II B R.AR.Y

OF THE

UNIVERSITY

Of ILLINOIS

590.5

v. 35
coo- 3

matural history
su&veV

5
S

.■3

NOTES ON AFRICAN BULBULS

FAMILY PYCNONOTIDAE: CLASS AVES

AUSTIN L. RAND

FIELDIANA: ZOOLOGY

VOLUME 35, NUMBER 6

Published by

iPT 2 \ i CB?£AGO NATURAL HISTORY MUSEUM

SEPTEMBER 26, 1958

NOTES ON AFRICAN BULBULS

FAMILY PYCNONOTIDAE: CLASS AVES

AUSTIN L. RAND

Chief Curator, Department of Zoology

FIELDIANA: ZOOLOGY

VOLUME 35, NUMBER 6

Published by

CHICAGO NATURAL HISTORY MUSEUM

SEPTEMBER 26, 1958

Library of Congress Catalog Card Number: 58-18825

PRINTED IN THE UNITED STATES OF AMERICA
BY CHICAGO NATURAL HISTORY MUSEUM PRESS

I/. 3^

2

African Bulbuls

The following notes on the classification and taxonomy of bulbuls
were brought together in the course of preparing the section on Afri-
can bulbuls for the continuation of Peters' Check-list.

It is a pleasure to acknowledge the help of various persons in lend-
ing specimens and in other ways, especially the help of Dr. Dean
Amadon and Dr. C. Vaurie of the American Museum of Natural
History, Mr. P. A. Clancey of the Durban Museum, Mr. H. Deig-
nan and Dr. H. Friedmann of the United States National Museum,
Mr. J. C. Green way of the Museum of Comparative Zoology, Dr.
R. E. Moreau of the Edward Grey Institute, Oxford, and Dr. J. G.
Williams of the Coryndon Museum.

Among the collections available in Chicago Natural History
Museum and especially important in this survey should be men-
tioned the van Someren collection of East African birds, acquired
in 1950, and the A. I. Good collection of Cameroon birds, the
property of the Cleveland Museum of Natural History but on
deposit here.

Generic limits. — The generic allocation of the greenish and yel-
lowish olive bulbuls has always been troublesome, but fortunately
we have Delacour's (1943, Zoologica, 28: 17-28) review of the fam-
ily at the genus and species level. My treatment for the African
forms differs from Delacour's in the following:

Neolestes, which Delacour does not include as a bulbul, seems best
included (see Chapin, 1953, Bull. Amer. Mus. Nat. Hist, 75A: 154).

Tylas, which Delacour also excludes from the bulbuls, I have
retained for lack of a better place for it.

Besides the other genera Delacour discussed, two others are occa-
sionally placed in the bulbuls:

Suaheliornis has been placed in the Pycnonotidae by some auth-
ors (Mackworth-Praed and Grant, 1955, Bds. E. and N. E. Afr., 2:
134). In some ways it recalls Phyllastrephus, but it seems still closer
to the Sylviidae genus Macrosphenus, which itself seems a hetero-
geneous group.

145

146 FIELDIANA: ZOOLOGY, VOLUME 35

Lioptilornis (genotype nigricapillus Vieillot) has been called a
bulbul by Vincent (1952, Checklist Bds. So. Afr., p. 65), but it
seems better considered a timaline, as Delacour (1946, L'Oiseau, 16:
30) and Chapin have considered it.

The genera and species of African bulbuls that I recognize are
as follows: Pycnonotus, 15 species (others in Asia); Calyptocichla, 1;
Baeopogon, 2; Ixonotus, 1; Chlorocichla, 5; Thescelocichla, 1; Phylla-
strephus, 22; Bleda, 3; Nicator, 3; Criniger, 4; Microscelis, 1 (others in
Asia); Neolestes, 1; and Tylas, 1; i.e., in Africa, I recognize 13 genera
(5 of which are monotypic) and 60 species. Sclater (1930, Syst. Av.
Aethiop.) recognized 25 genera (12 of them monotypic) and 70 spe-
cies, with certain of these species (now in Phyllastrephus) in the
Timiliidae and certain genera (Nicator, Neolestes) in the Laniidae.
Sclater recognized 15 species which I consider subspecies, and four
of his subspecies I consider species. One species, Phyllastrephus
orostruthus Vincent, 1933, was described after Sclater's Systema was
published.

Pycnonotus barbatus group

This group wholly or in part has been reviewed a number of
times, notably by Hartert (1906, Nov. Zool., 13: 389, 392), Zedlitz
(1916, Jour. f. Orn., 1916: 68), Sclater and Praed (1918, Ibis, 1918:
697), Gyldenstolpe (1924, K. Sv. Vet. Akad. Handl., (3), 1, no. 3,
p. 189), Sclater (1930, Syst. Av. Aethiop., p. 369), Friedmann (1937,
Bull. U. S. Nat. Mus., 153: 105), Chapin (1953, Bull. Amer. Mus.
Nat. Hist., 75A: 147), Meinertzhagen (1954, Birds of Arabia, p. 179),
and White (1956, Bull. Brit. Orn. CI., 76: 155-157).

There is a great diversity of opinion as to the specific limits in
this group. Some put them all in one species: others make a num-
ber of species of them. I recognize four species: xanthopygos, capen-
sis, nigricans, and barbatus. These four forms are nearly, if not
completely, geographical representatives. But no intergrading or
hybrid populations connecting them are known. The first three
species are monotypic. The fourth has many subspecies, some very
different from the others, but hybrid or intergrading populations
between them all are demonstrated so that all must be conspecific.

Pycnonotus xanthopygos Ehrenberg, 1833

Though sometimes considered conspecific with P. barbatus — and
stray individuals have been taken in Egypt, in the range of P. bar-
batus arsinoe (Meinertzhagen, 1954, Bds. Arabia, p. 179) — no inter-

RAND: AFRICAN BULBULS 147

mediate or hybrid populations appear to exist. In characters of
color and of eye wattle, xanthopygos is more like the South African
P. capensis or P. nigricans than it is the geographically adjacent
P. b. arsinoe. A specific status seems indicated.

Three races have been recognized at various times. A reading
of the literature and an examination of a small number of skins
inclines me to accept Meinertzhagen's (loc. cit.) view that no races
can be recognized.

For form of citation and date, see Zimmer (1926, Field Mus.
Nat. Hist., Zool. Ser., 16: 204, and references therein).

Pycnonotus nigricans Vieillot, 1818

Sclater (1930, Syst. Av. Aethiop., p. 370) considers harterti Zed-
litz, 1916, a race of nigricans, following his review of 1918 (Ibis,
p. 698), and a trinomial is still commonly used for P. nigricans in
South Africa. However, the type series of harterti was examined by
Gyldenstolpe (1924, Kungl. Sv. Vetenskapsakad. Handl., (3), 1,
no. 3, p. 190), who found it to be without eye wattles, closely related
to tricolor, not nigricans (see under P. b. tricolor), a view that Austin
Roberts put forward a number of times.

Apparently P. nigricans has no subspecies. The overlap in range
of P. nigricans with P. barbatus may be only seasonal, but as no in-
termediate populations are known in the comparatively well-worked
South African area, specific status seems indicated for nigricans.

Pycnonotus barbatus Desfontaine, 1789

Chapin's treatment (1953, Bull. Amer. Mus. Nat. Hist., 75A:
147-154) of this species is the best to date.

The birds that I include here have sometimes been included with
xanthopygos, nigricans and capensis under the oldest name, xantho-
pygos, and at the other extreme even in recent years have been di-
vided into several species. For instance, in 1930, Sclater divided
what is here considered barbatus into four species. Within this
species, in general appearance, the following three groups are evi-
dent:

barbatus group: under tail coverts white; northern Africa.

tricolor group: under tail coverts yellow; central and southern
Africa.

dodsoni group: under tail coverts yellow; back and breast more
or less patterned ; eastern Kenya.

148 FIELDIANA: ZOOLOGY, VOLUME 35

However, none of these can be kept as species, as barbatus and
tricolor intergrade or hybridize in Gabon, and tricolor and dodsoni
intergrade in three known areas in Kenya. Other attempts to further
subdivide the complex into species have been based on interpreta-
tions of intergrading populations.

White (1956, Bull. Brit. Orn. CI., 76: 155-157) most recently
has discussed the variation within this species. He outlines six
"primary subspecies" and some of the "secondary subspecies" and
ignores others. It seems that a review of the material available is
advisable.

It should be kept in mind in studying this species that the change
from fresh plumage to worn, sun-faded body plumage is very
great.

Pycnonotus barbatus barbatus Desfontaine, 1789

Our specimens from Morocco, Algeria, and Tangier measure:
Wing, <? 101, 102; 9 95, 96, 99. Tail, d" 90, 94; 9 88, 89, 89,
91 mm.

Pycnonotus barbatus inornatus Fraser, 1843

Our specimens from Senegal, Portuguese Guinea, Sierra Leone,
Liberia, and French Sudan (Bamako and Bandiagara) measure:
Wing, c? 90, 96, 97, 98, 99; 9 90, 91, 92, 92, 92, 95. Tail, & 83,
84, 86, 88; 9 77, 78, 79, 82, 84 mm.

The smaller size, especially the shorter tail, and the slightly
lighter bill are the main characters of this lightly differentiated
race.

Pycnonotus barbatus goodi Rand

Pycnonotus barbatus goodi, Rand, 1955, Fieldiana, Zool., 34: 333 — type locality
Garoua, northern Cameroon.

This is an intermediate between three quite different forms: arsi-
noe of Egypt and Darfur; inornatus of Upper Guinea; and nigeriae
of southern Nigeria and southern Cameroon; but as it is moderately
distinct from any of these, and occupies a considerable range, it
probably should have a name. This subspecies has somewhat the
same status as an intergrade and the same claim to nomenclatural
distinctions as does P. b. ngamii and P. b. fayi.

Specimens examined. — Cameroon, Garoua, 2; Nigeria, Zaria, 2
(AMNH); and Ashen (or Air), 3 (AMNH).

RAND: AFRICAN BULBULS 149

Pycnonotus barbatus arsinoe Lichtenstein, 1823

The range usually given is the Nile Valley south to Lake No
and east to Air (Agades and Asben). I have referred Air birds to
goodi, so the western edge of the range now extends only to Darfur
and Kordofan, following Lynes (Ibis, 1925, p. 120), who also mentions
specimens of this species from the Red Sea Province of the
Sudan.

This race differs from the preceding three in having the head
much darker (black), and a whitish mark back of the ear coverts.

We have but three Egyptian specimens. Measurements: Wing,
o* 92, 92; 9 87. Tail, & 83, 85; 9 83.

Apparently what happens where this form meets the tricolor
group (minor) of the Sudan and Belgian Congo has not been criti-
cally studied. Lynes (op. cit., p. 119) and Cave and Macdonald's
(1955, Bds. Sudan, p. 246) data would seem to indicate that the two
meet without intergrading in the Sudan, while they intergrade in
West Africa; this would indicate treatment as subspecies.

Note that the name arsinoe is often emended to arsinoe or arsi-
noe, but it was originally spelled without diacritical marks.

Pycnonotus barbatus schoanus Neumann, 1905

As compared with arsinoe, this race of Abyssinia is only moder-
ately distinct, in the generally darker back and the darker breast,
as usually described. The pale fleck back of the ear coverts is only
indicated at most. There seems little difference in size. Our speci-
mens of arsinoe measure: wing, d" 92, 92; 9 87; but Meinertzhagen
gives the measurements of 26 Egyptian specimens as wing 86-95.
Our Abyssinia specimens of schoanus measure: wing, d" 92, 93, 93,
95, 96, 99; 9 90, 91, 91, 92, 92, 93. They come from the following
areas: Eastern Amhara, 8 (Gumara River, Gondar, Gendoa River,
Lake Tsana); Gojjam, 1 (Jigga); Addis Ababa area, 2 (Mugger
River, Mod jo); Sidamo, 2 (Alata, Agara Salaam). The range thus
extends from Eritrea and Bogosland south over central (east to
Dire Daoua; Friedmann) and southwestern Abyssinia to Sidamo
(but not the southeast Province of Bale). Our single specimen from
Boma, Anglo-Egyptian Sudan (d\ wing 93 mm.), is worn but pre-
sumably is this form. Apparently no intergradation with the yellow-
vented forms that replace it to the south is recorded, but somaliensis,
the next race, links it to dodsoni of East Africa.

150 FIELDIANA: ZOOLOGY, VOLUME 35

Pycnonotus barbatus somaliensis Reichenow, 1905

There has been discussion as to the validity of this form, and its
characters. See Friedmann (1937, Bull. U. S. Nat. Mus., 153: 114),
who says it is known only from Zeila and from Somadu on the in-
land plateau.

The type series came from Zeila, which is on the coast in extreme
northeast British Somaliland, not far south of Djibouti.

Of specimens I refer to this race we have: French Somaliland, 2
(Djibouti); British Somaliland, 2 (Hullier); Abyssinia, from Bale
Province, 6 (Sheik Hussein, Webi River, Luku); from Arusi Prov-
ince, 1 (Abu Kasim). This is a fairly uniform series. Measurements:
Wing, tf 88, 89, 90, 92, 93, 94, 94; 9 83, 85, 86 mm. Thus the speci-
mens average smaller than schoanus, but this difference is hardly
distinctive. The bill is about the same length in schoanus (cf 17,
18, 18, 18.5, 19, 19, 19 mm.) and in somaliensis (d> 18, 18, 18.5, 18.5,
19, 19, 20 mm.), but in the latter it is somewhat more slender, as
noted in the literature.

However, in color our specimens show this to be a fairly distinct
race. Comparing fairly fresh-plumaged birds, such as those from
Djibouti and Bale, with similar plumaged birds ( schoanus) from
Amhara and Sidamo, the upper parts are paler, due to more pro-
nounced paler edgings to the feathers; the whitish fleck back of the
auriculars is larger and more conspicuous; the pale edgings to the
otherwise dark feathers of the lower breast are more conspicuous;
the dark of the upper breast tends to continue onto the lower breast
in a streaked pattern; and the tail feathers have vaguely defined but
larger pale tips.

Taken together, these characters make somaliensis a fairly dis-
tinct race. It is interesting that the more scaled pattern with incip-
ient streaks on the breast and the more pronounced white fleck back
of the auriculars are trends in the direction of the yellow-vented
dodsoni that replaces it to the south. Further, the only immature-
plumaged bird in the series, from Luku, Bale, a locality from which
we also have two white-vented adults, has pale yellow under tail
coverts. If this is constant in the immature it would be another
racial character for somaliensis and a further indication of the close
relationship with dodsoni.

The range of P. b. somaliensis is thus French Somaliland, western
British Somaliland, and southeastern Abyssinia (to eastern Arusi
and northeastern Bale) . It probably occurs also in southern Harar,

RAND: AFRICAN BULBULS 151

but not in the Hawash Valley (Dire Daoua, etc.) where Friedmann
(op. cit., p. 113) records schoanus.

Pycnonotus barbatus nigeriae Hartert, 1921

Compared with inornatus to the northeast, the birds in the area
from southern Nigeria and central and southern Cameroon to east-
ern Gabon are characterized by having darker heads, backs and
upper breasts; the white of the belly contrasting more sharply with
the dark breast; and under tail coverts usually faintly tinged yellow.
These are the birds I refer to nigeriae, restricting the name gabonen-
sis to the birds from Gabon, in which the yellow of the under tail
coverts is about half way between the nearly white condition of
nigeriae and the bright yellow condition of tricolor.

The range of nigeriae has been variously given as only in Nigeria,
or from Nigeria east to Gabon. This seems due to difference of
viewpoint: some authors seem to have regarded birds with any yel-
low tinge in the under tail coverts as gabonensis and only birds with
pure white under tail coverts as nigeriae, while the range of nigeriae
has been extended southward on the basis of the fact that the yellow
tinge of the under tail coverts is not observable in field identification.

Examining a series of six topotypes of nigeriae from Degama,
Lower Niger, I find the under tail coverts are slightly tinged yellow
or edged yellow in two specimens and fairly heavily in one other, as
much so as in birds from as far east as parts of western Gabon.
Birds that differ little, if at all, from the southern Nigeria birds in
the amount of yellow in the under tail coverts are represented in
our series as follows:

British Cameroon: Buea, 1.

French Cameroon: Yabassi, 1; Sannaga River, 2; Yaoundi, 1;
Edolowa, 6; Bitye, Ja River, 4; Tibati, 1; Ngaoundere, 1 (back
slightly paler than others).

Spanish Guinea: Corisco Island, 2.

Gabon, Cape Esterias, 5; Mouila, 3; Labamba, 1; M'Bigou, 3;
Tchibanga, 2; Mimongo, 8 (7 nigeriae-Uke, 1 tricolor-like).

In southeastern Cameroon and in certain localities in Gabon
occur mixed populations whose allocation to either nigeriae or gabo-
nensis may be questioned ; they seem closer to the latter, as discussed
under gabonensis. The pale back of the Ngaoundere bird men-
tioned above is an indication of intergradation with goodi of north-
ern Cameroon.

152 FIELDIANA: ZOOLOGY, VOLUME 35

Despite the acknowledged intergradation of the barbatus and the
tricolor group through gabonensis, the occurrence within the range
of nigeriae of occasional specimens of the tricolor type with bright
yellow under tail coverts, such as those recorded from near Ngaoun-
dere and from Lolodorf by Bannerman (1951, Bds. Trop. West Afr.,
8: 382), might be considered an argument for keeping tricolor and
barbatus as species. However, another interpretation of this occa-
sional occurrence of individuals resembling one subspecies within the
range of another is that it is a case of extreme variation of the local
population (Rand, 1948, Auk, 65: 416), a view that is adopted here.

Pycnonotus barbatus gabonensis Sharpe, 1871

This name I restrict to the birds with pale yellow under tail
coverts, more or less half way between the faintly yellow-tinged
white under tail coverts of nigeriae and the bright yellow ones of
tricolor. Except for the under tail coverts these three forms are
very similar. Birds that show such an intermediate condition are
represented in our collection from the following localities:

Cameroon: Sangmelima, 2; Yokadouma, 4.

Gabon: Fenian Vaz, 4.

Though the southeastern Cameroon birds are more yellowish
than most nigeriae specimens, they are not as yellow as exact in-
termediates would be. Of the four Fernan Vaz birds one is an exact
intermediate nigeriae X tricolor, one is nearly typical nigeriae, and
the other two fall between. This suggests that southeastern Came-
roon and parts of northwestern Gabon are on the north and west
edge of the band of direct intergradation of tricolor-nigeriae. Its
southern extent is probably in extreme southwestern French Middle
Congo (between Brazzaville and Dolisie; see Malbrant and Ma-
clatchy, 1949, Faune de L'Equat. Afr. Franc, 1, Oiseau, p. 301).
That this is a narrow band is indicated by the fact that our Middle
Congo specimens from Impfondo, Gamboma, and Djambala are
typical tricolor, and those from central and southern Gabon (Mouila,
Labamba, M'Bigou, Tchibanga) are nearly typical nigeriae.

Its range is thus central coastal Gabon and from extreme south-
ern French Congo to extreme southeastern Cameroon.

Our specimens suggest that this area is not characterized by a
uniformly intermediate but by a highly variable population perhaps
better considered a mixed population resulting from the hybridiza-
tion of two species, tricolor and barbatus. However, nigeriae is also
an inornatus-tricolor intergrade in characters other than the under

RAND: AFRICAN BULBULS 153

tail coverts; it has the dark general color of tricolor and a tinge of
yellow on the under tail coverts, both tendencies toward tricolor,
and it seems best to consider gabonensis as another population in-
termediate between subspecies.

The name gabonensis is thus applied to an intermediate popula-
tion, as intended by Sharpe in his original description.

Pycnonotus barbatus tricolor Hartlaub, 1862

This race differs sharply from all those preceding in having the
under tail coverts bright yellow. Otherwise it is very similar in color
and size to P. b. gabonensis and nigeriae, with which it intergrades
over a long narrow area in Gabon, as discussed under gabonensis.

The range is from northern South West Africa (an old record
only) and southern Angola to the French Middle Congo, the Cen-
tral Congo basin, Kivu Highlands, southern Uganda, eastern shore
of Lake Tanganyika, western Tanganyika, and northern Northern
Rhodesia. The type locality was restricted to northern Angola by
Zedlitz in 1916.

As mentioned above, this race intergrades on the northwest with
gabonensis; on the north it intergrades over a large area with minor;
to the east it intergrades with the small black-headed micrus, through
populations that look like fayi; to the east and south it intergrades
with the black-headed races micrus and layardi, through populations
that look like fayi, and through ngamii.

The following four names are synonyms:

Pycnonotus tricolor tanganjicae Reichenow, 1911, from the north
end of Lake Tanganyika (see Chapin, 1953, Bull. Amer. Mus. Nat.
Hist, 75A:150).

P. b. harterti Zedlitz, 1916, Jour. f. Orn., 64: 71; type locality
Huilla, Mossamedes; described as like tricolor but larger; forehead
and top of head darker; under wing coverts more brown ish- tinged ;
wing, d" 104-105; 9 96-102 (tricolor wing, d" 92-99; 9 88-90 mm.);
range, Mossamedes and Benguella, intergrading with tricolor far-
ther north.

Sclater and Praed (Ibis, 1918, p. 698) refer this form to the nigri-
cans group, saying that it has an eye wattle, though the feature is
not as pronounced as in nigricans. Roberts (1935, Ann. Trans. Mus.,
16: 130) says it is a race of tricolor.

Apparently both species occur in southern Angola. Hartert
(1906, Nov. Zool., 13: 391) recorded nigricans with bright eye wat-

154 FIELDIANA: ZOOLOGY, VOLUME 35

ties from Benguella, whence he also recorded two specimens with
bare, protruding, but dark eyelids; and I have examined specimens
of tricolor from Mossamedes.

There seems little reason to doubt that the name harterti was
actually based on specimens of the species tricolor, and Gyldenstolpe
(1924, Kungl. Sv. Vetenskaps. Handl., (3), 1, no. 3, p. 190), who
examined the type series, referred them to P. tricolor [=barbatus]
harterti, rather than nigricans.

For comparison I have the following specimens: Mossamedes
(Capelongo), 4; Benguella, 6 (Dondi, 4, Chitau, 2) and Loanda, 6
(Golunga-Alta, 4, Luanda, 2). Measurements: wing, Mossamedes
<? 100, 103; 9 99, 101; Benguella <? 95, 99; 9 93, 96, 97, 99;
Loanda & 93, 93, 96, 97, 97; 9 93. It is true that the southern
birds are slightly larger, but the difference is hardly distinctive. In
color one of the four Capelongo birds has a slightly more blackish
crown than do any of the Loanda birds, but the other three do not.
I find the under wing coverts a variable character.

One might have expected harterti to refer to an intermediate tri-
color-layardi population like ngamii, but it is so close to northern
Angola tricolor that it seems inadvisable to separate this population
from tricolor.

Pycnonotus tricolor vaughanjonesi White, Ibis, 1944, p. 146; type
locality, Mwinilunga, Northern Rhodesia; described as differing from
P. t. tricolor in having the fore part of the crown much more blackish
and in being distinctly larger (wing, d" 96-104; 9 88-100). It was
said to be much more similar to tricolor than to ngamii. Its range
was given as the Mwinilunga district of Northern Rhodesia and
northwest to the Katanga at Musonoie.

Pycnonotus tricolor limes Horniman, 1956, Ann. Mag. Nat. Hist.,
9, ser. 12, p. 366 — Mwinilunga, Northern Rhodesia. Apparently
this was described in ignorance of P. b. vaughanjonesi.

We have, of birds referred to this race, one topotype and two
from Musonoie (wing, cf 99, 102; 9 96). In color I cannot see that
they differ from many specimens of tricolor from Angola and the
Middle Congo. Their slightly larger size is equaled by birds from
other parts of the range of tricolor.

However, that this general area is in the northern edge of the
area of intergradation of tricolor-ngamii is shown by specimens from
Ndola (which I refer to ngamii) and from Kasempa and Balovale
(which are intermediate but which I refer to tricolor).

RAND: AFRICAN BULBULS 155

Specimens examined. — Angola, 12 (Mossamedes, 4 [AMNH];
Benguella, 6 [+2 AMNH]; Loanda, 6). French Middle Congo, 9
(Djambala, Gamboma, Impfondo). Belgian Congo, 7 (Katobwe,
Kabengera, Magira, Musonoie). Northern Rhodesia, 6 (Balovale, 2;
Kasempa, 2; Mwinilunga, 2). Uganda, 16 (Kigezi, 1; Toro, 8; Sem-
liki Valley, 7).

Wing measurements are as follows:

Angola: Mossamedes, o" 100, 103; 9 (?) 99, 101; Benguella, & 95,
99; 9 93, 96, 97, 98, 99; Loanda, o" 93, 93, 96, 97, 97; 9 93, 93.

French Middle Congo: <? 95, 95, 96, 99, 101; 9 87, 87, 90, 91.

Belgian Congo: Haut Luapula, 9 91, 92, 93; Musonoie, d" 102;
9 97.

Uganda: Kigezi, <? 91; Toro, <? 93, 97, 98, 100; 9 89, 91, 91, 91;
Semliki Valley, & 95, 97, 97; 9 89, 89, 92 mm.

Pycnonotus barbatus minor Heuglin, 1869

This race is fairly distinct when Sudan birds are compared with
tricolor from Angola, being slightly paler and more grayish above and
paler on the breast. The sometimes quoted smaller size is not evi-
dent (see measurements). It intergrades with tricolor over a broad
area, so that the outlining of ranges where they intergrade is arbi-
trary. The type locality is the upper White Nile and birds from as
far south as Lake Victoria, Lake Edward, Stanleyville, and extreme
southeastern Cameroon are referred here. The range thus is from
the southern Sudan and Dafur to southeastern Cameroon, northern
Congo, and northern Uganda.

Besides intergrading with tricolor to the south, this race inter-
grades with fayi to the east. Birds from the Mount Elgon area
show a definite approach to the larger, darker fayi. Intergradation
with the white-vented forms to the north has not been recorded.

Synonyms are:

Pycnonotus layardi phaeocephalus Mearns, Smiths. Misc. Coll.,
56, no. 20, p. 8, 1911; type locality, Kikandwa, Uganda.

This name was applied to the population that lies between Lake
Victoria and Lake Albert. It averages slightly larger than minor of
the Sudan and is slightly larger than some populations of tricolor.
It represents an intergrade between minor and tricolor and, espe-
cially in size, with fayi to the east. But the slight average differences
do not necessitate a name for the population. Of course this name
could almost equally well be considered a synonym of tricolor.

156 FIELDIANA: ZOOLOGY, VOLUME 35

Pycnonotus barbatus escherichi Grote, 1922, Jour. f. Orn., 70: 485;
type locality, Kumbe, eastern Cameroon [= northern Middle Congo,
4° 36' N., 16° 14' E., according to Times Atlas]; described as very like
tricolor but with grayish brown (not white or grayish white) under
wing coverts; and as differing from the very similar minor in having
a browner crown; wing, 95-97 mm. I have two specimens from the
Cameroon border just west of the type locality (Yokadouma and
Garoua Boulai) which I cannot separate from Uganda minor.
Though Sclater (1930, Syst. Av. Aethiop., p. 371) recognized it,
I agree with Chapin (1953, Bull. Amer. Mus. Nat. Hist., 75A: 147),
and Bannerman and Bates (Ibis, 1926, p. 796) in considering it a
synonym of the tricolor-minor intergrade.

Specimens examined. — Uganda, 22 (Mount Elgon area, 4; Bu-
soga, 2; Mengo, 7; Entebbe area, 6; Bunyoro, 3); Sudan, 17 (Torit);
French Cameroon, 2 (Yokadouma and Garoua Boulai).

Wing measurements are as follows:

Uganda: Mount Elgon area, d" 100, 101; 9 93, 94. Busoga,
d" 97; 9 92. Mengo, d" 90, 96, 98, 98; 9 91, 92. Entebbe area,
& 90, 91, 91, 97, 98. Bunyoro, d" 96, 98; 9 90.

Anglo-Egyptian Sudan: Torit, & 85, 87, 89, 94, 95, 96; 9 89,
90, 90, 91, 92, 95, 97.

Cameroon: d* 97; 9 93.

Pycnonotus barbatus spurius Reichenow, 1905

Sclater (1930, Syst. Av. Aethiop., p. 370) considered this name a
synonym of P. dodsoni, and Friedmann (1937, Bull. U. S. Nat. Mus.,
153: 109) considered it applicable to the race near dodsoni, which I
interpret as P. b. peasei. Hartert (1906, Nov. Zool., 13: 391) and
Benson (1946, Ibis, p. 45), however, show that a bird of the tricolor
group occurs in southern Abyssinia and while the original descrip-
tion is scant, the name spurius applies to this tricolor-Wke bird, with
which conclusion I agree.

Our specimens are very definitely of the tricolor group. They are
like fayi, but with head black (not brownish), and throat more exten-
sively dark, and blackish (not brownish). As Reichenow stated in
his original description, it is black-headed like layardi but the black
of the throat is more extensive. Also the upper parts are darker
than in layardi.

Measurements: Wing, & 95, 96; 9 91, 92. Tail, & 87, 90;
9 83, 85. Culmen, cf 19.5, 20; 9 18.5, 19 mm.

RAND: AFRICAN BULBULS 157

Our specimens come from Gedel Mountains (about 6° 50' N.,
39° 5' E.) and Alghe (about 5° 30' N., 38° 20' E.) in southern Abys-
sinia at 6000 to 8500 feet altitude. The three Gedel birds are very
similar in color. The two Alghe birds differ slightly in the direction
of the very different peasei (the dodsoni-tricolor group intermediate)
in having a trace of a pale fleck back of the auriculars and a slight
tendency, toward the lower edge of the dark breast, to break up into
streaking. However, this tendency is slight, though important as
indicating relationship.

The type locality is the general area Ennia Gallaland, about
Lat. 8° N. and Long. 42° E. The range is thus apparently a narrow
band from Bale to northern Boran. Apparently it is completely iso-
lated from that of others of the yellow-vented tricolor group.

The range of this race is evidently a small one, and we have
records of other, quite different races from nearby; thus schoanus
with white under tail coverts is recorded south to the Gardulla area
just west and north (about Lat. 5° 40' N., Long. 37° 30' E.). The
type locality of the mottled-breasted dodsoni is the Sillul River, very
close to Reichenow's Ennia Gallaland; we have specimens of dodsoni
from nearby eastern Arusi and northeastern Bale, and Benson re-
cords peasei from Yavello (about 5° N., 38° 10' E.).

The close proximity of the ranges of these representatives of these
three groups — barbatus (schoanus and somaliensis) , tricolor (spurius),
and dodsoni (dodsoni and peasei) — seems correlated with habitat, as
outlined by Benson (1946, Ibis, pp. 45-47) for a region in southern
Abyssinia where all three occur within an area of small radius:
spurius is largely a bird of highland areas of higher rainfall; peasei
a bird of juniper woods in the highland areas of lower rainfall and
the "thorn scrub" of the surrounding plains; schoanus a bird of the
"rather less arid, more luxuriant thorn" country.

Intergradation between the tricolor and the barbatus groups is not
indicated in this area. But partial intergradation between the barba-
tus and the dodsoni groups is shown by somaliensis (which see above)
and complete intergradation between the tricolor and the dodsoni
groups is shown by peasei (see below).

Pycnonotus barbatus fayi Mearns, 1911

This race is only lightly characterized compared with tricolor and
minor to the west. It differs from them chiefly in the more blackish
crown, side of head and throat, and in the slightly larger size attained
by some individuals in some areas. The under tail coverts also aver-

158 FIELDIANA: ZOOLOGY, VOLUME 35

age slightly brighter yellow; from spurius this race differs in that
the head is brownish black, not black, and the throat is less exten-
sively dark, and is brownish black, not black.

This name has been used for the population of the Kenya high-
lands only. It intergrades gradually with the minor-tricolor popu-
lation of Uganda, and to the east it intergrades abruptly with the
quite different dodsoni through the race peasei. This race, fayi, is
intermediate in color between the brown-headed minor-tricolor and
the black-headed micrus of eastern Tanganyika. Southward there
are other such intermediate populations. Chapin (1953, Bull. Amer.
Mus. Nat. Hist., 75A : 150) points out that such intermediates from
Tabora, Tanganyika Territory, also resemble fayi.

The range should thus be extended south through central Tan-
ganyika. Iringi birds have been referred to fayi, but our specimens
from there are definitely black-headed, showing that fayi does not
extend this far in Tanganyika.

Specimens examined. — Kenya: Kavirondo, 6; Nandi, 3; Burnt
Forest, 1; Kericho, 5; Naivasha, 1; Nakuru, Equatorial Farm, 10;
Kinangop, Aberdares, 1; Amala River, 3; Elgeyu, 1; Laikipia, 1;
Embu, 1; Nairobi area, 10; Narok, 2; Doingo Narok, 3. Tanganyika
Territory: Mwanza, 1.

Wing measurements are as follows:

Kenya: Kavirondo, & 92, 93, 98; 9 94, 96. Nandi, d" 98, 100,
101. Burnt Forest, cf 96. Kericho, d" 92, 92, 94; 9 89, 90. Nai-
vasha, cf 100. Nakuru, Equatorial Farm, & 96, 97, 99, 99, 100, 104;
9 93, 93, 94, 95. Amala River, d" 94, 97, 98. Elgeyu, & 95.
Laikipia, d* 98. Embu, & 94. Nairobi area, cf 92, 92, 94, 94, 96,
99; 9 90, 91, 92, 93, 95. Narok, d* 95; 9 89. Doingo Narok, cf 93,
94, 96.

Tanganyika Territory: Mwanza, cf 94.

Pycnonotus barbatus ngamii Ogilvie Grant, 1912

As fayi is the intergrade between minor plus the minor-tricolor
complex and the small, black-headed micrus of eastern Africa, so
ngamii is the intergrade between the brown-headed tricolor and the
large, black-headed layardi of South Africa.

In appearance this race is very similar to fayi, differing chiefly
in averaging slightly larger and slightly paler above; from tricolor it
differs in having the upper parts slightly paler and the crown more
blackish brown, contrasting more with the back. Its range is from

RAND: AFRICAN BULBULS 159

central northern Northern Rhodesia (Ndola) to Bechuanaland (Lake
Ngami).

Our specimens from Maun are practically topotypical, coming
from about 40 miles northwest of Lake Ngami.

Pycnonotus (Loidorusa) tricolor annectans Roberts, 1932, was de-
scribed with type locality as the Machile River, Northern Rhodesia.
White (1945, Ibis, p. 574) thinks the type more probably came from
the nearby Chobe Estuary, Bechuanaland, and suggests this as the
type locality. We have three specimens from Kabulabula, which is
on the Chobe River, and hence these specimens are practically topo-
typical. Compared with the Maun specimens the slight difference
in the paler back of two of the three specimens may well be due to
wear, and the two small series are so similar that annectans must
be considered a synonym of ngamii, as a number of authors have
suggested.

Specimens from Ndola average slightly smaller than Maun birds,
and the back is slightly paler, but they are otherwise very close to
ngamii, with which I include them. Apparently Ndola is very near
the northern edge of the range, for we have birds closest to tricolor
from just to the north and west (vaughnjonesi= tricolor).

Two Balovale (western North Rhodesia) birds do not admit of
definite allocation of the population; one is tricolor-like, with brown
crown; one is more like ngamii, with blackish brown crown.

There is the possibility that ngamii ranges to western Angola
and that the recorded northwestern South West African birds be-
long here.

Specimens examined. — Northern Rhodesia: Ndola, 10. Bechu-
analand: Maun, 2; Kabulabula, 3.

Wing measurements are as follows: Ndola: <? 99, 100, 100, 100,
101; 9 94, 98; sex? 92, 94, 102. Maun: <? 104, 106. Kabulabula:
<? 100, 104.

Pycnonotus barbatus layardi Gurney, 1879

This is the most southern population of the species and is charac-
terized by a black crown (not brownish black), contrasting with the
back. The size is large, but not more so than in some other more
northern populations of the yellow-vented group.

Compared with ngamii it has a black (not brownish black) crown
and more black in the sides of the head and throat.

160 FIELDIANA: ZOOLOGY, VOLUME 35

The range is from east Cape Province (except the highland range
of the next race) through Natal and Transvaal to eastern Northern
Rhodesia, Nyasaland and Portuguese East Africa, intergrading over
a wide area with the smaller, black-headed micrus to the north.

The variation in color of upper parts from series to series may be
due to effects of wear and of museum age. I attach no taxonomic
importance to the slightly paler back of Northern Rhodesia and
Nyasaland birds, and its reported occurrence in Portuguese East
African birds. The measurements given below suggest not only a
decrease in size northward from Transvaal but also a decrease in
size toward the coast in Natal and in Mozambique.

The name Pycnonotus layardi pallidus Roberts, 1912, was applied
to Boror, Portuguese East African birds characterized as like lay-
ardi but paler and with more yellow on the abdomen, a tinge extend-
ing up the center of the breast; the brown crop sharply defined, and
the throat lighter. Wing, & 94, 94 mm. We have a single male
from Lumbo, an area often included in the range of pallidus (when
recognized). The yellow on the abdomen is matched by individual
variation in other series. The general plumage is pale but the bird
is worn. It seems advisable to consider this a name for an inter-
grading population between layardi and micrus, and I synonymize
it with layardi.

Specimens examined. — Natal : Durban, 2. Transvaal : Pretoria, 2 ;
Crocodile River, 1; Hamanskraal, 1; Newington, 4. Southern Rho-
desia: Zimbabwe, 1. Northern Rhodesia: Lundazi, 4. Portuguese
East Africa: Lumbo, 1.

Wing measurements: Natal: & 94; 9 96. Transvaal: d" 97, 99,
100, 101, 102; 9 98, 94. Southern Rhodesia: & 100. Northern
Rhodesia: d" 95; sex? 94, 94. Nyasaland: d" 97. Portuguese East
Africa: d* 95. Culmen: <? 19-21; 9 19.5 mm.

Pycnonotus barbatus tenebrior Clancey, 1955

This race was described (Durban Museum Nov., 4: 207) from
5500 feet altitude, Mount Currie, near Kokstad, eastern Cape Prov-
ince, and the range was outlined as the highlands of eastern Cape
Province, Natal, and the Drakenberg Range in southern Basutoland.
This race of higher altitudes was characterized as being darker on the
upper parts, throat and flanks, and with less rich yellow under tail
coverts compared with layardi. At the same time Clancey suggested
that the coastal Pondoland and Natal populations might represent
yet another race, which he refrained from naming.

RAND: AFRICAN BULBULS 161

Through the courtesy of Mr. Clancey I have been able to exam-
ine the type and a topotype of tenebrior taken in April, 1955, and
four specimens of layardi from eastern Transvaal, taken in July and
August, 1952 and 1953, to avoid the effects of foxing. At first glance
the darkness of the Mount Currie birds is apparent. However, the
tenebrior specimens, taken in April, are moulting and have some worn,
faded feathers which are less different from the somewhat worn,
faded feathers of the specimens of layardi taken in July and August.
It is probable that strictly comparable material will show less differ-
ence, but it seems advisable to accept this darker, highland race.

Measurements of topotypes: c? wing, 102; culmen, 21, 21 mm.

Pycnonotus barbatus micrus Oberholser, 1905

This small, black-headed East African race differs from layardi in
its average smaller size and slightly darker back.

The northern edge of its range is in the Mount Kilimanjaro area
(type locality, Taveta, Kenya). Southward it intergrades in size
over a wide area with layardi (see measurements), but the slightly
darker coloration of the upper parts seems to prevail to southern
Tanganyika, and to this region I extend the range. Inland this race
intergrades with P. b. tricolor, increasing in size and with the black
of the head becoming brownish, locally, at least, becoming identical
with fayi, as in Tabora.

In addition to size, mentioned above, certain variations are notice-
able in the series I refer here, notably the much darker, colder upper
parts of the Iringi birds; these, however, were collected in 1952,
while the Kilimanjaro birds were collected in 1920; museum age
rather than geographical variation may well be the reason for the
color difference. The bills of these Iringi birds also average slightly
smaller than some other populations, as is also reported for Matengo
highland birds.

The type locality for micrus is Taveta and one might expect
birds from there to show some approach to the dodsoni type of birds,
peasei of farther north, but our topotype is definitely of the tricolor
group and similar to our Mount Kilimanjaro birds.

Pycnonotus tricolor naumanni Meise, 1934 (Orn. Monatsb., 42:
116), from the Matengo highlands was described as similar to micrus,
with a similar short bill (exposed culmen, o" 14-15 mm. compared
with 14-15 in micrus and 16.5-18.5 in layardi) but with a longer
wing ( d" 96-100, 9 90-94 mm.) and purer white under parts. I have

162 FIELDIANA: ZOOLOGY, VOLUME 35

no topotypes, but on size it is evidently one of the intermediate lay-
ardi-micrus populations and as Iringi highlands, Mahenge, and Dar-
es-Salaam birds seem best referred to micrus, I do the same with
naumanni.

Specimens examined. — Kenya : Taveta, 1. Tanganyika Territory :
Kilimanjaro, 4; Mount Meru, 2; Usambara, 2; Kilosa, 2; Mahenge,
1; Iringa, 4; Dar-es-Salaam, 1.

Measurements are as follows:

Kenya: sex? wing, 93; tail, 81; culmen, 19.

Tanganyika: Kilimanjaro: Wing, cf 89, 90, 91; 9 92. Tail, cf 78,
80, 80; 9 78. Culmen, cf 19, 19, 19; 9 19.5. Mount Meru: Wing,
cf 94; 9 87. Tail, d" 81; 9 78. Culmen, d* 19; 9 20. Usambara:
& wing, 92; tail, 78; culmen, 19. Kilosa: Wing, d" 94, 99. Tail, 81,
85. Culmen, 20, 20.5. Mahenge: 9 wing, 90; tail, 77; culmen 19.5.
Iringa: 9 wing, 94, 95, 95; tail, 82, 83, 83; culmen, 18, 18.5, 19.
Dar-es-Salaam: 9 wing, 91; tail, 78; culmen, 18.

Pycnonotus barbatus peasei Mearns, 1911

This race is the intergrade between Pycnonotus b. dodsoni and the
surrounding populations: micrus to the south; fayi to the east; and
spurius, the isolated member of the tricolor group of Abyssinia.

From spurius it differs in smaller size, the tendency toward white
tips on the tail feathers, the tendency toward white spots behind the
ear coverts, and the greater tendency to distinct mottling of the up-
per breast and more conspicuous pale edging of the back feathers.

From fayi it differs in a similar manner, with a greater discrep-
ancy in size (for fayi is larger than spurius) and in the darker throat.

From micrus it differs in a similar manner, but with less difference
in size (for micrus is smaller than fayi) and with a darker throat.

These differences are all in the direction of P. b. dodsoni, and
peasei stands about half way between the tricolor group and the very
different dodsoni.

The range is coastal Kenya from Sokoke to Vanga (but not in-
land in the Taru area where dodsoni takes its place) ; along the eastern
slopes of the Kenya Highlands; and in extreme southern Abyssinia
(Boran and probably eastward to southern Bale).

The Abyssinian population seems to be quite isolated and to
occupy a narrow band. Benson (1946, Ibis, p. 46) showed that it
occupies a habitat different from that of spurius. He recorded spu-
rius at Alghe; peasei at Yavello, about 45 miles south, and then

RAND: AFRICAN BULBULS 163

south to near Mega, another 65 miles. Friedmann recorded peasei
(under the name spurius; see Benson, 1946, Ibis, p. 46) from nearby
Sagon River, Tertale and Bodessa. The Kenya border is about 25
miles south of Mega and dodsoni ranges just north of the border at
Yebo and Chaffa to the west and to the border just east at Moyale.
Thus the range of peasei here may not be much more than 100 miles
wide. Its east-west extension may extend the length of the contact
of spurius and dodsoni.

The population of the Kenya Highland slopes extends at least
from the Mount Kenya area (North Waso Nyiro, Archer's Post) to
the upper Athi River. One would expect at least a tenuous con-
nection with the coastal Kenya population, but birds from the Kili-
manjaro area and Taveta are definitely micrus of more southern
range, while dodsoni of the Kenya lowlands extends south to the
Chyulu Hills and Tsavo, less than 60 miles away. Thus peasei, if
it exists in this area, does so in this very narrow belt. The Kenya
coastal belt of this species may also occupy a narrow range, for the
localities are all near the coast.

In the coastal Kenya population the variation is only moderate.
The narrow white tip to the tail feathers is present and up to 3 mm.
wide in about half of the specimens. In a few specimens it is indi-
cated only by a pale gray tip; in nearly half of the specimens the tail
is worn so that the original presence or absence of a white tip cannot
be determined. The distinctiveness of the white fleck back of the
ear coverts, the mottling of the breast, and the scaling of the back
vary somewhat, but not excessively.

The three specimens from Athi and Kaite Rivers are at the lower
limit of size of the coastal Kenya series but fall within its range of
variation in color. The two Abyssinia (Yavello) specimens are at
the upper limit of size of the coastal Kenya series, but also fall within
its range of color variation. It should be noted here that two speci-
mens of spurius from Alghe in Abyssinia, not far north of Yavello,
show a slight tendency toward peasei.

I have one series of birds, those from the North Waso Nyiro and
from Archer's Post, whose allocation to peasei or to dodsoni seems
doubtful. In size they fit better with peasei, and in color certain
specimens are very similar to average peasei, but other specimens
show characters that approach those of dodsoni. This population
definitely bridges the gap between the two and could be referred to
either form. Tentatively I place it here.

Sclater synonymizes littoralis van Someren with micrus. The
name Pycnonotus dodsoni littoralis van Someren (1923, Bull. Brit.

164 FIELDIANA: ZOOLOGY, VOLUME 35

On. CL, 43: 153), type locality Changamwe, near Mombasa, was
proposed for the coastal population of peasei. The specimens I have
examined and commented on above included two topotypes. "lit-
toralis" is no closer to micrus than it is to dodsoni, and I see no
reason, except the inconclusive one of geography, for separating it
from peasei. That peasei is the intergrade between the tricolor group
and dodsoni is indicated by material from all the localities listed,
except that from the Sokoke-Mongeya area, where it indicates that
peasei and dodsoni should be considered species. This is discussed
under dodsoni.

A case could be made for giving separate individual names to
each of the three separate populations I include here in peasei. P. b.
littoralis van Someren is available for the coastal Kenya bird which
is presumably the dodsoni-micrus intergrade; peasei would be the
name for the dodsoni-fayi intergrade on the slopes of the Kenya
highlands; and a new name would be needed for the dodsoni-spurius
intergrade of southern Abyssinia. However, this would be bestow-
ing names on populations that are externally indistinguishable, and
I think it better to group them under one name, which stresses their
common origin in dodsoni.

Specimens examined. — Abyssinia: Yavello, 2. Kenya: North
Waso Nyiro, 5; Archer's Post, 1; Upper Athi River, 2; Kaite River,
Ukamba, 1; Shimba Hills, Vanga, 1; Ganda, 6; Changamwe, 2;
Mombasa, 5; Rabai, 8; Sokoke Forest, 8; Mongeya Hills, 4.

Measurements are as follows:

Abyssinia: Wing, & 93; 9 88. Tail, & 78; 9 76. Culmen,
& 18; 9 18.

Kenya: Athi and Kaite Rivers: Wing, 9 79, 80, 82. Tail, 9 68,
71, 72. Culmen, 9 17, 17, 18. Archer's Post: <? wing, 80; tail, 70;
culmen, 18. North Waso Nyiro: Wing, d* 85, 89, 89; 9 81, 83.
Tail, cf 71, 75, 76; 9 67, 68. Culmen, tf 17.5, 18, 18; 9 17, 17.
Kenya coastal area: Wing, d1 (19) 82-92 (av. 86.1); 9 (10) 82-88
(av. 85.2). Tail, & (20) 68-79 (av. 73.5); 9 (10) 71-78 (av. 73.6).
Culmen, d* (20) 17.5-20 (av. 18.3); 9 (10) 17-18 (av. 17.6).

Pycnonotus barbatus dodsoni Sharpe, 1895

Differs from peasei in smaller size, in the more conspicuous white
fleck behind the ear coverts, in the more distinct mottling of the
breast, the more distinct pale edges of the back feathers, and the
larger white tip to the tail feathers.

RAND: AFRICAN BULBULS 165

Its range is Italian Somaliland (Higleleh, in the Haud of Ogaden
country) and eastern Abyssinia from parts of the upper Webi She-
beli River (Sillul, Lammo, Dada Uebi) southward over the lowlands
of Kenya on the coast to Kipini area and inland in the southeast to
Sokoke, Tsavo, and Chyulu Hills, and westward to Lake Rudolf,
extending into extreme southern Abyssinia at Yebo and Chaffa
(Friedmann, 1937, Bull. U. S. Nat. Mus., 153: 108).

The range is usually given to include British Somaliland. This
seems based on Elliot's record of two specimens collected at Hullier.
Elliot also collected one specimen at Higleleh, Ogaden, and says that
since Dr. Smith collected it at Lamoo [= Lammo] it would seem to
go to the coast also.

There are two errors here. First, while Elliot's Higleleh specimen
is the yellow-vented dodsoni, the two Hullier birds are the white-
vented somaliensis; secondly, Smith's locality, Lammo, is not on the
coast but is a river near the type locality, Sillul.

The populations included here are not all uniform. The Juba
River and the Sokoke area birds are smallest and have the subspe-
cific characters most pronounced, though the Juba River birds are
slightly paler. Lamu and Kipini birds average larger but are other-
wise similar; Marsabit birds also average larger, and perhaps a shade
duller and darker than any of the above, showing a very slight tend-
ency toward peasei; one of the two Tsavo birds is also large but agrees
in color with Sokoke birds. The two Chyulu birds are also large and
are slightly darker, less brownish than Sokoke and Tsavo skins, but
very similar to Marsabit birds. In fairly fresh plumage the Chyulu
birds have the scaling of the back slightly less distinct and the mot-
tling of the breast slightly less plentiful than in Juba-Sokoke birds,
a tendency toward peasei, but the white tips to the tail feathers are
well developed and the birds are more like dodsoni than peasei. The
Higleleh bird is in fairly fresh plumage and agrees well with Juba
River birds in color and size.

The name P. 6. chyulu van Someren, 1939, was given to birds
from the Chyulu Hills on the basis of darker general color and paler
yellow under tail coverts. The color of the under tail coverts of our
two topotypes is not as deep as in some other birds but falls within
the range of variation of several series. The slightly darker color
has been mentioned above.

Pycnonotus dodsoni teitensis van Someren, 1922, was applied to
Tsavo birds on the basis of their being larger than dodsoni. While
our two topotypes are somewhat large, the differences are not great,

166 FIELDIANA: ZOOLOGY, VOLUME 35

and this type of difference, similar but of larger size, occurs in at
least three other populations — Marsabit, southern Abyssinia, and
the Lamu-Kipini area — making nomenclatural treatment impossible.

I have treated dodsoni as conspecific with the tricolor group be-
cause in southern Abyssinia and on the slopes of the Kenya highlands
there are populations (peasei) that are intermediate between the two,
geographically and morphologically. Another similar population oc-
curs on the Kenya coast, from Vanga to Sokoke. However, from
Sokoke and the nearby Mongeya Hill I have a series that is obvi-
ously dodsoni, and another series that is just as obviously peasei.
The range of variation in each series is not great and though overlap
is not quite complete it is approached. Two entities seem involved.
That peasei is the more common, widespread form is indicated by its
being represented in our series by several collections made in several
years (1920, 1921, 1924, 1932), while that dodsoni is the more re-
stricted, though also common, form is indicated by the fact that the
series was all collected in one trip, in 1921. This suggests that where
desert scrub meets coastal forest the two races may come close to-
gether (as peasei and spurius and dodsoni do in southern Abyssinia;
see under peasei) and wide-ranging native collectors may have
brought together subspecies from two areas.

Two lines of nomenclatural treatment are thus open : On the basis
of these two series labeled from the same locality, to ignore the evi-
dence of intergradation of dodsoni and the tricolor group through
peasei and keep dodsoni a separate species, calling peasei a species
hybrid (and ignoring also the tendency toward intergradation of
dodsoni with the barbatus group through somaliensis) ; or, accepting
the demonstrated intergradation, to call dodsoni a subspecies and
postulate that further field work will show that peasei and dodsoni
do not actually live in the same areas in the Sokoke-Mongeya area
but are separated ecologically at least. This is the view adopted here.

Specimens examined. — Kenya: Tsavo, 2; Chyulu Hills, 2; Sokoke
Forest, 12; Kipini area, 4; Lamu Island, 1; Wajir, 2; Moyale, 2;
Marsabit, 15; Juba River (Serenli and Mandaira), 12. Abyssinia:
Higleleh in Haud, 8.30 N., 40.30 E., 1.

Measurements are as follows:

Kenya: Tsavo: Wing, d* 88; 9 77. Tail, d" 76; 9 65. Culmen,
c? 18; 9 16.5. Chyulu: Wing, d" 83; 9 85. Tail, cf 73; 9 74.
Culmen, d" 17; 9 18. Sokoke Forest and Mongeya Hill: Wing,
cf 76, 77, 78; 9 (9) 71-80 (av. 75.8). Tail, & 60, 60, 64; 9 (8)
63-66 (av. 64.3). Culmen, c? 16.5, 16.5, 17; 9 (9) 15-17 (av. 16.4).

RAND: AFRICAN BULBULS 167

Lamu and Kipini: d" wing, 80, 80, 81, 84, 85; tail, 70, 70, 71, 72;
culmen, 16, 18, 18, 18, 19. Wajir: 9 wing, 77, 80; tail, 64, 67; oil-
men, 15, 16. Moyale: Wing, c? 83; 9 81. Tail, & 69; 9 68.
Culmen, d" 18; 9 16.5. Marsabit: Wing, d* 77, 82, 82, 83, 84, 85,
85, 86, 86; 9 80, 80, 81, 81, 81, 82. Tail, & 68, 68, 69, 70, 70, 70,
71, 73; 9 66, 67, 67, 68, 68, 70. Culmen, d" 17, 17.5, 17.5, 18, 18,
18, 18.5; 9 16, 16, 17, 17, 17, 18. Juba River: Wing, c? 75, 79, 80,
80, 81, 82, 82; 9 74, 75, 75, 76, 77. Tail, d" 67, 67, 69, 69, 69, 70,
72; 9 64, 65, 65, 66, 67. Culmen, cf 16, 16, 16.5, 16.5, 16.5, 17, 17;
9 15, 15, 15.

Abyssinia: Highleleh: 9 wing, 75; tail, 65; culmen, 17.

Pycnonotus masukuensis Shelly

This is an East African bird with a scattered distribution on a
few mountain masses from Kenya and the eastern Congo to Nyasa-
land. The abundance of the species varies greatly with the locality.
In western Kenya it is a very rare bird (Jackson and Sclater, 1938,
Birds of Kenya . . . , p. 864), but in the Usambara Mountains it is
evidently very common (Friedmann and Loveridge, 1937, Bull. Mus.
Comp. Zool., 81 : 235, 236) and also in Iringa highlands, judging by
our series of 11 specimens collected in seven days.

The representatives at each end of the range look rather differ-
ent: kakamegae from Kenya has a gray throat and sides of head and
a darker gray crown; masukuensis of Nyasaland is nearly all greenish.
Friedmann and Loveridge (loc. cit.) have even suggested that they
are not conspecific. However, birds from intermediate localities
bridge some of the differences and there seems no reason why they
should not be considered conspecific.

Pycnonotus masukuensis kakamegae Sharpe, 1900

Known from western Kenya and eastern Ituri Forest (Chapin,
1953, Bull. Amer. Mus. Nat. Hist., 75A: 118), and Moreau writes
me that in the British Museum there are specimens from Mabali
Mountains, South Kigoma, western Tanganyika Territory, taken at
6000 feet. A greenish bird with throat, side of face and eye ring pale
gray; crown and nape dark gray; gray ear coverts lightly streaked
with whitish. In many ways a smaller, duller replica of P. tephro-
laemus kikuyuensis.

We have one specimen from Kakamega Forest, hence topotyp-
ical: 9 wing, 82; culmen, 15 mm.

168 FIELDIANA: ZOOLOGY, VOLUME 35

Pycnonotus masukuensis roehli Reichenow, 1905

This race has generally been considered to be restricted to the
Usambara Mountains and the Uluguru Mountains (Friedmann and
Loveridge, 1937, Bull. Mus. Comp. Zool., 81: 235), though Lynes
(1934, Jour. f. Orn., 82: 75) has already recorded three Dabaga
(Iringa) birds as roehli after comparing them with four topotypes,
and Moreau (1947, Ibis, p. 230) extends the range to Songea, Ma-
tengo highlands. Bangs and Loveridge (1933, Bull. Mus. Comp.
Zool., 75: 187), on the other hand, had specimens from Iringi Dis-
trict (Kigogo, Uzungwe Mountains) and from Rungwe District
(Madehani, Ukinga Mountains and Nkuka Forest, Rungwe Moun-
tains) and referred them all to P. m. masukuensis.

I have six specimens from the Usambara Mountains, three from
the Uluguru Mountains, and eleven from the Iringa Highlands.
They differ from our single specimen of the preceding race in the
more greenish gray crown and cheeks and the less extensive gray
throat; in the clearer green, less olive green back, and the much
more grayish green, not olive green under parts. As such it is a
rather distinct race.

The populations are not identical, however, as the above exam-
ples of their earlier taxonomic treatment suggest. The Usambara
birds have the grayest heads, the darkest green backs, and the gray-
est green under parts. Uluguru birds are similar but are slightly
brighter green above and have the crown slightly paler, greener gray.
The Iringa birds are still brighter green above, with more of a green-
ish wash on breast and abdomen and with more greenish in the
grayish crown.

It could be argued that the Iringa birds are entitled to separate
subspecies rank, as an "island" population with average, though
overlapping differences. However, they definitely represent a tend-
ency toward the next race, P. m. masukuensis, but are not yet half
way to it, and are adequately separable from the other two races by
the above diagnosis. It seems inadvisable to separate them and
then bring up the question of separating the Uluguru birds from the
Usambara birds, etc.

There is no difference in size between the populations: Wing,
cf (10), 82-87 (av. 85.1); 9 (6), 80-83 (av. 81.6). Tail, <? (10), 79-
84 (av. 81.7); 9 (6), 77-84 (av. 79.8). Culmen from base, <? 15-17;
9 15-16.5 mm.

RAND: AFRICAN BULBULS 169

Pycnonotus masukuensis masukuensis Shelley, 1897

The race of the Masuku Mountains of Nyasaland (type locality)
and the mountains of the Rungwe District of Tanganyika Territory.

I have seen no topotypical birds and assume that they are similar
to Rungwe birds, of which we have six.

A well-defined race, lacking the gray on the head; the crown
green, almost concolorous with the back; the side of the head olive
green, and the whole under parts pale yellowish olive (not gray
washed with greenish), slightly darker on the breast and paler on
the throat.

Measurements: Wing, c? 84, 86; 9 81, 81, 81, 81. Tail, d" 78, 82;
9 76, 76, 78, 80. Culmen, c? 16, 16; 9 15.5-16 mm.

White (1944, Ibis, p. 553) has shown that there is no valid North-
ern Rhodesia record.

Pycnonotus montanus Reichenow, 1892

I have not seen this species. Bannerman (1936, Bds. Trop.
W. Afr., 4 : 184) has examined the type of this and of concolor Bates,
1926, and considers the latter a synonym. Only these two specimens
were known until Serle (1950, Ibis, p. 375) collected additional speci-
mens in the Cameroon Highlands. Chapin (1953, Bull. Amer. Mus.
Nat. Hist., 75A: 118) considers that it may be a close ally of P. ma-
sukuensis, after which I place it.

Pycnonotus virens Cassin

The most recent reviews, by Rand (1951, Fieldiana: Zool., 32:
609) and Chapin (1953, Bull. Amer. Mus. Nat. Hist., 75A: 109),
show several points of disagreement on the way in which this species
should be split.

The receipt of additional material in Chicago Natural History
Museum allows the clarification of some points.

The species can be easily divided into three rather distinct
sections:

1. Upper Guinea: erythropterus; dark back and grayish olive
under parts.

2. Lower Guinea to Uganda: virens + holochlorus; back paler
and greener; under parts yellower.

3. Nyasaland to Kilimanjaro: zombensis -f- marwitzi + shimba
(and probably zanzibaricus, which I have not seen); like (2) but
under parts paler.

170 FIELDIANA: ZOOLOGY, VOLUME 35

How much farther to go is in part a subjective thing. I recog-
nize P. v. holochlorus (contra Chapin) for reasons already given (1951,
loc. cit.) . The average larger size of Angola birds is not great enough
to use taxonomically, as Chapin says. Our series of Gabon speci-
mens (practically topotypical P. v. virens) measures: wing, cf (7)
76-81; 9 (10) 70-75; tail, d* 70-75, 9 62-70. Our Angola birds
measure: wing, d" 78-83; tail, & 70-77 mm.

Of the third group I have seen no zanzibaricus. I have recently
received specimens from Rhodesia and the Uluguru Mountains to
compare with our Usambara, Kilimanjaro and Ganda specimens. In
1951 I followed Grant and Mackworth-Praed in grouping all these
as zombensis. Chapin tentatively recognized marwitzi and shimba.

Measurements of this material are as follows:

Kenya: Ganda: Wing, & 81; 9 76. Tail, d* 78; 9 73.

Tanganyika: Kilimanjaro-Amani area: Wing, cf1 76, 80, 81, 81,
82, 84, 86; 9 80, 80. Tail, d* 74, 75, 78, 78, 79, 80; 9 74, 75. Ulu-
guru Mountains: cf wing, 85; tail, 79.

Northern Rhodesia: 9 wing, 86; tail, 77. Sex? wing, 85, 87, 87,
87.5, 88, 90; tail, 78, 79, 79, 79, 80, 80, 82.

This shows that southern birds average slightly larger than do
northern ones.

On color I am unable to see differences between the Kilimanjaro
(topotypical marwitzi), Amani, and Ganda birds, and the Ganda
birds (topotypical shimba, for which the original description gave
wing 76-81) are not smaller.

The more southern birds (Uluguru and Rhodesia), zombensis, type
locality, Nyasaland, are very similar to the above on the under parts.
On the upper parts the back and especially the crown are slightly
more clear greenish olive, less brownish olive, the tail is darker olive
brown, and the wing coverts and outer edges of the wing coverts
show the most difference, being greener, less brownish olive.

The southern birds are more variable than the northern in char-
acters. The Uluguru bird is as greenish as the greenest Rhodesia
bird and these stand out rather distinctly from the northern birds.

From this it can be seen that shimba is a synonym of marwitzi,
but that marwitzi is a lightly differentiated southern race compared
with zombensis.

Pycnonotus gracilis gracilis Cabanis

Since discussing this species (Rand, 1951, Fieldiana: Zool., 32:
613) I have received five specimens from Gabon and one from the

RAND: AFRICAN BULBULS 171

French Middle Congo. Compared with Cameroon specimens they
are just as dark but average slightly less brownish on the flanks and
less brownish olive above. The difference in the color of the upper
parts is probably largely due to foxing of the browner specimens,
but there seems a slight geographical variation in the Gabon birds,
which average greener. There is considerable overlap in this, and
the difference is not great enough to be of taxonomic importance.
This difference is not an approach to the paler eastern race ugandae.

Pycnonotus ansorgei Hartert

I have already commented on this species (1951, Fieldiana: Zool.,
32 : 617) but have since seen additional specimens from lower Guinea
and also topotypical specimens from southern Nigeria.

Chapin (1953, Bull. Amer. Mus. Nat. Hist., 75A: 107) has sum-
marized what we know about the habits of this bird that is so puz-
zlingly like P. gracilis in both appearance and habits.

Pycnonotus ansorgei ansorgei Hartert, 1907

The type locality is Degama, southern Nigeria. P. muniensis
Grote, 1924, type locality Akonangi, Spanish Guinea, I now consider
a synonym. Chapin (loc. cit.) called Congo birds muniensis but
points out that there is no difference in size and says that between
upper Congo birds and topotypical ansorgei "any difference in color
is hard to see."

White (1956, Bull. Brit. Orn. CI., 76: 157) has compared nine
Owerri birds with seven from Cameroons to upper Congo, and finds
the latter less rusty, colder and grayer — a tendency toward P. a.
kavirondensis van Someren but with not great enough difference to
recognize any subspecies.

In the material before me the Gabon and Cameroon specimens
are somewhat variable but agree among themselves and differ from
four Degama (topotypical) specimens of P. a. ansorgei collected in
1902 in having the upper parts, especially the crown, averaging
darker. On the under parts, Gabon and Cameroon birds average
more gray on the breast and have the flanks browner. The under
tail coverts vary. The Gabon bird has them darker brownish; one
of the Cameroon birds has them pale olive brown, while the other
specimens show various stages of intermediacy between these ex-
tremes. By contrast the four Nigeria birds all have the under tail
coverts rather pale buffy olive. There is further pronounced varia-

172 FIELDIANA: ZOOLOGY, VOLUME 35

tion in this series: the Gabon bird is slightly greener above on crown
and back; and one of the Cameroon birds, from Lolodorf, approaches
closely in color the specimens from Nigeria.

Size, mentioned in the original description of muniensis, does not
seem diagnostic (see measurements).

It seems impractical to recognize any subspecies in lower Guinea.
White (loc. cit.) also records this subspecies from Sierra Leone
(York Pass and Bintumane Peak).

Specimens examined. — Gabon, 1 (Mimongo-Poingi) . Cameron, 5
(Sangmelima, Lolodorf, Kum, Nkol Enyeng). Nigeria, 4 (Degama,
AMNH).

Measurements are as follows:

Gabon: sex? wing, 68; tail, 61; culmen, 15.5.

Cameroon: Wing, & 73, 75, 80; 9 73. Tail, c? 65, 66; 9 60.
Culmen, d" 16-17 mm.

Nigeria (Degama) : Wing, <? 72, 75; 9 68, 69. Tail, & 60, 62;
9 58, -. Culmen, cT 16, 16; 9 16, 16 mm.

Pycnonotus curvirostris curvirostris Cassin

The slight difference between eastern and western populations in
color in Pycnonotus c. curvirostris is considered not of taxonomic
value (Rand, 1951, Fieldiana: Zool., 32: 618), and Chapin (1953,
Bull. Amer. Mus. Nat. Hist., 75A: 105) does not consider the slightly
longer tails of Angola birds as worthy of nomenclatural treatment.
Our series from Cameroon measures: Wing, d" 78-83; 9 75. Tail,
cT 66-74; 9 72. The specimens from Angola measure : Wing, cf 80-
85. Tail, & 73-81 mm.

Pycnonotus importunus Vieillot

The grayish-greenish South African Pycnonotus importunus and
the yellowish green East African insularis have been kept as two
species by such authors as Sclater (1930, Syst. Av. Aethiop., p. 393),
Friedmann (1937, Bull. U. S. Nat. Mus., 153: 121), and Delacour
(1943, Zoologica, 28: 23); but Roberts (1922, Ann. Transvaal Mus.,
8 : 226) had already pointed out that the forms are geographical rep-
resentatives, probably conspecific, a treatment followed by him in
his later works and adopted by Chapin (1953, Bull. Amer. Mus.
Nat. Hist., 75A: 108). Certainly the structure of the two groups
is the same, and the race oleaginus is the connecting form; importu-
nus Vieillot, 1815, must be used for the specific name.

RAND: AFRICAN BULBULS 178

The major trends in variation are: A greenish bird in the south
intergrades with the yellowish-green bird of Portuguese East Africa;
northward the yellow decreases in intensity, apparently gradually,
with consequent difficulty of drawing subspecific lines and diversity
of opinion as to the number of races to recognize. The buffy or
ochraceous under wing coverts (vs. clear yellow) give another char-
acter for northern coastal birds, and the yellow eye ring in the adult
for Kenya highland populations.

Pycnonotus importunus fricki Mearns, 1914

This is the race of the Kenya highlands; from Endoto Mountains
(topotypical fricki) and from the south central area (Kitunga, 7000
feet, Thika, junction Thika and Tana Rivers [Friedmann]).

It is a yellowish-olive-green form with buffy-yellow under wing
coverts and is distinguished from all the following races by having
a yellow eye-ring. It may average larger also.

In reviewing this group Friedmann (1937, Bull. U. S. Nat. Mus.,
153: 121, 122) was unable to confirm the various color differences
mentioned in the original description for kitungensis Mearns but
kept it separate on the basis of its larger size (fricki, wing <? 87 mm. ;
kitungensis, <? 90, 91, a not very convincing difference). Sclater
(1930, Syst. Av. Aethiop., p. 393) has pointed out that the two
are probably not separable and in 1938 Jackson and Sclater (Bds.
Kenya . . ., p. 879) remarked that until more specimens are collected,
further discussion is impossible.

I have seen no specimens of this race. A yellow eye-ring appears
to be a sign of immaturity in P. i. subalaris and P. i. hypoxanthus,
but the four known specimens of this race have been recorded
as adult by Dr. Friedmann. Friedmann (1937, op. cit., p. 122) has
commented that a specimen from the Tana River, at the mouth of
the Thika River, has the eye ring only faintly indicated.

Pycnonotus importunus somaliensis Reichenow, 1904

A pale form, with under wing coverts buffy yellow; no yellow
eye-ring. Measurements: Wing, & (10) 82-91 (av. 88.4); tail, 77-
82; culmen, 18.5-20. Wing, 9 80; tail, 72; culmen, 17.

This is the bird from the northeast corner of the species' range,
in the Juba River area (mid-Juba River and Barawa, on the coast
250 kilometers north of the mouth of the Juba).

We have a series (16 d\ 1 9) from Serenli and Gebier on the
Juba River.

174 FIELDIANA: ZOOLOGY, VOLUME 35

Pycnonotus importunus subalaris Reichenow, 1903

Differs from somaliensis in being slightly darker, more olive on the
breast and slightly yellower on the abdomen and under tail coverts.

The range is from Manda and Lamu Islands south to Vanga
along coastal Kenya, and inland to the mid-Tana River (Garissa),
Taveta, in Kenya, and Mount Kilimanjaro, Tanganyika Territory.

Specimens examined. — Manda Island, 2; lower Tana River (An-
asa, Kiaponi, Kipini), 6; Malinda, 2; Sokoke-Mongeya area, 26;
Changamwe, 3; Mombasa, 3; Rabai, 4; Tanganyika border, 500 feet
altitude, 1; Samburu, 1; Taveta, 1; Mount Kilimanjaro, 1; mid-
Tana (Garissa), 1.

A single specimen from Makueni, 80 miles south of Machakos in
central Ukamba (altitude 3000 feet), is large (wing 94), suggesting
increase in size inland. Pycnonotus i. fricki is also large.

Measurements of Sokoke-Mongeya birds: Wing, d" (10) 81-89
(av. 85.5); 9 (10) 77-88 (av. 83.8). Tail, d1 74-81 (av. 77.3); 9 71-
81 (av. 75.4). Culmen, d" 18-20; 9 17.5-19 mm.

The only immature, a male from Malinda, has a ring of yellow
feathers around the eye, apparently a character of immaturity in
some races of this species.

Though in series these specimens separate rather well from so-
maliensis on the characters given, there is considerable individual
variation giving overlap with somaliensis, so that some specimens
are of doubtful allocation on characters; for example, a few of the
individuals from Sokoke-Mongeya area (nearly topotypical suba-
laris), and the Samburu specimen and the most northern birds,
from Manda Island, show an approach to somaliensis. However,
the birds from such peripheral localities as the lower Tana, Garissa,
Taveta, and Tanganyika-Kenya border show the characters of suba-
laris more plainly. The single rather richly colored specimen from
Mount Kilimanjaro ( 9 wing 82 mm., which should be kilimandja-
ricus Sjostedt, 1908; type locality Kibonoto, Kilimanjaro, 1300 me-
ters) can be matched in color and size with specimens from the
Sokoke-Mongeya area, so I consider kilimandjaricus a synonym of
subalaris.

Sclater (1930, Syst. Av. Aethiop., p. 393) accepted kilimandja-
ricus as valid, and referred to Gyldenstolpe (1926, Ark. Zool., 19A,
no. 1, p. 58), who, however, through lack of material had not been
able to form an opinion of its racial validity beyond saying it was
very similar to subalaris.

RAND: AFRICAN BULBULS 175

Pycnonotus importunus insularis Hartlaub, 1861

Differs from subalaris in having the under wing yellow, not buffy
or ochraceous yellow.

The range is Zanzibar and the opposite coast from Dar-es-Salaam
north to Pangani River (Friedmann).

We have two Zanzibar specimens ( 9, wing, 83, 88; tail, 75, 79
mm.). In body plumage one is somewhat darker and more richly
colored than subalaris, the other in general color can be matched
rather well by a number of subalaris specimens from Sokoke, but the
difference in the under wing color is evident in both.

Vincent (1935, Ibis, p. 373) extends the range of insularis south
through Mozambique but our two specimens from Lumbo are con-
siderably more yellowish below than are Zanzibar examples and I
do not include them with insularis.

Pycnonotus importunus hypoxanthus Sharpe, 1876

Differs from insularis in being still more yellow below. Measure-
ments: wing, cf 92, 9 82; tail, d1 86, 9 77; culmen, <? 20, 9 17.

Its range is from southern Tanganyika Territory, through Portu-
guese East Africa to the Zambesi, and inland to Tete, southern
Nyasaland, and Northern Rhodesia.

Vincent has shown that the Kilosa, Tanganyika Territory, record
is an error (1935, Ibis, p. 374). Though he extends the range of
insularis south through Mozambique our two specimens from Lumbo
are certainly much yellower below than two topotypical (Zanzibar)
insularis and Moreau records that Mafia Island and Mikindani,
southern Tanganyika, birds are also yellower below than insularis,
and I refer them to this form.

Specimens examined. — Lumbo, Portuguese East Africa, 4 (2
CNHM, 2 Coryndon Mem. Mus.).

One of the four specimens before me has the rounded tip of the
first primary and the rather pointed tail feathers characteristic of
immaturity in this group. It also has a very narrow ring of yellow
feathers encircling the eye. The only other immature in the series
of more than 70 specimens of this species that I have seen (a male of
subalaris from Malindi, Kenya) also has a similar yellow eye ring.
Presumably in most races it is an immature character which appears
in the adult in fricki.

176 FIELDIANA: ZOOLOGY, VOLUME 35

Pycnonotus importunus oleaginus Peters, 1868

This race is the connecting link between the yellowish forms of
the insularis group to the north, and the green and grayish impor-
tunus to the south. It differs from hypoxanthus in being greener
above and in being considerably duller and less yellowish below
(d1 wing, 86, 90; tail, 89; culmen, 18, 18; tarsus, 22 mm.).

The range is northeast Zululand and Portuguese East Africa
south of the Zambesi mouth (Clancey).

Grant and Mackworth-Praed (1948, Bull. Brit. Orn. CI., 68: 59)
found that the bird marked as the type in the Berlin Museum is not
the type, so that the allocation of the name depends on the descrip-
tion alone (Peters, 1868, Jour, f . Orn., 16 : 133) ; type locality Lorenco
Marques. It reads (my translation from the Latin) : an Andropadus,
olive green, below paler, obsolete undulations, abdomen yellow green,
inner edge of bend of wing, tibia, under wing citrine; rectrices 1 and 2
margined internally with a yellow fringe; iris yellow, bill and feet
black; bill 19.5, wing 84, tail 87, tarsus 21 mm.

Grant and Mackworth-Praed (loc. cit.) say this description of the
under parts does not apply to any bulbul adult or young but prob-
ably to some young shrike. They propose dropping this name as in-
determinate. South African ornithologists have followed this
course.

On the contrary, on examining the evidence, I find the agreement
between the description and the bird for which the name has been
used is very close, both in color, including that of the eye, and in
measurements. The only discrepancy is the "obsolete undulations."
Now "obsolete undulatus" can be variously interpreted but obviously
the pattern is not very distinct and I submit that it could have been
caused by a disarrangement of the feathers. In view of the many
points of close agreement, and only one small, understandable dis-
crepancy, I consider the old, widely used oleaginus a valid, identi-
fiable name from the description, and Andropadus importunus mentor
(Clancey, 1952, Ann. Natal Mus., 12: 251; type locality Shimula's
Pont, Pongola River, N. E. Zululand), which was proposed to re-
place it, a straight synonym.

We have no specimens in Chicago but Mr. Clancey has lent me
six specimens from Kosi Bay, North East Zululand, Newington,
Eastern Transvaal; Lebombo Mountains and Big Bend, Swaziland,
and Manhica, Mozambique.

RAND: AFRICAN BULBULS 177

Pycnonotus importunus noomei Roberts, 1917

Differs from the preceding in the great reduction of yellow on the
under parts, which are grayish olive, and the darker upper parts.

Clancey (1952, Ann. Natal Mus., 12: 251) questioned the valid-
ity of noomei, compared with iynportunus, but more recently he
has secured additional material and this caused him to recognize two
races in South Africa, but with different ranges from those given by
Roberts, and after examining his material, which he kindly lent me,
I agree.

Six specimens from Pondoland (Mntafufu River, Mzamba, Fos-
sil Head) and Natal (Durban, Pietermaritzburg) belong to noomei,
which ranges from southern Zululand inland to Transvaal and south
to east Griqualand (eastern Cape).

Pycnonotus importunus importunus Vieillot, 1818

Like noomei but paler and more grayish olive below. Six speci-
mens seen from Uitenhage, Albany, and Fish River Mouth, all east-
ern Cape.

This form seems restricted to southern and eastern Cape Province.

Pycnonotus latirostris Strickland

The two West African races have long been recognized as dis-
tinct, but the literature dealing with the East African populations
contains not only diversified views as to races to be recognized, but
as to their characters. The most recent reviews are by Sclater
(1930, Syst. Av. Aethiop., p. 394), Friedmann (1937, Bull. U. S.
Nat. Mus., 153: 122), and Chapin (1953, Bull. Amer. Mus. Nat.
Hist., 75A: 113).

Our material indicates that the four races, from Upper Guinea,
Lower Guinea, Uganda, and the Kenya highlands are all fairly dis-
tinct. Confusion between the last two may have arisen through the
presence of a considerable zone of overlap.

In each area from which we have large series there are a few birds
that fall considerably outside of the rest of the individual variation
and resemble some other race, but I do not consider them either as
wanderers of another race, or as indicating that two species exist.

Pycnonotus latirostris congener Reichenow, 1897

This Upper Guinea race is quite distinct on the basis of its darker

back, and especially its darker, less reddish tail. It is a small race

(wing, d" 81, sex? 83; tail, d* 72, sex? 77 mm.).

Specimens examined. — CNHM, Liberia, 1; Sierra Leone, 1.

178 FIELDIANA: ZOOLOGY, VOLUME 35

Pycnonotus latirostris latirostris Strickland, 1844

This race has the back lighter and the tail lighter, more reddish
brown than congener. Its size is the same (o* wing, 80-90; tail,
68-80).

Our specimens of this race are all from western Africa. The
Gabon birds are slightly darker above than Cameroon birds but not
different enough to separate, and northern Angola birds (Heinrich
coll.) are very similar. The average larger size of Angola birds is
slight (see measurements).

The name efulenensis Sharpe (1904, Ibis, p. 636), type locality
Efulen, sounds in part as if intended for a latirostris-congener inter-
grade with darker brown tail than the former, but it is said to be
based on immatures. However, our Cameroon birds (including one
Efulen bird) and one Nigerian (Ondo Province) bird all seem very
similar.

Specimens examined. — CNHM, Nigeria, Ondo Province, 1; Came-
roon, 9; Gabon, 6; northern Angola (near Camabatela), 7.

Pycnonotus latirostris eugenius Reichenow, 1892

This is a moderately distinct race, differing from latirostris in the
richer, more yellowish olive under parts, the brighter yellow sub-
malar streaks with less dusky between their anterior ends, and the
slightly greener upper parts, especially the crown (which is less gray-
ish olive). The postulated larger size of this race is slight and most
evident in the longer tail (Uganda: cf wing, 80-90; tail, 77-83 mm.
West Kenya: d" wing, 82-93; tail, 81-90. For details see below).

Our Ruwenzori specimens agree fairly well with other Uganda
specimens, though Gyldenstolpe (1924, K. Svenska. Vetensk. Akad.
Handl., (3), 1, no. 3, p. 185) refers birds from the west side of the
Semliki Valley to latirostris.

The West Kenya birds are not of certain allocation to this race
or the next, being intermediate in size and in color. They are arbi-
trarily referred here.

Specimens examined. — CNHM, Uganda, Ruwenzori, 7; Toro, 7;
Ankole, 3; Kigezi, 1; Mabira, 2; Bugoma, 2. Kenya, Mount Elgon
area, 8; north Kavirondo, 8; Kericho, 8.

Pycnonotus latirostris saturatus Mearns, 1914

Distinguished from eugenius by the greener upper parts, espe-
cially the crown, the more greenish (less warm olive-tinged) under

RAND: AFRICAN BULBULS 179

parts, the yellow sub-malar stripes, which are slightly more conflu-
ent at the anterior end (i.e., more yellow in the chin), and the average
larger size, especially the longer tail. The postulated paler rectrices
are not evident in this material (d" wing, 89-96; tail, 83-92; for de-
tails see below).

This is the form of the Kenya highlands. As mentioned above,
west Kenya birds are intermediate on measurements and could
equally well be referred to this race, and the same is true of color.
Thus, the Rift Valley is not actually a sharp dividing line between
the two races but an area of intergradation over a considerable area.

About half of the Kenya highlands birds are clearly greener than
the Uganda birds, with about half showing little difference. How-
ever, color plus size, especially tail length, make this a fairly well-
defined race. Greenish tones such as these are often affected by
foxing, but it seems not to have been the case in this instance.
Specimens taken in 1906 are as green as those taken in 1949.

I have seen no specimen of the race australis Moreau (1941, Bull.
Brit. Orn. CI., 62: 29), Mbisi, western Tanganyika Territory, but
Moreau writes me that he does not now recognize it. This gives an
interrupted range for saturatus.

Pycnonotus latirostris pallidus Mearns is considered unseparable
from saturatus. It is said to differ from all the other races in being
paler, especially on the middle of the abdomen (Friedmann).

It appears to be known only from the original four specimens
from Mount Gargues, an isolated mountain mass north of Mount
Kenya on the extreme northeast edge of the range.

Dr. Friedmann has lent me three of the four specimens. These
are fairly uniform in appearance and have the large size and the
greenish-tinged plumage that characterize saturatus.

While pallidus is paler than average saturatus it falls within the
average range of the latter, and the complete overlap makes it seem
necessary to consider pallidus a synonym of saturatus.

Specimens examined. — Mount Kenya area, 11; Nairobi area, 8;
Mount Gargues, 3 (USNM).

Measurements of Pycnonotus latirostris are as follows:

P. I. congener. — Liberia, Sierra Leone: Wing, d" 81; sex? 83. Tail,
d1 72; sex? 77. Culmen, c? 16.5; sex? 16.

P. I. latirostris. — Nigeria, Cameroon and Gabon: Wing, d* (7)
80-90 (av. 85.8); 9 77, 82, 82, 83. Tail, tf (7) 68-80 (av. 76.4);
9 69, 75, 76. Culmen, & 17-19; 9 1&-17. Angola: Wing, d" 86,

180 FIELDIANA: ZOOLOGY, VOLUME 35

90, 90, 91, 92; 9 77, 79. Tail, & 83, 85; 9 77, 79. Culmen, d1 17.5-
19; 9 18.

P. I. eugenius.— Ruwenzori: Wing, <? 85, 88, 89, 89, 89; 9 78, 81.
Tail, c? 78, 82, 83, 84, 84; 9 74, 77. Culmen, cf 16.5, 17, 18, 18;
9 16.5. Uganda, other localities: Wing, d* (8) 80-90 (av. 85); 9 78,
80, 82, 86, 86. Tail, & (7) 77-83 (av. 79.4); 9 73, 74, 77, 78, 80.
Culmen, <? 17-19; 9 15-17. Kenya, Mount Elgon area: Wing,
d" 89, 91, 93; 9 84. Tail, d1 83, 85, 90; 9 81. Culmen, & 17, 17.5,
18. Northern Kavirondo: Wing, d" 82, 86, 88; 9 82, 88, 89. Tail,
d"81, 83; 9 76,78,82,83. Culmen, d* 17-17.5; 9 16-17. Kericho:
Wing, d" 89, 89, 89, 93; 9 82, 83, 86. Tail, c? 81, 84, 88; 9 77, 80,
80. Culmen, d* 16, 17, 17; 9 16, 16.5, 17.

P. I. saturatus— Mount Kenya area: Wing, c? 90, 91, 92, 93, 93,
94, 96; 9 85, 90. Tail, d" 86, 87, 87, 88, 90, 90, 91; 9 80, 88. Cul-
men, d" 17-18; 9 18. Nairobi area: Wing, d* 89, 89, 90, 92, 93, 95;

9 85. Tail, d" 83, 83, 85, 87, 90, 92; 9 83. Culmen, cf 16.5-18;

9 16. Mount Gargues: Wing, 9 [= cf] 92; 9 85,85. Tail, 9 [= cf]
91; 9 80> 83. Culmen, 9 [= <?] 17; 9 16, 16.

Pycnonotus gracilirostris

This is a forest species ranging from Senegal to Kenya and An-
gola. There is one undisputable race, percivali, of the Kenya high-
lands. For the rest, there is variation from locality to locality but
the difference between populations is such as to make delimitation
of races difficult. Bannerman's (1936, Bds. Trop. W. Afr., 4: 188)
and Chapin's (1953, Bull. Amer. Mus. Nat. Hist., 75A: 116) differ-
ent treatments illustrate this.

The following is perhaps the best treatment:

Pycnonotus gracilirostris gracilirostris Strickland, 1844

This is the Upper Guinea race. Bannerman gives its range as
east to Uganda, but Chapin says it is restricted to Upper Guinea,
reaching its eastern limits in southern Nigeria and Fernando Po.

It is said to be characterized by being paler than the next form,
lighter green above and lighter gray below. Judging from our single
specimen from Liberia it is only lightly characterized.

Pycnonotus gracilirostris congensis Reichenow, 1916

This is the lower Guinea form, from Cameroon to northern An-
gola and east through the Belgian Congo forests.

RAND: AFRICAN BULBULS 181

It is darker than the preceding. Our specimens show that Angola
birds are slightly darker above, especially with the tail more blackish
than Cameroon birds, but they are best grouped together. In both
Cameroon and Gabon a few birds are more olive below than the
others, but individual variation is not apparent in the Angola birds.
Angola birds have the tail slightly longer than do other specimens.

Measurements. — Cameroon : Wing, o" 83, 85, 85, 86, 88. Tail, d"
73, 75, 75, 76, 77. Gabon : Wing, d< 81, 83, 84, 86, 86, 87. Tail, & 71,
72, 75, 76, 78, 79. North Angola: Wing, d" (10) 82-88 (av. 85.6).
Tail, c? (8) 76-82 (av. 79).

Pycnonotus gracilirostris chagwensis van Someren, 1915

The Uganda race, intergrading with the next in western Kenya
and presumably with the preceding in the eastern Belgian Congo.
Presumably, southern Sudan birds and Kungwe-Mahare highland
birds belong here.

On the under parts this race differs little from congensis, but
above it is lighter olive, and the tail especially is lighter and browner.

Compared with Angola birds, Uganda birds are rather conspic-
uously different. Part of the difference in the back color may be due
to foxing of the older series, but that is hardly likely to be true of
the tail color. Compared with Cameroon birds the difference is less,
but still evident.

The characters of this race are not simply those of intergradation
between congensis and the greener percivali, and it seems advisable
to recognize chagwensis.

Measurements.— Uganda: Wing, & (10) 82-89 (av. 85.3); 9 (9)
81-85 (av. 82.7). Tail, <? (9) 75-82 (av. 77.7); 9 (6) 76-83 (av.
79.1). Culmen, c? 18-20; 9 18-20 mm.

Pycnonotus gracilirostris percivali Neumann, 1903

This race of central Kenya is fairly distinct on the basis of its

greener upper parts, crown, back, and tail, and its paler gray under

parts.

Measurements. — Mount Kenya area (Chuka: Meru) : Wing, cf 89,

89; 9 89. Tail, & 81, 83; 9 81. Culmen, 19-20. Nairobi area

(Kiambu and Karura Forest): Wing, d" 85, 87; 9 85. Tail, cf 78,

79; 9 78. Culmen, 19.

Pycnonotus tephrolaemus

The birds included here are all mountain forest birds with many
isolated populations. They have often been grouped into three spe-

182 FIELDIANA: ZOOLOGY, VOLUME 35

cies: tephrolaemus for the yellow-bellied forms (3 subspecies) from
Cameroon and from eastern Belgian Congo to the Kenya highlands;
nigriceps for the gray-breasted forms from the peripheral Tangan-
yika-Northern Rhodesia highlands localities (5 subspecies); and
chlorigulus for the yellow-throated form from the central Tanganyika
(Iringa) highlands. Though these groups do not show intergrada-
tion, their distribution, habitats, and habits are all similar, and
Moreau (1943, Ibis, p. 391) has already suggested they may be con-
specific, which course I follow.

The general similarity of all these birds, plus distribution pattern
and habit similarities, justify the grouping of forms with both
greater and lesser striking, non-intergrading color characters.

Pycnonotus tephrolaemus tephrolaemus Gray, 1862

From Mount Cameroon, 5800-6500 feet altitude, we have a series
of eight topotypes. Measurements: Wing, & 87, 88, 89, 89, 90; 9 83,
83, 84. Tail, c? 76, 78, 79, 80, 80; 9 73, 74, 77. Culmen, cf 20-21;
9 19-20 mm.

Amadon (1953, Bull. Amer. Mus. Nat. Hist., 100: 423) reports
that Fernando Po birds are the same.

Pycnonotus tephrolaemus bamendae Bannerman, 1923

This form, of the Cameroon highlands, is described as darker
green below; gray of head darker and continued farther onto chest,
and tail longer. Measurements: Wing, o" 87-94; 9 85-90. Tail, cf
85-87; 9 78-83 mm. (Bannerman, 1936, Bds. Trop. W. Afr., 4:
183). I have seen no specimens.

Pycnonotus tephrolaemus kikuyuensis Sharpe, 1891

The form of the mountains of the eastern Belgian Congo, Uganda,
and the Kenya highlands.

It differs from the Cameroon birds chiefly in having the gray ear
coverts distinctly streaked with white, and the white eye ring wider
and more conspicuous.

Specimens examined from Uganda (Ruwenzori, 15; Ankole, 4)
and Kenya (Elgon area, 3; Mau Plateau, 22; Eldoma Ravine, 1;
Laikipia, 1; Kinangop, 2; Navasha, 2; Elgeyu, 1; and Mount Kenya,
7) are very similar in both color and measurements. Wing, cT (10)
83-94 (av. 89.8); 9 (10) 83-87 (av. 85.2). Tail, & (10) 78-89 (av.
84.1); 9 (10) 78-85 (av. 81.9). Culmen, d" 16.5-19; 9 16.5-19.

RAND: AFRICAN BULBULS 183

Pycnonotus tephrolaemus nigriceps Shelley, 1890

The form of the mountain forests of central northern Tanganyika
(Mount Kilimanjaro, type locality, and Ngorongoro-Oldean-Hanang) .

This form is very different in color from the preceding, with back
darker green, crown blackish (not gray), eye ring gray (not white),
breast and abdomen gray tinged yellowish-green on flanks and center
of abdomen (not entirely greenish yellow).

Ngorongoro birds (8), compared with Kilimanjaro (3) and Mount
Meru (1) birds, differ slightly in having the greenish wash of flanks
and center of abdomen slightly duller and less extensive, but the
difference is slight and not worth nomenclatural designation. Meas-
urements: Wing, cT 88-92 (90.7); 9 85, 86, 87, 88. Tail, <? (8) 79-
85 (av. 82); 9 75, 79, 80, 83. Culmen, d" 18-20; 9 18-19.5 mm.

Pycnonotus tephrolaemus usambarae Grote, 1919

This form of South Pare (Moreau, 1940, Ibis, p. 457) and the
Usambara Mountains is a quite distinct race, similar to nigriceps
except for having the crown and hind neck gray bordered laterally
with black (not generally blackish), and in averaging more green in
the flanks; slightly more greenish wash on the abdomen; and average
smaller bill. Six specimens from Usambara were examined. Meas-
urements: Wing, c? 86, 88, 90, 91; 9 85, 86. Tail, & 82, 87, 88;
9 78, 83. Culmen, cf 16-18; 9 15.5, 16.5 mm.

Pycnonotus nigriceps percivali (Hartert, 1922, Bull. Brit. Orn. CI.,
42 : 50) from Usambara Mountains is a synonym, as has been pointed
out by Sclater and Moreau (1932, Ibis, p. 680).

Pycnonotus tephrolaemus neumanni Hartert, 1922

I have not seen this Uluguru Mountains form, which was de-
scribed as differing from P. t. nigriceps (and thus also from usam-
barae) in having no gray eye-ring, the ear coverts solid gray without
stripes, and the throat and jugulum darker gray (<? wing, 93; tail,
89 mm.).

Pycnonotus tephrolaemus kungwensis Moreau, 1941

This Kungwe-Mahara highland form I have not seen. It was
described as most like fusciceps but with the sides of face, lores, and
throat pale gray and the white around eye ill-defined (wing, 94;
tail, 85; culmen from base, 17).

184 FIELDIANA: ZOOLOGY, VOLUME 35

Pycnonotus tephrolaemus fusciceps Shelley, 1893

This, the form from the mountains north of Lake Nyasa (in Tan-
ganyika Territory), Nyasaland, Northern Rhodesia and northwest-
ern Portuguese East Africa, is one of the gray-breasted races, with a
gray crown; the white eye ring is incomplete, represented by distinct
white upper and lower segments, and the ear coverts are solid gray,
unstreaked.

Specimens from Northern Rhodesia (11) and southwestern Tan-
ganyika Territory (Poroto Mountains, Mount Rungwe) (10) are very
similar in color. Measurements: Wing, d" (7) 95-98 (av. 97.2);
9 (10) 87-94 (av. 90.4). Tail, & (7) 87-95 (av. 92); 9 (10) 83-90
(av. 87.5). Culmen, <? (7) 16.5-20 (av. 18.8); 9 16-19 mm.

Pycnonotus tephrolaemus chlorigulus Reichenow, 1899

We have a series of five of this form of the Iringa highlands. It
also occurs on Nguru Mountain (schusteri Reichenow, 1913, being a
synonym) and Kiboriani north of Mpwapwa.

In details of plumage it is very similar to fusciceps, with a gray
crown, darker over eye, and incomplete but conspicuous white eye
ring, but it is very distinct from that form in the brighter green upper
parts, the yellowish olive area on lower throat and upper breast, and
the slightly more yellowish green flanks. In size it is similar but
with an average slightly smaller bill. Measurements: Wing, d1 93,
97, 98; 9 89, 91. Tail, <? 90, 92, 93; 9 87, 89. Culmen, <? 18, 18,
18.5; 9 17, 18.

The yellowish-green lower throat and upper chest are unique in
the species.

Pycnonotus milanjensis

This is an East African species of mountain forests occupying an
interrupted range from Kenya to southern Rhodesia, where suitable
habitat occurs.

The extreme southern birds (Mlanje, Nyasaland, and Southern
Rhodesia) have a gray crown, and the birds from Cholo, Nyasaland,
north have a green crown and a generally yellower or more golden
plumage. Within this green-headed group, the southern birds are
duller and greener, and the northern birds are brighter, yellower
golden, with intermediate-colored populations at intermediate lo-
calities.

It seems to be advisable to recognize three subspecies as follows :

RAND: AFRICAN BULBULS 185

Pycnonotus milanjensis milanjensis Shelley, 1894

Type locality, Milanji (=Mlanje), Nyasaland.

Crown gray; body plumage greenish.

Range: Mlanje, Nyasaland, eastern Mashonaland, Southern
Rhodesia and in Portuguese East Africa.

Specimens examined. — Mlanje (2400 feet), Nyasaland (1), and
Vumba, Southern Rhodesia (2).

Pycnonotus milanjensis olivaceiceps Shelley, 1896

Type locality, Mount Chiradzulu, Nyasaland.

Though this race has been generally rejected, it seems perfectly
valid; like milanjensis but crown olive green (not gray), less gray in
throat, plumage generally more yellowish olive.

Range: From Cholo, Nyasaland, to Rungwe Highlands, south-
west Tanganyika Territory.

Specimens examined. — Nyasaland, 2 (Masuku Mountains and
Mzimba District) ; Tanganyika Territory, 2 (Rungwe) .

Pycnonotus milanjensis striifacies Reichenow and Neumann,
1895

Type locality, Marangu, Kilimanjaro.

Like olivaceiceps but plumage generally more yellowish- or golden-
tinged, above and below.

Range: Southern Kenya through eastern Tanganyika Territory
to Iringa Highlands.

Compared with olivaceiceps from Nyasaland and Rungwe the
birds from the northern part of the range are as different as most
moderately distinct, widely accepted subspecies. A bird from the
Iringa Highlands (Kigogo, Uzungwe Mountains) is intermediate but
best placed here.

Within this subspecies the samples I have of the various popula-
tions from Kenya to Uluguru Mountains show slight differences:
birds from Kilimanjaro average slightly more golden; from Mount
Meru they are slightly darker; from Chyulu Hills they are slightly
paler and duller; from Tavete they are slightly clearer yellow; from
Usambara they are very similar to Chyulu birds; from Uluguru they
are very similar to Kilimanjaro birds.

However, these differences are only average and slight, and are
complicated by individual variation and variation apparently due to

186 FIELDIANA: ZOOLOGY, VOLUME 35

museum age of skins. It seems inadvisable to try to separate
any of these populations into very finely split subspecies. Hence
I consider P. m. chyulu van Someren (1939, Jour. E. Afr. Ug. Nat.
Hist. Soc, 14: 64), type locality Chyulu Hills, as a synonym, though
there is a little something to the variation it was intended to point out.

Specimens examined. — Kenya: Teita, 6; Chyulu Hills, 6. Tan-
ganyika Territory: Mount Meru, 2; Mount Kilimanjaro (Moshi), 12;
Usambara Mountains, 4; Uluguru Mountains, 7; Uzungwe Moun-
tains, 2.

Baeopogon indicator Verreaux

I have seen no specimens from Upper Guinea so I accept the two
races there, leucurus Cassin, Liberia, and togoensis Reichenow (with
ussheri a synonym), Togoland, as outlined by Bannerman (1936,
Bds. Trop. West Afr., 4 : 168) . Of the Lower Guinea to Kenya popu-
lations I have: Cameroon, 25; Gabon, 2; Angola, 2; Uganda, 6;
Kenya, 1. These birds are customarily divided into two subspecies:
indicator Verreaux, type locality Gabon, ranging from northern
Angola to Cameroon to the Kasai and the Upper Congo near Stanley-
ville; and chlorosaturata van Someren, type locality Kyetume Forest,
Uganda, ranging from the eastern Congo forests to Sudan and
Kenya, characterized by being like indicator but more greenish olive
above, very much darker gray below, each feather with lateral olive
green edges, and abdomen darker ochraceous (original description).
Chapin has already pointed out (1953, Bull. Amer. Mus. Nat. Hist.,
75A: 123) that the differences between the Cameroon and Uganda
birds is really slight.

Our four adult Uganda birds, compared with two Gabon birds,
average slightly lighter, brighter, greener above, and paler gray (not
darker) below, but the differences are slight, and both Uganda and
Gabon birds fall within the range of variation of our series of southern
Cameroon birds in all these characters. There is no difference in size
(wing, d> Uganda, 101-106; Cameroon, 99-108; Gabon, 104, 106).
It seems inadvisable to recognize chlorosaturata on the slight aver-
age difference. Two Angola females (wing, 101, 106 mm. ; tail, 79, 82)
are apparently in first year plumage with rather narrow pointed rec-
trices. Compared with Cameroon birds, they are slightly clearer
gray, with less greenish wash or bufly tinge below, and colder, less
olive green above. Part at least of this difference is probably due to
the freshness of the Angola birds (collected in 1954).

RAND: AFRICAN BULBULS 187

Chlorocichla falkensteini Reichenow

The Cameroon and the Angola populations of this west African
species are generally kept separate as two subspecies. Chapin (1953,
Bull. Amer. Mus. Nat. Hist., 75 A: 138), while recognizing the two
subspecies and characterizing the Cameroon bird, viridescentior
Sharpe, as having deeper green upper parts and more gray on the
chest and flanks than C. f. falkensteini Reichenow of the Loango
Coast, etc., says that the differences in color are almost negligible.

We have seven adult Angola birds, collected in 1908 (Ansorge),
1952 (Beatty), and 1954 (Heinrich), and nine Cameroon birds col-
lected in 1924-27 (Bates) and 1941-52 (Good). The older birds
in both series show the characters attributed to falkensteini and the
recently collected birds those of viridescentior. Between birds of
comparable age of skin from the two localities I can see no color
differences. Evidently this species changes considerably through
museum age (foxing).

The following measurements show no differences except a very
slight average smaller bill size of Cameroon birds:

Angola: d* Wing, 88, 93, 93, 93, 94, 94. Tail, 78, 79, 81, 81,
82, 85. Culmen, 19 (5), 20 (1). 9 Wing, 87. Tail, 78. Culmen, 19.

Cameroon: 9 Wing, 92, 92, 92. Tail, 83, 83, 84, 85. Culmen,
18, 19, 19. 9 Wing, 85, 87, 87, 88, 92. Tail, 74, 75, 80, 82. Cul-
men, 17 (2), 18 (1), 19 (1).

It is impossible to recognize any races in this species.

The original description is sometimes quoted from a source that
is incorrect (Correspondenzbl. Afr. Gesellsch., no. 10, p. 179), for
Grant and Mackworth-Praed (1955, Bull. Brit. Orn. CI., 75: 24)
show that this was published in December, 1874, while another de-
scription (Jour. f. Orn., 22: 458) was published in October, 1874,
and therefore takes precedence.

Chlorocichla simplex Hartlaub

When I reported on a series of Liberia birds (1951, Fieldiana:
Zool., 32: 619) I had but two Cameroon birds. Now with an addi-
tional 30 birds from Cameroon and three from Angola certain other
points emerge.

There is considerable variation in color in the series of Cameroon
birds. About half the series has the under parts generally distinctly
washed with yellowish, the flanks, and especially the under tail
coverts buffy, and the upper parts slightly brownish-olive; the other

188 FIELDIANA: ZOOLOGY, VOLUME 35

half tends to have the under parts less yellow-tinged, flanks less
buffy, under tail coverts paler, and upper parts more dusky. These
last include all the worn birds and I assume that most if not all of
this difference is due to wear and fading in life, though the effects of
wear on feather structure is pronounced on only a few specimens.

The birds in first year plumage are very similar to the adult ex-
cept for the more pointed tail feathers, the more olive-brown outer
margins to the remiges, a few brownish-margined lesser upper wing
coverts, and the retention of some fuzzy under tail coverts.

Birds from Liberia (12) and Angola (3) agree well with the corre-
sponding stage of plumage of Cameroon birds.

There is no difference in size of adults:

Liberia: Wing, tf1 104-109; 9 95-100. Tail, <? 91-95; 9 85-91.
Culmen, tf1 22-23; 9 20-22.

Cameroon: Wing, <? 97-110; 9 95-100. Tail, d1 89-99; 9 86-
90. Culmen, d" 20-22.5; 9 19-21.

Angola: Wing, <? 110; 9 101, 102. Tail, <? 98; 9 93. Culmen,
tf1 22, 22; 9 21.

Immatures are little smaller than adults: Wing, d" 99, 102, 102;
9 94, 97. Tail, cf 91, 92, 96; 9 88, 90.

I have seen no birds from the eastern edge of the range (eastern
Belgian Congo).

Chlorocichla flavicollis Swainson

There are three well-defined subspecies or groups of subspecies in
this species: Upper Guinea to Central Cameroon, birds with yellow
throat and brownish under parts {flavicollis, with adamauae a syno-
nym) ; southern Cameroon to Gabon and most of the Belgian Congo,
birds with whitish throat and pale grayish olive breast and flanks
(soror and ?simplicicolor); from Angola to Northern Rhodesia and
north to Kenya, birds which are intermediate between the first two
in characters but not in geographical range (flavigula + pallidigula) .

Some further variation permits additional subspecies to be recog-
nized within these groups, but some variation within the groups is too
slight to use in establishing subspecies.

Chlorocichla flavicollis flavicollis Swainson

This Upper Guinea race is very distinct from the other forms in
its large size, bright yellow throat and under parts heavily washed
with brown.

RAND: AFRICAN BULBULS 189

C. f. adamauae Reichenow, 1910, from Ngendero [Genderu]
Mountains, north Cameroon, was described as differing from flavi-
coUis in the less pale, grayer tone of the under side. Bannerman
(1936, Bds. Trop. W. Afr., 6: 163) examined the type in the Berlin
Museum and while considering it a doubtful race recognized it and in-
cluded a Nigerian specimen in it, on the basis of the grayer under
side.

I have two specimens from Galim, Cameroon, that should be
adamauae, coming as they do from just east of the type locality,
and the American Museum has lent me one specimen of C. f. flavi-
collis from Senegambia. The two Galim birds are very similar to
each other, and on the upper parts are very similar to the Senegam-
bia bird. On the under parts the Galim birds differ slightly in the
rather more intensely yellow throat. The breast and abdomen of
one Galim bird are about the same, those of the other slightly browner
(not grayer) than those of the Senegambia birds. The postulated
characters of adamauae are thus not substantiated.

It should be noted that the quite different race soror has been
taken at Tibati, only about 30 miles south of Galim.

Measurements are as follows:

Senegambia: Sex? wing, 116; tail, 106; culmen, 24; tarsus, 25.

Cameroon: Wing, d" 122; 9 119. Tail, <? 114; 9 100. Culmen,
c? 23; 9 22. Tarsus, tf1 25; 9 24 mm.

Chlorocichla flavicollis simplicicolor Grote

This is a doubtful race on the northern edge of the range of the
race soror with apparently a very limited range. Bannerman (1936,
Bds. Trop. West Afr., 4: 163) has examined the type and accepts it
as a race, like soror in having a white throat but much browner
below.

Chlorocichla flavicollis soror Neumann

This is the race of Cameroon, Gabon (type locality), and most
of the Belgian Congo (see Chapin, 1953, Bull. Amer. Mus. Nat.
Hist., 75A: 142). It is quite different from flavicollis in the nearly
white, not bright yellow throat, and the much less brownish under
parts.

Chapin (loc. cit.) has pointed out minor size variation in this sub-
species, and our specimens (Gabon, 9; Moyen Congo, 5; Cameroon,
11) also show this: Gabon birds are smallest (wing, d* 101-107),

190 FIELDIANA: ZOOLOGY, VOLUME 35

Moyen Congo birds slightly larger (wing, cf 103-109), and Came-
roon birds largest (wing, d" 110-111 mm.). Cameroon birds are also
warmer, more brownish-tinged on the breast and flanks and slightly
warmer-tinged above than the others.

We have specimens of soror from Tibati, while we have flavicollis
from Galim, only about 30 miles north, without any signs of inter-
grading, which suggests that the two may possibly be found to meet
as species.

Chlorocichla flavicollis flavigula Cabanis

Sclater (1930, Syst. Av. Aethiop., p. 379) synonymized soror (type
locality in Gabon) with flavigula (type locality Angola) and recog-
nized both pallidigula Sharpe (type locality Entebbe, Uganda) and
shelleyi Neumann (type locality Muansa, south shore of Lake Vic-
toria). Bannerman (1936, Bds. Trop. West Afr., 4: 162) recognizes
soror and lumps flavigula and pallidigula, while Chapin (1953, Bull.
Amer. Mus. Nat. Hist., 75A: 142) recognizes soror, flavigula, and
pallidigula, with shelleyi a synonym of pallidigula.

Our material from Angola (6), Northern Rhodesia (4), southeast-
ern Congo (2), Uganda (34), and Kenya (3), compared with Gabon
birds (topotypical soror), indicates that Angola flavigula is a quite
distinct race compared with soror, on the basis of the yellower throat,
darker and more greenish olive (less grayish) chest and flanks, and
more greenish olive back. The Gabon soror were taken in 1952, the
Angola flavigula in 1954 ; so there is no question of foxing.

Kenya birds and eastern Uganda birds are again quite different
from Angola birds for the throat is still more yellow, and the breast,
flanks and upper parts are warmer brownish olive. (There is the
possibility that part of this difference is due to foxing, the Kenya-
Uganda birds being considerably older.)

Intermediate populations geographically (Northern Rhodesia,
eastern Congo [Kivu], and western Uganda) are variably interme-
diate in characters, and some western Uganda birds, especially those
from Bugoma and Bwamba, show an approach to soror.

It seems advisable to follow Chapin and make a rather arbitrary
division of these populations into a southern, less brownish-tinged,
less yellow-throated race, flavigula (Angola-southern Belgian Congo-
Northern Rhodesia), and a northeastern race, pallidigula (northern
end of Lake Tanganyika to Uganda and Kenya), with these charac-
ters more pronounced.

RAND: AFRICAN BULBULS 191

Chlorocichla flavicollis pallidigula Sharpe

The Kenya (Kavirondo) birds have the characters of this race
most pronounced (see under flavigula) and surprisingly look like exact
intermediates between the quite different soror and flavicollis, which
occur within 30 miles of each other in Cameroon without demon-
strated intergradation.

Chlorocichla flaviventris Smith

This is a bird of eastern and southern Africa, with apparently one
well-characterized race, a bird with brownish upper parts and tail, in
South Africa (Natal and Zululand) ; and a cline from a pale greenish-
backed bird in Bechuanaland to a darker brownish-backed bird with
a darker crown in Kenya — a cline complicated by variation in size,
in part correlated with considerable areas and in part local. Obvi-
ously, subjective views will affect the number of names used to des-
ignate this variation. At least eight are available. I adopt three:
one for the South African subspecies, and one for each end of the
cline, giving three subspecies, which accords with the treatment by
Sclater and Chapin.

In this species foxing is pronounced, older skins being browner
olive, less greenish olive on the upper parts, and this must be taken
into account when making comparisons or considering descriptions.

Chlorocichla flaviventris centralis Reichenow

This includes the northwest populations of the species, from
northern Portuguese East Africa and Tanganyika to Jubaland, to
parts of which four names have been applied: centralis Reichenow,
1887 (type locality Loeru, Tanganyika Territory) ; mombasae Shelley,
1896 (Mombasa), described as darker above than centralis; meru-
ensis Mearns, 1914 (Meru Forest near Mount Kenya), described as
like mombasae but top of head darker, back more greenish olive, and
under parts yellower (wing, type, o\ 101); and chyuluensis van
Someren (1939, Jour. E. Afr. and Ug. Nat. Hist. Soc., 14: 68), de-
scribed as darker above and paler below with more grayish flanks,
compared with all the above.

Our material indicates but slight color differences in specimens
from coastal Kenya (Rabai, Sokoke, etc.), Jubaland, Teita Range,
Nairobi area (Karura and Kiambu Forests) and Meru near Mount
Kenya, with individual variation in the larger series more than
bridging slight average differences. I have seen no Chyulu speci-
mens and follow Mackworth-Praed and Grant (1940, Ibis, p. 325)

192 FIELDIANA: ZOOLOGY, VOLUME 35

and Chapin (1953, Bull. Amer. Mus. Nat. Hist., 75A: 139) in con-
sidering them indistinguishable.

The Kilosa birds, from somewhat south and east of the type
locality of centralis, show a definite tendency toward occidentalis in
the paler and greener back, though with crown still definitely darker
and browner than back, and paler yellow under parts. These ap-
proach certain Northern Rhodesia birds in color but have a longer bill.

In size there are slight differences between certain populations.
Coastal Kenya and Kilosa (1700 feet) birds are smaller than the
inland Kenya birds in wing and tail.

However, as the color differences represent the northern end of
a cline and the size differences in wing and tail are small, it seems
advisable to group all these populations under one name.

Specimens examined. — Tanganyika Territory: Kilosa, 2. Kenya
Colony: Coastal (Rabai, Sokoke, etc.), 59; Teita Range, 3; Nairobi
area, 3; Meru, near Mount Kenya, 1. Jubaland: Juba River, 2.

Measurements are as follows: Kilosa: Wing, cf 105; 9 102.
Tail, & 103; 9 99. Culmen, tf 24; 9 24.

Coastal Kenya: Wing, o" (10) 99-108 (av. 105.4); 9 (10) 96-102
(av. 98.1). Tail, c? 92-101 (av. 96); 9 84-94 (av. 89.2). Culmen,
<? 23-25; 9 21-23.

Jubaland: d" Wing, 103, 105. Tail, 92, 96. Culmen, 24, 24.5 mm.

Teita Range: Wing, & 111, 112; 9 104. Tail, d" 101, 101; 9 93.
Culmen, tf 24, 24; 9 22.

Nairobi-Mount Kenya area: d" Wing, 110, 110. Tail, 101, 101.
Culmen, 24, 24.

Chlorocichla flaviventris occidentalis

The populations included here have had three names applied to
them in part: occidentalis Sharpe, 1881 (type locality Angola); zam-
besiae Shelley, 1896 (Zambesia, probably near Victoria Falls accord-
ing to Sclater), described as like centralis but upper parts strongly
washed with olive (not yellowish); dilutior White (1946, Ibis, p. 80;
Balovale, Northern Rhodesia), described as like occidentalis but paler
yellow below, breast more strongly washed with olive, above duller,
less yellow olive.

These populations represent the southern and southwestern half
of a color cline, distinguished at the extremes (Kenya vs. Bechuana-
land birds) by the southern birds having the back much paler and
greener (less brownish), the crown less dark brownish and less con-

RAND: AFRICAN BULBULS 193

trasting with the back, the under parts paler and less vivid yellow,
the flanks less brownish-tinged and the under wing clearer yellow.
The Bechuanaland and southern Angola birds are very distinct, espe-
cially above, from the Kenya birds in color, but progressively from
western Northern Rhodesia, northern Angola and eastern Northern
Rhodesia the color approaches that of the birds from the southern
part of the range assigned to C. f. centralis. In size, all these south-
ern populations differ from centralis in the shorter bill (c? 20-21, one
23, vs. 23-25 mm.), which apparently is not clinal; in wing length
there is local variation and some populations average slightly smaller
than any centralis, while others fall within the range of some centralis
populations, and none quite reach the size of interior Kenya centralis.

Specimens examined. — Angola, 9 (Luanda, 2; Dondi, 1; Huila, 6);
Bechuanaland, 8; Northern Rhodesia, 11 (including a topotype of
dilutior).

Measurements are as follows: Angola (north) : Wing, d" 98; 9 94.
Tail, <? 93; 9 92. Culmen, d" 21; 9 21. Angola (south) : Wing, &
103-110. Culmen, 21-22.

Bechuanaland: Wing, <? 100, 101, 104, 108; 9 95, 99, 100, 101.
Tail, & 96, 98, 98, 99; 9 90, 92, 94, 94. Culmen, tf 21, 21, 21;
9 20, 20, 21.

Northern Rhodesia: Wing, <? 95, 99, 101; 9 95, 95, 96, 97, 101.
Tail, & 91, 95, 100; 9 88, 90, 92, 92, 96. Culmen, <? 20, 21, 21;
9 21, 21, 23.

Chlorocichla flaviventris flaviventris Smith

I have not seen this form of Natal and Zululand, but from de-
scriptions it is very distinct, being olive brown above with the tail
more reddish brown than the back. This race appears to be more
like Kenya centralis than like the geographically intermediate popu-
lations (occidentalis) .

Thescelocichla leucopleura Cassin

Though this species (type locality Gabon) is generally considered
to show no geographical variation, our four Liberia specimens sug-
gested that Upper Guinea birds might be separable as a subspecies
for which the name nivosa Hartlaub, 1855, is available, on the basis
of less yellow below, grayer breast, and less olive in the upper parts.
Chapin could find no great difference between Upper and Lower
Guinea birds, as I wrote in 1951 (Fieldiana, Zool., 32: 620). We

194 FIELDIANA: ZOOLOGY, VOLUME 35

have added to our collections a considerable number of Cameroon
and Gabon birds, and four from Western Uganda (Bwamba Valley) .

Though no clearly defined races are apparent, the Cameroon
birds average yellower below and less gray on the chest and more
olive above than do the Upper Guinea specimens, but the latter fall
within the range of variation of the Cameroon series. The Bwamba
birds are grayer on the chest, like Upper Guinea birds, but are yel-
lower below and also fall within the variation of the Cameroon series.
The Gabon birds (topotypes or nearly so) are similar to the Cameroon
birds except for averaging slightly less yellow below. There is no
geographical variation in size.

It seems impractical to separate any subspecies on these vague
differences.

Specimens examined. — Liberia, 4; Cameroon, 44; Gabon, 11;
Uganda (Bwamba), 4.

Measurements (of Cameroon birds) : Wing, d1 (10) 111-117 (av.
114.7); 9 (10) 105-112 (av. 109.5). Tail, <* (10) 101-106 (av. 103.1);
9 (10) 95-102 (99.4). Culmen, & 23-25; 9 22-25 mm.

Phyllastrephus

I retain the general limits and the general arrangement of this
group as outlined by Delacour and Chapin. Contra Chapin I fol-
low Delacour in not recognizing Pyrrhurus (or Prosphorocichla) as a
monotypic genus for scandens.

Phyllastrephus scandens

Phyllastrephus scandens scandens Swainson

There are two specimens in the Good-Cleveland collection of this
Upper Guinea form from Galim, northern Cameroon, which measure:
wing, tf 112; 9 108.

Phyllastrephus scandens orientalis Hartlaub

This northern Belgian Congo-Uganda, etc. race differs chiefly
from scandens of Upper Guinea in the under parts being whitish (not
pronounced yellowish buff), and in the lower back being more gray-
ish olive, less greenish.

Specimens examined. — Uganda, Bwamba valley, 8.

Measurements: Wing, <? 101, 101, 107, 110; 9 93, 97, 98, 101.
Tail, d" 92, 98, 99, 99; 9 88, 91, 93, 95. Culmen, 21-22 mm.

RAND: AFRICAN BULBULS 195

Phyllastrephus scandens acedis Oberholser

This southern Cameroon-Gabon-western Belgian Congo race is
most like orientalis but differs in having the breast and flanks darker
and more grayish, the crown and back darker, and, most obvious,
the rufous tail decidedly darker.

Specimens examined. — Southern Cameroon, 6.

Measurements: Wing, d* 100, 102; 9 95, 98, 100, 103. Tail, <?
91, 92; 9 86, 89, 90, 91. Culmen, d1 20.5, 22; 9 20-22 mm.

Phyllastrephus scandens upembae Verheyen, 1953

Described from Upemba National Park of the Belgian Congo as
like orientalis but with tail more nut brown and lighter, and with
feet darker, almost slate blue. Schouteden (1954, Ann. Mus. Roy.
Congo Beige, (4), 4, fasc. 1, p. 90) refers southern Kasai birds here.
I have not seen specimens.

Phyllastrephus terrestris

Sclater (1930, Syst. Av. Aethiop., p. 382) suggests that this is a
race of strepitans, and Delacour (1943, Zoologica, 28: 25) considers
them conspecific. However, van Someren (1932, Nov. Zool., 37: 342)
records both forms from southern costal Kenya (whence we now have
both forms, from the van Someren collection), and Friedmann and
Loveridge (1937, Bull. Mus. Comp. Zool., 81 : 230, 231) record both
from Kilosa, Tanganyika Territory. There is no question that the
two are distinct species, with terrestris having a range from Kenya
to Cape Colony and west to Northern Rhodesia and Bechuanaland.

There is considerable geographical variation in P. terrestris, as
follows: A large, dark bird in South Africa; a pale bird, little smaller,
in Bechuanaland and Northern Rhodesia; and a slightly smaller and
considerably more rufous brown bird in southeastern Kenya. A fur-
ther variation is a more olive and yellowish population in the central
Kenya area (4000 to 4600 feet altitude).

There has been much uncertainty as to limits of subspecies in
this species, and some names which have been applied to interme-
diate populations make difficult the allocation of names.

In view of the well-marked differences between the four forms, as
shown by our material, it seems probable that intermediate areas
will possess intergrading populations, with individuals not clearly
referable to any subspecies.

196 FIELDIANA: ZOOLOGY, VOLUME 35

Phyllastrephus terrestris bensoni van Someren

This is a race of very restricted distribution, confined to the lower
Meru-Chuka forest near Mount Kenya, from 4000 to 4600 feet,
whence van Someren described it in 1945 (Bull. Brit. Orn. CI.,
66:11).

We have a single female from the van Someren collection, Chuka,
4600 feet altitude, June 14, 1944 (wing 75, tail 85, culmen 20, tar-
sus 21). It differs strikingly from P. t. suahelicus in the olive, not
rufous brown, upper parts; the grayish olive wash on breast and
flanks, and under tail coverts; and the yellowish tinge of lower
breast and abdomen.

Phyllastrephus terrestris suahelicus Reichenow

The type locality is Musa, Bagamoyo, Tanganyika Territory.
Other names that are involved are intermedins Gunning and Rob-
erts, 1911, type locality Umbelluzi River, Portuguese East Africa,
and rhodesiae Roberts, 1917, type locality Machile River, northwest
Rhodesia. The treatment of these races and their ranges has varied
considerably. Most recently Peters and Loveridge (1953, Bull. Mus.
Comp. Zool., 1 10 : 112) have synonymized intermedins with suahelicus,
while Clancey (1952, Ann. Natal Mus., 12: 250) recognizes interme-
dins and suggests that further races in the area should be recognized.

Our material from Kenya, Tanganyika Territory, Portuguese
East Africa, Northern Rhodesia, and Bechuanaland, representing
populations which have been variously allocated in the past, indi-
cates that while series are variable, and each series seems slightly
different from every other, a broad pattern of variation emerges:
the northern birds are warmer, more brownish on the back, and
especially the upper tail coverts and tail are more reddish brown;
also, the flanks are more buffy, while the southern birds are more
olive above, with grayer flanks.

The extremes clearly are different enough to stand as two sub-
species. That additional subspecies are recognizable seems doubtful.
However, where to draw the line between the two subspecies and
what names to allocate is not at once so apparent.

Two birds from Kilosa, Tanganyika Territory, are the nearest
to topotypical suahelicus I have. These are the birds that Peters
and Loveridge (loc. cit.) compared when including intermedins and
Nyasaland birds in suahelicus. These Kilosa birds are not as brown
as average Kenya birds but they do fall within the range of varia-
tion of our considerable Kenya series. Consequently I apply the

RAND: AFRICAN BULBULS 197

name suahelicus, the only one available, to the browner end of the
series of populations. A specimen from north Mozambique is sim-
ilar to Kilosa birds, especially in the redder upper tail coverts, and
extends the range from southeastern Kenya to northern Portuguese
East Africa.

Birds from farther south I group in the more olive end of the
series of populations and include under the next subspecies.

Specimens examined. — Kenya, 25 (Sokoke, Rabai, Ganda) ; Tan-
ganyika Territory, 2 (Kilosa; MCZ); Portuguese East Africa, 1
(northern Mozambique; AMNH).

Measurements are as follows:

Kenya: Wing, d" (9) 85-91 (av. 88); 9 (10) 74-81 (av. 77.8).
Tail, <? (8) 82-90 (av. 85.7); 9 73-81 (av. 76.4). Culmen, d" 21-24;
9 18-20 mm.

Tanganyika Territory: Wing, & 91; 9 83. Tail, cf 87; 9 84.
Culmen, c? 24; 9 20 mm.

Phyllastrephus terrestris intermedius Gunning and Roberts

Under the preceding subspecies, suahelicus, I have outlined the
characters and the names available for this subspecies (intermedius
1911 and rhodesiae 1917), in which I group the birds from southern
Portuguese East Africa to Bechuanaland, and the edge of the Belgian
Congo (Ndola in Northern Rhodesia).

Our Bechuanaland birds average slightly duller above than our
Northern Rhodesia birds, but the difference is slight and there is
considerable overlap. Roberts called both these rhodesiae. Our speci-
mens from Beira, nearby Gorongoza District, and Tete, Portuguese
East Africa, agree better with Northern Rhodesia birds than with
Kenya birds, especially on the less rufous upper tail coverts, though
they may be slightly paler above. All these birds are best grouped
together as belonging to the more olive end of the intermedius-sua-
helicus series and listed under intermedius. Chapin (1953, Bull.
Amer. Mus. Nat. Hist., 75A: 161) extends the range (as P. t. rho-
desiae) to Gambos in Angola.

It seems advisable to emphasize that Noom6, quoted in a num-
ber of places as the type locality, is actually the name of a collector,
F. 0. Noome\

Specimens examined. — Northern Rhodesia: Lundazi district, 8;
Ndola, 1. Southern Rhodesia: Mount Selinda, 1 (MCZ). Bechuana-
land: 13 (Kabulabula, Maun, Kasane). Portuguese East Africa: 6

198 FIELDIANA: ZOOLOGY, VOLUME 35

(Beira, 1, AMNH; Gorongoza district, 1, AMNH; Tete, 4, MCZ).
Angola: 5 (Huila, Jan).

Measurements are as follows:

Northern Rhodesia: Wing, & 85, 87, 91, 92, 93, 95; 9 77, 81.
Tail, <? 85, 87, 92, 93, 95, 96; 9 83, 84. Culmen, & 22.5-24;
9 19, 20.

Southern Rhodesia: Wing, 85. Culmen, 23.

Bechuanaland: Wing, d" (8) 85-96 (av. 92.1); 9 79, 81, 81, 86.
Tail, d" (8) 86-101 (av. 93.6); 9 87, 89, 90, 91. Culmen, c? 21-24;
9 20-22 mm.

Portuguese East Africa: Wing, 92, 85, 81, 81, 78, 77. Tail, 95,
89, 85, 83, 80. Culmen, 20-25 mm.

Angola: & Wing, 91-94. Tail, 94-100. Culmen, 23-23.5.

Phyllastrephus terrestris terrestris Swainson

This form, from southern and eastern Cape Province north to
Transvaal and part of Zululand, is, according to our specimens, very
little larger in wing length than intermedins, but it does have a larger
bill and is much darker olive brown on the upper parts and tail and
has much darker, olive-colored flanks and a conspicuous wash of this
on the breast.

Without topotypical montanus Roberts, of northern Drakensberg
and Zoutpansberg, I follow Vincent (1952, Check List of Bds. South
Africa, p. 65) in considering it a synonym. It was characterized by
Roberts as dark like terrestris but with more rufous, and smaller
perhaps.

Specimens examined. — Cape Province, 2. Natal, 2 (AMNH
and MCZ).

Measurements are as follows:

Cape Province: & Wing, 92, 95. Tail, 91, 95. Culmen, 25, 25.

Natal: Wing, d* 95; 9 87. Tail, & 93; 9 88. Culmen, d1 25;
9 23.

Phyllastrephus strepitans Reichenow

I have pointed out, under terrestris, that strepitans is a species,
distinct from terrestris. It is plainly separable by the more rufous
coloration and the smaller size. There is considerable sexual dimor-
phism in size, and this must be considered when comparing the species.

The many synonyms of this species have been brought together
and dismissed by Friedmann (1937, Bull. U. S. Nat. Mus., 153: 116:

RAND: AFRICAN BULBULS 199

Calamocichla schillings! Reichenow, 1904; P. sharpei Shelley, 1880;
P. rufescens Hartlaub, 1892; P. s. fricki Meams, 1914; and P. pauper
Sharpe, 1895).

As Friedmann has pointed out, wear and fading changes the ap-
pearance of these birds greatly. Juba River birds tend to have the
breast whiter and flanks paler than southern coastal Kenya birds,
and van Someren (1932, Nov. Zool., 37: 344) has kept these separate
from coastal Kenya birds as pauper. However, individual variation
is such that I agree with Friedmann in recognizing no races. No
differences in size are evident. In van Someren's measurements his
"larger" race averaged 78 mm., while his smaller race averaged
77 mm. The two Kilosa birds fall within the range of variation of
Kenya birds.

Specimens examined. — Abyssinia: 2 (Sheik Hussein, Bale).
Kenya: southern coast, 39 (Changamwe, Rabai, Mazeras, Ganda,
Malinda, Kipini, Sokoke, Tana River, Voi, Teita, Simba); Juba
River, 34; Kenya, northern and western, 7 (Kaptir, Suk; Marsabit;
Muressi, Turkwell). Uganda: Moroto Mountains, 1. Tanganyika:
Moshi, 3; Mkomosi, 1; Kilosa, 2 (MCZ).

Measurements are as follows:

Juba River: Wing, tf (10) 73-84 (av. 78.6); 9 (10) 72-79 (av.
74.6). Tail, & 75-88 (av. 81.2); 9 75-85 (av. 77.7). Culmen,
cf 19-22; 9 19-21 mm.

Sokoke, coastal Kenya: Wing, & 72-82; 9 69-80.

Tanganyika: Moshi, Wing, d" 78, 79. Kilossa, Wing, d" 81;
9 76.

Abyssinia: Wing, cf 83; 9 76. Tail, d* 90; 9 83.

Phyllastrephus cerviniventris

This seems to be a bird of evergreen forest, and consequently
with an interrupted distribution in its predominantly East African
range. Jackson and Sclater (Bds. Kenya . . . , 1938, 2: 868) ques-
tioned its occurrence in Kenya Colony, but their doubts are quite
without foundation, though it is a rare bird in the northern part of
its range.

Our Northern Rhodesia specimens are variable, an extreme being
a Chinga stream bird that has dark, dusky breast and flanks. Com-
pared with them three Nyasaland (topotypical) specimens fall within
the range of variation of the Lundazi birds.

Through the kindness of Dr. Friedmann I have examined the
type of lonnbergi (d" wing, 81; tail, 81 mm.) and another Taveta

200 FIELDIANA: ZOOLOGY, VOLUME 35

specimen, as well as a Taveta and a Moshi specimen in Chicago
Natural History Museum. The type of lonnbergi differs from the
two Taveta specimens in the slightly darker back, the slightly more
brownish gray crown, and the slightly deeper fulvous under parts,
but the differences are very slight. The Moshi specimen is paler and
less fulvous below than any of the three Kenya specimens. Allow-
ing for individual variation and the change of appearance caused by
wear and fading, I would group all these birds together.

On the under parts these four northern birds fall within the range
of variation of our Nyasaland-Northern Rhodesia series. On the
upper parts there is a slight difference, the back being slightly
browner olive, and the crown more brownish gray in the northern
birds. However, in this material the difference is slight and I rec-
ognize no races.

Specimens examined. — Kenya: 3 (Taveta, 2 [1, USNM]; Tha-
raka District, 1 [type of lonnbergi, USNM]). Tanganyika Territory:
Kilimanjaro, Moshi, 1. Northern Rhodesia: Lundazi District, 10.
Nyasaland:3 (USNM).

Measurements are as follows:

Kenya: & Wing, 81, 87. Tail, 81, 85. Culmen, 21.

Tanganyika: 9 Wing, 80. Tail, 78. Culmen, 18.

Northern Rhodesia (allocated to sex by size) : Wing, d" 87, 87,
88, 88, 90; 9 74, 78, 78, 78, 81. Tail, & 83, 85, 85, 86, 89; 9 74, 75,
75, 79. Culmen, cf 21; 9 19.

Nyasaland: Wing, 80, 80, 85; tail, 80, 80, 81; culmen, 20-21.

Phyllastrephus fulviventris Cabanis

This seems to be one of the least known African bulbuls. We
have a series of 11 birds from Angola (Gabela and Mucosa, Heinrich
coll.). They were taken in tangled secondary tropical woods.

White (Ibis, 1944, p. 553) shows that the old Northern Rhodesia
record is an error and the bird does not occur there. He writes me
that he suspects it should be considered conspecific with P. cervini-
ventris.

Phyllastrephus poensis Alexander

This is a mountain species (type locality Fernando Po) whose
range on the African mainland has been extended to several of the
Cameroon mountains by Serle (1950, Ibis, p. 373). Apparently
Correia did not collect it on Fernando Po, as Amadon (1953, Bull.
Amer. Mus. Nat. Hist., 100: 423) does not list it.

RAND: AFRICAN BULBULS 201

Phyllastrephus albigularis adametzi Reichenow, 1916, from Ba-
menda, Cameroon Highlands, is a name that has had unfortunate
treatment. Bannerman (1936, Bds. Trop. West Afr., 4: 177) exam-
ined the type and considered it a doubtful race of P. albigularis.
Serle (1950, Ibis, p. 374) reported that Stresemann examined the
type and told him it was a specimen of P. poliocephalus. Chapin
(1953, Bull. Amer. Mus. Nat. Hist., 75A: 177) accepts this. But in
1954, Serle (Ibis, 1954, p. 63) writes that a printer's slip was respon-
sible for the 1950 allocation, and that adametzi is a synonym of
poensis, not of poliocephalus.

Specimens examined. — Mount Cameroon (4,000 feet), 1 9 .

Measurements: Wing, 78; tail, 77; culmen, 20.

Phyllastrephus hypochloris Jackson

Chapin (1953, Bull. Amer. Mus. Nat. Hist., 75A: 162) has com-
mented on the resemblance of this species to members of the genus
Andropadus, (here included in Pycnonotus), with reasons of structure
and habits for placing it in Phyllastrephus.

Specimens examined. — Uganda, 8 (Ankole, Bwamba Valley, Ma-
bira Forest).

Measurements: Wing, <f 75, 76, 81, 83; 9 67, 69, 73. Tail,
cf 73, 74, 76, 80; 9 70, 70, 72. Culmen, d" 19-20; 9 18-19.

Phyllastrephus baumanni Reichenow, 1895

A synonym is eburneus Bannerman, 1923, following Bannerman's
later treatment. This seems to be a rather rare shy bird from Upper
Guinea. We have a single specimen, and I am struck with its simi-
larity to hypochloris of the eastern Congo and Uganda. They have
the same proportions, the same color pattern (agreeing closely, in
this group that is notorious for similarity of plumage), the same rictal
bristles, and the same rather broad and short bill that is dark horn
and light horn in color.

Though they agree so closely, the differences existing are such as
characterize a good subspecies — chiefly the grayer throat, breast and
abdomen (rather than more olive) of baumanni. Possibly they should
be considered conspecific, but I retain this form as a species.

Phyllastrephus poliocephalus Reichenow

This is a bird of the Cameroon Mountains. We have no speci-
mens. Chapin (1953, Bull. Amer. Mus. Nat. Hist., 75A: 173) has

202 FIELDIANA: ZOOLOGY, VOLUME 35

suggested that it may be conspecific with flavostriatus and from the
descriptions this is quite possible and would accord with the treat-
ment of P. tephrolaema, a somewhat similar case. However, I am
retaining it as a species.

Phyllastrephus flavostriatus Sharpe

This is a bird of the evergreen forest of the highlands in the more
tropical parts of Africa, hence with a much interrupted range, but it
approaches sea level in Portuguese East Africa. Without adequate
material I largely follow Chapin's (1953, Bull. Amer. Mus. Nat.
Hist., 75A: 172 ff.) treatment of races with a few comments.

Several interesting points are apparent in the geographical varia-
tion: the northern races have a straighter bill than the southern
forms; P. alfredi of northern Nyasaland and adjacent areas has the
crown brownish like the back, thus contrasting with the gray-
headed condition of the rest of the races.

Vincent (1933, Bull. Brit. Orn. CI., 53: 134) described as new
P. alfredi itoculo, a brown-headed bird that occurred in Portuguese
East Africa along with P. f. litoralis (=P. f. tenuirostris) , and since
the two occurred together he kept P. alfredi as a species. However,
Grant and Mackworth-Praed (1940, Bull. Brit. Orn. CI., 60: 52),
after examining the type of itoculo, concluded that it is a synonym
of P. munzneri, one of the P. fischeri group, and is not closely related
to alfredi. Thus there is no reason for considering alfredi a separate
species. For details of range see Benson (1953, Checklist Bds. Ny-
asaland, p. 50).

Our specimens of alfredi, from Upper Chiri River (3), and Jombo,
Mukutu (2), Northern Rhodesia, differ somewhat, for the latter is
more brownish olive above (less grayish brown) and the yellow
streaking of the under parts is brighter.

P. /. kungwensis, rather similar to graueri, is separated from it
by the more different olivaceogriseus. For comments see Moreau
(1945, Ibis, p. 101).

P. /. tenuirostris Fischer and Reichenow, 1884, Lindi, has as a syn-
onym P.f. litoralis Vincent (1933, Bull. Brit. Orn. CI., 53: 133), type
locality near Netia, Mozambique Province, according to Grant and
Mackworth-Praed (1940, Bull. Brit. Orn. CI., 60: 63). The range
of this form was extended northward to the Usambara Forest by
Friedmann and then by Sclater and Moreau (Ibis, 1932, p. 676),
but it appears to have been unrecorded for Kenya until de Schauen-

RAND: AFRICAN BULBULS 203

see (1951, Proc. Acad. Nat. Sci. Phila., 103: 48) recorded one from
Fort Hall.

In the van Someren collection in Chicago is a single Kenya female
from Mount Kasigau, southwest of Voi, taken November 18, 1938
(wing, 85; tail, 83 mm.). Compared with an Usambara male this
Kenya bird is somewhat paler grayish below, with yellow streaking
slightly more vivid, and slightly grayer back.

Phyllastrephus debilis Sclater

Sclater and Moreau (Ibis, 1932, pp. 678, 679, and 1933, p. 436)
have given us the best review of this small East African species.
Van Someren (1943, Bull. Brit. Orn. CI., 64: 12-15) would break
this into two species: P. debilis with two races, debilis and rabai;
and P. albigulus with two races, albigulus and shimbanus. However,
Sclater and Moreau (loc. cit.) and Moreau's field work show that
rabai intergrades with albigulus; so only one species with three sub-
species is recognized here. The iris is brown or white, varying with
the subspecies.

Phyllastrephus debilis rabai Hartert and van Someren

P. albigulus shimbanus van Someren (1943, Bull. Brit. Orn. CI.,
64, p. 12), type locality Shimba Hills, is a synonym.

The iris is whitish cream or whitish yellow, as marked on the
labels of Rabai, Shimba, and Uluguru specimens.

Birds from Sokoke, Ganda and Rabai (type locality) are fairly
uniform. Birds from the Shimba Hills, which van Someren charac-
terized as a new race, shimbanus, intermediate in character between
rabai and albigulus, are in effect intermediate. Our specimens show
the back slightly darker and greener, the gray crown slightly darker,
and the breast and flanks slightly duskier, compared with rabai; the
iris is marked on the label as cream, as in rabai. In comparison with
albigulus the back is similar, but the crown is distinctly gray, not
green, the flanks and breast are considerably less grayish or grayish
olive, the abdomen has more of a yellowish tinge, and the iris is
white, not brown.

Sclater and Moreau (loc. cit.) have shown that rabai occurs in
the lower and eastern Usambara Mountains and intergrades with
albigulus there. Presumably these are like the Shimba birds. While
differences exist, the Shimba birds are only slightly different from
Rabai birds and it seems advisable to consider shimbanus a synonym.

204 FIELDIANAI ZOOLOGY, VOLUME 35

Uluguru birds are slightly paler gray on the head, slightly paler
green on the back, and slightly duskier on the flanks than Rabai
birds, but again the differences are slight and separation of them
seems inadvisable.

Moreau (1940, Ibis, p. 457) records this race from Mount Mafi.

Specimens examined. — Kenya, 42 (Sokoke, Ganda, Rabai, Shim-
ba Hills). Tanganyika Territory, 3 (Kimboza Forest, Uluguru
Mountains) .

Measurements: There is no variation in size, as the following
summary shows: Wing, & Sokoke, 61-70; Ganda, 69, 69, 70; Rabai,
64-69; Shimba, 65-69; Uluguru, 69 mm.

The details, from the largest (Sokoke) series, are as follows:
Wing, d1 (10) 61-70 (av. 65.2); 9 (7) 61-66 (av. 63). Tail, d" (7)
61-65 (av. 62.8); 9 56, 57, 59, 60, 62. Culmen, d1 16-17; 9 15-16
mm. Tarsus, 17-18.

Phyllastrephus debilis albigulus Grote

This race was described as a Macrosphenus (Sylviidae), and
Sclater, in his Systema, placed it in Suaheliornis (Timaliidae). The
short tarsus and general characters place it here.

Apparently this race is restricted to the Usambara and Nguru
Mountains. It intergrades with the preceding form, rabai, in the
lower altitudes and eastern part of the Usambara Mountains (see
Moreau, Ibis, 1933, pp. 436, 437), the intergradation being complete
within 8 miles and 2,000 feet altitude. For Nguru record see Moreau
(1940, Ibis, p. 457).

This race is quite distinct from rabai: the iris is brown (not whit-
ish); the crown green (not gray); the back darker green; the breast
darker and more grayish; the flanks darker, more grayish olive; and
the abdomen less yellowish streaked and tinged.

Specimens examined. — Lushoto Forest, Amani, 3.

Measurements: Wing, d" 66; 9 63, 66. Tail, & 59; 9 58, 60.
Culmen, d" 16; 9 15. Tarsus, 17 mm.

Phyllastrephus debilis debilis Sclater

Sclater (1932, Ibis, p. 679) says that this form is "a good deal
yellower and less gray on the underparts" than the other forms.

Apparently the race is known only from the type locality, In-
hambane, Portuguese East Africa, and Nchingidi, Rondo Plateau,
Tanganyika Territory (Peters and Loveridge, 1942, Bull. Mus.
Comp. Zool., 89:248).

RAND: AFRICAN BULBULS 205

Phyllastrephus lorenzi Sassi

Chapin (1953, Bull. Amer. Mus. Nat. Hist., 75A: 171) says that
this species is known from but three specimens taken in the eastern
Congo (Ituri) Forests, but Schouteden (1954, Ann. Mus. Roy. Congo
Beige, (4), 4, fasc. 1, p. 98) lists additional specimens from Uele
and Kivu.

Phyllastrephus fischeri

This species, ranging from southern Sudan, Kenya, and Uganda
to Portuguese East Africa and Nyasaland and thence west through
the southern Congo and Rhodesia to Angola, is in part a mountain
bird, with a much interrupted range. It is clearly divisible into three
well-defined groups on the basis of color differences, as follows:

fischeri: Lowlands of East Africa from Tana River to northern
Portuguese East Africa.

placidus: Mountain forests from northern Kenya (Marsabit) to
Nyasaland (an interrupted range).

sucosus and cabanisi: Forests from southern Sudan, Uganda, to
Northern Rhodesia and east to Angola; separable into two popula-
tions on size.

P. f. fischeri is a brownish olive bird above, browner on the crown;
under parts pale olive gray with belly paler and more yellowish;
throat white; sucosus and cabanisi have the upper parts greenish
with crown like back, and under parts very much more yellowish
green. These two forms are different enough so that the interme-
diate placidus with crown more or less brownish olive, back more
olive, with a slight yellow wash on the throat, and rest of under
parts intermediate yellowish olive can be given a definite range and
most specimens identified.

Differences in color between various populations within each of
these populations exist. But to define and delimit most of these
populations is a fruitless task. First, though it appeared from a
preliminary inspection that several subspecies could be separated on
color, it was soon apparent that make of skin affected size and sharp-
ness of demarcation of the white throat area, and also the depth of
the olive gray wash on the breast. Second, museum age of skin,
"foxing," as shown by Nairobi area birds taken at various times over
the years, causes an amazing change; the older skins are more olive
brown on the back, more brownish on the crown, and more reddish
brown on the wings and duller below. With this in mind, and with

206 FIELDIANA: ZOOLOGY, VOLUME 35

as great differences between samples due to age as any apparently
due to geographical variation, I found it impractical to separate any
populations on color, and any trends indicated are only tentatively
set forth.

Size variation can be more objectively evaluated, though the pro-
nounced difference between the sexes and the possibility of erroneous
sexing makes a hazard in evaluating small samples. Size variation
exists in all three groups outlined above. In one, sucosus-cabanisi,
the difference between northern (smaller) and southern (larger) birds
is such as is usually used in separating subspecies, and these I accept
as two subspecies. In fischeri, the size difference, again larger south-
ward, is at a level where some taxonomists would use it in recogniz-
ing two subspecies while others would not; in the third, placidus, the
variation in size is considerable, but the distribution of the popula-
tions differing in size is peculiar, a single mountain (Kenya) having
the largest-sized populations with medium-sized birds around it and
small birds in the south (not north as in the other races) . Any allo-
cation of populations to two races seems unreasonable, as concealing
more facts than it reveals.

A further factor that has as yet received little attention is that
each of the three types of plumage has a long narrow range, roughly
parallel in part at least. This implies that intermediate populations
would occur over long, narrow belts. Though Moreau (1937, Bull.
Brit. Orn. CI., 57: 128) is of the opinion that the lowland and the
mountain forms are two species that occur in the Usambara area and
do not meet because of habitat preferences (lowland vs. mountain
forest), this might not apply elsewhere. Chapin (1953, Bull. Amer.
Mus. Nat. Hist., 75A: 164) has mentioned what seem intermediate
trends in the Lake Nyasa and in the Mbulu area, and I discuss them
and another intermediate western Kenya population below.

With all this in mind, four subspecies, each with varying popu-
lations within it, have been adopted as outlined earlier. They are:

Phyllastrephus fischeri fischeri Reichenow

Synonyms are: grotei Reichenow, 1910; munzneri Reichenow,
1916; sokokensis van Someren, 1923; P. alfredi itoculo Vincent, 1933
(Bull. Brit. Orn. CI., 53: 134).

This is the race of the lowland forests of East Africa. There is
an increase in size from north to south, as the following measure-
ments show:

Kenya (Sokoke): Wing, & 86-92 (for details see below).

RAND: AFRICAN BULBULS 207

Southeastern Tanganyika Territory (Mikindani):1 Wing, cf 92,
93, 97.

Portuguese East Africa:1 Wing, d" 96.

However, as discussed above, I do not recognize this as sufficient
for setting up a southern, larger subspecies, though there are three
names available, with grotei Reichenow, 1910, the oldest (not munz-
neri as has been used), for the southern birds.

The type locality of fischeri has been given as Muniuni, near the
mouth of the Juba River, and this appears to have been the basis of
quoting the range as coastal Kenya "from Juba River" southward.
Actually Muniuni, the type locality as Reichenow spelled it, is on
the Tana River near its mouth, according to Reichenow, and I can
find no evidence for the occurrence of this form north of Witu, just
north of the Tana.

P. p. sokokensis van Someren, 1923, when described, was com-
pared only with placidus and its range as given included the actual
type locality of fischeri.

The types of grotei and munzneri are in Berlin, where Grote and
Stresemann examined them and decided they were of the fischeri type
of plumage, and the type of itoculo is in the British Museum, and
belongs also with them. (Moreau, 1947, Bull. Brit. Orn. CI., 67:
88-90, recognized the larger southern birds as a distinct subspecies,
meeting fischeri in the Usambara area.)

The range extends south to near Netia, Portuguese East Africa,
the type locality of itoculo.

Specimens examined. — Kenya, 70 (Sokoke, Rabai, Ganda, Shim-
ba Hills).

Measurements are as follows:

Kenya: Sokoke: Wing, d" (10) 85-91 (av. 88.9); 9 (10) 77-82
(av. 79.1). Tail, & 83-93 (av. 88.7); 9 79-83 (av. 81.1). Culmen,
d" 21-24; 9 19-22 mm.

Tanganyika Territory: Mikindani: Wing, 83, 92, 93, 97 (topo-
types of grotei). Sanyi, Mahenge: Wing, 96 (from Moreau).

Portuguese East Africa: Wing, d" 96; 9 84, 84, 85 (from Moreau).

Phyllastrephus fischeri placidus Shelley

Synonyms are: keniensis Mearns, 1914, Mount Kenya; cognatus
Grote, 1919, Mlalo, Wilhelmstal ; marsabit van Someren, 1930 (Jour.
East Afr. Ug. Nat. Hist. Soc., no. 37, p. 197), Marsabit; chyuluensis

1 Moreau, 1947, Bull. Brit. Orn. CI., 67: 88-90.

208 FIELDIANA: ZOOLOGY, VOLUME 35

van Someren, 1939 (Jour. East Afr. Ug. Nat. Hist. Soc., 14: 66),
Chyulu Hills.

The range is from the highlands of Kenya east of the Rift Valley
(Marsabit, Mount Kenya, Chyulu Hills) south to Portuguese East
Africa (Mount Namuli) and Nyasaland (Moreau, 1937, Bull. Brit.
Orn. CI., 57: 127) and to Mlanje (Benson, 1953, Checklist Bds.
Nyasaland, p. 50).

The most obvious character that groups these birds together is
the olive yellowish under parts.

When topotypes of all four of the names synonymized were com-
pared with topotypes of placidus from Kilimanjaro, as well as the
other specimens listed beyond, at first it appeared that several series
could be separated on color; but it was soon apparent that the make
of skin affected the size and sharpness of demarcation of the white
throat area and also the apparent depth of the dusky wash on the
breast. Further, museum age of skin ("foxing") as shown by Nai-
robi area birds causes considerable change, the older skins being more
olive brown on the back, more brownish on the crown, and more
reddish olive on the wings. With this in mind, and allowing for in-
dividual variation, I can only suggest the following apparent trends:
birds from the southern part of the range, especially Rungwe, Tan-
ganyika Territory, are slightly greener on the crown than northern
populations. Chapin (1953, Bull. Amer. Mus. Nat. Hist., 75A: 164)
has already commented on this for the Lake Nyasa area birds.

There are very definite trends in variation in size. The Mount
Kenya birds are largest, with Marsabit and Nairobi birds next, and
more southern birds smaller again.

However, any subspecies based on the varying tenuous charac-
ters of these scattered populations seems unsatisfactory. If the
southern birds were to be separated, presumably cognatus Grote,
given to the Usambara birds, would be applicable, despite the un-
fortunate location of the type locality, near the central part of the
intergrading series rather than at the southern end.

Specimens examined. — Kenya, 49 (Marsabit, 11; Chyulu Hills, 4;
Teita Hills, 2; Kasigua, Voi, 1; Nairobi area, 19; Mount Kenya
[Chuka, Meru, 12]. Tanganyika Territory, 12 (Kilimanjaro, 2;
Usambara, 5; Iringa District, 1; Rungwe District, 4).

Measurements are as follows:

Marsabit: Wing, <? 86, 89, 91; 9 77, 77, 77, 79, 80, 80, 81.
Tail, cf 80, 85, 87; 9 73, 73, 75, 75, 78, 78, 82. Culmen, c? 19-22;
9 18-20.

RAND: AFRICAN BULBULS 209

Mount Kenya: Wing, cT 92, 92, 92, 93, 93, 94, 96; 9 82, 84, 86,
86. Tail, & 85, 86, 87, 88, 89, 90, 93; 9 77, 79, 81. Culmen, <? 21-
23; 9 19-20.

Nairobi area: Wing, o" (10) 85-92 (av. 89.1); 9 (7) 75-85
(av. 80.8).

Kilimanjaro: Wing, sex? 80, 83.

Chyulu Hills: Wing, d* 87, 89; 9 78, 80.

Usambara: Wing, cf 83, 85, 86; 9 82, 85.

Iringa highland : Wing, 9 79.

Poroto Mountains, Rungwe District: Wing, d" 82; 9 76, 77.

Phyllastrephus fischeri sucosus Reichenow

Synonyms: modestus Reichenow, 1908 (in Semliki); dowashanus
Madarasz, 1910 (Ngare-Dowash [=Mara River]).

The range is from the highlands of southern Sudan, Uganda,
Kenya, west of the Rift Valley, nearby central northern Tanganyika
(Loliondo) and the eastern Belgian Congo to Baraka on Lake Tan-
ganyika in the Kivu.

The most constant characters separating this from the preceding
races are the much more intense yellowish throat and yellowish olive
under parts; the greener crown, more like the greenish back, is an-
other character but it is much affected by foxing.

The type locality of sucosus is Bukoba, Uganda. The name mo-
destus, from the Semliki area of the Belgian Congo, was shown to be
a synonym by Gyldenstolpe (1924, K. Sv. Vet.-Akad. Handl., (3),
1, no. 3, p. 177). P. dowashanus Madarasz from Ngare-Dowash
(=Mara River), Kenya, was considered to be a synonym of placidus
by Sclater in 1930 (Syst. Av. Aethiop., p. 384), but in 1938 (Jackson
and Sclater, 1938, Birds of Kenya, etc., 2: 869) he allocated it to
sucosus. This latter is probably correct, for Kericho birds, which
are the nearest we have to topotypes, are closest to sucosus examples
from Uganda though somewhat intermediate (see below) .

P. cabanisi (sometimes under the name sucosus) has often been
kept as a species. Chapin (1953, Bull. Amer. Mus. Nat. Hist., 74A:
165) reunites it with fischeri, and our specimens from Kericho and
from Eldama Ravine certainly show an approach to the placidus of
the Nairobi area in the paler under parts, though I consider them
closer to sucosus. There is also a slight increase in size in the Kenya
populations of this subspecies, another approach to placidus to the
east.

210 FIELDIANA: ZOOLOGY, VOLUME 35

Specimens examined. — Uganda, 8 (Toro, Bwamba Pass, Ankole,
Mengo, Mabendi). Kenya, 28 (North Kavirondo, Mau slopes, El-
dama Ravine, near Cherangari Hills, Kericho).

Measurements are as follows:

Uganda: Wing, & 82, 82, 83; 9 72, 73, 74, 75, 77. Culmen,
<? 18-20; 9 18.

Kenya: North Kavirondo: Wing, sex? 72, 75, 76, 77, 77, 78, 82,
82, 83, 85. Kericho: Wing, <? 84, 86, 87, 88; 9 76, 81, 82. Mau
slopes: Wing, <? 91; 9 73, 79, 83. Eldama Ravine: Wing, 9 77,
82 mm.

Phyllastrephus fischeri cabanisi Sharpe, 1881

Synonyms: sylvicultor Neave, 1909, Dikuwle River, Katanga.

The range includes Angola, southern Belgian Congo, Northern
Rhodesia, and parts of southwestern Tanganyika Territory (Ufipa)
(Chapin). This race is like sucosus but larger (see measurements).

The type locality of cabanisi is Angola. Chapin (1944, Ibis,
p. 544) has explained the earlier misunderstanding about the name
cabanisi and the reason why sylvicultor from the Katanga was later
given to the same bird.

Our Angola birds are topotypical cabanisi; our specimens from
extreme northwestern Northern Rhodesia are similar in color but a
little smaller; specimens from Lundazi, eastern Northern Rhodesia,
are a little more dusky on flanks and breast, but as they are about
the same size and the crown is as green as the back, they are best
grouped here.

This subspecies is represented by fresh skins only, in our collec-
tion. Compared with the older skins by which sucosus is repre-
sented, conspicuous color differences are apparent, especially the
clearer yellowish olive gray of the under parts and clearer green,
less brownish olive, of the upper parts of the fresh skins. Undoubt-
edly most, if not all, of these differences are due to museum age of
skins, and I use size in separating these populations, with the reser-
vation that some of the observed color differences may be good.

Specimens examined. — Angola, 10 (Luhanda, near Quela; Duque
de Braganca). Northern Rhodesia: northwestern, 6 (Mwinilunga,
Solwezi) ; Lundazi, 5.

Measurements are as follows:

Angola: Wing, d" 92, 93, 95, 95; 9 82, 86, 86, 87, 87, 90. Tail,
<? 89, 91, 95, 96; 9 83, 83, 84, 85, 90, 91. Culmen, d" 21-22;
9 19-20.

RAND: AFRICAN BULBULS 211

Northern Rhodesia: northwestern: Sex? 80, 81, 87, 88, 90, 91.
Lundazi: Sex? 82, 89, 91, 92, 92 mm.

Phyllastrephus icterinus

Chapin (1944, Ibis, pp. 543-545) has shown that icterinus and
xavieri (formerly almost universally called cabanisi, which name
properly is applied to a western race of fischeri) are two species,
occurring together over much of their range. Chapin also shows
that P. sethsmithi Hartert and Neumann, 1910, Budongo, Unyoro,
which has usually been considered a larger eastern race of icterinus,
is actually a synonym of xavieri.

The two species are separable on size only, comparing sex with
sex, but sexual dimorphism is great and the larger males of icterinus
are distinguishable from the smaller females of xavieri in museum
skins only by the sexing (wing, icterinus, d* 71-80; 9 65-70; xavieri,
o" 83-91; 9 72-77 [Chapin]). It is probable that more overlap
occurs than these figure show and that certain specimens could not
be allocated with certainty.

Though Chapin says icterinus does not reach farther east than
Semliki Valley of the Belgian Congo, we have two adult females
from Budonga that are certainly icterinus on size. Two other speci-
mens from Budonga, one labeled a male, the other a female, taken
about the same time, I refer to icterinus, though they could be female
xavieri.

Evidently the relative abundance of this species and xavieri varies
greatly. In our collection we have 39 specimens of xavieri but none
of icterinus from Bugoma Forest, while from nearby Budonga we
have the above-mentioned four icterinus only. However, that xavieri
occurs at Budonga is shown by its being the type locality of seth-
smithi, the type a male, having a wing of 90 mm. Furthermore,
from Cameroon we have 44 specimens of icterinus and only five
of xavieri.

I have seen no Upper Guinea specimens and accept Chapin's allo-
cation of Upper Guinea birds to P. i. icterinus Bonaparte, 1850, type
locality Guinea, and P. i. tricolor Cassin, 1857, type locality Muni
River for the lower Guinea birds, from Cameroon south and east,
based on their duller, less yellowish under parts. Compared with
Uganda birds our Cameroon birds perhaps average slightly duller
yellow on the under parts; otherwise they are very similar.

212 FIELDIANA: ZOOLOGY, VOLUME 35

Specimens examined. — Uganda, 7 (Budonga, Semliki); Gabon, 4;
Moyen Congo, 1; Cameroon, 44; Nigeria, 1.

Measurements are as follows:
Uganda: Wing, d" 74, 75; 9 68, 68. Tail, <? 69, 71; 9 59, 62.
Culmen, & 18, 18.5; 9 16.5, 17. .

Cameroon: Wing, & (10) 75-82 (av. 78); 9 (10) 69-74 (av. 72.3).
Tail, & 68-76 (av. 69.8); 9 59-69 (av. 64.3). Culmen, d" 19-20;
9 17-19.

Phyllastrephus xavieri Oustalet, 1892

This species has been known in part as P. cabanisi (which is prop-
erly the name of a western race of P. fischeri) and in part as P. icteri-
nus sethsmithi (which is a synonym of P. x. xavieri), until 1944, when
Chapin (1944, Ibis, pp. 543-545) straightened out the nomenclature
and showed that icterinus and xavieri were two closely related species
with a considerable overlap in their ranges (see under P. icterinus).
There are two races.

Phyllastrephus xavieri serlei Chapin (1949, Bull. Brit. Orn. CI., 69 :
70), type locality Kumba, British Cameroons, with a range in British
Cameroons, I have not seen. It is described as like P. x. xavieri but
paler yellowish below. Measurements: Wing, d" 84-87; 9 71-77.
Tail, d" 74-79; 9 61-68 mm. (Chapin).

P. x. xavieri Oustalet (1892, Naturaliste, (2), 6: 218), type local-
ity Bangui on the Ubangi River, is the form ranging from the forests
of French Cameroon across the forests of the upper Congo to the
Bugoma and Budongo Forests of western Uganda.

P. icterinus sethsmithi Neumann and Hartert, 1910, was described
from Budongo; the type, a male with a wing of 90 mm., was consid-
ered as a race of icterinus and the name was so used for many years.
However, Chapin showed that it definitely applied to this larger
species. Chapin has suggested that Uganda birds, being yellower
below than Congo birds, might represent a valid race to which the
name sethsmithi would apply. Compared with Cameroon birds our
Uganda birds do average yellower below, slightly darker olive green
above, and slightly darker brown on the tail. However, French
Cameroon birds appear intermediate between the pale yellow Brit-
ish Cameroon birds, P. xavieri serlei, and the more richly colored
eastern birds, and it seems advisable to use but two names in divid-
ing this one type of variation in the species — serlei for the western

RAND: AFRICAN BULBULS 213

birds and xavieri for the intermediates and the eastern birds — con-
sidering sethsmithi a synonym of P. x. xavieri.

Specimens examined. — Uganda, 42 (Bugoma, 39; Bwamba, 3);
Cameroon, 5 (Sangmelima, Lolodorf).

Measurements are as follows:

Uganda: Wing, d" (10) 86-92 (av. 88.7); 9 (7) 73-77 (av. 75.1).
Tail, <? 80-86 (av. 82.9); 9 66-72 (av. 69.2). Culmen, d" 21-23;
9 18-19.

Cameroon: Sex? Wing, 81, 84, 85, 89, 90. Tail, 71, 75, 78, 79, 80.
Culmen, 20-22.

Phyllastrephus madagascariensis Gmelin

This, the type of the genus Bernieria, has only one character
which could be used as a generic character, the more elongate and
more hooked bill of the male. It also has a slightly heavier tarsus.
Otherwise it is very like P. xavieri, and the bill of the smaller female
is very like that of the male of xavieri; additional differences, of more
importance than minor color characters, are the less elongate rictal
bristles and the green (not brownish) tail of madagascariensis.

There are two quite distinct races, separable on intensity of col-
oration, and this is correlated with humidity.

Phyllastrephus madagascariensis madagascariensis Gmelin,
1789

This is the dark, vivid yellow and olive bird of the forests of
eastern Madagascar.

Specimens examined. — Eastern Madagascar, 4 CMaroantsetra,
N. E., Forest of Sianaka, Manombo).

Measurements: Wing, <? 85, 90; 9 73, 73. Tail, d" 73, 82; 9 62,
65. Culmen, & 28, 28; 9 22, 22. Tarsus, d" 24; 9 23.

Phyllastrephus madagascariensis inceleber Bangs and Peters,
1926

This is the much paler bird, above and below, of the drier coun-
try, the northern half of western Madagascar and the rain forest on
Mount d'Ambre.

Specimens examined. — Northwestern Madagascar, 2 (Mount
d'Ambre, near Maromandia; and near Tsarakibany).

Measurements: Wing, d* 87; 9 75. Tail, d" 78; 9 68. Culmen,
o" 27; 9 24. Tarsus, tf 24; 9 22.

214 FIELDIANA: ZOOLOGY, VOLUME 35

Phyllastrephus zosterops Sharpe

From an ecological point of view, P. zosterops seems to stand in
the same relationship to P. madagascariensis as does P. icterinus to
P. xavieri of the rain forests of Africa, in each case a larger and a
smaller species of the same genus, with a large part of their ranges
in common and rather similar habitat. However, while madagascari-
ensis seems close to xavieri and xavieri is so similar to icterinus that
some unsexed specimens are difficult to identify, zosterops is not
nearly so similar to either madagascariensis or icterinus. Indeed
Salomonsen (1934, Ann. Mag. Nat. Hist., (10), 14: 68) puts zosterops
in a different genus (Oxylabes) from that in which he places mada-
gascariensis (Bernieria), and earlier Sharpe erected a genus for zos-
terops (Xanthomixis Sharpe, 1881). Delacour (1943, Zoologica, 28:
18, 25) has, however, grouped them under Phyllastrephus, as he has
also tenebrosa, xanthophrys and cinereiceps. While not entirely happy
with this solution, in lieu of a better one I am following it, listing
them at the end of the genus which at least has the advantage of
keeping them together. Oxylabes, with which some of these have
been associated, has as genotype 0. madagascariensis and is a quite
different bird, as Delacour (loc. cit.) has said; possibly it is a warbler.

The geographical variation in P. zosterops has been reviewed by
Delacour (1932, Oiseau, 2: 68) and by Salomonsen (op. cit., pp. 68-
71), and commented on by me (1936, Bull. Amer. Mus. Nat. Hist.,
72 : 453-455) . The variation outlined exists, but when material from
intermediate localities has been collected perhaps fewer names will
be used to indicate it. Salomonsen has overstressed the isolation of
his subspecies. The species undoubtedly has a continuous range in
the body of the humid forests that stretch along eastern Madagascar,
with an isolated population on Mount d'Ambre in the extreme north
and possibly others in peripheral forest areas, as near Fianarantsoa.
In general, the named subspecies will probably prove to represent
stages on general trends which seem as follows:

(1) fulvescens, Delacour, 1931, Mount d'Ambre, north Madagas-
car; a very pale race.

(2) andapae, Salomonsen, 1934, northern end of continuous for-
est; a somewhat darker form.

(3) "maroantsetrae," Salomonsen, 1934, just south of andapae;
darker again than andapae. It seems unnecessary to use a name for
this stage on the cline and I synonymize it with zosterops.

(4) zosterops, Sharpe, 1875, of the central and southern rain forest
of eastern Madagascar; the darkest race, especially dark in the Sia-

RAND: AFRICAN BULBULS 215

naka-Fanovana area and becoming pale southward (i.e., more like
"maroanlsetrae").

(5) ankafanae, Salomonsen, 1934, apparently of the forests at
higher altitudes near Fianarantsoa in southeast; a pale, less greenish
race, which I have not seen.

Bleda syndactyla

Bleda syndactyla syndactyla Swainson, 1837

Without Upper Guinea material I accept Chapin's (1953, Bull.
Amer. Mus. Nat. Hist., 75 A: 179) conclusion as to the validity of
this slightly browner race.

Bleda syndactyla multicolor Bocage, 1880

This race, characterized by slightly greener upper parts, ranges
from southern Nigeria and Cameroon to Angola and eastern Congo.
Though Nigeria birds are sometimes referred to the Upper Guinea
race, our single Nigeria specimen is not separable from Cameroon
birds. However, Marchant (1954, Ibis, p. 374) calls Ondo-Benin
birds syndactyla and extends the range of multicolor only to Owerri.
The populations included here are not completely uniform. Angola
birds are slightly clearer green above and have the outer edges of the
remiges clearer green than do Gabon birds. This is apparent in our
freshly collected Angola birds when they are compared with recently
collected Gabon and Cameroon birds, and is still more apparent
when they are compared with older Cameroon skins. This indicates
that part, but not all, of the observed difference is due to foxing.
This geographical variation in color presumably is due to Cameroon
birds showing an approach to the browner-backed B. s. syndactyla of
Upper Guinea.

From the description of Criniger multicolor Bocage there is no
doubt that it applies to this species, and the description is that of a
bird in immature plumage, as Chapin (1953, Bull. Amer. Mus. Nat.
Hist., 75A: 179) has pointed out. Thus, ogowensis Neumann, 1914,
from Gabon, the name usually used for this subspecies, becomes a
synonym.

One of the Gabon specimens, an adult male, from Cape Esterias,
near Libreville, is aberrant in having the yellow and green pigments
much reduced, and the upper parts and the flanks much more gray-
ish olive, while the throat and abdomen are pale, yellowish white.
Another male from the same locality is normal.

216 FIELDIANA: ZOOLOGY, VOLUME 35

Specimens examined. — Cameroon, 9; Gabon, 2; Angola, 6; Nige-
ria, 1 (If on, Prov. Ondo).

Measurements are as follows:

Cameroon: Wing, cf 114, 116, 120; 9 109, 111, 111. Tail, d" 101,
103, 103; 9 92, 92, 96. Culmen, o" 26-29; 9 25-27.

Angola: Wing, & 110, 112, 116, 119; 9 107, 110. Tail, & 102,
107; 9 100. Culmen, & 28-30; 9 24, 25.

Bleda syndactyla woosnami Ogilvie Grant, 1907

The type locality is in Uganda, hence all our material is practi-
cally topotypical. In addition to the considerably richer yellow
under parts these birds are smaller than West African birds.

Specimens examined. — Uganda, 27 (Mabira, Mengo, Bugoma,
Budongo, Toro, Kigezi, Bwamba); Belgian Congo, 1 (Kivu).

Measurements: Wing, d" (10) 105-115 (av. 109.4); 9 (8) 97-101
(av. 100). Tail, & 91-98 (av. 94.5); 9 (8) 84-88 (86.6). Culmen,
d1 25-27; 9 22-24 mm.

Bleda eximia

Bleda eximia eximia Hartlaub, 1855

This upper Guinea race is not represented in the Chicago
Natural History Museum collection.

Bleda eximia notata Cassin, 1856

This lower Guinea race, ranging from southern Nigeria to the
lower Congo, differs from the preceding in the brighter yellow spot
before the eye and wider yellow tips to the tail feathers (Chapin).

Specimens examined. — Cameroon, 11; Gabon, 15.

Measurements: Wing, d" (10) 93-101 (av. 97.5); 9 88, 89, 91.
Tail, cf (10) 82-92 (av. 87.7); 9 77, 80, 81. Culmen, & 20-23;
9 19-20 mm.

Bleda eximia ugandae van Someren, 1915

The duller spot in front of the eye, compared with notata, seems
a very good character. Compared with notata, the size is also larger.

Specimens examined. — Uganda, 4 (Ankole, Mabira) ; Belgian Con-
go, 1 (Ituri); French Middle Congo, 1 (Impfondo).

Measurements: Wing, <? 105, 107; 9 97, 102. Tail, d* 96, 98;
9 90, 91. Culmen, d* 25; 9 22.

RAND: AFRICAN BULBULS 217

Bleda canicapilla Hartlaub

Bannerman gave the range as from Sierra Leone to Togoland,
but expected it would also be found in southern Nigeria, whence our
specimen comes.

Specimens examined. — Nigeria, 1 (Prov. Benin, Uyere).

Measurements: cf , wing, 95; tail, 91; culmen, 23; tarsus, 28 mm.

Nicator vireo Cabanis, 1875

The type locality is Chinchoxo, Portuguese Congo. In 1952 Serle
described N. v. tando Serle (Bull. Brit. Orn. CI., 72: 95) with type
locality N'Dalla Tando, northern Angola, on the basis of lighter gray
forehead, brighter green occiput, mantle, back, rump, and upper tail
coverts, lighter gray breast and belly, and paler yellow under tail
coverts. He makes no mention of the number of specimens exam-
ined. In 1953, Chapin (Bull. Amer. Mus. Nat. Hist., 75A: 185-186)
recognized no subspecies but he commented on four Angola speci-
mens, saying that they are somewhat lighter below than are Came-
roon and Ituri specimens.

Specimens examined. — Cameroon, 7 d\ 2 9; Gabon, 1 <?;
Bwamba, Uganda, 2 d" ; northern Angola, 6 d" .

The Angola birds have a very slightly longer tail than do the
other birds. Measurements: d" tail: Bwamba, 82, 84; Cameroon,
79-85; Gabon, 80; Angola, 80-87.

In color I can see no difference between Cameroon and Angola
birds, not even an average one. The Bwamba birds are slightly
lighter green above than average West African birds, but they fall
within the range of variation, being very similar to paler Angola and
Cameroon birds.

Evidently no subspecies can be recognized.

Nicator gularis Hartlaub and Finsch

Though it is quite true that gularis of eastern Africa and chloris
of West Africa have separate ranges, these do not meet, no inter-
grades are known, and the differences on a number of characters are
so large that it seems advisable to keep chloris and gularis as two
species, following Chapin.

Criniger ndussumensis Reichenow

These birds always have a red-brown tail and upper tail coverts.
They look very much like C. c. calurus of Cameroon but the bill is

218 FIELDIANA: ZOOLOGY, VOLUME 35

slightly shorter and considerably more slender; there is a more pro-
nounced grayish-white spot in front of the eye; the rictal bristles and
nuchal hairs are less well developed; the flanks average darker, and
the under tail coverts are usually more buffy (less plain yellowish).

The bird recognized here as a species has been called a subspecies
of calurus in the Ituri area and an extreme variant of calurus in the
Cameroon area by Chapin (1948, Auk, 65: 444; and 1953, Bull.
Amer. Mus. Nat. Hist., 75A: 132), but Berlioz (1955, Bull. Mus.
Nat. Hist. Nat., (2), 27: 189) and White (1956, Bull. Brit. Orn. CI.,
76 : 158) have held that two species are represented.

I have borrowed the material in the American Museum that
Chapin collected, and that in the U. S. National Museum, which,
combined with that in Chicago, gave a series of nearly 150 speci-
mens of calurus and ndussumensis. Evidently ndussumensis is much
the scarcer, for I have only 13 specimens of it. This material indi-
cates a complex situation.

In the Cameroon and Gabon material every specimen is of un-
equivocal allocation. The correlation of the several small details
makes it extremely probable that we are dealing not with individual
variants but rather with species, as Berlioz and White have claimed.

In the eastern Belgian Congo, however, the picture is different.
A Moera bird and a Semliki Valley (lower Butagu River) bird agree
with Cameroon birds. On the other hand, specimens from the more
northern or western localities of Irumu, Medje, Avakubi and Ga-
mangui are closer to calurus, here represented by the greenish-tailed
race C. c. emini. A series of birds from Augumu, 190 km. west of
Lake Edward, are, as Chapin reported (op. cit., p. 136), partly emini,
but they show a considerable approach to ndussumensis on one or
another character. Some thin-billed birds are otherwise emini, and
some thick-billed birds have more color characters of ndussumensis,
but none has the complete ndussumensis characters; notably none
has as rufous a tail.

Chapin (loc. cit.) has studied this material and concluded that
ndussumensis is a race of calurus, with a small range in the Semliki
area, and that it intergrades with C. c. emini extensively in surround-
ing areas. His extensive field work in this area has convinced him
that the two forms represent but one species there. My findings on
his material agree with his conclusions for this area.

This situation in which two sympatric forms behave as species in
part of the range and hybridize in another part recalls the somewhat
similar situation but with most of the forms allopatric, in Pycnonotus

RAND: AFRICAN BULBULS 219

barbatus. The northern African races, arsinoe, goodi, nigeriae, gabo-
nensis, tricolor, and minor, form an intergrading connecting series
from the Sudan-Lake Chad area, Nigeria, Cameroon, Gabon, and
Congo, to Sudan again, where minor and arsinoe are said to live to-
gether as species. In the case of Pycnonotus barbatus the area of
overlap is so small, compared with the area where they represent
each other as subspecies, that I have called them all subspecies of
one species. In the present case Criniger calurus and C. ndussumen-
sis live together over a considerable area (White says from south-
eastern Nigeria to eastern Congo) while the birds hybridize only in
a very small area in the eastern Congo, so that the treatment of the
two as species seems preferable.

Specimens examined. — Cameroon, 8 d", 1 9 (Bitye, Sangmelima,
Lolodorf, Djoum). Gabon, 2 sex? (Mimongo-Poingi, Mouila-Mount
Tandou). Belgian Congo, 2 (Moera, 1; Butagu River, Semliki Val-
ley, 1;AMNH).

Measurements are as follows:

Cameroon: Wing, d" (7) 90-95 (av. 91.5); 9 87. Tail, d" (8)
74-80 (av. 77.1); 9 76. Culmen, <? 19-20; 9 18. Tarsus, 20-21.

Gabon: d1 wing, 82; tail, 74; culmen, 18.5; tarsus, 19.5.

Eastern Congo: d" wing, 96; tail, 82; culmen, 20; tarsus, 21.

Criniger calurus

The measurements of the two races that have been confused with
ndussumensis are given below for comparison.

Criniger calurus calurus Cassin, 1857

This is the brown-tailed form from southern Nigeria to the mouth
of the Congo River. Cameroon birds give the following measure-
ments: Wing, tf (10) 87-95 (av. 91.6); 9 (10) 85-90 (av. 87.6).
Tail, cf (10) 77-86 (av. 82.1); 9 (10) 74-82 (av. 78.3). Culmen,
c? 21-23; 9 20-22. Tarsus, d" 20-23; 9 21-22 mm.

Criniger calurus emini Chapin, 1948

This is the green-tailed bird from the Middle Congo River to
Uganda. Measurements of Uganda birds, where the species ndussu-
mensis does not occur are: Wing, d" (10) 89-97 (av. 92.4); 9 (10)
83-88 (av. 86.3). Tail, & (9) 83-90 (av. 85.4); 9 (10) 72-81 (av.
75.8). Culmen, cf 20-22; 9 19-20. Tarsus, & 22; 9 20-22.

The smaller bill of Uganda birds is a slight approach in length but
not in thickness to the smaller, more slender bill of ndussumensis.

220 FIELDIANA: ZOOLOGY, VOLUME 35

Tylas

Though Tylas has been excluded from the bulbuls by Delacour,
there doesn't seem to be a better place to put it. Delacour's sug-
gestion that it is a thrush, possibly, is not very convincing.

The genus contains one species, T. eduardi Hartlaub, in which
two subspecies are recognizable: T. e. eduardi of the east, and T. e.
albigularis Hartlaub of western Madagascar (see Delacour, 1932,
L'Oiseau et Rev. Franc. Orn., II, p. 70; and Rand, Bull. Amer.
Mus. Nat. Hist., 72:457).

Publication 855