Systematic notes on Palearctic birds. No. 32, Oriolidae, Dicruridae, Bombycillidae, Pycnonotidae, Nectariniidae, and Zosteropidae

Amencan Museum

iovitates

PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY
CENTRAL PARK WEST AT 7QTH STREET, NEW YORK 24, N.Y.

NUMBER 1869 JANUARY 31, 1958

Systematic Notes on Palearctic Birds. No. 32
Oriolidae, Dicruridae, Bombycillidae,
Pycnonotidae, Nectariniudae,
and Zosteropidae

By CHARLES VAURIE

The following notes were made during a study of the families named
above for a contemplated check list of the birds of the Palearctic re-
gion. Among the species included in this paper, those discussed or
reviewed in greater detail are: Oriolus oriolus, Hypocoltus ampelinus,
Microscelis amaurotis, Pycnonotus leucotis, and Zosterops japonica.

I would like to express my gratitude to Mr. H. G. Deignan, Dr. G. F.
Mees, Dr. A. L. Rand, and Dr. R. W. Storer who have kindly helped
me in various ways. Mr. Deignan has read the manuscript dealing with
the Pycnonotidae, and he and Dr. Rand have furnished me with their
notes on the Palearctic members of this family. Dr. Mees sent me his
notes on the Palearctic Zosteropidae, and Dr. Storer lent material col-
lected by Dr. Koelz from the collection of the Museum of Zoology of
the University of Michigan.

ORIOLIDAE
Oriolus oriolus

The Golden Oriole has been reviewed by Meinertzhagen (1923,
Ibis, pp. 57-61) who recognizes two subspecies: nominate oriolus Lin-
naeus, 1758, type locality, Sweden; and kundoo Sykes, 1832, type local-
ity, Deccan, peninsular India. These two are very clearly differentiated.
Nominate oriolus is larger, the wing length of 20 fully adult males of
each race measured by me being 149-162 (154.5) in nominate ortolus

2 AMERICAN MUSEUM NOVITATES NO. 1869

as against 136-144 (141) in kundoo. The wing formula shows also a
clear-cut difference worthy of emphasis, although it has been ignored
by virtually all authors. This difference consists in the fact that the
second outer primary is longer than the fifth in nominate oriolus, often
by as much as 1 cm., whereas the reverse is true in kundoo. In addi-
tion, the fourth is slightly shorter than the third in nominate oriolus,
whereas it is equal to the third or slightly longer in kundoo.

There are also some clear-cut differences in coloration: In adult
males, the region behind the eye is black in kundoo, but in nominate
ortolus the black is restricted to the lores; in kundoo both sexes show
larger yellow areas in the tail and on the primary coverts; and female
kundoo is yellow or yellowish above and below when fully adult,
whereas female nominate oriolus is greenish above, grayish on the
throat, and whitish on the abdomen.

Nominate oriolus breeds throughout the greater part of continental
Europe, as well as in northwest Africa though in small numbers, in
Siberia east to the Yenisei, Iran, southwestern Transcaspia, and north
of kundoo in Russian Turkestan. The latter ranges from Russian and
Chinese Turkestan southward through the northern and eastern half
of Afghanistan to northern Baluchistan and India. It has been sug-
gested that kundoo and nominate oriolus are perhaps full species, and
both have been reported as breeding in the western Tian Shan in
Chinese Turkestan. However, according to Ludlow (1933, Ibis, pp.
658-659) the two replace each other altitudinally in this region, nom-
inate oriolus breeding at higher altitudes than kundoo which inhabits
the plains.

In view of the fact that these two races are so well differentiated, it
is not desirable to recognize other forms which have been described
but differ only very slightly. These forms are: caucasicus Zarudny, 1918,
type locality, northern Iran; turkestanica Zarudny and Kudashey, 1918,
type locality, Russian Turkestan; and sibiricus Hans Johansen (1944,
Jour. Ornith., vol. 92, p. 29), type locality, Krasnoyarsk region, Siberia.
An additional form has been described. It was named baltistanicus by
Koelz (1939, Proc. Biol. Soc. Washington, vol. 52, p. 72) with type
locality, Baltistan, northern Kashmir, but it appears to be a synonym
of kundoo.

Caucasicus and turkestanica were separated on size differences, cau-
casicus as smaller than nominate oriolus, and turkestanica as larger
than kundoo. Zarudny gave no measurements in his description of
caucasicus, but Stresemann (1928, Jour. Ornith., vol. 76, p. 346) states
that three adult males from northern Iran have a wing length of 145,

1958 VAURIE: PALEARCTIC BIRDS, NO. 32 j 3

147, 148, as against 150-161 in males from Europe, and recognizes
caucasicus. However, I have measured males from Europe with a wing
length of 149, and Niethammer (1937, Handbuch der Deutschen Vogel-
kunde, Leipzig, vol. 1, p. 42) has measured some from Germany with
a wing length of 147. It seems to me, therefore, that while birds from
northern Iran are somewhat smaller, the difference is too slight to
warrant separation.

Zarudny states that the wing length of turkestanica is 1 cm. longer
than that of kundoo. This is not confirmed by specimens that I have
measured from Turkestan, Afghanistan, and India. In males, specimens
from Turkestan measure 140, 140, 141; from Afghanistan, 140, 141, 141,
143, 144; and from peninsular India, 138, 138, 139, 141, 141, 142, 142,
144. Dementiev (1933, L’Oiseau, p. 749) states that males from Turkes-
tan measure 141-148, and it is probable that a cline of increasing size
runs from India northward to Turkestan; but, judging by the measure-
ments available, this cline does not seem sufficiently sharp to warrant
the recognition of turkestanica. Korelov (1954, Birds of the Soviet
Union, vol. 5, pp. 143-153) has synonymized caucasicus with nominate
ortolus and turkestanica with kundoo. He does not mention sibiricus
which was separated from nominate oriolus by Johansen on the
ground that it is larger, “males 153-165, as against 150-160,” and purer
and brighter yellow, less dull and greenish. The overlap in measure-
ments is very great, and I doubt that the difference in coloration is
sufficiently well marked and constant to warrant the recognition of
stbirtcus. Many males that I have examined from Europe were neither
dull nor greenish. As stated above, it seems misleading to recognize
such slightly differentiated forms as caucasicus, turkestanica, and sibiri-
cus. It is sufficient to mention the trends in the geographical variation
represented by these forms without resorting to nomenclatural separa-
tion.

Baltistanicus was based on six specimens (only one of which was
adult). It was separated by Koelz from kundoo and turkestanica on
color differences and as being smaller than the latter. I have not seen
specimens from Baltistan, but I cannot discern any color differences
whatever between specimens in comparative plumage from other parts
of Kashmir, Gilgit, Afghanistan, Turkestan, and northern and penin-
sular India. Whistler (1942, Jour. Bombay Nat. Hist. Soc., vol. 43, p.
36), who has examined some, states, “Specimens from Baltistan in my
own collection are not separable from O. 0. kundoo’ and believes
baltistanicus was based on individual variants. Koelz states that the
three specimens he used for comparison, and believes are turkestanica,

4 AMERICAN MUSEUM NOVITATES NO. 1869

consist of one female with a wing length of 154 collected on September
10 in Afghanistan and two males with a wing length of 160 collected
on September 20 and 21 in Ladak. However, as shown above, birds
from Turkestan do not have such an extraordinarily long wing, and
these measurements correspond with those of nominate oriolus from
the eastern part of its range. I did not examine the two males, but I
find that the female is nominate oriolus as shown incontrovertibly by
its typical coloration and wing formula. I believe it is likely that Koelz
misidentified the two males also and that they are likewise nominate
ortolus. All three specimens were collected at the height of the migra-
tion, and nominate oriolus, though it winters in Africa, occurs on pas-
sage as far east as Baluchistan and northwestern India.

DICRURIDAE

The Dicruridae are not a Palearctic family, but nevertheless three of
the 20 species are represented in the Paleartic region: the King Crow
(macrocercus), the Ashy Drongo (leucophaeus), and the Spangled
Drongo (hottentottus). The first invades the Palearctic region most ex-
tensively, with two subspecies out of a total of seven occurring there.
These are albirictus, which breeds as far west as southeastern Iran, and
cathoecus, which breeds as far north as southern and perhaps central
Manchuria. Three of the 14 subspecies of leucophaeus breed in the
Palearctic region: longicaudatus in Afghanistan, hopwoodi in Sikang
and up to 13,000 feet in northern Yunnan, and leucogenis in northern
China and southern Manchuria. In hottentottus only one of its 30 odd
subspecies breeds in the Palearctic region, this subspecies (brevirostris)
breeding in China north to northern Hopeh.

The family has been reviewed by me in detail (1949, Bull. Amer.
Mus. Nat. Hist., vol. 93, pp. 199-342), and a study of its evolution has
been published by Mayr and me (1948, Evolution, vol. 2, pp. 238-265).
I have now reviewed the Palearctic forms again and made the following
notes.

Dicrurus macrocercus

In 1949 I did not include Iran within the range of albirictus, men-
tioning only a record of Zarudny (1911, Jour. Ornith., vol. 59, p. 221)
who had listed this species as breeding in Persian Baluchistan. I was
not then familiar with the localities where Zarudny had collected,.but
the two specimens that I listed (p. 236) as being from “Northern
Baluchistan” were collected by him on July 23 and 24, 1898, on the
Rud i Bampur in Persian Baluchistan, not Baluchistan proper. These
two specimens, one of which is, I find, a very young bird, establish with

1958 VAURIE: PALEARCTIC BIRDS, NO. 32 5

certainty that the species breeds in Iran. The nearest point at which the
Bampur River approaches the frontier of Baluchistan proper is about
200 kilometers.

Koelz (1954, Contrib. Inst. Reg. Explor., no. 1, p. 15) has described
a new subspecies from Assam, naming it tstpi. Its type locality is Palas-
bari in the Brahmaputra Valley. He states that in ts¢pz the bill is larger
and the tail averages longer than in albirictus from the Himalayas, and
also that “The gloss is more green; the white rictal patch smaller; the
inner surface of wing feathers darker” and that tsipi differs from the
smaller cathoecus of China by having a larger bill but approaches the
latter in the color of its gloss and wings. Thanks to the courtesy of Dr.
R. W. Storer, I have been able to examine 10 topotypes of tsipz col-
lected by Koelz, some of which are paratypes, and I find that tsp: is a
very poorly differentiated intermediate between cathoecus and albiric-
tus. I believe it should be synonymized with the latter.

Koelz gave no measurements, but I find that among the 10 topotypes
the six that are adult measure: males, wing, 154, 155, 156, 161; tail, 175,
175, 177, 186; bill, 26, 27, 28, 29; females, respectively, 147, 151; 163,
172; 25, 27. As stated in my review of the family, albirictus measures:
males, wing, 146-159 (153.3); tail, 157-185 (170.4); bill, 24.5—29 (26.27);
females, wing, 148-150 (149.5); tail, 150-176 (165.5); bill, 25-28 (26.4).
Cathoecus measures: males, wing, 142-150 (146.0); tail, 140-154 (145.4);
bill, 25-28 (26.3); females, wing, 134-149 (143.3); tail, 136-150 (141.3);
bill, 24.5-27 (25.8). We see that the wing and tail measurements of
tstpi are similar to those of albirictus but that the measurements of the
bill in all three forms overlap too much to be of diagnostic value. In
the three forms the range of individual variation in the thickness and
width of the bill is the same.

As Koelz states, the coloration of tsipi (gloss and color of the wing
lining) approaches that of cathoecus, but the gloss is not a very good
character, as it is very similar in cathoecus and albirictus, though
averaging a little duller and more greenish in cathoecus. The best
difference in coloration is the color of the wing lining which is pale in
albirictus and distinctly darker in cathoecus. In two of the six adult
tsipi, the color of the wing lining is about as dark as in cathoecus, and
in the others it is slightly paler, but not so pale as in albirictus. We see,
therefore, that tszpz is similar in size to albirictus but closer to cathoecus
in coloration.

As I remarked in my monograph, cathoecus and albirictus are con-
nected by intermediate and intergradating populations in northern and
western Burma. The specimens collected by Koelz now show that the

6 AMERICAN MUSEUM NOVITATES NO. 1869

zone of intergradation extends farther west and includes Assam, at least
up to the Brahmaputra Valley. Judging by the specimens collected by
Koelz and other older specimens from the valley, it seems best to refer
the populations of Assam to albirictus. They are not sufficiently well
differentiated to warrant their nomenclatural separation.

Dicrurus leucophaeus

The exact status of the Ashy Drongo in peninsular and southern
India is not known. Ticehurst (1936, Ibis, p. 276) has stated that it “has
not been proved to breed in India proper outside the Himalayas. Birds
obtained in the Peninsula of India are, so far as is known, winter
visitors there.” Salim Ali was unable to throw light on this question in
the most recent work on the birds of southern India (1953, The birds
of Travancore and Cochin, London, Oxford University Press, p. 106),
stating that it is “presumably only a winter visitor’ to southern India.
In my 1949 paper I stated that there was some evidence that the bird
breeds south of the Himalayas, but this evidence is so meager and so
uncertain that it is not conclusive.

We are left, however, with the fact that some specimens collected in
southern and peninsular India are smaller than the birds breeding in
the Himalayas. In the specimens I measured, adults from peninsular
and southern India had a wing length of 128-146 (135.88) as against
137-147 (141.17) in males and 135-143 (139.0) in females collected in
the Himalayas. The birds from the peninsula and south are also some-
what paler and more bluish, less slaty, but, as I mentioned in 1949, the
difference in coloration is slight, and the measurements overlap. In
view of the fact that we do not know the origin of these smaller and
paler birds, and all the differences are slight, after all, it seems wiser
to me now not to recognize any subspecies, at least until we are certain
that the bird breeds south of the Himalayas. I consider therefore that
beavant Vaurie (described in my 1949 paper with type locality eastern
Afghanistan) is best synonymized with longicaudatus Jerdon, 1862,
type locality, Nilgiris. Another synonym is minimus Baker, 1918, type
locality, Ceylon, based on winter visitors to Ceylon of small size, while
beavani was based on large birds.

Dicrurus hottentottus

Vorobiev (1954, Birds of the Ussuri Region, Moscow, Akademia
Nauk, pp. 213-214) has published a record of D. hottentottus brevt-
rostris, collected on November 9, 1947, in southern Ussuriland. He
states also that D. macrocercus cathoecus was collected on May 26, 1944,

1958 VAURIE: PALEARCTIC BIRDS, NO. 32 7

in southern Ussuriland. These two records are the first for Ussuriland
and are interesting because they are the northernmost for these two
species, though Swinhoe stated long ago that cathoecus reaches the
Amur River on spring migration, but I do not know the basis for his
statement.

In 1955 (Ibis, pp. 153-155) I published a short study of the geo-
graphical variation of hottentottus in India. Differences in coloration
are very slight and not constant, and any nomenclatural separation
must be based on size differences alone. However, a critical study of the
measurements shows that the variation is clinal and that there is much
overlap in measurements. It is best, therefore, to synonymize chrishna
Gould, 1836, type locality, Nepal, and londae Koelz (1939, Proc. Biol.
Soc. Washington, vol. 52, p. 70), type locality, Londa, southern Bombay
Presidency, with nominate hottentottus Linnaeus, the correct type
locality of which, as shown by Stresemann (1952, Ibis, pp. 517, 521), is
Chandernagor, southern Bengal.

In 1949 I recognized chrishna but not londae. The latter was sub-
sequently recognized by Ali (loc. cit.) but without giving reasons for his
decision. As stated in my 1955 paper, I have examined the type and
paratypes of londae, and the measurements of these and other speci-
mens from southern India overlap a great deal the measurements of
topotypical nominate hottentottus. The difference in average is very —
trivial; the former measure 156-168 (163.4) and topotypes of nominate
hottentottus, 162-170 (165.0).

BOMBYCILLIDAE
Bombycilla garrulus

The Bohemian Waxwing varies geographically, but this variation is
slight or relatively slight, and the question has been raised whether
subspecies can be recognized. The three currently accepted are nomi-
nate garrulus Linnaeus, 1758, type locality, Sweden, for the populations
of western Eurasia; centralasiae Poliakov, 1915, type locality, Zmein-
ogorsk, Russian Altai, for those of eastern Eurasia; and pallidiceps
Reichenow, 1908, type locality, British Columbia, for those of North
America.

Stegmann (1931, Jour. Ornith., vol. 79, p. 184) recognized central-
asiae, stating that it was “merklich grauer und dunkler” than nominate
garrulus but the word “dunkler” seems to be an error, where “blasser’’
was meant, as it is generally admitted that if centralasiae differs at all
from nominate garrulus it differs from it by being paler, not darker.
Stegmann later came to doubt the validity of centralasiae, as shown

8 AMERICAN MUSEUM NOVITATES NO. 1869

by his correspondence with Meise who did not recognize centralasiae
(1934, Abhandl. Ber. Mus. Dresden, vol. 18, no. 2, p. 35), and also with
Hartert and Steinbacher (1934, Die Voégel der palaarktischen Fauna,
suppl. vol., p. 223) who state that Stegmann informed them that in
large series not less than 50 per cent of the specimens from the east are
identical with those of the west. Finally, Spangenberg in the “Birds of
the Soviet Union” (1954, vol. 6, p. 62) has synonymized centralasiae
with nominate garrulus, while Hartert and Steinbacher koe: cit.) also
have questioned that pallidiceps is valid.

Good series examined by me show, however, that it is possible to
separate the three races, though the differences are not perfectly con-
stant and are slight between centralasiae and nominate garrulus, but
better indicated in the case of pallidiceps. I find that nominate garrulus
is the most richly colored, more vinaceous above and below, while
centralastae is the palest, especially on the rump and upper tail coverts.
Pallidiceps is grayer, less vinaceous above, than nominate garrulus, and
darker and grayer on the flanks and abdomen, and quite distinctly
darker and grayer above and below than centralasiae. Nominate gar-
rulus and pallidiceps are similar in size, while centralasiae averages
slightly larger. The wing lengths of 10 adult males of each are: nomi-
nate garrulus, 113-119 (116.2); pallidiceps, 113-120 (116.5); central-
asiae, 116-122 (118.7).

Hypocolhius ampelinus

The systematic position of the Gray Hypocolius has puzzled so many
authors that scarcely any two have placed it in the same family. It was
described by Bonaparte as Hypocolius ampelinus and placed by him
in the waxwing family, and Delacour and Amadon (1949, Ibis, pp. 427-
429) conclude, after a discussion of its characters and habits, that Bona-
parte was correct, an opinion supported by Meinertzhagen (1954, Birds
of Arabia, London, Oliver and Boyd, p. 175).

Its distribution is not very well known. Hypocolius ampelinus occurs
on the African coast of the Red Sea but, as Meinertzhagen (loc. cit.)
states, it is uncertain whether it breeds there or occurs only as a vagrant.
If a vagrant, it probably arrives from Arabia, where it is resident and
breeds in the west and perhaps other parts of the peninsula. It breeds
in Iraq in the valleys of the ‘Tigris and Euphrates from Mosul south to
Fao. The population of Iraq is migratory, or partly so, migrating per-
haps to northeastern Africa, as suggested by Meinertzhagen (1924, Ibis,
p. 613), though it is possible that most birds go no farther than Arabia,
where they have been reported from Aden, Oman, and other localities.

1958 VAURIE: PALEARCTIC BIRDS, NO. 32 9

It breeds also in southern Iran in Khuzistan and along the coast of the
Persian Gulf, according to Zarudny, and it may breed in Baluchistan
and southern Afghanistan where it is reported from the Makran coast,
Kalat, and the region of Kandahar. Finally it straggles, but very rarely,
to Sind and has once reached Bombay.

Koelz (1939, Proc. Biol. Soc. Washington, vol. 52, p. 64) has described
as orientalis a series of four adult males and two females collected at
Kandahar on October 21-24, 1937, stating that they differed from the
- material in the Rothschild Collection from Fao in southern Iraq
“which may be considered the type locality of ampelinus”’ by being
“darker above, especially on the crown where in males there is a dis-
tinct bluish cast; they have a smaller bill and longer wing.’”’ However,
the material examined by Koelz was not in comparative condition,
and I do not believe that orientalis is valid, a conclusion reached
already by Meinertzhagen (1954, loc. cit.) who states that a specimen
from Kandahar in his collection “exactly matches Iraq specimens.”

The specimens collected by Koelz in Afghanistan are in extremely
fresh plumage, some of them just completing the molt, whereas the
specimens from Fao are old, very faded, and extremely worn skins col-
lected in June and July, 1885 and 1893. The tips of their wings are
badly worn, except for one specimen in which the wing is of the same
length as in the specimens from Afghanistan. A male in fresh plumage
collected on November 12, 1953, in northeast Arabia and which, there-
fore, was not seen by Koelz, matches the specimens from Afghanistan in
coloration and size, including the size of the bill. In the specimens from
Fao, the bill averages slightly larger.

As stated above, the specimens of orientalis were collected at the end
of October and cannot be assumed to be local birds. In fact it is quite
uncertain that the species breeds in Afghanistan or Baluchistan. ‘Tice-
hurst (1926, Jour. Bombay Nat. Hist. Soc., vol. 31, p. 694) states that it
is only a rare straggler in Baluchistan, and Whistler (1944, Jour. Bom-
bay Nat. Hist. Soc., vol. 44, p. 517) states that the only records for
Afghanistan are the specimens collected by Koelz, though, as we have
seen, at least one other specimen has been collected in Afghanistan.

The type locality of ampelinus Bonaparte is “California,” and
Heuglin (1868, Ibis, pp. 181-183) who was the first reviser, corrected it
to “the low Abyssinian coast-region.”” Heuglin had specimens from
Massaua on the coast of Eritrea, and we know for a fact that the species
occurs on the Red Sea coast of Africa, north to at least Jebel Elba on
the border of Egypt and the Sudan, from where it is reported by
Meinertzhagen (1954, loc. cit.). However, Grant and Mackworth-Praed

10 AMERICAN MUSEUM NOVITATES NO. 1869

(1943, Bull. Brit. Ornith. Club, vol. 64, pp. 25-26), and Koelz, as men-
tioned above, have shifted it to Irag, Grant and Mackworth-Praed
restricting it to the “River Tigris’ [Mosul] and Koelz to Fao. Berlioz
(1956, Bull. Mus. Hist. Nat. Paris, ser. 2, vol. 28, pp. 177-179) has
discussed the question of the correct type locality and inclines for
Mosul, but as it is not certain, as Berlioz recognizes, that the specimens
in the Paris Museum from “Sennaar’’ were collected in Iraq and are
the types of ampelinus, I see no conclusive reason to change the type
locality as it was fixed by Heuglin, the first reviser.

Measurements of the wing length of adults are: Afghanistan, Kan-
dahar, males, 101, 104, 105, 105, females, 96, 101 in the specimens
collected by Koelz, and 101 in one male reported by Meinertzhagen
from Kandahar. Iraq, males, 98, 98, 100, 100 (all worn); females, 100
(worn), 103. Northeast Arabia, latitude 27° N., longitude 45° 50’ E.,
male, 103.

PYCNONOTIDAE
Microscelts amaurotis

The Brown-eared Bulbul ranges from Hokkaido southward through
the main and satellite islands of Japan to the Ryu Kyus, the small
islands north of Luzon, and the Borodinos, Bonin, and Volcano
Islands, breeding also in a restricted region (Koshun district) at the ex-
treme southern tip of Formosa. Its geographical variation is very
strongly clinal in character. A cline of decreasing size accompanied by
one of increasing saturation runs southward, though the former is
reversed in the southern Ryu Kyus in stejnegeri of Ishigaki, Iriomote,
and Yonaguni. The populations of the Borodinos (borodinone),
Bonins (squamiceps), and Volcanos (magnirostre) are not part of the
clines but seem to be more or less closely related to ogawae of Amami
O Shima; a “cline,” however, runs from the Bonins to the Volcanos.

Altogether 11 races are recognized by the “Hand-list of the Japanese
birds” (1942, pp. 47-49). To these must be added three races (batan-
ensis, fugensis, and camiguensis) found on the small islands north of
Luzon, and septentrionalis Dementiev and Gizenko (1950, Doklady
Akad. Nauk, new ser., vol. 70, p. 1081) which is based on three speci-
mens collected in extreme southern Sakhalin on November 6, 12, and
29, 1948.

The material that I have examined is fairly comprehensive and,
though it does not include specimens of four forms, permits a brief re-
view of the species. It seems to me that too many stages on the cline
have been described. The ones I believe it sufficient to recognize are the
following (not including the three races found north of Luzon):

1958 VAURIE: PALEARCTIC BIRDS, NO. 32 1]

1. Microscelis amaurotis hensoni Stejneger, 1892, type locality, Hok-
kaido, with septentrionalis Dementiev and Gizenko, 1950 (loc. cit.),
type locality, mouth of the Suzuya River, as a synonym. Henson breeds
only in Hokkaido and represents the northern end of the cline, being
the palest, grayest, least brown, and largest race. However, individuals
with a long wing are found on the Borodinos, Bonins, and Volcanos
(see below), and it is possible that these populations, or at least the last
two, are about as large as hensoni, but the material that I have meas-
ured of hensoni and from the Bonins and Volcanos is insufficient to be
certain. Hensoni does not have the largest bill, the bill being consid-
erably more massive, broader, and thicker in magnirostre and squami-
ceps, and even in stejnegeri, though there are no constant differences
in length, or only slight ones.

The three specimens collected in Sakhalin are the first record for
that island. They were described as a new subspecies because Demen-
tiev and Gizenko believe they are paler than hensoni and were local
birds, emphasizing that hensoni, though it shows altitudinal move-
ments, is not migratory. However, it is well known that hensoni is
migratory (see the ““Hand-list” and also Austin and Kuroda, 1953, Bull.
Mus. Comp. Zool., vol. 109, pp. 521-523, and Austin, 1948, Bull. Mus.
Comp. Zool., vol. 101, p. 199). Austin and Kuroda state, “A few birds
winter in the neighborhod of Sapporo, but most of the population
moves southward, most of them apparently across the Japan Sea to
Korea, where the Hokkaido race is a common winter visitor, but not
known to breed.” Hensoni winters also in Hondo and perhaps in the
Ryu Kyus where it is listed as a winter visitor by the “Handlist,”
though, as hensoni and nominate amaurottis (which is partly migratory)
are not very sharply differentiated, it is possible that visitors to the Ryu
Kyus are nominate amaurotis. I have also examined hensoni from Chu
Shan Island and Ningpo [= Ninghsien] at the mouth of the Yangtze.
As the northern populations of the species are migratory and the
species had never been reported before from Sakhalin (though Jap-
anese ornithologists had been active in the southern part of that
island), we cannot assume that the three birds reported by Dementiev
and Gizenko were local birds. It is possible that these birds, taken in
November, were individuals that had wandered north to southern
Sakhalin, rather than going south as is normal. In short, I believe that
it is best to consider septentrionalis a synonym of hensoni, at least until
it is shown that the species breeds on Sakhalin and breeding birds can
be compared.

2. Microscelis amaurotis amaurotis ‘Temminck, 1830, type locality,

12 AMERICAN MUSEUM NOVITATES NO. 1869

Japan, with matchie Momiyama, 1923, type locality, Hachijo, Seven
Islands of Izu, as a synonym. According to Austin (in Austin and
Kuroda, 1953, loc. cit.), matchie is “somewhat darker and browner’’
than nominate amaurotis [from Hondo] but the populations of
Hachijo, Tanegashima, and Yakushima are inseparable. I had only five
specimens from Hondo and two from Hachijo, but I cannot discern
any constant or appreciable difference between these specimens and
series from Tanegashima and Yakushima. In view of the clinal vari-
ation, it is possible that large series would show that the birds are
“somewhat darker’ on Hachijo, Tanegashima, and Yakushima than
those in Hondo, but the geographical variation of the species should be
considered as a whole, and the presence of slightly differentiated popu-
lations on these three islands does not necessarily warrant their nomen-
clatural separation.

3. Microscelis amaurotis ogawae Hartert, 1907, type locality, Amami
O Shima, central Ryu Kyus. This race, which is found on Amami and
Tokuno Shima, is clearly darker than nominate amaurotis, as well as
smaller (see below). It differs also by having a narrow but sharply de-
marcated band of chestnut across the base of the throat connecting the
chestnut areas on the ear coverts. In nominate amaurotis the base
of the throat is gray, or faintly banded by chestnut in some specimens,
but the band is very vague and more or less interrupted.

4. Microscelis amaurotis pryeri Stejneger, 1887, type locality, Oki-
nawa, central Ryu Kyus, with insignis Kuroda, 1923, type locality,
Miyako Shima, northern southern Ryu Kyus, as a synonym. This race
is another well-differentiated stage on the cline of increasing saturation
and decreasing size, and its chestnut band across the lower throat is
much broader than in ogawae, invading the upper breast. I did not ex-
amine specimens from Miyako, but it seems unnecessary to recognize
insignis which apparently differs from pryeri, according to Kuroda,
“only by being a trifle darker.” This author states that it is larger than
pryert and equal in size to stejnegeri, but this is not confirmed by my
measurements of these two races. According to Kuroda, three specimens
of insignis have a wing length of 116—129.5 and an exposed culmen of
21.5-23.5, but in pryert the wing length measures 110-129 and the
exposed culmen 22-24, while steynegeri measures, respectively, 116-131
and 24-27. It seems to me that the measurements given by Kuroda are
more similar to those of pryert. 3

5. Microscelis amaurotis stejnegerit Hartert, 1907, type locality, Ishi-
gaki, southern Ryu Kyus, with harterti Kuroda, 1922, type locality,
Botel Tobago, as a synonym. This race represents the extreme in the

1958 VAURIE: PALEARCTIC BIRDS, NO. 32 13

cline of increasing saturation but not of size, as stejnegeri has a slightly
longer wing than pryeri and its bill averages even larger than that of
hensoni at the northern end of the cline. Stejnegert is found on Ishi-
gaki, Iriomote, and Yonaguni in the southern Ryu Kyus, the small
islands off eastern Formosa, and on the tip of southern Formosa.
Harterti, based on three specimens, was described as larger than
stejnegert, but the measurements of these specimens as given by Kuroda
are identical or virtually so with those of topotypical stejnegert, and it
seems to me that harterti is not sufficiently distinct to warrant its recog-
nition.1 According to Kuroda, harterti measures: males, bill 32.5, 34,
wing 130.5, 132.5, tail 121, 126, tarsus 24, 24.5; and one female, respec-
tively, 31.5, 120, 112, 24. In adult stejnegeri the tail measures 110-124
and the tarsus 22.5-24.5, while (see below) the bill measures 30-34 and
the wing 116-131.

6. Microscelis amaurotis borodinone Kuroda, 1923, type locality,
Borodino Islands. This race was not examined, but apparently it does
not form part of the clines discussed above. Its nearest relative seems to
be ogawae on Amami O Shima, but it is said to be more olive above,
with darker ear coverts and with a darker band across the lower throat,
but the rest of the under parts is “generally paler.” Borodinone, judg-
ing by the measurements given by Kuroda, is larger than ogawae. He
gives the wing length of borodinone as 118-134.5. This race is re-
stricted to the small Borodino Islands, east of the Ryu Kyus.

7. Microscelis amaurotis squamiceps Kittlitz, 1831, type locality,
Bonin Islands. This race, restricted to the Bonins, seems also to have
some affinities to ogawae, but it is larger and has a broader but less
sharply defined band of chestnut across the lower throat. Its bill is
relatively short but more massive than in ogawae or any other race,
except magnirostre in which the bill is the largest and most massive
of any race of the species.

8. Microscelis amaurotis magnirostre Hartert, 1905, type locality,
Dionisio Island [— Minami Iwo Jima], Volcano Islands. This race, as
stated, differs by its very heavy bill and is similar in coloration to
squamiceps except that the band of chestnut across the base of the
throat is narrower and its general coloration is paler and grayer.

MEASUREMENTs: The sexes are identical in coloration and it seemed

1 Since the above was written, I have examined four male specimens from Kasho
To Island, off eastern Formosa, which suggest that harterti is a slightly differentiated
but valid race. These specimens are a little darker rufous on the breast than
stejnegerit and their wing length averages longer (128.5, as against 123.8 in male
stejnegeri), but I cannot see any difference in the length of the bill, tail, or tarsus.

14 AMERICAN MUSEUM NOVITATES NO. 1869

obvious that some of the specimens examined were not correctly sexed
in view of conspicuous discrepancies between specimens labeled as be-
ing of the same sex. Nevertheless, with a few exceptions I accepted
the specimens as they were sexed. The bill is measured from the skull.

M. a. hensoni: Males, 131, 137; 31.5, 33; females, 130, 137; 31, 32. Average
wing length, 134.

M. a. amaurotis: Hachijo, male, 136, 30. O Shima, female, 126 [27]. Hondo,
males, 130, 133, 139; 30, 31, 32; females, 122, 123; 29, 29. Yakushima, males,
129, 131, 134, 136, 139; 30, 31, 32, 32, 33; females, 119, 121, 124; 29, 29, 30,
Tanegashima, males, 128, 129, 131, 132, 137; 31, 32, 32, 32, 33; females, 120,
120, 126, 126, 127; 28, 28, 29, 30, 30. Average wing length, 129.

M. a. ogawae: Amami, males, 120 (type), 120, 121, 122, 124, 124, 125, 125;
29 (type), 29, 29, 30, 30, 30, 31; females, 116, 116, 118, 118, 119, 119, 120, 123,
124; 27, 28, 28, 28, 29, 29, 29, 29, 29. Average wing length, 121.

M. a. pryeri: Okinawa, males, 120, 122, 122, 123, 129; 29, 29, 29, 30, 31;
females, 110, 110, 110, 112, 114, 114, 115, 116, 119; 27, 27, 27, 27, 27, 28, 28,
28, 29, 29. Average wing length, 117.

M. a. stejnegeri: Ishigaki, males, 119, 120, 121, 124, 125, 126, 131 (type, bill
broken); 31, 32, 32, 33, 33, 34; females, 116, 117, 120; 30, 31, 32. Average wing
length, 122.

M. a. harterti: Kasho To, males, 124, 128, 131, 131; 30, 31, 32, 33; female,
120, 31. Average wing length, 126.8.

M. a. borodinone: Eleven adults, according to Kuroda (1923, Bull. Brit.
Ornith. Club, vol. 43, p. 122), wing, 118-134.5, “entire culmen,” 27-29.

M. a. squamiceps: Bonin Island, males, 124, 127, 140; 28, 30, 31; unsexed,
124, 128; 29, 30. Average wing length, 129.

M. a. magnirostre: Minami Iwo Jima, males, 128, 132, 134, 134, 135 (type);
32, 32, 33 (type), 33, 34; female, 128, 32. Average wing length, 131.8.

Pycnonotus leucotis and Pycnonotus leucogenys

These two bulbuls are closely related and are usually considered con-
specific. It was believed that they replace each other geographically,
but specimens collected by Meinertzhagen and by Koelz show that they
are sympatric in eastern Afghanistan. Meinertzhagen (1938, Ibis, p.
675) reported that he had collected humii (which I believe is con-
specific with leucotis) at Jalalabad on May 31, 1937, where he found
it common and breeding. A few days earlier, on May 25-26, 1937,
Koelz had collected a series of leucogenys near Jalalabad. He made no
mention on the labels whether these specimens were breeding birds
or not, but the presumption that they were is strong because the
breeding season of leucogenys is from April to July. Apparently,
leucogenys is resident, as it was collected on December 18 and 19, 1937,
by Koelz at Jalalabad.

Delacour (1943, Zoologica, vol. 28, pp. 17-28) considers that leucotis

1958 VAURIE: PALEARCTIC BIRDS, NO. 32 15

and leucogenys are separate species, and the specimens collected by
Meinertzhagen and Koelz support this opinion. Delacour stated that
leucogenys occurred in Afghanistan, perhaps on the authority of
Stuart Baker, but the first actual record for Afghanistan seems to be
the one published by Koelz in 1954 (see below). The only records for
Afghanistan were those of humii in the east published by Meinertz-
hagen and older records of nominate leucotis in the south.

The fact that leucotis and leucogenys are separate species is sup-
ported also by the fact that they are quite distinct morphologically
(see below). Their eggs are different. The difference could be racial,
but it seems significant that whereas the eggs of nominate leucotis and
hum are identical, those of leucogenys are distinctly smaller and
duller, paler, and “not so well and darkly marked” (see Stuart Baker,
1932, Nidification of birds of the Indian Empire, London, Taylor and
Francis, vol. 1, pp. 365-367).

Hum requires discussion, as some authors consider it an inter-
mediate connecting link between leucotis and leucogenys. It is so con-
sidered, because generally speaking its range connects that of the other
two and because it has a short crest, whereas leucotis has none and
leucogenys has a well-developed one. The bill also is about inter-
mediate in size and shape in humii. Nevertheless, as shown below,
humit is very much more similar to leucotis than to leucogenys and is
not truly intermediate. Moreover, as we have seen, it is sympatric with
leucogenys in eastern Afghanistan, and I may add that the specimens
of the latter collected by Koelz in eastern Afghanistan are typical and
show no hint whatever of being intermediate. It is quite probable that
humii and leucogenys are sympatric in other regions as well. The col-
lection of the American Museum of Natural History contains a speci-
men of humii collected on March 11, 1910, at Kalabagh on the Indus
River east of Bannu, by C. H. T. Whitehead, who wrote on its label
“occurs in company with M. leucogenys and intermedius [a race of
P. cafer] in gardens south of bungalow.”

We must consider the possibility that humii is a hybrid of leucotis
and leucogenys, as it is known that these two birds hybridize with cafer
and hence may interbreed also with each other. We lack such evidence,
however, and it is unlikely that humzi is a hybrid because, judging by
the literature, it occupies an extensive range where one of the putative
parents (leucotis) does not occur and, though I cannot be certain as I
have examined only one specimen, the authors consulted agree that
humit is a stable form.

Ticehurst (1926, Jour. Bombay Nat. Hist. Soc., vol. 31, p. 695)

16 AMERICAN MUSEUM NOVITATES NO. 1869

mentions a specimen of humii collected within the range of nominate
leucotis, stating that the Quetta Museum contains a specimen that was
“locally obtained.” However, Ticehurst gives no further information
concerning this specimen as it probably lacks adequate data. I be-
lieve that nominate leucotis and humii are not sympatric, because
the specimen mentioned by Ticehurst could have been a vagrant, as
these bulbuls, though not migratory, are known to wander after the
breeding season. It is quite possible also that it may have been an
escaped cage bird, as it is well known that in northwestern India these
bulbuls are caught and sold in large numbers as cage birds (see White-
head, 1909, Ibis, p. 113). .

The two species resemble each other but only superficially. They
differ as follows: Leucogenys is a little larger, more robust, has a
distinctly larger bill, with more strongly developed rictal bristles, and
is greenish on the back and rump in fresh plumage, not grayish brown
as in leucotis. In leucogenys the white area on the cheeks is more re-
stricted, and the coloration and feathering of its crown differ very
conspicuously from those of leucotis. In leucogenys they are brown
rather than pure black, and the feathers are greatly elongated, particu-
larly the posterior ones, forming a very conspicuous crest. These crest
feathers are narrow, lanceolate in shape, and are bordered by narrow
white edges, the white edges being broader and more conspicuous on
the feathers in front of and above the eye, thus forming a distinct white
superciliary streak. Leucotis has no crest, the feathers are not edged
with white, it has no white superciliary streak, and, as stated, its crown
is black rather than brown. Humii is identical with nominate leucotis,
with the exceptions that its bill is slightly larger and that it has a
crest, but this crest is much shorter than in leucogenys and its feathers
are black, not edged with white, and are broader and rounded. It
seems to me therefore that hum: is not intermediate, despite the fact
that the feathers of its crown are elongated. The similarity of its eggs
to those of nominate leucotis is mentioned above.

Pycnonotus leucotis

Pycnonotus leucotis (the White-eared Bulbul) can be divided into
three subspecies: humii Oates, 1889, type locality, Jalalpoor near
Jhelum, northwestern Punjab; nominate leucotis Gould, 1836, type
locality, Sind, with farahensis Koelz (1939, Proc. Biol. Soc. Washing-
ton, vol. 52, p. 64), type locality, southern Afghanistan, as a synonym;
and mesopotamiae Ticehurst, 1918, type locality, Basra, southern Iraq,

1958 VAURIE: PALEARCTIC BIRDS, NO. 32 17

with dactylus Ripley (1951, Postilla, no. 9, p. 8), type locality, Dam-
mam, Saudi Arabia, as a synonym.

Humii is discussed above. Its range is extreme northwestern Punjab
west of the Jhelum River, westward to eastern Afghanistan south of the
Hindu Kush, south to Kohat and the region of Bannu in North West
Frontier Province. Nominate leucotis ranges from western United
Provinces and the western border of Central Provinces westward,
south of P. leucogenys and humii in the Punjab, through Rajputana,
Sind, Baluchistan, and southern Afghanistan to about the border of
Iran where it grades into mesopotamiae. The latter ranges from
southern Iran westward to southern Iraq and down the coast of Arabia
to the region of Hofuf and differs chiefly and constantly from nominate
leucotis only in that it shows a very narrow yellow eye ring that is
lacking in that form. It is said also that mesopotamiae is rather darker
and grayer below, slightly larger, and has a bigger bill. However, the
differences in coloration of the plumage and in the size of the bill are
slight, and the measurements of the bill and wing overlap.

It is difficult to establish with certainty where mesopotamiae and
nominate leucotis replace each other, as the differences between them
are relatively very slight but, more or less arbitrarily, one can select
Persian Baluchistan as the region where they grade into each other.
Ticehurst (loc. cit.) stated that a specimen from Mand on the Persian-
Baluchi border is mesopotamiae and was puzzled because Hartert had
informed him that specimens from father west (from Persian Baluchis-
tan in the Rothschild Collection) are indistinguishable from typical
nominate leucotis. It is true that these specimens do not show a yellow
eye ring, but even in perfectly prepared skins the ring is very narrow
and not conspicuous and the specimens from Persian Baluchistan were
not well prepared. One cannot be certain, but, in view of the fact
that specimens from Persian Baluchistan average somewhat smaller in
wing length and in size of the bill, it seems reasonable to conclude that
the two races grade into each other in that region.

The validity of farahensis has been denied by Whistler (1944, note
Bombay Nat. Hist. Soc., vol. 44, p. 517) who compared specimens from
Kandahar and India. I have examined the type and paratypes of
farahensis and agree with Whistler that it is not separable from
neminate leucotis. These specimens consist of two from Farah and
seven from Kandahar and one of the two from Farah has the bill very
slightly larger than in the rest of the series. This specimen suggests
that at Farah (about 350 kilometers west of Kandahar and not far from

18 AMERICAN MUSEUM NOVITATES NO. 1869

the Persian border) we approach the zone of intergradation mentioned
above.

Meinertzhagen (1954, Birds of Arabia, London, Oliver and Boyd, p.
180) holds that dactylus Ripley is not sufficiently well differentiated
from mesopotamiae to warrant its recognition. He states: “I have com-
pared Bahrein and Dahran birds with a large series from Iraq. In most
cases they agree perfectly; in a few cases Dahran and Hufoof birds are
almost imperceptibly paler below, a character I do not consider
sufficient for validity.” Two specimens of each were examined by me
from Arabia and Iraq but, as the latter are soiled and sooty below,
they throw no light on the question whether or not the birds of Arabia
are paler and less smoky below than those of Iraq, as Ripley believes.
Judging, however, by the observations of Meinertzhagen, the difference
between the two populations is not constant and seems much too
slight to warrant the recognition of dactylus.

When Meinertzhagen discussed humit, which he believes is con-
specific with leucogenys, he remarked (1938, loc. cit.) that it was “Not
noted at Kabul or elsewhere, which is curious as it has been observed
frequently in Russian Turkestan.” Lest this statement lead to error,
I may remark that “leucogenys” may not occur north of the Hindu
Kush and that records for Turkestan are open to doubt. Apparently,
only two records exist for Turkestan and are those of birds [seen ?]
by Zarudny on June 16 and September 15, 1910, in Tadzhikistan. The
Russian ornithologists, however, strongly question these “records,”
stating (1954, Birds of the Soviet Union, vol. 6, p. 72) that the bird
has never been observed in Turkestan, and they doubt that it was
collected by Zarudny as they have been unable to find or trace
specimens.

Pycnonotus leucogenys

The White-cheeked Bulbul increases somewhat in size as its popula-
tions range farther west, but the geographical variation is clinal and
slight, and there are no differences in coloration. Koelz (1954, Contrib.
Inst. Reg. Explor., no. 1, p. 11) has, however, described as picru (type
locality, Afghanistan) the birds of the Punjab, Kashmir, and Afghanis-
tan on the basis that they are larger than those of Darjeeling, which
he believes is the type locality of leucogenys Gray, 1835. His measure-
ments and those I have taken are listed below. They show much over-
lap, and the birds of the Punjab are perfectly intermediate. The
populations differ slightly in average size, but many specimens are in-
distinguishable, and it seems best to me not to recognize any subspecies.

The correct type locality of leucogenys Gray (1835, in Hardwicke,

1958 VAURIE: PALEARCTIC BIRDS, NO. 32 19

Illustrations of Indian ornithology) seems to be Kashmir and not
Darjeeling. The plates in Hardwicke’s book were supplied with sci-
entific names by Gray, but no text accompanies the plates, and the
correct type locality becomes the first definite locality supplied by a
subsequent author. Apparently Blyth (1845, Jour. Asiatic Soc. Bengal,
vol. 14, p. 567) is the first author to mention a definite locality, stating
that leucogenys Gray is “common in the Himalaya, and in Kashmir.”
“Himalaya” is a general term, but Kashmir is a definite one, and I
accept it, therefore, as the correct restricted type locality. I believe that
Koelz was probably misled by Stuart Baker (1921, Jour. Bombay Nat.
Hist. Soc., vol. 27, p. 469; and 1922, Fauna British India, London,
Taylor and Francis, vol. 1, p. 389) who states that the type locality
is Darjeeling, but it is noteworthy that Baker in his corrected syn-
onymies (1930, op. cit., vol. 7, p. 80) made no further mention of
Darjeeling. To be sure, Baker made no mention of Kashmir, or any
other locality, in 1930, but it is probable that he had become aware
that Darjeeling was not the correct type locality.

Individual measurements of the wing length of males are: Koelz:
Afghanistan and Kashmir, 94, 94, 95, 96, 96, 97, 97, 97, 97, 97, 99, 99,
99 (96.7); Himalayan Punjab, 88, 88, 90, 90, 91, 92, 93, 93, 94, 94, 94,
95, 95, 95, 96, 96, 97, 98 (93.3); Nepal and Darjeeling, 86, 88, 89, 89, 89,
89, 89, 89, 90, 90, 91, 92 (89.3). Vaurie: Afghanistan, 94, 94, 95, 97, 97,
97, 98 (96); Kashmir (unsexed adults), 94, 98, 99 (97); Himalayan Pun-
jab, 91, 93, 94, 96, 99 (94.6); Sikkim (unsexed adults), 90, 91, 92, 97
(92.5).

NECTARINIIDAE
Nectarinia pulchella

The Beautiful Sunbird penetrates into the southern fringes of the
Palearctic region, breeding north to the northern Air Massif and the
Ennedi Massif in the southern Sahara. Nominate pulchella Linnaeus,
1766, type locality, Senegal, breeds across the grasslands north of the
Guinea forest from Senegal and Portuguese Guinea to Kordofan in the
Sudan, and, according to Niethammer (1955, Bonner Zool. Beitrage,
vol. 6, pp. 73-74), north to the Ennedi. In the Sudan, east of Kordofan,
it is replaced by lucidipectus Hartert (1921, Novitates Zool., vol. 28,
p- +123), type locality, Wad Medani, eastern Sudan, which ranges east-
ward to Eritrea, and south to southwestern Uganda (west to Lake
Albert in the Belgian Congo) to Lake Edward and about central Kenya.
In the Air Massif, south to the region of Kano in northern Nigeria, is
found aegra Hartert (1921, op. cit., p. 122), type locality, Timia, in
the Air north of Mt. Baguezane.

20 | AMERICAN MUSEUM NOVITATES NO. 1869

The validity of the two races described by Hartert has been ques-
tioned recently. Villiers (1950, Mém. Inst. Francais d’Afrique Noire,
no. 10, p. 380) states that he considers aegra to be “sans valeur,’ and
Cave and Macdonald (1955, Birds of the Sudan, London, Oliver and
Boyd, p. 349) do not recognize lucidipectus, stating that it “is based
on characters not clearly evident to us.’’ However, Hartert was a con-
servative taxonomist, and I find, after studying his and other com-
parative material, that aegra and lucidipectus are valid. They are
not very sharply differentiated but are easily distinguishable. In aegra
the bill is distinctly smaller than in nominate pulchella, and the red
breast patch of the males is slightly duller. The bill length (measured
from the nostril for greater accuracy) measures in 20 males from
Senegal: 14, 14, 14, 14.2, 14.5, 14.5, 14.5, 15, 15, 15, 15, 15, 15, 15, 15.5,
15.5, 15.5, 16, 16, 16.5 (15), and in nine males from the Air: 12.5, 13,
13, 13.5, 13.5, 13.5, 13.5, 14, 14, (13.4). The difference is certainly not
great, and there is a slight overlap in individual measurements, but in
aegra the bill is altogether distinctly smaller, narrower at the base, and
more slender throughout. This difference, taken together with the
difference in coloration, seems to warrant the recognition of the race.
In lucidipectus, on the other hand, the bill is as large as in nominate
pulchella, but as Chapin states (1954, Bull. Amer. Mus. Nat. Hist.,
vol. 75B, p. 269) “the red in the middle of the breast and the yellow
at the sides [are] noticeably brighter.” Chapin recognizes aegra also,
and Niethammer (loc. cit.) lucidipectus.

I did not study the other forms that are usually considered to be
races of pulchella. They are melanogastra, nectarinioides, and erlangeri
in east Africa. The ranges of the first two require further study, and
some authors hold they overlap. If so, two species are probably in.
volved.

y

ZOSTEROPIDAE
Zosterops japonica

The Japanese White-eye is distributed throughout the Japanese
Archipelago, including the Seven Islands of Izu, the islands between
Japan and Korea, the Bonins, Volcanos, Ryu Kyus, and the Borodinos.
It breeds also on Formosa, the islands off eastern Formosa, Luzon and
some small islands north and south of Luzon, China from Hopeh
south to Yunnan and northern Indochina, and on Hainan. It varies
geographically and in my opinion can be divided into 10 subspecies
(see below), but the Japanese ornithologists recognize additional ones
which I believe are not valid or much too slightly differentiated for

1958 VAURIE: PALEARCTIC BIRDS, NO. 32 9

their recognition to be warranted. The “Hand-list of the Japanese
birds” (1942, pp. 31-33) recognizes 11 races, not counting three extra-
limital races (simplex, hainana, and meyeni), and to these must be
added another form described by a Japanese author in 1951, making
a total of 15 forms that must be considered.

My study has been based on about 325 specimens from virtually all
parts of the range, and I have also had the benefit of receiving some
notes from Dr. G. F. Mees of Leiden who has made a. specialty of the
study of the zosterops of Asia. He has examined much material from
European collections not available to me.

The over-all pattern of the geographical variation is simple and may
be briefly described. On the main islands of Japan (nominate japonica),
and on Tanegashima and Yakushima (insularis), the populations are
relatively dark green above, washed with isabelline brown below, and
are moderate in size. The two races are exceedingly similar, differing
only very slightly in coloration and the fact that insularis has a some-
what longer bill. Austin (1953, in Austin and Nagahisa Kuroda, Bull.
Mus. Comp. Zodl., vol. 109, pp. 577-578) speaks of a color cline, the
populations becoming “progressively darker through the several races
recognized down the Ryukyu chain to Formosa.”’ However, I cannot
see any evidence whatever of this cline, at least south of TTanegashima
and Yakushima; in fact, in the Ryu Kyus (loochooensis) and Formosa
(semplex), the populations are very definitely paler above and below,
certainly not darker. On the Seven Islands of Izu (stejynegert), Bonins
and Volcanos (alant), Borodinos (daitoensis), and on Botel Tobago
and Kasho To off eastern Formosa and on Batan Island north of Luzon
(batanis) are found four distinctive races, the first two with long and
wide bills. Simplex is a small race with a short bill and is found in
China and on Formosa, and is, as stated, very definitely paler than
nominate japonica or insularis, and paler than loochooensis. The race
of Hainan (hainana) is barely separable from simplex and about as
poorly differentiated from the latter as insularis is from nominate
japonica, while the race of Luzon (meyenz) is the smallest and palest
of all. |

I have measured a large number of the specimens I examined (table
1 and list of individual measurements), as measurements have been
used most extensively in the diagnoses of the various forms—more so
than coloration. A large series of measurements shows, however, that
the range of variation is rather narrow, generally speaking, though of
course the differerice between the extremes (stejnegeri the largest, and
meyent the smallest) is very well marked.

22 AMERICAN MUSEUM NOVITATES NO. 1869

1. Zosterops japonica japonica ‘Temminck and Schlegel, 1847, type
locality, Japan, with the following synonyms: 1jimae Nagamichi Ku-
roda, 1917, type locality, Tsushima; and yesoensis Nagahisa Kuroda
(1951, Bull. Biogeogr. Soc. Japan, vol. 15, no. 2, p. 5), type locality,
Hokkaido. Ijimae was separated from insularis on the sole basis of
being larger, but I can see no differences whatever between the meas-
urements of it as given by Kuroda and those of a long series of
insularts. Kuroda states that the wing length of five adults is 59-62
and the length of the exposed culmen 12.5-13. In 10 adults of insularis
that I have picked at random the exposed culmen measures 11.5, 11.5,
12, 12, 12.5, 12.5, 12.5, 13, 13, 13, and it measures 12.5 in the type of
insularis, a male. I certainly agree with Mees, therefore, that “not the
slightest reason exists to maintain 2jimae.” Mees considers it a synonym
of insularis, but in the topotype of ijimae that I have examined the
coloration is identical with that of nominate japonica, and the differ-
ence in the length of the bill is not significant, as it measures only half
of a millimeter longer than that of the specimen of nominate japonica
with the longest bill measured. It seems best to me to synonymize tjimae
with nominate japonica, but the differences between the latter and
insularis are relatively so slight that ijimae could be synonymized as
well with insularis.

Nagahisa Kuroda had only one specimen when he described yesoen-
sts, which he says is paler than nominate japonica, but Austin (loc. cit.)
believes it is valid. Austin states, however, that the material he has
examined from Hokkaido is very old and may be faded, as it was
collected between 1854 and 1885. He implies he is not certain that
yesoensis is valid, and Mees mentioned to me that in current Japanese
publications the validity of yesoensis has been denied. I have not seen
specimens from Hokkaido, but I follow Mees who believes it is best
not to recognize it.

The range of nominate japonica embraces the following islands:
Hokkaido, Hondo, Shikoku, Kyushu, Dagelet, Tsushima, Iki, Quelpart,
and probably Sado and Oki where it has been reported and may breed.
Austin (1948, Bull. Mus. Comp. Zodl., vol. 101, p. 248) believes that the
species does not breed in Korea but is only a transient. Nominate
japonica is not truly migratory, or is only slightly so. It shows some
seasonal movements and wanders and has reached, in addition to
Korea, the islands of Tanegashima, Yakushima, Amami O Shima,
Okinawa and some surrounding islands, and the Seven Islands of Izu.
As the ‘Hand-list’’ mentions, it is difficult, however, to be certain
about the Seven Islands. Apparently, the smaller race of Japan is a
better singer than the native steynegert and is imported to the Seven

1958 VAURIE: PALEARCTIC BIRDS, NO. 32 23

Islands as a cage bird. It is possible that the specimens of nominate
japonica reported from these islands were birds that had escaped.

2. Zosterops japonica insularis Ogawa, 1905, type locality, Tanegash-
ima and Yakushima. The type, in the collection of the American
Museum of Natural History, is from Tanegashima. As stated above,
this race, which is restricted to Tanegashima and Yakushima, is only
slightly differentiated from nominate japonica. The best difference
seems to be in the length of the bill which averages about 2 mm. longer
in insularis. In addition, the birds of Tanegashima and Yakushima are
a little darker and duller green above, and a little darker, richer yel-
low on the throat, but all the color differences are very slight and not
constant.

3. Zosterops japonica loochooensis Tristram, 1889, type locality “Loo
Choo Islands,” with the following synonyms: yonakuni Nagamichi
Kuroda, 1923, type locality, Yonaguni Island, southern Ryu Kyus; and
irtomotensis Nagamichi Kuroda, 1923, type locality, Iriomote Island,
southern Ryu Kyus. The material of irtomotensis and yonakuni ex-
amined by Mees or myself is insufficient, but it suggests that it is best
not to recognize these forms.

Kuroda states that yonakuni resembles batanis, but this is not made
evident in the description which, in fact, suggests that yonakuni does
not differ appreciably from loochooensis. Kuroda states that yonakuni
differs from loochooensis by being darker on the flanks and by having
“on an average” a longer and stouter bill. Mees tells me, however,
that specimens from Yonaguni Island differ from loochooensis only by
“being very slightly greener and duller” on the back, and I find that
three specimens from Ishigaki Island (the population of which is
called yonakuni by the “Hand-list’’) are identical with loochooensis.

Kuroda separated itriomotensis on the ground that it has a shorter
tail than loochooensis, “36.5—40.5 as against 39-43.5,” but I have ex-
amined a number of specimens of loochooensis with a tail of 36, and
a paratype of irtomotensis is identical with the latter in size and
coloration with the exception that it shows some rusty spots in its
plumage. These spots are mentioned by Kuroda who states that in
triomotensis “the upper parts are generally patched or sometimes uni-
formly washed with rusty.” He is inclined to believe that this is a
subspecific character, but I doubt this, in as much as individuals that
are irregularly patched with rust are found in nominate japonica,
insularis, and loochooensis. The cause of these rusty spots or washes
is not clear, but they may be associated with feeding habits, as they
are found chiefly on the head and, more rarely, on the back. This is
not certain, however, as Kuroda mentions that in one paratype of

24 AMERICAN MUSEUM NOVITATES NO. 1869

irtomotensis the entire upper parts are uniformly rusty “without any
olive-green parts.” In the paratype examined irregular traces of rust
are present on the forehead, nape, and throat, and the sides of the
head are washed with rust but a little more so on one side. Ogawa
(1905, Annot. Zool. Japonenses, vol. 5, p. 187) mentioned that 28 of
his 55 specimens of insularis (Tanegashima and Yakushima) showed
rusty areas variable in size and distributed irregularly throughout the
plumage. Rusty spots or traces of them are shown by 12 of my 35
specimens from these two islands. They are chiefly found on the fore-
head and crown, less frequently on the ear coverts and throat, and are
present on the back in two specimens.

Mees did not examine triomotensis but wrote to me that he did not
believe it was valid. As shown above, it does not differ from loochooen-
sis in size, and a difference in coloration is open to question.

Loochooensis, including yonakunit and iriomotensis, has been ex-
amined by me or reported from the following islands: Amami, Kikai,
Tokuno, Okono Erabu, Yoron, the Iheya group, Okinawa, Kume, the
Kerama group, Miyako, Ishigaki, Iriomote, and Yonaguni.

4, Zosterops japonica daitoensis Nagamichi Kuroda, 1923, type
locality, Borodino Islands. I believe that this race, which is restricted
to the Borodinos, is related to loochooensis and not to batanis as
Kuroda believes, or at any rate is closer to loochooensis. It is similar
to the latter, but it is very slightly paler and more yellow-green above
and is more yellow on the lores and on the forehead at the base of the
bill. It appears also to be blacker in front of and below the eye, but
this character may not be constant, as I can match perfectly well in
this respect a paratype of daitoensis with specimens of loochooensis.

5. Zosterops japonica simplex Swinhoe, 1861, type locality, southern
China, with the following synonyms: peguensis Baker (1922, Ibis, p.
144), type locality, Moulmein, Burma; and taivaniana Momiyama,
1927, type locality, Formosa. The species is only a winter visitor to
Burma, and peguensis was based on winter visitors of stmplex. I agree
with Mees that the population of Formosa cannot be separated from
simplex, and Hartert (1923, Die Végel der paldarktischen Fauna,
Nachtrag, p. 33) had already reached the same conclusion. Stresemann
(1931, Mitteil. Zool. Mus. Berlin, vol. 17, p. 208) believes, however,
that the populations of Formosa and Hainan are so extremely similar
that they should be combined under the same name. The difference
between simplex and hainana is very slight, but in my opinion the
birds of Formosa are simplex.

Simplex differs from the preceding valid races by being paler, more
yellowish, above and slightly paler below than loochooensts or dattoen-

1958 VAURIE: PALEARCTIC BIRDS, NO. 32 25

sts and much paler than insularis or nominate japonica; it usually
shows a little yellow on the lores and on the forehead and is smaller
than all the preceding races. It ranges from Hopeh, at least from the
south, and Shantung southward to Shensi, Szechwan, eastern Sikang,
Yunnan, Kwangsi, Kwangtung, and northern Indochina, and, as stated,
Formosa. It is partly migratory, winter visitors reaching northern
Siam, eastern and southern Burma south to Tenasserim, Indochina,
and Hainan.

6. Zosterops japonica hainana Hartert, 1923, type locality, Hainan.
This race, which is restricted to Hainan, is very poorly differentiated,
but it averages a little smaller than simplex and is usually somewhat
darker and richer yellow on the throat.

7. Zosterops japonica meyeni Bonaparte, 1850, type locality, Philip-
pines. This race, which is found on Luzon, Calayan Island north of
Luzon, and on Lubang, Verde, and Banton Islands south of Luzon,
is of the same size as hainana but has a somewhat shorter bill. With
the exception of batanis, it is the palest of all the races and the most
yellowish above. ‘The yellow of its throat is pure and bright and shows
a tendency to invade the cheeks to a variable extent, being less sharply
defined than in the other races.

8. Zosterops japonica batanis McGregor, 1907, type locality, Batan
Island, north of Luzon. This race is similar to meyenz in coloration,
pale and bright yellow-green above, but the yellow area on the lores
and on the forehead is larger and purer in shade and is more extensive
on the throat also, the yellow pigment invading the upper breast.
Batanis is distinctly larger than meyeni and has a much larger bill,
thicker, wider at the base and about 4 mm. longer. In addition to
Batan Island, it is found also on the small islands of Botel Tobago
(Koto Sho) and Kasho To, off southeastern Formosa.

9. Zosterops japonica stejnegert Seebohm, 1891, type locality,
Hachijo, Seven Islands of Izu. This race, which inhabits the Seven
Islands of Izu, is the largest race of all and has a conspicuously large
and long bill. It is similar to nominate japonica in coloration, being
washed with isabelline brown below but is very slightly paler and
brighter green above.

10. Zosterops japonica alani Hartert, 1905, type locality, San Dioni-

sio Island [— Minami Iwo Jima], Volcanos. This race is similar to
nominate japonica above but distinctly paler below, and therefore
paler also than stejnegert, being washed with grayish or very pale
isabelline brown. It is intermediate in size, including the bill, between
the smaller nominate japonica and the large stejynegert. This race in-
habits the Volcano and Bonin Islands.

- TABLE 1

MEASUREMENTS OF ADULTS IN Zosterops japonica

Race and Population

Nominate japonica
Hondo and
migrants
Tsushima
insularis
Tanegashima

Yakushima

loochooensis
Kikai Shima

Amami O Shima

Tokuno Shima
Iheya Shima
Kerama Retto
Okinawa

Ishigaki

Iriomote
daitoensis

Daito Shima
simplex

Eastern China

Formosa

hainana
Hainan

meyent

Luzon
batanis

Batan Island

stejnegert
Seven Islands

alant

Minami Iwo Shima

Chichi Shima

* Unsexed.

N

me G1 sT

So © ~=7 00
OO QAAQ{WAQVWOWWOAQODA WOAWD A410 Oy

—_

Rm WA UD DF OO WO

—

Wing

59-62 (60.3)

59-62 (60.4)

61

Tail

Bill

40-42 (41)
38-43 (40.2)
42

58-62.5 (60.4) 38-43 (40.1)

57-61 (59.7)
58-63 (60.3)
59-62 (60.0)

58-60 (58.6)
58, 58, 58
56-61 (58.4)
56-61 (57.2)
57

57, 58

58, 59

58-61 (59.4)
55-61 (58.2)
57, 60, 60

57

60

55-58 (56.4)
55-58 (56.0)
54-57 (55.1)
53-56 (53.9)

52-54 (52.5)
51, 54

52-56 (53.3)
53, 54
51-54 (52.6)

57
60

63-66 (64.7)
61, 61

61-64 (62.5)

62-64 (63.1)
64.5, 65

26

38-41 (39.6)
38-42 (40.6)
38-44 (39.7)

36-39 (37.9)
38, 39, 40

39-42 (40.2)
36-42 (39.2)

41, 42

37-40 (39)
37-43 (39.8)
38, 39, 39
40

43
32-37 (34.6)
34-37 (35.3)
34-38 (35.9)
33-39 (36.2)

32-37 (35.0)

35, 36

32-35 (33.6)
32, 34
32-36 (34.1)

39

40

42-46 (44)
40, 44

45-46 (45.5)
44-47 (45.6)
48, 48

14-16 (15)
14-15.5 (14.9)
16.5

16-17.5 (16.9)
16-17.5 (16.9)
16.5-18 (17.2)
16-18 (17.1)

15-16 (15.4)
14.5, 15, 15.5
14.5-16.5 (15.5)
14-15.5 (14.9)
15.2

15, 15

14, 15.5
14-16 (15.1)
14-16 (15.1)
15, 15, 15.5
14.5

14.5

12.5-13.5 (13.1)
12-14 (12.8)

12. 5-14 (13.2)
12.5-14 (13.3)

12.5-14 (13.5)
14, 14

11-13 (12.0)
13, 13
12-14 (13.0)

16
16

19.5-20.5 (20.0)
20, 20

16.8-17 (16.9)
16.5-17.5 (16.9)
16, 16

1958 VAURIE: PALEARCTIC BIRDS, NO. 32 27

INDIVIDUAL MEASUREMENTS Not GIVEN IN TABLE 1

Nominate japonica

Wing, males, 59, 59, 60, 61, 61, 61, 62; females, 59, 59, 60, 62, 62. Tail, males,
40, 40, 41, 41, 41, 42, 42; females, 38, 38, 40, 42, 43. Bill, males, 14, 14.5, 14.5,
15, 15, 15.5, 16; females, 14, 14, 15, 15, 15.5.

Zosterops j. insularis

Tanegashima: Wing, males, 58, 59, 60, 60, 60, 61, 62, 62.5; females, 57, 59,
60, 60, 60, 61, 61. Tail, males, 38, 38, 38, 40, 41, 41, 42, 43; females, 38, 38, 39,
40, 40, 41, 41. Bill, males, 16, 16, 16.5, 16.5, 17, 17, 17.5, 17.5; females, 16, 16,
16.5, 16.5, 17, 17.5, 17.5. The type of insularis, a male, included in this series
measures 62.5, 43, 17.

Yakushima: Wing, males, 58, 58, 59, 60, 60, 60, 61, 62, 62, 63; females, 59,
59, 59, 60, 60, 60, 60, 61, 61, 62. Tail, males, 38, 39, 40, 41, 41, 41, 41, 41, 42,
42; females, 38, 38, 39, 39, 39, 39, 40, 41, 41, 44. Bill, males, 16.5, 16.8, 17, 17,
17, 17.2, 17.2, 17.5 18, 18; females, 16, 16.2 17, 17, 17, 17, 17, 17.8, 18, 18.

Zosterops j. loochooensis

Kikai Shima: Males, wing, 58, 58, 58, 58, 59, 59, 59, 60; tail, 36, 37, 37, 38,
38, 39, 39, 39; bill, 15, 15, 15, 15, 15.5, 16, 16, 16.

Amami O Shima: Wing, males, 56, 58, 58, 58, 58, 58, 59, 60, 61; females, 56,
56, 56, 57, 57, 57, 57, 58, 61. Tail, males, 39, 39, 40, 40, 40, 40, 41, 41, 42; females,
36, 37, 39, 39, 39, 40, 40, 41, 42. Bill, males, 14.5, 15, 15, 15.2, 15.5, 15.5, 16,
16.5, 16.5; females, 14, 14.2, 14.5, 14.5, 15, 15, 15.5, 15.5, 15.5.

Okinawa: Wing, males, 58, 58, 59, 61, 61; females, 55, 57, 57, 59, 60, 61.
Tail, males, 37, 39, 39, 40, 40; females, 37, 39, 39, 40, 41, 43. Bill, males, 14, 15,
15, 15.5, 16; females, 14, 15, 15, 15.5, 15.5, 16.

Zosterops j. simplex

China: Wing, males, 55, 55, 56, 56, 56, 57, 57, 57, 57, 58; females, 55, 55,
55, 55, 56, 56, 56, 57, 57, 58. Tail, males, 32, 33, 33, 33, 34, 35, 36, 36, 37, 37;
females, 34, 34, 34, 35, 35, 35, 36, 36, 37, 37. Bill, males, 12.5, 12.8, 13, 13, 13,
13.2, 13.2, 13.5, 13.5, 13.5; females, 12, 12.2, 12.5, 12.5, 12.5, 12.5, 13, 13.2,
13.5, 14.

Formosa; Wing, males, 54, 54, 54, 55, 55, 55, 55, 56, 56, 57; females, 53, 53,
53, 53, 54, 54, 54, 54, 55, 56. Tail, males, 34, 34, 35, 35, 36, 36, 37, 37, 37, 38;
females, 33, 34, 35, 36, 36, 37, 37, 37, 38, 39. Bill, males, 12.5, 12.5, 13, 13 13,
13.5, 13.5, 13.5, 14, 14; females, 12.5, 12.5, 13, 13, 13, 13.5, 13.5, 14, 14, 14.

Zosterops j. hainana
Hainan: Males, wing, 52, 52, 52, 52, 52, 52, 52.5, 53, 53, 53, 54; tail, 32, 34,
35, 35, 35, 35, 35, 36, 36, 36, 37; bill, 12.5, 13, 13, 13, 13, 13.5, 13.5, 14, 14, 14,
14. The type of hainana, included in this series, measures 53, 35, 13.

Zosterops j. meyent
Wing, males, 52, 52, 53, 53, 54, 56; unsexed, 51, 52, 53, 53, 54. Tail, males,
32, 33, 34, 34, 34, 35; unsexed, 32, 33, 35, 35, 36. Bill, males, 11, 11.5, 12, 12,
12.5, 13; unsexed, 12, 13, 13, 13, 14.

28 AMERICAN MUSEUM NOVITATES NO. 1869

Zosterops j. stejnegert
Males, wing, 63, 65, 65, 66; tail, 42, 43, 45, 46; bill, 19.5, 20, 20, 20.5.
Zosterops j. alant
Wing, males, 61, 62, 63, 64; females, 62, 63, 63, 63, 64, 64. Tail, males, 45,
45, 46, 46; females, 44, 45, 45, 46, 47, 47. Bill, males, 16.8, 17,17, 17; females,

16.5, 16.5, 16.8, 17, 17, 17.5. The type of alant, a male, included in this series,
measures 62, 46, 17.

Zosterops palpebrosa

The Oriental White-eye ranges from India and Yunnan eastward to
the southern and central Philippines and Flores and penetrates a little
way into the Palearctic region, reaching eastern Afghanistan in the
region east of Kabul and south of the Hindu Kush. It has been divided
into about 20 subspecies, four of which inhabit India: nominate
palpebrosa, salimalit, nilgirtensis, and egregia.

Mees is of the opinion that egregia Madarasz, 1911, type locality,
Ceylon, is the race most widely distributed in India, inhabiting the
Laccadives, Ceylon, and all India with the exception of the more arid
parts of the northwest and of the regions occupied by the three other
races. Of these, nilgirtensis inhabits the Nilgiris and other hills of the
southwest to Travancore, salimalii the Eastern Ghats north to the
Godavari, and nominate palpebrosa Nepal, Bengal, Orissa, and eastern
India eastward to Burma and Yunnan. Mees believes that occidentis
Ticehurst (1927, Bull. Brit. Ornith. Club, vol. 47, p. 88), type locality,
Simla, is a synonym of egregia and not of nominate palpebrosa,
stating that Ripley (1950, Jour. Bombay Nat. Hist. Soc., vol. 49, p. 411)
was incorrect when he synonymized occidentis (called occidentalis in-
advertently by Ripley) with nominate palpebrosa. He considers that
remota Koelz (1939, Proc. Biol. Soc. Washington, vol. 52, p. 76), type
locality, Jalalabad, Afghanistan; and amabilis Koelz (1950, Amer. Mus.
Novitates, no. 1452, p. 9), type locality, Kathiawar, are synonyms of
egregia. I follow Mees as his opinions seem well founded.