Turner (Proc. Roy. Soc.
Vict., xvi, 281-2, 1904) does not know Australian material. F. Wood Jones

GEOMETRIDAE. 1389

(Corals and Atolls, p. 354, 1910) records, “‘ Chloroclystis tenwilinea Warren”
from Cocos Keeling Is., not uncommon in June and August, the larva green,
with darker markings, and slightly hairy, feeding on the sticky flowers of Pisonia
and pupating among them. But I suspect this is an error of determination ;
the only Chloroclystis | have seen from that locality (1 2 in Mus. Brit., 3 99 in
Mus. Tring) is a remarkably lepta-like species.

Gymnoscelis Mabille.

Ann. Soc. Ent. Fr., (4), vii, 656, 1867.—Meyrick, Tr. Ent. Soc. Lond., 1892, p. 65, 1892.—Hampson,
Faun. Brit. Ind., Moths, iii, 388, 1895.—Turner, Proc. Roy. Soc. Vict., xvi, 224, 1904.—
Dietze, Biol. Eupith., 15, 1913.—Dolerosceles Meyrick, Tr. Ent. Soc. Lond., 1889, p. 480, 1889.

Differs from Chloroclystis in the absence of the proximal spurs of the hind
tibia. A small intermediate genus, Calluga Moore = Sillophora Warren, with
a single proximal spur (SC! of fore wing as in Chloroclystis) and divisible on 5
characters into several sections, may prove to be represented in Polynesia by
a race, or races, of C. costalis Moore, which is already known to be represented
in New Guinea by C. (S.) albiviridis Warren (1907), and in Queensland by G.
cissocosma Turner (1904); or of Calluga (Megatheca) catocalaria Snellen (1881,
Celebes) = C. purpurea Warren (1897, Penang) = C. lobata 3 Hampson (1912.
Ceylon), which is scattered over the islands of Malaya and Melanesia.

Of Gymnoscelis about sixty species are known, but the number is likely
to be greatly imcreased when they have been more thoroughly studied, since
_the majority are small and obscure; and often confusingly alike. On the
other hand, some of them clearly have a very wide distribution. Of the known
about five species are Palaearctic, and some ten are African, the rest Indo-
Australian, with a rather high percentage in Queensland—probably because the
Australian Geometridae are being carefully studied by a competent entomologist,
Dr. A. J. Turner, while in many other parts of the region the small species are
almost entirely neglected.

Specialised ¢ sections are rarer and less striking in this genus than in the
preceding genera. Perhaps the most striking is [ramba Moore, with its curiously
tufted middle and hind tibiae. This is represented in India and Malaya by the
type species G. tibialis Moore, and in Australia by G. lophopus Turner (1904)
= G. homogona Turner (1907), and may be expected to occur in many other
parts of the region.

140 INSECTS OF SAMOA.

20. Gymnoscelis erymna (Meyrick).

Eupithecia erymna 3 Meyrick, Tr. Ent. Soc. Lond., 1889, p. 192, 1886 (Tonga).

? 2 Dolerosceles erymna Meyrick, Tr. Ent. Soc. Lond., 1889, p. 480, 1889 (British New Guinea).
Gymmnoscelis erymna (part.) Meyrick, Proc. Linn. Soc. N. S. Wales, (2), v, 794, 1891 (Tonga).
Gymnoscelis concinna Swinhoe, Tr. Ent. Soc. Lond., 1902, p. 651, 1902 (Tonga).

? Gymnoscelis semipurpurea Rebel, 2 Beth. Jahrb. Hamb. Wiss. Anst., xxxii, 150, 1915 (Samoa).

Upolu: Apia, 1 9, v.1896 (P. de la Garde); 1 9, 13.1x.1923 (Swezey and
Wilder); 2 99, 24.v.1924; 1 9, 30.v.1924; 1 9, x.1925; Malololelei, 2 ¢Q,
1 Q, 22.11.1924: 1 gf, 24.11.1924; 1 3, 6 PQ, 2,000 feet, 25.11.1924.

Tutuila: Pago Pago, 1 g, v.1896 (P. de la Garde).

Besides specimens from Tonga and Samoa, the British Museum has a single
example from Pitcairn Is. Other localities from which the present species has
been recorded are either very doubtful or positively erroneous—cf. Meyrick,
Proc. Linn. Soc. N.S. Wales, (2), v, 794, 1891, Proc. Roy. Soc. Vict., xvi, 225-6,
1904, and Proc. Linn. Soc. N.S. Wales, xxxi, 684, 1906.

Meyrick’s supposed “9 type” (allotype) belongs to a different species,
having SC! of the fore wing running into C, the hind wing emarginate between
the radials, the postmedian line of both wings angled at R! instead of R*, ete.
The ¢ holotype is unfortunately worn, and it is no doubt the dissimilarity
between the spurious allotype and the two beautiful 22, which were subsequently
received, that misled Swinhoe into erecting a new species (G. concinna) for
the latter. In the true G. erymna = G. concinna as Meyrick rightly indicates,
SC! of the fore wing does not touch C; at least this is, without exception, the
case In over twenty examples examined by me, although in one Malololelei 2
the approach is so close that an occasional aberration with slight anastomosis
would not be an impossible development.

In view of the confusion that has arisen over this species, which is
shown by the Samoan series to be decidedly variable, I give some further
description.

Frontal cone rather sharp. Palpus 1) or scarcely over, second segment
with a small dark tuft beneath at its distal end, third segment short to medium,
distinct, rather robust, somewhat deflexed and appearing more so on account
of the longitudinal curvature of its upper side. Antennal ciliation of ¢ 1, of
© 3. Abdomen of $ elongate. Fore wing elongate, with costa straight except
at base and near apex ; antemedian double, generally with the interspace dark-

GEOMETRIDAE. 143

filled ; postmedian wavy, especially anteriorly, slightly oblique outward and
generally thickened (especially between the radials) from costa to R?, here with
a pronounced, though generally rather rounded elbow, posteriorly more puncti-
form, about parallel with termen; distal area always with a very noticeable.
though not sharply defined, pale longitudinal streak about the middle, more
or less completely interrupting the subterminal markings and sometimes almost
reaching the termen. Hind wing with termen strongly convex, almost regularly
curved from apex to near tornus, only behind M? straighter or very faintly
incurved ; SC? stalked, but generally shortly, not rarely (at least in the Q)
almost connate; postmedian line strong, generally thickened, with proximal
black vein-dashes, a more or less V-shaped indentation between the radials,
generally produced proximad by a longer black dash which converts it into a
Y, a rather variable angle outward at R?; pale subterminal mark weaker than
on fore wing,
weakly marked.

Most of the Samoan series are darker, or at least more brownish, than
the pale greenish-grey G. concinna, from which one gets the best conception of
the Tonga form, so far as at present known; but one Malololelei 2 approaches
that form quite closely. One ¢ has the wings much clouded, especially in the
median area of the fore wing (unfortunately discoloured at the base). A Q has
the median area of the fore wing and the narrow band beyond almost clear
white, confluent with the distal pale streak, which is also whitened, while the
rest of the terminal area, together with the basal, is dark ; the hind wing dark,
with narrow white band outside the postmedian and small whitish mid-sub-
terminal spot. The only ¢, taken 25.11.1924, is small, attenuated and pale. but
not in perfect condition.

G. senupurpurea Rebel was founded on 2 99 sent by Friedrichs, hence pro-
bably from Apia, although only “Samoa” is given. Notwithstanding the
close agreement of Rebel’s description with one of Buxton’s Upolu forms, |
have felt constrained to add a query to the citation because its eminent and
experienced author has definitely differentiated his G. semipurpurea from
G. wpolensis 2 by the “ non-dentate ” condition of the antenna. In the present
species—as in nearly all the allied Gymnoscelis material that I can examine—
the segments in both sexes bear at their distal ends bands of more or less pro-
jecting scales, giving to the proximal part of the antenna a vague suggestion
of an Equisetum, but tending progressively to project at a steeper angle distally,

shorter (macular), not rarely more or less obsolete. Under side

142 INSECTS OF SAMOA.

so as to suggest under a hand-lens pronounced teeth. But inasmuch as these
scales (which are dark and conspicuous in several of the species, including G.
lepta) can be more appressed, or may become abraded, much caution is needed
in judging of the exact conditions.

21. Gymnoscelis upolensis Rebel.
2 Betheft Jahrb. Hamb. Wiss. Anst., xxxii, 131, 1915 (Upolu: Apia).

Rebel founded this species on a pair of specimens taken at light by Dr. K.
Friedrichs, the 3 with a fore wing length of 8 mm., the palpus short, the abdomen
very long, the 9 smaller (a fore wing 6 mm.), with palpus twice as long ; “ antenna
in both sexes finely dentate.” The postmedian line of the fore wing has its
principal angle near the costa instead of—as in G. lepta—in the middle of the
wing; Rebel places the angle “at vem 7,” but I suppose its precise point to
be between 7 and 6 ( ==SC® and R!), as in the specimen which I am referring to
this species, as well as in the group generally (G@. wmparatilis Walker, etc.).

A @ with the fore wing almost 8 mm. in length and with a more reddish
suffusion than in Rebel’s types, much as in the reddish and not very variegated
forms of G. pumulata Hiibner, was taken by Buxton at Apia, 29.1.1924, and may
with a fair degree of probability be assigned to G. wpolensis. I think that the
same species, in a more brownish form and with a rather heavy postmedian
line, occurs in the Marquesas Is. (Nuka Hiva, 2 9g, 2 99, in Mus. Brit. ; “ 300
feet,” 1 9, C. L. Collenette, “St. George ” Expedition), and I doubt whether
it will prove more than a form of the species which I call G. imparatilis Walker
(1865): the latter is probably very variable and has been described under
many different names, although it is quite possible that at present two or more
species are confused together. A few years ago I provisionally separated :

(1) G. wmparatilis Walker, 1865 (Sarawak) = G. subtristigera Walker, 1866
(Sarawak) = (ab.) G. perangusta Warren, 1897 (Sarawak). Wings of the 3
very narrow, in the type—but not in the other cited forms—reddish beneath.
Distal area without prominent white admixture. Penang; Borneo; ? Moluccas.

(2a) G. indicata indicata Walker, 1866 (Sarawak) = G. confusata Walker,
1866 (Sarawak) = (ab.) G. semivinosa Warren, 1896 (Java) = ? (ab.) G. semaal-
hida Walker, 1866 (Borneo). Wings of $ rather less narrow. Distal area of
hind wing usually with much admixture of white in posterior half. Ceylon ;
India; Malaya; Java; Borneo; Formosa.

GEOMETRIDAE. 143

(2b) G. indicata grisea Warren, 1897 (Fergusson) = G. aenictopa Turner,
1907 (Queensland). Whitish admixture in distal areas generally less manifest,
more central (nom. vix conservandum?). % Sambawa; ? Kei Is.; ? Dammer I. ;
New Guinea; D’Entrecasteaux Is., Louisiade Archipelago; Bismarck Archi-
pelago ; Solomon Is. ; Queensland.

Unfortunately much of the available material, including some of Walker’s
types, is in very poor condition, and some of the names are based on 99. The
present notes are only offered as a possible basis for further research.
Structurally the group differs from G. lepta in the anastomosis of SC! with C.
The specimen of G. “ wpolensis ” before me resembles G. invparatilis rather than
G. wndicata, but on geographical grounds—if there be two species—it more
probably represents a race of the latter.

22. Gymnoscelis refusaria (Walker).

Acidalia refusaria Walker, List Lep. Ins., xxii, 767, 1861 (Sarawak).
Chloroclystis minima Warren, Nov. Zool., iv, 227, 1897 (N. Queensland).
Pasiphilodes maculilinea Warren, Nov. Zool., v, 429, 1898 (Kei Islands).
Gymnoscelis refusaria Swinhoe, Cat. Lep. Het. Oxf. Mus., 11, 350, 1900.

Upolu: Malololelei, 1 9, 23.11.1924.

The specimen is not very fresh, but seems evidently referable to this
distinct and not particularly variable species, which is rather short-winged in
comparison with its congeners, paler than usual, and has the markings weak
except for a few strong spots.

New for Polynesia, but the previously known range is very wide: Ceylon ;
Borneo; Kei Is.; New Guinea (Goodenough Bay); Rossel I.; Bismarck
Archipelago (Rook I. and New Hanover) ; N. Queensland. G. refusaria appears
to be predominantly an island species, but—like most of its congeners—is no
doubt much overlooked.

Pseudomimetis, gen. n.

Face with small cone below. Palpus moderate, second segment distally
with projecting scales above and beneath, third segment moderately strong.
Tongue developed. Antenna in ¢ simply ciliated. Hind tibia with two rather
unequal spurs. Abdominal crests slight. Fore wing without areole, SC!-SC*
stalked, SC! arising opposite or slightly before (very rarely just after) SC’,
anastomosing with C, R! separate, M! separate. Hind wing in ¢ not contorted ;

144 INSECTS OF SAMOA.

C anastomosing with SC to near end of cell, SC? stalked with R! ia 9, coincident
in g, M! connate or just separate.

Genotype: Pseudomimetis semiviridis (Warren) = Gymmnoscelis coquina ab.
semivwidis Warren, 1897.

Differs from Hybridoneara Warren in having the $ antenna not dentate-
fasciculate, in the less long stalking of SC!, in the hind wing of the ¢ not being
distorted, but with SC? coincident, the wings of the 2 not unusually broad,
M? of the hind wing not stalked.

The few Larentiids of the Chloroclystis group which have the areole sup-
pressed (or vestigial *), leaving all five subcostals of the fore wing to arise from
a common stalk, can be referred to four genera, distinguishable as follows :

1. Hind tibia with 3 spurs, SC1 of fore wing free. ; : . 2
Hind tibia with 2 spurs, SC! of fore wing anastomosing w ith Cs : 3
2. Hind wing of 3 much distorted, costally + with large brush of hair . . Mariaba.
Hind wing of 3 nearly amygdaloid, not cisvorteds above with a roundish
sex-patch beyond cell , : : : : . Onagrodes.
3. Hind wing of $ triangular, wings of 2 Oe peed : : : . Hybridoneura.
Wings normally shaped : : : : : ; : : . Pseudomimetis.

Onagrodes contains two known species, and probably also a third (found in
S.-W. Sumatra), of which the 3 has yet to be discovered. Though closely related,
the two former species differ in the exact position of the sex-patches of the 3,
that of the hind wing being placed between SC? and R!, near their origin, in
the Indian species and at the base of the radials in the Papuan. The former,
O. obscurata Warren (1896), is only known as occurring in the Khasi Hills; the
latter, O. recurva Warren (1907) (3) = 20. viridis Warren (9) was described
from a specimen from New Guinea (Owen Stanley Range), but is also known
to occur in both forms in the Bismarck Archipelago. In the present genus, the
structure of the scent-patch shows a rather close resemblance to that of Melitulias
Meyrick.

To Mariaba Walker (= Adeta Warren) belong also, if my suggested
synonymy is correct, only two known species, both with a wide distribu-
tion, though not as yet found in Polynesia: (1) I. convoluta Walker (1866,
Sarawak) = ? MW. ampla Warren (1899, Lombok); additional known localities
for the collective species are Assam to Malay Peninsula, and Dutch New Guinea.

* The latter is the case in several $g and at least one 2 of Hybridoneura, including the type
specinen of the genotype H. abnormis Wane: though overlooked by its describer.

GEOMETRIDAE. 145

(2) C. semifascia Warren (1903, British New Guinea) ; additional localities for
the collective species are the Sula Is., New Britain, and the Solomon Is. Both
species will probably be separable into races, but are still very sparsely repre-
sented in collections, and are further complicated by the extreme sexual
dimorphism. It may be added that “ Mariaba” medioplaga Swinhoe (1902)
is a Micrulia (Opistheploce), vide supra; “ Adeta” confusa Warren (1906),
probably a Hybridoneura 9, while “ Adeta”’ viridis Warren (1907) is, I suspect,
the 2 of Onagrodes recurva Warren (1907).

Hybridoneura Warren and Symnumetis Turner (= Neoscelis Hampson,
1903, nec Schock, 1897) must evidently be sexes of a single genus, though this
does not appear to have been suspected heretofore. The known specimens
of Symmametis (sens. str.) and Neoscelis are invariably 2, and the pattern of their
under sides shows a good deal in common with that of examples of Hybridoneura,
which are always g. H. abnormis Warren (1898) was found in the Khasi Hills,
and NV. cristata Warren (1897) = N. rivula Hampson (1903) has also been taken
there. A ¢ closely similar to H. abnormis was captured with N. metachlora
Hampson (1907) at Haputale, Ceylon. An undescribed Hybridoneura, apparently
representing a race of H. abnormis, was collected in the Hydrographer Mountains,
British New Guinea, with what appears to be WN. cristata. The @ forms which
have passed, comprehensively, as “ Neoscelis”’ cristata, and really vary very
little, have an extremely wide range: Ceylon, N. India, the Andaman Is.,
Penang, Bali, Dutch and British New Guinea, Rossel [., Feni I. (off New Iveland),
the Solomon Is. (Bougainville I.), and ? Fiji (only a worn example yet seen from
the last-named locality). From Turner’s description of his NV. muscosa, more-
over—the type of the genus Symmimetis (1907)—I have no doubt that Queens-
land must also be added, though I have not yet seen an Australian example.

23. Pseudomimetis semiviridis (Warren).

Gymnoscelis coquina ab. semiviridis Warren, Nov. Zool., iv, 228, 1897 (Queensland).
Gymnoscelis picta Warren, Nov. Zool., viii, 30, 1901 (Khasi Hills [err. loc. ?]).
Symmimetis sylvatica Turner, Tr. Roy. Soc. S. Austral., xlvi, 234, 1923 (N. Queensland).

Upolu: Vailima, 600 feet, 1 9, 1.11.1924.
There may possibly be some geographical variation in this species, of which

few examples are yet known, but there can be no doubt concerning their essential
homogeneity. [xcept for the unique “ Khasi” type of G. picta, I had only seen

146 INSECTS OF SAMOA.

specimens from Woodlark and Rossel Islands and N. Australia, and it is there-
fore justifiable to question the origin of the type in question. But, in view of
the range of such species as Micrulia tenualinea (supra), | cannot reject it utterly,
especially considering the high degree of accuracy shown in the labelling of
specimens in the Tring Museum. ‘The Woodlark and Rossel 9° have the blackish
patch on the end of the abdomen above rather more extended than in those
from Queensland and Samoa. Both forms vary in the depth of the ochreous
suffusion of the hind wing.

Warren’s Gymnoscelis semiviridis was extremely badly described, and the
species has not the remotest connection with G. coquina; but since the author
in question expressed a suspicion that it might prove to be a distinct species,
his name must be regarded as valid.

Ziridava Walker.

List Lep. Ins., xxvi, 1550, 1862.—Hampson, Faun. Brit. Ind., Moths, iti, 391, 1895 (Chloroclystis
sect. Liridava [sic])—Swinhoe, Ann. Mag. Nat. Hist., (6), xvi, 295, 1895 (Xiridava [sic]).

This genus has, perhaps with justice, been merged by some systematists,
in Chloroclystis, but seems to be a collateral development. Face with long and
sharp cone. First segment of palpus with strong scale-tuft beneath, second
segment with strong downcurved scale-tuft distally beneath. Wings elongate,
coloration not green, SC! anastomosing with C, but without the pronounced
tendency to remain coincident with it thereafter.

Two sections, according to the 3 antenna, have hitherto been known.

A. Antenna of $ pectinate: Z. rufinigra Swinhoe (1895, Khasi Hills), with
a race brevicellula Prout (1916, D’Entrecasteaux Is.; also known to occur
in New Guinea), and a second race, at present unnamed (N. Queensland).

B. (Sect. typ.). Antenna of ¢ dentate-fasciculate: Z. xylinaria Walker
(1862, Sarawak ; also found in N. India, Malaya, Java, ? Bali), with provisional
races subrubida Warren (1897, Celebes), leptomita Turner (1907, N. Queens-
land; also known to occur in New Guinea, D’Entrecasteaux Is., Louisiade
and Bismarck Archipelagos, and the Solomon Is.; possibly susceptible of
further subdivisions), and perhaps a fourth race (? sp. div.) found ia Flores.

To these must now be added Sect. C, with g antenna simple, for the
reception of the following species.

GEOMETRIDAE. 147

24. Ziridava dysorga, sp. 0.

dg, 25 mm. Head and body concolorous with wings. Palpus 234, first
and second segments mixed with blackish beneath. Hind tibia with the spurs
not extremely unequal. Abdomen less slender than in the 3 of the genotype ;
first somite (with extremity of thorax) black above; further irregular brown
belts ; crests well developed, tipped behind with black.

Fore wing with SC! anastomosing only shortly with C; light pinkish cin-
namon, with deeper shadings—along the costa more vinaceous, partly (especially
in the cell posteriorly) bright ochraceous buff; cell-dot minute, grey ; Imes in
part very weak and slender; antemedian double, oblique, slightly interrupted,
enclosing a narrow, glistening, grey band (in some lights vinaceous-tinged) :
postmedian similar, rather less definite, slightly more oblique, retracted and
(especially proximally) ill-defined costally, marked at hind margin by a dark
spot and between the radials by a heavy blackish patch; subterminal
obsolescent, but with twin blackish spots between the radials ; terminal spots
evanescent ; fringe somewhat olivaceous. Hind wing with termen waved;
SC? quite shortly stalked ; proximal area rather pale, with very fine lines ; post-
median lines obsolete anteriorly, from R! to abdominal margin almost straight,
entirely filled in with blackish; outer lines very indefinite, the subterminal
developing dark interneural dots, of which the posterior three are the stronger.

Under side with the markings more weakly reproduced.

2, 28 mm. SC! of fore wing anastomosing moderately with C. Paler,
nearly the colour of Asthena eurychora, the blackish markings of abdomen and
postmedian wanting, those of subterminal greatly weakened, indicated in
brownish.

Upolu: Malololelei, 2,000 feet, type g, 21.1v.1925 ; allotype 9, 2.vii.1924
(Armstrong).

A figure of this species, in colour, will appear in Part III, Fascicle 4.

Sauris Guenée.

A moderately large Indo-Australian genus, consisting of about seventy or
eighty known species, of which several are not yet described. The present
genus is of great interest on account of the multiplicity of the secondary sexual
divergences in the J, although pride of place in this respect must be accorded
to the genera Anisodes and Sterrha, discussed above. In Sauris and the nearly

148 INSECTS OF SAMOA.

world-wide group to which it belongs, the modifications are chiefly in the 3
hind wing, which is often extremely narrowed and contorted, its posterior part
commonly cut or folded into simple or composite lobes or pockets. The
Holarctic genera (Lobophora group) are in the main the least extremely
specialised, and have the palpi normal. In Sauris the latter are more often
abnormally long, as is frequently the case in the endemic Tatosoma of New
Zealand, which differs from Sawris in having three spurs on the hind tibia and
the areole double, whereas in Sauris the g hind tibia is spurless, while the 2
has two spurs on its hind tibia and a simple areole. Rhopalodes, of Tropical
America, has the areole double and all spurs present, but is remarkable in having
clubbed antennae. The Chilian representatives of the group (Lomopteryz,
Laqynopteryx, Hoplosauris and several others) show great diversity of structure,
including some of the most extreme reductions of the hind wing, but hitherto
have not received much in the way of detailed study. :

In a first rough analysis of Sauris I have recognised thirteen structure-
eroups, which include the following “genera” of Guenée and Warren:
Dystypoptila Warren, Holorista Warren, Pseudoschista Warren, Helminthoceras
Warten, Remodes Guenée, Anthierax (Warren, Nov. Zool., xii, 11, 1905, in-
descr.), Coptogonia Warren, and Anisocolpra Warren. Possibly one or two of
these may prove generically tenable, but the majority are clearly shape-sections
of an indivisible entity. It is noteworthy that the 2 hind wing shows three types
of venation, which may have some phyletic significance: (1) R?—M! connate
or just separate (S. lichenvas Meyrick, 1891, only 1 2 known to me); (2) R3—M!
stalked (the great majority); (3) R’—M! coimeident (S. hirudinata Guenée,
1858, S. muscosa Rothschild, 1915 [hirudinata subsp.?] and S. angusta
Warren, 1905).

S. elaica Meyrick and its Lifu relative (vide supra, p. 132) probably belong
to the section which is typified by S. remodesaria Walker (1862—Remodes auctt.,
nec Guenée), and have, in addition to the contorted hind wing, a more or less deep
cleft in the termen of the fore wing between the median veins, and tufts of
specialised scales and of hair on the tornal lobe thus formed. This section con-
tains also the following, besides a few species and races at present undescribed :
S. melanoceros Meyrick, 1889, British New Guinea; S. vetustata Walker, 1866,
Australia [Queensland]; S. auricula Warren, 1895, Perak; 8S. parviplaga
Warren, 1906, New Guinea; 8S. cirrhigera Warren, 1897 = S. nusta Swinhoe,
1902, New Guinea and eastward as far as the Bismarck Archipelago, perhaps

GEOMETRIDAE. 145

also on Ceram I. and Waigeu I. (We~ren’s type was from Woodlark I.); 8.
subnigrata Warren, 1905, Solomon Is. (praec. subsp. ?). Whether the new species
about to be described also belongs here must be settled when the ¢ is discovered,
but [ think it probable. No other section of Sawris need yet be specially con-
sidered in relation to the Polynesian fauna; the section Sauris only reaches
the Solomon Is. (S. abnormis nitidula Warren, 1899 and S. angusta Warren,
1905), Holorista New Guinea, Pseudoschista the Solomon Is. (P. dentatelnea
Warren, 1905, and a few unnamed forms), Helminthoceras Woodlark J. (H.
erigens Prout, 1925), Anthierax the Solomon Is. (A. subfulva Warren, 1905),
Coptogonia the Bismarck Archipelago (C. lucens fortunata Prout, 1925); C.
volcanica Butler, 1887 (= C. erebata Warren, 1903) belongs to the Solomon Is.,
C. aspricosta Prout, 1925, to New Ireland, and C. brunnescens Warren, 1896,
to British New Guinea, Fergusson I. (loc. typ.) and the Solomon Is.

25. Sauris mellita, sp. n.

2, 383-36 mm. Head and body buff. Palpus very long (nearly 4), terminal
segment long, not extremely slender; tinged above with olive. Legs in part
clouded with grey.

Fore wing with areole large, SC! from before its end, SC® just stalked with
SC2-4, R! about connate, R? slightly before middle of DC; base mixed with
buff (perhaps discoloured in the type); the rest whitish-green, with rather
thick wavy lines of deep grape green, about eleven more or less well expressed,
one or two others faintly indicated ; the lines of the median area in part marked
with dusky brown, especially behind M and its branches, culminating in rather
strong spots at SM? ; this area twice as broad at costa as at hind margin; the
three postmedian lines (especially the outermost) rather markedly indented
inward at SC* and outward at R!, the second and third coalescing behind middle ;
the bisected band beyond rather more mixed with white, especially distally ;
the first and second lines of distal area marked with dusky brown between the
radials and just behind M?; the line outside the pale subterminal indented in
cellule 7, with some dark scales at the point of the tooth; termen with six
large dark dots (on veins SC® to M2), a smaller dot at apex and at SM? ; fringe
pale. Hund wing with C anastomosing with SC nearly to end of cell, SC? with
R! and R? with M! variably stalked, R? from about centre of DC ; cream-buff to
chamois ; distal half of fringe more whitish.

Tit (8) 3

150 INSECTS OF SAMOA.

Under side coloured nearly as hind wing above, the fore wing more tinged
with honey-yellow.

Upolu: Malololelei, 2,000 feet, type 5.vu., paratype 14.vi.1924.

I scarcely think this can be a race of S. elavca Meyrick, of which the type
2 18 worn, the hind wing apparently grey, not buff, while the fore wing has
heavier black-brown markings.

A figure of this species, in colour, will appear in Part III, Fascicle 4.

Asthena Hiibner.

A rather small genus, consisting, as at present restricted, of only twenty-
two species, although some related groups—Brhastna Prout, with dentate and
excavated distal margins, Minoa Treitschke, with rather different venation and
different habitus and ecology, and Laciniodes, with heavier palpus—might
well be treated as subgenera. Of the number mentioned, sixteen species are
Palaearctic, Chinese, or Himalayan, and there is consequently some room for
doubt whether the remaining six have any really near connection with them ;
of these six, five inhabit New Guinea, while A. euthecta Turner (1904) occurs in
Queensland and Victoria.

The genotype, A. albulata Hufnagel (= A. candidata Schiffermiiller) is
rather abnormal, having the cell of the fore wing nearly $, the areoles generally
both small, with all the five subcostals stalked well beyond them ; and Turner,
in proposing (Tr. Roy. Soc. S. Austral., xlvi, 233, 1922) to remove his A. euthecta
to Minoa, would also remove nearly all the other Asthena, including even A.
anserarva Herrich-Schafler, A. nymphaeata Staudinger and A. lassa Prout (1926),
which, mdeed, not only have (like A. euthecta) normally developed areoles, with
SC! arising from the anterior margin of the distal one, but further differ vena-
tionally from A. albulata in that the cell is ? or scarcely longer. On the other hand,
another abnormal and rather discordant species (“ Autallacta” lida Warren,
1896) could remain with A. albulata in so far as regards the stalking of the five
subcostals, yet has the distal areole elongate and the cell even shorter than in
the anserarva group. Finally, I have one aberrant A. albulata in which SC*
arises just proximally to the end of the outer areole. It is evident, therefore,
that some other differential characters must be sought, or the assemblage be
Jeft undisturbed. I am content, for the present, to accept the latter alternative
Minoa should either be merged (as by Meyrick) or preferably restricted to its

GEOMETRIDAE. 151

genotype M. murinata, pending further research. The Indo-Australian genus
Poecilasthena Warren, and still more the Holarctic Oporinia Hiitbner, differ more
widely from Asthena; cf. Turner, tom. cit., pp. 229, 231, to which should be
added, in the case of Oporinia, that the frenulum is aborted, the resting posture
different, the uncus fully developed, the heads of the labides united, and that
the early stages are not comparable. Poecilasthena consists of : ten or eleven
Australian species, one of which (P. thalassias Meyrick) extends to New Guinea,
the South Moluccas, and even Malaya (subsp. div. ?), and has a close relative
in Upper Burma (P. burmensis Prout, 1926) ; two species found in New Guinea ;
and two or three which occur in New Zealand.

The new species described below shows the general scheme of markings
characteristic of the genus, or in particular of A. euthecta Turner, but even
more recalls the Psilocambogia section of Hois Hiibner (L. memorata Walker,
E. cymatodes Meyrick, etc.), a genus which differs essentially from Asthena in
having the areole simple and very small, or even entirely wanting, the first
subcostal stalked well beyond the fifth, and the genitalia (so far as investigated)
more Aiisodes-like. In the species about to be described, there are no signs
of definitely close affinity with any other known to me, and I shall not be sur-
prised if the g proves to possess some characteristic specialisations.

26. Asthena eurychora, sp. n.

2, 28-24 mm. Face not broadened. Palpus with termmal segment dis-
tinct, projecting slightly in front of face. Antenna with minute ciliation and
with single longer and stiffer hairs. Head, body, and wings bufi, more whitish
beneath.

Fore wing with cell short (not quite 2), SC! very proximal at origin, both
areoles very long, SC! separating proximally to the apex of the outer one, R!
shortly stalked with SC3-°, M? from before end of cell; rather glossy, very
pale yellowish (not quite so bright as cartridge buff), with the irroration and
markings dull cmnamon, sufficiently dense to give the wing a prevailing tone of
pinkish buff; cell-dot mixed with black; lines crenulate, slightly bent near
costa or (the outer ones) more evenly curved, mostly rather thick, in part mixed
with grey, especially the first postmedian, which appears a little less crenulate
than the rest ; two subbasal not very sharply defined, the outer the thicker ;
two antemedian, rather irregular, perhaps more slender than the rest; sug-

152 INSECTS OF SAMOA.

gestions of longitudinal connection between these and the postmedian at bases
of R8—M? ; postmedian double, the outer a little the more curved ; subterminal
tripie, the proximal two dark-marked just in front of R* ; termen with minute
blackish dots at ends of veins. Hind wing with termen very weakly bent at
R? ; C anastomosing with SC to about 2 cell, SC? very shortly stalked, M! very
well separated at origin from R® ; subbasal lines wanting, antemedian both well
proximal to cell-dot, weak anteriorly ; the rest nearly as on fore wing, the
postmedians somewhat outbent at middle, here subconfluent.

Under side almost unmarked.

Upolu: Malololelei, 3 99, 24.11.1924.

A coloured figure will be published in Part III, Fascicle 4.

GHOMETRINAE.

This gigantic assemblage, the BoarMiaNarE [sic] of Hampson, SELipo-
SEMIDAE of Meyrick, EnNnominae of Hulst, is comparatively weakly represented
in Melanesia, and apparently—except for the species of the Chogada group of
Cleora, which are often disproportionately dominant numerically—extremely
weakly in Polynesia. Defective though our data remain in several respects, it
will not be altogether without value to give here some statistical comparisons
of the prevalence of this subfamily in various faunistic regions with that of the
GEOMETRIDAE as a whole.

For the Palaearctic Region as understood by Staudinger and Rebel (Cat.
Lep. Palaearct. 1901)—+.e. the region in which the small and obscure Geometrids
such as Sterrha and Hupithecia have been the least inadequately studied in
relation to the rest, with a resultant prejudice to any apparent dominance of
‘ BOARMIINAE —there were recognised about 396 Boarmiids out of a total
of 1233, or about 32 per cent. In Vol. IV of Seitz’s “ Macrolepidoptera of the
World *’—where we find a wider conception of the Palaearctic Region, including
vast tracts of China in which investigation of the obscure genera in question has
scarcely even conimenced—lI find the percentage of GEOMETRINAE has risen to
38:5 (some 870 species out of nearly 2,260).

With reference to Boreal America, Barnes and McDunnough in 1917 cata-
logued 516 GEOMETRINAE out of 1,047 GEOMETRIDAE, or 47 per cent. For

GEOMETRIDAE. 153

South America there is as yet no list, but: here also the present subfamily cer-
tainly furnishes a very high percentage of the whole.

As regards tropical Africa also there is no list; for South Africa Janse
(Check-Lnst, 1917) gives 267 GEOMETRINAE out of 590 Geometrid species, namely
between 44 and 45 per cent.

in British India Hampson (Faun. Brit. Ind., Moths, ii, 1895) recognised
1,063 species * of the family, 480 of them (45 per cent.) being Boarmiid.
In the wonderful collection made by Capt. A. EH. Swann in the Kachin Hills,
Upper Burma (cf. Jowrn. Bomb. Nat. Hist. Soc., xxxi, 129-46, 309-22, 780-99,
932-50, 1926-27), the Boarmiids actually preponderated, numbering 175 species
out of 329 (53 per cent.) ; and this notwithstanding the fact that Captain Swann
paid particular attention to the small species.

The Malayan Subregion, when figures are available, may yield results
similar to those furnished by the Indian; in working out the Geometrids of
Dr. E. Mjéberg’s valuable collections from some of the Mountains of Sarawak
(Sar. Mus. Journ., 11, Pt. II, No. 9, pp. 169-210, 1926) I found the 124 species
to comprise 58 GEOMETRINAE, or nearly 47 per cent.

In Australia, out of about 950 described GEOMETRIDAE,t some 300 are
** BOARMIINAE,”’ just the same percentage as in Staudinger’s Palaearctic Region.
As regards New Zealand, Mr. Meyrick (Tr. N.Z. Inst., xlx, 266, 1917)
has already remarked on the very inadequate representation of the subfamily ;
Tillyard shows the figure to be 44 out of a total of 237 Geometrids, or not quite
20 per cent.

For New Guinea, the only papers yet published that can provide any basis
for statistical calculations are three by Warren (Nov. Zool., x, 343-414, 1903 ;
xill, 61-161, 1906; xiv, 97-186, 1907), describing the new species contained in
some wonderfully rich collections made by Meek in the Owen Stanley Range.
Although these papers do not enumerate the previously known species of which
specimens were obtained, the proportion of novelties was so high that I beleve

* Hampson’s wholesale “lumpings” probably do not appreciably modify percentages,
since they are impartially distributed between the subfamilies ; their influence, if any, will tell in
favour of this author’s BoarmiAnar. The most glaring examples that occur to me are his
Boarmia acaciaria and Abraxas sylvata (“ BOARMIANAE”’), and his Sauris hirudinata (° LAREN~
TIANAE ’’),

t Tillyard’s recent census (Ins. Austral. and N. Z., pp. 449-452, Nov. 1926) yields a total
of 916, but this is not quite complete, since Dr. Turner’s works on which it is based, admittedly
exclude a number of species which he has been unable personally to examine.

154 INSECTS OF SAMOA.

the figures afford a fair index to the numerical strength of the respective sub-
families in the district, and probably even in the mountains of New Guinea in
general. It appears that, out of 473 new species, 171 (36 per cent.) belong to the
GEOMETRINAE, which the author—as in all his papers—distributes amongst
several untenable “ subfamilies.”

A collection from Dampier [., Vulcan I., and the Admiralty Is. which I
analysed for Lord Rothschild when it was received from Meek several years
ago, contained representatives of about 140 species, of which only 34 were
Boarmiid (between 24 and 25 per cent.). Several of these were common and
widely distributed species, such as Cleora inflexaria Snellen and C. decisaria
Walker, Hetropis cessaria Walker and FE. sabulosa Warren, Catoria delectaria
Walker, Semiothisa goramata Rober = S. variegata Warren, Hyposidra talaca
Walker, Borbacha euchrysa Lower, Petelia medardaria Herrich-Schiffer, Bulonga
qriseosericea Pagenstecher and others, though sometimes redeemed from the
commonplace by some racial modification. Of outstanding endemic develop-
ments there was scarcely a trace.

When we turn to Samoa, we find only 4 Geometrinae among our 30
species, and I doubt whether any other islands of the South Pacific would provide
a much higher proportion, though the entirely dissimilar Hawaiian fauna gives
the GEOMETRINAE more than a 2:1 majority over the LARENTIINAE, which is
the only other Geometrid subfamily found there—the endemic genus Scotorythra
and its immediate offshoots having obtained almost undisputed sway. Meyrick’s
“ Descriptions of Lepidoptera from the South Pacific’ (Lr. Ent. Soc. Lond.,
1886, pp. 190-212), out of 29 Geometrids, include only one Geometrine—Boarmia
psychastis, of the New Hebrides, apparently a 2 ab. of Cleora decisarva—
together with a reference to another single 9, from Samoa, “ probably the same
species,” or rather, in the light of present knowledge, No. 28 infra. Nearly
every island produces at least one representative of this group, but otherwise
{ have searched almost in vain, through the collections to which I have access,
for GEOMETRINAE. Two genera have, indeed, been erected for single species
and call for brief notice here; but they seem to be offshoots (with ¢ speciali-
sations) of the group last mentioned.

Cerotricha (type C. licornaria Guen., Tahiti) is perhaps generically separable
from Cleora by its more slender build and extremely long antenna, in some
measure linking that genus with Scotorythra. But Alcis nausori Bethune-Baker
(Proc. Zool. Soc. Lond., 1905, p. 94, t. viii, f. 6, Viti Levu), which has also rather

GEOMETRIDAE. 155

long wings and long antenna, is an obvious Cleora of the Chogada group.
Warren’s note on Cerotricha (Nov. Zooi., i, 377, 1894) was based on miscon-
ceptions, and must be ignored altogether.

Aegitrichus (type A. lanaris Butler, 1886, Fiji) is still only known from the
much damaged ¢ type, and no definite opinion as to it can yet be expressed ;
but from all the appearances I should suppose it to be a Cleora (Chogada) with
highly specialised g hind wing, in which the posterior half is separated off into
an excessively hairy lobe ; all the other characters shown agree absolutely with
those of the present group.

Of the four Geometrinae now known to occur in Samoa, two are “ Chogada,”’
while the other two extend respectively the range of an Indo-Australian genus
(Nadagara), and that of a widely distributed Indo-Australian species (Orsonoba
cleha}, both of which were previously supposed to reach their eastern limit in
the Solomon Islands.

Cleora Curtis.

Brit. Ent., ii, 88, 1825.—Turner, Proc. Linn. Soc. N. S. Wales, xlii, 370, 1917.—Alcis Curtis, Brit.
Ent., iii, 113, 1826.—Prout, Ann. Transv. Mus., iii, 222, 1913.—Chogada Moore, Lep. Ceyl.,
m, 415, 1887.—Selidosema (? Hiibner, Verz., 299, 1825) Meyrick, Tr. Ent. Soc. Lond., 1892,
105.—Boarmia part. (Treitschke, Schmett. Eur., v, (2), 483, 1825) Hampson, Faun. Brit.
Ind., Moths, iti, 256, 1895.

In my preliminary dealings with the world’s GromMETRINAE—admittedly
one of the most difficult of all the Lepidopterous subfamilies—I have perforce
accepted in its broad outline Meyrick’s conception of the nearly world-wide
*“ oenus ” which he called Selidosema, but for which Cleora * is probably a prior
and certainly a safer name, S. plumaria Schiffermiiller (the type of Selidosema)
differmg im several particulars and being usually considered suz generis, as by
Lederer, Guenée, McDunnough and others. I am aware, however, that further
subdivision will sooner or later be necessary; especially illuminating is
McDunnough’s fine essay, “‘ Studies in North American Cleorina” (1920), in
which he treats Hampson’s BoarmMia as a tribe, and differentiates no less than
twenty-three natural genera, according to characters derived from the genitalia,

* Cleora Curtis, 1 Oct., 1825; Selidosema Hiibner, “late 1825 (post Aug.),’”’ sec. Sherborn
and Prout, if not even 1826.

156 INSECTS OF SAMOA.

venation, antenna, etc. It is to be feared that many years must elapse before
anything of the kind is done with the mass of exotic material that is awaiting
attention ; but fortunately the need is not urgent in connection with the present
work, for Moore’s Chogada can quite satisfactorily be left with Cleora, sens. str.,
as defined by McDunnough (loc. cit., p. 19). Its type (C. alenaria Walker)
differs chiefly from Cleora cinctarva in the rather longer palpus, longer primary
pectinations and strongly dilated hind tibia of the 3, and in the strongly marked
under side. Vein SC? of the fore wing in C. alienaria and its nearest allies is
much more rarely connected with SC* than in typical, Holarctic Cleora, but
varies in both groups. The facies of what I have called the “ Chogada group ”
(Indo-Australian and African) is nearly always unmistakable, but I cannot
accept it as generic, since there are intergrades in all the characters mentioned ;
e.g. C. leucophaea Butler, of Japan, with long pectinations, C. displicata Walker,
of Queensland, with non-dilated hind tibia, several species with intermediate
palpus, and a few—such as C. fortunata Blachier, of Teneriffe, with intermediate
facies.

The number of species which have been confused under the name C.
“acacaria Boisduval’’ must be very great. Many of them have been found
readily separable by wing-markings, when studied by good observers like Moore
and Turner ; but some, which I confess to having supposed to be merely geo-
graphical forms, or sometimes even aberrations, are proved by the genitalia to
be also totally distinct species. Mr. W. H. T. Tams, at the British Museum,
has kindly made preparations of a very large number of forms, and the results
are extraordinarily interesting. But the number that would require examina-
tion before any systematic revision could be made continues to increase rather
than to diminish, and we must reluctantly relegate to the future all but the
barest generalisations. »

The numerous African forms for the most part differ markedly from the
Indo-Australian, and show on the whole a somewhat simpler type of genitalia.
One unnamed Ceylon species, of which only a single example is yet known,
possibly originated from the same stem as the African forms, and as regards
its genitalia is curiously remote from the rest of the Indo-Australian thus far
examined. Only two of the Indo-Australian species seem to be really widely
distributed : C. injectarra Walker (Ceylon) = C. sublectaria Walker (Ceylon)

—unfortunately founded on an aberrant 2 (Moulmein)-—will prove the oldest

GEOMETRIDAE. 157

name; and C. decisarva Walker (Ceram) =C. callicrossa Meyrick (New Guinea)
= C. lacteata Warren (New Britain). The former has a race or exceedingly
close ally in Fiji, and it will not be at all surprising if both it and C. decisaria
are also found in Polynesia. I have before me a ¢ of a probable race of C.
decisarva from Lifu, and a small worn ¢ (apparently more typical) from the New
Hebrides, the latter presumably the ¢ of C. psychastis Meyrick (vide swpra).
Warren named several aberrations from the Kei Is. (Nov. Zool., v, 430, 1898),
and one from the Solomon Is. (Nov. Zool., xii, 432, 1905) ; westward the species
reaches Borneo and Java (Mt. Gedeh), and I have even seen one specimen from
Penang.

The Samoan and other Polynesian Cleora belong, I suppose, to the same
subgroup as C. anyectaria and C. decisaria, and, on McDunnough’s antennal
character, would not be quite so strictly congeneric with C. cinctaria as the
alienaria subgroup. They have still longer and more lax 3 pectinations, which
tend to curl irregularly round the shaft, and the slender secondary pectinations
have disappeared—at least so far as can be determined without laborious
anatomical research. I cannot see, however, that this difference is correlated
with any other structural one ; all the African species have the antennal structure
of the Polynesian, etc., while several of the Indo-Malayan—C. alienaria Walker,
C. gelidaria Walker, C. determinata Walker, C. concentraria Snellen and some
unnamed species together with the Australian C. dlustraria Walker and the New
Guinea C’. hospita Prout—have the doubly bipectinate antenna of C. cinctaria
Schiffermiiller. It should be added that the very common C. mflexaria Snellen,
which is an outher of the present group and occurs in most localities from Malaya
to the Solomon Is., shows variable and often asymmetrical antennal conditions
similar to those noticed by Turner (77. Roy. Soc. S. Austral., xxviii, 230, 1904 ;
Proc. Linn. Soc. N. S. Wales, xii, 333, 1917) under “ Hetropis ” hemiprosopa
Turner and “ #.” camelaria Guenée, except that in that group (the genus Catoria
of Moore) the geminate branches are of approximately equal width, whereas in
Cleora inflexaria one branch is extremely slender, as in C. cinctaria. This
species (C. inflezarva) will perhaps be found to inhabit Polynesia also, and it is
not impossible that one or more representatives of Catoria may likewise be
discovered there ; both the white group (camelaria group) and the green (delectaria
[Walker] group) are very widely distributed, and reach the Bismarck Archipelago
and the Solomon Islands.

158 INSECTS OF SAMOA.

27. Cleora samoana (Butler) (pl. V, figs. 1-18).
Boarmia samoana Butler, Proc. Zool. Soc. Lond., 1886, p. 433, 1886 (Samoa).
Boarmia acaciaria samoana Rebel, Denkschr. K. Akad. Wiss. Wien, Math.-Naturw. K1., \xxxv,

429, t. xviii, f. 5, 6, 15, 1910 (Samoa).

Upolu: Apia, Vailima, Malololelei, a series of 160 gg, 100 99, taken in
every month of the year (various collectors).

“ Tutuila,” iv.1918, 1 ¢ (Kellers). Tutuila: Pago Pago, 1 g, 8.1x.1923 ;
1 5 9.1.1923): 1) fh: 10.1x.1928 LO So, 2 OO: Ta rx 199862 BT ORS axe 923K
1 Q, 21.1x.1923 ; 6 Sg, 1.1924; 10 dg 3 G9, 11.1924 (Steffany).

Savaii: Palauli, 1 g, 1 9, 8.vili.1924; Salailua, 2 $o, 12.viii.1924.

Manw’a: Ofu, 1 g, 27.11.1926 (Judd). “Samoa,” 3 gg, 2 GP, 1920; 2 gd,
i.—vi.1921 (O’Connor).

It has already been noticed by Rebel that this is an excessively variable
species, and the long series of specimens, mostly of very small size, collected
by Woodford for the Tring Museum, showed similar variability. But the
material now brought together is so extraordinarily interesting that it has
been thought desirable to figure (pl. V) a representative series of the most
remarkable forms. In erder to produce a more striking effect, we have drawn
exclusively on the wonderful series collected at light by Mr. Hopkins on a single
night (Upolu: Vailima, 7.ix.1925); and even with this restriction, the task
of selection has been difficult, many of the minor variations having inevitably
been left out. Perhaps the most usual forms of the ¢ are whitish- or yellowish-
brown, more or less heavily dark-irrorated and mottled, with no extremely
sharp colour contrasts; the 99 are on an average much whiter, and in conse-
quence more definitely or sharply marked. The cell-marks are always large
or fairly so, on the under side dark, on the upper side forming faint rings or
ocelli (as in pl. V, figs. 15, 16, etc.), or strong ones (figs. 4, 18, 17, etc.), or solid
black spots (figs. 7, 8, 12, etc.). The ante- and postmedian lines are always
traceable, though varying greatly in distinctness ; the antemedian (as in allied
species) 1s often duplicated proximally in blackish or brown, and very occasion-
ally (as in fig. 15) these two lines may be fused into a narrow band; the
postmedian very occasionally (as in fig. 14) suggests the presence of a similar
companion line distally, but is usually succeeded merely by an indefinite buff,
cinnamon or tawny shade, as in most of the allies, and even this may be
absent. The median line is extremely variable, sometimes (except at the
costa) weak or wanting (see figs. 7, 12, 16), sometimes strong (figs. 14, 15), some-

GEOMETRIDAE. 159

times the strongest and blackest of all the markings (fig. 11) or developed into
a band (fig. 2). Even the tendency shown in some of the allies (e.g. C.
anjectarra Walker and C. immemorata Walker) for this strong median line to
throw out a longitudinal branch across the postmedian in front of R* occa-
sionally manifests itself (fig. 15), or is indicated by blurred or diffuse darkening
(fig. 3). As regards the general distribution of the colours, there is endless
variation ; definitely banded forms may have the median area white and the
rest dark or red-brown (cf. figs. 7 and 17), or—at least on the fore wing
and anterior part of the hind wing—the antithesis of this, the median area being
blackened and the rest more or less pale (see fig. 1); or the white areas may
be partly suffused (figs. 4, 13) or more proximal (fig. 11) or more restricted
(fig. 3), or—somewhat similarly—a dark central band may be restricted to the
area between median and postmedian. An extreme, rare ab. (fig. 6) is almost
entirely blackened, leaving only the base and apex of the fore wing pale. One
& (not figured) has the cell-spots abnormally enlarged. Other phases of varia-
tion, which it is impossible to discuss in detail, may be gathered from a study
of the plate or of the general tendencies of the polymorphism known in the
other most variable Indo-Australian Cleora—inflexaria Snellen.

Finally, mention must be made of some teratological variations. Apart
from the very great discrepancies in size—ranging from 30 mm. or less up to
46 mm., and apparently not seasonal—there are occasional deviations from the
normal shape. Particularly curious is. the fine 9 aberration (pl. V, fig. 11)
in which the right fore wing 1s somewhat shortened, and the right hind wing
definitely enlarged. In a somewhat analogous aberration of the 3g, with broader
black median shading, both hind wings are somewhat produced in the middle,
though not absolutely symmetrically.

Concerning the probable origin of C. samoana, little can be said. It has
nothing directly to do with C. “ acaciaria,” nor even with C. alienaria Walker,
the species which Hampson (Faun. Brit. Ind., Moths, i, fig. 137, 1895) figures
under that name; probably Rebel (loc. cit. swpra) was misled by Hampson’s
absurd “‘ lumping,” and made no independent investigation. From the rather
narrow wings, prevalence of brown colouring, pronounced outward curve of the
postmedian of the fore wing at the radials (one of the very few comparative
constants in its scheme), and strongly clouded or banded under side, one might
have supposed it to be a representative, if not even a subspecies, of C. injectaria
Walker, but the genitalia do not altogether bear out this assumption. I can

160 INSECTS OF SAMOA.

only say that our researches have not shown us anything less far from C. samoana
than this, excepting that the Cleora that occurs in the Tonga Is. is unequivocally
C. samoana. We give figures (Text-fig. 1) of the valvae of topotypical C.
injectaria (Ceylon), the subspecies (?) compactarva from Singapore and the more

A Cc

Text-Fic. 1.—Valvae of males of Cleora injectaria from Ceylon (A); C. injectaria compactaria
from Singapore (B) ; C. anjectaria subsp ? from Fiji (C); C. samoana (D). Drawn by W. H.
T. Tams.

divergent Fiji representative (at present unnamed), side by side with those of
C. samoana. Species with similar facies but still more dissimilar genitalia are
C. lichenina Butler * (Loyalty Is.), C. ummemorata Walker (New Caledonia),
and two small, unnamed species respectively found in Tahiti and the Marquesas Is.

28. Cleora hemiopa, sp. n. (Text-fig. 2 B).

3, 32-34 mm.; 9, 34-36 mm. Distinguishable from C. samoana as follows :
Antenna of 3 pectinate to scarcely §. Valva of 3 with entirely different

* Founded on a fairly large Q which has not yet been matched, but I think that Warren
and I are correct in associating it with the close relative (? form) of C. cmmemorata, to which I
here apply the name, and not with the other known Loyalty Islands Cleora, which is much nearer
to C. decisaria.

GEOMETRIDAE. 161

armature (Text-fig. 2, B). Fovea of g smaller, formed as in C. decisaria Walker,
z.e. on the upper side with a small circular ridge in posterior half (adjoining SM?)
enclosing a rosette of fine and small, sharply black scales, which radiate from
a white centre. Fore wing relatively
a trifle broader, with the termen ap-
preciably less oblique; much less
brown and more weakly marked than
in most specimens of C. samoana, of
a more dead white than the whitest
aberrations thereof, without fuscous or
blackish irroration, in both sexes al-
together recalling weakly marked C.
decisarva, though slightly more tinged
with brown; cell-mark only narrowly
dark-shaded proximally and distally ;
postmedian line more regularly crenu-
late throughout, and without the pro-
nounced loop outward at the radials, Texr-ric. 2.—Valvae of males of Cleora decisaria
in this again rather closely resembling a oe. a akg ee Gee oe
C. decisaria; 2 with a conspicuous ane

black dot in the position of the

specialised patch of the ¢ fovea (as in some C. decisaria 99). Hind wing
similarly pale and weakly marked, with large raised white scales in end of
cell and again beyond, much as in C. decisaria ; a small black cell-dot, not
developed into the usual ocellus of the genus. Under side dirty whitish, with
the cell-spot of the fore wing rather smaller than in C. samoana, that of the
hind wing small and weak, the postmedian of both wings little sinuous, weak
posteriorly, the subterminal dark shades narrow, weak except anteriorly (a
little stronger in the 99 than in the 4).

Upolu: 2 3g, in Zool. Mus. Berlin (Friedlander); Apia, 1 3, 16.v.1923
(Armstrong); type g, 14.v.1924 (Armstrong); 1 3, 24.v.1924; 1 3 (valva
figured), 27.vi.1924; allotype 9, 12.x.1922 (Armstrong); 1 @, 1.x1.1924.

Savai: Palauli, 8.viii.1924.

Apart from the very different armature of the valva (harpe of Pierce, Genit.
Geometr., etc., 1914), I should have regarded this as a race of C. decisaria Walker.
A figure of the valva of a Kei Is. C. decisaria is given for comparison. The

162 INSECTS OF SAMOA.

only species yet known to me having the type of fovea to which attention is
here called are: C. displicata Walker (Queensland : structurally distinct in the
non-dilated hind tibia of the 3, and in the absence of the tufts of raised scaling,
except the slight ones of the cell-spots) ; C. mjdbergi Prout (Borneo: a large
species, with entirely dissimilar facies and also lacking the raised scaling); and
the immediate circle of C. decisarva. The members of the latter agree so closely
in many respects that they must surely have had a common origin ; a specimen
from the Philippines, belonging to an unnamed species, has the central armature
of the valva somewhat similar to that of C. hemeopa, but has a longer and stronger
marginal hooked process than that in C. decisaria ; another (also unnamed)
example, from Lifu, has not been studied anatomically, but may well represent
a weakly marked race of C. decisarva, with a rather strong dark mark on the post-
median of the fore wing at, and just in front of, vein R%.

Orsonoba Walker.

List Lep. Ins., xx, 218, 1860.—Moore, Lep. Ceyl., iii, 394, 1887.—Hampson, Faun. Brit. Ind.,
Moths, iii, 211, 1895.—Turner, Proc. Linn. Soc. N. S. Wales, xliv, 286, 1919.

This genus was erected solely for P. clelia Cramer (under its synonym O.
rajaca Walker, t. cat., 219, Ceylon), which remains the only well-known and
widely distributed species. Its two closest allies are O. varvarva Leech (1897,
W. China), and O. aethocrypta Prout (1927, Upper Burma). A somewhat
more distant relative is O. zapluta Turner (1904, Queensland). “ Hrosia”
hyperbolica Swinhoe, which Hampson (overlooking the pectinate 2 antenna !)
pronounces to be “‘a dwarf (30 mm.) brownish ochreous female” of O. clelva,
and of which the type was taken at Karachi, and two kindred Mediterranean
species (Coenina dentatarca Swinhoe and O. paulusi Rebel) have narrower
wings and are better referred to the African genus Coenina Walker (vide Seitz,
Macrolep., iv, 349, 1915).

In the Malayan and Papuan Subregions, the branchings from Orsonoba,
which I assume from its range to be the phylogenetically older form, have pro-
ceeded along other lines. An endemic Australian genus, Proboloptera (? Meyrick)
Turner (Proc. Linn. Soc. N.S. Wales, xliv, 287, 1919), unknown to me in nature,
is said to differ from it in having the antenna pectinate to the apex, and the
frons with a rounded or conical corneous projection. Another genus, Xylino-
phylla Warren = Adelphocrasta Warren, is almost as widely distributed as

GEOMETRIDAE. 165

Orsonoba itself, from which it differs little except in the simple antennae of both
sexes and some details of shape. Its range, as at present known, is from Penang
to the Solomon Is., and I consider it quite likely to be met with in Polynesia.

It is perhaps necessary to point out that there 1s no possible justification
for transferrmeg the name Gonodontis Hiibner to the present genus, notwithstand-
ing that Warren (Proc. Zool. Soc. Lond., 1893, p. 398, 1893) gave the following
citation: “ Type, G. clelia Cramer.” The diagnoses of the Familia in which
Hiibner places it, and of the Coitus Gonodontis itself, require a pale (not con-
colorous) hind wing, a regular, but dentate, termen to the latter, and the presence
of annular cell-spots on both wings; hence it was quite certainly founded on
G. dentarva Hiibner (G. bidentata Clerck), to which were added, as a kind of after-
thought, the discordant species G. anceta Cramer and G. clelia Cramer. More-
over, Geyer (Zutr. Exot. Schmett., v, 14, 1837), in describing Polygoma cingillaria
and comparing it with “ Gonodontis clelia,’ expressly states that the latter
really belongs to a new genus together with P. cingillaria—t.e. he transfers it
to his Polygonia (nec Hiibner). Since Rogenhofer in 1885 resuscitated Gonodontis
for G. bidentata, in lieu of Odontopera Stephens, the name has very generally
been used in that sense, in which its validity has never been questioned.

29. Orsonoba clelia (Cramer).

Uitl. Kapell., in, (24), 172 and 174, t. 288 B, C, 1780 (Coromandel Coast)—Hampson, Faun. Brit.
Ind., Moth., iii, 212, 1895.

Orsonoba rajaca Walker, List Lep. Ins., xx, 219, 1860 (Ceylon).—Moore, Lep. Ceyl., ui, 395, t. 187,
f. 1, 1b, 1887.

Orsonoba pallida Butler, Ann. Mag. Nat. Hist., (5), vi, 125, 1880; Jl. Het., vi, 53, t. 114, f. 5,

1886 (N. E. Himalaya).

Arrhodia orthotoma Lower, Tr. Roy. Soc. S. Austral., xviii, 83, 1895 (Queensland, type in Mus.
Tring, erroneously identified by Warren as Proboloptera embolias Meyrick).

Upolu: Apia, 1 9, 10.v.1922 (Armstrong).

Tutuila: Pago Pago, 2 §g, x.1923, 1.1924 (Stefiany).

Previously known range : Ceylon, India, Burma, Penang, Hainan, Formosa,
Java, Flores, South Moluccas, Queensland, New Britain, Solomon Is. (Bougain-
ville, Florida, Tulagi, Gizo, Vella Lavella).

In this species sexual demorphism is pronounced and individual variability
moderate, but I have not yet detected any sign of the formation of races. Not-
withstanding its extremely wide distribution, O. clelia is very far from being a
pest, and the scantiness of the Samoan records is in keeping with most of

164 INSECTS OF SAMOA.

our previous experiences of it. The larva is said to feed on Convolvulaceae, and
to be olive-fuscous with small dorsal prominences on the second and eighth
abdominal segments.

Nadagara Walker.

List Lep. Ins., xxiv, 1092, 1862.—Hampson, Faun. Brit. Ind., Moths, ii, 195, 1895.—Turner,
Proc. Linn. Soc. N.S. Wales, xliv, 284, 1919.

An exclusively Indo-Australian genus of moderate extent, the species mostly
pretty uniform in structure and facies, but including a few aberrant forms which
render its demarcation rather difficult. Turner’s diagnosis (loc. cit.) 1s good,
but he had very little material and no g, and was unaware of some points in
which there is variability: (1) the long stalk of SC!-2 by no means always
anastomoses with C; (2) the ¢ hind tibia is dilated in about three-fifths of the
species (“not dilated” was quoted from Hampson). The rather acute apex
of the fore wing, and the more or less wavy distal margins are generally cha-
racteristic. MM? of the fore wing generally arises close to the hind angle of the
cell, but is occasionally (as in N. argyrosticha Turner) considerably more
proximal. In the same species, R? of the fore wing arises well before the middle
of DC, while NV. synodoneura Prout is still more advanced in this respect, having
the origins of R! and R? very close together ; in the latter species it may possibly
be a 3 character only, the © being still unknown.

The previously known range of the genus extended from India to the Bis-
marck Archipelago and the Solomon Is. The variable and interesting Samoan
species 1s, [ suppose, nearest to N. extractata Prout (Louisiade Archipelago).
Since the descriptions of the species are very scattered and the genus has never
been handled as a whole, I offer a provisional key, together with a few supple
mentary notes.

1. Palpus and face-cone moderate (huj. gen. 2). : 3 2
Palpus and face-cone long. . 3
2. Dark, with termen of fone wing Nene Anite . umbrifera Wileman (1910).
Not dark, termen not laine : : cunergera Warren (1906).
3. Postmedian line of fore wing at hind margin less fe) Costa
end of that of hind wing : 4
Postmedian line of fore wing Nerang well before cactal end
of that of hind wing . ; : 16
4, Fore wing with R2 arising near RL (2 Gulenen) ; . synodoneura Prout (1923).

Fore wing with R2 normal . 5 6 3 : ‘ 5

GEOMETRIDAE, 165

5. Fore wing with cell-dot enlarged . f 5 : . orbrpuncta Prout (1925).
Fore wing with cell-dot not enlarged —. : : . 6

6. Rufescent or ochraceous, generally larger : . ; 7
Grisescent, at least in ground-colour, not large ‘ : 10

7. Hind wing with termen dentate, bent in middle . odontias Prout (1925).
Hind wing with termen rounded, at most subcrenulate — . 8

8. Hind wing beneath with dark ae oa and Seats

cell-spot ‘ . sp. n., Samoa.

Hind wing beneath not so. Y

9. Fore wing yellowish, antemedian ine vertical Senet juventinaria Guenée (1858).
Fore wing reddish, antemedian line strongly oblique

posteriorly. : E , . epopsioneura Prout (1926).
10. Postmedian line of both wings denciculate 4 , ‘ ll
Postmedian line of hind wing not denticulate . : : 12
11. Postmedian lines meeting, their denticulation slight . (A) tractata Prout (1916).
Postmedian of hind wing at costa more proximal, denti-
culation of postmedian deeper : : ; . antractata Walker (1862)
12. Fore wing with blackish apical maculation . : 3 15
Fore wing without blackish apical maculation. : : 14
13. Fore wing variegated with reddish subapically 5 . (B) mordinata Walker (1862)
Fore wing not variegated with reddish subapically . . (B) vigaia Walker (1862).
' 14. Fore wing with postmedian line denticulate . : . scitilineata Walker (1862).
Fore wing with postmedian line not denticulate . } 15
15. Fore wing 18 mm. long, with apex produced . . . dohertyi Prout (1925).
Fore wing not more than 15 mm., with apex not produced comprensata Walker (1862).
16. Grisescent or variegated : : : : ; ; 17
Rufescent or avellaneous —. : 19
17. Fore wing with apex produced, poe rnedia (He cfierdle : 18
Fore wing with apex not appreciably produced, post-
median line duplicated proximally . : synocha Prout (1923)
18. Under side with strong subterminal dark shades, ait feos
apically : : . (A) eatractata Prout (1923).
Under side without strong cnlyesumtiell dork phades! . (A) wretracta Warren (1899).
19. A dark line indicated midway between subterminal and
termen f argyrosticha Turner (1919),
No dark line indicated Sdyey betel eninienaninall te
termen 5 : : 20
20. Wings subcrenulate, beneath Pal Taro 6 reprensata Prout (1916).
ibane wings smooth- -margined, redder, beneath realy
marked : 5 ‘ : d : : . extensipennis Prout (1923).

Note (A).—These three are probably races, in some measure linked up by others not yet
named ; JN. zrretracta is known to occur in the Bismarck Archipelago, and in Gizo and Tulagi
(Solomon Is., off Isabel), and the Queensland form would work down to it in our key ; N. extraciata
represents it in the Louisiade Archipelago, and N. tractata (Rook I.) may be a further race. The
new Samoan species is probably developed from this, the most easterly, group.

Nore (B).—I am coming increasingly to think that N. vigata and N. inordinata are merely
forms of a single species. The vigaia forms are most frequently (though not invariably) 9, the
mordinata forms more often 3. Distribution, according to present determinations, seems

it (3) 4

166 INSECTS. OF SAMOA.

somewhat confused: Ceylon, Nigiris, Bhutan, Java, and Sumatra for N. vigaia; Sikkim,
Assam, l‘ormosa, Malaya, Borneo, and Celebes for N. cnordinata.

Grouped first by the structure of the 9 hind leg and then regionally, the
forms of Nadagara present themselves as follows :

I. Hind tibsa of 3 not dilated :

Formosan: NV. umbrifera (huj. gen. 2).

Inpo-Matayan: UN. vigaia and N. wmordimata (see note B, supra), N. orbi-
puncta (N.W. India), VN. epopsioneura (N.E. India), NV. synocha (Malaya, Java,
Sumatra), NV. iitractata (Borneo, Singapore).

Papuan: NV. cuneigera (huj. gen. *%).

Il. Hind tabia of 3 dilated -

Matayan: JN. comprensata (Andaman Is., Nicobar Is., Penang, Borneo,
Toekan Besi), N. scotilineata (Borneo, Perak, Hainan), N. odontias (Perak),
N. reprensata (Java, Perak).

Motuccan: NV. doherty (Batjan).

MeLAnesiAn: JN. eatractata, N. tractata and N. vrretracta (see note B, supra),
N. extensipennis (British New Guinea, Rook I.), N. synodoneura (Dutch and
British New Guinea).

AUSTRALIAN: ? NV. argyrosticha Turner (Queensland)—3 unknown, but the
species almost certainly belongs to the same group as N. extensipenms and
N. synodoneura.

POLYNESIAN: sp. n., Samoa.

From the above outline, it is clear that there are still many lacunae in our
knowledge of the geographical distribution of the known species, as indeed is
the case with all but a few of the most conspicuous of the Indo-Australian
genera of Geometridae. I think, however, that we already know enough to
support the view that the higher “ secondary sexual’? developments in this
genus are eastern and the simpler forms. western, and the consequent expecta-
tion that any further Polynesian Nadagara that may be discovered will belong
to our Group I1—probably, if one may hazard a more definite guess, to the
vicinity again of NV. wretracta. It will be noticed that of the eight species recorded
from the Malay Peninsula and the Greater Sunda Is., four belong to each group,
whereas all the four known to occur in India are in Group I and all the Melanesian

GEOMETRIDAE. 167

in Group II, excepting only N. cuneigera, in which even the wing-pattern, as
well as the structure of the head, suggests an interloper. Unfortunately I have
been obliged to omit from these statistics one Bornean species, since I know
only 99 of it; this is N. guventinaria Guen. (Spec. Gen. Lép., ix, 103, 1858) =
pulchrilineata Walk. (List. Lep. Ins., xxiv, 1095, 1862).

The following, which were described under Nadagara, are to be excluded
from the genus: WN. grisea Butler (1883), NV. punctilinearva and N. albovenaria
Leech (1897), N. diversilineata Warren (1896), NV. obrussata Holland (1900), and
N. camura Joicey and Talbot (1917).

30. Nadagara hypomerops, sp. n.

42, 35-37 mm. Head and front of thorax reddish-brown, more or less
intermingled with darker colourmg. Palpus rather over 2, with the terminal
segment markedly elongate, especially in the 9. Thorax and abdomen pale
(pinkish-buff or pinkish-cinnamon), more or less mixed, especially in the 9,
with red-brown. Hind tibia in § dilated, with hair-pencil, the abdominal spine
long and slender.

Fore wing moderately broad, termen not extremely oblique, waved to
faintly crenulate, with the concavity between apex and SC° rather pronounced ;
stalk of SC!-2 anastomosing well with C; rather variable in colour, as far as the
postmedian line (especially in the central area) light ochreous in the 3, more
tawny in the 9, the distal area in the 3 inclining to pale vinaceous drab, in the
2 generally scarcely differentiated, in both sexes with some dark-grey irroration ;
costal edge pale, with dark dashes and longer streaks ; cell-dot blackish; ante-
median line variable, rarely well expressed (sometimes almost entirely obsolete,
nearly always so anteriorly), very oblique outward to cell-fold, then more or less
oblique inward ; postmedian white, finely dark-edged proximally, anteriorly about
2 mm. from termen, from R! becoming oblique, wavy and with very slight
inward curves between the radials and between R? and SM?, reaching hind marein
at about $; terminal line fine, becoming thicker and blacker in front of SC5 :
fringe slightly reddened, with indications of a pale line at base and pale tips,
sometimes dark-spotted opposite the veins. Hind wing with termen somewhat
crenulate in the 3, more so in the 9; concolorous with fore wing, extreme base
pale, in strongly marked specimens bounded by a continuation of the ante-
median line; cell-dot enlarged, but ill-defined ; postmedian line straightish,
httle beyond middle, continuing that of fore wing ; traces of a wavy subterminal.

168 INSECTS OF SAMOA.

Under side variegated and variable, always with strong dark strigulae ;
sround-colour whitish, showing chiefly in a broad band outside the postmedian
(which on fore wing is straightish and little more oblique than termen, reaching
costa about 5 mm. from apex), also on the fore wing in an apical dash, on the
hind wing in a terminal band between R* and tornus; the rest suffused with
shades of yellow-brown or red-brown, on the hind wing with a large, dark greyish-
brown apical patch, which is otherwise only developed in N. tractata Prout,
though foreshadowed in NV. antractata Walker ; cell-dots sharply black, on the
hind wing large.

The most outstanding aberration—l 9, 4 99—has a strong fuscous shade
proximally to the postmedian, broadest on the under side, and on the upper
side a narrower one distally to the antemedian. One or two 29 are slightly
transitional.

Upolu: Malololelei, type g, 16.vi.1924; 1 9 ab., 21.1v.1925; 1 9 ab.,
12.11.1924 ; 1 9 ab., 20.vi.1924 ; allotype 9, 21.vi.1924; 2 99, 25, 28.v1.1924 ;
1 © ab., 2.vn.1924 (Armstrong); 1 9, vu.1925 (Wilder); 2 99, 22.x1.1924 ;
1 2, intermediate, 2,000 feet, 5.xi1.1925; Apia, 1 typical 9, 1 9 ab., x.1924.

A coloured figure of this species will appear in Part III, Fascicle 4.

EXPLANATION OF TEXT-FIGURES AND PLATE.

Text-fig. 1. Valvae of males of Cleora injectaria from Ceylon (A); C. imjectaria compactaria
from Singapore (B) ; C. injectaria subsp % from Fiji (C); C. samoana (D). Drawn
by W. H. T. Tams.

2. Valvae of males of Cleroa decisaria from Key Islands (A); C. hemiopa from Upolu,
Samoa (B). Drawn by W. H. T. Tams.

Plate V. Aberrations of Cleora samoana, all taken at Vailima, Upolu, Samoa, 7th September,

1925 (G. H. E. Hopkins).

29

a a a hen
PRINTED IN GREAT BRITAIN BY WILLIAM CLOWES AND SONS, LTD., LONDON AND BECOLES.

(oxi leastsM1) stolid OONHONNZ pvoo!0

H. J. Cambell, photo.
PART III. ; PEATE V.

168 ; INSECTS OF SAMOA.

Under side variegated and variable, always with strong dark strigulae ;
sround-colour whitish, showing chiefly ina broad band outside the postmedian

(which on fore wing is straightish and little more oblique than termen, reaching -
costa about 5 mm. from apex), also on the fore wing in an apical dash, on the
hind wing ma terminal band between R? and tornus; the rest sufiused with

shades of yellow-brown or red-brown, on the hind wing with a large, dark greyish-
brown apical patch, which is otherwise only developed in JN. éractata Prout,

though foreshadowed in N: intractata Walker; cell-dots oe black, on the
hind wing large.

The most outstanding aberration—i g, 4 99—has a etons fuscous shade:
pr ximaliy to the postmediampyhrqyresp,on the under side, and on the upper
side a narrower one distally to the antemedian. One or two 99 are slightly
transitional. ) |

Upolu: ‘Malololelei, type 3, 16.vi.1924; 1 3 ab., 2l.iv. 1925; 1 2 ab...

)

12.11.1924 :,1 9 ab., 20.vi.1924 : all type Q, 21.v1. Wee ‘ QO, 25, 28.v1.1924 ;
i QO. ab. 2 vAbersationg of Cleora semoana, Butler, oes FWi ide) 2 90, 22.xi.1924 ;
1 9, anterelediate 2 2,000 feet, 5.xn.1925; Apia, 1 typical 9, 1 2 ab., x.1924.

A coloured figure of this species will appear in Part 111 Fascicle 4...”

EXPLANATION OF TEXT-FIGURES AND PLATE,

Text-fig. 1. Valvae of males of .Cleora wyectarza from Ceylon (A); C. injectaria compactaria
from Singapore (B) ; C. injectaria subsp ? from Fiji (C) ; C.samoana(D). Drawn
by W. H. T. Tams. ee.

2. Valvae of males of Cleroa decisaria from Key Islands (A); C. hemiopa from Upolu,.
Samoa (B). Drawn by W. H. T. Tams.

Plate V. Aberrations of Cleora samoana, all taken at Vailima, Upolu, Samoa, 7th September,

1925 (G. H. E. Hopkins). :

29

PRINTED IN GREAT BRITAIN BY WILLIAM CLOWES AND SONS, LTD., LONDON AND BEOOLES.

Sus

vie

7

BRIT. MUS. (NAT. HIST.) INSECTS OF SAMOA.

H. Ji: Gembell photo.
PART III. PLATE V.

my

PART MEN

DE

OF |

INSECTS OF SAMOA

owe OTHER SAMOAN TERRESTRIAL
ARTHROPODA

PROPOSED ARRANGEMENT :—
Part I. Orthoptera and Dermaptera.
coe ll; Hemiptera.
a aie Lepidoptera. |
,, IV. Coleoptera.
_,, V. Hymenoptera.
; pee VI. Diptera.
aes » VII. Other Orders of Insects.
_,, VIII. Terrestrial Arthropoda other than Insects.

‘The work will be published at intervals in the form of numbered fascicles.
fe Although individual fascicles may contain contributions by more than one
author, each fascicle will be so arranged as to form an integral portion of one or
_ other of the Parts specified above.

Fehr ey
a