(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Bulletin of the Museum of Comparative Zoology at Harvard College"

HARVARD UNIVERSITY 

Library of the 

Museum of 

Comparative Zoology 



Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 

Vol. LXXIX, No. 1 



SCIENTIFIC RESULTS OF AN EXPEDITION 
TO RAIN FOREST REGIONS IN EASTERN AFRICA 



NEW REPTILES AND AMPHIBIANS FROM EAST AFRICA 



By Arthur Loveridge 



CAMBRIDGE, MASS., U. S. A. 

PRINTED FOR THE MUSEUM 

February, 1935 



No. 1. — Scientific Results of an Expedition to Rain Forest Regions 

in Eastern Africa 

I 
New Reptiles and Amphibians from East Africa 

By Arthur Loveridge 

During 1933 and 1934, as a fellow of the John Simon Guggenheim 
Foundation, I was given the opportunity of investigating on behalf 
of the Museum of Comparative Zoology, the herpetologieal fauna of 
certain remnants of rain forest in East Africa. It is proposed to pub- 
lish a series of reports dealing with the material which was collected 
incidental to the investigation. Certain of the specimens, however, 
were recognized in the field as being undescribed forms, two others 
were subsequently differentiated; it appears advisable to publish 
descriptions of these without further delay. 

Full information as to altitudes, and other data concerning the type 
localities, will be published later in the Introduction to the reports. 
In the report dealing with Herpetology, more precise notes will be 
furnished as to the habitat, food, etc., of these new species. 

Fortunately with only three exceptions, adequate series were 
obtained of all the new things described below. The actual numbers 
taken or utilized in the descriptions, together with the type localities, 
are as follows : 

28 Testudo pardalis babcocki subsp. nov. Mount Debasien, Karamoja, Uganda. 

1 Typhlops kaimosae sp. nov. Kaimosi, Kakamega, Nyanza Province, K. C. 
62 Natrix olivacea uluguruensis subsp. nov. Nyange, Uluguru Mtns., T. T. 

1 Natrix olivacea pembana subsp. nov. Pemba Island, Tanganyika Territory. 

4 Coronella semiornata fuscorosea subsp. nov. foot of Mt. Mbololo, K. C. 

6 Aparallactus turneri sp. nov. Sokoki Forest, near Malindi, K. C. 

19 Agama agama kaimosae subsp. nov. near Kaimosi, Kakamega, K. C. 

20 Riopa tanae sp. nov. Kau, near the mouth of the Tana River, K. C. 
11 Riopa mabuiiformis sp. nov. Ngatana on the Tana River, K. C. 

44 Acontias percivali sp. nov. foot of Mount Mbololo, Taita, K. C. 

52 Chamaeleon bitaeniatus altaeelgonis subsp. nov. Kaburomi, 10,500 feet, 

Mt. Elgon, Uganda. 
30 Boidengerula taitanus sp. nov. Mt. Mbololo at 4,800 feet, K. C. 

3 Arthroleptides dutoiti sp. nov. Koitobos River, Mount Elgon, K. C. 
75 Hyperolius milnei sp. nov. Witu, Coast Province, Kenya Colony. 

The last is named in appreciation of the generous hospitality of 
Mr. R. D. Milne, who was not only my host but assisted me in hunting 
frogs during the five days which I spent near Witu. 



4 bulletin: museum of comparative zoology 

The Arthroleptides is named after its discoverer, Dr. A. C. du Toit 
of Cape Town University, who visited the eastern slopes of the moun- 
tain when I was on the western. The occurrence on Mount Elgon of a 
member of this hitherto monotypic genus is of extraordinary interest 
and significance. 

Testudo pardalis babcocki subsp. nov. 

Type. Museum of Comparative Zoology, No. 40,003. An adult 9 
from the western slopes of Mount Debasien, Karamoja, Uganda at 
5,500 feet, collected by Arthur Loveridge, November 23, 1933. 

Paraiypcs. Thirteen specimens in the Museum of Comparative 
Zoology from the following localities: Kenya Colony: Guaso Nyiro 
River; Meru River; Ithanga Hills; Wema, Tana River; Mount 
Mbololo, Taita. Tanganyika Territory: Arusha; Saranda; Pwaga; 
Tukuyu. South Africa: three without definite locality. 

Fourteen specimens in the British Museum (N.H.) from the fol- 
lowing localities: Sudan: Bahr el Gebel. Uganda: Mount Elgon 
between 9,000 and 10,000 feet. Kenya Colony :Ndogo, Lake Baringo. 
Tanganyika Territory : Mgana; "East Central Africa (Speke coll.)." 
Mashonaland: Mount Darwin. Bechuanaland : Magalapsi. South 
Africa : Algoa Bay ; and five without definite locality. 

Diagnosis. The southeast and east African Leopard Tortoise has 
long been confused with the typical form which was described by Bell 
as from "Promont. Bonae Spei." Fortunately Bell gave the length 
and depth of the type which places it within the range of the low- 
shelled form inhabiting southwest Africa. What was probably the 
type specimen, (which should be in the Zoological Museum, Cambridge, 
England,) was beautifully figured by Sowerby and Lear (1872, Tor- 
toises, Terrapins, and Turtles, p. 3, pis. ix and x). Seven examples of 
the typical form and twenty-eight of the new form have been utilized 
in assessing the ranges of variation of the two forms. 
Carapace high, the greatest height of shell being in- 
cluded in its greatest length from 1.61 to 2.07 times. 
(South, East and Central Africa to the Bahr el Gebel) . T. p. babcocki 
Carapace low, the greatest height of shell being in- 
cluded in the greatest length from 2.02 to 2.62 times. 
(Southwest Africa Protectorate and probably Cape 
Peninsula, as type locality is given as Cape of Good 
Hope) T. p. pardalis 



loveridge: new African reptiles and amphibians o 

It should perhaps be mentioned that only 2 of the 28 examples of 
the new form are shallower than 1.94 times. Both of these are British 
Museum measurements based on an adult (453 mm.) from Ndogo, 
Lake Baringo, Kenya Colony (2.00 times), and a juvenile (54 mm.) 
from Mgana, Tanganyika Territory (2.07 times); this last locality I 
have failed to locate on the maps available. 

The average of height into length is 1.81 for the 28 examples of the 
new form as against 2.18 times for 7 specimens referred to typical 
pardalis. As, however, 4 of the latter are British Museum material 
with only "South Africa" for data they are somewhat arbitrarily 
assumed to be of the typical race. 

Remarks. Ten years ago when I first saw two examples (adult of 
302 mm., and young of 55 mm. long) of the Leopard Tortoise from 
Kolmanskop and Aroab in the Southwest Protectorate I was struck 
by the low vaulted shell as compared with that of the reptile with 
which I was familiar in East Africa. They form a parallel to Kinixys 
belliana of the savannah and the depressed K. spekii of the more arid 
districts. 

Through the exceeding kindness of Mr. H. W. Parker, who has 
furnished me with detailed measurements and other particulars of all 
the Leopard Tortoises in the British Museum, I have become con- 
vinced that the high-vaulted type merits subspecific distinction and in 
designating it I am glad to associate the name of my friend Dr. Harold 
L. Babcock, author of "The Turtles of New England" and Curator 
of Reptiles in the Museum of the Boston Society of Natural History. 

Typhlops kaimosae sp. nov. 

Type. Museum of Comparative Zoology, No. 40,060 from Kaimosi 
Forest, near Friends' Africa Mission, Kakamega district, Nyanza 
Province, Kenya Colony, collected by Arthur Loveridge, March 7, 
1934. 

Diagnosis. Differs from T. p. punctatus and all other East African 
members of the genus Typhlops in possessing an ocular which is 
broadly in contact with the nasal shield below the preocular, thus 
separating the latter shield from the upper labials; no subocular as is 
present in T. p. gierrai. 

Differs from T. praeocidaris Stejneger of Leopoldville, Belgian Congo 
in the possession of 28 (instead of 24-26) midbody scale-rows; diameter 
of body being included in total length 43 (instead of 67) times; rounded 
snout (instead of with sharp cutting edge); habit; coloration. 



6 bulletin: museum of comparative zoology 

Description. Snout prominent, rounded, without obtuse horizontal 
keel; nostrils inferior; rostral large, more than half the width of head, 
extending almost to the level of the eyes, which are distinguishable; 
nasal semidivided, the suture extending from the first labial to the 
nostril; preocular present, narrower than the nasal, not more than half 
as broad as the ocular which is in contact with the nasal below the 
preocular so as to broadly separate the preocular from the labials; 
ocular in contact with the third and fourth labials; four upper labials. 
Diameter of body 43 times in the total length; tail broader than long, 
ending in a spine. Midbody scale-rows 28. 

Coloration. Above, uniformly black. Below, scarcely lighter than 
above except around the mouth and anus. 

Measurements. Total length 215 mm.; head and body 211 mm.; 
tail 4 mm. ; diameter at midbody 5 mm. 

Natrix olivacea and its races 

In 1925, and again in 1928, I drew attention to the fact that for 
the past forty years the common and widely-distributed Olive Water 
Snake was believed to possess invariably 19 longitudinal scale-rows 
at midbody. It is true that Fischer (1884, Jahr. Hamb. Wiss. Anst., 
1, p. 6) recorded an individual with 17 scale-rows from Masailand. 
Boulenger (1893, Cat. Snakes Brit. Mus., 1, p. 227), possibly thinking 
it a case of misidentification, omitted this record from his synonymy 
so that it has passed into oblivion. 

In the forested mountains of the Usambara, Uluguru and Rungwe 
there occurs a race with 17 scale-rows at midbody. Of sixty -three 
snakes collected by me in these mountains, no less than 89% possessed 
17 scale-rows, the rest 18 or 19 except for one with 16. When reporting 
on some of these in 1928, Dr. Barbour and I refrained from describing 
this race until the opportunity occurred for securing an adequate 
series of the typical form (which was described from Tete, Mozam- 
bique) from the coast and savannah regions round about. 

That opportunity occurred during my visit to East Africa last year, 
and I have been able now to assemble 94 specimens of the typical 
form, or their data, from all parts of the range. Omitting Fischer's 
record, as Masailand was a vague term applied to the country over 
which the Masai roamed and embracing several mountains, in all 
these 94 snakes only one (M. C. Z. 30,074 from Albertville, Belgian 
Congo) had 17 scale-rows while three from the Tana River had 18 
(though these displayed 19 slightly in advance of mathematical 



loveridge: new African reptiles and amphibians / 

midbody). I propose, therefore, to designate the smaller montane 
form with 17 scale-rows as Natrix olivacea uluguruensis. 

On the single example of the Olive Water Snake known from Pemba 
Island, however, there are only 15 scale-rows. As several mainland 
reptiles have distinctive races or representatives occurring on Pemba, 
I depart from my usual practice and designate this insular race 
Xatri.v olivacea pembana on the basis of a single specimen. It should 
be remembered that Pemba, unlike Zanzibar, is separated from the 
mainland by a channel of exceptional depth — 400 fathoms; eight 
times the depth of the channel between Zanzibar and the adjacent 
coast. 

The three forms of the Olive Water Snakes may be defined as fol- 
lows : 

Midbody scales usually in 19 longitudinal rows 

(17 rows in only 1 specimen of 94 examined). 

Tropical and South Africa o. olivacea 

Midbody scales usually in 17 longitudinal rows (19 

rows in only 7 specimens of 63 examined). 

Uluguru, Usambara and Rungwe Mtns., Tanganyika, o. uluguruensis 

Midbody scales in 15 longitudinal rows (only the type 

specimen known.) Pemba Island o. pembana 



Natrix olivacea uluguruensis subsp. nov. 

Natrix olivaceus (part) Barbour & Loveridge, 1928, Mem. Mus. Comp. Zool., 
50, p. 109; (part) Loveridge, 1933, Bull. Mus. Comp. Zool., 74, p. 231. 

Type. Museum of Comparative Zoology, No. 23,117. An adult 9 
from Nyange, Uluguru Mountains, Tanganyika Territory, collected 
by Arthur Loveridge, October 8, 1926. 

Paratypes. Sixty-one specimens from various localities in the 
Uluguru, Usambara and Rungwe Mountains as listed in the citations 
given above. 

Description. Scales in 19 rows on nape immediately behind head, 
17 rows at midbody, 17 rows at anus; ventrals 136; anal divided; 
subcaudals 74; labials 8, 4th and 5th entering the orbit; preocular 1; 
post oculars 3; temporals 1 + 2. 

Measurements. Total length 350 (250 + 100) mm. 

Remarks. This montane race attains smaller dimensions and lays 
fewer eggs than the typical form of the lowlands. See citations for 
further details. 



6 bulletin: museum of comparative zoology 

Natrix olivacea pembana subsp. nov. 
N atrix olivacea (part) Loveridge, 1925, Proc. Zool. Soc. London, p. 71. 

Holotype. Museum of Comparative Zoology, No. 19,112. A 9 from 
Chakechake, Pemba Island, Tanganyika Territory, collected by 
Salimu bin Asmani, October 6, 1923. 

Description. Scales in 17 rows on nape, 15 rows at midbody, 15 
rows at anus; ventrals 127; anal divided; subcaudals 56; labials 8, 
4th and 5th entering the orbit; preoculars 2; post oculars 3; temporals 
1 + 2. 

Measurements. Total length 288 (210 + 78) mm. 

Remarks. This snake is discussed in detail in the citation given 
above. 

CORONELLA SEMIORNATA FUSCOROSEA subsp. nOV. 

Type. Museum of Comparative Zoology, No. 40,555. An adult 
cf from the lower slopes of Mount Mbololo, Taita, Coast Province, 
Kenya Colony, collected by Arthur Loveridge, April 25, 1934. 

Paratypes. Museum of Comparative Zoology, Nos. 40,556-7, being 
two females with the same data as the type. No. 40,554, a juvenile 
from near the station at Tsavo, Coast Province, Kenya Colony. 

Diagnosis. The following key emphasizes the distinguishing charac- 
ters of the new form. 

Labials 8, 4th and 5th entering the orbit 1 ; preoculars 
1, very rarely 2; olive or plumbeous above, white below 
usually each ventral scale so heavily edged with black 

as to present a plumbeous appearance s. semiornata 

Labials 9, 5th and 6th entering the orbit; preoculars 2, 
rarely 1 ; bright brownish pink above, clear pink below. . s. fuscorosea 
Description. Snout not prominent, rostral much broader than deep; 
internasals as long as the prefrontals; frontal longer than its distance 
from the end of the snout, shorter than the parietals; loreal longer 
than deep (adults) or as long as deep (young); preoculars 2 (1 on left 
side of head in Nos. 40,554-5) the upper in contact with (or separated 
from) the frontal; postoculars 2; temporals 2-f- 2 (2 + 3 on left side 
of head in No. 40,554); upper labials 9, the 5th and 6th entering the 
orbit; four lower labials in contact with the anterior chin-shields, 

'This was the case in C. scheffleri Sternfeld from Kibwezi, which is a synonym of C. s. semi- 
ornata for they agree in all respects. A topotype of scheffleri (M.C.Z. 40,553), however, has 
9 upper labials, 5th and 6th entering the orbit on the left side of the head only, 2 preoculars on 
the right. This is the only intermediate in a very large series, or records, of semiornata examined. 



loveridge: new African reptiles and amphibians 9 

which are as long (or longer) than the ill-developed posterior pair. 
Midbody scales in 21 rows; ventrals 181 (183-198 in paratypes); 
anal divided; subcaudals 96 (82-92 in paratypes). 

Coloration. Above, uniformly bright brownish pink except for an 
ill defined dusky bar from the 5th and 6th labials across the frontal, 
a distinct one from the 8th and 9th labials across the parietals, and 
a well-defined, though shorter, one on the nape at a distance of 6 
(or 5) scale-rows behind the parietals. Below, uniformly pink. 

The paratype young one exhibits 64 faint bars or paired spots along 
the back. 

Measurements. Type c? . Head and body 472 mm., tail 186 mm. 
Paratype 9 (M.C.Z. 40,556). Head and body 500 mm., tail 150 mm. 

Aparallactus turneri sp. now 

Type. Museum of Comparative Zoology, No. 30,117. A d" from 
Sokoki Forest, near Malindi, Coast Province, Kenya Colony, col- 
lected by H. J. Allen Turner, Esq., in June, 1932. 

Paratypes. Museum of Comparative Zoology, Nos. 40,120-4 being 
five snakes from Peccatoni; Mkonumbi; and near Witu. All these 
localities being in Coast Province, north of the Tana River, south of 
Lamu Island and near the coast. Collected by Arthur Loveridge, 
May 24, 28, and 31, 1934 respectively. 

Diagnosis. Very closely related to A. wemeri Boulenger of the 
Usambara and Uluguru Mountains, Tanganyika Territory. It may be 
readily distinguished from that species by the fewer ventrals (120-139 
as against 141-163 in wemeri, the latter figures being based on scale 
counts of fifty specimens from seven localities, forty-three of the 
snakes are topotypes), different coloration, much smaller size (202 
mm. maximum as against 354 mm. maximum). 

Description. Diameter of eye greater than its distance from the 
oral margin; rostral twice as broad as deep, the portion visible from 
above not more than one-third its distance from the frontal; inter- 
nasals much shorter than the prefrontals; frontal once and a half as 
long as broad, much longer than its distance from the end of the snout, 
as long as the parietals; nasal entire, in contact with the preocular; 
a single (a pair in all paratypes) postocular in contact with the anterior 
temporal; temporals 1 + 1; 6 upper labials, second and third entering 
the orbit; first lower labial in contact with its fellow behind the 
symphysial (barely in contact in No. 40,124) ; two pairs of chin shields, 
the anterior broader and a little longer and in contact with three lower 



10 bulletin: museum of comparative zoology 

labials. Scales in 15 rows; ventrals 129 (120-139 in paratypes); anal 
entire; subcaudals 42 (31-37 in paratypes). 

Coloration in alcohol. Above, head and nape black except for white 
blotches, one anterior, one posterior, to the eye, the second blotch 
larger and extending upwards on to the anterior temporal; also a 
narrow white color just posterior to the parietals so separating the 
black of the head from the black of the nape (in all paratypes there 
are vestigial indications of a second light color immediately posterior 
to the black of the nape); back uniformly pallid, slightly pinkish, 
brown with the edges of each scale darker (in all paratypes there is a 
fine, hair-like, black line along the vertebral row of scales to the end 
of the tail). Below, uniformly white except for a slight encroachment 
of the black nape patch in two downward-pointing patches. 

Measurements. Type cf . Head and body 140 mm., tail 35 mm. 
Largest specimen (M.C.Z. 40,120) also a cf , head and body 167 mm., 
tail 35 mm. 

Agama agama kaimosae subsp. nov. 

Type. Museum of Comparative Zoology, No. 40,136. A cf from 
rocky heights three miles west of the Friends' Africa Mission Station 
at Kaimosi, Kakamega, Kenya Colony, collected by Arthur Loveridge, 
March 2, 1934. 

Paratypes. Museum of Comparative Zoology, Nos. 40,137-40,150 
and four unnumbered duplicates, being seven males and eleven females 
taken at the same locality as the type on March 2-9, 1934. 

Diagnosis. Most nearly related to A. a caudospina Meek with a 
cotype of which it has been compared. It agrees with that race in the 
greatly depressed body and numerous other characters, but differs in 
the less developed spinosity of the tail and its bright pink color (yellow- 
in caudospina) as well as in general coloration. 

Midbody scale-rows 86 (range in paratype males 76-86, with an 
average of 81; in females 78-88, with an average of 83); preanal 
pores 11 (range 10-13, with an average of 11 for the eight males). 

Coloration in life, cf • Above, snout and crown of head brown 
overlying red; nape, back and two-thirds of upper arm, reddish buff 
with grey vermiculations ; rest of arm and whole of hind limbs a bril- 
liant, somewhat metallic, purplish blue; tail bright pink. Below, 
edge of jaws grayish white; throat dull red with purplish-mauve tones 
overlying gray vermiculations; breast pinkish purple; abdomen dull 
purple overlaid with brown; whole of undersurface of limbs bright 



loveridge: new African reptiles and amphibians 11 

purplish blue; palms of hands as well as soles of feet and digits, grayish 
white; tail bright pink. 

9 . Above, gray with irregular sepia brown markings along the 
vertebral line; plumbeous on limbs and sides. Below, throat grayish- 
white with gray vermiculations ; limbs grayish-white except base of 
hind legs which are mustard yellow, as are the abdomen and underside 
of tail. 

Measurements. Type cf . Head and body 144 mm., tail 211 mm. 
Largest perfect paratype 9 . Head and body 113 mm., tail 152 mm. The 
largest cf and 9 measure 148 and 133 mm. respectively in length 
from snout to anus but the tails of both specimens are truncated as 
is the case with more than half of the series. 

RlOPA TANAE sp. nov. 

Type. Museum of Comparative Zoology, No. 40,251. An adult 9 
from Kau, near the mouth of the Tana River, Coast Province, Kenya 
Colony, collected by Arthur Loveridge, June 4, 1934. 

Paratypes. Museum of Comparative Zoology, Nos. 40,252-9 and 
six others all with the same data as the type; No. 40,261 from near 
Witu, just north of Kau; Nos. 40,262-3 from Golbanti, Tana River 
above Kau; Nos. 40,264-5 from Ngatana, Tana River above Golbanti. 

Diagnosis. A very slender, attenuated species with small, penta- 
dactyle limbs, the hind limb being included from 334 (young) to 6 
(adult) times in the distance between the axilla and groin. A pair of 
supranasals form a suture separating the rostral from the frontonasal, 
nostril in a single nasal which forms a horizontal suture with the 
supranasal. Midbody scales smooth, in 22 to 24 rows. 

Description. Snout but slightly depressed, not wedge-shaped as in 
sundevallii from the same region. Nostril in a single nasal which is 
separated from its fellow by a pair of supranasals ; frontonasal broader 
than long; prefrontals present, widely separated; frontal as long 
(slightly longer and slightly shorter in some paratypes) as the fronto- 
parietals and interparietal together, in contact with the 1st and 2nd 
supraoculars; parietals in contact behind the interparietal and each 
bordered along its posterior edge by 2 (left), 3 (right) or 4 (some 
paratypes) large scales; supraoculars 4; supraciliaries 8 (usually 7, 
rarely 6) ; loreals 2 ; preoculars 2 ; lower eyelid with a large, undivided 
opaque disk; upper labials 7 (8 in No. 40,253), 5th largest and below 
the eye (6th in No. 40,253 and on one side only in two others) ; ear- 
opening small, round, no larger than the nostril. Limbs very short, 



12 bulletin: museum of comparative zoology 

pentadactyle, the adpressed anterior limb just reaching the ear (in 
adult and young) ; the length of the posterior limb is contained 6 times 
(334 in young, the proportion varying regularly with age) in the dis- 
tance between the axilla and groin. Fingers short, the 3rd a trifle 
longer than the 4th; toes short, the 4th a trifle longer than the 3rd 
(equal to, or shorter than, in some paratypes), the 5th extending as 
far forward as the 2nd, further than the 1st; 10 (or 9) lamellae beneath 
the 4th toe. Scales smooth, in 22 (or 24) rows; preanals 3, slightly 
enlarged. 

Coloration in alcohol. Above, uniformly plumbeus, upper labials 
flecked with lighter; between the head and fore limb frequently 
several scales with light flecks (in life these were pale green and so 
arranged as to give the impression of vertical barring on the neck; 
only adult skinks, and not all of these, possess this barring). Below, 
throat and lower labials white heavily spotted with dark brown; 
belly grayish white but each scale with a dark spot; tail almost plum- 
beus as a result of the increase in size of these spots. (In some para- 
types the spots on the belly and tail are arranged along the lateral 
edges of the scales so as to produce the appearance of 8 longitudinal 
lines along the underside from hind chin to anus). 

Measurements. Type 9 • Head and body 89 mm., tail 58 mm., hind 
limb 11 mm., fourth toe 3 mm. 

1 RlOPA MABUIIFORMIS Sp. nov. 

Type. Museum of Comparative Zoology, No. 40,266. An adult c? 
from Ngatana, Tana River, Coast Province, Kenya Colony, col- 
lected by Arthur Loveridge, June 14, 1934. 

Paratypes. Museum of Comparative Zoology, Nos. 40,267-71 and 
five others with same data as type but collected June 14-19, 1934. 

Diagnosis. A very large Mabuia-like species with well developed 
pentadactyle limbs, the hind limb being included from 2 (young) to 
3 (adult) times in the distance between the axilla and groin. A pair 
of supranasals form a suture separating the rostral from the fronto- 
nasal, nostril so large as to nearly divide the single nasal shield, latter 
forming a horizontal suture with the supranasal. Midbody scales 
smooth, in 28 to 30 rows. Color of adults and young different. 

Description. Snout neither depressed nor wedge-shaped. Nostril 
in a single nasal which is separated from its fellow by a pair of supra- 
nasals; frontonasal broader than long; prefrontals present, widely 

1 Note. This species differs from Riopa as defined by Boulenger in that the frontal is decidedly 
broader than the supraocular region. 



loveridge: new African reptiles and amphibians 13 

separated; frontal decidedly broader than the supraocular region, as 
long as the frontoparietals and interparietal together, in contact with 
the 1st and 2nd supraoculars; parietals in contact behind the inter- 
parietal and each bordered along its posterior edge by 3 or 4 large 
scales; supraoculars 4 (5 on one side of No. 40,268 by subdivision); 
supraciliaries 8 (7 in seven paratypes); loreals 2; preoculars 2; lower 
eyelid with a large, opaque disk; upper labials 7, 5th largest and below 
the eye; ear-opening large with two small rounded lobules on its 
anterior border. Limbs well developed, pentadactyle, the adpressed 
anterior limb reaching well beyond the ear (in adult and young) ; the 
length of the posterior limb is contained 3 times (2 in young) in the 
distance between axilla and groin. Fingers long, the 3rd a trifle longer 
than the 4th; toes moderately long, the 4th longer than the 3rd by a 
claw length (in whole series), the 5th extending as far forward as the 
3rd, much further than the 1st; 16 (or 15) lamellae beneath the 4th 
toe. Scales smooth, in 30 (or 28) rows; preanals with a median pair 
slightly enlarged. 

Coloration in life, c? and 9 adults. Above, uniformly plumbeous 
but on closer inspection each scale is seen to be lighter at its base, the 
light area tends to increase in size towards the tail so that the scales 
on the tail are light centered with dark edges; both upper and lower 
labials white, each barred with brown or black posteriorly; scales on 
the sides white, heavily edged with black on their posterior border. 
Below, uniformly white. 

Young. Above, black, crown of head mottled with pale brown, 
occipital scale conspicuously white with dark centre; a vertebral 
stripe of pale brown, one scale in width, commences behind occipital 
scale and continues on to base of tail where it disappears, the vertebral 
stripe is flanked on either side by a dorso-lateral stripe of same color 
but two scales in width and commencing at the last supraocular; 
limbs uniformly black; tail transparent red, each scale edged with 
brown. Below, pure white, the internal organs visible through the 
scales; tail clear coral pink. 

Measurements. Type c? and 9 . Head and body 95 and 91 mm., 
tail 141 and 121 mm., hind limb 22 and 20 mm., fourth toe 65 and 
64 mm. 

Acontias percivali sp. nov. 

Acontias meleagris Loveridge (not of Linn6), 1923, Proc. Zool. Soc. London, 
p. 964. 

Type. Museum of Comparative Zoology, No. 40,174. An adult 9 



14 bulletin: museum of comparative zoology 

(largest of forty) from the foot of Mount Mbololo, Taita Mountains, 
Kenya Colony, collected by Arthur Loveridge, April 26, 1934. 

Paratypes. Museum of Comparative Zoology, Nos. 40,175-40,200 
being thirty-eight specimens with the same data as the type. Also 
Museum of Comparative Zoology, Nos. 31,079-31,080 and three skinks 
in the Coryndon Memorial Museum, Nairobi, collected at Voi near 
Mount Mbololo, by A. Blayney Percival in 1914. 

Diagnosis. Only separable from meleagris (Linne) of South Africa 
by its shorter tail and coloration. According to Boulenger (1887, 
Cat. Snakes Brit. Mus., 3, p. 427) the tail of meleagris is "not more 
than one fifth of the total," it is included 6.1 to 6.3 times in the total 
length in examples from the Transvaal and Orange River Colony in 
the Museum of Comparative Zoology. In forty examples of percivali 
it ranges from 8.4 to 10 times (average exactly 9). In alcohol percivali 
has a very broad dorsal band of deep black while below it is pure 
white; there is no dark spot on each scale resulting in a lineolate or 
gray appearance as in meleagris. 

It is important to note that the key character used by Boulenger to 
separate meleagris and plumbcus of Mozambique is inconstant though 
the two species are very distinct. In percivali about 25 specimens have 
the first supraocular larger than the second and third together, in 
13 specimens it is equal to, in 3 possibly smaller. In the Museum of 
Comparative Zoology examples of plumbeus the first supraocular is as 
large as, or smaller than, the other two; in our meleagris it is as large 
as, or much larger than, the other two. 

The range of meleagris is separated by 1,500 miles from that of 
percivali. 

Description. Head conical; snout obtuse, projecting; ear hidden; 
rostral enormous, covering the greater part of the snout; mental 
enormous, its posterior border reaching to below the eye; frontonasal 
hardly half as long as the rostral, much shorter than the frontal, 
which is broader than long; supraoculars 3, the first much larger than 
the two others together (equal to the two others occasionally) ; supra- 
ciliaries 4 (on right) or 3 (on left); interparietal much narrower than 
the frontal, longer than broad (sometimes as broad as long), narrower 
than the parietals (rarely as broad as); no labials enter the orbit. 
Midbody scale-rows 18 (16, 17 or 18 in paratypes, average for forty- 
two skinks 17.5) those of the two median dorsal rows transversely 
enlarged. A single, very large preanal plate. Limbs absent. Length 
of tail included 9.6 times in total length (8.4 to 10 times in paratypes). 

Coloration in life. Both adults and young are glossy black above 



loveridge: new African reptiles and amphibians 15 

except when about to shed the epidermis. At such times they are 
bluish gray above, a delicate^ pink below. Normally the young are 
reddish orange below, this shade sometimes persisting until they are 
two-thirds grown. Adults are pale chrome below; very occasionally 
an individual will be found with a brown centre to each lower scale 
resulting in the production of a faintly lineolate appearance. For 
coloration in alcohol see diagnosis above. 

Measurements. Type 9- Head and body 234 mm., tail 27 mm. 



Chamaeleon bitaeniatus altaeelgonis subsp. nov. 

Type. Museum of Comparative Zoology, No. 40,274. An adult cf 
from Kaburomi, 10,500 feet, on the western slopes of Mount Elgon, 
Uganda, collected by Arthur Loveridge, December 28, 1933. 

Paratypes. Museum of Comparative Zoology, Nos. 40,275-40,300, 
being twenty one males and thirty females with the same data as the 
type. 

Diagnosis. Most closely related to C. b. hohneli, from which it is 
distinguished only by its smaller size. See remarks below. 

Measurements. Type c? . Head and body 78 mm., tail 73 mm. 
M.C.Z. 40,275, Paratype 9 . Head and body 75 mm., tail 73 mm. 

Remarks. Kaburomi lies in the tree heath (Erica arborea) zone 
immediately below the alpine zone of Mount Elgon. On reaching it, 
I was immediately struck by the small size and different coloring of 
the chameleons as compared with those with which I was familiar 
at Sipi (circa 6,500 feet) and of which I had a good series, collected 
during the three weeks stay at Sipi. 

Almost all our specimens from both places were adult and breeding. 
A tabulation of their measurements by sexes immediately shows 
that at Kaburomi these chameleons attain to about three-quarters 
the size which they do at lower levels. 

Largest of 22 males from Kaburomi 151 mm. 

Largest of 22 males from Sipi 199 mm. 

Largest of 30 females from Kaburomi 148 mm. 

Largest of 30 females from Sipi 190 mm. 

The explanation is doubtless to be found in the relative scarcity 

of insect life at the higher altitude coupled with the fewer hours for 

feeding. For we observed at 10,500 feet it was so cold that reptiles 

remained lethargic or sluggish until about 11 a.m. by which time the 

sun had dispersed the mist and warmed the atmosphere. The difference 



16 bulletin: museum of comparative zoology 

in coloring was attributable to the altered character of the vegetation 
and different tones of color resulting therefrom. 

Woosnam (in Boulenger, 1909, Trans. Zool. Soc. London, 19, p. 
245) noted a similar difference between the chameleons at 6,000 and 
10,000 feet on Mount Ruwenzori. Speciation had gone further in 
this instance and Boulenger gave the upper zone form a name. It is 
now known as C. b. rudis, but is very different from the Elgon races. 
Parker (1932, Linn. Soc. Journ. Zool., 38, pp. 227-9) refers the upper 
zone form from Mount Kenya at 14,000 feet to C. b. schubotzi Stern- 
feld, discusses the derivations of some of the races of bitaeniaius and 
figures them. 

Boulengerula taitanus sp. nov. 

Type. Museum of Comparative Zoology, No. 20,001. An adult d 1 
from the absolute summit, 4,800 feet, of Mount Mbololo, Taita 
Mountains, Coast Province, Kenya Colony, collected by Arthur 
Loveridge, April 14, 1934. 

Paratypes. Museum of Comparative Zoology, Nos. 20,002-20,024, 
being twenty nine caecilians with the same data as the type except 
that they were taken from 4,000 feet to 4,800 feet and between April 
14 and 24, 1934. 

Diagnosis. In its large size and coloration resembling a Scoleeo- 
morphus and totally unlike Boulengerula with which, however, it 
agrees in structural characters, possessing two rows of teeth in the 
lower jaw and a lateral tentacle situated above, and about midway 
along, the upper jaw. 

Intermediate in number of annuli between B. uluguruensis Barbour 
& Loveridge and B. changamwensis Loveridge, agreeing with the former 
in the number of annuli but differing in that the tentacle is slightly 
nearer to the nostril than to the corner of the mouth, in uluguruensis 
this position is reversed. B. taitanus differs from changamwensis 
both in number of annuli and in position of the tentacle. It differs 
from both in size and coloration. The young, however, agree in color. 

Description. Habit moderate, vermiform. Snout obtusely pointed, 
projecting far beyond the lower jaw; eye indistinguishable; tentacle 
round, exsertile, with indications of a circular groove surrounding it, 
situated rather nearer to the nostril than to the corner of the mouth ; 
26 or 28 teeth round the upper jaw; 146 annuli (136-148 in the para- 
type series), these annuli are interrupted in the middle dorsal line 
except on the nape and posteriorly. 



loveridge: new African reptiles and amphibians 17 

Coloration in life. Type cf . Above, glossy black, each annular ring 
blue gray except along the vertebral line. Below, blue gray blotched 
with brown, except for the throat which is uniformly pink; the cir- 
cumanal area is pale blue gray. 

Very young specimens are flesh pink like adults of changamwensis. 
Immediately after being chloroformed they turn gray above, but 
remain pink below. 

Coloration in alcohol. Type cf . Generally plumbeous but the gray 
blue and pink areas become white. 

Measurements. Type cf. Total length 355 mm., midbody diameter 
7 mm. Paratypes. Total lengths 73-360 mm., midbody diameters 
1.8-7 mm. Diameters included in total lengths from 40.5 to 53.3 
times. 

Arthroleptides dutoiti sp. nov. 

Type. Museum of Comparative Zoology, No. 19,864. An adult, 
gravid 9 from the Koitobos (Koitobross) River, eastern slopes of 
Mount Elgon, Kenya Colony at about 7,200 feet, collected by Dr. A. 
C. du Toit, January 8, 1934. 

Paratypes. No. 150 of du Toit collection. An adult d 71 and young 
specimen with the same history as the type. 

Diagnosis. This very distinct little frog may be readily distinguished 
from the only other member of this hitherto monotypic genus as 
follows: 
Toes half-webbed ; buccal margin scarcely flattened, the 
orbits projecting beyond it when viewed from above. 
Color below, light violet brown slightly or moderately 

flecked with white. Adult 9 31 mm dutoiti 

Toes with only a rudiment of web at their base; buccal 
margin strongly flattened, extending well beyond the 
orbits when viewed from above. Color below, white, at 
most the throat violet brown in males. Adult 9 59 

mm., cf 74 mm martiensseni 

Description. Vomerine teeth absent. Head slightly broader than 
long; snout subacuminate, not depressed as in martiensseni; canthus 
rostralis sharply defined; loreal region concave; nostril equidistant 
between end of snout and anterior border of orbit ; tympanum distinct, 
not quite two-thirds the diameter of the eye; interorbital space about 
equal to the width of an upper eyelid. Tips of fingers and toes strongly 
dilated, the disks' having a median groove; digits without web; toes 



18 



bulletin: museum of comparative zoology 



half-webbed the membrane extending as a narrow fringe to the disk 
of the 1st, 2nd and 5th toes on their inner margin, almost to the disk 
on the 3rd, the disk and two distal joints of the 4th toe free of web ; 
the tibio-tarsal articulation of the adpressed hind limb reaches slightly 
beyond the tip of the snout (well beyond in the cf). 

Skin above rugose (corrugated and distinctly warty in the cf) 
pitted. In martiensseni the skin is smooth, how far preservation may 
have accentuated these differences it is difficult to say. 

Coloration in alcohol. (After formalin preservation in the field). 
Above, uniformly black except for the digital expansions which are 
more or less white-edged. Below and posterior aspect of thighs, light 
violet brown slightly flecked with white in the type, rather more 
abundantly flecked and mottled in the paratypes. 







, 


Young 


Measurements. 


9 Type 


cf Paratype 


Paratype 


Length from snout to anus 


31 mm. 


25 mm. 


10.5 mm. 


Breadth of head 


12.5 mm. 


10.5 mm. 


4.5 mm. 


Length of head 


10.5 mm. 


9 mm. 


4 mm. 


Length of hind limb 


59 mm. 


50 mm. 


15.5 mm. 


Length of fourth toe 


13.5 mm. 


11.5 mm. 


3.5 mm. 



Hyperolius milnei sp. nov. 

Type. Museum of Comparative Zoology, No. 20,025. A gravid 9 
from Witu, Coast Province, Kenya Colony, collected by Arthur 
Loveridge, May 31, 1934. 

Paratypes. Museum of Comparative Zoology, Nos. 20,026-20,050 
being seventy-four frogs with the same data as the type. Nos. 20,051-2 
from Golbanti, Tana River and No. 20,053 from Malindi, all near the 
coast in Coast Province, Kenya Colony. 

Description. A small species of stouter build than its allies of the 
parlceri-usaramoae group. Snout obtusely acuminate {not sharply 
pointed); distance from the end of the snout to the nostril equal to 
two-thirds the distance from nostril to anterior border of the eye; 
distance from the end of the snout to the anterior border of the eye 
longer than the orbital diameter; tympanum hidden. Fingers one- 
third webbed; 1st, 2nd, 3rd and 5th toes are webbed to the disk (or 
practically to the disk) on their inner aspect, on the 4th toe the disk 
and last two phalanges are free of web. The tibiotarsal articulation 
of the adpressed hind limb reaches to the eye (in females, to the nostril 
in only one female) or to the nostril (in males, very occasionally falls 



loveridge: new African reptiles and amphibians 19 

short). Skin smooth above, granular on the belly, no strong fold 
across the chest. Males with a large subgular vocal sack and strongly 
granular disk. 

Coloration in alcohol. Above, white, a dusky line from end of snout 
through nostril to orbit, a black spot on upper eyelid, half-a-dozen 
scattered, irregularly disposed, black spots on anterior portion of 
back (maximum number of spots about 20 in Malindi para type; 
these spots are entirely absent in many paratypes, principally males), a 
similar spot on knee and elbow (often absent); in addition there are 
numerous, minute, reddish-brown flecks on top of head, back, fore 
arm and tibia; thighs colorless. Below, transparently white, internal 
organs visible through the skin. 

Coloration in life. Above, an unusual shade of cabbage green, a 
black canthal line through nostril to orbit, a number of irregularly 
disposed, black spots on head and back in addition to numerous 
reddish-brown flecks on top of head, back, fore arm and tibia ; thighs 
colorless; a broad, but indistinct, subdermal band of palest yellow 
from eye to flank present in males but not observed in females ; fingers 
and toes orange. Below, semi-transparent, throat greenish in 9 , 
the eggs and internal organs showing distinctly, intestines imparting 
a greenish tinge to the sides; throat and breast pure white in d 71 with a 
certain amount of green or greenish blue periphally. 





9 


d" 


Measurements 


M.C.Z. 20,025 


M.C.Z. 20,026 


Length from snout to anus 


20 mm. 


21.5 mm. 


Breadth of head 


7 mm. 


7.25 mm. 


Length of head 


7 mm. 


7.25 mm. 


Length of hind limb 


30 mm. 


31 mm. 


Length of fourth toe 


6 mm. 


G.5 mm. 



Bulletin of the Museum of Comparative Zoology 
AT HARVARD COLLEGE 

Vol. LXXIX, No. 2 



SCIENTIFIC RESULTS OF AN EXPEDITION TO RAIN 
FOREST REGIONS IN EASTERN AFRICA 

II 
CRUSTACEA 



By Mary J. Rathbun 



With Two Plates 



CAMBRIDGE, MASS., U. S. A. 

PRINTED FOR THE MUSEUM 

July, 1935 



No. 2. — Scientific Results of an Expedition to Rain Forest Regions 

in Eastern Africa 

II 

Crustacea 
By Mary J. Rathbun 

The expedition produced two new species of Potamon, both of the 
Geothelphusa type. Also a specimen of Palaemon delagoae Stebbing 
previously known only from the type locality at the southern end of 
Portuguese East Africa. 

Brachyura 
POTAMONIDAE 

Potamon (Geothelphusa) harvardi spec. nov. 

Plate 1 

Type cT (M. C. Z. 8241) Sipi, W. Mt. Elgon, Uganda, 6000 ft. 18.XII. 33. 

Figured c? (M. C. Z. 8242), same data. 

19c? 6 9 (M.C.Z. 8243), do. 

1 cM 9 ovigerous (M.C.Z. 8240) Kaburomi, Mt. Elgon, Uganda. 28.XII. 

33. 
34 cf 28 9 6 juv. (M.C.Z. 8244, 8245) Kaimosi, Kakamega, Kenya 

Colony. 10-15. 11.34. 

This species belongs to the group of P. ((?.) granviki Colosi 1 and 
P. (G.) anthcus Colosi 2 . Carapace covered with coarse punctae. 
Antero-lateral margin narrowly but smoothly indicated; postero- 
lateral margin roughened with short, oblique rugae. Carapace deeply 
furrowed about the cardiac and posterior mesogastric region; a deep 
sulcus runs from the posterior cardiac region along the posterior 
margin of the branchial region. Cervical groove shallow, incomplete, 
ending forward in a row of coarse punctae directed toward the post- 
orbital pit. Epigastric lobes prominent and widely separated; tip of 

1 1924, Arkiv for Zoologi, Stockholm, 16 (1) p. 16, text fig. 11, pi. 1, fig. 5. 

2 1920, Boll. Mus. Zool. ed. Anat. comp., 35, p. 35; 1924, Arkiv for Zoologi, Stockholm, 
16 (l),p. 17, text fig. 12. 



24 bulletin: museum of comparative zoology 

nasogastric region roof -shaped. A faint, blunt ridge behind and 
adjacent to the orbit; a deep, narrow groove between the ridge and the 
orbital rim; the ridge fades out just beyond the outer angle of the orbit; 
in the triangle thus formed there is a well marked pit (pi. 1, fig. 4). 
The front, measured across the middle of its depth is H as wide as the 
carapace; its margin is broadly shallow at middle and its sides are 
distinctly oblique. Chelae similar to those of granviki (fig. lie) but 
narrower and more elongate. Merus of outer maxillipeds broader than 
long, the outer anterior angle prominently rounded. Groove on 
ischium nearest inner margin, very distinct but short. Anterior end 
of sternum akin to that of antheus 1 ; space in front of groove 5 times as 
wide as long, not counting the narrow point between maxillipeds. 
Groove at base of chelipeds longer than in granviki, about \]/2 times 
as long as the distance from the groove to the median line; in antheus 
the groove extends across the sternum. Distal end of sixth segment of 
male abdomen V/i times as long, proximal end twice as wide as long. 

Measurements. 

Length of carapace of male 30 mm. 
Width of carapace 42.8 
Width of front below 13.3 
Width of front above 16 
Length of major propodus of cheliped 46 
Length of major palm at middle 30 
Width of major palm at highest point 19 
Length of minor propodus 33.2 
Length of minor palm at middle 18.7 
Width of minor palm at highest point 1 1 . 



Potamon (Geothelphusa) perparvus Rathbun 

Potamon {Geothelphusa) perparvus Rathbun, 1921, Bull. Amer. Mus. Nat. 
Hist., 43, p. 425, pi. xxviii, fig. 2; pi. xxx; text fig. 12: Stanleyville, Belgian 
Congo. 

1 c? 2 9 (1 with young) (M.C.Z. 8239) Kaimosi, Kakamega, Kenya Colony. 
10-15.11.34. 

1 1920, Boll. Mus. Zool. ed. Anat. comp., 35, p. 35; 1924, Arkiv for Zoologi, Stockholm, 16 (1), 
p. 17, text fig. 12 



RATHBUN: AFRICAN CRABS 25 

POTAMON (GEOTHELPHUSA) AM.ALERENSIS Spec. nov. 

Plate 2 

5 c? 5 9 1 juv. (M.C.Z. 8237 holotype, 8238) Amaler River, Mt. Debasien, 
Uganda. 5000 ft. IX.33. 

Near P. (G.) bcrardi. Carapace wider. Epigastric lobes oblique, 
separated feebly from the postorbital ridge which is sinuous, being 
more advanced behind the inner half of the orbital rim; the ridge unites 
with the lateral border of carapace at an obtuse angle; it is serrate at 
the outer end and is continued downward and inward with a round 
turn to the outer margin of the orbit. Frontal margin divided into 
two shallow lobes, sides oblique. Lateral margin of carapace finely 
serrate. Mesogastric region with anterior end roof-shaped; behind 
this region two deep triangular depressions; short deep furrows either 
side of cardiac region converge slightly backward. Sides of carapace 
with rows of short oblique striae. Chelipeds of male unequal, minor 
palm about % as high as major; the major manus is stout, convex 
below; its fingers gape narrowly and bear uneven teeth. Merus of 
outer maxilliped similar to that of bcrardi. 

Measuremetits. 

Length of carapace of male holotype 17.4 mm. 

Width of carapace 25.5 

Length of carapace of female 20.5 

Width of carapace 31. 

Potamon (Geothelphusa) berardi (Audouin) 

Thelphusa berardi Audouin, 1826, Expl. Soram. Plates by Savigny, in Desc. 

de l'Egypte. Hist. Nat., 1, pt. 4, p. 82 (pi. 2, fig. 6 of Savigny). 
Potamon {Geothelphusa) berardi, Rathbun, 1905, Nouv. Arch. Mus. Hist. Nat., 

Paris, 7, p. 203; 6, pi. xviii, fig. 3 and 10, and synonymy. 

260* 17$ (M.C.Z. 8235) Butandiga, Mt. Elgon, Uganda. 8.1.34. 
40c? 42 9 (13 ovig.) (M.C.Z. 8236), Elgonyi, Mt. Elgon, Kenya Colony. 
7000 ft. 25.1.34. 

Potamon (Acanthothelphusa) niloticus (Milne Edwards) 

Thelpheusa nilotica H. Milne Edwards, 1837, Hist. Nat. Crust., 2, p. 12: The 

Nile. 



26 bulletin: museum of comparative zoology 

Potamon (Parathelphusa) niloticus Rathbun, 1905, Nouv. Arch. Mus. Hist. 

Nat., Paris, 7, p. 263, pi. xii, fig. 15. 
Potamon (Acanthothelphusa) niloticus Rathbun, 1933, Bull. Mus. Comp. Zool., 

Cambridge, 75, p. 258 and synonymy. 

1 9 ovig. (M.C.Z. 8234) Kaimosi, Kenya Colony. 111.34. 

Potamon (Potamonautes) didieri Rathbun 

Potamon (Potamonautes) didieri Rathbun, 1905, Nouv. Arch. Mus. Hist. Nat., 
Paris, 7, p. 170; 6, pi. xiv, fig. 9: Belgian Congo. 

4 c? 1 9 2 juv. (M.C.Z. 8226) Butandiga, Mt. Elgon, Uganda. 8.1.34. 
3 c? 5 9 (M.C.Z. 8227) Elgonyi, Mt. Elgon, Kenya Colony. 7000 ft. 
25.1.34. 

Potamon (Potamonautes) hilgendorfi (Pfeffer) 

Telphusa suprasulcata Hilgendorf, 1898, Deutsch-Ost-Afrika, 4, p. 8, pi., fig. 

5-5d: On the way to Kilimanjaro. 
Potamon (Potamonautes) hilgendorfi Rathbun, 1933, Bull. Mus. Comp. Zool., 

Cambridge, 75, p. 256 and synonymy. 

2 c? (M.C.Z. 8228) Mt. Mbololo, Taita, Kenya Colony. IV.34. 

1 juv. (M.C.Z. 8229) Amaler River, Mt. Debasien, Uganda. 5000 ft. 
IX.33. 

Potamon (Potamonautes) bottegoi de Man 

Potamon (Potamonautes) bottegoi de Man, 1898, Ann. Mus. Civ. Genova (2) 19, 
p. 262 [3], pi. hi; Rathbun, 1933, Bull. Mus. Comp. Zool., Cambridge, 
75, p. 258. 

1 9 (M.C.Z. 8230) Amaler River, Mt. Debasien, Uganda. 5000 ft. XI.33. 

1 c? (M.C.Z. 8231) Mombosasa, near Witu, Kenya Colony. V.34. 

3c? 3 9 (M.C.Z. 8232) Voi, Kenya Colony. 7.1 V.34. 

1 9 (M.C.Z. 8233) opposite Kilindini, Kenya Colony. 6. VI 1. 34. 

OCYPODLDAE 
Ocypode kuhlii de Haan 

Ocypode (Ocypode) kuhlii de Haan, 1835, Fauna Japon., Crust., Dec. 2, p. 58. 
Ocypode kuhlii, Rathbun, 1933, Bull. Mus. Comp. Zool., 75, p. 260, pi. vii. 

lc? (M.C.Z. 8248) Lamu Island, Kenya Colony. 7-12.IV.34. 
lc? (U.S.N. M.) Kitau, Manda Island, Kenya Colony. V.34. 
2c? (M.C.Z. 8249) Malindi, Kenya Colony. VI.34. 



RATHBUN: AFRICAN CRABS 27 

Uca inversa (Hoffmann) 

Gelasimus inversus Hoffman, 1874, Crust. Echinod. Madagasc, p. 19, pi. iv, 
fig. 23-26; De Man, 1891, Notes Leyden Mus., 13, p. 44. pi. iv, fig. 12. 

Id 1 (M.C.Z. 8251) Gongoni, Kenya Colony. 27.IV.34. 

GECARCINIDAE 

Cardisoma carnifex (Herbst) 

Cancer carnifex Herbst, 1796, Naturg. Krabben und Krebse, 2, p. 163, pi. xli, 

fig. 1: "Trankenbar." 
Cardisoma carnifex, Latreille, 1825, Encycl. Method., 10, p. 685. 

1 cf 2 9 (M.C.Z. 8252) Kitau, Manda Island. V.34. 

Anomura 
COENOBITIDAE 

Coenobita rugosus Milne Edwards 

Cenobita rugosa Milne Edwards, 1837, Hist. Nat. Crust., 2, p. 241: Indian 

Ocean. 
Coenobita rugosus, Alcock, 1905, Catal. Indian Dec. Crust., Part II. Anomura, 

Fasc. 1. Pagurides, p. 143, pi. xiv, fig. 3, 3a, and synonymy. 

1 aM 9 (M.C.Z. 8254) Opposite Kilindini. 6.VII.34. 

Coenobita rugosus var. jousseaumi Bouvier 

t 

Coenobita rugosa var. jousseaumi Bouvier, 1890, Bull. Soc. Philom., Paris (8) 
2, p. 146: Aden. 

Id* (M.C.Z. 8253) Lamu, Lamu Island, Kenya Colony. V.34. 

Macrura 

PALAEMONIDAE 

Macrobrachium patsa (Coutiere) 

Palaemon (Parapalaemon) patsa Coutiere, 1899, Bull. Mus. Hist. Nat. Paris, 7, 
p. 383: Madagascar. 1901, Ann. Sci. Nat., Zool., 12, p. 284, pi. 11, fig. 
xx-xxii. 

1 9 (M.C.Z. 8255) Tsavo, Kenya Colony. 2-4. IV.34. 



28 



bulletin: museum of comparative zoology 



Palaemon (Parapalaemon) dolichodactylus Hilgendorf 

Palaemon (s.s.) dolichodactylus Hilgendorf, 1878, Monatsb. d. K. Akad. Wiss. 
Berlin, p. 840, pi. iv, fig. 18: Mozambique. 

2 9 (M.C.Z. 8256) Ngatana, Tana River, Kenya Colony. VI.34. 



Palaemon delagoae Stebbing 

Text figures 1 and 2 

Palaemon delagoae Stebbing, 1912, Ann. South Afr. Mus., 15, p. 74 pi., 80: 
Delagoa Bay, Portuguese East Africa. 

1 9 (M.C.Z. 8257) Ngatana, Tana River, Kenya Colony. VI.34. 




Palaemon delagoae 
Fig. 1. Carapace and appendages (cheliped lacking), lateral view, x 2. 




Fig. 2. Three segments of abdomen and appendages, dorsal view, x 2. 

The specimen is soft shell and lacks chelipeds. It is larger than the 
type, the carapace measuring 43 mm. long and telson 14. 



EXPLANATION OF PLATES 



PLATE 1 



Rathbun — Crabs 



Plate 1 

Potamon (Geothelphusa) harvardi d" type 

Fig. 1. Dorsal view, nat. size. 

Fig. 2. Ventral view, showing chelae, nat. size. 

Fig. 4. Frontal view, showing front, eyes and anterior end of maxillipeds, 

x iy 2 . 

Fig. 5. Ventral view, showing front part of sternum, x 2. 

Fig. 6. Ventral view, for abdomen, x 1%. 

Fig. 3. Ventral view of smaller specimen, d 1 (M.C.Z. 8242) nat. size. 



BULL. MUS. COMP. ZOOL. 



Rathbun. African Crabs. Plate 1 




PLATE 2 



Rathbun — Crabs 



Plate 2 

Potamon (Geothelphusa) amalerensis <? holotype 

Fig. 1. Dorsal view xlj^. 

Fig. 2. Ventral view, showing chelipeds, xl)4- 

Fig. 3. Frontal view for eyes, groove behind eyes and anterior end of maxilli- 

peds, x 2. 
Fig. 4. Ventral view for front part of sternum, x 3. 
Fig. 5. Ventral view for abdomen, x 2. 



BULL. MUS. COMP. ZOOL. 



Rathbun. African Crabs. Plate 2 





he 



Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 

Vol. LXXIX, No. 3 



SCIENTIFIC RESULTS OF AN EXPEDITION 
TO RAIN FOREST REGIONS IN EASTERN AFRICA 

III 

MAMMALS 



By Glover M. Allen and Barbara Lawrence 

with field notes by 
Arthur Loveridge 



With Five Plates 



CAMBRIDGE, MASS., U. S. A. 

PRINTED FOR THE MUSEUM 
January, 1936 



No. 3. — Reports on the Scientific Results of an Expedition to 
Rain Forest Regions in Eastern Africa 

III 

Mammals 

By Glover M. Allen and Barbara Lawrence 

WITH FIELD NOTES BY ARTHUR LoVERIDGE 



CONTENTS p 

Page 

Introduction .■> 31 

List of species collected 34 

Systematic discussion 39 

Bibliography 126 



INTRODUCTION 

The collection dealt with in this report, was made by Mr. Arthur 
Loveridge while investigating distributional problems associated with 
rain forest areas in Uganda and Kenya. The enquiry was carried 
out on behalf of the Museum of Comparative Zoology with a fellow- 
ship granted by the John Simon Guggenheim Memorial Foundation 
of New York. 

The authors have collaborated in the identification and taxonomic 
work recorded under the headings of Distribution, Discussion and 
Coloration. The field notes contributed by the collector, are listed 
in the first person singular under Measurements, Breeding, Diet, 
Enemies, Habits, Folklore, Native names, etc. 

Altitudes, and other information regarding the localities in which 
collecting was carried on, will be found in the final paper of this series 
of reports, which will treat of the whole vertebrate terrestrial fauna 
of Mounts Debasien and Elgon in relation to that of the Usambara 
Mountains in Tanganyika Territory. 

The period of collecting mammals was from November 9, 1933, to 
June 30, 1934, during which time 1,024 skins and skulls representing 
133 species or races of mammals were secured. Of these 64 forms were 
new to the collections of the Museum of Comparative Zoology. A 
special feature of the collection was the topotypes, frequently in 



32 bulletin: museum of comparative zoology 

series, of 36 species apart from many others which were almost topo- 
typic. It has been found necessary to describe only one bat and a 
pouched rat as new, namely 

Nycteris nana tristis subsp. nov. 9 from Kaimosi, Kenya Colony. 
Saccostomus cricetulus sp. nov. cf 9 from Greeki River, Uganda. 

Attention, however, may be directed to such rarities as Petrodomus 
s. sangi, Nycteris aurita, Rhinolophus f. exsul, Perodicticus p. ibeanus, 
Cercopithecus neglectus, Cercocebus g. galeritus, Colobus b. rufomitratus, 
Zelotomys h. vinaceus, Atherurus turneri, Dendrohyrax a. bettoni and 
many others of which good series were obtained. 

When measurements are given serially they are always in the 
following order: — (1) length from snout to anus; (2) length of the 
tail without terminal hairs; (3) length of hind foot without claws; 
(4) length of ear from tip to notch. In the case of bats a fifth measure- 
ment is added: (5) length of wing from axilla to tip. All dimensions 
are in millimetres, and it is those of the largest male and largest female 
of the series which are supplied. 

We take this opportunity of expressing our thanks to Mr. Gerritt 
S. Miller Jr., and Dr. Remington Kellogg for loaning material 
from the National Collection, and Mr. J. K. Doutt of the Carnegie 
Museum; Mons. L. Chopard for identifying the hemimerid parasites 
and our colleagues, Drs. Joseph Bequaert and J. H. Sandground 
of the Harvard School of Tropical Medicine for their kindness in 
identifying the many parasites enumerated in the following pages. 
All the photographs, where not otherwise stated, were taken by 
Mrs. Loveridge. 

Special interest attaches to those species of mammals characteristic 
of heavy forest growth in which this collection is particularly rich. 
Their presence at various localities here marks the eastward or north- 
eastward limit of their range and is correlated with relict patches of 
forest, often isolated on mountains or in river valleys. No doubt 
these areas were formerly more extensive and supported a fauna 
whose species ranged more or less continuously across the continent 
though frequently breaking up into local races. Increasing aridity 
and the effect of human occupation, clearing and burning, has re- 
sulted in driving back the eastward outposts of the rain forest with 
a corresponding restriction in the numbers and ranges of these 
species. 

The following are rather characteristic of this fauna : — Sylvisorex 
gemmeus and S. mundus, long-tailed shrews; Rouscttus angolensis, a 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 33 

rousette bat here at about its eastern limit ; Nycteris nana tristis, the 
dwarf hollow-faced bat, and probably one or two others of the genus 
although some are more characteristic of slightly drier country ;Perodic- 
ticus potto ibeanus, the eastern potto, for which Kaimosi seems to be 
about the most eastern station; Cercocebus g. galcritus, the crested 
mangabey, of which an outpost colony on the lower reaches of the 
Tana River is the sole representative of this genus in Kenya ; Cercop- 
ithccus nictitans schmidti, an eastern representative of the white- 
nosed monkey, most of the other races of which are West African; 
Cercopithecus mitis kibonotensis and C. m. stuhlmanni, well-marked 
races of the blue monkey, the former found in the coastal forests, 
the latter in those of the Elgon region and adjacent areas westward; 
Colobns polykomos matschiei of the Elgon region, a close relative of 
the race of the central Kenya forests; C. badius rufomitratus, the red- 
capped colobus, of special interest since this type is apparently not 
found in Kenya except for this outpost in the region of the lower Tana 
River, whilst its nearest ally is probably the race gordonorum, a rare 
animal of south central Tanganyika ; Genetta servalina bettoni, a close- 
spotted genet found in the Elgon and Kaimosi forests which is clearly 
an eastern representative of 6?. servalina of West Africa; Nandinia 
binotata arborca, the tree civet, a very slightly marked form of the 
type common in the western forests; Anomalurus jacksoni, a gray 
species of scaly-tailed flying squirrel, confined to the heavy forests 
of Uganda eastward to Kaimosi in Kenya Colony; Heliosciurus rufo- 
brachium ny ansae, an outpost subspecies of a squirrel common in 
West African forests and meeting the range of H. multicolor elegans an 
eastern tree squirrel, on Mount Elgon; //. undulatus shindi, a related 
species of which this race is restricted to forested mountain tops in 
the Taita Hills near the coast; Protoxerus stangeri bea, the Kenya 
giant squirrel, again an eastern race of a forest squirrel common 
in parts of West Africa but only found in Kenya in the Kakamega 
forests near Kaimosi ; Claviglis saturatus, a forest-living dormouse ; 
Dendromus ruddi, recalling the unstriped D. messorius of the Came- 
roons; Oenomys bacchante editus, the rufous-nosed mouse, evidently 
allied to West African races of 0. hypoxanthus and represented on 
Mount Kenya by a similar subspecies. Lophuromys sikapusi ansor- 
gei, the pink-bellied mouse, is apparently here near the eastern 
limits of its range and giving place to the eastern species L. a. aquilus; 
Atherurus tumeri, the brush-tailed porcupine, allied to A. africanus, 
perhaps its eastern representative, reaches at Kaimosi, the north- 
eastern limit of the group's range in Africa; Hylochoerus meinertz- 



34 bulletin: museum of comparative zoology 

hageni, the forest pig, found in other forested areas of Kenya 
Colony as well. 

To this list doubtless others will be added such as Colomys, lately 
discovered at Elburgon, Kenya, but several genera such as Stochomys, 
Malacomys, Deomys still only known from West Africa, appear to 
be absent as are the galagos, Galago elegantulus and G. demidovvi. 
No doubt these are confined to the western portions of the continent. 
Additional interest is furnished by those areas in Kenya Colony 
where the western forest fauna meets with that of the steppe and 
thornbush districts of the east. 



LIST OF SPECIES COLLECTED* 

ERINACEIDAE Page 

Atelerix pruneri hindei (Thomas) 39 

MACROSCELIDIDAE 

Petrodromus (Cercoctenus) sultan sangi Heller 39 

Nasilio brachyrhynchus delamerei (Thomas) 40 

Elephantulus rufescens rufescens (Peters) 40 

SORICIDAE 

Sylvisorex gemmeus Heller 41 

Sylvisorex mundus Osgood 41 

Crocidura ny ansae ny ansae Neumann 41 

Crocidura hindei Thomas 42 

Crocidura turba zaodon Osgood 42 

Crocidura jacksoni jacksoni Thomas 43 

Crocidura hildegardeae hildegardeae Thomas 43 

Crocidura bicolor elgonius Osgood 44 

PTEROPIDAE 

Rousetttis (Lissonycteris) angolensis (Bocage) .' 44 

Rousettus leachi (Smith) 45 

Rousettus lanosus kempi Thomas 45 

Epomophorus icahlbergi wahlbergi (Sundevall) 45 

Eponwphorus labiatus minor Dobson 46 

EMBALLONURIDAE 

Taphozous perforatum haedinus Thomas 46 

♦Species in parenthesis were not collected but are discussed. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 35 

NYCTERIDAE Page 

Nycteris nana tristis subsp. nov 47 

Nycteris hispida (Schreber) 48 

Nycteris aurita (Andersen) 48 

Nycteris damarensis brockmani (Andersen) 49 

Nycteris thebaica rcvoili Robin 49 

MEGADERMIDAE 

Laviafrons rex Miller 50 

Cardioderma cor (Peters) 50 

RHIXOLOPHIDAE 

Rhinolophus hildcbrandtii Peters 51 

Rhinolophus eloquens Andersen 51 

Rhinolophus fumigatus exsul Andersen 52 

HIPPOSIDERIDAE 

Hipposideros coffer (Sundevall) 52 

Hipposideros ruber (Noack) 53 

VESPERTILIONIDAE 

Pipistrellus nanus (Peters) 53 

Glauconycteris argentata (Dobson) 53 

MOLOSSIDAE 

Mops (Allomops) osborni Allen 54 

Chaerephon hindei (Thomas) 55 

CANIDAE 

Thos mesomelas mcmillani Heller 55 

{Lycaon pictus lupinus Thomas) 55 

MUSTELIDAE 

Mellivora cape?isis sagulata Hollister 57 

Aonyx capensis hindei (Thomas) 58 

VIVERRIDAE 

Civettictis civetta schwarzi Cabrera 58 

Genetta servalina bettoni Thomas 60 

Genetta stuhlmanni stuhlmanni Matschie 60 

Genetta stuhlmanni erlangeri Matschie 61 

Nandinia binotata arbor ea Heller 62 

Galcrella sanguined ibeae (Wroughton) 62 

Herpestes ichneumon funestus (Osgood) 63 



36 bulletin: museum of comparative zoology 

Page 

Atilax paludinosus robustus (Gray) 63 

Ichneumia albicauda ibeana (Thomas) 64 

Helogale undulata rufula Thomas 65 

Mungos mungo colonus (Heller) 65 

FEL1DAE 

Felis (Leptailurus) capensis hindei Wroughton 65 

Felis ocreata nandae Heller 66 

LORISIDAE 

Perodicticus potto ibeanus Thomas 67 

GALAGIDAE 

Galago crassicaudatus lasiotis Peters 68 

Galago senegalensis albipes Dollman 69 

Galago senegalensis braccatus Elliot 69 

CERCOPITHECIDAE 

Cercopithecus nictitans schmidti Matschie 69 

Cercopithecus aethiops joknstoni Pocock 71 

Cercopithecus aethiops callidus (Hollister) 71 

Cercopithecus mitis kibonotensis Lonnberg 72 

Cercopithecus mitis stuhlmanni Matschie 73 

Cercopithecus neglectus Schlegel 74 

Cercocebus galeritus galeritus Peters 75 

Papio furax Elliot 75 

Papio ibeanus Thomas 78 

PITHECIDAE 

Colobus polykomos matschiei Neumann 78 

Colobus badius rufomitratus Peters 79 

ANOMALURIDAE 

Anomalurus jacksoni de Winton 80 

SCIURIDAE 

Heliosciurus rufobrachium ny ansae (Neumann) 81 

Heliosciurus undulatus shindi Heller 82 

Heliosciurus multicolor elegans Thomas 82 

Protoxerus stangeri bea Heller 83 

MYOXIDAE 

Claviglis parvus parvus (True) 83 

Claviglis saturatus (Dollman) 84 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 37 

CRICETIDAE Page 

Dipodillus pusillus (Peters) 86 

Tatera vicina vidua (Peters) 87 

Tatera nigricauda nigricauda (Peters) 87 

Tatera nigrita Wroughton 87 

RHIZOMYIDAE 

Tachyoryctes ruddi Thomas 88 

MURIDAE 

Dendromus insignia insignis Thomas 89 

Dendromus whytei pallescens Osgood 89 

Dendromus ruddi Wroughton 89 

Dendromus acraeus Wroughton 90 

Zelotomys hildegardeae vinaceus Heller 90 

Thamnomys surdaster polionops Osgood 91 

Thamnomys surdaster elgonis Thomas 91 

Oenomys bacchante ediius Thomas & Wroughton 92 

Rattus rattus kijabius (Allen) 93 

Aethomys kaiseri medicatus (Wroughton) 94 

Praomys tullbergi jacksoni (de Winton) 94 

Praomys taitae (Heller) 95 

Mastomys coucha tinctus (Hollister) 95 

M ostomy s coucha hildebrandtii (Peters) 96 

Leggada triton triton Thomas 96 

Leggada bella bella Thomas 97 

Leggada bella vicina Thomas 97 

Leggada grata grata Thomas 97 

Cricetom.ys gambianus elgonis Thomas 98 

Lophuromys aquilus aquilus (True) 99 

Lophuromys sikapusi ansorgei de Winton 100 

Saccostomus cricetulus sp. nov 100 

Acomys ignitus ignitus Dollman 102 

Acomys wilsoni wilsoni Thomas 103 

Dasymys helukus helukus Heller 103 

Pelomys fallax iridescens Heller 104 

Arvicanthis abyssinicus nubilans Wroughton 104 

Arvicanthis abyssinicus virescens Heller 104 

Lemniscomys griselda maculosus (Osgood) 105 

Lemniscomys striatus massaicus (Pagenstecher)) 105 

Rhabdomys pumilio diminutus (Thomas) 105 

Otomys tropicalis elgonis Wroughton 106 

Otomys angoniensis elassodon Osgood 106 



38 bulletin: museum of comparative zoology 

HYSTRICIDAE Page 

Hystrix galeata Thomas 107 

Atherurus turneri St. Leger 107 

THRYONOMYIDAE 

Choeromys grcgorianus (Thomas) 108 

LEPORIDAE 

Lepus victoriae kakumegae Heller 109 

SUIDAE 

(Hylochoerus meinertzhageni meincrtzhageni Thomas) 109 

BOVIDAE 

Damaliscus korrigum topi Blaine 110 

Cephalophus monticola musculoides Heller 110 

Sylvicapra grimmia deserti Heller Ill 

Sylvicapra grimmia lobeliarum Lonnberg 112 

Sylvicapra grimmia nyanzae Neumann 112 

Ourebia montana cottoni Thomas & Wroughton 113 

Raphiceros campestris neumanni (Matschie) 113 

Rhynchotragus kirkii kirkii (Gunther) 113 

Rhynchotragus kirkii nyikae Heller 114 

Kobus ellipslprymnus kuru Heller 115 

Kobus defassa ugandae Neumann 115 

Tragelaphus scriptus delamerei Pocock 116 

Tragelaphus scriptus massaicus Neumann 117 

Tragelaphus scriptus olivaceus Heller 117 

ELEPHANTIDAE 

(Loxodonta africana peeli (Lydekker)) 118 

PROCAVIIDAE 

Procaeia habessinica daemon Thomas 119 

Heterohyrax syriacus kempi (Thomas) 120 

Hetcrohyrax syriacus hindei (Wroughton) 121 

Dendrohyrax arbor eus stuhlmanni (Matschie) 123 

Dendrohyrax arborcus bettoni (Thomas & Wroughton) 123 

DELPHINIDAE 

Prodelphinus attenuatus (Gray) 124 

DUGONGIDAE 

{Dugong dugon (P. S. L. Miiller)) 125 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 39 

Systematic Discussion 

ERINACEIDAE 

Atelerix pruneri hindei (Thomas) 

Erinaceus hindei Thomas, 1910, Ann. Mag. Nat. Hist. (8), 5, p. 193: Kitui, 
Kenya Colony. 

2 <? 2 9 (M. C. Z. 31796-9) Voi, K. C. 18.iv.34. 

Native name. Kisesegcdi (Kitaita). 

Discussion. In 1922, J. A. Allen discussed at length the status 
of the names albiventris and pruneri, reaching the conclusion that 
the former was unidentifiable, and that the latter, synchronously 
published, should replace it for the typical race from Senaar. This 
decision rests largely on the fact that no locality can be assigned to 
the type of albiventris, "although the type appears to have been pre- 
served in the Munich Museum." Through the courtesy of the Direc- 
tor of the latter institution, Dr. Lorenz Muller, this specimen has 
been loaned for examination and proves to be the Senegalese species, 
which is distinct from A. pruneri. 

Coloration. This series illustrates what is probably the effect of 
age, in that the younger two have the spine tips clear white, con- 
trasting with the blackish bases, with a short buffy or pale-ochraceous 
area forming a transitional band between. In the older specimens 
many of the whitish tips become discolored buffy, and the black 
bases of a less intense browner shade. What seem to be new spines 
with fresh white tips are apparently coming in here and there. 

Measurements. cT. 180. 21. 24. 27 mm., 9 . 165. 18. 22. 24 mm. 

Parasites. The largest male was infested with a score or more of 
large ticks (Rhipicephalus armatus). These were not confined to the 
spinous regions but occurred on the belly where it might have been 
supposed that an insectivorous animal like a hedgehog would have 
attacked them. 

MACROSCELIDIDAE 

Petrodromus (Cercoctenus) sultan sangi Heller 

Petrodromus sultani sangi Heller, 1912, Smithsonian Misc. Coll., 60, No. 12, 
p. 12: Mount Mbololo, Kenya Colony. 

9 (M. C. Z. 31795) Mt. Mbololo, K. C. 25. iv. 34. 

Native name. Mwonunguomballa (Kitaita). 



40 bulletin: museum of comparative zoology 

Discussion. The specimen is a topotype and apparently the second 
to be recorded. In color it agrees closely with the typical form, sultan, 
but Hollister, however, believed the race separable on the basis of a 
narrower rostrum and smaller upper premolars. The latter character is 
borne out in comparison with a specimen representing typical sultan, 
from Amani, Tanganyika Territory. 

Measurements. 9 ■ 205. 184. 56. 36 mm. 



Nasilio brachyrhynchus delamerei (Thomas) 

Macroscelides delamerei Thomas, 1901, Ann. Mag. Nat. Hist. (7), 8, p. 155: 
Athi River, Kenya Colony. 

9 (M. C. Z. 31807) Voi, K. C. 7. iv. 34. 

Native name. Mwonungu (Kitaita). 

Discussion. This specimen is immature, having the complete milk 
dentition only, so that the diagnostic third lower molar of the per- 
manent set is not yet present. 

Coloration. The narrow eye ring, interrupted at the front, and 
not extended behind as a broad white mark to the base of the ear, 
as well as the richer chestnut of the back seem to confirm its refer 
ence to this genus rather than to Elephantulus, which was secured 
at the same locality. It is peculiar in having the backs of the hind 
feet buffy, instead of clear whitish as in specimens from farther 
inland. 

Measurements. 9 . 97. 80. 27 mm. Ear eaten by ants while the 
animal was lying dead in a snap-back rat trap. 



Elephantulus rufescens rufescens (Peters) 

Macroscelides rufescens Peters, 1878, Monatsb. Akad. Wiss. Berlin, p. 198: 
Ndi, Taita, Kenya Colony. 

1 o^ 2 9 (M. C. Z. 31803-5) Voi, K. C. 9-10. iv. 34. 

4 9 (M. C. Z. 31800-2, 31806) Mt. Mbololo, K. C. 24. iv. 34. 

Distribution. The series from Mount Mbololo, Taita Hills, are 
practically topotypes. 

Native name. Mwonungu (Kitaita). 

Measurements. <?. 115. 166. 29. 19 mm., 9 . 138. 125. 31. 20 mm. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 41 

SORICIDAE 

Sylvisorex gemmeus Heller 

Sylvisorex gemmeus Heller, 1910, Smithsonian Misc. Coll., 56, No. 15, p. 7: 
Rhino Camp, Lado Enclave. 

1 (M. C. Z. 31242) Kirui, K. C. 6. ii. 34. 
3 cM 9 (M. C. Z. 31239-40, 31257, 31259) Kaimosi, K. C. 16-20. 
ii. 34. 

Native names. Lunihi (Luragoli); luhui (Lutereki) for all shrews. 

Coloration. These five specimens are uniformly dark chocolate 
brown above, with white-tipped underparts. This color and the # long 
tail, exceeding the head and body, distinguish it readily. Hollister 
(1918, p. 39) has recorded a large series from Kaimosi and adds a 
table of measurements. 

Measurements, cf . 75. 87. 14. 7 mm., 9 . 65. 75. 14. 7 mm. 

Sylvisorex mundtjs Osgood 

Sylvisorex mundus Osgood, 1910, Publ. Field Mus. Nat. Hist., Zool. Series, 
10, p. 18: Kijabe, Kenya Colony. 

& 9 (M. C. Z. 31243-4) Butandiga, U. 14. i. 34. 

Distribution. In addition to the original locality (Kijabe), speci- 
mens are recorded by Hollister (1918, p. 39) from the west side of 
Mount Kenya at altitudes of from 7,000 to 10,000 feet. This pair 
from the western slopes of Mount Elgon (7,000 feet) apparently 
constitute the first record north and west of Mount Kenya. 

Native name. Namageba (Lugishu). 

Discussion. Osgood in his original description indicates its close 
relationship with S. granti of Mount Ruwenzori, in which, however, 
the tail is relatively longer. In S. mundus it is apparently shorter 
than head and body. The uniformly blackish-brown coloration above 
and below further distinguishes it at once from the pale-bellied 
S. gemmeus. 

Measurements, cf. 64. 57. 12. 7 mm., 9 . 64. 58. 12. 7 mm. 

Crocidura. nyansae nyansae Neumann 

Crocidura flavescens nyansae Neumann, 1900, Zool. Jahrb. Syst., 13, p. 544: 
Fort Lubwa, Usoga, Uganda. 

9 (M. C. Z. 31266) Sipi, U. 19. xii. 33. 
2 9 (M. C. Z. 31263-4) Kaimosi, K. C. 20. ii & 9. hi. 34. 



42 bulletin: museum of comparative zoology 

Distribution. Hollister (1918, p. 42) has previously recorded this 
large brown shrew from Kaimosi. 

Coloration. One of the specimens from Sipi is just beginning to 
acquire the new pelage, which appears on the forehead in a patch 
extending from the nose to between the ears, and again as an oval 
area in the centre of the back behind the shoulders. The new fur 
is much darker, more nearly seal brown than the rest, which has 
faded to a dull brown. 

Measurements. 9 . 160. 175. 35. 20 mm. 

Parasites. Fleas were removed from the fur of a Kaimosi shrew. 

Enemies. One was recovered, and preserved in alcohol, from the 
stomach of a Nose-horned Viper (Bitis nasicornis) at Kaimosi. 

Crocidura hindei Thomas 

Crocidura hindei Thomas, 1904, Ann. Mag. Nat. Hist. (7), 14, p. 237: Mach- 
akos, Kenya Colony. 

9 (M. C. Z. 31785) Ngatana, K. C. 18. vi. 34. 

Native name. Tungu (Kipokomo). 

Discussion. This shrew is in a short gray pelage, apparently im- 
mature. The shorter tail as compared with C. suahelae and the flat- 
tened skull with a total length of about 24 mm. distinguish it. 

Measurements. 9 . 90. 44. 13. 6 mm. 

Habitat. I captured this gray shrew beneath a rubbish heap in one 
of the native gardens. 

Crocidura turba zaodon Osgood 

Crocidura turba zaodon Osgood, 1910, Publ. Field Mus. Nat. Hist., Zool. 
Series, 10, No. 3, p. 21 : Nairobi, Kenya Colony. 

9 (M. C. Z. 31268) Greeki River, U. 6. xii. 33. 

3 9 (M. C. Z. 31254-6) Sipi, U. 20-22. xii. 33. 

2 cf 4 9 (M. C. Z. 31245-50) Butandiga, U. 8-14. i. 34. 

2 cf 2 9 (M. C. Z. 31251-3, 31265) Kaimosi, K. C. 9-26. ii. 34. 

Distribution. This seems to be the common shrew in northern 
Kenya Colony. The series of thirteen skins is largely from Mount 
Elgon, and so represents Crocidura turba kempi Dollman (type locality, 
Kirui, Mount Elgon), but this, as Hollister (1918, p. 55) first sug- 
gested, is clearly the same as C. t. zaodon of the Nairobi region. 

Native names. Etutwi (Karamojong) ; guchuru (Kisabei) ; namageba 
(Lugishu); lunihi (Luragoli); luhui (Lutereki). 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 43 

Coloration. In color the Greeki River specimen is in a pale reddish- 
brown pelage, with pale belly, while all of the others are of the usual 
dark blackish-brown, indicating the occasional occurrence in this as 
in some other species of the genus, of a reddish color phase. 

Measurements. <?. 87. 60. 14. 7 mm., 9 . 97. 51. 13. 9 mm. 

Enemies. Examples of this shrew were removed from the stomachs 
of Brown House Snakes (Boacdon lincatus) at Sipi and Butandiga. 
Of these the Sipi Shrew was made into a skin. 

Crocidura jacksoni jacksoni Thomas 

Crocidura jacksoni Thomas, 1904, Ann. Mag. Nat. Hist. (7), 14, p. 238: 
Ravine Station, Kenya Colony. 

c? (M. C. Z. 31784) Voi, K. C. 12. iv. 34. 
3^39 (M. C. Z. 31282-3, 31286-9) Peccatoni, K. C. 25. v. 34. 
1 (M. C. Z. 31790) Golbanti, K. C. 23. vi. 34. 

Native names. Nyonge (Kisagalla); tungu (Kipokomo). 

Coloration. A fairly uniform series, of a dull chestnut brown and 
gray, with faintly bicolor tail and pale whitish to grayish underside. 

Measurements, d\ 95. 56. 14. 8 mm., 9 . 85. 60. 12. 12 mm. 

Habitat. I dug the Voi specimen out of a mass of flood debris in 
the dry bed of the Voi River. 

Crocidura hildegardeae hildegardeae Thomas 

Crocidura hildegardeae Thomas, 1904, Ann. Mag. Nat. Hist. (7), 14, p. 240: 
Fort Hall, Kenya Colony. 

9 (M. C. Z. 31267) Sipi, U. 22. xii. 33. 

c? 9 (M. C. Z. 31241, 31258) Kaimosi, K. C. 8 & 15. ii. 34. 

9 9 (M. C. Z. 31793-4) Mt. Mbololo, K. C. 23. iv. 34. 

c? (M. C. Z. 31792) Peccatoni, K. C. 25. v. 34. 

tf 1 (M. C. Z. 31791) Wema, K. C. 19. vi. 34. 

Distribution. Apparently Heller did not secure this species at 
Kaimosi so that the pair from that locality and the female from 
Sipi, Mount Elgon, furnish interesting records north of those given 
by Hollister (1918, p. 64). 

Native names. Namageba (Lugishu); lunihi (Luragoli); luhni 
(Lutereki); munyongi (Kitaita); tungu (Kipokomo). 

Discussion. The skull length — about 18.5 to 19. mm. — is slightly 
but constantly less than in C. jacksoni. The two females from Mount 
Mbololo have slightly wider brain cases than the two males from 
Peccatoni and Wema. 



44 BULLETIN" MUSEUM OF COMPARATIVE ZOOLOGY 

Measurements, cf . (Wema) 87. 45. 12. 11 mm., 9 . (Kaimosi) 75. 
46. 12. 7 mm. The Mbololo females had identical measurements, 
presumably being subadults from the same nest, 9 9 . 67. 47. 12. 
7 mm. 

Habitat. I dug the Wema shrew from the galleries of a semi- aban- 
doned termitarium. 

Crocidura bicolor elgonius Osgood 

Crocidura bicolor elgonius Osgood, 1910, Ann. Mag. Nat. Hist. (8), 5, p. 369: 
Kirui, Mount Elgon, Kenya Colony. 

9 (M. C. Z. 31262) Butandiga, U. 14. i. 34. 
1 (M. C. Z. 31261) Kirui, K. C. 6. ii. 34. 
9 (M. C. Z. 31260) Kaimosi, K. C. 7. iii. 34. 

Native names. Namageba (Lugishu); lunihi (Luragoli); luhui 
(Lutereki). 

Coloration. The small shrews of this species are notable for their 
very short pelage. The Butandiga female is darker than the two 
others, which have a faintly brownish or chocolate tint instead of 
being dark seal brown. 

Measurements. 9 . (Kaimosi) 58. 43. 8. 7 mm. 



PTEROPIDAE 

ROUSETTUS (LlSSONYCTERIs) ANGOLENSIS (Bocage) 

Cynonyderis angolensis Bocage, 1898, Jorn. Sci. Math. Acad. Sci. Lisboa (2), 
5, p. 133: Pungo Andongo, Cahata, Quibula, Angola. 

9 (M. C. Z. 31149) Sipi, U. 21. xii. 33. 

9 (M. C. Z. 31147) Butandiga, U. 10. i. 34. 

Distribution. This bat has been recorded over the area from Angola 
to the Ruwenzori region and Tanganyika Territory, so that the above 
records from Mount Elgon extend its known range slightly to the 
northeast. 

Native names. Bebea (Kisabei); ebugut (Lugishu). 

Discussion. The Sipi specimen has a supernumerary lower molar 
lying to the inner side of the alveolar line, between the usual two 
molars. 

Measurements. 9 juv. (Butandiga) 112. 0. 18. 19. 230 mm. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 45 

Rousettus leachi (Smith) 
Pteropus leachi A. Smith, 1829, Zool. Journ., 4, p. 433: Cape of Good Hope. 
3^3? (M. C. Z. 31125, 31128-32) Sipi, U. 18. xii. 33. 

Distribution. This is the most northerly point from which typical 
leachi has been reported. 

Native names. Bebea (Kisabei); ebugut (Lugishu). 

Discussion. These six specimens are on the whole nearer typical 
leachi than aegyptiacus, and agree with the careful description of 
Andersen in having the palatal ridges 4, 3, 1, instead of 4, 4, 1, all 
three of those in which the ridges are preserved having three divided 
ridges behind the molars; the teeth are smaller; and the interorbital 
constriction equals the postorbital width instead of exceeding it. 

Measurements, tf . 150. 17. 18. 22. 300 mm., 9 . 130. 22. 20. 22. 230 
mm. 

Rousettus lanosus kempi Thomas 

Rousettus kempi Thomas, 1909, Ann. Mag. Nat. Hist. (8), 4, p. 543: Kirui, 
Mount Elgon, Kenya Colony. 

4t?19 (M. C. Z. 31123-4, 31126-7, 31148) Sipi, U. 21. xii. 33. 

Native names. Bebea (Kisabei); ebugut (Lugishu). 

Corrigenda. The series of fruit bats from Bagilo, Uluguru Moun- 
tains, Tanganyika Territory, previously referred to leachi (Allen and 
Loveridge, 1927, p. 420) are in reality kempi, their teeth being only 
minutely smaller than those of the Sipi series which are practically 
topotypes of kempi coming from the western (Uganda) side of the 
mountain instead of the southeastern (Kenya) slopes. There seems to 
be no doubt that the Elgon animal is only a very slightly differentiated 
race. 

Measurements, d 1 . 155. 25. 17. 23. 292 mm., 9. 130. 22. 22. 20. 
273 mm. 

Parasites. Numerous streblids were collected from the fur of these 
fruit bats. 

Habitat. These bats, as also the Sipi examples of the last two mem- 
bers of the genus, were all collected in the great cave below the mag- 
nificent Sipi Falls. 

Epomophorus wahlbergi wahlbergi (Sundevall) 

Pteropus wahlbergi Sundevall, 1846, Of vers. Kongl. Vet.-Akad. Forh., Stock- 
holm, 3, p. 118: Port Natal, i.e. Durban, Natal. 

& (M. C. Z. 31122) Sipi, U. 23. xii. 33. 

9 (M. C. Z. 31843) Lamu, Lamu Id., K. C. 7. v. 34. 



46 bulletin: museum of comparative zoology 

Distribution. According to Andersen's account, these specimens 
come from the range covered by the typical race. 

Native names. Ebugut (Lugishu). 

D'iscussion. Both specimens are young, yet well grown with the 
epiphyses of the finger joints nearly ankylosed with their respective 
diaphyses, so that their appearance of small size is deceptive. The 
skulls prove their youth, having shorter rostra and fuller brain cases 
than in later stages. The palatal ridges also show a slightly different 
condition in that the last one is proportionally farther forward than 
in mature animals, being slightly anterior to the middle of the post- 
dental palate. 

Measurements. <?. 100. 0. 17. 18. 190 mm., 9 . 91. 0. 15. 18. 200 mm. 

Epomophorus labiatus minor Dobson 

Epomophorus minor Dobson, 1880, Proc. Zool. Soc. London, p. 715: Zanzibar 
(fide Andersen). 

9 (M. C. Z. 31146) Kaimosi, K. C. 25. ii. 34. 

Distribution. The type locality of E. labiatus is "Abyssinia," so 
that this specimen from Kaimosi is from a locality nearly midway 
between the typical areas of that animal and E. minor. 

Native name. Linyinya (Luragoli and Lutereki). 

Discussion. Our example seems to be about intermediate in tooth 
dimensions between labiatus and minor as shown in Andersen's table 
but minutely nearer the limits assigned to minor, to which it is there- 
fore referred. Since the two forms only differ in size, the occurrence 
of an intermediate specimen seems to warrant relegating minor to 
subspecific rank. 

Measurements. 9 . 116. 0. 17. 18. 222 mm. 

Breeding. On February 25, 1934, .this female held a large fetus 
which was preserved. 



EMBALLONURIDAE 

Taphozous perforatus haedinus Thomas 

Taphozous perforatus haedinus Thomas, 1915, Journ. Bombay Nat. Hist. 

Soc, 24, p. 62: Chanler Falls, Guaso Nyiro, Kenya Colony. 
Taphozous perforatus Hollister, 1918, U. S. Nat. Mus. Bull. 99, p. 73. 

5 cf 6 9 (M. C. Z. 31847, 31849-54, 31874-6, 31878) Lamu Id., 
K. C. 7. v. 34. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 47 

Measurements, cf . 75. 26. 11. 19. 1S1 mm., 9 . SO. 26. 12. 18. 187 
mm. 

Breeding. The female whose measurements are given above was 
carrying a young male measuring 61. 16. 12. 14. 113 mm. 

Habitat. Great numbers of these tomb bats were living in some old 
buildings along the sea front. The windows had long been boarded 
up and the places used as warehouses. 



NYCTERIDAE 

Nycteris nana tristis subsp. nov. 

Type. Museum of Comparative Zoology, No. 31,156. An adult 
female, skin and skull, from Kaimosi, Kakamega district, Kenya 
Colony, collected by Arthur Loveridge, February 13, 1934. 

Description. Compared with skins from Lolodorf, Cameroon, rep- 
resenting typical N . nana (type locality, Benito River, French Congo), 
the East African race lacks the warm russet coloration of the fur of 
both surfaces, and is instead a uniform dark drab gray both above 
and below. The fur is of the same color from tip to base except on 
the nape and expecially about the bases of the ears, where it is slightly 
paler, a soiled grayish, basally. On the membranes the fur extends 
out on the propatagium from the axilla to about the end of the first 
third of the fore arm, and on the plagiopatagium to a line joining 
the elbow and the first third of the tibia. On the uropatagium the 
fur extends out as far as a line connecting the proximal ends of the 
tibiae. On the under side the extent is about the same, except that 
it does not quite reach the knees. 

Measurements. The specimen has practically the same dimensions 
as those of the West African race. The type measures: fore arm, 35.8 
mm.; tibia, 15.7; foot, 6.5; tail (about) 45; thumb, 11.5; third meta- 
carpal, 28.5; first phalanx, 16.5; fourth metacarpal, 29.7; fifth meta- 
carpal, 30.3 mm. 

The skull measures: greatest length, 16.6 mm.; basal length, 12.5; 
palatal length, 3.6; zygomatic width, 9.3; mastoid width, 8.0; width 
across frontal plate, 6.6; width outside last molars, 6.2; upper cheek 
teeth, 5.3; lower cheek teeth, 5.8 mm. 

Remarks. Through the kindness of Mr. J. Kenneth Doutt, of the 
Carnegie Museum, we have had the loan of two specimens represent- 
ing typical Nycteris nana, from Lolodorf, Cameroon. Both agree in 
their pronounced russet tint, contrasting with the dull gray hue of the 



48 bulletin: museum of comparative zoology 

eastern animal. It is a rather rare species, for, in addition to the 
original specimen from Benito River, it has apparently been recorded 
but twice: by Hollister, in 1918, who mentions two in the United 
States National Museum from Yala River, Kenya Colony, as form- 
ing a considerable extension of the known range into eastern Africa, 
and again by Cabrera and Ruxton (1926, Ann. Mag. Nat. Hist. (9), 
17, p. 591), who had a specimen from Luluabourg, Belgian Congo, 
that flew into a room. The Cameroon specimens, referred to above, 
are now recorded for the first time. 

Nycteris hispida (Schreber) 

Vespertilio hispida Schreber, 1774, Saugthiere, pi. lvi: Senegal. 

2 cf 2 9 (M. C. Z. 31153-5, 31157) Kaimosi, K. C. 9. ii. & 9. iii. 34. 
9 (M. C. Z. 31842) Ngatana, K. C. 14. vi. 34. 

Native names. Linyinya (Luragoli) ; nundu (Kipokomo) ; both gen- 
eral for small bats. 

Discussion. Externally these differ from somewhat similar specie? 
in having the fur extend laterally on to the membrane from the first 
third of the forearm to the knee, and on the interfemoral membrane 
slightly beyond to a line about joining the middle of the tibiae, though 
more thinly. At the sides of the lateral membrane the fur becomes a 
golden brown. 

Measurements, cf. (Kaimosi) 50. 48. 7. 22. 140 mm., 9 . (Wema, 
Ngatana) 48. 50. 7. 20. 124 mm. 

Habitat. Both this and the following species occur together for 
they were brought to my tent late at night by a native who had taken 
them in the village of Wema close to my camping ground. 

Nycteris aurita (Andersen) 

Petalia aurita K. Andersen, 1912, Ann. Mag. Nat. Hist. (8), 10, p. 547: Kilifi, 
Kenya Colony. 

9 (M. C. Z. 32059) Ngatana, K. C. 14. vi. 34. 

Distribution. In addition to the type, described by Andersen from 
Kilifi, which is just a hundred miles south of Ngatana, we have found 
two other specimens of this bat recorded, namely those in the United 
States National Museum listed by Hollister (1918, p. 74), one taken 
many years ago on the Tana River, by Chanler, and recorded by 
True as Nycteris hispida, the other secured by Heller on the Marsabit 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 49 

Road, northern Kenya Colony. Granvik (1924) it is true reports its 
occurrence in numbers in a cave on Mount Elgon, but it seems possible 
that this identification requires confirmation. 

Measurements. 9 . 52. 55. 8. 27. 138 mm. 

Nycteris damarensis brockmani (Andersen) 

Petalia damarensis brockmani K. Andersen, 1912, Ann. Mag. Nat. Hist. (8), 
10, p. 548: Upper Sheikh, British Somaliland. 

9 (M. C. Z. 31141) Voi, K. C. 7. iv. 34. 

Discussion. This single specimen seems to correspond in every 
particular to Andersen's description, its measurements being about 
the maximum of the extremes that he gives. In color the lower side 
is very pale, almost whitish. 

Measurements. 9 . 62. 58. 12. 34. 150 mm. 

Nycteris thebaica revoili Robin 

Nycteris revoili Robin, 1881, Bull. Soc. Philom. Paris, (7), 5, p. 90: Northern 
Somaliland. 

9 9 (M. C. Z. 31158-9) Elgonyi, K. C. 24. i. & 4. ii. 34. 
9 (M. C. Z. 31160) Kirui, K. C. 29. i. 34. 

Native name. Kubukabuk (Kitosh). 

Discussion. The form revoili seems to replace typical thebaica in 
Somaliland and Kenya Colony, and as Dobson once suggested, may 
prove to be only racially distinct from capensis to the south. It seems 
best, therefore, to regard revoili as a subspecies of the Egyptian 
thebaica, which it resembles in size. A cotype of revoili which has been 
used for comparison, does not differ except possibly in minute points 
that fall within the limits of individual variation. 

Measurements. 9 . (Kirui) 50. 52. 10. 33. 144 mm. 

Breeding. The female from Kirui on January 29 held a fetus ready 
for birth (preserved in alcohol). One of the Elgonyi bats was carrying 
two young on February 4, 1934. The measurements of one of these 
sucklings were as follows: cf . 36. 20. 9. 16. 63 mm. 

Habitat. The Kirui bat was not taken in Kemp's cave but in a low 
cavern on the opposite slopes of the valley. When I entered there 
were five bats in the colony, four flew out immediately on being 
disturbed so I only succeeded in netting the pregnant female which 
flew from one corner to another. 



50 bulletin: museum of comparative zoology 

MEGADERMIDAE 

Lavia frons rex Miller 

Lavia rex Miller, 1905, Proc. Biol. Soc. Washington, 18, p. 227: Taveta, 
Kenya Colony. 

<? 9 (M. C. Z. 31134-5) s. bank Greeki River, U. 7. xii. 33. 

c? (M. C. Z. 31133) Kaimosi, K. C. 1. iii. 34. 

c? (M. C. Z. 31836) Kitau, Manda Id., K. C. 15. v. 34. 

c? (M. C. Z. 31839) Mkonumbi, K. C. 29. v. 34. 

9 (M. C. Z. 31837) Golbanti, K. C. 22. vi. 34. 

Native name. Lumenwa (Karamojong). 

Discussion. The status of the forms described as rex and affinis 
seems still not to be finally settled, but we have provisionally con- 
sidered these specimens to represent the former. 

Measurements, d 71 . (Kaimosi) 70. 0. 16. 42. 190 mm., 9. (Greeki 
River) 71. 0. 18. 45. 162 mm. 

Enemies. The skin from Kitau was taken from a freshly swallowed 
bat recovered from the stomach of a Common Mamba {Dendraspis 
angusticcps) . 

Habitat. The pair from Greeki River were shot as they hung to- 
gether in a thorn tree near the south bank. The species was also 
seen at Sheila on Lamu Island. They were common in the acacia on 
Manda Island. The Mkonumbi bat was flitting from tree to tree 
calling with a bird-like note just after dark. After watching it for 
some time, I shone its eyes with an electric torch and shot it with a 
.22 cartridge loaded with dust shot. 

Cardioderma cor (Peters) 
Megaderma cor Peters, 1872, Monatsb. Akad. Wiss. Berlin, p. 194: Abyssinia. 
4 a* 8 9 (M. C. Z. 31825-35, 31838) Lamu Id., K. C. 8. v. 34. 

Distribution. This seems to be a rather uncommon species prin- 
cipally found in northeastern Africa. 

Coloration. The uniformly blue-gray pelage is a shade darker in the 
immature specimen. 

Measurements. <? . 75. 0. 18. 35. 175 mm., 9 . 75. 0. 18. 39. 174 mm. 

Parasites. Streblids were recovered from their fur. 

Habitat. At least a hundred of these bats were found to be occupy- 
ing a deserted house on one of Mr. C. E. Whitton's estates about 
half-an-hour's walk northwest of Lamu township. They hung from 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 51 

the beams and found ready egress through the large gaps in the rot- 
ting thatch. All were shot with a .22 rifle whose rifling was removed 
to fire dust shot. 

RHINOLOPHIDAE 

Rhinolophus hildebrandtii Peters 

Rhinolophus hildebrandtii Peters, 1878, Monatsb. Akad. Wiss. Berlin, p. 195: 
Ndi, Taita, Kenya Colony. 

9 (M. C. Z. 31840) Voi, K. C. 11. iv. 34. 

Distribution. Voi is only about fifteen miles distant from the type 
locality. 

Coloration. This specimen agrees in its pale buffy color with ex- 
amples from Tanganyika Territory, in this respect differing notice- 
ably from the series of the smaller, darker R. eloquens. 

Measurements. 9 . 80. 45. 12. 33. 195 mm. 

Rhinolophus eloquens Andersen 

Rhinolophus hildebrandti eloquens K. Andersen, 1905, Ann. Mag. Nat. Hist. (7), 
15, p. 74: Entebbe, Uganda. 

<? (M. C. Z. 31173) Mt. Debasien, U. 13. xi. 33. 
1+7^29 (M. C. Z. 31161-70) Kirui, K. C. 23. i. 34. 
9 (M. C. Z. 31171) Elgonyi, K. C. 23. 1. 34. 

Native names. Lumenwa (Karamojong) ; kurukuru (Kitosh). 

Discussion. Hollister (1918, p. 84) has shown that this smaller 
species is well separated in dimensions, with a forearm of about 
50 or 57 mm., from the larger R. hildebrandtii and probably should 
be considered as a distinct species, rather than a geographical race. 

Coloration. This fine series of well made skins shows a decided 
difference in color from the series of hildebrandtii from Tanganyika 
collected by Loveridge on an earlier expedition. They are decidedly 
darker, a general light blackish brown, instead of buffy brown (cf. 
Ridgway) and can easily be separated by this character, thus con- 
firming Hollister's judgment of their specific distinction. 

Measurements. & . (Kirui) 75. 30. 13. 29. 177 mm., 9 . (Kirui) 73. 
30. 12. 28. 181 mm. 

Parasites. One had an hemipterous polyctenid as well as a large 
tick {Ixodes simplex) on its breast, many had encysted flies (Ascodi- 
pteron) in their wings as well as numerous streblids. 



52 bulletin: museum of comparative zoology 

Habitat. All the Kirui series were taken in Robin Kemp's cave in 
the course of an hour. They simply swarmed in the low recesses at 
the back of this cave, the noise of their wings was like the running 
of the engine of a stationary car. When molested in one of these sub- 
sidiary caves they quickly migrated to another. We followed by 
crawling through the tunnel back into the main cave and thence 
to their fresh refuge. Associated with them, though in the proportion 
of one to ten, were some horseshoe bats {Hipposideros caffer) or perhaps 
the latter were more adroit at avoiding the butterfly net employed in 
capturing their companions. 

I shot the Debasien bat at dusk as it hung up in a thorn tree to 
rest. 

Rhinolophus fumigatus exsul Andersen 

Rhinolophus fwnigatus exsul K. Andersen, 1905, Ann. Mag. Nat. Hist. (7), 
15, p. 74: Kitui, Kenya Colony. 

Yng. (M. C. Z. 31172) Kirui, K. C. 5. ii. 34. 

Discussion. This seems to be an uncommon species. Only one 
example, and that without skull, was available to its describer in 
1905. It was one of the few species of the genus which the East 
African expeditions of the United States National Museum failed to 
find in that region. In addition to the male secured by Loveridge, 
the Museum of Comparative Zoology has a second from the Kenya 
forest, ten miles west of Chuka, collected by Dr. F. R. Wulsin in 
1914. The skull shows the minute upper and lower premolars present. 
The hairy noseleaf and the forearm of about 51-52 mm. are distinctive 
in comparison with other East African members of the genus. 

HIPPOSIDERIDAE 

Hipposideros caffer (Sundevall) 

Rhinolophus caffer Sundevall, 1846, Ofvers. Kongl. Vet.-Akad. Forh., Stock- 
holm, 3, p. 118: Near Port Natal, Natal. 

7 c? (M. C. Z. 31844-6, 32062-5) Tsavo, K. C. 30. hi. 34. 

Measurements. & . 52. 32. 8. 15. 136 mm. 

Habitat. I captured all these at night with a butterfly net in one 
of the rooms of the abandoned house, three hundred yards north of 
the station, which I occupied. By day the bats apparently sleep in 
the roof, but at night fly in and out of the rooms through the broken 
window panes, and rest by hanging on the cornices. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 53 

Hipposideros ruber (Noack) 

Phyllorhina rubra Noack, 1893, Zool. Jahrb. Syst., 7, p. 586: "Lugerrunjere" 
i.e. Ngerengere River, Tanganyika Territory. 

9 (M. C. Z. 31150) Kirui, K. C. 23. i. 34. 
d" 9 (M. C. Z. 31151-2) Elgonyi, K. C. 23. i. 34. 

Coloration. These three skins represent the brown phase. 
Measurements, cf . 61. 35. 8. 15. 147 mm., 9 . (Elgonyi) 55. 38. 8. 
15. 147 mm. 

Parasites. A nyeteribiid on one. 



VESPERTILIONIDAE 

Pipistrellus nanus (Peters) 

Vespertilio nanus Peters, 1852, Reise nach Mossambique, Saugethiere, p. 63: 
Inhambane, Mozambique. 

7 c? 3 9 (M. C. Z. 31136-45) Kaimosi, K. C. 26-28. ii. 34. 
2^69 (M. C. Z. 31868-73, 32060-1) Mt. Mbololo, K. C. 23. iv. 34. 
c? (M. C. Z. 31867) Golbanti, K. C. 22. vi. 34. 

Native name. Eososi (Kitaita). 

Coloration. The series from Kaimosi averages slightly darker than 
the other specimens listed above which are from the coastal region. 
They can, however, be matched by specimens from Tanganyika 
Territory to the south, so that the difference is probably due to fading. 

Measurements, cf . (Kaimosi) 43. 34. 5. 8. 100 mm., 9 . (Kaimosi) 
38. 30. 5. 8. 107 mm. 

Breeding. One of the females from Kaimosi was carrying a young 
one at her breast on February 26, 1934. 

Glauconycteris argentata (Dobson) 

Chalinolobus (Glauconycteris) argentatus Dobson, 1875, Proc. Zool. Soc. London, 
p. 385: Cameroon Mountains. 

7^89 (M. C. Z. 31879, 31881-7, 32052-8) Kikuyu, K.C. 14. iii. 34. 

Measurements, d 1 . 55. 46. 4. 9. 140 mm. 

Breeding. Four of the females were carrying young. 

Habitat. Dr. J. W. Arthur directed my attention to a circular 
patch on his lawn where the grass had been killed. It was about a 
foot in diameter and immediatelv beneath a solid cluster of these 



54 bulletin: museum of comparative zoology 

bats, which were hanging in a correspondingly circular area of an 
ornamental 'palm' {Dracaena sp.) at a height of twelve feet from the 
ground. 

After they had been photographed, during which process they 
showed no nervousness though the camera and operator were within 
six feet of them, I made one sweep of my net which resulted in the 
capture of thirty bats. Five were released one by one and photo- 
graphed in flight, unfortunately the results were not a success. Ten 
were skinned and donated to the Coryndon Museum at Nairobi in 
appreciation of the facilities and materials afforded for skinning. 
I reached Nairobi at 11 a.m. and my young Mgishu skinner started 
on the work almost immediately continuing until he had finished 
the last of the twenty-five at 5 p.m. 



MOLOSSIDAE 

Mops (Allomops) osborni Allen 

Mops (Allomops) osborni J. A. Allen, 1917, Bull. Amer. Mus. Nat. Hist., 37, 
p. 473: Kinshasa, near Leopoldville, Belgian Congo. 

9 (M. C. Z. 31863) Bellazoni, K. C. 5. vi. 34. 

5 <? 5 9 (M. C. Z. 31855-62, 31864-5) Ngatana, K. C. 14. vi. 34. 
also 5 in alcohol with the same data as these last. 

Native name. Nundu (Kipokomo). 

Coloration. On comparing these skins with a considerable series 
from Ujiji, collected by Loveridge on a previous expedition, they seem 
to average paler on the throat and chest. In the Ujiji series, which 
must represent the typical race, the throat is gray in most cases, 
and this color extends to the upper chest. In two or three, however, 
the whole central part of the under surface is white from the dark 
chin to the anus. 

In the Ngatana series only two out of the ten are dark throated, 
the others being pure white throughout the mid-ventral region. 
Notwithstanding this apparently average difference, the extremes 
are bridged by individual variation, so that it does not seem possible 
to recognize a coastal race. A few of the series have the entire dorsal 
surface of head and body mixed with whitish hairs, giving a slightly 
frosted effect, while others are uniformly brown above. 

Measurements. <?. 85. 52. 12. 19. 180 mm., 9.77.49.11.17.170 mm. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 55 

Habitat. These were obtained from the school house of a village, 
on the banks of the Tana, which had been abandoned by order of the 
medical officer of health. 

Chaerephon hindei (Thomas) 

Nyctinomus hindei Thomas, 1904, Ann. Mag. Nat. Hist. (7), 13, p. 210: Fort 
Hall, Kenya Colony. 

tf 1 cf (M. C. Z. 31866, 31877) Witu, K. C. 7. vi. 34. 

Coloration. Although both these specimens agree in the pattern 
of white markings, with the midventral area and a line from armpit 
to anus white, wings white, interfemoral dark, the older specimen is 
much more russet brown than the younger, which is a grayish choc- 
olate. 

Measurements, c? . 60. 41. 9. 17. 138 mm. 



CANIDAE 

Thos mesomelas mcmillani Heller 

Thos mesomelas mcmillani Heller, 1914, Smithsonian Misc. Coll., 63, No. 7, 
p. 6: Mtito Andei Station, Kenya Colony. 

<? <? (M. C. Z. 31957, 31960) Tsavo, K. C. 3. iv. 34. 

Coloration. These two jackals, from very near the type locality, 
agree with the coastal race in the pale sides, buff instead of rufous, 
and generally less rufous tints than the more inland race. 

Native name. Muzozo (Kitaita). 

Measurements, d 1 . 745. 330. 162. 106 mm. 

Diet. In the stomachs of both were numerous scraps of goat hide 
and hair, obviously trimmings picked up at the village; in addition 
there was much wild fruit. 

Habits. Both were shot as they visited the carcass of a hyrax which 
I had pegged beneath a baobab, the first at 1.45 a.m., the second at 
2.30 a.m. Their calls, which seemed to me to differ considerably 
from those of jackals in central Tanganyika, could be heard in each 
case for half-an-hour before the animal arrived. 

Lycaon pictus lupinus Thomas 

Lycaon pictus lupinus Thomas, 1902, Ann. Mag. Nat. Hist. (7), 9, p. 439: 
Nyuki River Swamp, Rift Valley, Kenya Colony. 



56 bulletin: museum of comparative zoology 

Native name. Kiwao (Kitaita). 

Habits, etc. On arrival at Kibwezi, March 22, 1934, I made en- 
quiries with a view to securing topotypes of the hunting dog (L. hueb- 
neri) described from here. Herr. Huebner, I learned, had been em- 
ployed on the Dwa Estate many years before the War. I walked 
out to Dwa, which is four miles from Kibwezi station, and was told 
by Mr. A. B. C. Smith, the manager of Dwa Estate, that one of his 
staff, Mr. Cushny, had, only a fortnight before, shot and killed ten 
and wounded an eleventh hunting dog from a pack of twelve. The 
place was fifteen miles away and Mr. Smith very kindly arranged 
for a native to be sent to the spot. On his return, this man reported 
that no skulls were to be found, and presumed that hyenas had cleaned 
up the spot. At irregular intervals packs hunt through the Estate, 
usually at night. It would be very difficult to arrange to hunt such 
mobile creatures which are here one day and miles away the next. 

At Tsavo, on April 4, 1934, I was awakened at daybreak, half-an- 
hour before sunrise, by the patter of feet outside the long-vacant 
house which I was occupying. Almost simultaneously a hoof sharply 
struck a tin can. Springing from bed, I stepped out on to the verandah. 
For a hundred feet around the house the ground is clear, beyond this 
the thornbush stretches like a blanket over the country, in every 
direction, so far as the eye can see. At the edge of the clearing stood 
a half-grown waterbuck calf; next moment it had disappeared into 
the bush. 

Returning to my room, I began to dress and was putting on my 
puttees when my ears were assailed by the terrified bleating of the 
young waterbuck. The cries came from the front of the house, down 
by the river. Snatching up a rifle, I raced down the slope, dodging 
the thornbushes as best I could. Ahead of me sounded a heavy splash 
followed by much splashing. Some animal dashed through the grass 
at the river's brink, it was a hunting dog. I took a snapshot at it 
but missed being very much out of breath. Up to this time I had not 
given hunting dogs a thought having supposed that a crocodile had 
seized the calf or its mother. Later my gunbearer reported that he 
had seen the mother make off to the south as I ran down the hill. 

The calf was standing in the middle of the Tsavo River, only its 
head and part of its back showed above the swift current. Some 
hunting dogs were in shallow water on the farther side but lost no 
time in scrambling out and up the bank. On my side two others, 
hidden in the rank grass, barked and yapped defiantly at me before 
taking to the thornbush. I followed them for a quarter-of-an-hour 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 57 

but they kept well to cover. Returning to the river we found the buck 
still in the same place, we sat down to watch it, momentarily fearing 
that it would be taken by a crocodile. After waiting ten minutes, 
however, we saw the creature swim upstream then struggle up the 
farther bank. The dogs had been hunting silently until now, but later 
in the morning we heard them calling, and concluded that they made 
a kill about a mile down the river about 11 a.m. 

According to the natives, hunting dogs visit Manda Island from 
time to time, harassing the dikdik, then return to the mainland. 



MUSTELIDAE 

Mellivora capensis sagulata Hollister 

Mellivora sagulata Hollister, 1910, Smithsonian Misc. Coll., 56, No. 13, p. 2: 
Mount Kilimanjaro, Tanganyika Territory. 

Native skin (M. C. Z. 31951) Mt. Mbololo, K. C. 14. iv. 34. 

Native names. Ekore (Karamojong) ; kisegi (Kitaita). 

Coloration. This skin agrees fairly well with Hollister's description 
and with other specimens from Tanganyika Territory. The buffy 
stripe continues from the forehead to the rump, while the remaining 
pale area of the back is darker with admixture of black hairs and 
others with rusty tips, perhaps in part due to staining. Where these 
animals live in red-soil country they become much stained with this 
earth. 

Folklore. The following tale was related by my Karamojong gun- 
bearer. "Once upon a time a Karamojong woman was out in the bush 
when she came upon a ratel which had been covered with dirt by 
termites. Its ears and tail were showing, however. Exclaiming 'What 
good fortune to find an animal already dead,' she took some branches 
and brushed off the dirt. It was a time of scarcity or famine so she 
concealed the ratel among her skin clothing and returned home. On 
her meeting some neighbors, they asked her, 'What have you there?' 
To deceive them, she replied, 'Oh, nothing, a dead creature which has 
become very rotten.' When they would have seen it, she refused and, 
entering her hut, fastened the door securely. 

This being done she instructed her child to build up the fire. Then 
told him to bring a knife with which to skin the ratel. The boy tried 
to cut the tough skin but failed. His mother ordered him to sharpen 
the knife. Then she held the ratel while he tried to cut into the skin 



58 bulletin: museum of comparative zoology 

at the throat. At this the ratel moved an ear. The boy cried, 'Mother, 
it is not dead.' 'Nonsense,' answered the mother. 'Do as I bid you.' 
The child then succeeded in making an incision whereupon the ratel, 
which had only been somnolent as a result of gorging itself upon honey 
and grubs, revived and attacked the child. The mother attempted to 
drive it off but the savage creature sprang upon her, scratching her 
head and face and biting her severely. 

She shrieked aloud for help, the neighbors tried the door but found 
it fastened too securely for them to open. The woman's husband, 
who had been herding goats, was returning to the village when he 
heard the uproar. Leaving the goats in care of another son, he hurried 
to the house and broke down the door. As he did so, the ratel dashed 
out, the assembled neighbors hurled their spears at it, but not a spear 
penetrated the tough hide and the ratel made good its escape." 

Compare this story with that of the Wakami relating to a civet as 
recorded under that species. 

Aonyx capensis hindei (Thomas) 

Lutra capensis hindei Thomas, 1905, Ann. Mag. Nat. Hist. (7), 15, p. 78: 
Fort Hall, Kenya Colony. 

& 9 (M. C. Z. 31621-2) Mvvahedio River, Kaimosi, K. C. 9. li. 34. 

Native names. Lizibi (Luragoli); inzibi (Lutereki). 

Discussion. Two beautiful specimens of the clawless otter are 
referred to hindei, although they differ from the type as described 
by Thomas, in having the ears rimmed with white above as is usual 
in the species, instead of uniformly dark. As noted by J. A. Allen 
both this and the race helios from the Sotik were based on single 
individuals, so that until series of specimens can be compared, the 
value of the characters claimed for them cannot be estimated. 

Measurements, cf. 740. 425. 120. 33 mm., 9 . 600. 400. 130. 30 mm. 

Parasites. Ticks (JSaemaphy salts leachii) were abundant in their fur. 



VIVERRIDAE 

Civettictis civetta schwarzi Cabrera 

Viverra civetta orientalis Matschie, (not V. orientalis Hodgson, 1842 = V. zibetha 
Linnaeus) 1891, Arch, fur Naturgesch., 1, p. 352: Zanzibar. 

Viverra civetta schwarzi Cabrera, 1929, Mem. R. Soc. Espanol. Hist. Nat., 
Madrid, 16, No. 1, p. 36, footnote: Bagamoyo, Tanganyika Territory. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 59 

Viverra civetta matschiei Pocock, 1933, Journ. Bombay Nat. Hist. Soc, 36, 
p. 429, footnote. 

c? (M. C. Z. 31612) Sipi, U. 18. xii. 33. 

9 9 (M. C. Z. 31104, 32271) Butandiga, U. 9. i. 34. 

9 (M. C. Z. 31611) Bukori, K. C. 18. i. 34. 

cf 9 (M. C. Z. 32203, 32257) Kaimosi, K. C. 15 & 23. ii. 34. 

Native names. Mugis (Kisabei); ndesi (Lugishu); kuteli (Kitosh); 
ndereet (Kimasai); ligunyuli (Luragoli); likunyuli (Lutereki); fungo 
(Kitaita). 

Coloration. The Sipi male is melanistie. 

Measurements, cf • (Kaimosi) 930. 400. 140. 55 mm., 9 - (Butandiga) 
890. 410. 125. 52 mm. 

Breeding. The big Butandiga female was suckling a large kitten, 
the latter measuring: 9 . 420. ISO. 75. 41 mm. The Kaimosi female 
was also obviously nursing a family on February 15, 1934. 

Diet. The Sipi civet was very emaciated and its stomach held only 
a mass of hair matted into the shape and size of an ordinary brown 
rat. One Kaimosi animal held a rat (Oenomys b. editus), the other 
a nosehorned viper (Bitis nasicornis) and some invertebrates. 

Parasites. Ticks (Haemaphy salts leachii) were collected from the 
Sipi, Butandiga and Kaimosi civets. 

Folklore. "Civets," said an Mkami to me, "rob our gardens of the 
maize, gorge upon the cobs, and then retire to the bush to sleep. 
They sleep so heavily with their mouths wide open that they deceive 
even the flies who, supposing them to be dead, assemble to lay their 
eggs about the gaping jaws. One day an Mkami discovered that his 
garden had been robbed, tracked the thief and found that it was a 
civet which had apparently succumbed from too heavy a repast. 
Wrapping the animal in banana leaves, he threw it over his shoulder 
and returned to the village where a beer-drink was in progress. As 
the Wakami ate civets in those days, and he was anxious to avoid 
sharing the meat with his neighbors, he entered his hut surreptitiously, 
closed the door, and built up a big fire. 

"His wife and children gathered round as he thrust the fungo on 
the fire to roast. Immediately the creature awoke and dashing wildly 
about the hut broke utensils and scattered the fire. Seizing a spear 
the man attempted to kill the terrified animal, but in the uproar, 
smoke, and confusion accidentally speared his own child. On realizing 
what had happened, the mother and other children began to wail. 
The neighbors, hearing the outcry, broke in the door and dragged the 



60 bulletin: museum of comparative zoology 

man away to a council of the old men who fined him heavily in goats 
for the murder." 

"So that is what came of his not sharing his find with his friends," 
added Salimu, with a grin. "He lost the civet, killed his child, and it 
cost him several goats. The only profit was for the village elders who 
had a good feast at his expense." 

Whether this moral tale was pure folklore, or whether it had some 
slight basis in fact, of course, I cannot say. It was related to me 
as an explanation of why the Wakami will not touch civet meat today. 
Strangely enough, my superior, and relatively cultured, Mganda cook 
had a great fondness for civet, toasting the meat over the fire. As 
his companions scorned to touch it, the carcasses of all the animals 
listed above fell to his share. In comparing the foregoing story with 
the Karamojong tale of a ratel, it should be remembered that the 
Wakami and Karamoja tribes are separated by seven hundred miles. 

Genetta servalina bettoni Thomas 

Genetta bettoni Thomas, 1902, Ann. Mag. Nat. Hist. (7), 9, p. 365: Lagari, 
Mau district, Kenya Colony. 

J 1 (M. C. Z. 32305) Sipi, U. 26. xii. 33. 

9 (M. C. Z. 32207) Butandiga, U. 12. i. 34. 

9 (M. C. Z. 32306) Kaimosi, K. C. 10. ii. 34. 

Native names. Lungiri (Lugishu) ; kidarongo (Luragoli) ; shingangayu 
(Lutereki). 

Discussion. This small-spotted genet is evidently an eastern rep- 
resentative of the servalina type of West African forests, and we have 
therefore regarded it as a subspecies of that animal. It furnishes 
another instance of the eastward extension of the forest fauna of the 
Congo basin, and is apparently a much less common species than 
G. s. stuhlmanni in the area here covered. Hollister (1918, p. 118) 
records two examples from the Kakamega region in the United States 
National Museum. 

Measurements, cf. 505. 445. 81. 41 mm., 9 . 412. 350. 77. 42 mm. 

Breeding. The Butandiga genet is only a little more than halfgrown. 

Diet. Rodent fur in the stomach of the Sipi specimen. 

Parasites. Ticks (Haemaphysalis leachii) were present on the 
Butandiga example. 

Genetta stuhlmanni stuhlmanni Matschie 

Genetta stuhlmanni Matschie, 1902, Verh. V. Int. Zool-Congr., Berlin, p. 1142: 
Bukoba, Tanganyika Territory. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 61 

3 d" 1 9 (M. C. Z. 32289-92) Sipi, U. 19-25. xii. 33. 
c? (M. C. Z. 32296) Butandiga, U. 9. i. 34. 
9 (M. C. Z. 32297) Kirui, K. C. 31. i. 34. 
1 <? 10 9 (M. C. Z. 32293-5, 32298-303, 32308-9) Kaimosi, K. C. 
9. i-17. ii. 34. 

Native names. Mwown (Kisabei); lunziri (Lugishu); kumondo 
(Kitosh) ; maragok (Kimasai) ; kidarongo (Luragoli) ; shitarongo 
(Lutereki). 

Coloration. Notwithstanding the great individual variation in color, 
this seems to be of a fairly definite sort. In skins from the same 
locality, the ground color is usually buffy, sometimes light gray 
without buffy tinge except that the pale rings of the tail are almost 
always somewhat tinged with buffy. The spots and broken stripes 
of darker, are either all black, or there is a variable amount of chestnut 
or rusty red hairs in the center of the black spot, these hairs pre- 
dominating in some and in the extreme cases forming the entire spot 
to the practical exclusion of black hairs. 

Measurements, d 1 . (Kaimosi) 520. 420. 75. 44 mm., 9 . (Kaimosi) 
500. 430. 85. 44 mm. 

Breeding. This largest female from Kaimosi, killed on February 16, 
1934, had two kittens, both of which were females, the bigger measur- 
ing 270. 222. 60. 35 mm. 

Diet. The stomach contents of Sipi genets were noted as follows: 
(1) rodent fur; (2) rodent remains, apparently those of several young 
rats together with the reddish feathers of a bird, or birds, possibly 
those of nestling Paradise Flycatchers (Tchitrea v. viridis); (3) a 
young rat (Otomys t. elgonis) and a round white forest fruit smaller 
than an average marble. The Butandiga genet held a rat (Tatera sp.) 
while Kaimosi specimens contained: (1) rodent fur; (2) a tree rat 
(Oenomys b. editus); (3) a swamp rat (Otomys t. elgonis). 

Parasites. Ticks (II 'aemaphy salts leachii) on a Sipi and Kaimosi 
genet, the Sipi animal also harboring a tapeworm (Taenia parva). 

Genetta stuhlmanni erlangeri Matschie 

Genetta erlangeri Matschie, 1902, Verh. V. Int. Zool.-Congr., Berlin, p. 1143: 
Kitui, Kenya Colony. 

9 (M. C. Z. 32326) Mt. Mbololo, K. C. 25. iv. 34. 
9 (M. C. Z. 32304) Lamu Island, K. C. 12. v. 34. 
9 (M. C. Z. 32325) Ngatana, K. C. 12. vi. 34. 

Native names. Ludindi (Kitaita); kanu (Kiswahili and Kipokomo). 



62 bulletin: museum of comparative zoology 

Coloration. These are referred to the paler coastal race, although 
it must be admitted that many specimens are difficult to place. The 
adult from Lamu Island has a pale gray ground color with black spots 
and, except for the white instead of buffy rings on the tail, it is prac- 
tically an exact match for the palest of the Kaimosi series of the 
typical form. The whitish rings on the tail characterize the two other 
specimens which are immature, whereas in all the Kaimosi series the 
rings are buffy. The same variation in the spots from black to red- 
centered or all-chestnut-red is seen in the coast form. 

Measurements. 9 - 500. 420. 87. 44 mm. 

Nandinia binotata arborea Heller 

Nandinia binotata arborea Heller, 1913, Smithsonian Misc. Coll., 61, No. 13, 
p. 9: Lukosa River, Kenya Colony. 

2 9 (M. C. Z. 32249-50) Sipi, U. 18 & 22. xii. 33. 
9 (M. C. Z. 32248) Butandiga, U. 13. i. 34. 
4(^19 (M. C. Z. 31101, 31623-4, 32193, 32251) Kaimosi, K. C. 
14-24. ii. 34. 

Type locality. The Kaimosi series are topotypes as they come from 
the forest through which the Lukosa (i.e. Yala) River flows. 

Native names. Moxoe (Kisabei); liwala (Lugishu); ninamugogo 
(Luragoli); kunamugogo (Lutereki). 

Coloration. This series is very uniform in its ground tint as well as 
in the rather light spotting, but is not very different in these char- 
acters from Cameroon examples of the typical form, though in the 
latter the general tone is slightly darker. 

Measurements, cf. 580. 560. 80. 41 mm., 9 . (Kaimosi) 530. 530. 80. 
38 mm. 

Diet. One stomach held a tree rat (Oenomys b. editus), another a 
swamp rat (Otomys t. elgonis) at Kaimosi, all the other stomachs were 
empty and free of parasites. 

Enemies. The bodies were eaten by the Bagishu who hunted them. 

Galerella sanguinea ibeae (Wroughton) 

Mungos sanguineus ibeae Wroughton, 1907, Ann. Mag. Nat. Hist. (7), 20, p. 
118: Fort Hall, Kenya Colony. 

9 (M. C. Z. 32332) Tsavo, K. C. 2. iv. 34. 

Discussion. The late Dr. J. A. Allen has advocated the use of the 
generic name Galerella for the mongooses of this group. Thomas's 
Myonax is perhaps not very different. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 63 

Measurements. 9 . 295. 280. 55. 25 mm. 

Diet. The stomach was empty, these animals being diurnal feeders. 

Habitat. At 8 a.m. I observed a pair of mongoose basking in com- 
pany of a hyrax (//. s. hindci) on the top of a huge rock, the hyrax 
observed me and promptly disappeared. I stalked its companions, 
the male mongoose rose on its hind legs, meerkat fashion, the better 
to see me, then followed the hyrax as I fired at the female. 

Herpestes ichneumon funestus (Osgood) 

Mungos ichneumon funestus Osgood, 1910, Publ. Field Mus. Nat. Hist., Zool. 
Series, 10, No. 3, p. 17: Naivasha, Kenya Colony. 

d 1 (M. C. Z. 32245) Sipi, U. 11. xii. 33. 
<? (M. C. Z. 32246) Kirui, K. C. 30. i. 34. 
2^49 (M. C. Z. 32240-4, 32247) Kaimosi, K. C. 17-26. ii. 34. 

Native names. Ekosemate (Karamojong) ; churunguru (Kisabei); 
serengeta (Lugishu); anamambi (Kitosh); mogoect (Kimasai); luna- 
mageki (Luragoli); linueli (Lutereki). 

Discussion. In spite of its large size, the dentition of this mongoose 
is noticeably weak in comparison with that of Atilax. 

Measurements, c? . (Kirui) 600. 460. 91. 35 mm., 9- (Kaimosi) 
540. 480. 82. 35 mm. 

Breeding. A kitten (d 71 . 275. 150. 53. 26 mm.) obtained at Kaimosi 
on February 20, 1934, had milk only in its stomach. 

Diet. At Sipi, rodent fur and the claw of a fowl. At Kaimosi, three 
animals each held a different species of rat, viz. Ocnomys b. editus, 
Praomys t. jacksoni and Arvicanthis a. nubilans. 

Parasites. Nematodes (Ascaris sp.) were collected from one Kaimosi 
mongoose. 

Atilax paludinosus robustus (Gray) 

Athylax robustus Gray, 1864, Proc. Zool. Soc. London, p. 558: White Nile. 
cf (M. C. Z. 32256) Butandiga, U. 9. i. 34. 
<? (M. C. Z. 31606) Kaimosi, K. C. 12. ii. 34. 

Native names. Linsi (Lugishu); mugoeet (Kimasai); lunamatu 
(Luragoli). 

Coloration. While both specimens agree closely in body color, the 
one from Butandiga has a much more uniformly black tail than the 
other, in which the color of the tail is about that of the body. 

Measurements, d\ (Kaimosi) 620. 360. 117. 39 mm. 

Diet. The stomachs of both held the remains of small rodents. 



64 bulletin: museum of comparative zoology 

Parasites. Ticks (Haemaphysalis leachii) were collected on the 
Butandiga mongoose. 

Ichneumia albicauda ibeana (Thomas) 

Herpestes albicaudus ibeanus Thomas, 1904, Ann. Mag. Nat. Hist. (7), 13, 
p. 409: Stony Athi, Kenya Colony. 

9 (M. C. Z. 32255) Kirui, K. C. 31. i. 34. 

lc?49 (M. C. Z. 31601, 32252-4, 32258) Kaimosi, K. C. 12. ii-2. hi. 34. 

9 (M. C. Z. 31958) Kibwezi, K. C. 28. iii. 34. 

cT (M. C. Z. 31959) Tsavo, K. C. 3. iv. 34. 

Distribution . This race occurs on Mount Mbololo where a freshly 
taken pelt was seen. 

Native names. Waranyet (Kimasai); usambaruo (Lugishu); kinueli 
(Luragoli); linueli (Lutereki); mwalasangali (Kitaita). 

Coloration. Heller (1913, Smithsonian Misc. Coll., 61, No. 13, 
p. 11) distinguished the white-tailed mongoose of the coastal area 
as fcrox with Changamwe, near Mombasa, as type locality. Hollister 
(1918, p. 130) is inclined to doubt its validity and places it in the 
synonymy. The two specimens from Kibwezi and Tsavo listed above, 
are, however, distinctly lighter than those from Kaimosi and Kirui, 
with less of the long black tips to the hairs of the dorsal region, so 
that possibly it may be recognizable when series are compared. 
Melanism is common in the entire area, making an estimate of shades 
of dark or light rather more difficult. Two of the series, from Kirui 
and Kaimosi, are unusually dark, even the tails being almost wholly 
black with the exception of the gray-based hairs at their proximal 
end. 

Measurements, cf. (Tsavo) 560. 410. 101. 38 mm., 9 . (Kirui) 570. 
450. 123. 45 mm. 

Diet. At Kaimosi one mongoose had a rat (Otomys t. elgonis) and 
the scales of a viper (Bitis nasicornis) in its stomach, another a few 
small bones, snake's eggs, what were apparently lizard's eggs, together 
with a mass of beetle and other insect remains. The stomach of the 
Kibwezi mongoose was crammed with grasshoppers. The Tsavo 
animal had some scraps of meat but mostly vegetable matter. I shot 
this mongoose as it was sniffing at the carcass of a hyrax at 12.30 a.m. 
A second mongoose came at 4 a.m. but I missed it. 

Parasites. Nematodes (Physaloptera sp). were found in the stomach 
of a Kaimosi mongoose. 



ALLEN AND LAWKENCE : AFRICAN MAMMALS 65 

Helogale undulata rufula Thomas 

Helogale undulata rufula Thomas, 1910, Ann. Mag. Nat. Hist. (8), 5, p. 194: 
Rogoro, Kikuyu, Kenya Colony. 

9 (M. C. Z. 32132) Tsavo, K. C. 31. hi. 34. 

9 (M. C. Z. 32330) Mt. Mbololo, K. C. 26. iv. 34. 

Native name. Munuru (Kitaita) . 

Coloration. The small mongooses of this genus seem rather vari- 
able in coloration, but it is not altogether clear as yet how far this 
is an individual matter. 

Measurements. 9 . (Mbololo) 460. 165. 43. 20 mm. 

Breeding. 9 juv. (Tsavo) 180. 140. 43. 16 mm. 

Habitat. The young mongoose was shot in the head with dust-shot 
from a .22 at a range of ten feet as it peered at me from the base of a 
thornbush into which it had run. 

Mungos mungo colonus (Heller) 

Crossarchus fasciatus colonus Heller, 1911, Smithsonian Misc. Coll., 56, No. 17, 
p. 16: Southern Guaso Nyiro, Kenya Colony. 

9 (M. C. Z. 32331) Voi, K. C. 17. iv. 34. 

Native name. Munuru wa sanga (Kitaita) . 

Distribution. In addition to the single specimen from Voi, a cranium 
was picked up in a cave near Kemp's Cave above Kirui, southeast 
face of Mount Elgon. The Banded Mongoose seems to be much less 
common in Kenya Colony than it is farther south. 

Measurements. 9 juv. 342. 222. 72. 21 mm. 



FELIDAE 

Felis (Leptailurus) capensis hindei Wroughton 

Felis capensis hindei Wroughton, 1910, Ann. Mag. Nat. Hist. (8), 5, p. 205: 

Machakos, Kenya Colony. 
Felis capensis kempi Wroughton, 1910, Ann. Mag. Nat. Hist. (8), 5, p. 206: 

Kirui, Mount Elgon, Kenya Colony. 

c? (M. C. Z. 31620) Kaburomi, U. 28. xii. 33. 
9 (M. C. Z. 31950) Golbanti, K. C. 25. vi. 34. 

Native names. Korobat (Kisabei); lutuku (Lugishu); indamamweli 
(Luragoli and Lutereki). 



66 bulletin: museum of comparative zoology 

Coloration. Hollister (1918, p. 175) has indicated that there seem 
to be no good reasons for recognizing a Mount Elgon race of serval, 
distinct from that of the lowlands to the southeast. The male secured 
by Loveridge at 10,500 feet on Mount Elgon is quite like one in the 
Museum collection from Kijabe, representing hindei with a similarly 
ochraceous-buff background above. On the other hand, the Golbanti 
skin from the coast is very much paler in the general tone, with pale 
buft'y sides and feet, shading to nearly ochraceous buff in the very 
middle line of the back. It is a very good match for another specimen 
from Mwanza, south shore of Lake Victoria, so that the series at 
hand shows no clear distinction of a paler coastal and a richer-colored 
inland form, though possibly of a paler, more southern one. We have 
for the present, therefore, tentatively referred both to hindei. 

Measurements. d\ 870. 280. 185. 86 mm., 9 . 820. 300. 190. 88 mm. 

Breeding. Regretfully one noted that the Golbanti serval, dis- 
turbed in long grass and shot on the run, was nursing a family. 

Diet. Feet and fur of a mole rat (Tachyoryctes ruddi) in the Kabu- 
romi animal, two rats {M ostomy s c. hildebrandti) and two Harlequin 
Quail (Coturnix delegorguei) in the Golbanti specimen. 

Parasites. Hippoboscid flies (Hippobosca longipennis) were very 
abundant in the fur of the female. 

Felis ocreata nandae Heller 

Felis ocreata nandae Heller, 1913, Smithsonian Misc. Coll., 61, No. 13, p. 14: 
Lukosa River, Nandi Escarpment, Kenya Colony. 

2^69 (M. C. Z. 32263-70) Near Lukosa R., Kaimosi, K. C. 
14. ii-7. iii. 34. 

Native names. Lug alio (Luragoli) ; skitarongo (Lutereki). 

Coloration. This topotypic series shows much variation in the tint 
of the occiput and ears, much more mixed with black in some than in 
others, while the backs of the ears are nearly red or in others black- 
ish. In two new-born kittens the pattern of indistinctly-marked 
cross-stripes and spots of black is much more obvious than in the 
adults which tend to lose all the body markings, and develop a nearly 
uniform buffy-gray coat with faint indications of the stripes; in the 
kittens the fine longitudinal black stripes are clearly marked but dis- 
appear in the adults. 

Measurements. <? . 505. 300. 117. 57 mm., 9 . 570. 352. 130. 64 mm. 

Breeding. On March 6, 1934, two kittens, apparently only born 
that day were brought in, they measured: cf. 152. 70. 36. 15 mm., 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 67 

9 . 155. 70. 36. 15 mm. On February 25, a half-grown kitten was 
collected. 

Diet. Four swamp rats (Dasymys h. helukus and Otomys t. clgonis) 
were recovered from the stomachs of three cats, those of all the rest 
were empty. 

Enemies. The natives frequently find the kittens of the wild tabby 
and take them to their huts where they remain for a time. One 
such, judging by its cropped ears, had evidently been in captivity for 
some time. 

LORISIDAE 

Perodicticus potto ibeanus Thomas 

Perodidicus ibeanus Thomas, 1910, Abstr. Proc. Zool. Soc. London, No. 81, 
p. 17: Kakamega Forest, Kenya Colony. 

3 cf 2 9 (M. C. Z. 31117-21, 31720) Kaimosi, Kakamega Forest, 
K. C. 23. ii.-9. iii. 34. 

Distribution. The potto has never been recorded from Mount 
Elgon, but the two Bagishu skinners from Sipi, western slopes of 
Elgon, were most emphatic that it is known to them. 

Native names. Likene (Luragoli) ; shakami (Lutereki). 

Coloration. Considerable variation in color is displayed by this 
series of topotypes; Hollister (1924, p. 11) pointed out the same thing 
with regard to his series of five from this locality. The general ground 
color varies from a distinctly grayish buff* to a pale ochraceous, much 
mixed with black above, especially on the shoulders. The white tip- 
ping of the hairs, if present, results in a frosted appearance, but this 
is characteristic of immaturity according to Schwarz (1931, Ann. 
Mag. Nat. Hist. (10), 8, p. 249). It might be pointed out that the 
correct spelling of the type locality is Kakamega, not Kakumega, 
and it is in Kenya Colony, not Uganda as stated by Schwarz. 

Measurements, c? . 340. 65. 75. 24 mm., 9 . 350. 74. 80. 29 mm. 

Diet. Dr. P. J. Darlington has kindly examined the stomach con- 
tents of one of these animals and reports it to consist of: Fragments 
and whole examples of some fifteen to twenty ants; pieces of several 
beetles, probably Scarabaeidae; hair; cotton or plant fibres; plant 
fragments resembling grasses. 

Notes. On the night of our arrival at Kaimosi, February 7, I heard 
what I took to be a potto in a tree near my tent. A few nights later 
I definitely heard one in a tree over the tent, the next night at 8.30 p.m. 
it was startled into defaecating on the tent by a noisy party of natives 



68 bulletin: museum of comparative zoology 

returning homewards along the road. I ran outside with a flashlight 
and shone its eyes. These glowing eyes seemed very close, so fearful 
of damaging a valuable specimen I backed off thirty feet further 
before firing with No. 8 shot from an open barrel. The animal made 
off and daylight revealed that I was really too far away. 

I then offered a reward of two shillings (50 cents U. S.) for the 
first potto to be brought in but had to wait a fortnight before one 
turned up. This potto was alive and practically uninjured having 
received only a slight bite from a dog on its right hind foot. It was 
slow and clumsy on the ground but clambered quickly about trees. 
When put on small trees it invariably clambered downwards, head 
down, and went in search of a larger tree. Another potto showed 
that they can really make quite a respectable pace on the ground 
when so inclined. If molested, they would turn and snap with aston- 
ishing rapidity. 

The best account of the habits of this creature will be found in 
Pitman (1931, "A Game Warden among his Charges." London, 
pp. 158, 274). 

Parasites. A tick (Ixodes ugandanus) in the fur of one, and a nema- 
tode in the stomach. 



GALAGIDAE 

Galago crassicaudatus lasiotis Peters 

Galago lasiotis Peters, 1877 (1876), Monatsb. Akad. Wiss. Berlin, p. 912: 

Mombasa, Kenya Colony. 
Galago crassicaudatus lasiotis Schwarz, 1931, Ann. Mag. Nat. Hist. (10), 7, 

p. 41. 

d" 9 (M. C. Z. 31721-2) Mt. Mbololo, K. C. 19 & 28. iv. 34. 

Native name. Mwongagi (Kitaita). 

Coloration. Hollister (1924, p. 12) mentions a specimen from this 
locality which had a white tail tip. The amount of white at the tip 
is subject to variation. One of our specimens has the terminal third 
of the tail dark blackish brown, while in the other the distal third 
is a mixture of pale gray and dark brown but no white. 

Measurements. & . 260. 310. 50. 45 mm., 9 . 290. 370. 85. 45 mm. 

Habitat. The male was shot within a few hundred feet of the male 
Galago s. braccatus in the rain forest capping the mountain at 
4,800 feet. Both were shot the same night at about 7.30 p.m. by shin- 
ing their eyes, a rather difficult business for these galagoes turn away 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 69 

their heads or conceal their eyes very quickly after the light is turned 
upon them. The second specimen was killed by a native near the 
foot of the mountain at about 1,500 feet. 

Galago senegalensis albipes Dollman 

Galago braccatus albipes Dollman, 1909, Ann. Mag. Nat. Hist. (8), 4, p. 549: 

Kirui, Mount Elgon, Kenya Colony. 
Galago senegalensis albipes Schwarz, 1931, Ann. Mag. Nat. Hist. (10), 7, p. 41. 

tf" (M. C. Z. 31288) Sipi, U. 21. xii. 33. 
d" 9 (M. C. Z. 31286-7) Kirui, K. C. IS & 20. i. 34. 

Distribution. This, according to Schwarz, is the form of senegalensis 
inhabiting the " uplands of Kenya Colony west of the Rift Valley." 
In addition to the specimens listed above, the Museum of Compara- 
tive Zoology has one from Mwanza and another from thirty miles 
south of Tabora, Tanganyika Territory. These are so closely similar 
that they must be regarded as representing the same race, having 
uniformly gray backs, and the limbs (especially the hinder pair) 
faintly washed with buffy. 

Native name. Karara (Lugishu); kaerarut (Kimasai); makwinyet 
(Kitosh). 

Measurements, cf. (Kirui) 200. 270. 70. 46 mm., 9. 180. 230. 
65. 42 mm. 

Galago senegalensis braccatus Elliot 

Galago braccatus Elliot, 1907, Ann. Mag. Nat. Hist. (7), 20, p. 187: Tsavo 

River, near Mount Kilimanjaro, Kenya Colony. 
Galago senegalensis braccatus Schwarz, 1931, Ann. Mag. Nat. Hist. (10), 7, 

p. 41. 

9 (M. C. Z. 31724) Tsavo River, K. C. 31. iii. 34. 
rf" (M. C. Z. 31725) Mt. Mbololo, K. C. 19. iv. 34. 

Native name. Monarang (Kitaita). 

Coloration. The topotype has the outer side of the limbs and the 
middle of the belly washed with pale yellow. 

Measurements, cf . 160. 170. 26. 30 mm., 9 . 170. 240. 60. 42 mm. 

CERCOPITHECIDAE 

Cercopithecus nictitans schmidti Matschie 

Cercopithecus schmidti Matschie, 1892, Zool. Anz., 15, p. 161: Forest between 
Mengo and Mjongo, Uganda. 



70 bulletin: museum of comparative zoology 

Lasiopyga ascanius kaimosae Heller, 1913, Smithsonian Misc. Coll., 61, No. 17, 
p. 10: Upper Lukosa River, near the mission station of Kaimosi, Kenya 
Colony. 

3cf 10 9 (M. C. Z. 31984, 31988-96, 31998-32000) Upper Lukosa 
River, Kaimosi, K. C. 19. ii.-3. iii. 34. 

Distribution. This species has not been recorded from Mount 
Elgon but I observed two red-tailed monkeys, which I feel reasonably 
certain were of this race, in company of a troupe of blue monkey (C. 
m. stuhlmanni) at Sipi. 

Native names. Enhondo (Luragoli); ikhondo (Lutereki). 

Measurements, c? . juv. 220. 295. 76. 26 mm., 9. 490. 6S0. 125. 
25 mm. 

Breeding. One of the females shot on March 3 contained an embryo, 
two others were carrying young of din'erent ages, these measured : 
cf. 185. 271. 67. 27 mm., <?. 195. 295. 78. 22 mm. 

Enemies. Both Bagishu and Watereki eat these monkeys and from 
the skill with which the Watereki boys and men climb lianas in pur- 
suit of them, I judge that they frequently hunt them. 

Habits. I was puzzled, when hunting these monkeys, to note the 
absence of large males and the preponderance of juveniles. The 
Watereki who accompanied me, explained that after the first rush 
of the troupe the males invariably descended to the ground and made 
off through the underbrush unless the hunters were accompanied by 
dogs. We tested this statement and found it to be perfectly true. 
Still there remained the large proportion of young monkeys which, 
when surprised, made off with shrill piping cries. Undoubtedly the 
species is very prolific but there seems to be a tendency for the adult 
females to remain concealed in the foliage when the alarm is given 
and the youngsters make off. I also observed that where the cover 
in their immediate vicinity is scanty, the females make a dash for 
the nearest liana-smothered tree — which are plentiful enough in 
this magnificent forest — and there seek concealment. Familiar 
as I am with the ways of colobus and blue monkey, I am confident 
that neither of these will remain concealed to the extent practised 
by Schmidt's Monkey. I have known them to remain quietly with- 
out giving a sign as the lianas were roughly shaken to the accom- 
paniment of hoarse cries and shouts, nor even move until a climber 
was within twenty feet of them. On March 3, I shot eight with ten 
shots, losing none. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 71 

Cercopithecus aethiops johnstoni Pocock 

(Cercopithecus pygerythrus) johnstoni Pocock, 1907, Proc. Zool. Soc. London, 

2, p. 738: Moshi, south side of Mount Kilimanjaro, Tanganyika Territory. 

Cercopithecus aethiops johnstoni Schwarz, 1926, Zeitschr. f. Saugetierk., l,p.40. 

3 9 (M. C. Z. 31947-8, 31975) Kibwezi, K. C. 26. iii. 34. 
<? (M. C. Z. 31961) Golbanti, K. C. 22. vi. 34. 

Distribution. Also seen at Tsavo. 

Native name. Tsavow (Kitaita). 

Discussion.. The coastal specimens should be nearly like the sup- 
posed races tumbili (from Ndi) and contingua (from Changamwe). 
Both these, however, are regarded by Schwarz in his revision as 
synonyms of johnstoni, this race extending from the coast to the Rift 
Valley in Kenya Colony. 

Coloration. These skins are fairly uniform in color, though that 
of the Golbanti male is distinctly more buffy yellow along the sides 
than are those of the Kibwezi females. 

Measurements, d" . 390. 500. 118. 33 mm., 9 . 450. 530. 127. 33 mm. 

Habitat. These bold little guenons were shot from a large troupe 
feeding in a wild fig tree close to the station at Kibwezi. The Gol- 
banti monkey was one of two animals in a tree close to the rest house; 
on picking it up I was surprised to find a cord tightly about, and 
almost cutting into its waist. As nobody in the village claimed to 
have lost a monkey it seems possible that it had been caught in a 
snare and gnawed itself free. 

Cercopithecus aethiops callidus (Hollister) 

Lasiopyga pygerythra callida Hollister, 1912, Smithsonian Misc. Coll., 59, 
No. 3, p. 1 : South end of Lake Naivasha, Kenya Colony. 

cT juv. 9 (M. C. Z. 31997, 32001) Mt. Debasien, U. i. xii. 33. 
9 juv. (M. C. Z. 31614) Kind, Mt. Elgon, K. C. 1. ii. 34. 

Native names. Akwadogot (Karamojong) ; chokea (Kisabei); musoni 
(Lugishu). 

Measurements. 9 . 515. 560. 115. 33 mm. 

Breeding. This female was nursing a young male which was chloro- 
formed on January 1, 1934, when it measured 212. 270. 68. 28 mm. 

Notes. These guenons are common along the gallery forest fringing 
the Amaler River from 5,000 feet down to the plains of western Deba- 
sien. Several times troupes slept in the trees in the vicinity of our 
camp. On November 30 several were secreted in a small but dense 
clump of foliage in a tall tree at the edge of our camp clearing. They 



72 bulletin: museum of comparative zoology 

remained close until stones rattling through the foliage caused a small 
monkey to dash out, it was followed shortly afterwards by a monkey 
carrying a young one on her breast. I could have shot either but 
refrained. There were plenty of other monkeys in the trees round 
about but very wary. 

On returning to camp at noon on December 1, I was told that 
monkeys had been in the vicinity all morning; just at that moment a 
monkey left the tree from which we had dislodged the trio the pre- 
vious evening. Thinking that it was a solitary male I shot it, it dropped 
stone dead; when I picked it up I found that it was a nursing female. 
Shortly afterwards a young monkey started crying in a nearby tree. 
Blazio, a Baganda cook who was a splendid climber, brought it down 
from the very topmost branches; as he ascended we saw the other 
young monkey leave the tree in full view. It seemed certain that this 
nursing female was being accompanied by her youngster from a pre- 
vious birth. 

From the very first the baby monkey, which I should judge was 
about a fortnight old, took to sucking milk from an improvised teat 
attached to a small bottle. At first he took half-a-bottle at a time 
but within a fortnight this was increased to two bottles. It was an 
amusing sight to see him standing on his hind legs and holding on to 
his bottle with both hands as he rapidly absorbed its contents. What 
a commotion was raised when the first bottle was finished before the 
second bottle could be substituted! Three weeks after capture he ate 
biscuits and banana with gusto. 

At first we found this monkey rather a nuisance as he wanted to be 
carried the whole time and squeaked continually to be picked up if 
left alone. While carried he looked at surrounding objects with great 
interest, turning his head this way and that. On safari he was tied 
up in a white cotton sugar bag, dropped into a haversack which was 
carried by a native. He appeared to appreciate this manner of travel- 
ling for he never cried and seemed willing to forego regular meal times 
so long as the motion of being carried continued. He sucked the 
fingers of his right hand continuously. Unfortunately he developed 
a complaint of the digestive tract which was rather stubborn and it 
seemed best to chloroform the little fellow. 

Cercopithecus mitis kibonotensis Lonnberg 

Cercopithecus albogularis kibonotensis Lonnberg, 1908, in Sjostedt, "Wiss. 
Ergeb. Schwed. Zool. Exped. Kilimandjaro, Meru umgeb. Massaisteppen." 
1, No. 2, p. 3: Kibonoto, Mount Kilimanjaro, Tanganyika Territory. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 73 

Lasiopyga albogularis maritima Heller, 1913, Smithsonian Misc. Coll., 61, 

No. 17, p. 8: Mazeras, Kenya Colony. 
Lasiopyga albogularis kima Heller, 1913, Smithsonian Misc. Coll., 61, No. 17, 

p. 9: Mount Mbololo, Taita district, Kenya Colony. 
Cercopithecus leucampyx kibonotensis Schwarz, 1928, Ann. Mag. Nat. Hist. 
(10), 1, p. 655. 

c? (M. C. Z. 31935) Kibvvezi, K. C. 24. hi. 34. 
c? (M. C. Z. 31937) Mt. Mbololo, K. C. 14. iv. 34. 
2 J* 3 9 (M. C. Z. 31936, 31962-3, 31973, 32288) Ngatana, K. C. 
11-15. vi. 34. 

Native names. Ngima (Kitaita); chima (Kipokomo). Being cor- 
ruptions of the Kiswahili kima. 

Discussion. The specimen from Mount Mbololo is a topotype of 
Heller's race, kima, but though Hollister (1924, p. 36) treated kima 
as well as the supposedly paler race maritima as distinct, Schwarz 
(1928, p. 655) relegates both to the synonymy of kibonotensis. 

Coloration. In our Mbololo specimen the white collar is much less 
nearly complete than is the case with the Kibwezi monkey where it 
lacks about 35 mm. of forming a complete ring. 

Measurements. & . (Mbololo) 595. 660. 150 40 mm., 9 . 470. 670. 
120. 37 mm. 

Parasites. Nematodes (Abreviata sp.) were present in Ngatana 
specimens. 

Enemies. Deadfall traps are employed by the Wapokomo to kill 
these monkeys which they prize as an article of diet. 

Habits. These monkeys are very abundant on the summit of Mount 
Mbololo and, strangely enough, are much less wary than the other 
creatures of the forest, be they squirrels, blue duiker or hyrax. It 
may be that they are no longer hunted for food by the Wataita. With 
a little cautious stalking, one might be certain almost any morning 
of surprising a party feeding, always a pleasant sight, and watch them 
until detected by some member of the troupe when all would go bound- 
ing away for a short distance. Vocally they were very quiet, their 
presence only being betrayed by movements of the foliage or the 
dropping of rejected fragments. Occasionally the piping whistle of 
a young monkey would be heard or more rarely still the deep grunt 
of a male who had detected the intruder from afar. 

Cercopithecus mitis stuhlmanni Matschie 

Cercopithecus stuhlmanni Matschie, 1893, Sitz. Ges. naturf. Freunde Berlin, 
p. 225: North of Kingawana, between Lakes Albert Edward and Albert, 
Belgian Congo. 



74 bulletin: museum of comparative zoology 

Cercopithecus leucampyx elgonis Lonnberg, 1919, Rev. Zool. Afr., 7, p. 134: 
Mount Elgon. 

9 9 (M. C. Z. 31986, 32003) Sipi, U. 21. xii. 33. 
9 9 (M. C. Z. 32002, 32004) Butandiga, U. 10. i. 34. 
c? 9 9 (M. C. Z. 31615, 31985, 31987) Elgonyi, K. C. 25. i. & 
4. ii. 34. 

Distribution. Mount Elgon apparently marks the northeastern 
limit of the range of this subspecies. 

Native names. Sivul (Kisabei); ekobe (Lugishu); sibolit (Kimasai); 
kikutusi (Kitosh); imwawihondo (Lutereki). 

Discussion. Schwarz (1933, Zeit. fur Saugetierk., 8, p. 278) has 
shown that mitis should replace leucampyx as the specific name for 
the monkeys of this group. Previously he stated that Lonnberg's 
elgonis is a synonym of stuhlmanni, a conclusion in which we would 
concur after comparing the seven Elgon skins listed above with one 
from the forests of eastern Ruwenzori. 

Measurements, cf. juv. 305. 423. 105. 77 mm., 9 . (Sipi) 570. 730. 
140. 38 mm. 

Breeding. On January 10, I fired at a monkey hiding in dense 
foliage and it proved to be a nursing female with a young one ( 9 . 
285. 362. 82. 30 mm.), both were killed outright with No. 3 shot in the 
head. There was a fetus in a female killed at Elgonyi on February 4, 
1934. 

Enemies. The blue monkey is much hunted by the Bagishu for its 
flesh; at Sipi a small crowd collected while the animals were being 
skinned. The bodies were carried off by a wrangling party of natives 
who all but fought over the division of the meat, later one encoun- 
tered individuals carrying off limbs carefully wrapped in banana 
leaves. At Butandiga the carcasses were similarly in great demand. 

Habits. As with allied races it was found that these blue monkeys 
associated with troupes of colobus, apparently for mutual protection. 

Cercopithecus ndglectus Schlegel 

Cercopithecus neglectus Schlegel, 1876, Mus. des Pays-Bas, Simiae, p. 70: 
White Nile(?). 

J* (M. C. Z. 31616) Kirui, K. C. 5. ii. 34. 

Distribution. The type of this species, now in the British Museum, 
was secured by Petherick on the White Nile, but Schwarz doubts if 
this is the actual place of its origin. Elliot refers to other specimens 
in the British Museum from north of Lake Rudolf. The general range 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 75 

is thus across the great forest westward to the mouth of the Congo, 
whence, near Brazzaville, came the type of brazzae, regarded by 
Schwarz as a synonym. Our specimen appears to be the first recorded 
from Mount Elgon, Kind's village being on the southern slopes. 

Discussion. The skull, compared with that of C. m. stuhlmanni, 
shows a number of minor differences, particularly in the width of the 
posterior narial opening and in the form of the audital bullae which 
converge to form a deep keel anteroventrally. 

Coloration. This Elgon example has the tail black mixed with 
whitish hairs quite to the tip, instead of being all black as in two 
specimens from the Cameroons and a third from southeastern Congo. 
Possibly after all, brazzae may prove to be a recognizable western race. 

Measurements. cT. 595. 630. 148. 38 mm. 

Note. This monkey was killed by a native, aided by his dog, and 
brought into camp in the flesh. 

Cercocebus galeritus galeritus Peters 

Cercocebus galeritus Peters, 1879, Monatsb. Akad. Wiss. Berlin, p. 830, pis. iB 
and iii: Miatola, mouth of Osi and Tana Rivers, Kenya Colony. 

cf (M. C. Z. 31934) Weraa, Ngatana, K. C. 16. vi. 34. 

Distribution. Very little is known of this mangabey, which seems 
to be confined to the gallery forest along the lower reaches of the 
Tana River. It is the only East African representative of the genus 
east of the general bounds of the central African forests. 

Native name. Garawa (Kipokomo). 

Measurements, cf . 600. 620. 158. 39 mm. 

Diet. According to the Wapokomo these animals raid their rice 
crops but only when such are in close proximity to the forest. 

Habits. At dawn one hears the deep-toned bark of the old males 
soon to be followed by the squeals of the younger mangabeys. The 
troupes keep much to the ground in the forest though ascending the 
trees to feed, or when disturbed, to see who is coming. Having located 
the danger, they drop to the ground and run off to the accompani- 
ment of an uproar such as one associates with baboons. Sections of 
the forests were waterlogged and here I encountered the mangabeys 
in the trees as I waded about in the knee-deep water. 

Papio furax Elliot 

Papiofurax Elliot, 1907, Ann. Mag. Nat. Hist. (7), 20, p. 498: Lake Baringo, 
Kenya Colony. 



76 bulletin: museum of comparative zoology 

d 1 (M. C. Z. 32006) Kirui, K. C. 6. ii. 34. 
9 (M. C. Z. 31619) Kaimosi, K. C. 19. ii. 34. 

Native names. Echum (Karamojong) ; nyanya (Kimasai); ngugi 
(Luragoli and Lutereki). 

Discussion. These are referred to P.furax, although until a general 
review of the baboons can be made, it is not possible to make a final 
estimate of the validity of some of the forms. According to Hollister 
(1924, p. 18) this species "is readily distinguishable by the shortness 
of the rostral portion of the skull" from other East African forms. 
In the Kaimosi skull the distance from the rim of the orbit to the tip 
of the premaxillary is 99 mm. 

Coloration. This Kaimosi specimen, an adult female, has the cen- 
tral area of the backs of hands and feet blackish with a few ochraceous 
hairs, and bordered by a paler area. The general color is ochraceous 
mixed with black. 

Measurements. & juv. 340. 255. 115. 52 mm., 9 . 710. 350. 181. 
52 mm. 

Diet. Green vegetable matter in both stomachs, maize being recog- 
nizable. 

Folklore. The following tale is from the Maragoli of Kaimosi. 
One day a man named Madugi was digging a pit in the forest in which 
to trap animals. As he was digging, some baboons passed by and 
hailed him. He answered " Yoo," then the baboons asked why was 
he digging. "I am digging for rats," replied Madugi. "Perhaps you 
are digging a trap in which to catch us and other animals," suggested 
the baboons. The man denied this, asserting that he was only digging 
for rats. However the baboons would not believe him, saying, "We, 
ourselves, know that indeed it is a trap which you are digging, come 
now we will fight with you over this." "With what kind of sticks 
shall we fight?" asked Madugi. The baboons answered, "We will 
fight with tsikhuvu (a species of shrub with many leaves)." The man 
agreed for he knew that tsikhuvu could not harm anyone. 

Madugi then procured a branch of tsikhuvu and concealed a sword 
among the leaves. When he was ready he invited the baboons to 
select one of their number to fight him, and this being done, cried, 
" Come on and fight, you may begin." The baboon, armed with a 
luhuvu (singular, i.e. one branch) hit the man and Madugi struck 
back, his hidden sword piercing the baboon. "W r ith what are you 
cutting me," cried the baboon. Madugi lied, saying, "Nothing, this 
is only l-sikh^u.'' They continued the fight and as before the baboon 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 77 

was wounded by the concealed weapon. This time the baboon cried 
out, "You are cutting me with a sword." "True," said Madugi, 
" I have a sword." As soon as they heard this all the other baboons 
fled and Madugi killed the baboon with whom he had been fighting. 
Then the man returned home and boasted of the fight in which he 
had killed a baboon by deceit and treachery. 

One day, a long time ago, a certain woman took her baby with her 
when she went to work in her garden. She did this because she bad 
no little nurse girl with whom to leave the baby. On arriving at the 
garden she laid the sleeping child in the shade of a tree. Then she 
proceeded to plant the grain. Later, when the mother returned to 
get the child she failed to find it. She thought that perhaps she had 
forgotten just where she had placed it, and was hunting round about 
when a baboon calling to her, said, " I, myself, stole your child." 
Then the mother answered, " Oh, please give me my child, I am ready 
to return home." The baboon answered, "No, I, myself, will not 
give you your child because I like to look at it." 

The distracted woman returned home to tell her husband. On 
hearing the news the man called together the villagers and said, " My 
wife took our child with her when she went to work in the garden, and 
now a baboon has stolen the child." The neighbours replied, "Ask 
your wife to show us this baboon." The woman accompanied them 
and pointed out the tree into which the baboon had climbed with 
the child. On seeing the people peering up into the tree, the baboon 
called to them, saying, "What will you do if I kill this child." The 
people answered, "You had better give up the child." The baboon 
replied, "Leave the child and let me play with it." On hearing this 
some of the neighbours said, " It would be well to shoot the baboon 
with an arrow and recover the child." Others objected, saying, "It 
would be unwise to kill the baboon, let him play with the child until 
he is tired." The mother of the child also agreed, saying, " It would 
be best to leave the child with him." It was now about six o'clock in 
the evening when the people started back to their homes leaving the 
child with the baboon. 

About two hours later the baboon brought the child and placed it 
outside the door of the mother's hut, then it took some water and 
poured it over the child so that the baby cried. The baboon called 
out, "I, myself, have brought your child." Then the mother knew 
that her baby had been returned alive and there was great rejoicing 
in the home that night. 



78 bulletin: museum of comparative zoology 

Papio ibeanus Thomas 

Papio thoth ibeanus Thomas, 1893, Ann. Mag. Nat. Hist. (6), 11, p. 46: Lamu, 
Kenya Colony. 

<? (M. C. Z. 31949) Voi, K. C. 11. iv. 34. 
Skull (M. C. Z. 32287) Kitau, Manda Id., K. C. 15. v. 34. 

Type locality. This baboon does not occur at Lamu township on 
Lamu Island except in captivity. It is, however, exceedingly abundant 
on the mainland which was formerly known as Lamu district, or 
loosely " Lamu' ' in the vague way that " Zanzibar' ' was also applied 
to the mainland opposite though with better justification. 

Distribution. Also seen at Kibwezi, Tsavo and Witu. In the dom 
palm forests just north of Witu they were both numerous and 
bold, finding a plentiful supply of fruit in the mango trees scattered 
through the bush, eloquent testimony to the fact that much of this 
region was under cultivation in the slave era. 

Native name. Fubi (Kitaita). 

Discussion. The topotypic skull from Manda Island, Lamu dis- 
trict, is that of a very old male. As the result of an injury, or possibly 
caused by an abscess resulting from a broken canine, considerable 
necrosis has taken place in the right suborbital region. The edges 
of the bone have become somewhat smoothed with a slightly irregular 
bony deposit at the front and rear edges of the lesion. In this old 
male, the sutures have all fused, even those of the nasals are quite 
obliterated. The teeth are worn but in good condition, except that 
the crown of the middle molar below the wound has gone, and the 
right canine is broken off. The last molar of the left side is missing 
and its alveolus wholly resorbed. 

The specimen from Voi is provisionally regarded as the same 
although the rostrum of the skull is considerably shorter, being 109 
mm. from orbit to tip of premaxillary as against 124 mm. in the big 
male from Kitau. Its coat is pale buffy gray, only slightly mixed with 
black, but it is rather worn and perhaps faded. 

Measurements. & . 800. 620. 315. 70 mm. 

PITHECIDAE 

Colobus polykomos matschiei Neumann 

Colobus matschiei Neumann, 1899, Sitz. Ges. naturf. Freunde Berlin, p. 15: Kwa 

Kitoto, near Kisumu, Kenya Colony. 
Colobus occidentalis matschiei Hollister, 1924, U. S. Nat. Mus., Bull. 99. p. 45. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 79 

Colobus abyssinicus elgonis Granvik, 1924, Lunds Univers. Arsskr. N. F., 21, 
No. 3, p. 4: Mount Elgon, Kenya Colony. 

c? ((M. C. Z. 31105) Mt. Debasien, U. 16. xi. 33. 
a* 9 (M. C. Z. 31106, 31617) Elgonyi, K. C. 25. i. 34. 
<? (M. C. Z. 32005) Nandi Forest, K. C. 14. ii. 34. 

Distribution. Said not to occur at Sipi and Butandiga on western 
Elgon but in the forests above the latter locality. Formerly present 
in the Sipi forests but exterminated by the Bagishu hunters. 

Native names. Echumwa (Karamojong) ; mangesia (Kisabei); 
lilubis (Lugishu); mongcsiet (Kimasai); kendubisi (Kitosh); 
induviri (Luragolis); nduviri (Lutereki). 

Coloration. Both Elgonyi specimens have the white brow band 
wide and continuous; in the Debasien animal it is a trifle narrower. 
In the former the shoulder stripe nearly reaches the white of the 
cheeks, but in the latter it is separated from the cheeks by a wider 
space. 

Measurements. & . (Elgonyi) 650. 640. 180. 40 mm., 9 . 630. 630. 
168. 38 mm. 

Breeding. Immediately after I shot the Elgonyi female, a big young 
guereza, which had been concealed higher up in the same tree unknown 
to me, made off. Though not entirely weaned it was well able to take 
care of itself. 

The juvenile male (370. 420. 125. 32 mm.) from the Nandi Forest 
twenty-five miles east of Kaimosi, was being hugged by its mother, 
despite its large size. They separated when disturbed. Its stomach 
held only vegetable matter as far as one could see; there was no trace 
of milk. 

Enemies. Some Karamojong refused the meat of the Debasien 
colobus but it was eaten by a party of Acholi in search of work. 

Habitat. The Debasien colobus was shot from a party of half-a- 
dozen in the large wood between the last two hills at the western 
foot of the mountain. They were fairly plentiful and exceedingly 
tame on the mountain above 7,000 feet. 

In the Elgonyi forest, though their deep throaty cries were heard 
at least once a day, generally towards sunset, or in response to a 
shot, I should say that they are not common. A four-hour hunt only 
revealed two. 

Colobus badius rufomitratus Peters 

Colobus rufomitratus Peters, 1879, Monatsb. Akad. Wiss. Berlin, p. 829, pis. 
ia and ii: Muniuni, Kenya Colony. 



80 bulletin: museum of comparative zoology 

Colobus badius rufomitratus Schwarz, 1928, Zeitschr. fiir Saugetierk., 3, p. 95. 
3 d 1 1 9 (M. C. Z. 31938-41) Wema, Ngatana, K. C. 13 & 16. vi. 34. 

Distribution. The four specimens of this rare monkey are from the 
forests near Wema and Ngatana villages about thirty miles in a direct 
line from the mouth of the Tana River, therefore not far from the 
type locality. The type, a male, is in the Berlin Museum; in addition 
there are a pair from the lower Tana in the British Museum which, 
according to Schwarz, agree perfectly with the type. 

Native name. Mbalawasi (Kipokomo). 

Discussion. Schwarz (1928, p. 95) in his review of the red colobi, 
regards rufomitratus as a geographic race of C. badius, the distribution 
of which as a species, is now more discontinuous in eastern Africa 
than that of the black-and-white group for there seems to be no rep- 
resentative of it in Kenya Colony except in this limited area of gallery 
forest along the lower Tana. 

Coloration. Elliot (1912, 3, p. 123) in his Review of the Primates, 
states that tephrosceles, the Ruwenzori representative of this form, 
differs, among other points, in not having any black on the head 
between the lateral tufts. The Ngatana skins, however, are also 
without black in this area, the entire crown being uniformly rufous. 

Measurements, cf. 600. 650. 168. 35 mm., 9 . 535. 605. 157. 35 mm. 
All the males have the same length from snout to anus. 

Diet. The Wapokomo stated that these monkeys rarely descend 
to the ground and never molest their crops of corn and rice which 
abutted on the forests in which the animals were shot. 

Enemies. Neither the Wapokomo nor other natives would touch the 
meat of these guerezas. Probably freedom from molestation accounts 
for the quiet way in which these monkeys sunned themselves when it 
was reasonably quiet. 



ANOMALURIDAE 

Anomalurus jacksoni de Winton 

Anomalurus jacksoni de Winton, 1898, Ann. Mag. Nat. Hist. (7), 1, p. 251: 
Entebbe, Uganda. 

3 9 (M. C. Z. 32335-7) Sipi, U. 18-26. xii. 33. 

Distribution. It is not generally known that this scaly-tail occurs 
in Kenya Colony but a native-made skin was seen at Kaimosi, the 
animal having been killed locally. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 81 

Native names. Chebkowe (Kisabei) ; hapa (Lugishu) ; ekima (Kitosh) ; 
viusiondet (Kimasai); lisimba (Lutereki). 

Coloration. Two of these flying squirrels are quite as described by 
de Winton, dark gray above, paler on the forehead, and blackish 
towards the edge of the membranes; a black muzzle and band through 
the eye, and a blackish spot about the base of the ear. The third has 
the ear spot and crown of the head rusty reddish, slightly mixed with 
grayish, while the hind feet, tibia, as well as the surrounding mem- 
brane, and a small area posterior to the forearm, are also rusty. A 
slight tinge of the same is present along the middle half of the tail, 
and shows faintly in its blackish terminal tuft. 

Measurements. 9 . 380. 210. 62. 41 with a "wing" of 196 mm. 

Diet. The stomachs of two were full of a mealy substance and free 
of parasites. 



SCIURIDAE 

Heliosciurus rufobrachium nyansae (Neumann) 

Sciurus nyansae Neumann, 1902, Sitz. Ges. naturf. Freunde Berlin, p. 56: 
Kwa Kitoto, near Kisumu, Kenya Colony. 

9cfl9 (M. C. Z. 30761-2, 30764-70) Sipi, U. 13-20. xii. 33. 

9 (M. C. Z. 30773) Butandiga, U. 15. i. 34. 

3 <? 1 9 (M. C. Z. 30763, 30771-2) Elgonyi, K. C. 25. i. & 5. ii. 34. 

4 tf 3 9 (M. C. Z. 30760, 30774-8) Kaimosi, K. C. 19. ii.— 7. iii. 34. 

Distribution. Said by the natives to occur at Kaburomi (10,500 feet) 
in the alpine zone of Mount Elgon, but rare at this altitude. This 
squirrel, which is so common in the Elgon forests, like the local race 
of Protoxerus at Kaimosi, seems to represent a northeastward out- 
post of the species, which is characteristic of the rain forest areas. 

Native names. Kwereru (Kisabei) ; kau (Lugishu) ; einuna (Kitosh) ; 
gererut (Kimasai). 

Measurements, <? . (Sipi) 255. 240. 50. 17 mm., 9 - (Sipi) 242. 240. 
40. 17 mm. 

Breeding. At Sipi, on December 12, 1933, a single large embryo was 
preserved from the uterus of one female, while two embryos were 
present in a second. At Kaimosi, on March 7, 1934, I heard a subdued 
cry of kwek-kwek in the undergrowth near my tent and surprised two 
young squirrels about the size of a little male (145. 140. 46. 14 mm.) 
shot on February 25, 1934. 

The breeding season is evidently early in the year in this locality 



82 bulletin: museum of comparative zoology 

for Heller collected very young squirrels at Kaimosi towards the end 
of January. 

Parasites. Nematodes were numerous in the stomachs of Sipi and 
Kaimosi squirrels, those from the former locality have been identified 
by Dr. J. H. Sandground as Protospiura muricola and Strongyluris sp. 

Heliosciurus undulatus shindi Heller 

Heliosciurus rufobrachiatus shindi Heller, 1914, Smithsonian Misc. Coll., 63, 
No. 7, p. 7: Summit of Mount Umengo, Taita Hills, Kenya Colony. 

2 cf (M. C. Z. 32333-4) Mt. Mbololo, K. C. 14 & 19. iv. 34. 

Distribution. Mount Mbololo being just across the valley from 
Mount Umengo, these specimens are practically topotypes of the 
race described by Heller on the basis of a single squirrel. His notes 
are quoted by Hollister (1919, p. 12) to the effect that this form is 
confined to the remnant of forest capping the extreme summits of 
the Taita Hills where it is rare, for the type was the only individual 
seen during a fortnight's stay on Umengo. 

Native names. Heller states that shindi is the Wataita name of this 
animal whereas my information is that shindi is Kisagalla, and orosh 
the Kitaita name. 

Coloration. The two additional specimens listed above, bear out 
the characters originally claimed for this race, viz. that the underside 
differs from that of typical undulatus in its much paler buffy tint. 
The throat and upper chest as well as the lower abdomen are much 
mixed with whitish, forming conspicuous pale areas, as well. The 
squirrels of this group resemble multicolor in having a broad orange 
band to the hairs of the dorsum, concealed by the succeeding black 
band and pale tip. 

Measurements. d\ 238. 285. 53. 18 mm. 

Habitat. Both were shot within thirty yards of the same spot at 
4,800 feet on the very summit of Mount Mbololo, and were the only 
ones seen during a fortnight's camping on the summit. 

Heliosciurus multicolor elegans Thomas 

Heliosciurus multicolor elegans Thomas, 1909, Ann. Mag. Nat. Hist. (8), 4, 
p. 103: Mount Elgon, Kenya Colony. 

c? 9 (M. C. Z. 31282-3) Karita River camp, Karamojo, U. 9. xi. 33. 
4 c? 19 (M. C. Z. 31277-80, 31285) Mount Debasien, U. 14. xi-1. xii. 33. 

Distribution. This squirrel was not met with on Mount Elgon; it is 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 83 

probably found on the eastern slopes where the dry conditions approxi- 
mate more to those of Mount Debasien. 

Native names. Eles (Karamojong). 

Coloration. This is a pale squirrel with white underparts, and a 
pattern of dorsal hairs resembling that of Ileliosciurus undulatus in 
having a blackish base, followed by a wide orange-rufous band, then 
a black band and a pale tip. The orange-rufous band is concealed 
by the terminal parts of the coat. 

Measurements. &. 223. 224. 38. 14 mm., 9 . 215. 250. 46. 12 mm. 
Both from Mount Debasien, the Karita River squirrels being sub- 
adult. 

Diet. The Karita River specimens betrayed their presence by the 
dropping of acorns as they fed in a big tree beneath which my tent 
was pitched. 

Disease. The genital region of one male was diseased, charac- 
terized by a huge swelling which the native skinner reported as being 
full of aqueous matter. 

Habitat. All the Debasien squirrels were shot in gallery forest along 
the Amaler River or adjacent dry watercourses above 5,000 feet. 

Protoxerus stangeri bea Heller 

Protoxerus stangeri bea Heller, 1912, Smithsonian Misc. Coll., 59, No. 16, p. 2: 
Lukosa River, Kakamega Forest, Kenya Colony. 

9^59 (M. C. Z. 30746-59) Kaimosi, K. C. 9. ii.-7. iii. 34. 

Distribution. This fine series of topotypes was secured in the Kaka- 
mega Forest by the Lukosa River and neighborhood of Kaimosi. 
As mentioned by Hollister, the discovery of this race constituted an 
important eastward extension of the range of this western squirrel. 

Native names. Kisila (Luragoli); shiseera (Lutereki). 

Measurements, c? . 320. 317. 67. 25 mm., 9 . 315. 310. 65. 22 mm. 

Breeding. On February 12, 1934, a young cf. 157. 133. 43. 12 mm. 
was brought in. It had a pure white tail, totally unlike the adults. 

Diet. One male was shot as he descended a tree, head downwards, 
with a large apple-like fruit, measuring 50 mm. in diameter, in his 
mouth. 

MYOXIDAE 

Claviglis parvus parvus (True) 

Eliomys parvus True, 1893, Proc. U. S. Nat. Mus., 16, p. 601: Tana River, 
between the coast and Hameye, Kenya Colony. 



84 bulletin: museum of comparative zoology 

Graphiurus parvus parvus Hollister, 1919, U. S. Nat. Mus. Bull. 99, p. 154. 
<? (M. C. Z. 32066) Ngatana, K. C. 15. vi. 34. 

Native name. Neki (Kipokomo). 

Discussion. This solitary example of the dormice of this group is 
referred to True's species, of which it is a topotype, although it seems 
likely that it may prove to be conspecific with microtis when the inter- 
relationships of the dormice can be worked out more carefully. 

Coloration. Its short gray fur, white hind feet and less gray under- 
side are distinguishing characters from the group to which saturatus 
belongs. 

Measurements, cf. 83. 75. 15. 12 mm. 

Habitat. Captured in my tent which was pitched beneath a wild 
fig tree surrounded by grasslands and swamps, near Wema village. 

Claviglis saturatus (Dollman) 

Graphiurus microtis saturatus Dollman, 1910, Ann. Mag. Nat. Hist. 
(8), 5, p. 204: South face of Mount Elgon at 9,000 feet, Kenya Colony. 

9 (M. C. Z. 31589) Mt. Debasien, U. 21. xi. 33. 

2 & 12 9 5 yng. (M. C. Z. 31571-88, 31597) Butandiga, U. 8-14. i. 34. 

d> (M. C. Z. 31598) Elgonyi, K. C. 5. ii. 34. 

2^69 (M.C.Z. 31590-6, 31599) Kaimosi,K.C. 19. ii.-5.iii. 34. 

Distribution. The dormouse from Elgonyi is almost topotypic as it 
comes from the south face of Mount Elgon at 7,000 feet. 

Native names. Isene (Lugishu and Kitosh); kererut (Kimasai); 
kizuguzu (Luragoli); shitzigutzu (Lutereki). 

Coloration. In contrast to the series from Mount Elgon, which are 
unstained below, the specimen from Mount Debasien has the whole 
ventral side of the throat and upper chest, lips, lower cheeks and fore 
arms stained a brilliant rusty, probably from some food such as pollen 
on which it had been feeding. 

There is a slight amount of variation in the color of the dorsum, 
some grayer and darker, some faintly more ochraceous, and a few 
with a decided brownish tinge, perhaps a result of fading. The dark 
bases of the belly hairs show through prominently. As usual, many 
exhibit broken tails with the terminal hairs of the stump white. This 
large series includes a number of small, youngish examples. 

Measurements, cf. (Kaimosi) 98. 80. 17. 13 mm., 9. (Debasien) 
105. 90. 16. 15 mm. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 85 

Breeding. At Butandiga, on January 14, 1934, a single very young- 
dormouse (50. 30. 11.5 mm.) was brought in, its tail being rat-like, 
not heavily furred like those of the adults. Four were taken from 
another nest by a native who did not bring them to me until the fol- 
lowing day after two had died. An examination of their stomachs 
showed that they were still being suckled. The two live ones were 
hungry and twittering like small birds, warm cow's milk was offered 
them in the palm of one's hand and they commenced to lap almost 
immediately. This diet, however, even when diluted did not suit 
them too well for they developed digestive troubles and succumbed 
some three weeks later. 

Diet. The natives claim that these dormice are very fond of bananas. 

Enemies. It was also said that they were very abundant in the 
Maragoli country (Kaimosi and vicinity) but were being driven away 
or exterminated by the introduction of roof rats (Rattus r. kijabins). 

Habitat. During the fortnight that I was at Elgonyi I constantly 
urged the natives to bring in dormice as I was only two thousand feet 
below the type locality. But beyond stating that they were familiar 
with it as a dweller in their bananas at the foot of the mountain, they 
failed to secure any. 

My tent was pitched in a clearing of the forest about fifteen feet 
from the base of a giant, twisted, forest tree that reminded me of 
one at Madehani above Lake Nyasa. From that tree I had obtained 
a series of Clamglis s. collaris (Allen & Loveridge, 1933, p. 122) where 
they were associated with a tree rat (Hylomyscus weileri). For several 
nights after our arrival I visited this tree at Elgonyi with a flashlight 
but without result. I sent a native up to search for signs of rodents 
but he failed to find either tracks or nests. After we had been in camp 
a week we purchased a bunch of unripe bananas and hung them 
against the trunk of this tree to ripen. 

The following day, my wife drew my attention to a couple of ban- 
anas that had been slightly eaten. After a careful examination I 
dismissed the idea of fruit bats being the culprits, then, discovering 
a dropping wedged between two of the bananas, postulated a dor- 
mouse as the robber. I removed the bunch to another place, nailed 
a rat trap to the branch where the bunch had hung and baited the 
trap with a fragment of banana. Next morning the bait was gone but 
the trap unsprung. The day after the bait was again taken and the 
trap sprung. The third morning furnished the same result as the 
second nor could I detect hairs upon the cloth which was wound 
round the wire of the trap to prevent injury to the hoped-for specimen. 



86 bulletin: museum of comparative zoology 

An hour before sunset on the third day I was approaching the trap 
to reset it when I heard a rustle among the dry leaves among the 
buttress roots of the giant tree. Looking down, I saw the bright eyes 
of a rat {Praomys t. jacksoni) watching me, it was within eighteen 
inches of my feet and had doubtless just emerged from a cavity at 
the base of the tree into which it promptly disappeared. I removed 
the trap from above and set it at the entrance of the hole, then went 
to dinner. The meal over, I returned to remove a Praomys from 
the trap before resetting it. Five minutes later I heard a snap and 
going to the trap took out another Praomys. Having reset the trap, 
I put another down at a similar opening on the far side of the tree. 
I had not left ten minutes when both traps went off almost simul- 
taneously. In one was a third Praomys which, like its predecessors, 
had been killed by a blow on the back of the neck. In the other was 
a live dormouse held by one foreleg. 



CRICETIDAE 

Dipodillus pusillus (Peters) 

Gerbillus pusillus Peters, 1878, Monatsb. Akad. Wiss. Berlin, p. 201: Ndi, 

Taita, Kenya Colony. 
Dipodillus percivali Dollman, 1914, Ann. Mag. Nat. Hist. (8), 14, p. 488: 

Voi, Kenya Colony. 

1 & 12 9 (M. C. Z. 32088-91, 32094-7, 32100-4) Voi, K. C. 9. iv. 34. 
2d 1 3 9 (M. C. Z. 32092-3, 32098-9, 32105) Mt. Mbololo. 23. iv. 34. 

Distribution. The series from Voi are topotypes of percivali which, 
as previously supposed by Hollister (1919, p. 25), is synonymous with 
pusillus of Peters. The Mbololo series are almost topotypic of the 
latter. 

Native names. Monguru (Kisagalla); corrupted to mongulu 
(Kitaita). 

Measurements. cT. (Voi) 98. 100. 20. 9 mm., 9 . (Mbololo) 85. 110. 
18. 10 mm. 

Breeding. About half the series of eighteen gerbils are immature. 

Enemies. One was recovered from the stomach of a sand boa {Eryx 
c. loveridgei). 

Tatera victna vicina (Peters) 

Gerbillus vicinus Peters, 1878, Monatsb. Akad. Wiss. Berlin, p. 200: Kitui, 
Ukamba, Kenya Colony. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 87 

Tatera mombasae Wroughton, 1906, Ann. Mag. Nat. Hist. (7), 17, p. 493: 
Takaungu, Kenya Colony. 

<? 2 9 (M. C. Z. 32116-7, 32119) Voi, K. C. 10. iv. 34. 
<? (M. C. Z. 32118) Mt. Mbololo, K. C. 28. iv. 34. 

Native names. Gue (Kisagalla); ngorobo (Kitaita). 

Discussion. The much more prominently tufted tail-tip and the 
slightly smaller size (hind foot 31-34 mm.) distinguish this gerbil from 
T. nigricauda which occurs in the same area. 

Measurements. d>. (Mbololo) 125. 170. 32. 18 mm., 9 . (Voi) 124. 
110. 27. 22 mm. 

Tatera nigricauda nigricauda (Peters) 

Gerbillus nigricaudus Peters, 1878, Monatsb. Akad. Wiss. Berlin, p. 200: Ndi, 
Taita, Kenya Colony. 

9 (M. C. Z. 32182) Kibwezi, K. C. 27. iii. 34. 

3^19 (M. C. Z. 32113-4, 32180-1) Voi, K. C. 11. iv. 34. 

c? (M. C. Z. 32115) Peccatoni, K. C. 24. v. 34. 

9 (M. C. Z. 32179) Ngatana, K. C. 13. vi. 34. 

3 <? (M. C. Z. 32120-1, 32178) Malindi, K. C. 29. vi. 34. 

Native name. The Wapokomo only apply the Kiswahili panya to 
this gerbil; the Wasagalla and Wataita do not differentiate it from the 
last species. 

Discussion. The adults agree in having the long tail black on its 
upper surface to the very tip, with only a slight admixture of paler 
huffy hairs towards the base. They differ from vicina in the much 
shorter hairs of the terminal tail tuft, and in the larger hind foot 
(36-43 mm.). The color of the face differs in being duller, a mixed 
grayish instead of clear bright buff in the paler areas. The character 
of the tail tuft in the young of both species is like that of the respective 
adults. 

Measurements. <?. (Voi) 160. 181. 34. 20 mm., 9 . (Ngatana) 160. 
182. 40. 22 mm. 

Breeding. Two of the Malindi series are juvenile, measuring 85. 101. 
31. 15 mm. 

Diet. The Kibwezi gerbil was trapped with cheese, all the others 
with bread. 

Tatera nigrita Wroughton 

Tatera nigrita Wroughton, 1906, Ann. Mag. Nat. Hist. (7), 17, p. 491 : Masindi, 
Unyoro, Uganda. 



88 bulletin: museum of comparative zoology 

9 (M. C. Z. 31351) Kirui, K. C. 18. i. 34. 
9 (M. C. Z. 31352) Kaimosi, K. C. 9. hi. 34. 

Distribution. These records constitute a slight eastward extension 
of the range which was formerly confined to Uganda. 

Native names. Oonga (Lugishu); muragutyet (Kimasai); emata 
(Kitosh). 

Discussion. These two specimens agree in being very much darker 
above than is usual in East African species of the genus Tatera. Their 
tails are dark above and buffy below, as described by Wroughton. 
The fore claws are notably long and stout, the skull with broader 
molars and larger bullae (11-11.5 mm. in length) as compared with 
those of T. nigricauda to which species they otherwise bear a slight 
resemblance. Although nigrita has been described as a small gerbil, 
our two examples are as large as T. n. nigricauda. 

Measurements. 9 . (Kirui) 162. 160. 22. 23 mm. 



RHIZOMYIDAE 

Tachyoryctes ruddi Thomas 

Tachyorydes ruddi Thomas, 1909, Ann. Mag. Nat. Hist. (8), 4, p. 546: Kirui, 
Mount Elgon, Kenya Colony. 

8^59 (M. C. Z. 30779-83, 30785-92) Sipi, U. 12-18. xii. 33. 

2 cf (M. C. Z. 30793, 30801) Kaburomi, U. 28. xii. 33. 

1 & (M. C. Z. 30799) Bukori, Kitosh, K. C. 18. i. 34. 

4J39 (M. C. Z. 30784, 30794-8, 30800) Kirui, K. C. 20. i. 34. 

Native names. Bumoongivet (Kisabei); unungwet (Kimasai); ifuko 
(Lugishu and Kitosh). 

Coloration. One young one is grayish, another black as were all the 
young ones rejected by Loveridge. This is the normal color of the 
young according to Hollister (1919, p. 41) as distinct from the brown 
pelage of the adults, though one adult has remained black. 

Measurements. & . (Sipi) 215. 64. 28. 10 mm., 9 . (Kirui) 210. 65. 
30. 8 mm. 

Enemies. The Wanderobo women are expert in handling these big 
mole rats for they are eaten as a matter of course. They can be picked 
up by the tail quite safely though to the accompaniment of shrill 
squeaks. They bite fiercel} r with their huge teeth at a stick or other 
object with which they may be restrained. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 89 

MURIDAE 

Dendromus insignis insignis Thomas 

Dendromus insignis Thomas, 1903, Ann. Mag. Nat. Hist. (7), 12, p. 341: 
Nandi, Kenya Colony. 

7 cf 4 9 (M. C. Z. 31211-2, 31214-22) Sipi, U. 13-22. xii. 33. 
9 (M. C. Z. 31208) Butandiga, U. 8. i. 34. 
c? 9 (M. C. Z. 31174, 31176) Kirui, K. C. 25. i. & 9. ii. 34. 
9^59 (M. C. Z. 31181-9, 31193, 31195, 31 197-8, 31203) Kaimosi,K.C. 
14-20. ii. 34. 

Distribution. Hollister (1919, p. 48) has already recorded this very 
common species of tree mouse from Kaimosi. Though given rank 
as a distinct species, it seems likely that this striped-back group is 
more closely related to the South African form. 

Native names. Chapchorogos (Kisabei); mandiosi (Lugishu); chep- 
toragopsi (Kimasai); mulubendi (Kitosh); kisie (Luragoli); isorodoni 
(Lutereki) . 

Measurements, cf. (Sipi) 90. 102. 21. 15 mm., 9. (Sipi) 90. 104. 
22. 15 mm. 

Enemies. One was recovered from the stomach of a tree viper 
(Atheris squamigera) at Kaimosi. 

Dendromus whytei pallescens Osgood 

Dendromus whytei pallescens Osgood, 1910, Publ. Field Mus. Nat. Hist., Zool. 
Series, 10, No. 2, p. 7: Ulukenya Hills, Kenya Colony. 

& 9 (M. C. Z. 31223-4) Sipi, U. 14 & 22. xii. 33. 

<? 2 9 (M. C. Z. 31210, 31230, 31232) Butandiga, U. 8-14. i. 34. 

4 + 19 (M. C. Z. 31175, 31180, 31225-7) Kirui, K. C. 6-9. ii. 34. 

c? 2 9 (M. C. Z. 31196, 31228-9) Kaimosi, K. C. 10 & 20. ii. 34. 

Distribution. This is a much smaller species than D. i. insignis 
but occurs alongside it as do also the following species. 

Measurements, d 71 . (Kaimosi) 77. 90. IS. 12 mm., 9. (Butandiga) 
76. 86. 18. 13 mm. 

Dendromus ruddi Wroughton 

Dendromus ruddi Wroughton, 1910, Ann. Mag. Nat. Hist. (8), 5, p. 275: 
Malakisi, Mount Elgon, Kenya Colony. 

2 d* 3 9 (M. C. Z. 31231, 31234-7) Butandiga, U. 8-14. i. 34. 

Type locality. The village of Malakisi is not on Mount Elgon but 
lies at the southern foot to the west of Bukori. I passed through it 



90 bulletin: museum of comparative zoology 

when motoring from Mbale to Bukori as it is on the mainroad skirt- 
ing the mountain. 

Native name. Tsuru (Lugishu). The Bagishu distinguish between 
this unstriped species and the others which display a dorsal line. 

Measurements, c? . 68. 90. 16. 12 mm., 9 . 60. 70. 15. 17 mm. 

Enemies. On January 13, three very young unstriped tree mice and 
seven adult striped D. acraeus were brought in alive. As we already 
had as much skinning to accomplish as we could handle, I put these 
mice into a large can with quantities of soft grass, bread, and cheese. 
Later in the evening I heard squeaks proceeding from the tin and 
on examining them found that all the three young had been killed, 
and partly devoured, by the adults. 

Dendromus acraeus Wroughton 

Dendromus acraeus Wroughton, 1909, Ann. Mag. Nat. Hist. (8), 4, p. 541: 
Kirui, Mount Elgon, Kenya Colony. 

1+5^29 (M. C. Z. 31204-7, 31209, 31233) Butandiga, U. 8-14. i. 34. 
3 young (M. C. Z. 31177-9) Kirui, K. C. 9. ii. 34. 
5 C? 2 9 (M. C. Z. 31190-2, 31194, 31199-202) Kaimosi, K. C. 14-20. 
ii. 34. 

Distribution. This species of small size and with an obsolete black 
median line, seems to be about as common as D. i. insignis and occurs 
in the same localities. 

Measurements, c? . (Kaimosi) 68. 80. 16. 10 mm., 9 . (Kaimosi) 73. 
105. 18. 12 mm. 

Zelotomys hildegardeae vinaceus Heller 

Zelotomys hildegardae (sic) vinaceus Heller, 1912, Smithsonian Misc. Coll., 
59, No. 16, p. 10: Mount Mbololo, Taita Hills, at 3,000 feet, Kenya 
Colony (amended, see below). 

d 1 9 (M. C. Z. 32086-7) Mt. Mbololo at 3,000 ft., K. C. 24-25. iv. 34. 

Type locality. In the original description, this reads: "Ndi, Mount 
Mbololo." As Ndi is a small railway station on the plains some dis- 
tance from the foot of the mountain, and an unlikely spot in which 
to find a rain-forest species, Mr. Heller was communicated with and 
confirms the supposition that "Mount Mbololo, near Ndi" was 
intended, as the mountain was so little known at that time. 

Discussion. These two topotypes agree precisely with the original 
description. Their close color resemblance to the local race of Mas- 
tomys has been remarked upon by Heller and Thomas though the 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 91 

former states that in life vinaceus is distinguishable by the pinkish 
tint of the paler parts of the tail. 

Osgood (1910, p. 7), when erecting the genus, gave the mammary 
formula as 2 — 2 = 8, but Thomas (1915, p. 4S1) found three pairs of 
pectorals in Congo specimens making 3 — 2 = 10. One of the above 
specimens is an adult female with well-developed mammae of which 
we are able to discover but two pectoral pairs, so that there is probably 
some variation in this respect. 

Nothing was learned of the habits of this interesting animal, which 
is still rare in collections. The slightly proclivous upper incisors, and 
rather shortened skull, may indicate that it is somewhat of a bur- 
rower. 

Measurements, tf 1 . 130. 105. 23. 16 mm., 9 . 130. 108. 22. 15 mm. 

Thamnomys surd aster polionops Osgood 

Thamnomys surdaster polionops Osgood, 1910, Publ. Field Mus. Nat. Hist., 
Zool. Series, 10, No. 2, p. 8: Ulukenya Hills, Kenya Colony. 

c? (M. C. Z. 32171) Mt. Mbololo, K. C. 17. iv. 34. 

Coloration. This single specimen is a decided shade darker gray, 
and a trifle less deep in its rusty ochraceous back, than a series from 
Tanganyika Territory considered as representative of typical sur- 
daster. Hollister (1919, p. 58), commenting on a series from Mount 
Mbololo, regards them as "somewhat intermediate between typical 
■polionops and the Kilimanjaro form" but "clearly nearest" to the 
former. 

Measurements, cf . 122. 170. 26. 15 mm. 

Diet. Trapped with bread as a bait at 4,800 feet. 

Thamnomys surdaster elgonis Thomas 

Thamnomys surdaster elgonis Thomas, 1910, Ann. Mag. Nat. Hist. (8), 5, p. 282 : 

Malakisi, south foot of Mount Elgon, Kenya Colony. 
Thamnomys surdaster discolor Thomas, 1910, Ann. Mag. Nat. Hist. (8), 5, 

p. 283: Kakamega Forest, Kenya Colony. 
Thamnomys surdaster insignis Dollman, 1911, Ann. Mag. Nat. Hist. (8), 7, 

p. 528: south face of Mount Elgon at 9,000 feet, Kenya Colony. 

2 9 (M. C. Z. 31384-5) Mt. Debasien, U. 22-23. xi. 33. 
d 1 juv. (M. C. Z. 31213) Sipi, Mt. Elgon, U. 20. xii. 33. 
<? (M. C. Z. 31386) Butandiga, U. 10. i. 34. 
9 (M. C. Z. 31387) Kirui, K. C. 28. i. 34. 
9 (M. C. Z. 31238) Elgonyi, K. C. 5. ii. 34. 
<? 9 (M. C. Z. 31382-3) Kaimosi, K. C. 19. ii. & 7. iii. 34. 



92 bulletin: museum of comparative zoology 

Distribution. The Kirui specimen is almost topotypic of elgonis, 
the Elgonyi of insignis, while the pair from Kaimosi are topotypes 
of discolor. We fail to see any reasons, taxonomic or geographic, 
for keeping them distinct. 

Native names. Araragct (Kisabei); sungama (Lugishu and Kitosh). 

Discussion. This thicket mouse of the moist, forested Elgon-Kaka- 
mega region is distinguished by the contrasted dark central marking 
on the metatarsals. In other respects it is very similar to typical 
surdaster, of which it should undoubtedly be considered a local race. 

Coloration. The two mice from Mount Debasien, as well as one 
from Mount Kenya and another from Nyeri in the Museum collection, 
differ from the rest of the series in wholly lacking any trace of the 
ochraceous buffy line bounding the white of the underparts from 
cheeks to ankle. Instead the line of demarcation is abrupt and the 
sides of the body buffy gray, backs of the feet pale ochraceous buff, 
and head and shoulders slightly grayer than the back. In skull char- 
acters and other respects, however, they show no significant differ- 
ences from the rest of the series. 

Measurements, cf. (Kirui) 112. 161. 22. 16 mm., 9. (Butandiga) 
125. 172.23. 17 mm. 

Breeding. The very young male (72. 90. 17. 13 mm.) was brought 
in alive at Sipi on December 20, 1934. 

Diet. The Elgonyi mouse was trapped with unripe banana as bait. 

Oenomys bacchante editus Thomas & Wroughton 

Oenomys bacchante editus Thomas & Wroughton, 1910, Trans. Zool. Soc. 
London, 19, p. 509: Mubuku Valley, Mount Ruwenzori, Uganda. 

7 cf 4 9 (M. C. Z. 31626-34, 31650-1) Sipi, U. 13-20. xii. 33. 
2 c? 2 9 (M. C. Z. 31635-8) Butandiga, U. 8-10. i. 34. 
5 d" 5 9 (M. C. Z. 31639-49) Kaimosi, K. C. 10-19. ii. 34. 

Native names. Bunwe (Lugishu); indioro (Luragoli); nangcti (Lute- 
reki). 

Coloration. This large series from Mount Elgon and Kakamega is 
referred to editus, following Hollister's (1919, p. 64) determination 
of those from the latter region in the United States National Museum. 
As was the case with his series, there is much individual variation 
in the depth and extent of the rufous areas. Two of the Butandiga 
rats have the outer and inner sides of the ears deep bright rufous, 
others from the same locality are dark-eared. An extreme variant 
from Mount Elgon has the whole of the posterior half of the back 
tinged with rufous. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 93 

Measurements, d". (Kaimosi) 180. 200. 30. 20 mm., 9. (Kaimosi) 
175. 185. 30. 20 mm. 

Enemies. At Kaimosi, where rusty-nosed rats were very abundant, 
they were recovered from the stomachs of the following animals and 
reptiles: Civet (Civettictis c. schwarzi), genet (Genetta s. stuhlmanni), 
tree civet (Nandinia b. arbor -ea), mongoose (Herpestes i. funestus), 
mamba (Dendraspis jamesoni) and nose-horned viper (Bitis nasi- 
cornis) . 

Rattus rattus kijabius (Allen) 

Mus kijabius J. A. Allen, 1909, Bull. Amer. Mus. Nat. Hist., 36, p. 169: Kijabe, 
Kenya Colony. 

<? juv. (M. C. Z. 31336) Mt. Debasien, U. 18. xi. 33. 
2^29 (M. C. Z. 31332-5) Sipi, U. 12. xii. 33. 
^29 (M. C. Z. 31337-9) Elgonyi, K. C. 24. i. 34. 
cT (M. C. Z. 32177) Golbanti, K. C. 23. vi. 34. 

Distribution. As Hollister (1919, p. 68) has pointed out, this house 
rat differs from the race inhabiting the southern United States and 
the Mediterranean region. It may prove to be a native variety or 
possibly have been introduced. 

Native navies. Miri (Karamojong) ; mabaja (Lugishu); gummabat 
(Kisabei); mabageet (Kimasai); kimbeba (Kitosh); lijunga (Luragoli); 
lichungu (Lutereki); ndeba (Kitaita); panya (Kipokomo and Kis- 
wahili). 

Coloration. Typically this is a dark, slaty-bellied rat with a dark 
gray back, not much mixed with brown. The male from Golbanti, 
Tana River, however, is pure white below to the roots of the hairs. In 
this respect it corresponds to the rural form of house rats in parts of 
Asia where the slaty-bellied forms are usually found about towns. 

Measurements. <? . (Sipi) 188. 206. 31. 23 mm., 9 . (Sipi) 180. 192. 
30. 22 mm. 

Breeding. This large Sipi female held nine big embryos on December 
12, 1933. 

Diet. The Debasien rat was trapped with meat bait in the tent 
after it had disturbed me several times by knocking things over. 
According to the Maragoli, when these rats arrived in the Kakamega 
country they drove out the dormice {Claviglis saturatus) which were 
formerly very common there. 

Enemies. A very big rat, measuring seven and a half inches from 
snout to anus, was recovered from the stomach of a thirty -two inch 
House Snake (Boaedon lineatus) at Kaimosi, a smaller rat was taken 



94 bulletin: museum of comparative zoology 

from a second House Snake in the same locality and a big one from a 
Puff Adder (Bitis arietans) at Bukori. 

Habitat. In corroboration of the authors' views given above, it is 
interesting to note that the white-bellied rat was taken in the rice 
fields situated some distance from the village of Golbanti. 

Aethomys kaiseri medicatus (Wroughton) 

Mus medicatus Wroughton, 1909, Ann. Mag. Nat. Hist. (8), 4, p. 540: Mumias, 
Kenya Colony. 

1 + c? 9 (M. C. Z. 31340-1, 31353) Mt. Debasien, U. 14-29. xi. 33. 

Distribution. Hollister (1919, p. 71) has recorded many examples 
of this subspecies from Uganda as well as from the Kakamega region. 

Native name. Lonang (Karamojong). 

Coloration. Two of the series are immature, the adult female agrees 
in its buffy sides and mixed buffy-and-black middorsal area with 
skins of this race from the Guaso Nyiro country. 

Measurements. & imm. 125. 121. 21. 19 mm., 9 . 160. 130+. 30. 
21 mm. 

Diet. The male was taken in a trap baited with mealie porridge. 

Praomys tullbergi jacksoni (de Winton) 

Musjacksoni de Winton, 1897, Ann. Mag. Nat. Hist. (6), 20, p. 318: Entebbe, 
Uganda. 

5 o* 5 9 (M. C. Z. 31389-93, 31415, 31476-9) Sipi, U. 14-20. xii. 33. 

9 (M. C. Z. 31394) Butandiga, U. 12. i. 34. 

2 cf 2 9 (M. C. Z. 31396-9) Elgonyi, K. C. 5-7. ii. 34. 

c? (M. C. Z. 31388) Kirui, K. C. 25. i. 34. 

9 (M. C. Z. 31395) Kaimosi, K. C. 25. ii. 34. 

Native names. Morong (Kisabei); changwaset (Kimasai); imbagula 
(Lugishu); isakula (Lutereki). 

Coloration. There is more or less individual variation in the amount 
of russet in the pelage, which seems to increase with age till in some 
specimens it is nearly clear rufous about the base of the tail. 

Measurements. &. 130. 145. 25. 21 mm., 9 . 130. 143. 25. 20 mm. 
Both these rats were from Elgonyi; at Sipi they averaged much 
smaller, viz. d". 120. 126. 24. 20 mm., 9 . 122. 135. 25. 18 mm., even 
so this male was much larger than the general run of males taken at 
Sipi. 

Diet. Trapped with banana at Elgonyi as related under Claviglis 
saturatus; with a cheese bait in my tent at Kaimosi. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 95 

Enemies. At Kaimosi one was recovered from the stomach of a 
mongoose (Herpestes i. fimestus), another from a tree viper (Atheris 
squamigcra) on February 24, 1934. 

Habitat. So abundant was this species in the forest surrounding 
the clearing where I camped above Sipi, that we took nine in ten 
traps put out the first night of trapping. 

Praomys taitae (Heller) 

Epimys taitae Heller, 1912, Smithsonian Misc. Coll., 59, No. 16, p. 9: Mount 
Mbololo, Taita Hills, Kenya Colony. 

cT 2 9 (M. C. Z. 32131, 32124-5) Mt. Mbololo, K. C. 18. iv. 34. 

Distribution. As this rat has only been taken from the Taita Hills 
it may have rather a restricted range. 

Discussion. This species seems to be quite distinct from the larger 
P. tullbergi and its races. P. taitae has a smaller body and feet and a 
much shorter skull in which the line of supraorbital beading is present 
as a barely indicated ridge. 

Measurements, cf juv. 62. 48. 17. 10 mm., 9 ad. 95. 120. 22. 18 mm. 

Breeding. This female and her two young were dug from among 
loose mold and dead leaves which had drifted between the huge but- 
tress roots of a giant tree at the lower edge of the forest at about 
4,000 feet. 

Mastomys coucha tinctus (Hollister) 

Rattus coucha tinctus Hollister, 1918, Smithsonian Misc. Coll., 66, No. 10, 
p. 1 : Kaimosi, Kenya Colony. 

3 d" 7 9 (M. C. Z. 31400-9) Mt. Debasien, U. 20-29. xi. 33. 

c? juv. (M. C. Z. 31780) Greeki River, U. 7. xii. 33. 

9 (M. C. Z. 31414) Kirui, K. C. 1. ii. 34. 

4 9 (M. C. Z. 31410-3) Kaimosi, K. C. 20. ii-8. iii. 34. 

Native names. Loyokomur (Karamojong) ; ikaria (Lugishu). 

Coloration. Hollister (1919, p. 89) defines this as a large race of 
dark tint whose underparts are only slightly paler than its sides. 
Most of the series listed above are not fully adult but the hair of their 
bellies, though with the usual gray bases, is rather contrastingly 
white-tipped. The rats forming the Mount Debasien series are also 
mostly subadult, but scarcely differ from the Kaimosi topotypes, 
though possibly a very little paler grayish above. 

Measurements, cf • (Debasien) 130. 108. 23. 19 mm., 9 . (Kaimosi) 
130. 112. 22. 18 mm. 



96 bulletin: museum of comparative zoology 

Breeding. The male from Greeki River, December 7, 1933, was 
one of four young (83. 80. 22. 17 mm.) trapped the same day but 
damaged by ants. 

Enemies. One was recovered from the stomach of a tree viper 
{Atheris squamigera) at Kaimosi on February 21, 1934. 

Mastomys coucha hildebrandtii (Peters) 

Mus hildebrandtii Peters, 1878, Monatsb. Akad. Wiss. Berlin, p. 200: Ndi, 
Taita, Kenya Colony. 

<? (M. C. Z. 32189) Kibwezi, K. C. 27. iii. 34. 
<? cf (M. C- Z. 32130, 32183) Peccatoni, K. C. 25. v. 34. 
c? 5 9 (M. C. Z. 32126-28, 32184-5, 32187) Ngatana, K.C. 12-19. vi. 34. 
9 9 (M. C. Z. 32186, 32129) Golbanti, K. C. 23. vi. 34. 
c? (M. C. Z. 32189) Malindi, K. C. 30. vi. 34. 

Distribution. The specimens from the Tana region are probably 
best referred to hildebrandtii. The immatures are less blue gray than 
are rats from near Mombasa which represent the race durumae. 

Native name. Panya (Kipokomo, who do not appear to have speci- 
fic names for rodents). 

Measurements. <?. (Malindi) 155. 182. 31. 21 mm., 9 . (Ngatana) 
138. 122. 22. 17 mm. 

Breeding. At Golbanti, June 23, 1934, a female, which had sixteen 
breasts in milk, together with her thirteen young (68. 60. 19. 13 mm.) 
was dug from a nest of grass situated about a foot beneath the sur- 
face, in clay, at the edge of a rice field. The sedges, which formerly 
covered the site of the nest, had been cut down and spread over the 
ground like a carpet. In addition to two entrances to the nest there 
were a number of blind alleys, some of which were probably for drain- 
age purposes. 

Enemies. Two in the stomach of a serval (Felis c. hindei) near 
Golbanti. 

Parasites. Mites were numerous on the mother. 

Leggada triton triton Thomas 

Leggada triton triton Thomas, 1909, Ann. Mag. Nat. Hist. (8), 4, p. 548: 
Kirui, Mount Elgon, Kenya Colony. 

3 <? 11 9 5 young (M. C. Z. 31482, 31498, 31530-6, 31552-61) 
Sipi, U. 14-22. xii. 33. 
3 cf 10 9 (M. C. Z. 31537-47, 31550-1) Butandiga, U. 8-13. i. 34. 

1 (M. C. Z. 31515) Kirui, K. C. 6. ii. 34. 
7 d" 4 9 (M. C. Z. 31548-9, 31562-70) Kaimosi, K. C. 9-21. ii. 34. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 97 

Distribution. Occurring alongside L. g. grata in all these localities, 
where both species are abundant. The topotype is without sex or 
measurements as it was obtained by a skinner, who could neither read 
nor write, who was sent to Kirui to ask for it specifically by its native 
name. 

Native names. Chepchom (Kisabei); mbuhu (Lugishu); livudu 
(Luragoli); shivudu (Lutereki). 

Measurements. <? . (Kaimosi) 85. 54. 15. 12 mm., 9 . (Kaimosi) 85. 
55. 14. 12 mm. 

Breeding. At Sipi, on December 20, 1933, three nestling young were 
brought in, on the 22nd another. 

Enemies. At Butandiga a female had a truncated tail, healed, 
though half was missing. One of these pygmy mice was recovered 
from the stomach of a house snake (Boaedon lineatus). 

LEGGADA BELLA BELLA Thomas 

Leggada bella Thomas, 1910, Ann. Mag. Nat. Hist. (8), 5, p. 87: Machakos, 
Kenya Colony. 

3 (M. C. Z. 31512-4) Kirui, K. C. 6. ii. 34. 

Distribution. This white-bellied pygmy mouse was apparently far 
less common in the Elgon region than the other two species, triton 
and grata. These three were obtained by Loveridge's skinner under 
the conditions mentioned above. 

Leggada bella vicina Thomas 

Leggada bella vicina Thomas, 1910, Ann. Mag. Nat. Hist. (8), 5, p. 88: Taka- 
ungu, near Mombasa, Kenya Colony. 

c? 9 (M. C. Z. 32190-1) Kitau, Manda Id., K. C. 16 & 18. v. 34. 

Distribution. Said to occur in houses at Lamu, Lamu Island, though 
Loveridge failed to obtain any during the week that he was there. 
Measurements, d". 57. 40. 11. 8 mm., 9 . 54. 48. 11. 10 mm. 
Habitat. Captured in a ruined hut. 

Leggada grata grata Thomas 

Leggada grata Thomas, 1909, Ann. Mag. Nat. Hist. (8), 4, p. 549: Mubuku 
Valley, Mount Ruwenzori, Uganda. 

5^993 young (M. C. Z. 31481, 31483-97, 32192) Sipi, U. 19-22. xh. 33. 
5^29 (M. C. Z. 31499-505) Butandiga, U. 8. i. 34. 
4 + 29 (M. C. Z. 31506-511) Kirui, K. C. 1-6. ii. 34. 
9 (M. C. Z. 31517) Elgonyi, K. C. 4. ii. 34. 
2 cf 10 9 (M. C. Z. 31518-29) Kaimosi, K. C. 9-20. ii. 34. 



98 bulletin: museum of comparative zoology 

Native names. Limwani (Luragoli); sinamutali (Lutereki). 

Coloration. This is a small grayish species with a buffy line separat- 
ing the dorsal coloring from the white of the belly. 

Measurements, tf . (Butandiga) 70. 55. 13. 10 mm., 9. (Sipi) 72. 
54. 12. 10 mm. 

Breeding. At Butandiga, on January 8, 1934, a female, measuring 
62. 52. 13. 10 mm., held three fetuses, measuring 9 . 41. 16. 8. and 
ear ? mm. 

Enemies. At Kaimosi one of these pygmy mice was recovered from 
the stomach of an European Kestrel (Falco t. tinnunculus) , two from 
House Snakes (Boaedon lineatvs) and one from a tree viper (Atheris 
squamigera). 

A Sipi male had no external trace of a right hind limb though 
within the apparently uninjured skin it was present to the knee; the 
condition would appear therefore to have been congenital rather than 
resulting from an attack. Despite this handicap the little animal was 
in good condition. Living must be particularly favourable for this 
species at Sipi, despite the great variety of larger rodents occurring 
there, for natives brought in twice the number preserved during the 
first week of mv stav. 

Cricetomys gambianus elgonis Thomas 

Cricetomys gambianus elgotiis Thomas, 1910, Ann. Mag. Nat. Hist. (8), 5, 
p. 198: South face of Mount Elgon at 10,000 feet, Kenya Colony. 

2c?19 (M. C. Z. 32219-21) Sipi, U. 18-22. xii. 33. 

1 tf 1 9 (M. C. Z. 32222-3) Kirui, K. C. 25. i. 34. 

6 cf 2 9 (M. C. Z. 32212-8, 32224) Elgonyi, K. C. 28. i-3. ii. 34. 

6 cf 4 9 (M. C. Z. 32225-32, 32261-2) Kaimosi, K. C. 14-21. ii. 34. 

Distribution. I was also shown one at Kaburomi, 10,500 feet, in the 
alpine zone, which rat was said to have been killed locally. The fact 
that it was decomposed, however, raises the possibility that it had 
been brought up the mountain from Sipi. 

Native names. Keraing (Kisabei) ; Hvunzi (Lugishu) ; evunge (Kitosh) ; 
unget (Kimasai); kikomi (Luragoli); shekome (Lutereki). 

Measurements, & . (Kaimosi) 400. 440. 70. 46 mm., 9. (Kirui) 
380. 405. 67. 45 mm. 

Breeding. At Kaimosi, on February 14, 1934, an embryo measur- 
ing 100. 35. 19. 12 mm., was preserved. The following day a native 
brought in two naked nestlings of which the male measured 115. 55. 
22. 10 mm. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 99 

Diet. The Bagishu aver that the giant rat only eats grass. I checked 
and rechecked this statement as it is in direct contradiction to what 
the Wakami told me respecting the race (C. g. osgoodi) inhabiting 
the Uluguru Mountains (cf. Allen & Loveridge, 1927, p. 436). 

Parasites. Two species of orthopteran parasites (Hemimerus han- 
seni & //. talpoides) were collected in their fur at Sipi, the former 
only at Kirui and Kaimosi. The rats were dirty at Sipi but in fine 
clean condition at Kaimosi despite the numerous Hemimerus. 

Enemies. The bodies of these giant rats were esteemed a delicacy 
and greatly in demand by the Wasabei, Bagishu, Elgon Masai, and 
Watereki all of whom trap them with deadfalls as a regular thing. 
One frequently met with the traps in the forest. 

Folklore. The Maragoli say that when a male is killed all the females 
in the vicinity will die. 

Lophuromys aquilus aquilus (True) 

Mus aquilus True, 1892, Proc. U. S. Nat. Mus., 15, p. 460: Mount Kiliman- 
jaro, Tanganyika Territory. 

Lophuromys rubecula Dollman, 1909, Ann. Mag. Nat. Hist. (8), 4, p. 551: 
Elgonyi, Mount Elgon, Kenya Colony. 

16 (M. C. Z. 31447-62) Sipi, U. 12-13. xii. 33. 

c? 4 9 (M. C. Z. 31441-5) Butandiga, U. 8-13. i. 34. 

9 (M. C. Z. 31473) Goletatomi, K. C. 31. i. 34. 

3 d" 6 9 (M. C. Z. 31464-72) Kaimosi, K. C. 10-19. ii. 34. 

Distribution. This harsh-furred mouse also occurs in the alpine 
zone at Kaburomi, 10,500 feet, where I rejected a damaged specimen. 
The species did not occur at Elgonyi where I was camped so that it 
appears probable that Kemp obtained the type of rubecula rather 
higher than my camp. I sent natives up a day's march above Elgonyi 
and they secured one at a place they called Goletatomi, which is 
topotypic of rubecula. 

Native names. Jamasiku (Kisabei); siku (Lugishu); chemasoget 
(Kimasai); lidulu (Luragoli); liguve (Lutereki). 

Coloration. The wide individual differences in intensity of coloring 
on the underside, as well as those due to season, nullify any attempt 
to break up this species into races on such grounds since intergrada- 
tion occurs between extreme types. The extremes in the case of the 
ventral surface are a pinkish buff on the one hand and a bright vina- 
ceous on the other. 

We therefore concur with Hollister (1919, p. 110) that there are 



100 bulletin: museum of comparative zoology 

insufficient grounds for considering rubecula of Mount Elgon as racially 
distinct from aquilus of Mount Kilimanjaro. 

Measurements, cf. (Kaimosi) 125. 75. 20. 16 mm., 9. (Kaimosi) 
143. 75. 21. 18 mm. 

Enemies. A mouse, apparently referable to this species, was found 
in the stomach of a nose-horned viper (Bitis nasicomis). 

Lophuromys sikapusi ansorgei de Winton 

Lophuromys ansorgei de Winton, 1896, Proc. Zool. Soc. London, p. 607: Mum- 
ias, Kenya Colony. 

<? cT (M. C. Z. 31446, 31463) Sipi, U. 12-13. xii. 33. 

Discussion. Judging from the fact that only two of these mice were 
obtained at Sipi, it would appear that it is much less common there, 
or has a more restricted habitat, than L. a. aquilus. There can be 
little doubt that this is an eastern representative of L. sikapusi of 
the clearings and forest edges of West Africa. We are therefore regard- 
ing its status as that of a subspecies. 

Coloration. As compared with L. a. aquilus, the pelage of ansorgei 
is of more even coloring, a paler olive brown and lacking the minute 
ticking while the belly is of a clearer chestnut tint. 

Measurement, cf. 101. 54. 20. 16 mm. 

Saucostomus cricetulus sp. nov. 

Type. Museum of Comparative Zoology, No. 31, 475. A subadult 
male, skin and skull, from the south bank of Greeki River, Sabei 
district, due north of Mount Elgon, Uganda, collected by Arthur 
Loveridge, December 5, 1933. 

Description. In color this is a much darker gray than neighboring 
members of the genus, with almost none of the buffy tints of back 
and sides. Dorsal surface of the body from the muzzle to the root of 
the tail, a uniform " deep mouse gray" to " dusky drab" of Ridgway, 
becoming faintly tinged along the sides and cheeks with " pale ochra- 
ceous buff." The individual hairs of the back are of a "deep neutral 
gray" basally, this color gradually passing into a narrow subter- 
minal band of very pale buffy (under a lens appearing soiled whitish), 
succeeded by a black tip. The ears on both inner and outer surfaces 
are a uniform dark blackish brown, slightly contrasting with the 
surrounding dark gray of the head, and conspicuously edged with 
clear white. Dorsal surface of the tail like the back, becoming slightly 
paler below with the admixture of short whitish hairs among the 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 101 

black. Backs of the hands and feet white as far as the wrists and ankles. 
Below, the hair of the chin is whitish to the roots; elsewhere that of 
the entire under surface of the body, and of the legs to the wrists and 
ankles is "deep neutral gray" at base, tipped with whitish, the gray 
bases everywhere showing through conspicuously, giving an effect 
of dark grayish tinted with bluish. 

The skull does not differ noticeably from that of Saccostomus isiolae 
Heller, its eastern neighbor of the dry Guaso Nyiro country. The 
nasals equal or minutely exceed the posterior extension of the pre- 
maxillae as in the latter, and in the coastal mearnsi Heller (type from 
Changamwe, Kenya Colony), instead of conspicuously exceeding 
them as in umbriventcr Miller of the Sotik region to the southward. 
Of these closely related forms, the molar teeth are smallest and the 
molar rows nearly parallel in the last-named but in the others and 
in cricetulus the teeth are slightly heavier, while the divergence of the 
rows is most noticeable in the new animal. The supraorbital ridges 
are most prominent in umbriventcr, less so in the others. The posterior 
palatal pits are of about the same size in all, except that in the type 
specimen of mearnsi they are unusually large. 

Measurements. In external measurements this pouched mouse 
does not apparently differ from the neighboring forms. The collec- 
tor's measurements are as follows, those of the male type (No. 31475) 
preceding those of the female paratype (No. 31474): — head and 
body, 133, 133 mm.; tail, 50, 55 mm.; hind foot, 20, 22 mm.; ear 20, 
21 mm. 

The skull of the type and that of the paratype are slightly damaged 
at the posterior end, but that of the former shows the following: — 
greatest length, 30 mm.; palatal length, 18.7; zygomatic width, 16.6; 
width of brain case above squamosal roots, 13.7; interorbital width, 
4.5; length of nasals, 13.0; upper molar row, 6.7; width across molar 
rows anteriorly, 7.7; same posteriorly, 6.5; lower molar row, 6.6. 

Remarks. The trapping of these two specimens extends the known 
range of the genus slightly to the northwestward. In their uniformly 
dark gray coloring and conspicuous white edges of the ears they 
differ strikingly from the neighboring forms of the genus, while the 
shortened tail and white feet further combine to give them a close 
external likeness to the larger Asiatic species of Cricetulus, which 
has suggested the specific name. They were captured in the usual 
type of habitat for the genus, namely open grass- and bush-covered 
country, and in this case near the banks of a river. 

Through the kindness of Mr. Gerrit S. Miller, Jr., and Dr. Reming- 



102 bulletin: museum of comparative zoology 

ton Kellogg, of the U. S. National Museum, we have had for com- 
parison the type and a topotype of the form mearnsi Heller and most 
of the original series of umbriventcr Miller, which, with a series of 
topotypical isiolae Heller from Guaso Nyiro country, have served as a 
basis for comparison. These three were described as separate species, 
but externally are all practically identical and are without doubt 
very closely related. In cranial characters, however, umbriventcr 
(which seems to be hardly, if any, darker underneath than the two 
others) is perhaps distinguishable on the basis of its slightly weaker 
molar teeth and the nearly parallel alignment of the molar rows, as 
well as by the relatively longer nasals, exceeding the premaxillaries 
posteriorly, and the narrower interorbital space with somewhat more 
prominent supraorbital ridges. These characters are at best of no 
more than subspecific value, and all three, if recognizable at all, are 
subspecifically related. Hollister (1919, p. 114) in reviewing the East 
African forms has already suggested this, but did not attempt to say 
of what species they should be considered races. 

The South African S. mashonae is said to be at once distinguishable 
by the well-developed antero-external cusp of the second upper molar, 
which in these more northern forms is extremely small; moreover, 
S. campestris, of which these might be thought races, is at once dis- 
tinguished by the pure white belly, the hairs without dark bases. 
In the absence of specimens from intermediate localities, it seems 
best at present to regard the Kenya Colony Saccostomus as a separate 
species {mearnsi) of three described races, while the form here named is 
so different in its coloration, that its subspecific relation to them is 
doubtful, and we have provisionally given it specific rank. 



Acomys ignitus ignitus Dollman 

Acomys ignitus Dollman, 1910, Ann. Mag. Nat. Hist. (8), 6, p. 229; Voi, 
Kenya Colony. 

a* & (M. C. Z. 32107-8) Tsavo, K. C. 31. iii. & 4. iv. 34. 
d* 9 (M. C. Z. 32106, 32110) Voi, K. C. 7. iv. 34. 

cT (M. C. Z. 32109) Kitau, Manda Id., K. C. 15. v. 34. 

Coloration. The three spiny mice from Tsavo and Voi differ slightly 
amongst themselves, the topotype from Voi being an intense rusty on 
the sides while the two from Tsavo are a clear, rich ochraceous. The 
Manda Island animal is slightly paler with its dull dorsal area more 
extensive, a difference which is probably due in part to immaturity. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 103 

Measurements, <? . (Tsavo) 120. 88. 15. 15 mm., 9. 100. 72. 16. 
16 mm. 

Breeding. At Voi, on April 7, 1034, a very young male (68. 45. 11. 12 
mm.) was taken. 

Enemies. At Voi one was recovered from the stomach of a Lizard- 
Buzzard (Kaupifalco m. monogrammicus) . 

Habits. One of the Tsavo specimens was taken in a rat trap between 
6 and 9 a.m., if the trapper can be believed, for he averred that he 
examined the trap at 6 a.m. daybreak. The abdominal fur had slipped 
already when found at 9 a.m., the weather was extremely hot and the 
trap exposed to the sun's rays. 

Acomys wilsoni wilsoni Thomas 

Acomys wilsoni Thomas, 1892, Ann. Mag. Nat. Hist. (6), 10, p. 22: Mombasa, 
Kenya Colony. 

cf 9 (M. C. Z. 32111-2) Wema, Ngatana, K. C. 12 & 19. vi. 34. 

Native name. Mgonachekede (Kipokomo). 

Coloration. These two skins represent the brighter-colored coastal 
race of this small short-tailed spiny mouse. 

Measurements, <? . 87. 45. 12. 11 mm., 9 . 75. 50. 12. 12 mm. 

Dasymys helukus helukus Heller 

Dasymys helukus Heller, 1910, Smithsonian Misc. Coll., 54, No. 1924, p. 2: 
Sirgoit, Uasin Gishu Plateau, Kenya Colony. 

9 (M. C. Z. 31357) Butandiga, U. 8. i. 34. 
<? 9 (M. C. Z. 31354-5) Kirui, K. C. 1. ii. 34. 
4 cf 5 9 (M. C. Z. 31356, 31358-65) Kaimosi, K. C. 8-22. ii. 34. 

Native names. Bunwc (Lugishu); inya (Luragoli and Lutereki). 

Coloration. These swamp rats form a uniformly dark, shaggy-haired 
series with the exception of one of the Kirui specimens which is decid- 
edly more tawny or olive brown. 

Measurements, cf. (Kirui) 180. 142. 29. 20 mm., 9 . (Kaimosi) 165. 
137. 30. 24 mm. 

Parasites. Numerous fleas and mites preserved from the fur of a 
Kaimosi rat. 

Enemies. One was recovered from the stomach of a wild cat {Felts 
o. nandae) at Kaimosi. 



104 bulletin: museum of comparative zoology 

Pelomys fallax iridescens Heller 

Pelomys fallax iridescens Heller, 1912, Smithsonian Misc. Coll., 59, No. 16, 
p. 12: Mount Mbololo, Taita, Kenya Colony. 

9 (M. C. Z. 32123) Mt. Mbololo, K. C. 26. iv. 34. 

Distribution. This is one of the mammals that Loveridge hoped to 
obtain in his brief visit to the type locality, but only trapped a single 
specimen. According to Hollister (1919, p. 125) it was abundant at 
the time of Heller's visit for the latter secured no less than thirty- 
seven. 

Measurements. 9 . 130. 115. 29. ? ear. mm. 

Arvicanthis abyssinicus nubilans Wroughton 

Arvicanthis abyssinicus nubilans 1909, Ann. Mag. Nat. Hist. (8), 4, p. 539: 
Kisumu, 3,600 feet, Kenya Colony. 

5^49 (M. C. Z. 31342-50) Sipi, U. 12-19. xii. 33. 

6 c? 4 9 (M. C. Z. 31296-9, 31310-5) Butandiga, U. 8. i. 34. 

5 <? 5 9 (M. C. Z. 31291-2, 31316-23) Kirui, K. C. 22-23. i. 34. 

- 5^69 (M. C. Z. 31293-5, 31324-31) Kaimosi, K. C. 10-14. ii. 34. 

Native names. Myera (Kisabei, Lugishu and Kitosh); manyaret 
(Kimasai); engeki (Luragoli); injhi (Lutereki). 

Measurements, cf . (Kaimosi) 165. 130. 29. 18 mm., 9 . (Kirui) 161. 
122. 27. 18 mm. 

Enemies. Recovered from the stomachs of a Crested Eagle (Lophae- 
tus occipitalis), mongoose (Herpestes i. funestus), Hissing Sand Snake 
(Psammophis sibilans) and Black-lipped Cobra (Naja melanoleuca) , 
at Butandiga and Kaimosi. 

Arvicanthis abyssinicus virescens Heller 

Arvicanthis abyssinicus virescens Heller, Smithsonian Misc. Coll., 63, No. 7, 
p. 11: Voi, Kenya Colony. 

<? (M. C. Z. 32207) Golbanti, K. C. 22. vi. 34. 

Discussion. This specimen is referred to the coastal race which, 
however, is not very different from its neighbors farther inland. 

Measurements, cf . 150. 136. 30. 18 mm. 

Diet. Trapped with bread as a bait. 

Note. Very small ants cut the hairs from the rump of this rat as it 
lay dead in the trap. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 105 

Lemniscomys griselda maculosus (Osgood) 

Arvicanthis dorsalis maculosus Osgood, 1910, Publ. Field Mus. Nat. Hist., Zool. 
Series, 10, p. 17: Voi, Kenya Colony. 

<? (M. C. Z. 32122) Wema, Ngatana, K. C. 15. vi. 34. 

Distribution. This seems to be an uncommon species on the coast of 
Kenya Colony though abundant enough in the Voi-Taita region. 

Native name. Kadzora (Kipokomo). 

Measurements. d\ 130. 132. 26. 18 mm. 

Breeding. Striped Grass Rats were breeding at Wema in June for 
about a score of nestlings, averaging 57. 32. 14. 7 mm., were brought 
in from the village of Wema, Tana River. 

Lemniscomys striatus massaicus (Pagenstecher) 

Mus (Lemniscomys) barbarus L. var. massaicus Pagenstecher, 1885, Jahrb. 
Wiss. Anstalt, Hamburg, 2, p. 45: Lake Naivasha, Kenya Colony. 

1+3^29 (M. C. Z. 31687-92) Mt. Debasien, U. 21-29. xi. 33. 

c? (M. C. Z. 31664) Sabei, U. 9. xii. 33. 

3 cf 5 9 (M. C. Z. 31652-9) Sipi, U. 12-20. xii. 33. 
2 cf 2 9 (M. C. Z. 31660-3) Butandiga, U. 8. i. 34. 
7^4? (M. C. Z. 31676-86) Kirui, K. C. 22. i-1. ii. 34. 

2 9 (M. C. Z. 31665-6) Elgonyi, K. C. 4. ii. 34. 

4 <? 5 9 (M. C. Z. 31667-75) Kaimosi, K. C. 8-19. ii. 34. 

Distribution. Another was trapped on the south bank of the Greeki 
River but the sun was so hot that the fur had slipped when it was 
brought in at 4 p.m. This species, so common in the Mount Elgon 
region, is an upland form apparently absent from the coast. 

Native names. Kimwarces (Kisabei) ; oluvende (Lugishu) ; aluvende 
(Kitosh); chemotet (Kimasai) ; livende (Luragoli and Lutereki). 

Measurements, o 71 . (Sabei and Sipi) 136. 141. 25. 16 mm., 9 • (Sipi) 
131. 133.25. 16 mm. 

Parasites. Nymphal ticks (Ixodes sp.) were common on these rats 
at Mount Debasien. 

Enemies. One was recovered from the stomach of a White Nile 
Chanting Hawk (Melierax m. metabatcs) on Debasien. 

Rhabdomys pumilio diminutus (Thomas) 

Isomys -pumilio diminutus Thomas, 1892, Proc. Zool. Soc. London, p. 551: 
Mianzini, east of Lake Naivasha, Kenya Colony. 

9 (M. C. Z. 31693) Madangi, U. 3. i. 34. 



106 bulletin: museum of comparative zoology 

Distribution. This striped grass rat was only encountered in the 
alpine zone of Mount Elgon where it was common enough at 12,000 
feet. 

Native name. Oluvende (Lugishu). 

Measurements. 9 . 125. 88. 21. 14 mm. 

Otomys tropicalis elgonis Wroughton 

Otomys irroratus elgonis Wroughton, 1910, Ann. Mag. Nat. Hist. (8), 5, p. 207: 
Elgonyi, Mount Elgon, Kenya Colony. 

7o , 89 (M. C. Z. 31269-72, 31377-81, 31426-31) Sipi, U. 13-16. xii. 33. 
1 (M. C. Z. 31371) Kaburomi, U. 28. xii. 33. 
9 (M. C. Z. 31376) Madangi, U. 3. i. 34. 
<? 9 (M. C. Z. 31373-4) Butandiga, U. 8. i. 34. 
6 d 1 3 9 (M. C. Z. 31375, 31416, 31418, 31420, 31432-4, 31436, 31439) 
Kaimosi, K. C. 8-14. ii. 34. 

Native names. Urusti (Kisabei); mbole (Lugishu); ivole (Luragoli); 
livole (Lutereki). 

Measurements, cf. (Sipi) 180. 85. 28. 23 mm., 9 . (Kaimosi) 182. 
86. 26. 22 mm. 

Breeding. At Sipi, on December 20, 1933, a native brought in a 
female suckling her two young. When I lifted her out of the gourd 
she made no attempt to escape and the young, though quite large, 
remained attached to the mother's teats as she was transferred to 
the ground to be photographed. Afterwards I removed the family to 
the 'bush' and let them go. At Butandiga, on January 8, 1934, a 
suckling male, measuring 83. 40. 21. 13 mm., was brought in. At 
Kaimosi, on February 15, 1934, a female was suckling two young. 

Enemies. Elgon swamp rats were recovered from the stomachs of a 
variety of creatures at Kaimosi, namely a genet (Genetta s. stuhl- 
vianni), tree civet (Nandinia b. arborea), mongooses (Herpestes i. 
funestus and Ichneumia a. ibeana), twice from wild cats (Felis o. 
nandae), and a raamba (Dendraspis jamesoni). 

Otomys angoniensis elassodon Osgood 

Otomys angoniensis elassodon Osgood, 1910, Publ. Field Mus. Nat. Hist., Zool. 
Series, 10, p. 10: Naivasha, Kenya Colony. 

5 cf 5 9 (M. C. Z. 31368-70, 31372, 31421-5, 31438) Kaburomi, U. 

28. xii. 33. 
2 9 (M. C. Z. 31366-7) Goletatomi, K. C. 31. L 34. 
3^29 (M. C. Z. 31417, 31419, 31435, 31437. 31440) Kaimosi, 

K. C. 8-14. ii. 34. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 107 

Native names. Murusi (Kisabei, perhaps not different from that 
recorded above for 0. t. elgonis); mburustit (Kimasai). 

Coloration. This is a somewhat grayer looking animal than 0. t. 
elgonis and possessing a pale grayish throat and chest. It seems to 
largely replace elgonis in the alpine zone of Mount Elgon but both 
occur together at Kaimosi where they are perhaps equally common 
though, strangely enough, Heller obtained seventeen elgonis but no 
elassodon in this locality. 

Measurements, cf. (Kaimosi) 167. 182. 26. 22 mm., 9 ■ (Kaimosi) 
170. 80. 23. 21 mm. 

Enemies. This swamp rat is eaten by the Wanderobo at Kaburomi. 
Its fur figured largely in the droppings of servals (Felis c. hindei) 
which were very abundant in the alpine-meadow zone of Mount 
Elgon. One was recovered from the stomach of a buzzard (Buteo 
r. augur) at Kaburomi, another from that of a harrier (Circus macrou- 
rus) at Kaimosi. 

HYSTRICIDAE 

Hystrix galeata Thomas 

Hystrix galeata Thomas, 1893, Ann. Mag. Nat. Hist. (6), 11, p. 220: Lamu, 
Kenya Colony. 

d* (M. C. Z. 32284) Sipi, U. 22. xii. 33. 

Native names. Sabidet (Kimasai); esegesi (Kitosh); isegcsi (Lugishu 
and Lutereki); rungu (Luragoli); sasa (Kitaita). 

Discussion. This specimen is immature with only two cheek teeth 
erupted on each side. It can only tentatively be identified with typi- 
cal galeata rather than with one of the more or less nominal races that 
have been described from Kenya Colony and Tanganyika Territory 
in recent years. 

Measurements, cf . 570. 80. 85. 36 mm. 

Diet. Its stomach was distended with maize, so finely masticated 
that I had to rely for its identification on the Bagishu who speared 
this porcupine. 

Enemies. Eaten by the Bagishu and Wasabei. 

Atherurus turneri St. Leger 

Atherura turneri St. Leger, 1932, Ann. Mag. Nat. Hist. (10), 10, p. 231: Kai- 
mosi, Kakamega, Kenya Colony. 

1 + 5 cf 4 9 (M. C. Z. 32274-83) Kaimosi, K. C. 8-11. ii. 34. 



108 bulletin: museum of comparative zoology 

Native names. Kahegenya (Luragoli) ; shihekenye (Lutereki). 

Discussion. This fine series of topotypes presents a fairly uniform 
appearance, with dark blackish-brown spines, pale whitish at the 
base, and with the tassel of the tail white. Externally they are similar 
to A. africana of western Africa, but the skulls differ conspicuously 
in almost lacking any sign of inflation of the premaxillary and anterior 
frontal region. Instead, the rostrum is narrow with a narrow and 
inconspicuous ascending arm of the premaxillary which hardly appears 
in dorsal view, whereas in the other species this part of the base of the 
rostrum is broad with a wide arm of the premaxillary broadly visible 
in the dorsal aspect of the skull. In addition the frontal region of the 
Kaimosi animal is much flattened, but swollen and inflated in africana. 

Evidently turneri represents a much more primitive state, and 
instead of being a race of africana, as might have been supposed, may 
be retained as a distinct species until intermediate forms are found. 

Measurements, tf. 470. 200. 72. 39 mm., 9 . 460. 205. 62. 31 mm. 

Diet. The stomachs contained finely gnawed vegetable matter. 

Habitat. I often came across holes in the forest near the river. 
The natives said that these were made by aquatic porcupines, that 
they lived in them and that they did not climb trees, the latter might 
be inferred from their build and feet which appear unsuitable for 
climbing. The Watereki hunt them with dogs and spears. 



THRYOXOMYIDAE 

Choeromys gregorianus (Thomas) 

Aulacodus gregorianus Thomas, 1894, Ann. Mag. Nat. Hist. (6), 13, p. 202: 
Luiji Reru River, Konu, Kikuyu, Kenya Colony. 

c? juv. (M. C. Z. 31276) Kirui, K. C. 25. i. 34. 
2 <? 9 juv. (M. C. Z. 31273-5) Kaimosi, K. C. 9 & 20. ii. 34. 

Native names. Megore (Kimasai) ; csabolet (Kitosh) ; isiavale (Lura- 
goli and Lutereki). 

Discussion. These specimens are very uniform in coloring, but not 
fully adult, as is shown by the fact that the last molars have not 
erupted. 

Measurements. <? . 340. 73. 62. 29 mm., 9 . 290. 85. 59. 27 mm. 

Remarks. The large cane rat (T. sivinderianus) also occurs at 
Kaimosi, several huge ones were offered for sale but were so highly 
esteemed as food that the hunters would not sell them under five 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 109 

shillings ($1.25) each. Nor would they consent to have the skin and 
skull removed for a shilling and the meat returned to them. 

LEPORIUAE 

Lepus victoriae kakumegae Heller 

Lepus kakumegae Heller, 1912, Smithsonian Misc. Coll., 59, No. 16, p. 19: 
Lukosa River, Kakamega Forest, Kenya Colony. 

tf (M. C. Z. 32273) Kirui, K. C. 28. i. 34. 
cf (M. C. Z. 32272) Elgonyi, K. C. 5. ii. 34. 

Native names. Nduyu (Kitosh); irangut (Kimasai); kifuyo (Lura- 
goli); shikalla (Lutereki). 

Coloration. This is a richly colored hare, the extensively black 
upper side of whose tail was pointed out by Heller as being one of 
the characters which distinguish it from typical victoriae. 

Measurements, cf. (Elgonyi) 480. 75. 110. 90 mm. 

Parasites. Fleas and ticks were swarming on the Elgonyi animal. 

Habitat. These hares were common enough on the grass-grown hill- 
sides of southern Elgon as well as in the alpine zone, but so active that 
I never got a shot at one. Both the animals listed above were snared 
by natives, and brought in alive, then chloroformed. 

SUIDAE 
Hylochoerus meinertzhageni meinertzhageni Thomas 

Hylochoerus meinertzhageni Thomas, 1904, Nature, 70, p. 577: Nandi Forest, 
Kenya Colony at 7,000 feet. 

Type locality. Mr. F. N. Hoyt of the Friends' Africa Mission at 
Kaimosi, informs me that the type specimen of this giant forest hog 
was killed by natives on the mission property just below my camp 
site, i.e. at Kaimosi. It was given, or sold, by the natives to Mr. A. 
B. Chilsom, Mr. Hoyt's predecessor, who in turn parted with it to 
Major Meinertzhagen when the latter was passing through Kaimosi. 

The type locality is, therefore, in one sense erroneous. On the other 
hand the hog undoubtedly came from either the Kakamega or Nandi 
Forests in the vicinity of Kaimosi. 

During my stay at Kaimosi a native offered me a skin for sale. He 
stated that it had been killed in the Nandi Forest but as it lacked 
head, feet and skull I did not accept it. I saw several Tereki war 
shields which appeared to be made from the tough hides of these giant 
forest hogs. 



110 bulletin: museum of comparative zoology 

BOVIDAE 

Damaliscus korrigum topi Blaine 

Damaliscus korrigum topi Blaine, 1914, Ann. Mag. Nat. Hist. (8), 13, p. 333: 
Near Malindi, Kenya Colony. 

c? skull (M. C. Z. 32068) Between Malindi & Golbanti, K. C. 2. v.34. 

Distribution. This topotype was shot on the way to Lamu Island 
south of the Tana. Unfortunately the necessities of three days' inces- 
sant travelling through torrential downpours caused the hair to slip 
despite all precautions. I was particularly sorry to lose it as it was 
the first skin of any game animal that I have ever shot which I failed 
to preserve. 

Measurements. <?. 2,000. 430. 510. 195 mm. 

Parasites. Ticks. 

Cephalophus monticola musculoides Heller 

Cephalophus monticola musculoides Heller, 1913, Smithsonian Misc. Coll., 61, 
No. 7, p. 9: Kakamega Forest, Kenya Colony. 

skull & 5 tf 1 9 (M. C. Z. 31102-3, 31609-10, 31967, 32196, 32286) 

Elgonyi, K. C. 31. i-6. ii. 34. 
d" juv. (M. C. Z. 32285) Kaimosi, Kakamega, K. C. 23. ii. 34. 

Native names. Ikirungu (Lugishu); kabenyet (Kimasai); kasendi 
(Luragoli); shiseri (Lutereki). 

Discussion. Lydekker and Blaine (1914, p. 96) regard musculoides 
and other East African forms as races of melanorheus (1846) from 
Fernando Po, treating monticola (1789) of Cape Colony as a distinct 
full species. Hollister (1924, p. 80) follows Heller in regarding all as 
races of the older monticola. 

Granvik (1924, p. 31) records a male from Mount Elgon under the 
name of aequatorialis Matschie, described from Chagwe, Uganda. 
There is no difference to be discerned between our Elgon series and 
the topotype of Heller's musculoides, distinguished by larger size and 
lighter underparts which contrast with the flanks. It would appear 
that Granvik's identification was incorrect or, alternatively, that 
musculoides is a synonym of aequatorialis. 

The horns of one adult male from Elgonyi are peculiar in that they 
have not developed sufficiently to be visible externally. Their bony 
cores are low knobs with flattened tops which fit into small pockets 
in the skin instead of penetrating it. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 111 

Measurements. Heller's measurements were 525. 85. 170. 55 mm., 
so that our Elgonyi animals are all larger still, viz. cf . 643. 104. 165. 
54 mm., 9 . 550. 90. 165. 56 mm. 

Breeding. At Elgonyi, on February 1, 1934, a native brought in a 
new-born duiker (cf . 310. 55. 125. 43 mm.), and at Kaimosi, Feb- 
ruary 23, 1934, another slightly older (d\ 385. 70. 130. 47. mm.). 
Both said that their dogs had brought them the little creatures. 

Parasites. Ticks {Amblyomma varicgata and nymphal Rhipi- 
cephalus sp.) were recovered from several duikers of the Elgonyi 
series, while Haemaphy salts parmata were present on the Kaimosi 
buck. 

Enemies. At Madangi, 11,500 feet, the skull was removed from the 
dried body of a pygmy duiker found at the very edge of a cliff. (Un- 
fortunately it was left behind by the skinner, so lost). The tiny ante- 
lope had evidently been cornered by a dog and disembowelled. Though 
these animals were very common in the alpine meadows they were 
so wild that it was difficult to get a shot at one. The reason for their 
wariness was obvious for all day long and almost every day, bands 
of natives accompanied by large dogs, some of which were belled, 
made the hills and valleys ring with their shouts and cries as they 
harried the game. At Kaimosi these duiker have apparently become 
exceedingly rare. At Elgonyi the skull cap of a duiker which had 
been killed by a leopard, was collected. 



Sylvicapra grimmia deserti Heller 

Sylvicapra grimmia deserti Heller, 1913, Smithsonian Misc. Coll., 61, No. 17, 
p. 4: Voi, Kenya Colony. 

9 & fetus (M. C. Z. 31954, 32327) Lamu Id., K. C. 9. v. 34. 

Native name. Nguruvu (Kiamu) . 

Discussion. These skins are referred to the coastal race though 
they differ little, if at all, in general coloration from inland specimens. 

Measurements. 9 . 840. 110. 260. 107 mm. 

Breeding. On May 9, 1934, this animal held a fetal 9 • 345. 45. 
146. 57 mm. which was practically ready for birth. 

Habitat. These duiker, which doubtless owe their introduction on 
the island to man, are said to be exceedingly common on the sand- 
hills east of Sheila where I shot this animal as it unexpectedly dashed 
out of a bush. 



112 bulletin: museum of comparative zoology 

Sylvicapra grimmia lobeliarum Lonnberg 

Sylvicapra grimmia lobeliarum Lonnberg, 1919, Rev. Zool. Africaine, 7, p. 181: 
Mount Elgon (at high altitude among lobelias), Kenya Colony. 

c? (M. C. Z. 31608) Kaburomi, K. C. 30. xii. 33. 

Distribution. Our specimen is topotypic having been shot above 
the Kaburomi eamp in the alpine zone of Elgon at 11,000 feet. 

Native name. Ekisi (Lugishu). 

Discussion. This adult male has the dark forehead which is men- 
tioned by Lonnberg as a principal character of this montane race. 

Measurements, a\ 870. 100. 272. 107 mm. 

Diet. Shot while grazing at 8 a.m. Its stomach was full of grass; 
no internal nor external parasites observed. 

Sylvicapra grimmia nyanzae Neumann 

Sylvicapra abyssinica nyanzae Neumann, 1905, Sitz. Ges. naturf. Freunde 
Berlin, p. 89: Kwa Kitoto, Kavirondo, Kenya Colony. 

9 juv. (M. C. Z. 31618) Kirui, K. C. 28. i. 34. 

Native name. Ekisi (Kitosh). 

Discussion. This young animal, coming from below 7,000 feet on the 
southern face of Mount Elgon, is referred to nyanzae tentatively on 
geographic grounds. 

Measurements. 9 . juv. 500. 50. 186. 70 mm. 

Parasites. Ticks were preserved from its fur. 

Habits. Many years ago, the late Dr. S. L. Hinde drew attention 
to the custom of captive duikers killing fowls in the same enclosure. 
They pulled off the heads and lapped the blood. In 1914, Mr. L. S. B. 
Leakey told me of his captive duiker doing the same thing. The 
introduction of a salt lick into the enclosure put an end to the killing 
of the fowls. This result seemed to show that a desire for salt on the 
part of the antelope led to what was supposed to be a perversion of 
character due to captivity. 

While I was walking along the western foot of Mount Debasien, I 
paused on the main trail to shoot a bird in a nearby tree. An excla- 
mation from my gunbearer, however, caused me to turn in time to 
see a female duiker land with a bound in the road almost on top of a 
dove, which I had previously observed to be feeding there. They 
disappeared together, the dove over, the duiker into, the rank grass 
which flanked the track. The boys averred that the antelope had 
pounced upon the bird in an effort to seize it. (November 27, 1933). 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 113 

On the north bank of the Greeki River, therefore also in Karamojo, 
I had just shot and bagged a guineafowl when, with loud cries, a 
dozen of the birds rose from long grass a hundred yards away and 
settled in some small thorn trees. My gunbearer said that he had 
seen a duiker bound through the grass into the middle of the covey. 
Certainly the birds were so agitated by what had disturbed them 
that I was able to approach within range and shoot one of their 
number. (December 5, 1933). 

Ourebia Montana cottoni Thomas & Wroughton 

Ourebia cottoni Thomas & Wroughton, 1908, Ann. Mag. Nat. Hist. (8), 1, 
p. 178: Sirgoit Rock, Uasin Gishu Plateau, Kenya Colony. 

Ourebia microdon Hollister, 1910, Smithsonian Misc. Coll., 56, No. 2, p. 4: 
South of Nzoia River, Uasin Gishu Plateau, Kenya Colony. 

cf juv. d" 9 (M. C. Z. 31100, 31607, 31971) w. foot Mt. Debasien, U.29. xi. 33. 

Measurements. d\ 1,000. 80. 270 (with hoof 315). 112 mm., 9 . 1,400. 
100. 280 (with hoof 325). 100 mm. 

Breeding. The juvenile (780. 70. 225 (with hoof 260) 90 mm.) was 
being weaned. It sprang up from its lair in long grass and was shot 
with No. 8 from a twelve bore at a distance of thirty feet. 

Parasites. Nymphal ticks Ambly omnia sp. were numerous about the 
genitalia. 

Raphiceros campestris neumanni (Matschie) 

Pediotragus neumanni Matschie, 1894, Sitz. Ges. naturf. Freunde Berlin, p. 
122: Northern Ugogo, Tanganyika Territory. 

9 juv. (M. C. Z. 32329) Peccatoni, K. C. 25. v. 34. 

Distribution. This record must be near the northern bounds of the 
species on the coast. Most of the thirty examples listed by Hollister 
(1924, p. 94) are from inland localities on the Kenya Plateau. 

Measurements. Very young 9 . 315. 65. 125. 52 mm. 

Rhynchotragus kirkii kirkii (Giinther) 

Neotragus kirkii Giinther, 1880, Proc. Zool. Soc. London, p. 222: Brava, 
Italian Somaliland. 

5 cf 3 9 & foetus (M. C. Z. 31946, 31968-70, 32195, 32259-60, 32328) 
Kitau, Manda Id., K. C. 16-19. v. 34. 

Distribution. According to Heller (1913, Smithsonian Misc. Coll., 
61, No. 7, p. 4) typical kirkii is found along the coast south to the 



114 bulletin: museum of comparative zoology 

Tana River; the series shot by Loveridge on Manda near Larau Island 
are thus approaching the southern limits of the range of the typical 
form. 

Native name. Tibi (Kipokomo). 

Discussion. Their cranial dimensions average noticeably smaller 
than those of the race nyikae as described by Heller. The adult males 
of the former have an upper tooth row of about 32-32.5 mm. as 
against 37.5-39 mm. in nyikae. 

Measurements, cf . 620. 45. 188. 68 mm., 9 . 600. 40. 191. 63 mm. 

Breeding. On May 17, 1934, a fetal 9 . 264. 20. 110. 45 mm. was 
removed. The same day two extremely young dikdik were seen but 
they were both able to run well. 

Enemies. According to the natives, Hunting Dogs from the adja- 
cent mainland invade the island from time to time and harass the 
dikdik for days before taking their departure. 

Habitat. Dikdik are said not to occur on Lamu Island but on Manda 
they are the dominant mammal. One would encounter at least a 
dozen in the course of an hour's walk. At one spot nearly three miles 
north of Kitau they might truly be said to be as common as rabbits 
in an English pasture; instead of being in pairs, small parties of them 
would start up on every side and go bounding away. 

At eventide it was a pleasant sight to see these diminutive ante- 
lopes grazing. As the time of my visit coincided with the breaking 
of the rains, the acacia were in fresh verdure and presented many 
miniature park-like spots where the ground was clothed with blades 
of fresh green grass less than six inches high. At the Kitau end of 
the island a pair of dikdik would be found feeding in each glade at 
sunset. At night, as I slept across the entrance of the tent, I was 
awakened several times by the explosive snort of one of these animals 
which had wandered to within thirty feet of my bed. 

Rhynchotragus kirkii nyikae Heller 

Rhynchotragus kirki nyikae Heller, 1913, Smithsonian Misc. Coll., 61, No. 7, 
p. 3: Ndi, Taita, Kenya Colony. 

c? 9 (M. C. Z. 31945, 31956) Tsavo, K. C. 4. iv. 34. 

c? (M. C. Z. 31953) Karawa, near Malindi, K. C. 26. vi. 34. 

Distribution. The Tsavo specimens are almost topotypic for Tsavo 
is only nineteen miles west of Ndi, the latter being thirteen miles west 
of Voi. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 115 

Discussion. These three dikdik agree in having cranial dimensions 
slightly larger than those of typical kirkii, the tooth rows also being 
larger as mentioned above. 

Measurements, cf. (Karawa) 630. 40. 190. 64 mm., 9 . (Tsavo) 640. 
66. 192. 72 mm. 

Breeding. At Tsavo, on April 4, 1934, the female carried a fetal 
cf. 200. 13. 76. 29 mm., which was preserved in alcohol. 

Habitat. The Tsavo pair were shot in the dry scrub within a hun- 
dred yards of Tsavo station, with a right and left of No. 3 shot from 
the twelve bore. 

Kobus ellipsiprymnus kuru Heller 

Kobus ellipsiprymnus kuru Heller, 1913, Smithsonian Misc. Coll., 61, No. 13, 
p. 6: Taveta, Kenya Colony. 

c? (M. C. Z. 31974) Wema, Ngatana, K. C. 11. vi. 34. 

Native name. Kuyo (Kipokomo). 

Discussion. Neither Lydekker and Blaine (1914, p. 230) nor Hol- 
lister (1924, p. 104) venture an opinion as to the validity of this sup- 
posed race; the latter, however, suggests that it should be compared 
with the race kulu described by Matschie (1911) from Maliwe, west 
of Kilwa in southern Tanganyika Territory. 

Measurements, cf. 1,500. 396. 475. 195 mm. 

Breeding. This fine animal was accompanied by about twenty 
females but no young. They were in a marsh just a mile east of Wema, 
and a quarter of a mile from the north bank of the Tana River. 

Kobus defassa ugandae Neumann 

Kobus unctuosus ugandae Neumann, 1905, Sitz. Ges. naturf. Freunde Berlin, 
p. 92: Maianda Valley, northern Uganda. 

c? (M. C. Z. 31600) Elgonyi, K. C. 28. i. 34. 

Native names. Amosemos (Karamojong) ; saramet (Kimasai) ; ekhuro 
(Kitosh). 

Discussion. This fine male, not wholly mature, conforms well to 
ugandae as redescribed by Lydekker. Of the various races briefly 
characterized by Matschie and Lonnberg, this may be closely similar 
to tjaderi Lonnberg (1907, northwestern Laikipia Plateau, west of 
the junction of the Guaso Nyiro and Guaso Hanek), nzoiae Matschie 
(1910, Uasin Gishu Plateau) and fulvifrons Matschie (1910, east of 
Kitosh, between the Guaso Masa and Nzoia Rivers) but until the 



116 bulletin: museum of comparative zoology 

validity of these supposed races can be determined, Lydekker's use 
of ugandae for all the defassa waterbuck of this area seems best. 

Measurements, cf . 1,770. 325. 520. 213 mm. 

Parasites. Through an oversight, the numerous cestodes which 
were removed from its stomach were not preserved. 

Habitat. Shot at 10 a.m. while grazing alone on the mountainside 
at 8,000 feet, far from any water; nevertheless waterbuck tracks were 
common in the dense thickets close by. 

Tragelaphus scriptus delamerei Pocock 

Tragelaphus delamerei Pocock, 1900, Ann. Mag. Nat. Hist. (7), 5, p. 95: Sayer, 
northeastern limits of Laikipia Plateau, Kenya Colony. 

9 (M. C. Z. 31613) Elgonyi, K. C. 1. ii. 34. 

Distribution. Bushbuck were frequently heard barking in the forests 
on the western slopes of Mount Debasien, but none was seen. They 
would be referable to the race laticeps Matschie (1912) described 
from the northwestern base of Debasien, should that race be recog- 
nizable. Bushbuck were often heard at Sipi and Butandiga on the 
western slopes of Mount Elgon, which is the type locality of heter- 
ochrous Cabrera (1918), but were so harried by the native hunters 
that none was encountered. 

Native name. Aderit (Karamojong). 

Coloration. This female from Elgonyi agrees perfectly with one 
from Mwanza on the south shore of Lake Victoria, in the rich rufous 
tone of the dorsal surfaces, narrow black spinal stripe with slight 
admixture of white hairs, almost complete suppression of the trans- 
verse body stripes, which, however, may be faintly traceable, three 
or four in number, while the lateral row of white spots is largely 
absent. Males are much darker, blackish brown above instead of 
red, with the white markings similarly suppressed or altogether 
absent. This more intensely colored race, therefore, inhabits the high- 
lands of Kenya Colony from the Elgon region and northern border 
of Tanganyika at least to the southern end of Lake Victoria. 

Measurements. 9 . 1,300. 150. 357. 142 mm. 

Breeding. She was gravid with a small hairless fetal cT 1 . 203. 30. 
59. 24 mm. 

Diet. Shot while grazing at 6 p.m., the stomach held grass only, no 
parasites being observed. 

Enemies. In certain parts of western Elgon almost every third man 
(Bagishu) one met was dressed in a bushbuck (T. s. heterochrous) skin. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 117 

Tragelaphus scriptus massaicus Neumann 

Tragelaphus massaicus Neumann, 1902, Sitz. Ges. naturf. Freunde Berlin, 
p. 96: Upper Bubu River, northwestern Irangi, Tanganyika Territory. 

rf" (M. C. Z. 31955) Kibwezi, K. C. 28. iii. 34. 
c? skull (M. C. Z. 31976) Voi, K. C. 10. iv. 34. 

Discussion. Mr. A. B. C. Smith, long resident at Kibwezi, told 
Loveridge that the horns of the local bushbuek were very short, 
not exceeding fifteen inches in length he should think. 

Measurements. d\ 1,290. 200. 350. 413 mm. 

Habitat. One of a pair that were lying up in long grass within a 
couple of hundred yards of the station. 

Parasites. Ticks (Rhipicephalus maculatvs) were found in the fur. 

Tragelaphus scriptus olivaceus Heller 

Tragelaphus scriptus olivaceus Heller, 1913, Smithsonian Misc. Coll., 61, No. 13, 
p. 1: Maji ya chumvi, Kenya Colony. 

9 (M. C. Z. 31972) Lamu Island, K. C. 9. v 34. 

Distribution. A female was also seen on Manda Island. Hollister 
(1924, p. 128) appears to regard this as a valid form, representing 
a pale coastal race. Whether or not it is really separable from some 
of the pale races described from Somaliland and Ethiopia remains 
to be seen, but it may stand for the present. 

Native name. Kungu (Kiamu) . 

Coloration. The female from Lamu is obviously paler than speci- 
mens of massaicus from Tanganyika Territory. It is more ochraceous 
buff on the sides, the central dorsal area less contrastingly brown; 
about four pairs of transverse body stripes are more clearly, though 
not sharply, marked; a lateral row of white spots becomes more nu- 
merous on the haunches; there is an elongate white mark in advance 
of the root of the tail on each side. 

Measurements. 9 . 1,123. 80. 335. 130 mm. 

Breeding. I regretted to find that this animal was in milk. 

Habitat. On several occasions, we came upon the tracks of bush- 
buck in the sandhills east of Sheila. I had followed up one of these 
for a mile when a doe sprang out on the further side of a bush which 
I was passing, I shot it, though not fatally, and had to follow it for 
a mile before getting it with the second shot. These animals have 
so far adapted themselves to local conditions that they pause and 
look back when topping each high dune. It is supposed that captive 
specimens escaped, or that the species was deliberately introduced 
on to the island in bygone days. 



118 bulletin: museum of comparative zoology 

ELEPHANTIDAE 

Loxodonta africana peeli (Lydekker) 

Elephas africanus peeli Lydekker, 1907, Proc. Zool. Soc. London, p. 393: 
Aberdare Mountains, Kenya Colony. 

Field Notes. Signs of elephant were numerous on Mount Debasien, 
a small skull was found in a ravine at about 7,000 feet and another 
in the patch of forest at the western foot of the mountain. At the 
time of my visit, however, (November, 1933) they had departed for 
the swamps in the vicinity of Lake Gedge, Karamojo. 

When the country flanking the Greeki River, due north of Mount 
Elgon, is flooded, elephant must be very plentiful if one may judge 
by the abundance of spoor. One of my skinners found a pair of tusks, 
which had been exposed by the recent burning of the grass in which 
they had been hidden. In accordance with regulations these were 
turned over to the District Officer and the reward for 'found ivory' 
paid to the discoverer. 

During our safari through the semiflooded coastal plain between 
the Tana River and Malindi, evidence of the presence of elephants 
was of hourly occurrence. Nevertheless only one was seen, and that 
by the porters whom I had preceded by an hour. They halted and so 
did the elephant which, after looking at them, made off. 

Folklore. I am indebted to Miss Parker of the mission at Kaimosi 
for the substance both of the following story and the Maragoli tales 
about baboons already recorded. 

Once upon a time a man set out to dig a pit in which to trap an 
elephant. As he was engaged in digging, two elephants accompanied 
by their young came along the path and hailed the man, whose name 
was Shikhuyakluya. "What are you digging?" they asked. "I am 
digging for rats," was the reply. The elephants knew better, however, 
and challenged him with the remark: "It would be as well for you to 
brush us." (This being in reference to a sacrificial ceremony). 

The man picked up a nearby reed as he enquired of an elephant: 
"Shall I brush you with this?" and received a reply in the affirmative. 
As he began to brush them with the reed, he chanted these words: 
"Be well, keep your eyes open and see that you do not fall into this 
hole." After brushing the parents, he turned to the children, but to 
them he muttered in a low voice so that their parents would not hear 
"May your eyes be closed so that you cannot see clearly. Fall into 
this pit." When the little elephants heard this they said to their 
father and mother: "Shikhuyakluya said for us to fall into the pit." 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 119 

Shikhuyakluya denied the accusation, saying: "Perhaps the children 
did not hear correctly." Then the old elephants agreed, saying: 
"Our children are very young and probably did not understand." 
The ceremony being finished the elephants departed on their way. 

The man recommenced work on the pit; on finishing it, he covered 
the hole with branches and leaves before taking his departure. Long 
afterwards, the elephants returned that way having forgotten all 
about the pit during the intervening time. They followed their usual 
path until one of the young animals fell into the hole. Only then did 
the old ones remember, and because they did not want Shikhuyakluya 
to get their little one they covered him up with earth from the heap 
still lying beside the path. Thus did they bury the little elephant alive 
and then continued on their way with their one remaining child. 



PROCAVIIDAE 

Procavia habessinica daemon Thomas 

Procavia daemon Thomas, 1910, Ann. Mag. Nat. Hist. (8), 5, p. 199: Elgonyi, 
southern slopes of Mount Elgon at 7,000 feet, Kenya Colony. 

Procavia daemon varians Granvik, 1924, Lunds Univers. Arsskr., N. F., 21, 
No. 3, p. 26: Eastern slopes of Mount Elgon at 7,000 feet, Kenya Colony. 

2 c? (M. C. Z. 31097, 32477) Kirui, Mt. Elgon, K. C. 20 & 23. 
i. 34. 
1 + 5c^ 7 9 (M C. Z. 31096, 31098, 31603-5, 32197-202, 32461-2) Elgonyi, 
K. C. 24-31. i. 34. 

Native names. So far as could be ascertained, the Masai at Elgonyi 
did not differentiate between this and the following species which 
they called mutunyet. 

Coloration. This fine series of topotypes is fairly uniform, although 
the extremes of variation differ obviously ; on the one hand the darkest 
individuals with the crown dark blackish brown, minutely peppered 
with grayish, and the dorsal surfaces much mixed with longer black 
hairs; on the other, the palest skins with the pale rings on the hairs 
of the forehead nearly twice as broad, the black hairs on the dorsum 
much fewer, giving the back a much more buffy tone. The lower side 
is bright ochraceous buffy in all as is also the dorsal spot. 

Measurements. d\ 540. 0. 70. 35 mm., 9 . 580. 0. 72. 39 mm. 

Breeding. At Elgonyi, on January 23, 1934, two of the largest of 
the series each held three fetuses, a third only two of an average 
size of 165. 0. 31. 17 mm. 



120 bulletin: museum of comparative zoology 

Parasites. Fleas swarmed on some individuals while many 
others were infested with nematodes (Crossophorus collaris and 
Trichuris sp.) 

Habitat. Plentiful on the rock-strewn slopes of the escarpment ; 
many were seen in the vicinity of Kemp's Cave both on the cliff face 
and in trees growing from it. 

Heterohyrax syriacus kempi (Thomas) 

Procavia brucei kempi Thomas, 1910, Ann. Mag. Nat. Hist. (8), 5, p. 200: 
Elgonyi, Mount Elgon at 7,000 feet, Kenya Colony. 

9 (M. C. Z. 32476) Kaburomi, U. 28. xii. 33. 
9 (M. C. Z. ) Butandiga, U. 9. i. 34. 

d" 9 (M. C. Z. 32473, 32475) Kirui, K. C. 22-23. i. 34. 
c? (M. C. Z. 32474) Elgonyi, K. C. 28. i. 34. 

Distribution. All these localities are on, or close to, Mount Elgon, 
the first above Loveridge's camp at 11,000 feet, the second and fourth 
at 7,000 feet, while Kirui is at the southern foot about 5,000 feet. 

Native names. Aduhwa (Karamojong) ; kwenera (Kisabei); Icigen- 
erwa (Lugishu); mutunyet (Kimasai); lihenele (Luragoli); haterera 
(Lutereki). 

Coloration. This montane form averages a more uniform gray 
beneath, the chest and underside of forelimbs in particular contrast- 
ing with those of H. s. hindei in which race they are usually pure white 
to the roots of the hair though occasional specimens display a cloudi- 
ness which would render them difficult to distinguish from kempi. 
The character of head pelage being darker than the dorsal pelage, 
utilized by Hahn (1934, pp. 276-277) to distinguish the two forms was 
flatly contradicted by our material for the four adults of kempi con- 
form to Hahn's definition of hindei in this respect. 

Measurements. <?. (Kirui) 470. 0. 67. 31 mm., 9 . (Kirui) 325. 0. 
52. 30 mm. 

Parasites. The stomach of the Kaburomi specimen was full of grass 
but free of parasites so far as could be seen, which came as somewhat 
of a surprise for in Tanganyika Territory most hyraxes support a 
heavy infestation of nematodes and cestodes. 

Enemies. At Sipi (6,500 feet) I was told that hyrax were formerly 
plentiful but had "been driven away by the numerous goats." I 
thought this answer rather imaginative till I visited their haunts 
above Kaburomi in the alpine zone. There I saw a herd of goats, the 
property of one of the cave-dwelling Wanderobo, clambering about a 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 121 

cliff face which had almost certainly been the home of hyrax in former 
times. 

It is more probable, however, that it was the constant hunting by 
voracious Bagishu natives that reduced their numbers at Sipi; even 
at the remote spot where they occurred above Kaburomi, snares and 
traps had been set in their runways by Wanderobo who, like the 
Bagishu, look upon them as a regular source of food. 

Habitat. At Kaburomi they occurred in a rocky gorge which was 
very overgrown with creepers and relatively heavily wooded with tree 
heath; the ravine was about an hour's walk above my camp. 

One of these hyrax was seen just below Jackson's Summit, Mount 
Elgon, at about 13,000 feet, at an unusually early hour. It was 6.40 
a.m., just twenty minutes before the sun rose over the mountain and 
so cold that there was frost on all the vegetation and ice on an adja- 
cent lake. 

Droppings of hyrax were found among the rocks of the Nandi 
Escarpment near Kaimosi. The animal is well known to both Mara- 
goli and Watereki, but has evidently been hunted so persistently that 
all my efforts to secure one during my stay there, failed. 

At Kirui as well as in their type locality, Elgonyi, these animals 
occur on the rocks of the escarpment alongside P. h. daemon. 

Heterohyrax syriacus hindei (Wroughton) 

Procavia brucei hindei Wroughton, 1910, Ann. Mag. Nat. Hist. (8), 5, p. 107: 

Fort Hall, Kenya Colony. 
Procavia brucei maculata Osgood, 1910, Publ. Field Mus. Nat. Hist., Zool. 

Series, 10, p. 6: Lukenya Mountain, Ulukenya Hills, Kenya Colony. 

10 rf" 4 9 (M. C. Z. 31952, 31964, 32338-9, 32385-8, 32463-8) Kibwezi, 

K. C. 24 & 27. iii. 34. 
3 & 5 9 (M. C. Z. 31965-6, 32349-51, 32389-90, 32469) Tsavo, K. C. 

31. iii. 34. 
3 tf 1 8 9 (M. C. Z. 31942-3, 32340-8) Mt. Mbololo, K. C. 19-27. iv. 34. 

Native name. Ngivu (Kitaita). 

Discussion. Hahn (1934, p. 271) considers all the described forms 
of Heterohyrax to be races of the single species syriacus. We are, 
however, not at all convinced that the very small pumila, and its 
subspecies rudolfi, are not a distinct specie?. 

Coloration. The fine series collected is fairly uniform in appearance, 
though the dorsal spot may vary in color from clear white with a 
trace of ochraceous tipping, to nearly entirely ochraceous. 



122 bulletin: museum of comparative zoology 

Measurements, c? . (Kibwezi) 470. 0. 69. 32 mm., $. (Kibwezi) 
480. 0. 65. 30 mm. 

Breeding. At Kibwezi neither of the two females shot on March 24, 
1934, was gravid, but both shot on the 27th were; one held a large 
fetus measuring 145 mm. from snout to anus with a hind foot of 
24 mm., the other held two fetuses of much smaller dimensions. 

On Mount Mbololo, April 26, 1934, a female held a single fetus 
measuring 203. 0. 33. 22 mm.; on the 27th, each of two females held 
two fetuses ready for birth and measuring from 185. 0. 35. 21 mm., 
to 9 . 215. 0. 36. 22 mm. 

Parasites. Both cestodes (Anoplocephala opatula) and nematodes 
(Crossophorus collaris and Setaria hyracis) were numerous in the 
Kibwezi series. 

Habitat. At Kibwezi, a young Mkamba offered to guide me to a 
locality where hyrax were numerous. For a couple of miles he led 
me southwest of the railway line till we reached scrub forest growing 
on the lava which covers so much of this neighborhood. The head- 
quarters of the animals were in volcanic craters which were below 
the general ground level; in these hollows were numerous masses of 
lava and big caverns. Besides shooting some among these rocks, 
others were obtained in trees as they squatted on horizontal limbs or 
in crotches of the main stem up to heights of thirty feet from the 
ground. In such situations they were only seen from 8 to 10 in 
the early morning; as soon as the sun became powerful (11 to 12 noon) 
they descended and were to be detected on gnarled branches in the 
bush at not more than six feet from the ground. The path 
that traversed this wilderness of thicket was much used by natives 
from 6 till 10 a.m. so that the hyrax were accustomed to the sight of 
human beings and appeared to study them with a mildly interested 
curiosity. If, however, one moved from the path, at the first footfall 
on the dead leaves which carpeted this dry scrub, most of these hyrax 
vanished. It was rarely necessary to leave the path and the first seven 
were secured with seven shots (No. 5 from twelve bore). Finding the 
males were in such a heavy preponderance (ratio of 5 to 2), I returned 
three days later and shot seven more but found the proportion re- 
mained the same (5 to 2). Whether this inequality of the sexes is due 
to the females being more wary than the males, or whether it is for- 
tuitous, I cannot say. I judge, however, that the Wakamba do not 
molest the hyrax to any appreciable extent; not one of them came to 
camp asking for their bodies. My Bagishu skinners on the other hand 
made themselves ill with a surfeit of hyrax meat. 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 123 

At Tsavo, hyraxes were found on rocky hills a mile and a half 
north of the station, the environment and temperature conditions 
being vastly different from those obtaining at Kibwezi. 

On Mount Mbololo they occurred in a great jumble of rocks and 
tangled undergrowth at the foot of a precipice on the southern (?) 
side of the mountain at an altitude of about 3,500 feet. 

Dendrohyrax arboreus stuhlmanni (Matschie) 

Procavia stuhlmanni Matschie, 1892, Sitz. Ges. naturf. Freunde Berlin, p. Ill: 
Bukoba, Tanganyika Territory. 

c? (M. C. Z. 31602) Kirui, K. C. 23. i. 34. 
tf (M. C. Z. 31099) Elgonyi, K. C. i. 34. 

Coloration. These two tree hyrax from Mount Elgon are provi- 
sionally referred to this race, following Hahn's 1934 revision, although 
they are a trifle darker brown than a topotype from Bukoba, Tan- 
ganyika Territory. The crown of the head is contrastingly dark 
blackish brown, with little admixture of buffy-ringed hairs, such as 
prevail over the rest of the dorsum. The sides of the body are barely 
paler though the bases of the hairs along the sides are much less dark 
than in the median region. On the belly the hair is clear whitish to 
the bases. The cheeks are mixed pale and blackish brown, the latter 
prevailing, and there is a slight trace of a pale supraorbital spot. 

Measurements. cT. (Elgonyi) 550. 0. 75. 36 mm. 

Dendrohyrax arboreus bettoni (Thomas & Schwann) 

Procavia bettoni Thomas & Schwann, 1904, Abstr. Proc. Zool. Soc. London, 
No. 6, p. 23, April 26: Rogoro, Kikuyu, Kenya Colony. 

Procavia {Dendrohyrax) scheffleri Brauer, 1913, Sitz. Ges. naturf. Freunde 
Berlin, p. 131: Teleki River, Kibwezi, Kenya Colony. 

Procavia {Dendrohyrax) vilhelmi Lonnberg, 1916, Arkiv for Zoologi, 10, No. 
12, p. 26: Donyo Sabuk, Kenya Colony. 

c? (M. C. Z. 31944) Mt. Mbololo, K. C. 17. iv. 34. 

Distribution. This single specimen is doubtless referable to this 
race although the locality, Mount Mbololo, is considerably nearer 
to the coast than Rogoro, whence came the type. It is, however, 
only about seventy-five miles southeast of Kibwezi, type locality 
of scheffleri which Hahn (1934, p. 267) refers to the synonymy. 

Native name. Mbelele (Kitaita). 



124 bulletin: museum of comparative zoology 

Discussion. Though in color the skin of this specimen is practically 
indistinguishable from those of stuhlmanni from Mount Elgon, the 
skull differs notably in the much larger incisors and longer rostrum. 
The greatest length of the nasals is 29 mm. against 23.5 mm. in a 
skull of stuhlmanni of comparable size and age. The median point of 
the frontals lies on the same transverse plane as the front of the orbit 
in front of the lachrymal bone, whereas in the two skulls of stuhl- 
manni from Elgon it falls in advance of the orbit and even in front of 
the anterior root of the zygoma. 

Measurements, cf. 500. 0. 75. 23 mm. Weight 6 pounds. 

Voice. Shortly after my arrival at Ivibwezi station at 2.22 a.m., 
when preparing to sleep on the platform, I heard the raucous cries 
of these hyrax in close proximity to the station; later I ascertained 
that they came from the big trees fringing the river. The cry is like 
an exaggerated and prolonged call of the Square-marked Toad 
(Bufo r. regularis) and tails off like that of the Galago (G. senegalensis 
lasiotis) ; in other words it begins with a sound like a watchman's 
rattle combined with a sawing note and ends with a 'quek-quek' 
crescendo. Immediately after sunrise, I set off and spent a couple of 
hours searching for them in the dense tangle among the big trees 
growing north of the line, but without sighting one. 

Enemies. According to my native employees (Bagishu, Mganda 
and Karamojong), the flesh of this animal was "as bitter as quinine" 
and though they had looked forward to eating it, after a trial they 
rejected most of the meat. 

Habitat. Very abundant in the cap of forest on the summit (4,800 ft.) 
of Mbololo. The trees were such a height, however, and the animals 
so wary that it was only after persistent hunting that I succeeded 
in bringing one down after firing both barrels of a twelve bore loaded 
with No. 3 shot. Even then, after falling from this great height, it 
made off through the undergrowth and might well have escaped had 
not my gunbearer pluckily pursued and secured it. 

DELPHINIDAE 

Prodelphinus attenuatus (Gray) 

Steno attenuatus Gray, 1846, in "Voyage of the Erebus and Terror," Zool. 
p. 44, pi. xxviii: Cape of Good Hope. 

Cranium (M. C. Z. 31734) Malindi, K. C. 30 vi. 34. 

Discussion. This well-preserved cranium, picked up on the sea- 



ALLEN AND LAWRENCE: AFRICAN MAMMALS 125 

shore, agrees essentially with True's description of this species. 
Although the tips of the premaxillaries are slightly beachworn, the 
maxillary bones show forty-nine tooth sockets on each side. 



DUGONGIDAE 

Dugong dugon (Muller) 

Trichechus dugon P. S. L. Muller, 1776, in Linne, Vollstandiger Natursystems, 
Suppl., p. 21: "Vorgeburge der Guten Hofnung an, bis an die philippin- 
ischen Inseln." 

Halicore dugung Erxleben, 1777, Syst. regn. animal., Classis I, Mammalia, 
p. 599 

Discussion. Miiller's name dugon clearly antedates Erxleben's 
dugung for the same animal. 

Field Note. Dugong are said by the natives to be fairly common 
along the Lamu coast. While I was staying at Lamu, some fishermen 
took one of these animals in their nets and brought it in their boat for 
my inspection. The meat has a high commercial value, nevertheless, 
I was surprised that they refused my offer of ten shillings ($2.50) to 
be allowed to skin the animal, retaining skin and skull only. Possibly 
Moslem laws influenced their decision. 



126 bulletin: museum of comparative zoology 



BIBLIOGRAPHY 

Allen, Glover M. and Loveridge, A. 

1927. "Mammals from the Uluguru and Usambara Mountains, Tangan- 
yika Territory." Proc. Boston Soc. Nat. Hist., 38, pp. 413-441. 

1933. "Reports on the Scientific Results of an Expedition to the South- 
western Highlands of Tanganyika Territory. II. Mammals." Bull. 
Mus. Comp. Zool., 75, pp. 45-140, pi. i. 

Granvik, Hugo , 

1924. "Mammals from the Eastern Slopes of Mount Elgon, Kenya 
Colony." Lunds Univers. Arsskrift (2), 21, No. 3, pp. 1-32, pi. i-ii 
and text figs. 

Hahn, Herbert 

1934. "Die Familie der Procaviidae." Zeitschr. f. Saugetierkunde, 9, pp. 
207-358, figs. 1-69, pis. xiii-xvi. 

Hollister, Ned 

1918. "East African Mammals in the United States National Museum. 

I. Insectivora, Chiroptera, and Carnivora." U. S. Nat. Mus. Bull. 
99, pp. 1-194, figs. 1-3, pis. i-lv. 

1919. "East African Mammals in the United States National Museum. 

II. Rodentia, Lagomorpha, and Tubulidentata." U. S. Nat. Mus. 
Bull. 99, pp. 1-184, fig. 1, pis. i-xliv. 

1924. "East African Mammals in the United States National Museum. 

III. Primates, Artiodactyla, Perissodactyla, Proboscidea, and 
Hyracoidea." U. S. Nat. Mus. Bull. 99, pp. 1-164, fig. 1, pis. i-vii. 

Lydekker, Richard and Blaine, G. 

1914. "Catalogue of the Ungulate Mammals in the British Museum 
(Natural History)." 8vo, London, 2, pp. i-xiii and 1-295, figs. 
1-33. 

Schwarz, Ernst 

1928a. "Notes on the Classification of the African Monkeys in the genus 

Cercopithecus, Erxleben." Ann. Mag. Nat. Hist. (10), 1, pp. 

649-663. 
1928b. "The Species of the genus Cercocebus, E. Geoffroy." Ann. Mag. 

Nat. Hist. (10), 1, pp. 664-670. 
1928c. "Bemerkungen uber die roten Stummelaffen." Zeitschr. f. Siiu- 

getierk., 3, pp. 92-97. 
1929. "On the local Races and Distribution of the Black and White 

Colobus Monkeys." Proc. Zool. Soc. London, pp. 585-598, map. 
1931a. "On the African Long-tailed Lemurs or Galagos." Ann. Mag. 

Nat. Hist. (10), 7, pp. 41-66. 
1931b. "On the African Short-tailed Lemurs or Pottos." Ann. Mag/ Nat. 

Hist. (10), 8, pp. 249-256. 



EXPLANATION OF PLATES 



PLATE 1 



Allen and Lawrence — African Mammals. 



PLATE 1 

Fig. 1. Eastern Potto (Perodicticus potto ibeanus) 

This arboreal relative of the lemurs progresses on all slender boughs in a 
sloth-like fashion with its back downwards, it also descends tree trunks head 
foremost. 

Fig. 2. A Potto turning to descend to the ground 

Natives assert that if a potto is seized by the nape, the points of its cervical 
vertebrae, which protrude through the skin, are brought into defensive action 
by a backward jerk of the head. Both photographs taken at Kaimosi of an 
animal captured in vicinity, 



BULL. MUS. COMP. ZOOL. 



Allen and Lawrence. African Mammals. Plate 1. 





PLATE 2 



Allen and Lawrence — African Mammals. 



PLATE 2 

Fig. 1. Scalytail Flying Squirrel (A no malurus jacksoni) 

These large squirrels, measuring two feet from nose to tip of tail, are com- 
mon to the forests of both Kakamega and Mount Elgon. The figured specimen 
is dead, having just been shot through the chest by the arrow of an Mgishu 
hunter. 

Fig. 2. An Arboreal mouse (Thamnomys surdaster elgonis) 

This specimen was photographed at Kaimosi, Kakamega, so is a topotype 
of the synonym Thamnomys s. discolor. This animal was trapped by utilizing 
banana as a bait; nests found in wild bananas were attributed to these mice by 
Watereki natives. 



BULL. MUS. COM P. ZOOL. 



Allen and Lawrence. African Mammals. Plate 2. 





PLATE 3 



Allen and Lawrence — African Mammals. 



PLATE 3 

Fig. 1. Young gray forest dormice (Claviglis saturatus) 

These delightful little creatures, taken in mid-January from their nest at 
Butandiga on the western slopes of Mount Elgon, are shown feeding on bread 
and milk; the lid of a shaving-stick container serves them as a bowl. 

Fig. 2. An adult pygmy mouse (Leggada grata grata) 

Many species of pygmy mice were encountered during the trip. After re- 
peated, but unsuccessful, attempts to photograph some in their natural sur- 
roundings, the tail of one was held between the fingers whose owner hoped that 
the mouse would not retaliate with its sharp little teeth. Sipi, Mount Elgon, 
Uganda. 



BULL. MUS. COMP. ZOOL. 



Allen and Lawrence. Akrican Mammals. Plate 3. 





PLATE 4 



Allen and Lawrence — African Mammals. 



PLATE 4 
Fig. 1. Swamp rat (Otomys tropicalis elgonis) and young 

Swamp rats are abundant on western Elgon where they are preyed upon by 
genets, tree civets, mongoose and wild cats. The animal in the photograph is 
actually suckling two young. She was afterwards released. Sipi, Mt. Elgon. 

Fig. 2. A brush-tailed porcupine (Atherurus turneri) 

A good series of topotypes of this recently-described species of forest-dwell- 
ing porcupine was obtained at Kaimosi, Kakamega, Kenya Colony, where the 
genus reaches the eastern limits of its distribution in Africa. 



BULL. MUS. COMP. ZOOL. 



Allen and Lawrence. African Mammals. Plate 4. 





PLATE 5 



Allen vnd Lawrence — African Mammals. 



PLATE 5 
Fig. 1. A young waterbuck (Cobus ellipsiprymnus kuru) 

In the text will be found an account of how one of these young antelope was 
driven into Tsavo River by a pack of hunting dogs. The photograph, however, 
was taken at Witu, and is reproduced here by the courtesy of Mr. R. D. Milne. 

Fig. 2. A dugong (Dugong dugon) at Lamu Island, K. C. 

Dugong are relatively common along the Lamu coast where they are highly 
prized by the natives for their flesh. Fishermen offered to sell one to Mr. Lov- 
eridge and brought it in their boat for his inspection. We are indebted to 
J. MacDougall, Esq., of Lamu, for the photograph. 



BULL. MUS. COMP. ZOOL. 



Allen and Lawrence. African Mammals. Plate 5. 





n -l 



Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 

Vol. LXXIX, No. 4 



°r Comparative J 
Harvard lta lv » Pslt 



SCIENTIFIC RESULTS OF AN EXPEDITION 
TO RAIN FOREST REGIONS IN EASTERN AFRICA 

IV 
BIRDS 



By James Lee Peters and Arthur Loveridge 



With Two Plates 



CAMBRIDGE, MASS., U. S. A. 

PRINTED FOR THE MUSEUM 
January, 1936 



No. 3. — Reports on the Scientific Results of an Expedition 
to Rain Forest Regions in Eastern Africa 

IV 

Birds 
By James Lee Peters and Arthur Loveridge 

CONTENTS Page 

Introduction 129 

List of Species collected 131 

Systematic discussion 138 

Bibliography 204 

INTRODUCTION 

The collection on which the following report is based, was made 
by the junior author while investigating the herpetological fauna of 
certain rain forest areas in Uganda and Kenya. This was carried out 
on behalf of the Museum of Comparative Zoology with a fellowship 
granted by the John Simon Guggenheim Foundation of New York. 

Altitudes and other information regarding the localities where 
collecting was done, will be furnished in the final report. In addition 
to giving the itinerary this paper will deal with the conclusions 
arrived at regarding these forest faunae in relation to those of the 
Usambara and Uluguru Mountains in Tanganyika Territory. 

In view of the intensive study already directed to the ornithology 
of this region by van Someren and Granvik, less attention was directed 
to this group of vertebrates than would have been the case otherwise. 
The period of bird collecting was from November 9, 1933 to June 29, 
1934, during which time 530 skins representing 228 species, or races, 
of birds were secured. Twenty of these were migrants; such are dis- 
tinguished by an asterisk in the following pages. 

So well is the avifauna of this region represented in the Museum 
of Comparative Zoology that only eight forms proved new to the col- 
lections. Only one race and one species appeared to require descrip- 
tion, viz. 

Tyto capcnsis libratus from Kaimosi, K. C. 



130 bulletin: museum of comparative zoology 

Zosterops silvanus from Mount Mbololo, K. C. and this has been 
done. 1 

Besides discussions on variation and synonymy, field notes are 
included on the nesting habits and diet of certain species including 
eight forms additional to those of which skins were collected. In 
this connection attention is directed to what is, according to a letter 
from Mr. W. L. Sclater, apparently the first record of the eggs of 
Ixobrychus sturmii. The statements in the literature as to the laying 
of blue eggs by this species appear to be erroneous, originally based 
on supposition and subsequently copied by one author after another. 
In connection with these field notes occasionally it has been found 
necessary to use the personal pronoun, in such instances it refers to 
the junior author as collector. 

Parasites are treated under a separate heading following the species 
in, or on, which they were found. Here we should like to express our 
indebtedness to our colleagues, Dr. J. Bequaert and Dr. J. H. Sand- 
ground for marking the determinations of the ecto- and endopara- 
sites respectively. 

We take this opportunity of thanking Messrs. Claude H. B. Grant, 
Herbert Friedman and W. L. Sclater for their kindness in loaning 
or comparing specimens of certain green pigeons, thrushes or white- 
eyes. Mr. Sclater's monumental work, the Systema Avium Aethi- 
opicarum, has been of the greatest assistance. Reference to other 
works consulted will be found in the bibliography at the end of this 
paper. 

The order of families adopted is that of Wetmore (1934) while the 
arrangement of species is according to Peters (1931 and 1934) to the 
end of the Burhinidae, thereafter following that of Sclater (1924 and 
1930). As this rearrangement may prove somewhat confusing and 
render difficult the locating of a species in the text, we provide an 
indexed list of species, as follows. 

1 Peters and Loveridge, 1935, Proc. Biol. Soc. Washington, 48, p. 77. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 131 

List of Species Collected* 
PELECANIDAE Page 
Pelecanus rufescens Gmelin 138 

PHALACROCORACIDAE 

Halietor africanus africanus (Gmelin) 139 

ARDEIDAE 

Ardea cinerea cinerea Linne 139 

Camerodius albus melanorhynchus (Wagler) 139 

Ixobrychus sturmii (Wagler) 139 

CICON1IDAE 

Ibis ibis (Linne) 140 

Anastomus lamelliger us lamelligerus Temminck 141 

Leptoptilos crumeniferus (Lesson) 141 

THRESKIORNITHIDAE 

Threskiornis aethiopica aethiopica (Latham) 141 

Hagedashia hagedash nilotica Neumann 142 

ANATIDAE 

Sarkidiornis melanonota (Pennant) 142 

Anas sparsa leucostigma Riippell 142 

Anas punctata punctata Burchell 143 

ACCIPITRIDAE 

Elanus caeruleus caeruleus (De?fontaines) 142 

Milvus migrans parasitus (Daudin) 142 

Melierax musicus polioptcrus Cabanis 144 

Melierax metabates metabates Heuglin 144 

Buteo rufofuscus augur Riippell 145 

Buteo vulpinus vulpinus (Gloger) 145 

Kaupifalco monogrammicus (Temminck) 145 

Butastur rufipennis (Sundevall) 146 

Lophaetus occipitalis (Daudin) 146 

Aquila wahlbergi Sundevall 147 

Haliaeetus vocifer damans C. L. Brehm 147 

Necrosyrtes vionachus monachus (Temminck) 148 

*Circus macrourus (S. G. Gmelin) 148 

*Circaetus gallicus gallicus (Gmelin) 148 

Circaetus fasciolatus Gurney 148 

* An asterisk denotes a migrant. 



132 bulletin: museum of comparative zoology 

FALCONIDAE Page 

Polihierax semitorquatus deckeni Zedlitz 149 

*Falco tinnunoulus tinnunculus Linne 149 

Falco tinnunculus carlo (Hartert & Neumann) 149 

PHASIANIDAE 

Francolinus sephaena grantii Hartlaub 150 

Francolinus clappertoni gedgii Ogilvie-Grant 150 

Francolinus hildebrandti hildebrandti Cabanis 150 

Francolinus squamatus zappeyi Mearns 151 

Pternistis afer leucoparaeus (Fischer & Reichenow) 151 

NUMIDLDAE 

Numida meleagris major Hartlaub 151 

Guttera edouardi sethsmithi Neumann 152 

RALLIDAE 

Lvmnocorax flavirostra (Swainson) 152 

Sarothrura rufa elizabethae van Someren 152 

Sarothrura pulchra centralis Neumann 152 

Sarothrura elegans reichenovi (Sharpe) 153 

CHARADRIIDAE 

Hoplopterus spinosus (Linne) 153 

Charadrius pecuarius pecuarius Temminck 1 53 

BURHINIDAE 

*Burhinus oedicnemus oedicnemus (Linne) 153 

Burhinus vermiculatus vermiculatus (Cabanis) 153 

PTEROCLIDIDAE 

Pterocles decor atus decor atus Cabanis 154 

COLUMBIDAE 

Treron waalia (Meyer) 154 

Treron calva salvadorii (Dubois) 155 

Treron calva wakefieldii Sharpe 155 

Columba arquatrix arquatrix (Temminck) 155 

Streptopelia semitorquata semitorquata (Riippell) 155 

Streptopelia decipicns perspicillata (Fischer & Reichenow). ... 156 

Streptopelia capicola tropica (Reichenow) 156 

Aplopelia larvata larvata (Temminck) 156 

Turtur chalcospilos chalcospilos (Wagler) 156 

Tympanistria tympanistria fraseri Bonaparte 157 



PETERS AND LOVERIDGE: AFRICAN BIRDS 133 

PSITTACIDAE Page 

Poicephalus rufiventris rufiventris (Riippell) 157 

Poicephalus meyeri saturatus Sharpe 157 

MUSOPHAGIDAE 

Turacus hartlaubi (Fischer & Reichenow) 157 

Turacus leucolophus (Heuglin) 158 

Musophaga rossac rossae Gould 158 

Corythaeola cristata yalensis (Mearns) 158 

Corythaixoides leucogaster (Riippell) . 159 

CUCULIDAE 

Lampromorpha klaasi (Stephens) 159 

Centropus monachus fischeri Reichenow 1 59 

Centropus superciliosus superciliosus Hemprich & Ehrenberg ... 1 59 

TYTONIDAE 

Tyto capensis libratus Peters & Loveridge 160 

STRIGIDAE 

Strix woodfordii suahelica (Reichenow) 160 

Bubo africanus africanus (Temminck) 160 

Bubo lacteus (Temminck) 161 

CAPRIMULGIDAE 

*Caprimulgus europaeus europaeus Linne 161 

*Caprimulgus europaeus meridionalis Hartert 161 

*Caprimulgus europaeus unwini Hume 161 

Caprimulgus ?iubicus taruensis van Someren 162 

Caprimulgus fossil clarus Reichenow 162 

COLIIDAE 

Colius striatus mombassicus van Someren 163 

Colius striatus jebelensis Mearns 163 

TROGONIDAE 

Apaloderma narina narina (Stephens) 1 63 

ALCEDINIDAE 

Ceryle rudis rudis (Linne) 164 

Halcyon senegaloides ranivora Meinertzhagen 164 

MEROPIDAE 

Melittophagus pusillus cyanostictus (Cabanis) 164 

M ditto phagus lafrcsnayii oreobates Sharpe 164 



134 bulletin: museum of comparative zoology 

CORACIIDAE Page 

*Coracias garrulus garrulus Linne 165 

Coracias caudatus lorti Shelley 165 

PHOENICULIDAE 

Phoeniculus purpureus marwitzi (Reichenow) 166 

Phoeniculus bollei jacksoni (Sharpe) 166 

BUCEROTIDAE 

Bycanistes bucinator (Temminck) 166 

Bycanistes subcylindricus (Sclater) 166 

Lophoceros erythrorhynchus erythrorhynchns (Temminck) 167 

Lophoceros flavirostris flavirostris (Riippell) 167 

Lophoceros jacksoni Grant 167 

Lophoceros melanoleucos geloensis Neumann 167 

INDICATORIDAE 

Indicator indicator (Sparrman) 168 

CAPITONIDAE 

Tricholaema lacrymosum lacrymosum Cabanis 168 

Tricholaema melanocephalum stigmatothorax Cabanis 169 

Gymnobucco bonapartei cinereiceps Sharpe 169 

Trachyphonus darnaudii damaudii (Prevost et Des Murs) .... 169 

Trachyphonus darnaudii boehmi Fischer & Reichenow 169 

PICIDAE 

Campethcra nubica pallida (Sharpe) 169 

Dendropicos fuscescens massaicus Neumann 169 

Dendropicos fusccscens hemprichii (Ehrenberg) 170 

Dendropicos lafresnayi lepidus (Cabanis & Heine) 170 

Yungipicus obsoletus ingens Hartert 170 

EURYLAEMIDAE 

Smithomis capensis meinertzhageni van Someren 170 

ALAUDIDAE 

Mirafra poccilosterna massaica (Fischer & Reichenow) 170 

HIRUNDINIDAE 

*Hirundo rustica rustica Linne 1 70 

Hirundo angolensis arcticincta Sharpe 171 

Hirundo rufula emini Reichenow 171 

Ptyonoprogne rufigula rufigula (Fischer & Reichenow) 171 



PETERS AND LOVERIDGE: AFRICAN BIRDS 135 

CAMPEPHAGIDAE Page 

Campephaga quiscalina martini Jackson 17.1 

Coracina caesia pura (Sharpe) 172 

DICRURIDAE 

Dicrurus adsimilis divaricatus (Lichtenstein) 172 

Dicrurus adsimilis fug ax Peters 172 

ORIOLIDAE 

*Oriolus oriolus oriolus (Linne) 172 

Oriolus monacha rolleti Salvadori 173 

CORVIDAE 

Corvus albus Miiller 173 

PARIDAE 

Par us albiventris albiventris Shelley 173 

Anthoscopus musculus (Hartlaub) 1 74 

TIMELIIDAE 

Argyra rubiginosa heuglini Sharpe 174 

PYCNONOTIDAE 

Pycnonotus tricolor dodsoni Sharpe 174 

Pycnonotus tricolor fayi Mearns 175 

Phyllastrcphus sucosus sucosus Reichenow 175 

Phyllastrcphus fischeri placidus (Shelley) 1 75 

Arizelocichla milanjensis striifacies (Reichenow & Neumann) . . 175 

Stelgidocichla latirostris eugenia (Reichenow) 176 

Eurillas virens holochlorus van Someren 176 

TURDIDAE 

T urdus olivaceus helleri (Mearns) 1 76 

Turdus tcphronotus Cabanis 177 

*Monticola saxatilis (Linne) 178 

*Oenanthe oenanthc ocnanthe (Linne) 178 

*Oenanthe leucomela leucomela (Pallas) 178 

*Oenanthe isabellina (Temminck) 178 

Pinarcchroa sordida rudolphi Madarasz 178 

Cossypha heuglini heuglini Hartlaub 179 

Cichladusa guttata rufipennis Sharpe 179 

Pogonocichla margaritifcra helleri Mearns 179 



136 bulletin: museum of comparative zoology 

SYLVIIDAE Page 

*Sylvia borin (Boddaert) 179 

* Sylvia atricapilla atricapilla (Linne) 180 

*Phylloscoptis trochilus trochilus (Linne) 180 

Seicercus ruficapilla minulla (Reichenow) 180 

Seicercus umbrovirens mackenzianus (Sharpe) 180 

Calamonastes simplex simplex (Cabanis) 181 

Apalis flavocincta (Sharpe) 181 

Apalis porphyrolaema porphyrolaema Reichenow & Neumann . 181 

Apalis rufifrons rufidorsalis (Sharpe) 181 

Sylvietta brachyura leucopsis (Reichenow) 181 

Sylvietta leucophrys leucophrys (Sharpe) 181 

Eremomela pusilla elgonensis van Someren 182 

Ercmomela badiceps turneri van Someren 182 

Camaroptera brcvicaudata aschani Granvik 182 

Camaroptera brevicaudata, griseigula Sharpe 182 

Cisticola chubbi Sharpe 183 

Cisticola galactotes haematocephala Cabanis 183 

Cisticola troglodytes troglodytes (Antinori) 183 

Melocichla meritalis orientalis (Sharpe) 184 

Prinia mistacea imimdabilis van Someren 184 

Prinia mistacea tenella (Cabanis) 184 

Prinia somalica erlangeri Reichenow 184 

Prinia leucopogon reichenowi (Hartlaub) 184 

Prinia bairdii melanops (Reichenow & Neumann) 185 

MUSCICAPIDAE 

Alseonax minimus inter positus van Someren 185 

Alseonax minimus murinus Fischer & Reichenow 185 

Bradornis pallidus subalaris Sharpe 185 

Bradornis pallidus suahelicu-s van Someren 186 

Dioptrornis fischeri Reichenow 186 

Batis molitor puella Reichenow 186 

Batis minor suahelicus Neumann 186 

Platysteira peltata jacksoni Sharpe 186 

Erranornis longicauda teresita (Antinori) . . .• 187 

Tchitrea viridis viridis (Miiller) 187 

MOTACILLIDAE 

*Motacilla cinerea cinerea Tunstall 187 

Anthus sordidus longirostris Neumann 187 

Anthus richardi lacuum Meinertzhagen 187 

* Anthus trivialis trivialis (Linne) 188 

Macronyx croceus croceus (Vieillot) 188 



PETERS AND LOVERIDGE : AFRICAN BIRDS 137 

LANIIDAE Page 

Lanius e.vcubitorius princeps Cabanis 188 

Lanius mackinnoni Sharpe 189 

*Lanius collurio Linne 189 

Corvinella corvina chapini Friedmann & Bowen 189 

Laniarius ferrugineus major (Hartlaub) 189 

Laniarius liihderi castaneiceps Sharpe 189 

Dryoscopus ajjinis (Gray) 190 

Tschagra australis emini (Reichenow) 190 

Tschagra senegala armena (Oberholser) .' 190 

Tschagra senegala orientalis (Cabanis) 190 

Tschagra jamesi mandana (Neumann) 190 

Rhodophoneus cruentus cathemagmenus (Reichenow) 191 

PRIONOPIDAE 

Prionops cristata omoensis Neumann 191 

STURNIDAE 

Cinnyricinclus leucogaster verreauxi (Bocage) 191 

Lamprocolius chalybeus sycobius Hartlaub 192 

Lamprocolius splendidus splendidus (Vieillot) 192 

Lamprotornis purpuropterus purpuropterus Riippell 192 

Onychognathus morio shelleyi (Hartert) 193 

Onychognathus morio ruppellii (Verreaux) 193 

Spreo superbus (Riippell) 193 

Buphagus erythrorynchus caffer Grote 193 

NECTARINIIDAE 

Nectarinia formosa centralis van Someren 194 

Nectarinia kilimensis kilimensis Shelley 194 

Cinnyris bifasciatus tsavoensis van Someren 194 

Cinnyris venustus igneiventris Reichenow 194 

Cinnyris venustus falkensteini Fischer & Reichenow 195 

Cinnyris reichenowi reichenowi Sharpe 195 

Chalcomitra amethystina kalckreuthi Cabanis 195 

Chalcomitra amethystina doggetti (Sharpe) 195 

Chalcomitra senegalensis aequatorialis (Reichenow) 195 

Chalcomitra verticalis viridisplendens (Reichenow) 195 

Cyanomitra olivacea ragazzii (Salvadori) 196 

Anthreptes collaris ugandae van Someren 196 

ZOSTEROPIDAE 

Zoster ops virens jacksoni Neumann 196 

Zosterops virens stuhlmanni Reichenow 197 

Zost crops silvanus Peters & Loveridge 197 



138 bulletin: museum of comparative zoology 

PLOCEIDAE Page 

Dinemellia dinemelli dinemelli (Riippell) 197 

Plocepasser mahali melanorhynchus Bonaparte 198 

Passer griseus gongonensis (Oustalet) 198 

Gymnoris pyrgita massaica Neumann 198 

Ploceus insignis insignis (Sharpe) 198 

Ploceus melanogaster stephanophorus (Sharpe) 198 

Ploceus nigricollis nigricollis (Vieillot) : 198 

Ploceus nigricollis melanoxanthus (Cabanis) 199 

Ploceus bojeri bojeri (Cabanis) 199 

Euplectes nigroventris Cassin 200 

Euplectes capensis xanthomelas Riippell 200 

Uraeginthus bengalus brunneigularis Mearns 200 

Uraeginthus bengalus ugogoensis Reichenow 200 

Uraeginthus bengalus ugandae Zedlitz 201 

Granatina ianthinogaster ianthinogaster (Reichenow) 201 

Colius passer ardens suahelica (van Someren) 201 

Spermophaga ruficapilla ruficapilla (Shelley) 201 

Pytilia melba kirki Shelley 201 

Estrilda astrild nyanzae Neumann 202 

Estrilda nonnula nonnula Hartlaub 202 

FRINGILLIDAE 

Poliospiza angolensis reichenowi (Salvadori) 202 

Polios piza striolata ugandae van Someren 202 

Emberiza flaviventris flaviventris Stephens 203 



Systematic Discussion of the Species Collected 

PELECANIDAE 

Pelecanus rufescens Gmelin 

Pelecanus rufescens Gmelin, 1789, Syst. Nat., 1, pt. 2, p. 571: West Africa. 

<? (M. C. Z. 168501) Wema, Tana River, K. C. 11 June 1934. 
9 (M. C. Z. 168502) Golbanti, Tana R., K. C. 23 June 1934. 

Native name. Hajowa (Kipokomo). 

Parasites. Numerous nematodes (Contracaectim spicidigerum) in 
stomach of 9 . 



PETERS AND LOVERIDGE: AFRICAN BIRDS 139 

PHALACROCORACIDAE 

Halietor africanus africanus (Gmelin) 

Pelecanus africanus Gmelin, 1789, Syst. Nat., 1, pt. 2, p. 577: Africa. 

9 (M. C. Z. 168503) Wema, Tana River, K. C. 13 June 1934. 

Native name. Samaluve (Kipokomo). 

Diet. The tadpole of a frog, with hind limbs already developed, 
was recovered from the stomach of this bird. 



ARDEIDAE 

Ardea cinerea cinerea Linne 

A rdea cinerea Linne, 1758, Syst. Nat., ed. 10, 1, p. 143: Europe, (restricted 
type locality, Sweden. Cf. Hartert, 1920, Vog. pal. Fauna, 2, p. 1229). 

9 (M. C. Z. 168504) Wema, Tana River, K. C. 14 June 1934. 

Diet. Four large shrimps, Palaemon (Parapalaemon) dolichodactylus 
Hilgendorf, had just been swallowed and were preserved. 

Parasites. Three nematodes {Porrocaccum serpcntulusf) were 
recovered from the stomach. 

Casmerodius albus melanorhynchos (Wagler) 
Ardea Melanorhynchos Wagler, 1827, Syst. Av., Addit.: Senegambia. 

<? (M. C. Z. 168505) Wema, Tana River, K. C. 12 June 1934. 

Native name. Nzari (Kipokomo). 

Variation. Bill brownish black, somewhat yellow at the base of the 
lower mandible. This is the condition described for September in 
West Africa (Bannermann, 1930, 1, p. 64). Some of the nuptial 
plumes have been lost and the remainder are somewhat stained and 
abraded. 

Diet. In the stomach was a tadpole and a lungfish (Protopterus 
aethiopicus). 

Ixobrychus sturmii (Wagler) 
Ardea Sturmii Wagler, 1827, Syst. Av., Ardea sp. 37: Senegambia. 
1 egg near Witu, K. C. 2 June 1934. 

Native name. Flangus (Kipokomo). During the return journey 
from Lamu to Malindi, the big rains were in full force and the coastal 



140 bulletin: museum of comparative zoology 

plains in a semi-flooded state. This was particularly the case in the 
Tana delta for the river had overflowed its banks. 

Frequently one disturbed a small heron feeding about the pools 
of rainwater lying in one's path, or in the water-logged grass-lands. 
In size it appeared to me to be about that of Butorides a. atricapillus 
but it struck me as being a species with which I was unfamiliar. I 
made a note at the time that it was "a bright blue little heron with 
richly coloured yellow legs which are very conspicuous in flight." In 
all I flushed a score of these birds but never shot one as I was usually 
carrying a frogging-net at the time or, if I had my gun, the cartridge 
held shot too large for so small a bird at close quarters. 

Breeding. On May 29, at Mkonumbi, a nest was found in the spiky 
fronds of a small dom palm, little larger than a bush, growing from 
knee-deep water in an extensively flooded area. The nest held a single, 
almost white, egg. Supposing that the clutch consisted of at least 
two, I left the egg untouched. Returning on the morning of the 30th 
I observed the bird leave the nest before I was within gunshot, she 
flew straight away. There was still only one egg. Again I left it, 
hoping to return, but the arrival of the porters necessitated my leav- 
ing for Witu. 

On June 2, at Witu, I disturbed one of these herons as I approached 
a small pond. On going to the spot from which the bird had flown, I 
found a nest, consisting of a shallow platform of sticks, on the flat 
branch of a dead mimosa. The nest was two feet above the water 
and in a very exposed position, the water at this spot was about two 
feet deep. The nest held one white egg measuring 35 x 26 mm. On 
close examination a faint bluish tinge can be seen on the shell but 
nothing like the rich blue shade exhibited by a pair of eggs of 
B. a. atricapillus which I collected at Kilosa on July 7, 1921. These 
latter measured 34 x 28 mm. 



CICONIIDAE 

Ibis ibis (Linne) 

Tantalus Ibis Linne, 1766, Syst. Nat., ed. 12, 1, p. 241: Egypt. 

9 9 (M. C. Z. 168506-7) Wema, Tana River, K. C. 11 & 15 June 1934. 

Native name. Gole (Kipokomo). 

Large flocks of Wood Ibis were to be seen feeding in the flooded 
marshes on either side of the river. During the middle of the day 



PETERS AND LOVERIDGE: AFRICAN BIRDS 141 

single birds or small parties might be seen resting in the trees along 
the banks or in the forests. 

Anastomus lamelligerus lamelligerus Temminck 

Anastomus lamelligerus Temminck, 1823, PL Col., livr. 40, pi. 236: Senegal. 
9 (M. C. Z. 168508) Wema, Tana River, K. C. 11 June 1934. 
Native name. Kokola (Kipokomo). 

Leptoptilos crumeniferus (Lesson) 

Ciconia crumenifera Lesson, 1831, Traite d'Orn., livr. 8, p. 585: Senegal. 

9 (M. C. Z. 168509) Western foot of Mt. Debasien, U. 16 November 1933. 

Native name. Achulutu (Karamojo). 

Breeding. This bird was shot as it circled overhead, it planed for 
half-a-mile and was picked up dead with an S. S. G. bullet in its chest. 
The oviduct held one large egg still lacking a shell but measuring 
74 x 53 mm. ; in addition there were four others, spherical in shape and 
measuring 40 mm., 35 mm., 28 mm. and 20 mm. respectively. 

On November 23, I was descending a ravine on the mountain at a 
point about two flight miles distant from where I had shot the bird. 
As we approached an enormous mvuli (Chlorophora sp.) tree growing 
in this ravine we were momentarily startled by a deep grunting note. 
It was something like the warning call of a guereza, something like 
the moo-moo of a cow. My gunbearer remarked: "A coward, hearing 
that would probably suppose that it was a lion and run away." On 
the noise being repeated I traced it to a colony of marabou which 
were nesting in the mvuli. About six birds were distinguished, they 
were sitting or standing on their relatively small nests; the latter were 
widely scattered and not easy to detect among the dense foliage at 
such a height. The position of the tree in relation to sea level was in 
the neighbourhood of 4,500 feet. 



THRESKIORNITHIDAE 
Threskiornis aethiopica aethiopica (Latham) 
Tantalus aethiopicus Latham, 1790, Ind. Orn., p. 706: "Aethiopia" = Egypt. 
Native name. The Sacred Ibis is known as Nycngc to the Wapokomo. 



142 bulletin: museum of comparative zoology 

Hagedashia hagedash nilotica Neumann 

Hagedashia hagedash nilotica Neumann, 1909, Ornis, 13, p. 193: Kimo, north- 
west of Addis Ababa, Ethiopia. 

9 (M. C. Z. 168510) Mt. Debasien, U. 16 November 1933. 

Corrigenda. The bird from Mwaya, Lake Nyasa, identified as 
nilotica by Bangs and Loveridge (1933, p. 148) is not adult. It should, 
therefore, probably be referred to H. h. erlangeri Neumann. 

Distribution. Granvik (1934, p. 4) has recorded this form from the 
western slopes of Mt. Elgon where he found it uncommon. At my 
camp at 5,000 feet on the western slopes of Mt. Debasien, the noisy 
cries of the Nile Valley Hadadas were a familiar sound at dawn; they 
were then apparently returning from their fishing grounds. The bird 
that was shot had been fishing in the Amaler River. 

Diet. Its stomach held much mud and the remains of crabs. 



ANATIDAE 

Sarkidiornis melanonota (Pennant) 
Anser melanonotos Pennant, 1769, Indian Zool., p. 12, pi. xi: Ceylon. 

c? (M. C. Z. 168511) Wema, Tana River, K. C. 19 June 1934. 

Anas sparsa leucostigma Ruppell 

Anas leucostigma Ruppell, 1845, Syst. Uebers. Vog. N.-O. Afr., pp. 130, 138, 
pi. xlviii: Abyssinia (i.e. Ethiopia). 

c? 9 (M. C. Z. 168512-3) Sipi, w. Mt. Elgon, U. 15 December 1933. 

Native names. Tarratetabe (Kisebei^ ; Liduvi (Lugishu). 

Breeding. On December 11, the evening of the day of my arrival 
at Sipi, two Black Duck flew past, one carrying a piece of sedge or 
grass in its bill. On the 15th, a native brought in alive this pair of 
birds which he had snared. To prevent their escape, he had cruelly 
broken their wings and a leg. I purchased and chloroformed them 
immediately, refusing to buy any more birds during my stay. 

The following day a youngster brought in a duck egg. Supposing 
it to be a domestic product, I paid little attention to him, referring 
him to the cook who purchased for the commissariat. Next morning a 
couple of eggs were served up for breakfast, but on opening them 
feathered ducklings were found within. On enquiry, I then learned 



PETERS AND LOVERIDGE: AFRICAN BIRDS 143 

that the vendor of the duck's egg had later produced six more, and 
that he was the same lad who had brought in the ducks. It seems 
probable, therefore, that the seven eggs represented a clutch and that 
the birds brought to me had been snared at their nest. If so, it fur- 
nishes further evidence of the destruction of bird life involved in 
purchasing from natives. 

Anas punctata punctata Burchell 

Anas punctata Burchell, 1822, Travels, 1, p. 283, note: Zak River, Cape 
Province. 

d> (M. C. Z. 168514) Karawa, near Malindi, K. C. 26 June 1934. 

This bird was one of a flock of a dozen paddling and feeding on 
shallow, extensively flooded flats; two other parties of less than half-a- 
dozen duck were seen. If van Someren (1930, p. 63) is correct in 
supposing that Hottentot Teal nest in Kenya in May and June, then 
it looks as if these birds had already reassembled in flocks. 



ACCIPITRIDAE 

Elanus caeruleus caeruleus (Desfontaines) 

Falco caeruleus Desfontaines, 1789 (1787), Hist. (Mem.) Acad. Paris, p. 503, 
pi. xv : Algiers. 

9 (M. C. Z. 168515) Kaimosi, K. C. 23 February 1934. 

Breeding. Non-breeding bird in immature plumage with stomach 
empty. 

Milvus migrans parasitus (Daudin) 

Falco parasitus Daudin, 1800, Traite d'Orn., 2, p. 150: South Africa, ex 
Levaillant. 

d> (M. C. Z. 168516) Karita River, Suk, U. 9 November 1933. 
c? 9 (M. C. Z. 168517-8) Mt. Debasien, U. 17 November 1933. 

Native names. Chuli (Karamojo), Gisirire (Lugishu). 

Variation. The pair from Mt. Debasien were killed by a single shot 
as they sat together in a tree. The 9 is immature with horny-black 
bill quite unlike the yellow bills of the adults. In coloration also she 
is much more rufous; particularly about the head as described by 



144 bulletin: museum of comparative zoology 

Granvik (1934, p. 7) for Mt. Elgon birds. The wing measurements 
are as follows : — cf cf 425 mm., 430 mm. 9 409 mm. 

Diet, etc. When descending the mountain on the morning of Novem- 
ber 7, I found many scores of kites in the trees in the vicinity of the 
abandoned Karamojong village. So numerous were they that every 
large tree in the vicinity held birds, as many as five and six to a tree. 
They appeared to be resting in enjoyment of the early sunshine. As 
they were not there on the previous evening, it seems possible that 
they had been attracted to the district by extensive grass fires which 
had been raging on the plain 1,200 feet below. The only food found in 
their stomachs, however, was a large green scarab in that of the male. 

Melierax musicus poliopterus Cabanis 

Melierax poliopterus Cabanis, 1869, in von der Decken, Reise, 3, Abth. 1, 
p. 40: Umba River, East Africa (Kenya-Tanganyika boundary). 

9 (M. C. Z. 168519) Kitau, Manda Island, K. C. 19 May 1934. 

Diet, etc. Shortly after daybreak I was attracted by the warning 
cries of baboons (Papio ibeanus) such as they make when they have 
discovered a leopard. I followed them up into very dense bush but 
they kept retiring so that I did not even get a glimpse of them. At 
the spot where they had been, however, a persistent mewing drew 
my attention to this Chanting Hawk which was perched on the sum- 
mit of a wide-spreading acacia. The bird continued calling, undeterred 
by my arrival or the preceding commotion. In the tree as well as in 
an adjacent one were large hawk nests, but both were empty. On 
examining the stomach of the hawk, I found it crammed, even at this 
early hour, with locust-like grasshoppers of a species of which I had 
collected a series at Lamu. 

Melierax metabates metabates Heuglin 

Melierax metabates Heuglin, 1861, Ibis, p. 72: White Nile between 6° and 
7° N. lat. 

d" (M. C. Z. 168520) Mt. Debasien, U. 29 November 1933 

Breeding. Granvik (1934, p. 12) found this species breeding on 
July 14 at Lotonok; he also collected the White Nile Chanting Hawk 
at Kacheliba and Kolosia in the vicinity of Mt. Debasien. 

Diet. A striped rat (Lemniscomys striatus massaicus), remains of 
another rodent, also a striped skink {Mabuya striata) were recovered 
from the stomach of this bird. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 145 



Buteo rufofuscus augur (Riippell) 

Falco (Buteo) Augur Riippell, 1836, Neue Wirbelth., Vog., p. 38, pi. xvi: 
Ethiopia. 

<? 9 9 (M. C. Z. 168521-3) Kaburomi, Mt. Elgon, U. 28-30 

December 1933. 
J> juv. (M. C. Z. 168524) Kirui, s. Mt. Elgon, K. C. 24 January 1934 

Native names. Nabuye for white phase, yet/male for melanistic birds 
(Lugishu). 

Variation. The adult 9 from Kaburomi has the underparts en- 
tirely deep black and the paler portions of the back and interscapular 
area also black. See Granvik (1934, p. 10) for further occurrence of 
melanism in Elgon birds. 

Breeding. On January 23, I watched a young Augur Buzzard float 
down from the crags above Kemp's Cave to the valley below — 
evidently its first attempted flight. My local gunbearer urged me to 
shoot it but I refused. Next day it was brought up the mountain 
by a native who had killed it. 

Diet . The stomach contents of the three Kaburomi birds were : — 
(1) Swamp rat (Otoinys tropicalis elgonis) and Variable Skink (Mabuya 
v. varia). (2) Five skinks (M. v. varia) and four chameleons (C. bitaen- 
iatus altacclgonis). (3) A chameleon (C. b. altacdgonis) . 

Habitat. Augur Buzzards were quite the commonest large birds in 
the alpine meadow zone. They frequent the cliffs just below Jackson's 
Summit (13,650 feet). Nearby four buzzards were seen dashing at 
each other in the air, eventually two went off, evidently driven away 
by the others. 

Parasites. Lice (Laemobothrion titan) were abundant on the fledgling. 

Buteo vulpinus vulpinus (Gloger) 

Palco vulpinus "Licht." Gloger, 1833, Das Abandern der Vogel, p. 141 : 
Africa. 

9 (M. C. Z. 168525) Elgonyi, Mt. Elgon, K. C. 30 January 1934. 

Kaupifalco monogrammicus (Temminck) 
Falco monogramicus Temminck, 1824, PI. Col., livr. 53, pi. cccxiv: Senegal. 

& J 1 (M. C. Z. 168526-7) Bukori, Kitosh, K. C. 18 January 1934. 
c? 9 (M. C. Z. 168528-9) Voi, Seyidie Prov., K. C. 7 & 11 April 1934. 



146 bulletin: museum of comparative zoology 

Native name. Sichit (Masai). 

Variation. On laying out the series of twenty-two skins of the 
Lizard-Buzzard in the collection of the Museum of Comparative 
Zoology, we fail to discover any constant characters to separate 
forms though the series covers practically all parts of the range. 

In size, males of the southern and eastern birds (including Uganda), 
range from 204-214 mm. in wing measurements. Males of the western 
(Cameroon and Portuguese Guinea) birds range from 202-216 mm. 
Females of the eastern birds are from 211-222 mm., of the western 
215-236 mm. 

The two skins from the arid thorn -bush region of Voi are notice- 
ably paler than the pair of birds from the almost equally dry Kitosh 
country which, however, is adjacent to the forested slopes of Mount 
Elgon. One of these Bukori birds is the darkest in the whole series 
but the other is no darker than specimens from elsewhere. 

We find ourselves, therefore, in entire agreement with the conclu- 
sions reached by Claude Grant and Mackworth-Praed (1934, p. 130). 

Diet. Mouse fur in stomach of one Bukori bird. A young chameleon 
(C. d. roperi) and remains of many grasshoppers in the cT from Voi; 
a spiny mouse (Acomys ignitus) in the 9 from this locality. 

Enemies. Both Bagishu and Jaluo eagerly eat these birds but they 
are scorned by the Karamojong. 

Butastur rufipennis (Sundevall) 

Poliornis rufipennis Sundevall, 1851 (1850), Ofv. K. Vet.-Akad. Forh., 7, 
p. 131; Near Khartoum. 

9 (M. C. Z. 168530) W. foot of Mt. Debasien, U. 2 December 1933. 

Besides an old skin from the Lafresnaye collection and a pair from 
Morogoro, Tanganyika Territory, collected by Loveridge in 1918, no 
other examples of the rare Grasshopper-Buzzard have been received 
by the Museum. Van Someren (1922, p. 42) mentions a skin from 
Singo, Uganda as well as some of Loveridge's Morogoro series. 

Lophaetus occipitalis (Daudin) 

Falco occipitalis Daudin, 1800, Traite d'Orn., 2, p. 40: Anteniquoi Country. 
(i.e. Knysna Dist., Cape Province, apud Sclater). 

cf (M. C. Z. 168531) Elgonyi, Mt. Elgon, K. C. 5 February 1934. 
<? (M. C. Z. 168532) Kaimosi, Kakamega, K. C. 21 February 1934. 

Diet. A rat (Arvicanthis a. nubilans) was in the stomach of the 
Crested Eagle from Kaimosi. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 147 

Aquila wahlbergi Sundevall 

Aquila Wahlbergi Sundevall, 1851 (1850), Ofv. K. Vet.-Akad. Forh., 7, 
p. 109: "Caffraria superiori, prope 25° lat." (Type from Mohapoani Berg, 
Bechuanaland.) 

a* (M. C. Z. 168533) Voi, K. C. 10 April 1934. 

Diet. In addition to the feet and feathers of a big bird, the stomach 
of this Wahlberg's Eagle held half-a-dozen large grasshoppers. 

Haliaeetus vocifer clamans C. L. Brehm 

Haliaetos clamans C. L. Brehm, 1853, Journ. f. Orn., 1, p. 199, note: No 
definite locality. (Shoa, Ethiopia designated by Friedmann). 

tf 1 (M. C. Z. 168534) Mt. Debasien, U. 16 November 1933. 

a* (M. C. Z. 168535) Tana River nr. Ngau, K. C. 21 June 1934; 

Native names. Egwele (Karamojong) ; Chalipoko (Kipokomo). 

Variation. For designation of the type locality and discussion on 
the Northern Fish Eagle see Friedmann (1930, p. 66), also Banner- 
mann (1930, p. 268) and van Someren (1932, p. 269). 

The wings of these two males, when pressed down on the ruler, 
measure 498 and 511 mm. respectively. They are, therefore, consider- 
ably smaller than our series of the typical form from near Mwanza, 
Lake Victoria and Ujiji, Lake Tanganyika. These give 540-563 mm. 
for three adult males, 565 mm. for an adult female. 

From the measurements that he gives, it would seem that the birds 
from Rhino Camp, west Nile, referred to the typical form by Fried- 
mann, should rather be relegated to clamans. This would be in con- 
formity with Bannermann's remarks on the ranges of the forms. 
It would also appear that the birds from Kenya and Uganda listed by 
Friedmann as v. vocifer are in reality somewhat intermediate between 
the typical form and clamans. 

Diet. The Debasien bird was in a tree in dry bush at an altitude 
of 4,500 feet when I shot it at sunset. A most unexpected place for 
this riverine and lake-haunting bird. Still more astonishing was the 
statement of my Mgishu skinner who avowed that its stomach held 
termites ! 

Parasites. These birds were shot in the hope of obtaining the 
peculiar hippoboscid fly which has been described from it. None were 
found on either though lice were plentiful enough on the Ngau bird 
after its recoverv from the river into which it fell when shot. 



148 bulletin: museum of comparative zoology 

Folklore. Wapokomo women wear a claw of the Northern Fish 
Eagle dependent from their waist to insure fertility or an easy delivery. 

Necrosyrtes monachus pileatus (Burchell) 

Vultur pileatus Burchell, 1824, Travels, 2, p. 195, note: Hopetown District, 
Cape Province. 

d" cf imm, 1 unsexed (M. C. Z. 168536-8) Mt. Debasien,U.2 Decem- 
ber 1933. 

Habitat. The Southern Hooded Vulture was common on the plains 
at the western foot of the mountain as well as in the vicinity of our 
camp (5,000 feet). According to Granvik (1923, p. 65; 1934, p. 5) the 
species is common on Mt. Elgon at 7,500 feet. 

*Circus macrourus (S. G. Gmelin) 

Accipiter macrourus S. G. Gmelin, 1771 (1770), Nov. Comm. Acad. Petrop., 
15, p. 439, pis. viii & ix: Voronezh, southern Russia. 

9 (M. C. Z. 168539) Madangi, Mt. Elgon, U. 4 January 1934. 
9 (M. C. Z. 168540) Bukori, Kitosh, K. C. 18 January 1934. 
9 9 (M. C. Z. 168541-2) Kaimosi, K. C. 12 February and 3 March 1934. 

Diet. Rodent fur in the stomachs of the Bukori and a Kaimosi 
bird, a swamp rat (Otomys a. elassodon) in the juvenile Kaimosi 
harrier. 

*Circaetus gallicus gallicus (Gmelin) 

Falco gallicus Gmelin, 1788, Syst. Nat., 1, p. 259: France 

9 (M. C. Z. 168543) w. foot Mt. Debasien. U. 2 December 1933. 

This is very definitely the European bird and not C. g. pectoralis. 
Diet. Its stomach held a green snake (Chlorophis neglectus). 
Parasites. Among its feathers were hippoboscid flies (Lynchia sp.). 

Circaetus fasciolatus Gurney 

Circaetus fasciolatus Gurney, 1861, Ibis, p. 130: (Natal, ex Gray, 1848, List 
Bds. Brit. Mus., p. 18, where a nomen nudum). 

<? (M. C. Z. 168544) Ngatana, Tana River, K. C. 13 June 1934. 

Native name. Kifoanzongo (Kipokomo). 

Variation. Our material is too scanty to give a definite opinion, 
but it would appear as if cinerascens of South Africa is conspecific. 
The only other example of this rare Banded Harrier-Eagle in the 



PETERS AND LOVERIDGE: AFRICAN BIRDS 149 

Museum is a 9 collected by Loveridge near Mkindo River, central 
Tanganyika Territory, September 7, 1921. 

Diet. The stomach of the Ngatana bird held two snakes (Boaedon 
lineatus and Philothamnus s. semivariegatus) . 



FALCONIDAE 

PoLIHIERAX SEMITORQUATUS DECKENI Zedlitz 

Poliohierax semitorquatus deckeni Zedlitz, 1914, Journ. f. Orn., 62, p. 675: 
Southern Somaliland. 

<? (M. C. Z. 168545) Tsavo, K. C. 5 April 1934. 

Affinities. This Pigmy Falcon, erroneously sexed 9 by the native 
skinner, is obviously a male by its gray back. Our material is in- 
sufficient to assist in arriving at a decision as to whether two or four 
races can be recognized. The wing measurement of this bird is 1 15 mm. 
so that both on size and locality it should be referable to deckeni as 
defined by Bowen (1931, pp. 257-262). 

*Falco tinnunculus tinnunculus Linne 

Falco Tinnunculus Linne, 1758, Syst. Nat., ed. 10, p. 90: Europe (restricted 
type locality, Sweden, apud Hartert). 

1 (M. C. Z. 168546) Madangi, U. 2 January 1934. 
9 (M. C. Z. 168547) Kaimosi, K. C. 12 February 1934. 

Native name. Nab uho (Lugishu). 

Diet. A Variable Skink (Mabuya v. varia) in stomach of Madangi 
bird. It was a common sight to see these European Kestrels hovering 
over the alpine meadows of Mount Elgon at 12,000 feet. Doubtless 
in this region the numerous skinks furnish them with an abundance 
of food. The 9 held a Pigmy Mouse (Leggada b. bella) and locusts in 
its stomach. 

Falco tinnunculus carlo (Hartert & Neumann) 

Cerchneis tinnunculus carlo Hartert & Neumann, 1907, Journ. f. Orn., 55, 
p. 592: Bissidimo, near Harrar, Ethiopia. 

c? (M. C. Z. 168548) w. slopes Mt. Debasien, U. 18 November 1933. 

Corrigenda. This Mountain Kestrel has been compared with the 
Ujiji series (Bangs & Loveridge, 1933, p. 149) since studied by Claude 
Grant and Mackworth-Praed (1934, p. 79) who consider that the pale 



150 bulletin: museum of comparative zoology 

male (M. C. Z. 148228) shot in company with the Mountain Kestrels, 
is really an immature example of the Egyptian Kestrel (F. t. rupi- 
colaeformis) on migration. 

Diet. Shot while chasing a small bird around a stockade fence; its 
stomach held locusts and grasshoppers. 



PHASIANIDAE 

Francolinus sephaena grantii Hartlaub 

Francolinus grantii Hartlaub, 1866 (1865), Proc. Zool. Soc. London, p. 665, 
pi. xxxix, fig. 1 : Unyamwezi, Tanganyika Territory. 

c? (M. C. Z. 168549) Karita River, Suk, U. 9 November 1933. 
a" <? 9 9 (M.C.Z. 168550-3) s. bank Greeki River, U. 4 December 1933. 
d> c? (M. C. Z. 168554-5) Voi, K. C. 12 April 1934. 
<? tf (M. C. Z. 168556-7) Kitau, Manda Id., K. C. 16 & 18 May 1934. 
d> (M. C. Z. 168558) Golbanti, Tana R., K. C. 25 June 1934. 

Variation. In the series from Greeki River alone, we note a vari- 
ation of 10 mm. in the wing length of the males. In so far as they 
relate to grantii, our material bears out the conclusions of Claude 
Grant and Mackworth-Praed (1934, pp. 170-1). 

Francolinus clappertoni gedgii Ogilvie-Grant 

Francolinus clappertoni gedgii Ogilvie-Grant, 1891, Ibis, p. 124: Plains near 
Mount Elgon. 

9 (M. C. Z. 168559) w. foot Mt. Debasien, U. 20 November 1933. 
9 9 9 (M. C. Z. 168560-2) n. bank Greeki River, U. 5-6 December 1933. 
9 (M. C. Z. 168563) s. bank Greeki River, U. 7 December 1933. 

The Greeki River birds are topotypes of this race, and their habits 
as observed on the plains through which the river flows agree well 
with those noted by van Someren (1926, p. 51). 

Francolinus hildebrandti hildebrandti Cabanis 

Francolinus (Scleroptera) Hildebrandti Cabanis, 1878, Journ. f. Orn., 26, pp. 
206, 243, pi. iv, fig. 2: Voi, Taita district, Kenya Colony. 

9 (M. C. Z. 168564) Voi, K C. 12 April 1934. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 151 



Francolinus squamatus zappeyi Mearns 

Francolinus schuetti zappeyi Mearns, 1911, Smiths. Misc. Coll., 56, no. 20, 
p. 4: East shore of Lake Victoria, Kenya Colony. 

9 (M. C. Z. 168565) Kaimosi, K. C. 17 February 1934. 

This bird is almost topotypic, Kaimosi being fifty miles northeast 
of the lake. 

Pternistis afer leucoparaeus (Fischer & Reichenow) 

Francolinus (Pternistes) leucoparaeus Fischer & Reichenow, 1884, Journ. f. 
Orn., 32, p. 263: Kipini, at the mouth of the Tana River, Kenya Colony. 

c? (M. C. Z. 168566) Golbanti, K. C. 23 June 1934. 

Golbanti is about twenty-five miles in a straight line from the 
t3 r pe locality, Kipini, which is actually at the mouth of the Ozi, 
branch of the Tana River. 

Variation. Van Someren (1925, p. 95) is definitely WTong instating 
that the cheeks of the adult are black. Not only in the original de- 
scription are they stated to be white, but in our almost topotypic 
bird thev are feathered white. 



XUMIDIDAE 

Numida meleagris major Hartlaub 

Numida ptilorhyncha var. major Hartlaub, 1884 (1883), Abh. naturwiss. 
Ver. Bremen, 8, p. 217: Wakkala, in the Bari country near Gondokoro. 

<? 9 (M. C. Z. 168567-8) Mt. Debasien, U. 21-22 November 1933. 
d" d" 9 9 (M. C. Z. 168569-72) n. bank Greeki River, U. 5-6 Decem- 
ber 1933. 
d> J> <? (M. C. Z. 168573-5) s. bank Greeki River, U. 7 December 1933. 
9 (M. C. Z. 168576) Elgonyi, K. C. 1 February 1934. 

Distribution. The Debasien birds were shot at the very edge of the 
rain forest at 7,000 feet, the Elgonyi bird on the deforested hillside at 
7,000 feet. The north bank of the Greeki River is in Karamojo, the 
south bank in the Sabei country. The Uganda Tufted Guineafowl 
were in the habit of roosting in the big trees and scattered clumps of 
acacia which fringed the river banks. 



152 bulletin: museum of comparative zoology 

Guttera edouardi sethsmithi Neumann 

Guttera cristata sethsmithi Neumann, 1908, Bull. Brit. Orn. Club, 23, p. 13: 
Budongo Forest, Unyoro, Uganda. 

<? 9 (M. C. Z. 168577-8) Elgonyi, K. C. 5 February 1934. 
d" 9 (M. C. Z. 168579-80) Kaimosi, K. C. 26 February 1934. 

RALLIDAE 

Limnocorax flavirostra (Swainson) 

Gallinula flavirostra Swainson, 1837, Birds of West Africa, 2, p. 244, pi. 
xxviii: Senegal. 

Nest and 3 eggs near Witu, K. C. 2 June 1934, 

Breeding. It should be stated that a Black Crake was not seen to 
leave this nest but birds of this species were seen within a quarter of 
a mile of the rain-formed pool in which the nest was situated. The 
eggs are a shade larger than the largest in Pitman's fine series. 

The nest, roughly measuring 180 x 150 mm. across and 60 mm. in 
outside depth, was entirely composed of dry sedges and placed in 
the centre of a clump of fresh green sedges growing in knee-deep 
water. It held three fresh eggs, measuring 36 x 23 mm., of a buffy 
white ground color slightly tinged with pink; the whole more or less 
uniformly necked with small dots or dashes of pale brown and pale 
violet. 

Sarothrura rufa elizabethae van Someren 

Sarothrura rufa elizabethae van Someren, 1919, Bull. Brit. Orn. Club, 40, 
p. 20: Kakamegoes, Kavirondo a 71 , and Kisumu 9, Kenya Colony. 

7 <?, 3 9 (M. C. Z. 168581-90) Kaimosi, K. C. 10-23 February 1934. 

Distribution. These skins are practically topotypes as Kaimosi 
is an outlyer of the Kakamega forest. The series compares very well 
with the male from Mkarazi (not Mkaraji as spelt by Friedmann, 
1928, p. 76) Uluguru Mountains, Tanganyika Territory in the col- 
lection. 

Sarothrura pulchra centralis Neumann 

Sarothrura pulchra centralis Neumann, 1908, Bull. Brit. Orn, Club, 21, p. 
45: Mswa, west shore of Lake Albert. 

3 d* (M. C. Z. 168591-3) Kaimosi, K. C. 9-13 February 1934. 

It is interesting to find three species of Sarothrura occurring in 
the same locality. All were carefully snared by the same native. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 153 

Sarothrura elegans reichenovi (Sharpe) 

Corethrura reichenovi Sharpe, 1894, Cat. Bds. Brit. Mus., 23, p. 121: Cameroon 
(= Buea, Cameroon Mt., ex Reichenow, 1892, Journ. f. Orn., p. 178). 

d" (M. C. Z. 168594) Kaimosi, K. C. 26 February 1934. 

Affinities. This bird constitutes the first record for Kenya Colony. 
In default of adequate comparative material we follow van Someren 
(1922, p. 23) in referring it to reichenowi though it would seem possible 
that Uganda and Kenya birds may belong to an undescribed race. 



CHARADRIIDAE 

Hoplopterus spinosus (Linne) 
Charadrius spinosus Linne, 1758, Syst. Nat., ed. 10, 1, p. 151: Egypt. 

c? 9 (M. C. Z. 168595-6) Wema, Tana River, K. C. 13-18 June 1934. 
Native name. Tirangole (Kipokomo). 

Charadrius pecuarius pecuarius Temminck 

Charadrius pecuarius Temminck, 1823, PI. col., livr. 31, pi. clxxxiii: Cape of 
Good Hope. 

c? (M. C. Z. 168597) Kitau, Manda Id., K. C. 17 May 1934. 



BURHINIDAE 
*Burhinus oedicnemus oedicnemus (Linne) 
Charadrius Oedicnemus Linne, 1758, Syst. Nat., ed. 10, 1, p. 151: England. 

c? (M. C. Z. 168598) N. bank of Greeki River, U. 5 December 1934. 

Burhinus vermiculatus vermiculatus (Cabanis) 

Oedicnemus vermiculatus Cabanis, 1868, Journ. f. Orn., 16, p. 413: No locality 
given (= Lake Jipe, near Taita, Kenya Colony). 

d> 9 9 (M. C. Z. 168599-601) Wema, Tana River, K. C. 19 June 1934. 

Both females are young birds and were killed on a sandbar exposed 
by a fall in the river. 



154 bulletin: museum of comparative zoology 

PTEROCLIDIDAE 

Pterocles decoratus decoratus Cabanis 

Pterocles decoratus Cabanis, 1868, Journ, f. Orn., 16, p. 413: No locality given 
( = Lake Jipe, near Taita, Kenya Colony). 

& (M. C. Z. 168602) Tsavo, K. C. 31 March 1934. 

Tsavo is only seventy miles northeast of the type locality. 



COLUMBIDAE 

Treron waalia (Meyer) 

Columba waalia F. A. A. Meyer, 1793, Syst.-Sum. Uebers. Zool. Entdeck., 
p. 128: Tcherkin, near Lake Tsana, Ethiopia. 

J* <? 9 (M. C. Z. 168603-5) Mt. Debasien, U. 20 &25 November 1933. 

Variation. There is little doubt but that with adequate series from 
different parts of the range of the species, at least one geographical 
race will have to be separated. We refrain from doing so for two 
reasons, the first is insufficiency of our material, particularly from 
West Africa; the second, lack of topotypical material of cinereiceps 
Neumann (Journ. f. Orn., 52, 1904, p. 341 : middle Gelo River, Jambo- 
land). The wings of four males from the Ethiopian highlands (Lake 
Tana and Hawash districts) measure 174-186 mm. 

Our two males from Mount Debasien and three more from Kenya 
and Uganda in the British Museum, measured for us by Mr. C. H. B. 
Grant, give 159-178 mm., a male from the Dindar River, eastern 
Sudan, has a wing of 162 mm., but this measurement is not full since 
the bird is in molt and the second longest primary is not yet grown 
out, the measurement is therefore to the tip of the outer primary 
which is worn. A male and an unsexed bird from British Somaliland 
have wings measuring 171 and 174 mm. respectively, and an unsexed 
bird from West Africa, 172 mm. 

A female from the Hawash district has a wing 174 mm., our Debasien 
bird 161, and two from the Upper Blue Nile 159-163 mm., but as in 
the Dindar River male the state of their primaries prevents measure- 
ments of full value. 

While it appears likely that there is a smaller race in Kenya, Uganda 
and perhaps the Sudan and western Ethiopia, we are by no means 
certain that Neumann's name really applies to it, the status of the 



PETERS AND LOVERIDGE: AFRICAN BIRDS 155 

Somaliland and West African birds cannot be settled, but we hope 
that the information given here may enable some future investigator 
to clear up the question. 

We do not believe that the African green fruit pigeons (Vinago) 
merit generic separation from the In do-Malayan species (Treron). 

Treron calva salvadorii (Dubois) 

Vinago salvadorii Dubois, 1897, Proc. Zool. Soc. London, p. 784: Eastern and 
central tropical Africa (restricted to the western shores of Lake Tanganyika 
by Hartert and Goodson). 

c? c? 9 (M. C. Z. 168606-8) Mt. Debasien, U. 16-30 November 1933. 
<? (M. C. Z. 168609) Nyenye, U. 8 December 1933. 
9 (M. C. Z. 168610) Kaimosi, K. C. 26 February 1934. 

Treron calva wakefieldii Sharpe 

Treron wakefieldii Sharpe, 1874 (1873), Proc. Zool. Soc. London, p. 715, pi. 
Iviii, fig. 2: Mombasa, Kenya Colony. 

& 9 (M. C. Z. 168611-2) Wema, Tana River, K. C. 20 June 1934. 

Affinities. We follow Grote (1931, p. 140) in considering wake- 
fieldii only subspecifically distinct from the gray-tailed calva group. 
We are not, however, quite prepared to accept his view that all the 
mainland forms should be regarded as subspecies of the Malagasy 
australis. 

CoLUMBA arquatrix arquatrix Temminck 

Colombo (sic) arquatrix Temminck, 1809, in Knip, Les Pigeons, les Colombes, 
p. 11, pi. v: South Africa-Anteniquoi country (i.e. Knysna Dist., Cape 
Province, apud Sclater). 

c? & (M. C. Z. 168613-4) Mt. Debasien, U. 22 & 27 November 1933. 

Shot in the rain forest at 7,000 feet. Recorded from Mount Elgon 
by Granvik (1934, p. 19). 

Streptopelia semitorquata semitorquata (Ruppell) 

Columba semitorquata Ruppell, 1837, Neue Wirbelth., Vog., p. 66, pi. xxiii, 
fig. 2: Tarentola Mountains, Ethiopia. 

9 (M. C. Z. 168615) Butandiga, U. 31 January 1934. 
c? (M. C. Z. 168616) Elgonyi, K. C. 30 January 1934. 

Variation. We agree with Friedmann (1930, p. 216) that the 
characters claimed for clgonensis Granvik, based on a female from 



156 bulletin: museum of comparative zoology 

the eastern slopes of the mountain at 7,000 feet, are individual, and 
that our birds from the west (6,000 feet) and south (7,000 feet) do 
not substantiate them. 



Streptopelia decipiens perspicillata (Fischer & Reichenow) 

Turtur perspicillata Fischer & Reichenow, 1884, Journ. f. Orn., 32, p. 179: 
Nguruman, Masailand, Tanganyika Territory. 

& (M. C. Z. 168617) Voi, K. C. 10 April 1934. 

Variation. This bird has a wing length of 142 mm. which places 
it as a capicola. On the other hand its cheeks are distinctly pale gray 
which is one of the principal characters relied upon for distinguishing 
the decipiens group from capicola. In its coloration it closely resembles 
a bird from Kabare, Bukoba which, according to Friedmann (1930, 
p. 221) is within the range of decipiens permista. It in no way re- 
sembles specimens attributed to perspicillata from Shinyanga, Usuku- 
ma and Kisumu, Kavirondo. 



Streptopelia capicola tropica (Reichenow) 

Turtur capicola tropica Reichenow, 1902, Orn. Monatsb., 10, p. 139. East 
Africa (Type from Songea, Tanganyika Territory). 

<? & (M. C. Z. 168618-9) Mt. Debasien, U. 22 & 29 November 1933. 

Previously recorded from Kacheliba and Mount Elgon by Granvik 
(1934, p. 21). 

Aplopelia larvata larvata (Temminck) 

Columba larvata Temminck, 1810, in Knip, Les Pigeons, les Colombes, p. 
71, pi. xxxi; Anteniquoi (i.e. Knysna Dist., Cape Province, apud Sclater). 

9 (M. C. Z. 168620) Mt. Mbololo, K. C. 19 April 1934. 



Turtur chalcospilos chalcospilos (Wagler) 

Columba Chalcospilos Wagler, 1827, Syst. Av., Columba, sp. 83: South Africa 
( = Eastern Cape Province) . 

d> (M. C. Z. 168621) Elgonyi, K. C. 24 January 1934. 

<? (M. C. Z. 168622) Kibwezi, K. C. 24 March 1934. 

9 (M. C. Z. 168623) Kitau, Manda Id., K. C. 17 May 1934. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 157 

Tympanistria tympanistria fraseri Bonaparte 

Tympanistria fraseri Bonaparte, 1855 (after April 15), Consp. Av., 2, p. 67: 
Fernando Po. 

<? (M. C. Z. 168624) Sipi, U. 12 December 1933. 
PSITTACIDAE 

PoiCEPHALUS RUFIVENTRIS RUFIVENTRIS (Riippell) 

Pionus rufiventris Riippell, 1845, Syst. Uebers., p. 83, pi. xxxii: Shoa, Ethiopia. 

J> 9 (M. C. Z. 168625-6) Tsavo, K. C. 4 April 1934. 

Affinities. In the absence of any typical Ethiopian material, we 
follow Friedmann (1930, p. 291) in referring these birds to the typical 
form and regarding simplex Reichenow of Tanganyika as a synonym. 

POICEPHALUS MEYERI SATURATUS (Sharpe) 

Poeocephalus saturatus Sharpe, 1901, Bull. Brit. Orn. Club, 11, p. 67: Northern 
Ankole, Uganda. 

9 (M. C. Z. 168627) Bukori, K. C. 18 January 1934. 

Affinities. This race has been discussed at length by Granvik 
(1923, p. 74 and 1934, p. 29) who obtained good series on the east 
and northeast of Mt. Elgon. 

MUSOPHAGIDAE 

Turacus hartlaubi (Fischer & Reichenow) 

Corythaix hartlaubi Fischer & Reichenow, 1884, Journ, f. Orn., 32 p. 52: Mt. 

Meru, Tanganyika Territory. 
Turacus hartlaubi crissalis Mearns, 1915, Smiths. Misc. Coll., 65, No. 13, p. 3: 

Mt. Mbololo, Taita, Kenya Colony. 

o* (M. C. Z. 168628) Sipi, IT. December 1933. 
9 (M. C. Z. 168629) Butandiga, U. 10 January 1934. 
9 9 (M. C. Z. 168630-1) Mt. Mbololo, K. C. 16 & 20 April 1934. 

Affinities. We have reexamined the same series on which Friedmann 
commented (1930, p. 251) and have nothing further to add. Both 
topotypes of crissalis have black thighs as defined by Mearns but do 
not differ in this respect from birds in the long Usambara series. 



158 bulletin: museum of comparative zoology 

Turacus leucolophus (Hartlaub) 

Corythaix leucolophus Hartlaub, 1857, Syst. Orn. W. Africa, p. 159: Belina 
Mts., between lat. 4° and 5° N. in the Bari-Nile region. 

3 d\ 1 9 (M. C. Z. 168632-5) Mt. Debasien, U. 15-30 November 1933. 

Citatation. The original reference to the Journ. f. Orn., 1855 with 
citation of Heuglin as author of this species, furnished by Sclater 
(1924, p. 193) is entirely erroneous. Turacus leucolophus Heuglin 
1856 is a nomen nudum. The name was first correctly introduced by 
Hartlaub as given above. 

In this connection we wish to express our indebtedness to Dr. H. 
Friedmann for looking up certain references which were unavailable 
to us here. 

Habits. Granvik (1934, p. 27) has stated that this species was 
difficult to obtain on Mount Elgon on account of its shyness. It 
was found that Debasien birds exhibited the same wary alertness 
making it extremely difficult to approach within gunshot. 

MUSOPHAGA ROSSAE ROSSAE Gould 

Musophaga rossae Gould, 1851, Proc. Zool. Soc. London, p. 93: said to have 
come from the western coast of Africa, (i.e. Loanda, fide Grant, 1915, 
Ibis, p. 413 ). 

9 (M. C. Z. 168636) Mt. Debasien, U. 14 November 1933. 
c? 9 9 (M. C. Z. 168637-9) Elgonyi, U. 24-29 January 1934. 

Affinities. We are by no means convinced that this bird should 
not be regarded as a race of violacea as was done by Grote (1922, 
p. 398). In view of the fact, however, that a western race of rossae, 
named savanicola by Grote, may occur alongside violacea in Northern 
Cameroon, we treat them for the present as distinct. 

Parasites. Lice (Laemobothrium tinnunculi) were preserved from the 
feathers. 

Corythaeola crist ata yalensis (Mearns) 

Musophaga cristata yalensis Mearns, 1915, Smiths. Misc. Coll., 65, No. 13, 
p. 5: Yala River, Kavirondo, Kenya Colony. 

4 d" , 4 9 (M. C. Z. 168640-7) Yala River, K. C. 26-28 February 1934. 

Variation. These topotypes, when compared with our series of 
West African birds, do show the rather fine differences which have 
been pointed out by van Someren (1922, p. 48). 



PETERS AND LOVERIDGE: AFRICAN BIRDS 159 

CORYTHAIXOIDES LEUCOGASTER (Ruppell) 

Chizaehris leucogaster Ruppell, 1842, Mus. Senckenb., 3, p. 127: Southern 
Ethiopia. 

• (? (M. C. Z. 168648) Voi, K. C. 7 April 1934 

Affinities. Roberts (1926, p. 218) has proposed that this species 
be made the type of a monotypic genus for which he proposes the 
name Criniferoides. While this genus is probably as well differen- 
tiated as some other recent generic separations in the family, such as 
Ruwenzorornis and Proturacus, we scarcely think that further divi- 
sion is desirable. 

CUCULIDAE 

Lampromorpha klaasi (Stephens) 

Cuculus klaasi Stephens, 1815, in Shaw, Gen. Zool., 9, p. 128: Platte River 
(ex Levaillant). 

& juv. (M. C. Z. 168649) Mt. Debasien, U. 30 November 1933 

Centropus monachus fischeri Reichenow 
Centropus ftscheri Reichenow, 1887, Journ. f. Orn., 35 p. 57: Niakatschi. 

d> d" (M. C. Z. 168650-1) Kaimosi, K. C. 14 & 20 February 1934. 

Affinities. In referring these birds to fischeri, of which we have 
only a juvenile from the Mabira Forest, Uganda, we are in agree- 
ment with the findings of van Someren (1932, p. 274). 

Habitat. Not uncommon in the undergrowth and bushes fringing 
the forest and millpond. One of these birds was found drowned, 
floating at night upon the millpond. Several times individuals were 
disturbed at one side of the pond and with clumsy flight made off 
to the sedges on the further shore. 

Centropus superciliosus superciliosus Hemprich & Ehrenberg 

Centropus superciliosus Hemprich & Ehrenberg, 1828, Symb. Phys., folio R., 
pi. xi: Southern Arabia. 

c? (M. C. Z. 168652) Lamu, Lamu Island, K. C. 11 May 1934. 

Affinities. Friedmann (1930, p. 280) has given a full discussion on 
the status of the various races of this bird. In the absence of typical 
material from Arabia, we follow Friedmann in not recognizing 
C. s. furvus Friedmann (n. n. for C. s. intermedins van Someren, 



160 bulletin: museum of comparative zoology 

preoccupied. Type locality, Mombasa) though it would seem dis- 
tinctly advisable that the question of its status be reexamined when 
an adequate series of Arabian skins is available. Van Someren (1932, 
p. 275) refuses to accept Friedmann's view and maintains that furvus 
should be regarded as distinct. 



TYTONIDAE 
Tyto capensis libratus Peters & Loveridge 

Tyto capensis libratus Peters & Loveridge, 1935, Proc. Biol. Soc. Washington, 
48, p. 77: Kaimosi, Kakamega district, Nyansa Province, Kenya Colony. 

Holotype 9 (M. C. Z. 168653) Kaimosi, K. C. 21 February 1934. 



STRIGIDAE 
Strix woodfordii suahelica (Reichenow) 

Syrnium suahelicum Reichenow, 1898, in Werther, Die Mittl. Hochl. des 
nordl. Deutsch-Ost-Afr., p. 272: Tanganyika Territory. 

& 9 (M. C. Z. 168654-5) Mt. Debasien, U. 18 November 1933. 
c? (M. C. Z. 168656) Kaimosi, K. C. 7 March 1934. 

Native name. Ekeruk (Karamojong). 

Affinities. Van Someren (1922, p. 45) has recorded this species 
from Moroto, northeast of Mt. Debasien, and both he and Granvik 
(1934, p. 40) referred their Mt. Elgon birds to this race. Our material 
is undoubtedly of this form but it seems likely that suahelica Reich- 
enow, 1898 may have to become a synonym of nigricantia Sharpe, 
1897, described from Mpwapwa, Tanganyika Territory. 

Diet. Both Debasien birds had their stomachs distended with the 
remains of small beetles. 

Habitat. The Debasien birds were shot at about 9 a.m. in gallery 
forest fringing the banks of a dry river bed on the western slopes at 
an altitude of 7,000 feet. 

Bubo africanus africanus (Temminck) 
Strix africana Temminck, 1823, PI. col., livr. 9, pi. 1 : Cape of Good Hope. 
<? 9 (M.C.Z. 168657-8) Kaimosi, K. C. 6 March 1934. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 161 

Diet. The stomachs were full of the elytra of small beetles, among 
them those of a cockchafer were recognizable; in addition one bird 
held the fur of a small rodent. 

Enemies. These owls were shot about 8 a.m. from some mighty 
mimosas where they were being mobbed by a party of small birds. 

Bubo lacteus (Temminck) 

Strix lactea Temminck, 1824, PL col., livr. 1, pi. 4: Senegal. 

c? 9 (M. C. Z. 168659-60) Wema, Ngatana, K. C. 13 June 1934. 

Native name. Hwichi (Kipokomo). 

Diet. The stomach of one held the red legs of an Allen's Reed Hen 
(Porphyrula alleni) and its plumage; that of its mate contained 
feathers of what was apparently the same species. 



CAPRIMULGIDAE 

*Caprimulgus europaeus europaeus Linne 

Caprimulgus europaeus Linne, 1758, Syst. Nat., ed. 10, 1, p. 193: Europe and 
America (restricted type locality, Sweden, apud Hartert). 

C? (M. C. Z. 168661) Kaimosi, K. C. 16 February 1934. 

Measurements. Wing to end of second primary, as outer ones in 
moult, 185 mm. 

*Caprimulgus europaeus meridionalis Hartert 
Caprimulgus europaeus meridionalis Hartert, 1896, Ibis, p. 370: Greece. 
9 (M. C. Z. 168662) Kibwezi, K. C. 29 March 1934 
Measurements. Wing 174 mm. 

*Caprimulgus europaeus unwini Hume 

Caprimulgus unwini Hume, 1871, Ibis, p. 406: Agrore valley, Hazara district, 
northwest India. 

c? (M. C. Z. 168663) Voi, K. C. 13 April 1934. 

Variation. We have compared this bird with an adult and an 
immature unwini from Turkestan. We find that the African bird 
possesses the large white wing spot characteristic of this race with 
which it also agrees in its paler coloration, paler than in either of the 
two European migrant races. 



162 bulletin: museum of comparative zoology 

Measurements. Wing 197 mm. 

Parasites. Hippoboscid flies (Pseudolynchia rufipes) were collected 
from its plumage where they were very abundant. 

Habitat. This bird was one of several disturbed in the bed of the 
Voi River where they were resting upon the sand. It flew about a 
hundred feet, then settled upon a small branch at a height of six feet 
from the ground. 

Caprimulgus nubicus taruensis van Someren 

Caprimulgus nubicus taruensis van Someren, 1919, Bull. Brit. Orn. Club, 
40, p. 25: Tsavo, Kenya Colony. 

9 Juvenile (M. C. Z. 168664) Kibwezi, K. C. 29 March 1934. 
9 and egg (M. C. Z. 168665) Tsavo, K. C. 2 April 1934. 

Breeding. I was particularly on the lookout for this bird when at 
Tsavo as I wished to procure a topotype, for the species was un- 
represented in the collection of the Museum of Comparative Zoology. 

On March 31, I disturbed a bird at 8 a.m. that was sitting on a 
single fresh egg at the base of a small scraggy thorn bush in this 
desert-like region. Returning at the same hour two days later, I 
found the bird incubating the solitary egg, I shot the bird and col- 
lected the egg which measures 26 x 19 mm. It has a white ground 
color, almost obscured around the upper pole by dull purple over- 
lying brown; the lower half is blotched with brown but scarcely any 
purple. The egg was lying in a circular patch of sand about 100 mm. 
in diameter; the surrounding area without the circle was littered 
with gravel, twigs, etc. 

Habitat. The Kibwezi bird was one of a pair sitting on the road 
a mile east of the station about 8 p.m. I shone a torch in its eyes and 
then shot it. The Southern European Nightjar was obtained on the 
same stretch of road in the same way and at the same time. 

Caprimulgus fossii clarus Reichenow 

Caprimulgus clarus Reichenow, 1892, Journ, f. Orn., 40, p. 29: Bukoba, Lake 
Victoria, Tanganyika Territory. 

cf 9 (M. C. Z. 168666-7) Lamu, Lamu Id., K. C. 8 & 9 May 1934. 

Breeding. A bird of this species which I disturbed on May 9, 
attracted my attention by the antics with which it attempted to 
draw me away from the vicinity of its young. Flying up to within 
ten feet of me, it would alight, then flop along the ground with one, 



PETERS AND LOVERIDGE: AFRICAN BIRDS 163 

sometimes with both, wings outspread as if injured. The two young 
were obviously only just out of the eggs, dainty little balls of fluff. 

Parasites. The male bird had numerous hippoboscid flies (Pseudo- 
lynchia rufipes) among its feathers. 



COLIIDAE 

Colius striatus mombassicus van Someren 

Coitus striatus mombassicus van Someren, 1919, Bull. Brit. Orn. Club, 40, 
p. 26: Changamwe, Kenya Colony. 

9 (M. C. Z. 168668) Lamu Island, K. C. 11 May 1934. 
9 (M. C. Z. 168669) Malindi, K. C. 29 June 1934. 

Variation. In his description of this race van Someren states that 
his series was comprised of birds from along the coast between Mom- 
basa and Lamu. We find these two females agree closely with our 
Mombasa birds being grayer than the more sandy brown affinis de- 
scribed from Dar es Salaam further south. 

Colius striatus jebelensis Mearns 

Colius striatus jebelensis Mearns, 1915, Proc. U. S. Nat. Mus., 48, p. 394: 
Gondokoro, Uganda. 

<? & (M. C. Z. 168670-1) Mt. Debasien, U. 20 & 25 November 1934. 
9 juv. (M. C. Z. 168672) Kaimosi, K. C. 1 March 1934. 

Variation. With this small series we prefer to follow Friedmann 
(1930, p. 323) in referring the Debasien coly to jebelensis of which he 
regards ugandensis van Someren (1919, p. 26: Chagwe, Uganda) as a 
synonym. The correct determination of the juvenile bird from 
Kaimosi is less certain, perhaps it should be identified with kikuyuensis 
van Someren. 

TROGONIDAE 

Apaloderma narina narina (Stephens) 

Trogon narina Stephens, 1815, in Shaw, Gen. Zool., 9, p. 14: Anteniquoi, 
{i.e. Knysna district, Cape Province, apud Sclater). 

9 9 (M.C.Z. 168673-4) Mt. Debasien, U. 21 & 29 November 1933. 
9 (M. C. Z. 168675) Sipi, Mt. Elgon, U. December 1933. 

Distribution. The Mount Debasien birds were shot at 5,000 and 
7,000 feet in the big trees fringing the Amaler River and one of its 



164 bulletin: museum of comparative zoology 

tributaries. One might have expected the Elgon bird to have been 
A. n. brachyurum Chapin, but the measurements of all three are those 
of the typical form, viz. wings 125, 126, 125 mm., tails 178, 171, and 
177 mm. in order as listed. 



ALCEDINIDAE 

Ceryle rudis rudis (Linne) 
Alcedo rudis Linne, 1758, Syst. Nat., ed. 10, 1, p. 116: Egypt. 

tf (M. C. Z. 168676) Wema, Tana River, K. C. 18 June 1934. 
Native name. Tiglili (Kipokomo). 

Halcyon senegaloides ranivora Meinertzhagen 

Halcyon senegaloides ranivora Meinertzhagen, 1924, Bull. Brit. Orn. Club, 
44, p. 44: Pangani River, Tanganyika Territory. 

9 (M. C. Z. 168677) Kitau, Manda Id.; K. C. 16 May 1934. 

Affinities. In the absence of topotypical material of senegaloides 
Smith for comparison, we refer this bird to the northern race of which 
we have a d" and 9 from Dar es Salaam and Mombasa respectively. 

Halcyon chelicuti variegata (Vieillot) 

Alcedo Variegata Vieillot, 1820, Tabl. Encyc. Meth., Orn., pt. 1, p. 397: 
Senegal. 

Breeding. On May 8, 1934, a pair of these kingfishers had a nestful 
of young in the stump of a coconut palm just outside Lamu, Lamu 
Island. The stump was only four feet in height and situated about a 
hundred feet from the entrance of my tent. Though there was a hole 
in the top (through which the rain entered until I stopped it with a 
coconut), the birds always entered and left by a woodpecker-like hole 
in the side at a height of three and a half feet from the ground. The 
birds were remarkably tame with so much activity going on around 
their home, if, however, anyone was passing the stump when one of 
the parents was returning with food, the bird generally sheered off 
and, perching on the branch of a nearby mango tree, patiently waited 
till the coast was clear. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 165 

CORACIIDAE 

*Coracias garrulus garrulus Linne 

Coracias garrulus Linne, 1758, Syst, Nat., ed. 10. 1, p. 107: Europe; restricted 
type locality, southern Sweden. 

cT 9 (M. C. Z. 168678-9) Kibwezi, K. C. 24 March 1934. 
9 (M. C. Z. 168680) Tsavo, K. C. 2 April 1934. 

Diet. The cf European Roller from Kibwezi had a large spider in 
its stomach ; the 9 had her stomach filled to capacity by seven very 
large grasshoppers. 

Coracias caudatus lorti Shelley 
Coracias lorli Shelley, 1885, Ibis, p. 399: plateau south of Berbera, Somaliland. 

<? (M. C. Z. 168681) Wema, Tana River, K. C. 18 June 1934. 
Native name. Nchorwe (Kipokomo). 

MEROPIDAE 

Melittophagus ptjsilltjs cyanostictus (Cabanis) 

Merops cyanostictus Cabanis, 1869, in von der Decken, Reise, 3, p. 34: 
Mombasa, Kenya Colony. 

d" (M. C. Z. 168682) Mt. Debasien, U. 15 November 1934. 
d* (M. C. Z. 168683) Voi, K. C. 13 April 1934. 

Melittophagus lafresnayii oreobates Sharpe 
Melittophagus oreobates Sharpe, 1892, Ibis, p. 320: Sav<§, Mount Elgon, Uganda. 
9 (M. C. Z. 168684) Elgonyi, Mt. Elgon, K. C. 22 January 1934. 

Type locality. The type locality according to Sharpe is at 6,000 
feet and in the description of Jackson's itinerary he states that it 
is on the north side of Mount Elgon. This places it on the Uganda 
side of the mountain and I have been able to locate it on an early 
map, "compiled from sketches." I feel reasonably certain that Save 
corresponds to Sabei, (pronounced Sabe) which agrees in position 
and altitude. Elgonyi is also about 6,000 feet but is on the southeast 
and Kenya portion of the mountain. 

Breeding. On January 22, 1934, two clutches, composed of three 
eggs each, were found at Elgonyi. The clutches were situated at 



166 bulletin: museum of comparative zoology 

about a foot and one and a half feet respectively from the entrance 
of the burrows made in a high bank of red earth, the burrows being 
at a height of six feet from the ground. The female was shot as she 
left the nest-hole. The eggs, measuring 23 x 19 mm., were the usual 
white, and perfectly fresh. 



PHOENICULIDAE 

Phoeniculus purpureus marwitzi (Reichenow) 

Irrisor erythrorhynchus marwitzi Reichenow, 1906, Orn. Monatsb., 16, p. 171: 
Mkalama, Kondoa Irangi district, Tanganyika Territory. 

c? (M. C. Z. 168685) Mt. Debasien, U. 25 November 1934. 

Variation. The coloring of the body plumage is slightly inter- 
mediate between marwitzi and niloticus (Neumann) as might be ex- 
pected. We refer it to marwitzi on account of the greenish underside 
of the wings. 

Phoeniculus bollei jacksoni (Sharpe) 

Irrisor jacksoni Sharpe, 1890, Ann. Mag. Nat. Hist., (6), 6, p. 503: Kikuyu 
country, Kenya Colony. 

d" 9 imm. 9 imm. (M. C. Z. 168686-8) Kaimosi, K. C. 26 February 
& 1 March 1934. 

Distribution. The Kikuyu White-headed Kakelaar was also seen 
in the forest at Elgonyi, Mount Elgon. 



BUCEROTIDAE 

Bycanistes bucinator (Temminck) 

Buceros bucinator Temminck, 1824, PI. col. livr. 48, pi. clxxxiv: Cape of 
Good Hope, South Africa. 

& d* (M. C. Z. 168689-90) Wema, Tana River, K. C. 15 & 16 June 1934. 

Bycanistes subcylindricus (Sclater) 

Buceros subcylindricus P. Sclater, 1870, Proc. Zool. Soc. London, p. 668, 
pi. xxxix; West Africa. 

& (M. C. Z. 168691) Sipi, U. 22 December 1933. 
c? tf 1 (M. C. Z. 168692-3) Elgonyi, K. C. 25 & 29 January 1934. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 167 

Native name. Lingasa (Lugishu). 

Breeding. Granvik (1923, p. 98) gives an interesting account of 
the breeding of the Black-and-white Casqued Hornbill on Mount 
Elgon. 

Enemies. The Bagishu relish this bird as an article of diet, requests 
for the carcasses of those skinned being numerous. 

LOPHOCEROS ERYTHRORHYNCHUS ERYTHRORHYNCHUS (Temminck) 
Buceros erythrorhynchus Temminck, 1823, PI. col. livr., 36, sp. 19: Senegal. 

d" d" (M. C. Z. 168694-5) Karita River, U. 9 November 1933. 
a" (M. C. Z. 168696) Tsavo River, K. C. 5 April 1934. 

LOPHOCEROS FLAVIROSTRIS FLAVIROSTRIS (Riippell) 

Buceros flavirostris Riippell, 1835, Neue Wirbelth., Vog., p. 6, pi. ii: Taranta 
Mountains, Ethiopia. 

J> 9 (M. C. Z. 168698-9) Tsavo River, K. C. 5 April 1934. 

Lophoceros jacksoni Grant 

Lophoceros jacksoni Ogilvie Grant, 1891, Ibis, p. 127: Turkwel, Suk, Kenya 
Colony. 

9 (M. C. Z. 168697) Tsavo River, K. C. 5 April 1934. 

Distribution. It will be observed that the above record is south- 
easterly for this species, but another example in the collection, whose 
identification has been confirmed, was obtained by Loveridge (1923, 
p. 910) in central Tanganyika. Van Someren collected the very 
similar deckeni at Tsavo and it will be noted that Loveridge shot two 
other species there on the same day. 

Lophoceros melanoleucos geloensis Neumann 
Plate 1, fig. 1 

Lophoceros melanoleucos geloensis Neumann, 1905, Journ. f. Orn., 53, p. 187: 
Schekho on the Upper Gelo River, southwest Ethiopia. 

c? (M. C. Z. 168700) Sipi, U. 22 December 1933. 
c? cf 9 (M. C. Z. 168701-3) Kaimosi, K. C. 25 February 1934. 

Native name. Gizombe (Lugishu). 

Variation. We feel fairly confident that the brown plumage cited 
as a subspecific character differentiating suahclieus and stegmanni is not 



168 bulletin: museum of comparative zoology 

of taxonomic importance being probably due to the bleaching effects 
resulting from dwelling in open country, while birds inhabiting the 
wetter rain-forest regions are darker. Professor Neumann (in lift.) 
now considers stegmanni synonymous with geloensis. 

Coloration. The three birds from Kaimosi were killed with a single 
shot, the gular skin of the males was black, that of the female greenish. 

Breeding. As the junior author was passing down a forest trail at 
Sipi on December 21, 1933, a Gelo River Crowned Hornbill put its 
head out of a knot hole high up on a tree trunk. It quickly withdrew 
it again but as he continued watching the bird emerged and flew 
away. Though the bark of the tree was perfectly smooth, the trunk 
was only a foot in diameter so was climbed by the native gunbearer. 
He found nothing but rotted wood at the bottom of the hole so it 
seems probable that the bird was only prospecting for a nesting site. 



INDICATORIDAE 

Indicator indicator (Sparrman) 

Cuculus indicator Sparrman, 1777, Phil. Trans., 67, p. 43, pi. i: Great Fish 
River, near Somerset East, Cape Province. 

d" c? & (M. C. Z. 168704-6) Mt. Debasien, U. 14, 18, 21 November 1933. 
<? (M. C. Z. 168707) Tsavo, K. C. 5 April 1934. 

Remarks. The Black-throated Honey Guide has been recorded 
from Mount Elgon and Tsavo by van Someren (1922, p. 53). Lover- 
idge thought it was more abundant on Mount Debasien than in any 
other part of East Africa in which he had collected. The species is 
beloved by the Karamojong who are passionately fond of honey. 



CAPITONIDAE 

Tricholaema lacrymosum lacrymosum Cabanis 

T richolaema lacrymosa Cabanis, 1878, Journ. f. Orn., 26, p. 205: Adi River, 
i.e. Athi River, Kenya Colony (vide p. 240). 

c? d 1 9 (M. C. Z. 168708-10) Voi, K. C. 7 April 1934. 

9 9 (M.C. Z. 168711-2) Mt. Mbololo, K. C. 25 & 28 April 1934. 

Diet. These birds were gorging on wild figs which were just ripening 
at the time. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 169 

Tricholaema melanocephalum stigmatothorax Cabanis 

Tricholaema stigmatothorax Cabanis, 1878, Journ. f. On., 26, p. 205: Ndi, 
Taita, Kenya Colony {vide p. 240). 

c? (M. C. Z. 168713) Tsavo, K. C. 4 April 1934. 

Distribution. This Brown-throated Barbet is almost topotypic as 
Tsavo station is probably less than fifteen miles from Ndi. 

Gymnobucco bonapartei cinereiceps Sharpe 

Gymnobucco cinereiceps Sharpe, 1891, Ibis, p. 122: Mount Elgon. 

7 cf , 2 9 (M. C. Z. 168714-22) Elgonyi, K. C. 22 January to 
4 February 1934. 

Trachyphonus darnaudii darnaudii (Prevost et Des Murs) 

Micropogon darnaudii Provost et Des Murs, 1850, in Lefebvre, Abyss., p. 133: 
Kordofan, Anglo-Egyptian Sudan. 

9 (M. C. Z. 168723) Karita River, U. 9 November 1933. 

Trachyphonus darnaudii bohmi Fischer & Reichenow 

Trachyphonus Bohmi Fischer & Reichenow, 1884, Journ. f. Orn., 32 p. 179: 
Barawa, Juba River, Italian Somaliland (see Zedlitz, 1915, Journ. f. 
Orn., p. 16). 

9 (M. C. Z. 168724) Voi. K. C. 7 April 1934. 



PICIDAE 

Campethera nubica pallida (Sharpe) 

Dendromus pallidus Sharpe, 1902, Ibis, p. 638: Lamu, Kenya Colony. 

c? 9 (M. C. Z. 168725-6) Kitau, Manda Id., K. C. 17-18 May 1934. 

Distribution. Manda Island in Lamu district is only separated from 
Lamu Island by a narrow channel. The Pale Nubian Woodpecker 
was twice seen on Lamu Island by Loveridge who shot the Kitau 
birds when they were in an acacia tree growing on the beach. 

Dendropicos fuscescens massaicus Neumann 

Dendropicus guineensis massaicus Neumann, 1900, Journ. f. Orn., 48, p. 206: 
Ndala-lani, Lake Nguruman. 

c? (M. C. Z. 168727) Kibwezi, K. C. 29 March 1934. 



170 bulletin: museum of comparative zoology 

Dendropicos fuscescens hemprichii (Ehrenberg) 

Picus hemprichii Ehrenberg, 1828, in Hemprich & Ehrenberg, Symb. Phys. 
Av., fol. r. : Akiko, Ethiopia. 

<? (M. C. Z. 168728) Lamu, Lamu Id., K. C. 11 May 1934. 
9 (M. C. Z. 168729) Kitau, Manda Id., K. C. 16 May 1934. 

Dendropicos lafresnayi lepidus (Cabanis & Heine) 

Ipoclonus lepidus Cabanis & Heine, 1863, Mus. Hein., Th. 4, Heft 2, p. 118: 
Ethiopia. 

c? (M. C. Z. 168730) Kaimosi, K. C. 25 February 1934. 

Yungipicus obsoletus ingens Hartert 

Iyngipicus obsoletus ingens Hartert, 1900, Novit. Zool., 7, p. 33: Nairobi, 
Kenya Colony. 

c? (M. C. Z. 168731) Mt. Debasien, U. 30 November 1933. 

EURYLAEMIDAE 

Smithornis capensis meinertzhageni van Someren 

Smithornis capensis meinertzhageni van Someren, 1919, Bull. Brit. Orn. Club, 
40, p. 24: Lerundo, Kavirondo, Kenya Colony. 

c? & (M. C. Z. 168732-3) Kaimosi, K. C. 23 February & 4 March 1934. 

ALAUDIDAE 

Mirafra poecilosterna massaica (Fischer & Reichenow) 

Megalophonus massaicus Fischer & Reichenow, 1884, Journ. f. Orn., 32, 
p. 55: Klein- Arusha, Tanganyika Territory. 

& 9 9 (M. C. Z. 168734-6) Tsavo, K. C. 3, 4, 5 April 1934. 

HIRUNDINIDAE 

*Hirundo rustica rustica Linne 

Hirundo rustica Linne, 1758, Syst. Nat., ed. 10, 1, p. 191: Europe; restricted 
type locality, Sweden (Hartert). 

tf 1 c? (M. C. Z. 168737-8) Kibwezi, K. C. 24 March 1934. 
c? (M. C. Z. 168739) Tsavo, K. C. 3 April 1934. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 171 

HlRUNDO ANGOLENSIS ARCTICINCTA Sharpe 

Hirundo arcticincta Sharpe, 1891, Ibis, p. 119: Mount Elgon at 7,000 feet, 
Kenya Colony. 

3 c? 3 9 (M. C. Z. 168740-5) Elgonyi, Mt. Elgon, K. C. 23-29 
January 1934. 

Variation. This topotypical series of six swallows shot, like the 
types, at the Elgon caves, has been compared by us with five exam- 
ples of the typical form of which three are topotypes and the remaining 
two from Kisenyi, Belgian Ruanda. These undoubtedly show that 
the northeastern bird, including one from Kome, Mwanza, averages 
paler on the abdomen. This is the view now held by Granvik (1934, 
p. 113) as opposed to his first action (1922, p. 118) in synonymizing 
arcticincta with the typical form. 

Breeding. Only two pairs of Uganda Swallows were found in Kemp's 
Cave but three or four pairs in caves on the opposite escarpment. 
A couple were observed continually returning to a nest, attached to 
the roof at a height of twenty feet from the cave floor. By means 
of a tree brought to the spot, a native was able to ascend to the nest 
but found that it was only in process of being built. (January 23, 
1934). 

Hirundo rufula emini Reichenow 

Hirundo emini Reichenow, 1892, Journ. f. Orn., 40, p. 30: Bussisi, southern 
shore of Lake Victoria, Tanganyika Territory. 

c? 9 (M. C. Z. 168746-7) Voi, K. C. 7 April 1934. 

c? 9 (M. C. Z. 168748-9) Mt. Mbololo, K. C. 14 April 1934. 

Ptyonoprogne rufigula rufigula (Fischer & Reichenow) 

Cotyle rufigula Fischer & Reichenow, 1884, Journ. f. Orn., 32, p. 53: Lake 
Naivasha, Kenya Colony. 

Breeding. Four nests of the Red-throated Rock Martin holding 
young which were being fed by their parents, were found attached to 
the roof of Kemp's and another cave below Elgonyi on the southern 
side of Mount Elgon, Kenya Colony, January 23-24, 1934. 

CAMPEPHAGIDAE 

Campephaga quiscalina martini Jackson 

Campophaga martini Jackson, 1912, Bull. Brit. Orn. Club, 31, p. 18: Ravine, 
i.e. Nandi, Kenya Colony. 

1 (M. C. Z. 168750) Label detached. 1933-1934. 



172 bulletin: museum of comparative zoology 

This Uganda Purple-throated Cuckoo Shrike was probably taken 
at Kaimosi, Kenya Colony, but its label is lost. The species has been 
recorded by van Someren (1922, p. 107) from Kakamega (in which 
district Kaimosi lies) and south Elgon. 

Coracina caesia pura (Sharpe) 

Graucalus purus Sharpe, 1891, Ibis, p. 121: Mount Elgon. 

<? (M. C. Z. 168751) Butandiga, U. 10 January 1934. 
c? (M. C. Z. 168752) Kirui's, K. C. 20 January 1934. 

Distribution. It is not certain whether Jackson shot the type on 
the Uganda or Kenya side of the mountain so that both of these 
birds may be considered topotypes for both were shot on Mount 
Elgon. 

DICRURLDAE 

Dicrurus adsimilis divaricatus (Lichtenstein) 

Muscicapa divaricata Lichtenstein, 1823, Verz. Doubl. Zool. Mus. Berlin, 
p. 52: Senegal. 

<? (M. C. Z. 168753) Karita River, U. 9 November 1933. 

Affinities. The wing of this bird measures 124 mm. If the sexing 
is correct it comes very near the coast form, for van Someren (1922, 
p. 125) gives 130-136 mm. for Uganda males though Granvik (1934, 
p. 116) says 117-124 mm. for his birds from the Elgon region. 

Dicrurus adsimilis fugax Peters 

Dicrurus fugax Peters, 1868, Journ. f. Orn., 16, p. 132: Tete and Inhambane, 
Mozambique. 

d> (M. C. Z. 168754) Lamu, Lamu Id., K. C. 11 May 1934. 

Affinities. The wing of this bird measures 115 mm. but it is im- 
mature; van Someren (1922, p. 125) gives 117-127 mm. for coastal 
males. 

ORIOLIDAE 

*Oriolus oriolus oriolus (Linne) 

Coracias Oriolus Linne, 1758, Syst. Nat., ed. 10, 1, p. 107: Europe, Asia. = 
Sweden apud Hartert. 

9 (M. C. Z. 168755) Voi, K. C. 7 April 1934. 

9 (M. C. Z. 168756) Mt. Mbololo, K. C. 24 April 1934. 



PETERS AND LOVERIDGE : AFRICAN BIRDS 173 

Diet. These European Golden Orioles were feeding upon the freshly- 
ripened wild figs when shot. 

Oriolus monacha rolleti Salvadori 

Oriolus Rolleti Salvadori, 1864, Atti Soc. Italiana Sci. Nat. Milano, 7, Ri- 
unione a Biella, p. 161: White Nile between lat. 4° and 5° N. 

d 1 J 1 (M. C. Z. 168757-8) Mt. Debasien, U. 13 & 29 December 1933. 
9 (M. C. Z. 168759) Sipi, Mt. Elgon, U. 22 December 1933. 

In the third volume of the Catalogue of Birds of the British 
Museum, 1877, the original citation for this form was given as 
"Atti R. Accad. Torin. 7, p. 151;". All subsequent authors have 
copied this, inserting 1864 as the year of publication; Reichenow 
in Vogel Afrikas gave a correct page number, but otherwise did his 
part in perpetuating the erroneous citation. We have checked Salva- 
dori's description with specimens identified as 0. rolleti and made the 
same comparison with 0. larvatus Licht. and 0. brachyrhynchvs 
Swains, as did Salvadori. We should not have been surprised to find 
that the original description referred to something entirely different 
than the common conception of rolleti, since no one seems to have 
looked at Salvadori's article for at least sixty-eight years; nevertheless 
the name seems to be correct! v allocated. 



CORVIDAE 

Corvultur albicollis (Latham) 
Corvus albicollis Latham, 1790, Ind. Orn., 2, p. 151: Africa. 

cf 9 (M. C. Z. 168760-61) Mt, Debasien, U. 16 November 1933. 

PARLDAE 

Parus albiventris albiventris Shelley 
Parus albiventris Shelley, 1881, Ibis, p. 116: Ugogo, Tanganyika Territory. 
d 1 (M. C. Z. 168762) Mt. Debasien, U. 27 November 1933. 

Distribution. Dr. van Someren (1922, p. 205) has recorded this 
species from Mount Moroto and Mount Elgon to the north and south 
of Debasien. 

Measurements. Length of wing, 80 mm., which is rather small for 
males of the typical race for which van Someren gives 83-86 mm. 



174 bulletin: museum of comparative zoology 

On the other hand, Granvik (1923, p. 227) gives 72-87 for four Elgon 
and upland males (is it possible that 72 is a misprint for 82?), later he 
(1934, p. 125) obtained four more males from Mount Elgon whose 
measurements were 77-85 mm. This would make the grounds for 
accepting Friedmann's small coastal race curtus rather slender; 
Sclater (1930, p. 642, footnote) regarded it as doubtfully distinct. 
P. a. curtus was based on a young female from Taveta and he gives 
the range for the males as 75-77 mm. That of the females appears to 
be 72-75 mm. 

Anthoscopus musculus (Hartlaub) 

Aegithalus musculus Hartlaub, 1882, Orn. Centralbl., p. 91: Lado (see Journ. 
f. Orn., 30, 1882, p. 326). 

d" (M. C. Z. 168763) s. bank Greeki River, U. 7 December 1933. 

Distribution. This locality is in the Sabei district at the north of 
Mount Elgon, about six miles in a direct line from the nearest moun- 
tain foothills. 

TIMELIIDAE 

Argyra rubiginosa heuglini Sharpe 

Argyra heuglini Sharpe, 1883, Cat. Birds Brit. Mus., 7, p. 391: Zanzibar. 

c? 9 + 1 (M. C. Z. 168764-6) Kitau, Manda Id., K. C. 19 May 1934. 

Synonymy. For the confusion in names applicable to this race see 
van Someren (1922, p. 235) but saturata Sharpe must be relegated to 
the synonymy of heuglini Sharpe as has already been done by Sclater 
(1930, p. 356, footnote). 

PYCNONOTIDAE 

Pycnonotus tricolor dodsoni Sharpe 

Pycnonotus dodsoni Sharpe, 1895, Proc. Zool. Soc. London, p. 488: Sillul, 
Ogaden country, Ethiopia. 

c? 9 (M. C. Z. 168767-8) Voi, K. C. 7 April 1934. 
<? cT tf (M. C. Z. 168769-71) Mt. Mbololo, K. C. 25-28 April 1934. 
cf (M. C. Z. 168772) Kitau, Manda Id., K. C. 16 May 1934. 

Measurements. Those of the wings fall within the range given for 
this race by van Someren (1932, p. 347) in his revision of the forms 
found in East Africa. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 175 

Pycnonotus tricolor fayi Mearns 

Pycnonotus layardi fayi Mearns, 1911, Smiths. Misc. Coll., 56, No. 20, 
p. 7: Fay's Farm, Njabini, Kenya Colony. 

<? 9 9 (M. C. Z. 168773-5) Butandiga, U. 6 & 10 January 1934. 

Variation. While two of these birds have the darker head of fayi, 
the third might well be referred to minor Heuglin, which latter race 
they should be according to van Someren (1932, p. 347). Thus they 
bear out Sclater's (1930, p. 372) remarks that Mount Elgon is the 
meeting place of the two forms. Butandiga is situated on the western 
slopes at 7,000 feet. 

Phyllastrephus sucosus sucosus Reichenow 

Phyllastrephus cabanisi sucosus Reichenow, 1903, Journ. f. Orn., 54, p. 544: 
Bukoba, Tanganyika Territory. 

9 (M. C. Z. 168776) Kaimosi, K. C. 16 February 1934. 

Distribution. Sclater (1930, p. 383) in giving the range says "per- 
haps Elgon." There is an Elgon example from the Jackson collection 
in the Museum of Comparative Zoology which is inseparable from 
our Kaimosi bird. Elsewhere van Someren (1932, p. 344) has shown 
that its range extends eastward to Molo in Kenya Colony. 

Phyllastrephus fischeri placidus (Shelley) 

Xenocichla placida Shelley, 1889, Proc. Zool. Soc. London, p. 363: Kiliman- 
jaro, Tanganyika Territory. 

5 & 1 9 (M. C. Z. 168777-82) Mt. Mbololo, K. C. 16-20 April 1934. 

Synonymy. Van Someren (1932, p. 343) agrees that keniensis 
Mearns from Mount Kenya is a synonym of this race. 

Arizelocichla milanjensis striifacies (Reichenow & Neumann) 

Xenocichla striifacies Reichenow & Neumann, 1895, Orn. Monatsb., 3, p. 74: 
Marangu, Kilimanjaro, Tanganyika Territory. 

<? <? 9 (M. C. Z. 168783-5) Mt. Mbololo, K. C. 18-19 April 1934. 

Distribution. Van Someren (1932, p. 345) has taken this race in 
the Taveta forest on the Lumi River. 

Measurements. Wings. & 90-92 mm., 9 86 mm. These are wholly 
comparable to birds from the Usambara and Uluguru Mountains, 
Tanganyika Territory. 



176 bulletin: museum of comparative zoology 

Stelgidocichla latirostris eugenia (Reichenow) 

Andropadus eugenius Reichenow, 1892, Journ, f. Orn., 40, p. 53: Bukoba, 
Tanganyika Territory. 

c? (M. C. Z. 168786) Sipi, U. 22 December 1933. 
9 9 (M. C. Z. 168787-8) Elgonyi, K. C. 22 & 25 January 1934. 
9 (M. C. Z. 168789) Kaimosi, K. C. 16 February 1934. 

Breeding. Testes of the Sipi bird were much enlarged. 

Eurillas virens holochlorus van Someren. 

Eurillas virens holochlorus van Someren, 1922, Novit. Zool., 29, p. 189: 
Sezibwa, Uganda. 

9 imm. (M. C. Z. 168790) Kaimosi, K. C. 1 March 1934. 

Affinities. Apropos the remarks by Sclater and Moreau (1932, p. 
683) on the Usambara birds referred to E. v. zombensis (Shelley) by 
Friedmann, we have laid out extensive series of E. v. virens (Cassin) 
from the Cameroons and observe them to be noticeably darker than 
the Usambara skins. Elgon birds, on the other hand, appear to occupy 
an intermediate position, conforming in their characteristics to those 
given by van Someren when describing holochlorus. Thus, so far as we 
are able to judge by our material, it does uphold van Someren's 
(1932, p. 346) later views. 

Probably the Usambara birds should have been called E. v. mar- 
witzi Reichenow, described from Kilimanjaro, but neither Friedmann 
nor we have any Nyasaland material by which to ascertain whether 
marwitzi is separable from zombensis. 

TURDIDAE 

Turdus olivaceus helleri (Mearns) 

Planesticus helleri Mearns, 1913, Smithsonian Misc. Coll., 61, No. 10, p. 1: 
Mount Mbololo, 4,000 feet, Kenya Colony, 
d* imm., cf 9 ad. (M. C. Z. 168791-3) Mt. Mbololo, K. C. 14-20 April 1934; 

Affinities. In a footnote, Sclater (1930, p. 441) suggests that helleri 
is closely related to, if not identical with, T. o. roehli Reichenow of the 
Usambara Mountains. Through the courtesy of Dr. H. Friedmann, 
we have been able to compare the type of helleri with these topotypes 
and also with topotypes of roehli. As a result we have no hesitation 
in saying that helleri is readily distinguishable by the sharply-defined 



PETERS AND LOVERIDGE: AFRICAN BIRDS 177 

black crown, the more extensive white area on the posterior under- 
pays, and the consequent reduction of the chestnut area. Both our 
adult birds agree with Mearns' type except that the upper parts are 
more olivaceous, the type having become somewhat foxed in the 
interval since it was collected. 

Turdus tephronotus Cabanis 

Turdus tephronotus Cabanis, 1878, Journ. f. Orn., 26, pp. 205, 218, pi. iii, 
fig. 2: Ndi, Taita district, Kenya Colony. 

d* (M. C. Z. 168794) Lamu Id. K. C. 11 May 1934. 
d* 9 (M. C. Z. 168795-96) Kitau, Manda Id., K. C. 15 May 1934. 
also two young in alcohol with same history. 

Measurements. The wings are cf 106, 100. 9 100 mm. Two males 
from the Tana River, Kenya Colony measure 103, 102 mm., 9 102 
mm. A pair from Dodoma and vicinity, central Tanganyika Terri- 
tory are: cf 112. 9 104 mm. Males from Sagon River and Wobok, 
Ethiopia, Endoto and s. of Milele in Kenya measure 109, 108, 107 
and 105, a female from this last locality was 100 mm. 

Enemies. As I (Loveridge) was walking through the dense acacia 
scrub an hour before sunset, my attention was attracted by the cries 
of two Bare-eyed Thrushes fluttering in a nearby acacia. Scarcely 
pausing to think, though the idea flashed through my mind that they 
might be indulging in courting antics, I dropped both birds with a 
right and left shot. 

My gunbearer ran to get them, stooping to pass beneath the low, 
spreading branches of the tree. Having picked up a bird he stood up 
on the far side, then called to me to come quickly as there was a snake 
in the tree. I ran under the acacia, then standing up, asked where the 
snake was. "Directly over your head," he replied. Looking up I saw 
not two feet above me a five-and-a-half-foot mamba (Dendraspis 
angusticeps). Moving to a distance so as not to injure the reptile, 
I shot it dead with a charge of dust shot. The gunbearer then said 
that he had noticed the bird flapping its wings in the snake's face just 
before I fired. 

On reaching camp I examined the stomach contents of the mamba 
and found that it had first swallowed a bat (Laviafrons rex), then two 
well-grown, though still nestling, thrushes. Returning to the spot 
I located the thrushes' nest in the next tree to that in which I had s! ot 
the snake. Undoubtedly the parent birds were engaged in an attempt 
to drive it away. 



178 bulletin: museum of comparative zoology 

*Monticola saxatilis (Linne) 

Turdus saxatilis Linne, 1766, Syst. Nat., ed. 12, 1, p. 294: Mountains of 
Switzerland, Austria, and Prussia; Switzerland, (apud Hartert). 

<? (M. C. Z. 168797) Voi, K. C. 7 April 1934. 

*Oenanthe oenanthe oenanthe (Linne) 

Motacilla oenanthe Linne, 1758, Syst. Nat., ed. 10, 1, p. 186: Europe; Sweden 
(apud Hartert). 

d" 9 (M. C. Z. 168798-9) Mt. Debasien. U. 20 & 29 November 1933 

*Oenanthe leucomela leucomela (Pallas) 

Motacilla leucomela Pallas, 1771, Novi Comm. Sci. Petrop., 14, p. 584, 
pi. xxii: Lower Volga. 

cf (M. C. Z. 168800) Mt. Debasien, U. 20 November 1933. 

Synonymy. Sclater (1930, p. 452, footnote) states that pleschanka 
Lepechin, though antedating leucomela, is, not strictly a binomial name. 

*Oenanthe isabellina (Temminck) 

Saxicola isabellina Temminck, 1829, PI. Col. livr. 79, pi. 472, fig. 1: Nubia. 

1+9 (M. C. Z. 168801-2) Tsavo, K. C. 31 March & 3 April 1934. 

Migration. Previously recorded from Tsavo by van Someren (1922, 
p. 242) who (1931, p. 11) states that they move northwards in March. 

Pinarochroa sordida rudolphi Madarasz 

Pinarochroa rudolphi Madarasz, 1912, Orn. Monatsb., 20, p. 175: Mount Elgon. 

9 (M. C. Z. 168803) Kaburomi, U. 29 December 1933. 
9 9 9 (M. C. Z. 168804-6) Madangi, U. 3 January 1934. 

Synonymy. Sclater (1930, p. 411, footnote) refers this form to 
the synonymy- of ernesti from Mount Kenya. We have compared 
these topotypic birds, all shot in the alpine zone of Mount Elgon above 
10,500 feet, with topotypes of ernesti from Mount Kenya in the col- 
lection of the Museum of Comparative Zoology. We find that the 
Elgon birds have darker ear-coverts while their superloreal stripes 
are more brown and less gray, in three of the Elgon series the breast 
plumage is richer brown, less gray. We therefore consider that Gran- 
vik (1923, p. 251) was justified in maintaining it as distinct. This 
author gives a good description of the behaviour and habitat of these 
interesting hill-chats. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 179 

CoSSYPHA HEUGLINI HEUGLINI Hartlailb 

Cossypha heuglini Hartlaub, 1886, Journ. f. Orn., 34, p. 36: "Keren" errore 
pro Wau, Bahr el Ghazal, Sudan (vide Heuglin, Orn. Nordost-Afr., 1, p. 
375). 

<? (M. C. Z. 168807) Mt. Debasien, U. 18 November 1933. 

ClCHLADUSA GUTTATA RUFIPENNIS Sharpe 

Cichladusa rufipennis Sharpe, 1901, Bull. Brit. Orn. Club, 12, p. 35: Lamu, 
Kenya Colony. 

<? tf <? (M. C. Z. 168808-10) Kitau, Manda Id., K. C. 15-18 May 1934. 

Distribution. These birds are practically topotypes, Kitau being 
separated from Lamu Island by a relatively narrow channel on the 
one hand and from the mainland (formerly called Lamu, more re- 
cently Lamu district) by a still narrower one. 

Breeding. Several of the characteristic mud nests were found in 
acacias May 19, 1934. One held young, from another the young 
appeared to have flown, two others were inaccessible. 

POGONOCICHLA MARGARITIFERA HELLERI Meams 

Pogonocichla cucullata helleri Mearns, 1913, Smiths. Misc. Coll., 41, No. 20, 
p. 1 : Mount Mbololo, Kenya Colony. 

& <? (M. C. Z. 168811-2) Mt. Mbololo, K. C. 20-21 April 1934. 

Synonymy. Sclater (1930, p. 487, footnote) synonymizes this form 
with P. m. guttifer (Reichenow & Neumann) from Kifinika, Kiliman- 
jaro. As both topotypes are juveniles in spotted plumage we are 
unable to offer an opinion. The Mbololo birds were shot in the forest 
at 4,800 feet while the type of guttifer was taken at over 9,000 feet. 



SYLVIIDAE 

*Sylvia borin (Boddaert) 

Motacilla borin Boddaert, 1783, Table PL Enl., p. 35: France (ex Daubenton, 
PI. Enl., 579, fig. 2). 

c? (M. C. Z. 168813) Elgonyi, K. C. 22 January 1934. 
9 9 (M. C. Z. 168814-5) Kaimosi, K. C. 1 & 5 March 1934. 

Migration. The Garden Warbler is a common migrant to East 
Africa, vide van Someren (1931, p. 16) who states that most of them 



180 bulletin: museum of compaeative zoology 

depart by the second week in April though he has obtained stragglers 
as late as the first week in May. 

*Sylvia atricapilla atricapilla (Linne) 

Motacilla atricapilla Linne, 1758, Syst. Nat., ed. 10, 1, p. 187: Europe; restricted 
type locality, Sweden. 

1 (M. C. Z. 168816) Kirui's, K. C. 20 January 1934. 
c* & 4 9 (M. C. Z. 168817-21) Elgonyi, K. C. 23-30 January 1934. 

Migration. Both van Someren (1931, p. 15) and Sclater and Moreau 
(1933, p. 19) have commented recently on the occurrence of the 
European Blackcap in Kenya and Tanganyika respectively. 

*Phylloscopus trochilus trochilus (Linne) 

Motacilla trochilus Linne, 1758, Syst. Nat., ed. 10, 1, p. 188: England (vide 
Hartert, 1907, Vog. pal. Fauna, 1, p. 507). 

<? (M. C. Z. 168822) Mt. Debasien, U. 20 November 1933. 
c? (M. C. Z. 168823) Elgonyi, K. C. 30 January 1934. 
9 9 9 (M. Z. 168824-6) Kaimosi, K. C. 2 & 5 March 1934. 

Migration. According to van Someren (1931, p. 14) the Willow 
Wren arrives in Kenya in September and departs in April though 
stragglers may occasionally be taken as late as July. 

Seicercus ruficapilla minulla (Reichenow) 

Chloropeta minulla Reichenow, 1905, Orn. Monatsb., 13, p. 181: Mlalo, 
near Wilhelmstal (now Lushoto), Usambara Mountains, Tanganyika 
Territory. 

9 (M. C. Z. 168827) Mt. Mbololo, K. C. 16 April 1934. 

Distribution. Sclater (1930, p. 506) is mistaken in stating that this 
race is known only from the type locality, for Friedmann (1928, p. 93) 
recorded ten specimens of Loveridge's collecting from Bagilo, Nyingwa 
and Vituri in the Uluguru Mountains to the south as well as from 
Amani in the eastern Usambaras (Mlalo is in the western). More 
recently Moreau (1933, p. 20) has taken it at Amani and also in the 
Nguru Mountain. This Mbololo bird apparently constitutes the first 
record of its occurrence in Kenya Colony. 

Seicercus umbrovirens mackenzianus (Sharpe) 

Cryptolopha mackenziana Sharpe, 1892, Ibis, p. 153: Kikuyu, Kenya Colony. 

c? (M. C. Z. 168828) Sipi, Mt. Elgon, U. December 1933. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 181 

Calamonastes simplex simplex (Cabanis) 

Thamnobia simplex Cabanis, 1878, Journ. f. Orn., 26, pp. 205, 221: Ndi, Taita 
district, Kenya Colony. 

J* (M. C. Z. 168829) Tsavo, K. C. 4 April 1934. 

Distribution. The type locality is only about fifteen miles from 
Tsavo from which place van Someren (1922, p. 228) has already 
taken the Sooty Scrub Warbler. 

Apalis flavocincta (Sharpe) 

Euprinoides flavocinclus Sharpe, 1882, Journ. f. Orn., 30, p. 346: Adi, i.e. 
Athi River, Kenya Colony. 

c? (M. C. Z. 168830) Kitau, Manda Id., K. C. 18 May 1934. 

Affinities. Though Sclater (1930, p. 524) treats this bird as a race 
of flavida, van Someren (1932, p. 367) has adduced good reasons for 
regarding it as a full species. 

Apalis porphyrolaema porphyrolaema Reichenow & Neumann 

Apalis por-phyrolaema Reichenow & Neumann, 1895, Orn. Monatsb., 4, 
p. 75: Eldama, Kenya Colony. 

c? (M. C. Z. 168831) Sipi, Mt. Elgon, U. 23 December 1933. 

Breeding. The testes of this bird were very large. 

Apalis rufifrons rufidorsalis (Sharpe) 

Dryodromas rufidorsalis Sharpe, 1897, Bull. Brit. Orn. Club, 6, p. 48: Tsavo 
River, Kenya Colony. 

<? (M. C. Z. 168832) Tsavo River, K. C. 2 April 1934. 

Sylvietta brachyura leucopsis (Reichenow) 

Sylviella leucopsis Reichenow, 1879, Orn. Centralbl., p. 114: Kibaradja, Tana 
River, Kenya Colony. 

<? (M. C. Z. 168833) Tsavo River, K. C. 5 April 1934. 

Sylvietta leucophrys leucophrys (Sharpe) 
Sylviella leucophrys Sharpe, 1891, Ibis, p. 120: Mount Elgon. 

& (M. C. Z. 168834) Sipi, Mt. Elgon, U. 22 December 1933. 



182 bulletin: museum of comparative zoology 

Eremomela pusilla elgonensis van Someren 

Eremomela elegans elgonensis van Someren, 1920, Bull. Brit. Orn. Club, 40, 
p. 92: Kibingei, south Elgon, Kenya Colony. 

3^19 (M. C. Z. 168469-72) Mt. Debasien, U. 25 & 29 November 1933. 

Distribution. These records form a slight extension of the range 
northwards in Uganda. 

Eremomela badiceps turneri van Someren 

Eremomela badiceps turneri van Someren, 1920, Bull. Brit. Orn. Club, 40, 
p. 92: Yala River, Kakamega, Kenya Colony. 

9 (M. C. Z. 168835) Kaimosi, K. C. 5 March 1934. 

Distribution. This bird is actually a topotype as Turner's camp site 
was only a hundred yards from where Loveridge camped in 1934 and 
the latter did much of his collecting in the forest along the banks of 
the Yala River. 

Measurements. Wing 48 mm. 

Camaroptera. brevicaudata aschani Granvik 

Camaroptera brevicaudata aschani Granvik, 1934, Rev. Zool. Bot. Air., 25, 
p. 101: northeastern slopes of Mount Elgon, Kenya Colony. 

9 (M. C. Z. 168836) Mt. Debasien, U. 29 November 1933. 

Affinities. Granvik (1934, p. 103) has recorded a bird from Kache- 
liba, at the southeastern foot of Debasien, as an intermediate between 
aschani and abyssinica of Zedlitz. Our bird, collected at 5,000 feet 
on the western slopes of the mountain, may be in the same position. 
It is in post nuptial moult and does not appear to be immature but 
lack of adequate comparative material renders it inadvisable for us to 
attempt to discuss the status of the races. 

Measurements. Wing 51.5 mm. 

Camaroptera brevicaudata griseigula Sharpe 

Camaroptera griseigula Sharpe, 1892, Ibis, p. 158: Voi River, Taita, Kenya 
Colony. 

tf c? (M. C. Z. 168837-8) Kibwezi, K. C. 27 March 1934. 

Measurements. Wings 54, 55 mm. See Granvik (1934, p. 101) for 
discussion of type. 



PETERS AND LOVERIDGE : AFRICAN BIRDS 183 

Cisticola chubbi Sharpe 

Cisticola chubbi Sharpe, 1892, Ibis, p. 157: Kimangitchi, Mt. Elgon. 

d" (M. C. Z. 168839) Kaburomi, U. 28 December 1933. 
cf (M. C. Z. 168840) Kaimosi, K. C. 23 February 1934. 

Affinities. Kaburomi is on western Elgon at about 10,500 feet. 
After comparing these birds with nigriloris Shelley from northwest 
Lake Nyasa, one wonders why Lynes did not reduce the latter to 
subspeeific relationship with chubbi. Doubtless he had good reasons 
for treating them as distinct. 

Cisticola galactotes haematocephala Cabanis 

Cisticola haematocephala Cabanis, 1868, Journ. f. Orn., 16, p. 412: Mombasa, 
Kenya Colony. 

4 c? (M. C. Z. 168841-4) Kitau, Manda Id., K. C. 17-18 May 1934. 
c? 9 9 nest & eggs (M. C. Z. 168845-7) Wema, K. C. 11-14 June 1934. 

Native name. Dowe (Kipokomo). 

Measurements. A male from Mombasa has a wing of 58 mm., the 
four males from Manda Island measure 50-56 mm. 

Breeding. At Wema, Tana River, on the night of June 14, 1934 a 
hen bird was netted on her nest in long grass after several attempts to 
shoot her during the preceding day. Unfortunately she broke one egg 
and punctured a second with her claws. The nest, which was attached 
to long grasses at a height of eighteen inches from the ground, mea- 
sured 120 mm. high by 70 mm. in diameter, being slightly domed. 
Both within and without it was composed of dry grasses kept together 
to a slight extent by vegetable down. It held three fresh eggs, measur- 
ing 17 x 13 mm., of a pale pink ground color blotched with dark red, 
these red blotches were chiefly concentrated in a circle round the 
larger pole. 

Cisticola troglodytes troglodytes (Antinori) 

Drymoica troglodytes Antinori, 1864, Cat. Descr. Coll. Uccelli, p. 38: Djur, 
Bahr el Ghazal, Sudan. 

4 c? 1 9 (M. C. Z. 168848-52) Mt. Debasien, U. 16-30 November 1933. 

Variation. These birds in November plumage bear out the remarks 
of Granvik (1934, p. 109) on their dress. Van Someren (1932, p. 362) 
doubts the correctness of referring Suk and Turkana birds, with 
which he would presumably include these Karamojo skins, to the 
typical form. 



184 bulletin: museum of comparative zoology 

Melocichla mentalis orientalis (Sharpe) 

Cisticola orientalis Sharpe, 1883. Cat. Birds Brit. Mus., 7, p. 245: Pangani 
River, Tanganyika Territory. 

cf (M. C. Z. 168853) Kibwezi, K. C. 28 March 1934. 

Prinia mistacea immutabilis van Someren 

Prinia mistacea immutabilis van Someren, 1920, Bull. Brit. Orn. Club, 40, 
p. 93: Lake Nakuru, Kenya Colony. 

c? <? <? (M. C. Z. 168854-6) Mt. Debasien, U. 29 November 1932l 

Distribution. Granvik (1934, p. 110) has recorded this race from 
Mount Elgon. 

Prinia mistacea tenella (Cabanis) 

Drymoica tenella Cabanis, 1869, in von der Decken, Reise, 3, p. 23: Mombasa, 
Kenya Colony. 

Breeding. At Witu on June 1, 1934 a Mombasa Tawny-flanked 
Longtail completed her clutch and commenced sitting. The nest, 
which was attached to tall mpimbi grass stems at a height of three and 
a half feet from the ground, measured 115 mm. high by 70 mm. in 
diameter, being composed of dry grass lined with finer grasses. It 
held three fresh eggs, measuring 17 x 11 mm., of a pale sea-green 
ground color blotched with pale brown, chiefly around the larger pole, 
and scribbled and splashed with sepia brown to a lesser extent. Van 
Someren (1922, p. 219) has obtained this race at Witu as well as on 
the islands of Lamu and Manda. 

Prinia somalica erlangeri Reichenow 

Prinia somalica erlangeri Reichenow, 1905, Orn. Monatsb., 13, p. 24: Gurra 
country, S. Somaliland. 

c? 9 (M. C. Z. 168857-8) Kitau, Manda Id., K. C. 18 May 1934. 

Synonymy. P. intermedia Jackson from the Northern Guaso Nyiro is 
regarded as a synonym. 

Prinia leucopogon reichenowi (Hartlaub) 

Burnesia reichenowi Hartlaub, 1890, Journ. f. Orn., 38, p. 151: Njangalo, 
northeastern Belgian Congo. 

c? 9 (M. C. Z. 168859-60) Kaimosi, K. C. 12 & 16 February 1934. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 185 

Prinia bairdii melanops (Reichenow & Neumann) 

Burnesia melanops Reichenow & Neumann, 1895, Orn. Monatsb., 3, p. 75: 
Mau, Kenya Colony. 

9 (M. C. Z. 168861) Kaimosi, K. C. 1 March 1934. 



MUSCICAPIDAE 

Alseonax minimus inte-rpositus van Someren 

Alseonax minimus interpositus van Someren, 1930, Journ. E. Africa & Uganda 
Nat. Hist. Soc, No. 37, p. 194: Molo, Kenya Colony. 

cf <? 9 (M. C. Z. 168862-4) Mt. Debasien, U. 20-25 November 1933. 

d" (M. C. Z. 168865) Butandiga, K. C. 6 January 1934. 

d" 9 (M. C. Z. 168866-7) Elgonyi, K. C. 22-25 January 1934. 

<? (M. C. Z. 168868) Kaimosi, K. C. 1 March 1934. 

Variation. In the absence of topotypic material of A. m. murinus 
from Mount Meru for comparison, we follow Granvik (1934, p. 70) in 
referring our Elgon birds to this race though it appears to be based on 
rather slender grounds. 

Alseonax minimus murinus Fischer & Reichenow 

Alseonax murina Fischer & Reichenow, 1884, Journ. f. Orn., 32, p. 54: foot of 
Mount Meru, Tanganyika Territory. 

c? (M. C. Z. 168869) Mt. Mbololo, K. C. 19 April 1934. 

This bird is a juvenile and in consequence its identification some- 
what questionable, being based in part on geographical considerations. 

Bradornis pallidus subalaris Sharpe 

Bradornis subalaris Sharpe, 1873, Proc. Zool. Soc. London, p. 713, pi. lviii, 
fig. 1 : Mombasa, Kenya Colony. 

9 (M. C. Z. 168870) Kibwezi, K. C. 24 March 1934. 
c? 9 (M. C. Z. 168871-2) Lamu Id., K. C. 11 May 1934. 

<? (M. C. Z. 168873) Kitau, Manda Id., K. C. 16 May 1934. 

Breeding. The Kibwezi bird is a young one in speckled plumage but 
appears to belong to the coastal race without question and not to 

suahelicus. 



186 bulletin: museum of comparative zoology 

Bradornis pallidus suahelicus van Someren 

Bradornis pallidus suahelicus van Someren, 1921, Bull. Brit. Orn. Club, 41, 
p. 104: Londiani, Kenya Colony. 

1 + 4 cM 9 (M. C. Z. 168874-9) Mt. Debasien, U. 14-29 November 1933. 

Distribution. Granvik (1934, p. 71) obtained a good series of the 
Uganda Pale Flycatcher on Mount Elgon. 

Dioptrornis fischeri Reichenow 

Dioptrornis Fischeri Reichenow, 1884, Journ. f. Orn., 32, p. 53: Mount Meru, 
Tanganyika Territory. 

t?39 (M. C. Z. 168880-3) Kaimosi, K. C. 16 February & 2 March 1934. 

Batis molitor puella Reichenow 

Balis puella Reichenow, 1893, Jahrb. Hamburgischen Wiss. Anstalten, 10, 
part 1, p. 125: No locality. (Type from Bussisi, s. shore of Lake Victoria, 
Tanganyika Territory, in Berlin Museum). 

d* 9 (M. C. Z. 168884-5) Mt. Debasien, U. 29 November 1933. 
c? (M. C. Z. 168886) Butandiga, U. 15 January 1934. 

Variation. The throat of the 9 is pure white, in this respect being 
similar to a female from Mount Elgon referred to by Granvik (1934, 
p. 74) as lacking the chin spot. We identify these birds with puella 
rather than with B. o. perkeo Neumann on account of the larger size 
of their wings, viz. c? d 1 62, 60 mm.; 9 60 mm. 

Batis minor suahelicus Neumann 

Batis minor suahelicus Neumann, 1907, Journ. f. Orn., 55, p. 353: Kahe, 
near Kilimanjaro, Tanganyika Territory (fide type in Berlin Museum). 

& (M. C. Z. 168887) Kitau, Manda Id., K. C. 16 May 1934. 

Distribution. This bird has previously been recorded as occurring 
on Manda and Lamu Islands by van Someren (1932, p. 296). 

Platysteira peltata jacksoni Sharpe 

Platystira jacksoni Sharpe, 1891, Ibis, p. 445: Sotik, Kenya Colony. 

cf 9 9 (M. C. Z. 168888-90) Elgonyi, K. C. 22-30 January 1934. 

Distribution. Granvik (1934, p. 75) discusses the plumage variation 
in a series of these flycatchers which he obtained on the eastern slopes 
of Mount Elgon. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 187 

Corrigenda. The two birds (M. C. Z. 148620-1) from Igale, Poroto 
Mountains, referred to typical peltata by Bangs & Loveridge (1933, 
p. 190), are actually jacksoni conforming with that race in the steely 
blue reflections of their plumage. 

Erranornis longicauda teresita (Antinori) 

Elminia teresita Antinori, 1864, Cat. Descr. Coll. Ucc, p. 50: Djur, Bahr el 
Ghazal, Sudan. 

cf (M. C. Z. 168891) Elgonyi, K. C. 31 January 1934. 

Distribution. The Bahr el Ghazal Blue Flycatcher occurs in the 
dry scrub and Baobab bush below 7,000 feet where I found it as 
common as did Granvik (1934, p. 75) on the northeast slopes of the 
mountain. 

Tchitrea viridis viridis (Miiller) 

Muscicapa viridis Miiller, 1776, Natursyst., Suppl., p. 171: Senegal. 

cf cT (M. C. Z. 168892-3) Mt. Debasien, U. 16 & 20 November 1933. 

Variation. One of these Paradise Flycatchers has the back and tail 
reddish brown, the tail only half grown, measuring 130 mm., the 
other with a tail of 330 mm., has both back and tail almost pure white. 

MOTACILLIDAE 

*MOTACILLA CINEREA CINEREA Tunstall 

Motacilla cinera Tunstall, 1771, Orn. Brit., p. 2: Great Britain. 

& <? cf (M. C. Z. 168894-6) Mt. Debasien, U. 14 & 15 November 1933. 

Migration. Granvik (1934, p. 61) obtained the Gray Wagtail on 
Mount Elgon in November and December 1926. 

Anthus sordidus longirostris Neumann 

Anthus nicholsoni longirostris Neumann, 1905, Orn. Monatsb., 13, p. 77: 
Gardulla, Lake Gandjula (i.e. Lake Abaya), Ethiopia. 

c? (M. C. Z. 168897) Mt. Mbololo, K. C. 26 April 1934. 

Anthus richardi lacuum Meinertzhagen 

Anthus richardi lacuum Meinertzhagen, 1920, Bull. Brit. Orn. Club, 41, p. 22: 
Lake Naivasha, Kenya Colony. 

9 (M. C. Z. 168898) Malindi, K. C. 29 June 1934. 



188 bulletin: museum of comparative zoology 

Measurements. This is a particularly small bird with wing of 74.5 
mm., its head is damaged and its plumage worn so that the identifi- 
cation is none too certain. 

*Anthus trivialis trivialis (Linne) 

Alauda trivialis Linne" 1758, Syst, Nat., ed. 10, 1, p. 166: Sweden. 

c? (M. C. Z. 168899) Mt. Debasien, U. 28 November 1933. 

Migration. Granvik (1934, p. 62) collected the Tree Pipit on Mount 
Elgon also in November and December 1926. 

Macronyx croceus croceus (Vieillot) 

Alauda crocea Vieillot, 1816, Nouv. Diet. d'Hist. Nat., 1, p. 365: "Java" 

(Senegal, fide Swainson.) 

1 (M. C. Z. 168900) Mt. Debasien, U. 16 November 1933. 
c? (M. C. Z. 168901) Elgonyi, K. C. 28 January 1934. 

Breeding. Granvik (1934, p. 63) found the Yellow-throated Long- 
claw nesting on Mount Elgon on June 23 (1923, p. 200). 



LANIIDAE 

Lanius excubitorius princeps Cabanis 

Lanius princeps Cabanis, 1850, Mus. Hein., Heft 1, p. 73, note: Sources of 
the Nile, Sudan. 

<? (M. C. Z. 168902) N. bank Greeki River, U. 6 November 1933. 

Distribution. The actual spot where these birds were very abundant 
and our specimen shot, is known as Nabugut, though no village marks 
the site. 

Synonymy. Sclater (1930, p. 611, footnote) synonymized this race 
with the typical form, but van Someren (1922, p. 123; 1932, p. 311) 
upholds it as a bird of migratory habits and smaller wing measure- 
ments. Our specimen has a wing of 111 mm. We have a pair of birds 
from south Ankole (wings 114, 114 mm.) taken by Loveridge's native 
collector at the same time as van Someren's collector secured the 
birds referred by him to "excubitorius Rchw." (1922, p. 123). As 
Reichenow is not the author of excubitorius it seems obvious that 
that the subspecific name bohmi was accidentally omitted. Pre- 
sumably these are the same birds which later (1932, p. 311) he re- 
ferred to L. e. princeps. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 189 



Lanius mackinnoni Sharpe 

Lanius mackinnoni Sharpe, 1891, Ibis, p. 444, pi. xiii: Kavirondo, Kenya 
Colony. (Type from Bugemaia vide Ibis, 1891, p. 596). 

2d , 39 (M. C. Z. 168903-7) Kaimosi, K. C. 16 February 1934. 



*Lanius collurio Linne 

Lanius collurio Linne, 1758, Syst.[Nat., ed. 10, 1, p. 94: Europe (i.e. Sweden, 
fide Hartert). 

9 (M. C. Z. 168908) Tsavo, K. C. 2 April 1934. 
c? c? (M. C. Z. 168909-10) Voi, K. C. 10 & 13 April 1934. 

Migration. Van Someren (1931, p. 21) refers to large flocks in 
the Tsavo-Taru area which departed northward on April 20. 



Corvinella corvina chapini Friedmann & Bowen 

Corvinella corvina chapini Friedmann & Bowen, 1933, Proc. Biol. Soc. Wash- 
ington, 46, p. 121 : Kibigori, Kavirondo, Kenya Colony. 

1 (M. C. Z. 168911) Kitale, K. C. 25 December 1928. 

This Yellow-billed Strike was given to me for presentation to 
the museum by Mr. Jeffries of Kitale. He has noted on the label 
"Bill yellow, legs greyish. Wing 125 mm." 

Laniarius ferrugineus major (Hartlaub) 

Telephonus major Hartlaub, 1848, Rev. Zool., p. 108: Elmina, Gold Coast. 

c? 9 (M. C. Z. 168912-3) Kaimosi, K. C. 23 February 1934. 

Measurements. Wing of c? , 99 mm., 9 93 mm. See also Granvik 
(1934, p. 120) who obtained this race on Mount Elgon. 

Laniarius luhderi castaneiceps Sharpe 

Laniarius castaneiceps Sharpe, 1891, Ibis, p. 445: Mount Elgon. 

tf (M. C. Z. 168914) Elgonyi, K. C. 4 February 1934. 

Synonymy. This race, which Sclater (1930, p. 621, footnote) fol- 
lowing Jackson, disallowed, is upheld by van Someren (1932, p. 306) 
and substantiated by our measurements which are 80 and 85 mm. 
for the wings of these two males from Mount Elgon, while the wings 
of two males from the Cameroons measure 89 and 91 mm. 



190 bulletin: museum of comparative zoology 

Dryoscopus affinis (Gray) 

Hapalophus affmis G.R. Gray, 1837, Mag. Nat. Hist., 1, (n.s.) p. 489: Zanzibar. 

9 (M. C. Z. 168915) Lamu Id., K. C. 9 May 1934. 
<? (M. C. Z. 168916) Mkonumbi, K. C. 29 May 1934. 

Distribution. Van Someren (1922, p. 120) mentions birds from 
Lamu which were presumably part of the rich collections made for 
him in this region by Mr. H. J. Allen Turner. 

TSCHAGRA AUSTRALIS EMINI (Reichenow) 

Telephonus australis emini Reichenow, 1893, Orn. Monatsb., 1, p. 60: Bukoba, 
Tanganyika Territory. 

c? (M. C. Z. 168917) Mt. Debasien, U. 21 November 1933. 

Measurements. Wing SO. 5 mm. Granvik's (1934, p. 121) experience 
of collecting opposite sexes of this and the succeeding species in the 
same locality, is corroborated by two of our examples of the genus. 

TSCHAGRA SENEGALA ARMENA (Oberholser) 

Pomalorhynchus senegalus armenus Oberholser, 1906, Proc. U. S. Nat. Mus., 
30, p. 809: Taveta, Kenya Colony. 

9 (M. C. Z. 168918) Mt. Debasien, U. 13 November 1933. 
Measurements. Wing 85 mm. 

TSCHAGRA SENEGALA ORIENTALIS (Cabanis) 

Pomatorhynchus orientalis Cabanis, 1869, in von der Decken, Reise 3, p. 27: 
Mombasa, Kenya Colony. 

Breeding. At Peccatoni near Witu, on May 28, 1934, a nest of the 
Mombasa Black-headed Tchagra was found in a low dom palm at a 
height of four feet from the ground and in a very exposed position. 
It held a single egg, a second egg was laid about 9 a.m. on the 29th. 
On the 30th I visited the nest several times before 7.15 a.m. but the 
bird had not managed to complete her clutch so I left her in peace 
as camp had been struck and I could not delay further. 

Tschagra jamesi mandana (Neumann) 

Telephonus jamesi mandanus Neumann, 1903, Orn. Monatsb., 11, p. 183: 
Manda Island, Kenya Colony. 

d" (M. C. Z. 168919) Lamu, Lamu Id., K. C. 11 May 1934. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 191 

Breeding. At Kitau, Manda Island, on May 17, 1934, a topotype of 
this race was seen on its nest. Several attempts to surprise her leav- 
ing the nest so as to get a shot, failed, for she slipped off into the 
surrounding thickets too quickly and would not return so long as I 
waited in the vicinity. The nest was built, or one might say 'slung' 
between two dry twigs, one of acacia, the other of a climbing euphorbia 
which was literally smothering the acacia in which the nest had been 
placed at a height of seven feet from the ground. The nest, measuring 
on the outside 50 mm. high by 100 mm. in diameter, was 30 mm. deep 
by 70 mm. in diameter on the inside. Externally it was composed of 
skeleton leaves and shreds of grass-like bark, internally it was roughly 
lined with a few coarse grasses, fibres and a single feather. 

It held three fresh eggs, measuring 20 x 13 mm., of a white ground 
heavily flecked with brown except for the upper pole which is capped 
with brown almost to the exclusion of the ground color. 

Rhodophoneus cruentus cathemagmenus (Reichenow) 

Laniarius cathemagmenus Reichenow, 1887, Journ. Orn., 35, p. 63: Loeru. 

cf (M. C. Z. 168920) Tsavo, K. C. 4 April 1934. 

Affinities. We follow van Someren (1932, p. 305) in regarding this 
Rosy-patched Shrike as a race of cruentus instead of a full species as 
is done by Sclater (1930, p. 638). This action is justified by the com- 
parative material in the museum collection. 

PRIONOPIDAE 

Prionops cristata omoensis Neumann 

Prionops cristatus omoensis Neumann, 1905, Journ. f. Ornith., 53, p. 216: 
Omo River, Ethiopia. 

1 + c? (M. C. Z. 168921-2) Karita River, U. 9 November 1933. 
? <? (M. C. Z. 168923) Mt. Debasien, U. 29 November 1933. 

Measurements. Wings of Nos. 168921-2, 126 and 124 mm., No. 
168923, 114 mm. 

STURNIDAE 

ClNNYRICINCLUS LEUCOGASTER VERREAUXI (Bocage) 

Pholidauges verreauxi Bocage, 1870, in Finsch & Hartlaub, V6g. Ost-Afr., 
p. 867: Caconda, Angola. 

3 ef 1 9 (M. C. Z. 168924-7) Mt. Debasien, U. 25 & 30 November 1933. 
9 (M. C. Z. 168928) Voi, K. C. 7 April 1934. 



192 bulletin: museum or comparative zoology 

Synonymy. The status of the race lauragrayae Bowen has recently 
received attention from van Someren (1932, p. 313) and Sclater and 
Moreau (1933, p. 206), the former unable to support it, the latter 
rather doubting its validity. 

After measuring an extensive series in the Museum of Comparative 
Zoology, we are quite unable to endorse Bowen's claim that smaller 
size is a characteristic of birds inhabiting the area which he assigns 
to lauragrayae. Conversely some of our East African birds are ac- 
tually larger than South African specimens, the measurements not 
even overlapping. Furthermore we concur with Sclater that the 
amount of white in the tail feathers is too variable a character to 
receive consideration. 

Lamprocolius chalybeus sycobius Hartlaub 

Lamprocolius sycobius Hartlaub, 1859, Journ. f. Orn., 7, p. 19:Tete, Mozam- 
bique. 

<? 9 9 (M. C. Z. 168929-31) Voi, K. C. 7 April 1934. 

Synonymy. Van Someren (1921, Bull. Brit. Orn. Club, 41, p. 124) 
proposed the name pestis for birds which are intermediate in size 
between typical chalybeus of the Sudan and Ethiopia, and sycobius of 
Mozambique. Sclater (1930, p. 657, footnote) relegates pestis to the 
synonymy of sycobius an action with which we are in agreement not- 
withstanding the defence offered by van Someren (1932, p. 315). 
The measurements of the three Voi birds listed above are of little 
significance for the cf (wing, 113 mm.) and a 9 (wing, 118 mm.) 
are young while the other 9 is moulting. 

Lamprocolius splendidus splendidus (Vieillot) 

Turdus Splendidus Vieillot, 1822, Tabl. Encyc. Meth. Orn., pt. 2, p. 653. 
Malimba, Portuguese Congo. 

c? (M. C. Z. 168932) Kaimosi, K. C. 25 February 1934. 

Lamprotornis purpuropterus purpuropterus Riippell 

Lamprotornis purpuropterus Riippell, 1845, Syst. Uebers., pp. 64, 75, 
pi. xxv : Shoa, Ethiopia. 

cf 9 (M. C. Z. 168933-4) Mkonumbi, K. C. 29 May 1934. 

Distribution. Though well to the east of the distribution of this 
race as defined by Sclater (1930, p. 661) we have no hesitancy in 



PETERS AND LOVERIDGE : AFRICAN BIRDS 193 

identifying these birds as purpuropterus after comparing them with 
skins from Ethiopia, the Ithanga Hills, Kenya Colony and Ukerewe 
Island, Lake Victoria, Tanganyika Territory. 

Onychognathus morio shelleyi (Hartert) 

Amydrus morio shelleyi Hartert, 1891, Kat. Vogelsamml. Mus. Senck., p. 75 
note: Ugogo, Tanganyika Territory. 

d 1 (M. C. Z. 168935) Mt. Mbololo, K. C. 25 April 1934. 

Onychognathus morio ruppellii (Verreaux) 

Amydrus ruppellii Verreaux, 1856, in Chenu, Encycl. d'Hist. Nat. Ois., 5, 
p. 166: Ethiopia. 

<? (M. C. Z. 168936) Butandiga, U. 4 January 1934. 
<? <? (? (M. C. Z. 168937-9) Elgonyi, K. C. 29 & 31 January 1934. 

Synonymy. In 1919 van Someren described montanus from Mount 
Elgon and Sclater (1930, p. 665) gives its range as from 9,000 to 10,000 
feet. Our birds are from the west and south slopes of Mount Elgon 
at about 7,000 feet. Granvik (1934, p. 129) refers two females which 
he collected on Elgon to ruppellii and states that he does not consider 
montanus to be distinct from ruppellii. It is very probable that this 
author does not appreciate the significance of altitude in this con- 
nection. 

Our birds are clearly not montanus for they are smaller, not larger 
than ruppellii, nor are their bills different from Ethiopian examples 
of ruppellii. Three males of the latter from Maraco, Ethiopia, measure: 
Bills, 28, 27, 26 mm. Wings, 163, 161, 162 mm. Our four Elgon males 
measure: Bills, 28, 29, 29, 25.5 mm. Wings, 152, 152, 155, 149 mm. 

Spreo superbus (Riippell) 

Lamprocolius superbus Riippell, 1845, Syst. Uebers., pp. 65, 75, pi. xxvi: 
Shoa, Ethiopia. 

d 1 (M. C. Z. 168940) Tsavo, K. C. 4 April 1934. 

BUPHAGUS ERYTHRORYNCHUS CAFFER Grote 

Buphagus erythrorynchus caffer Grote, 1927, Orn. Monatsb., 35, p. 13:Selala 
River, Transvaal. 

c? (M. C. Z. 168941) Kaimosi, K. S. 26 February 1934. 



194 bulletin: museum of comparative zoology 



NECTARINIIDAE 

Nectarinia formosa centralis "Neumann" van Someren, 1916, Ibis, p. 446: 
"Scrub country", i.e. Lusasa, Uganda. 

cf (M. C. Z. 168942) Kaburomi, U. 30 December 1933. 

Variation. This specimen appears to constitute the first record for 
any race of Malachite Sunbird on Mount Elgon. The bird has strong 
coppery reflections on the crown, cheeks and anterior underparts 
which at first led us to suppose that it must be N. f. cuprconitens 
Shelley, but close examination shows that this color is not symmet- 
rically distributed and furthermore it is easily washed off with a little 
plain water. This stain therefore is probably from the nectar of the 
flowers from which the bird feeds. The question naturally arises, is 
not the copper red gloss on the plumage of the type of cupreonitens of 
similar origin? 

In the absence of topotypical material of cupreonitens we have 
been able to make comparison only with aenigularis and centralis 
and conclude that while the latter is not too well characterized it 
may usually be distinguished by the slightly greener, less blue, gloss 
on the posterior underparts. On this basis our specimen is centralis 
though it agrees with aenigularis in having a more strongly decurved 
bill. 

Nectarinia kilimensis kilimensis Shelley 

Nectarinia kilimensis Shelley, 1884, Proc. Zool. Soc. London, p. 555: Kili- 
manjaro, about 5,000 feet, Tanganyika Territory. 

9 (M. C. Z. 168943) Kaburomi, U. 30 December 1933. 
9 (M. C. Z. 168944) Kaimosi, K. C. 2 March 1934. 

Cinnyris bifasciatus tsavoensis van Someren 

Cinnyris bifasciatus tsavoensis van Someren, 1922, Novit. Zool. 29, p. 196: 
Tsavo, Kenya Colony. 

c? 9 (M. C. Z. 168945-6) Voi, K. C. 10 & 11 April 1934. 

Voi is less than thirty miles distant from Tsavo, the type locality. 

Cinnyrus venustus igneiventris Reichenow 

Cinnyris igneiventris Reichenow, 1899, Orn. Monatsb., 7, p. 171: Karagwe, 
Uganda. 



PETERS AND LOVERIDGE : AFRICAN BIRDS 195 

9 (M. C. Z. 168947) Sipi, U. 14 December 1933. 

<? (M. C. Z. 168948) Kaburomi, U. 30 December 1933. 

c? (M. C. Z. 168949) Butandiga, U. 6 January 1934. 

Cinnyris venustus falkensteini Fischer & Reichenow 

Cinnyris Falkensteini Fischer & Reichenow, 1884, Journ. f. Orn., 32, p. 56: 
Lake Naivasha, Kenya Colony. 

d" 9 9 (M. C. Z. 168950-2) Kaimosi, K. C. 16 Feb. to 5 March 1934. 

Cinnyris reichenowi reichenowi Sharpe 
Cinnyris reichenowi Sharpe, 1891, Ibis, p. 444: Sotik, Kenya Colony. 

6 c? c? (M. C. Z. 168953-8) Kaimosi, K. C. 25 Feb. and 5 March 1934. 

Chalcomitra amethystina kalckreuthi Cabanis 

Cinnyris {Chalcomitra) kalckreuthi Cabanis, 1878, Journ. f. Orn., 26, pp. 205, 
227: Mombasa, Kenya Colony. 

<? & (M. C. Z. 168959-60) Lamu, Lamu Id., K. C. 11 May 1934. 

Chalcomitra amethystina doggetti (Sharpe) 
Cinnyris doggetti Sharpe, 1902, Ibis, p. 116: Ravine, Kenya Colony. 

<? 9 (M. C. Z. 168961-2) Mt. Debasien, U. 29 November 1933. 

cf (M. C. Z. 168963) Elgonyi, K. C. 31 January 1934. 

c? (M. C. Z. 168964) Kibwezi, K. C. 29 March 1934. 
<? 9 (M. C. Z. 168965-6) Voi, K. C. 10 & 12 April 1934. 

Chalcomitra senegalensis aequatorialis (Reichenow) 

Cinnyris aequatorialis Reichenow, 1899, Orn. Monatsb., 7, p. 171: Bukoba, 
Tanganyika Territory. 

<? (M. C. Z. 168967) Mt. Debasien, U. 20 November 1933. 

Chalcomitra verticalis viridisplendens (Reichenow) 

Cinnyris viridisplendens Reichenow, 1892, Journ. f. Orn., 40, pp. 54, 132: 
Bukoba, Tanganyika Territory. 

d" (M. C. Z. 168969) Kirui, Kitosh, K. C. 20 January 1934. 
c? 9 (M. C. Z. 168970-1) Kaimosi, K. C. 16 Feb. & 1 March 1934. 



196 bulletin: museum of comparative zoology 

Cyanomitra olivacea ragazzii (Salvador!) 

Eleocerthia ragazzii Salvadori, 1888, Ann. Mus. Civ. Genova, 26, p. 247: 
Ferkerie-ghem Forests, Shoa, Ethiopia. 

& (M. C. Z. 168968) Kaimosi, K. C. 5 March 1934. 

Anthreptes collaris ugandae van Someren 

Anthreptes collaris ugandae van Someren, 1921, Bull. Brit. Orn. Club, 41, p. 113: 
Maraquet (i. e. Marakwet), Kenya Colony. 

c? (M. C. Z. 168972) Elgonyi, K. C. 31 January 1934. 
5 tf 1 <? (M. C. Z. 168973-7) Kaimosi, K. C. 1 & 5 March 1934. 



ZOSTEROPIDAE 

Zosterops virens jacksoni Neumann 

Zosterops jacksoni Neumann, 1899, Orn. Monatsb., 7, p. 23: Mau, Kenya 

Colony. 
Zosterops yalensis van Someren, 1922, Novit. Zool., 29, p. 191: Kaimosi, 

Kenya Colony. 
Zosterops elgonensis van Someren, 1922, Novit, Zool., 29, p. 191: Bukedi, 

Mount Elgon, Uganda. 

1 + c?2 9 9 (M. C. Z. 168978-81) Kaburomi, Elgon, U. 28 December 1933. 
2 9 9 (M. C. Z. 168982-3) Elgonyi, Elgon, K. C. 23 January 1934. 
d 1 4 9 9 (M. C. Z. 168984-8) Kaimosi, K. C. 1 & 5 March 1934. 

Synonymy. Sclater (1930, p. 675, footnote) refers yalensis and 
elgonensis to the synonymy of jacksoni. In 1932 van Someren (1932, 
p. 349) defends their validity, even claiming full specific rank for 
yalensis. Later Granvik (1934, p. 134) when studying his series of 
white-eyes from Mount Elgon fails to recognize elgonensis which he 
thinks is a straight synonym of jacksoni. 

It will be observed that our Kaimosi series are topotypes of yale?isis 
while our Elgon birds are from near the type locality of elgonensis. 
The measurements are as follows: — Kaburomi, 58-63 mm., Elgonyi, 
56-59 mm., Kaimosi, 56-58 mm. We purposely omit giving them by 
sex as native sexing is liable to question, except when the native 
skinners are of long service and known to be trustworthy. The 
Kaburomi birds are a trifle larger than those from lower levels yet 
van Someren's measurements for jacksoni are 62-65 mm., for elgonensis 
he gives 56-61, for yalensis 58-62 (thirty skins). 



PETERS AND LOVERIDGE: AFRICAN BIRDS 197 

The Kaburomi birds which were shot in the tree heaths just above 
the forest limits at 10,500 feet are perhaps a trifle paler than those 
from lower levels. There is so much variation shown in the series 
from the various localities that we concur in Sclater's action con- 
sidering that the Zosterops from this region have been subjected to 
far too critical dividing. 

Breeding. At Kaimosi on March 5, 1934, the oviduct of a female 
bird was found to hold a fully-formed white egg though without 
shell, it measured 16 x 11 mm. 

Zosterops virens stuhlmanni Reichenow 

Zosterops stuhlmanni Reichenow, 1892, Journ. f. On., 40, p. 54: Bukoba, 
Tanganyika Territory. 

2 d" 2 9 (M. C. Z. 168989-92) Mt. Debasien, U. 11-30 November 1933. 

Distribution. This is rather a surprising locality for this race but 
after careful comparison with a topotype and series of stuhlmanni 
we find them only a shade less yellowish green above than that form, 
distinctly more yellowish than our Elgon birds referred to jacksoni. 

Zosterops silvanus Peters & Loveridge 

Zosterops silvanus Peters & Loveridge, 1935, Proc. Biol. Soc. Washington, 48, 
p. 77: Mt. Mbololo, 4,800 feet, Taita, Kenya Colony. 

2 tf" & 2 9 9 (M. C. Z. 168993-6) Mt. Mbololo, K. C. 21 April 1934. 

Distribution. Known only from the type series listed above, one of 
which is now in the British Museum (N. H.). The type is a cf 
(M. C. Z. 168994). Closely related to Z. winifredae Sclater & Moreau 
with the type of which one of these has been compared. 



PLOCEIDAE 

DlNEMELLIA DINEMELLI DINEMELLI (Riippell) 

Textor dinemelli Riippell, 1845, Syst. Uebers., p. 72, pi. xxx: Shoa, Ethiopia 
(ex Horsfield mss.). 

c? (M. C. Z. 168997) Tsavo, K. C. 5 April 1934. 

c? (M. C. Z. 168998) Marareni, nr. Malindi, K. C. 26 June 1934. 

Measurements. Van Someren (1932, p. 317) states that birds from 
Taita, near Tsavo, are intermediate between the nominate form and 



198 bulletin: museum of comparative zoology 

bohmi with wings from 115-124 mm. Our Tsavo bird is wholly typical 
with a wing of 109 mm., the Malindi bird is 116 mm. 

Plocepasser mahali melanorhynchus Bonaparte 

Plocepasser melanorhynchus Bonaparte, 1851, Consp. Gen. Av., 1, p. 444: 
Shoa, Ethiopia (ex Ruppell, 1845, Syst. Uebers., p. 78). 

d" (M. C. Z. 168999) Karita River, U. 9 November 1933. 
1 (M. C. Z. 169000) Tsavo, K. C. 4 April 1934. 

Passer griseus gongonensis (Oustalet) 

Pseudostruthus gongonensis Oustalet, 1890, La Naturaliste, p. 274: Gongoni, 
near Mombasa, Kenya Colony. 

9 (M. C. Z. 168473) Lamu, Lamu Id., K. C. 11 May 1934. 

Gymnoris pyrgita massaica Neumann 

Gymnoris pyrgita massaica Neumann, 1908, Bull. Brit. Orn. Club, 21, p. 70: 
Kikuyu, Kenya Colony. 

9 (M. C. Z. 168474) Voi, K. C. 7 April 1934. 

Ploceus insignis insignis (Sharpe) 

Sycobrotus insignis Sharpe, 1891, Ibis, p. 117, pi. vi, fig. 1: Mount Elgon. 

d 1 (M. C. Z. 168475) Elgonyi, K. C. 25 January 1934. 

Corrigenda. In treating of this species, van Someren (1932, p. 319) 
states: "Sclater admits three races, and a race of a race!" and pro- 
ceeds to make deductions from what is most obviously a slip such 
as occur in van Someren's own paper where, for example, he breaks 
up Poliospiza angolensis into races and a race of a race (1932, p. 330). 

Ploceus melanogaster stephanophorus (Sharpe) 

Heterhyphantes stephanophorus Sharpe, 1891, Ibis, p. 117, pi. vi, fig. 2: Mau, 
Kenya Colony. 

cf (M. C. Z. 168476) Kaimosi, K. C. 1 March 1934. 

Ploceus nigricollis nigricollis (Vieillot) 

Malimbus nigricollis Vieillot, 1805, Ois. Chant., p. 74, pi. xlv, Malimba, 
Portuguese Congo. 

<? (M. C. Z. 168477) Kaimosi, K. C. 1 March, 1934. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 199 

Synonymy. We have carefully compared this skin with a large 
series from Cameroon, Gaboon, Angola and the eastern Belgian 
Congo and fail to find any justification for the reviving of the race 
vacillans van Someren (type locality: Budongo, Uganda) as has 
been attempted by van Someren (1932, p. 320). 

Ploceus nigricollis melanoxanthus (Cabanis) 

Hyphanturgus melanoxanthus Cabanis, 1878, Journ. f; Orn., 26, pp. 205, 232: 
Mombasa, Kenya Colony. 

9 (M. C. Z. 168479) Lamu, Lamu Id., K. C. 9 May 1934. 
9 (M. C. Z. 168478) Kitau, Manda Id., K. C. 18 May 1934. 

Ploceus bojeri bojeri (Cabanis) 

Hyphantornis bojeri Cabanis, 1869, in von der Decken's, Reise, 3, Vogel, p. 32: 
Mombasa, Kenya Colony (ex. mss. Finsch & Hartlaub). 

c? (M. C. Z. 168480) Kkonumbi, K. C. 29 May 1934. 
d> (M. C. Z. 168481) Malindi, K. C. 29 June 1934. 

Affinities. Sclater (1930, p. 748) treats bojeri as a race of aureoflavus 
but van Someren (1932, p. 321) has shown that this cannot be. 
Sclater (1930, p. 748, footnote) refers alleni Mearns from Meru 
River, Mount Kenya to the synonymy. Previously van Someren 
(1922, p. 140) thought it barely recognizable as being rather larger 
than the coast birds. A re-examination of the type and a good series 
convinces us that there is no differentiation in size but the form 
appears to be recognizable on the basis of deeper coloration, par- 
ticularly on the top of the head and below. On the other hand, our 
topotypic material of bojeri is very scanty, doubtless both Sclater 
and van Someren had more extensive series on which to base their 
conclusions. 

Breeding. On June 7, 1934, a nest of the usual type but with the 
slightest of tunnel-entrances, was found overhanging the Tana River 
in the vicinity of Ngau. It held two eggs of a sea-green ground color 
mottled all over with gray or gray-brown. 

On June 14, 1934, three nests were found in close proximity, each 
attached to a spray of thorn growing from knee-deep water in a 
flooded area close to the Tana River at Ngatana. Each nest held two 
eggs, two of these clutches were similar in type but the third had a 
ground color of white, or faintly bluish white, so thickly overlaid 
with fawn as to give the eggs the appearance of being almost uniform 



200 bulletin: museum of comparative zoology 

fawn. All three clutches held embryos. Nearby another nest held 
naked young; a fifth was lying on the water, waterlogged except for 
its entrance in which crouched a large, though featherless, nestling. 
I restored it to a vertical position before leaving it. 

Euplectes nigroventris Cassin 

Euplectes nigroventris Cassin, 1848, Proc. Acad. Nat. Sci. Phila., p. 66: Zan- 
zibar. 

Breeding. On June 2, 1934, a nest of the Zanzibar Red Bishop 
was found near Witu. The nest, which was attached to reeds growing 
in a pond, was some three feet above the level of the water. It mea- 
sured about 150 mm. high by 50 mm. in diameter being of the usual 
domed type with side entrance and composed of fine grasses. It held 
three fresh eggs, measuring 18 x 13 mm., of a pale greenish blue 
ground color but while two were only very sparsely and minutely 
dotted with black, the third was well blotched all over with gray or 
black. 

On June 19, 1934, at Wema, Ngatana, two nests were found con- 
taining young. 

Euplectes capensis xanthomelas Riippell 

Euplectes xanthomelas Riippell, 1840, Neue Wirbelth., Vog., p. 94: Temben and 
Simen, Ethiopia. 

9 (M. C. Z. 168482) Kibwezi, K. C. 24 March 1934. 

This bird was captured as it fluttered on the ground, its feathers 
being so gummed together by sticky seeds that it is doubtful whether 
it would ever have got them freed. 

Uraeginthus bengalus brunneigularis Mearns 

Uraeginthus bengalus brunneigularis Mearns, 1911, Smiths. Misc. Coll., 56, 
No. 20, p. 6: Wambugu, Mt. Kenya, Kenya Colony. 

9 (M. C. Z. 168490) Kibwezi, K. C. 24 March 1934. 

Uraeginthus bengalus ugogoensis Reichenow 

Uraeginthus bengalus var. ugogoensis Reichenow, 1911, Mitt. Zool. Mus. Berlin, 

5, p. 228: Ugogo, Tanganyika Territory. 
Uraeginthus bengalus littoralis van Someren, 1922, Novit. Zool., 29, p. 160: 

Mombasa, Kenya Colony. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 201 

Uraeginthus bengalus loveni Granvik, 1923, Journ. f. Orn.,71, p. 181: Mombasa, 
Kenya Colony. 

c? cf (M. C. Z. 168491-92) Kitau, Manda Id., K. C. 16-17 May 1934. 

Synonymy. We have compared these birds with topotypes of both 
ugogoensis and littoralis and cannot see that van Someren (1932, 
p. 327) has any grounds for reversing Sclater's (1930, p. 804, foot- 
note) decision as to the disposition of the latter in the synonymy of 
the former. 

Uraeginthus bengalus ugandae Zedlitz 

Uraeginthus bengalus ugandae Zedlitz, 1911, Journ. f. Orn., 59, p. 606: Entebbe, 
Uganda. 

cf (M. C. Z. 168493) Kaimpsi, K. C. 24 March 1934. 

Previously recorded from this locality by van Someren (1922, p. 
160). 

Granatin a. ianthinogaster ianthinogaster (Reichenow) 

Uraeginthus ianthinogaster Reichenow, 1879, Orn. Centralbl., p. 114: Massa, 
Tana River, Kenya Colony. 

9 (M. C. Z. 168494) Kibweai, K. C. 26 March 1934. 

Coliuspasser ardens suahelica (van Someren) 

Penthretia laticauda suahelica van Someren, 1921, Bull. Brit. Orn. Club, 41, 
p. 121: Nairobi River, Kenya Colony. 

cf (M. C. Z. 168483) Mt. Debasien, U. 21 November 1933. 

This Kenya Red-naped Whydah is in fresh off-season plumage, 
whether correctly sexed or not. 

Spermophaga ruficapilla ruficapilla (Shelley) 

Spermospiza ruficapilla Shelley, 1888, Proc. Zool. Soc. London, p. 30: Bellima, 
Uele district, Belgian Congo. 

cf (M. C. Z. 168484) Kaimosi, K. C. 26 February 1934. 

Pytilia melba kirki Shelley 

Pytelia kirki Shelley, 1903, Bull. Brit. Orn. Club, 13, p. 76: Lamu, Kenya 

Colony. 

cf 9 (M. C. Z. 168485-6) Tsavo, K. C. 4 April 1934. 

9 (M. C. Z. 168487) Kitau, Manda Id., K. C. 17 May 1934. 



202 bulletin: museum of comparative zoology 

Affinities. Van Someren (1932, p. 325) considers Tsavo birds are 
intermediate between kirki and belli and this is also the case with 
our skins from that locality. The male having red lores but less 
barring on the abdomen. 

We use the name kirki, which Sclater (1930, p. 788, footnote) 
refers to the synonymy of soudancnsis, as our Kitau bird is prac- 
tically topotypic of kirki while the type locality of soudanensis is 
uncertain, and the status of the forms still somewhat unsettled. 

Estrilda astrild nyanzae Neumann 

Estrilda astrild nyanzae Neumann, 1907, Journ. f. Orn., 55, p. 596: Bukoba, 
Tanganyika Territory. 

Juv. (M. C. Z. 168488) Butandiga, U. 13 January 1934. 

Estrilda nonnula nonnula (Hartlaub) 

Astrilda nonnula Hartlaub, 1883, Journ. f. Orn., 31, p. 425: Kudurma, Bahr el 
Ghazal, Sudan. 

9 juv. (M. C. Z. 168489) Butandiga, U. 11 January 1934. 



FRINGILLIDAE 

Poliospiza angolensis reichenowi (Salvadori) 

Serinus reichenowi Salvadori, 1888, Ann. Mus. Civ. Genova, 26, p. 272: 
Cialalaka, Shoa, Ethiopia. 

9 (M. C. Z. 168495) Voi, K. C. 7 April 1934. 

Measurements. Wing 62 mm., compared with examples from Alaba, 
Guaso Nyiro and Kapiti Plains which measure from 67 to 69 mm. 

Poliospiza striolata ugandae van Someren 

Poliospiza striolata ugandae van Someren, 1921, Bull. Brit. Orn. Club, 41, 
p. 114: Mount Elgon. 

9 (M. C. Z. 168496) Kaburomi, U. 30 December 1933. 
9 (M. C. Z. 168497) Madangi, U. 3 January 1934. 

Both these birds being from Mount Elgon, they may be considered 
topo typical. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 203 

Emberiza flaviventtris flaviventris Stephens 

Plate 2, fig. 1. 

Emberiza flaviventris Stephens, 1815, Gen. Zool., 9, part 2, p. 374: Cape of 
Good Hope. 

Nest, eggs, 9 (M. C. Z. 168498) Mt. Debasien. U. 25 November 1933. 
c? 9 (M. C. Z. 168499-168500) Tsavo, K. C. 4 April 1934. 

Breeding. On November 20, 1933, a nest of the Golden-breasted 
Bunting was found in the fork of a thorn-tree at a height of about 
five feet from the ground. The nest, measuring 60 mm. in depth by 
100 mm. in outside diameter, was 30 mm. in depth by 55 mm. in 
diameter inside. Externally it was composed of shredded grasses, 
tendrils and rootlets, inside it was neatly lined with hair. When 
found it held a single egg, on revisiting the nest on November 25, 
I found only two eggs, these showed slight traces of incubation and 
measured 20 x 15 mm. Their ground color was faintty bluish white 
on which was a circumpolar ring of faint purple and black scribblings, 
elsewhere only a few scattered dots. The female was shot. 



204 bulletin: museum of comparative zoology 

BIBLIOGRAPHY 

Bangs, O. and Loveridge, A. 

1933. "Reports on the Scientific Results of an Expedition to the South- 
western Highlands of Tanganyika Territory. Part III. Birds." 
Bull. Mus. Comp. Zool., 75, pp. 143-221, pi. i. 

Bannermann, D. A. 

1930. "The Birds of Tropical West Africa." 1, pp. i-lxxv, 1-376, figs. 
1-119, pis. i-x. London. 

Friedmann, Herbert 

1928. "A Collection of Birds from the Uluguru and Usambara Moun- 
tains, Tanganyika Territory." Ibis, pp. 74-99. 

1930. "Birds collected by the Childs Frick Expedition to Ethiopia and 
Kenya Colony." Bull., U. S. Nat. Mus. no. 153, pp. i-xiii, 1-516, 
figs. 1-22, pis. i-xii. 

Grant, C. H. B. and Mackworth-Praed, C. W. 

1933. "On the Races and Distribution of the African and Arabian Kes- 
trels of the Falco tinnunculus group with descriptions of two new 
races." Bull. Brit. Orn. Club, 54, pp. 75-83. 

1934. "On the races of the Lizard-Buzzard, Kaupifalco monogrammicus 
Temm." Bull. Brit. Orn. Club, 54, pp. 130-131. 

1934. "Eastern races of Francolinus sephaena (Smith)." Bull. Brit. 
Orn. Club, 54, pp. 170-173. 

Granvik, Hugo 

1923. "Birds collected by the Swedish Mount Elgon Expedition 1920." 

Journ. f. Orn., 71, Sonderheft, pp. 1-280, pis. i-xi. 
1934. "The Ornithology of North Western Kenya Colony with special 

regard to the Suk and Turkana Distrikt." Rev. Zool. Bot. Afr., 

25, pp. 1-190, pis. i-iv. 

Grote, Hermann 

1922. "Bermerkungen tiber einige neue afrikanische Formen." Journ. 
f. Orn., 70, pp. 397-404. 

1931. "Die Gliederung des Formenkreises Treron australis (L.)." Anz. 
Orn. Ges. Bayern, 2, pp. 140-141. 

Peters, James L. 

1931. "Check-List of Birds of the World." 1, pp. i-xviii, 1-345. Cam- 
bridge, Mass. 

1934. "Check-List of Birds of the World." 2, pp. i-xvii, 1-401. Cam- 
bridge, Mass. 

Peters, James L. and Loveridge, A. 

1935. "New Birds from Kenya Colony." Proc. Biol. Soc. Washington, 
48, pp. 77-78. 



PETERS AND LOVERIDGE: AFRICAN BIRDS 205 

Roberts, Austin 

1926. "Some Changes in Nomenclature, New Records of Migrants and 
New Forms of S. African Birds." Ann. Transv. Mus., 2, pp. 217- 
225. 

SCLATER, W. L. 

1924. "Systema Avium Aethiopicarum." Part I, i-iv, 1-304. London. 
1930. "Systema Avium Aethiopicarum." Part II, i-xi, 305-922. London. 

Sclater, W. L. and Moreau, R. E. 

1932 "Taxonomic and Field Notes on some Birds of North-eastern 
and Tanganyika Territory." Parts I-V. Ibis, 1932, pp. 487-522, 

1933. pis. vi-vii; pp. 656-683. 1933, pp. 1-33; pp. 187-219, pi. vi; pp. 
399-440; 

Someren, V. G. L. van 

1916. "List of Birds collected in Uganda and British East Africa, with 

notes on their nesting and other habits. "Parts I— i.1. Ibis, pp. 

193-252, pis. iv-vi; pp. 373-472, pis. viii-xiii. 
1918. "Notes on a collection of Birds from Lamu and district, made by 

Mr. H. J. Allen Turner in April 1916." Journ. E. A. &Uganda 

Nat. Hist. Soc, no. 6, pp. 251-261. 
1918. "A further con ribution to the Ornithology of Uganda (West 

Elgon and district)." Novit. Zool., 25, pp. 263-290. 
1922. "Notes on the Birds of East Africa." Novit. Zool., 29, pp. 1-246, 

pis. i-vi. 

1925. "The Birds of Kenya and Uganda." Part II. Journ. E. A. & 
Uganda Nat. Hist. Soc, No. 23, pp. 95-104. 

1926. "The Birds of Kenya and Uganda." Part III. Journ. E. A. & 
Uganda Nat. Hist. Soc, No. 25, pp. 29-60. 

1930. "The Birds of Kenya and Uganda." Part VIII. Journ. E. A. & 
Uganda Nat. Hist. Soc, Nos. 38-39, pp. 34-65, pis. 

1931. "Catalogue of the European and Asiatic Migrants to Kenya and 
Uganda with brief outline of the subject of Migration of Birds." 
Spec. Supp. No. 4 to Journ. E. A. & Uganda Nat. Hist. Soc. 
October, 1931, pp. 1-40. 

1932. "Birds of Kenya and Uganda, being addenda and corrigenda to 
my previous paper in "Novitates Zoologicae", XXIX, 1922." 
Novit. Zool., 32, pp. 252-380, pis. i-iv. 

Wetmore, Alexander 

1930. "A Systematic Classification for the Birds of the World." Proc 
U. S. Nat. Mus., 76, Art. 24, pp. 1-8. 

1934. "A Systematic Classification for the Birds of the World, Revised 
and Amended." Smiths. Misc. Coll., 89, No. 13, p. 1-11. 



EXPLANATION OF PLATES 



PLATE 1 



Peters and Loveridge — African Birds. 



PLATE 1 
Fig. 1. Forest at Sipi, Mount Elgon, Uganda. 

Gunbearer ascending to hole from whence a Gelo River Crowned Hornbill 
(Lophoceros melanoleucos geloensis) had flown. December 21, 1933. Photo 
M. V. Loveridge. 

Fig. 2. Weaver (Ploceus cucullatus feminina) colony. 

These nests were in an acacia overhanging a stagnant pool at Kananyait, 
western foot of Mount Debasien, Karamoja. December 3, 1933. Photo 
M. V. Loveridge. 



BULL. MUS. COMP. Z03L. 



Peters and Loveridoe. African Birds. Plate 1 . 





PLATE 2 



Petebs and Loveridge — African Birds. 



PLATE 2 

Fig. 1. Nest and eggs of Golden-breasted Bunting. 

This bird (Emberiza flaviventris flaviventris) had built among the four-inch 
thorns of a dry acacia, Mount Debasien. November 25, 1933. Photo M. V. 
Loveridge. 

Fig. 2. Mount Debasien from the west. 

This mountain, whose rocky summit attains an altitude of about 10,050 
feet, is situated in the vast dessicating region of Karamoja. Forest-loving 
birds are only to be found along the river beds and in the small remnant of 
forest nestling against the rocky heights. November, 1933. Photo M. V. Lov- 
eridge. 



BULL. MUS. COMP. ZOOL. 



Peters and Loveridge. African Birds. Plate 2. 









& ' ' 



H-CL 



Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 

Vol. LXXIX, No. 5 



SCIENTIFIC RESULTS OF AN EXPEDITION 
TO RAIN FOREST REGIONS IN EASTERN AFRICA 

V 
REPTILES 



By Arthur Loveridge 



With Nine Plates 



CAMBRIDGE, MASS., U.S.A. 
PRINTEDFORT HE MUSEUM 
November,, 1936 \ 



No. 5. — Reports on the Scientific Results of an Expedition 
to Rain Forest Regions in Eastern Africa 

V 

Reptiles 
By Arthur Loveridge 

CONTENTS Page 
Introduction 

Material 209 

Acknowledgements 210 

Additions to the Fauna 211 

Summary of taxonomic alterations 212 

List and index to species collected 213 

Systematic Discussion 

Crocodiles 217 

Chelonians 218 

Snakes 225 

Lizards 281 

Chameleons 329 

Bibliography 336 

INTRODUCTION 

The collection on which this report is based was made by the 
author, as a Fellow of the John Simon Guggenheim Foundation, with 
a view to elucidating the present-day distribution of the montane, 
sylvicoline fauna of certain mountains in eastern Uganda and Kenya 
Colony. 

This subject will be dealt with in the concluding contribution to 
the series of reports, in a paper which will also contain the itinerary 
and full information regarding localities and altitudes. 

The period of collecting was from November 8, 1933, to July 9, 1934, 
during which time 2,280 reptiles representing 123 species were pre- 
served. This total, which excludes certain species received as a gift, 
comprises 1 species of crocodile, 5 of tortoises and turtles, 57 of snakes, 
50 of lizards and 10 forms of chameleons. In all 18 of these and 1 genus 
were new to the collections of the Museum of Comparative Zoology, 



210 bulletin: museum of comparative zoology 

but a dozen others were previously represented by only 1 or 2 examples 
sometimes with poor data. 

One might single out for special mention such rarities as : Leptotyph- 
lops boulengeri, Mehelya nyassae, Chamaetortus a. aulicus, Rhamphio- 
phis rubropunctatus, Micrelaps bicoloratus, Bunocnemis modestus, 
Lygodactylus /. scheffleri, Geocalamus acutus, Eremias neumanni, 
Mabuya irregularis, Chameleon tavetensis. Five of these were known 
previously only from the type. In all, topotypes of more than a score 
of species were collected. 

Originally I had planned to visit the Sokoki Forest, on the coast 
south of Malindi. As, however, that veteran collector and well-known 
naturalist, Mr. H. J. Allen Turner, camped there in June, 1932, I 
availed myself of his typically generous offer to work over the herpeto- 
logical material (now in the Coryndon Memorial Museum, Nairobi) 
which he had secured. The resulting locality records are included 
under the heading "Distribution," but his specimens do not figure in 
the statistics given above. 

ACKNOWLEDGMENTS 

I should like to take this opportunity to express my gratitude to 
the John Simon Guggenheim Memorial Foundation for the grant which 
made the expedition possible. Also to Dr. Thomas Barbour, of the 
Museum of Comparative Zoology, for the interest and encouragement 
he has given in furthering the prosecution of this work. 

As before, I am indebted to Doctors Joseph Bequaert and J. H. 
Sandground of the Harvard School of Tropical Medicine for identi- 
fying the ecto- and endoparasites recorded in the following pages. 
My entomological colleague, Mr. Nathan Banks, has been most kind 
in determining the more difficult remains of insects which figured in 
the stomach contents of these reptiles. 

Many herpetologists and others have aided by answering questions, 
making comparisons with types in their care, and in other ways. Due 
appreciation is expressed in the text, but I may mention here that 
Messrs. E. R. Dunn, Malcolm A. Smith, H. W. Parker, C. R. S. Pit- 
man, K. P. Schmidt, O. G. Stull-Davies and V. G. L. van Someren have 
all assisted at one time or another during the writing up of these notes. 

Lastly, but by no means least, I am grateful to my wife for taking 
all the photographs used to illustrate this paper. Many were taken 
under trying conditions of heat, tropical glare, and often with rest- 
less reptilian subjects. 



loveridge: African reptiles 211 



ADDITIONS TO THE FAUNA 

As a result of this study, the following species, or races, collected on 
the expedition, have been described for the first time: 

Testudo pardalis babcocki, Mount Debasien, Uganda. 

Typhlops kaimosae, Kaimosi, Kenya Colony. 

Coronella semiornata fuscorosea, Mount Mbololo, Kenya Colony. 

Aparallactus turneri, Sokoki Forest, Kenya Colony. 

Dendraspis Jameson kaimosae, Kaimosi, Kenya Colony. 

Cnemaspis africanus elgonensis, Sipi, Mount Elgon, Uganda. 

Hemidactylus mandensis Kitau, Manda Id., Kenya Colony. 

Lygodactylus picluratus mombasicus, Kilindini, Mombasa Id., Kenya Colony. 

Agama agama kaimosae (see below), Near Kaimosi, Kenya Colony. 

Riopa mabuiiformis, Ngatana, Tana River, Kenya Colony. 

Riopa tanae, Kau, Tana River, Kenya Colony. 

Acontias percivali, Foot of Mount Mbololo, Kenya Colony. 

Chamaeleon bitaeniatus altaeelgonis, Kaburomi, Mt. Elgon, Uganda. 

In addition to these new forms, the undermentioned are recorded 
from Uganda or Kenya Colony for the first time: 

New for Uganda 

Chlorophis carinatus Andersson, of Cameroons. 
Cnemaspis quattuorseriatus (Sternfeld), of Ruanda. 
Algiroides alleni Barbour, of Kenya Colony. 
Eremias spekii sextaeniata Stejneger, of Kenya Colony. 
Mabuya irregularis Lonnberg, of Kenya Colony. 

New for Kenya Colony 

Kinixys spekii Gray, of "Central Africa." 

Typhlops pallidus (Cope), of Zanzibar. 

Leptotyphlops longicauda (Peters), of Mozambique. 

Mehelya nyassae (Giinther), of Nyasaland. 

Chlorophis carinatus Anderson, of Cameroons. 

Coronella coronata (Schlegel), of the Gold Coast. 

Prosy mna ambigua stuhlmanni (Pfeffer), of Tanganyika Territory. 



212 bulletin: museum of comparative zoology 

Aparallactus uluguruensis Barbour and Loveridge, of Tanganyika Territory. 
Zonurus tropidosternum Cope, of Tanganyika Territory. 
Eremias neumanni Tornier, of Ethiopia. 

The genera Prosymna and Zonurus have their ranges extended north- 
wards into Kenya, though the former is known to occur in British 
Somaliland. 



SUMMARY OF TAXONOMIC ALTERATIONS 

The following subspecies are revived from the synonymy of their 
respective species: 

Prosymna ambigua stuhlmanni, (Pfeffer). 
Lygodactylus fischeri scheffleri, Sternfeld. 
Mabuya quinquetaeniata obsti, Werner. 

Parker was correct in treating Riopa modestum as a full species, and 
I was in error in assuming it to be a race of E. sundevallii (Smith). 
Chamaeleon tavetensis Steindachner becomes a race of fischeri. 

The undermentioned are considered to be synonyms : 

Kinixys jordani Hewitt = Kinixys spekii Gray 

Pelusios sinualus zuluensis Hewitt =Pelusios sinuatus (Smith) 

Pelusios sinuatus leptus Hewitt = Pelusios sinuatus (Smith) 

Typhlops boulengeri Bocage = Typhlops p. punctatus (Leach) 

Glauconia braueri Sternfeld = Typhlops braminus (Daudin) 

Meizodon regularis Fischer = Coronella coronata (Schlegel) 

Coronella regularis praeornata Angel =Coronella coronata (Schlegel) 
Coronella semiornata mossambicae Cott = Coronella s. semiornata Peters 

Scaphiophis calciatii Calabresi =Scaphiophis raffreyi Bocourt 

Rhinocalamus meleagris Sternfeld = Micrelaps bicoloratus Sternfeld 
Aparallactus concolor boulengeri 

Scortecci =A. uluguruensis Barb. & Love. 

Dendraspis sjostedtii Lonnberg = Dendraspis angusticeps (Smith) 

Agama agama turuensis Loveridge =Agama agama elgonis Lonnberg 

Agama agama kaimosae Loveridge = Agama planiceps caudospinosa Meek 

Zonurus parkeri Cott = Zonurus tropidosternum Cope 

Euprepes (Euprepis) taitanus Peters = Mabuya planifrons (Peters) 

Ablepharus carsonii Boulenger = Ablepharus wahlbergii (Smith) 



loveridge: African reptiles 213 



LIST OF SPECIES COLLECTED* 

CROCODYLIDAE Page 

Crocodylus niloticus Laurenti 217 

TESTUDINIDAE 

Kinixys spekii Gray 218 

Testudo ■partialis babcocki Loveridge 220 

{Tcstudo tornieri Siebenrock) 221 

CHELONIIDAE 

Chelonia mydas (Linnaeus) 221 

PELOMEDUSIDAE 

Pelusios sinuatus (Smith) 222 

Pelusios nigricans nigricans (Dondorff) 223 

Pelomedusa galeata (SchoepfT) 225 

TYPHLOPIDAE 

(Typhlops jmnctatus punctatus (Leach) ) 225 

Typhlops kaimosae Loveridge 226 

Typhlops schlegelii mucruso (Peters) 226 

Typhlops uniiaeniatus unitaeniatus Peters 227 

Typhlops pallidas (Cope) 227 

(Typhlops braminus (Daudin)) 228 

Typhlops lumbriciformis (Peters) 228 

LEPTOTYPHLOPIDAE 

Lcptotyphlops boulengeri (Boettger) 230 

Leptotyphlops longicauda (Peters) 231 

Lcptotyphlops conjuncta (Jan) 232 

BOIDAE 

Python sebae (Gmelin) 232 

Eryx colubriniis loveridgei Stull 233 

(Eryx colubriniis rufescens AM) 235 

COLUBRIDAE 

Natrix olivacea olivacea (Peters) 236 

Boaedon lineatus Dumeril & Bibron 237 

(Boaedon olivaceus (Dumeril) ) 240 

Lycophidion capense capense (Smith) 241 

Lycophidion capense acutirostre Giinther 242 

*Species in parentheses are discussed, though not collected. 



214 bulletin: museum of comparative zoology 

COLUBRIDAE continued Page 

Mehelya ?iyassae (Gunther) 243 

Chlorophis carinatus Andersson 243 

Chlorophis hoplogaster (Gunther) 245 

Chlorophis ncglectus (Peters) 246 

Chlorophis irregularis (Peters) 247 

Philothamnus semivariegatus semivariegatus Smith 247 

Hapsidophrys lineata Fischer 249 

Thrasops jacksonii jacksonii Gunther 249 

Coronclla semiornata semiornata Peters 250 

Coronella semiornata fmcorosea Loveridge 252 

Coronella coronata (Schlegel) 253 

Prosymna ambigua stuhlmanni (Pfeffer) 254 

(Scaphiophis albopunctatus Peters) 255 

DASYPELTINAE 

Dasypeltis scaber Linnaeus 256 

BOIGINAE 

Crotaphopcltis hotamboeia hotamboeia (Laurenti) 257 

Chamactortus aulicus aulicus Gunther 259 

Hemirhagerrhis kelleri Boettger 260 

Rhamphiophis rubropunctatus (Fischer) 261 

Rhamphiophis rostratus Peters 262 

Psamviophis sibilans (Linnaeus) 262 

Psammophis subtaeniatus Peters 263 

Psammophis punctulatus Dumeril & Bibron 264 

Psamviophis biseriatus Peters 265 

Thelotornis kirtlandii (Hallowell) 265 

Dispholidus typus (Smith) 266 

Calamelaps unicolor (Reinhardt) 266 

Micrelaps bicoloratus Sternfeld 267 

AparaUactus turneri Loveridge 268 

Aparallactus capensis Smith 268 

AparaUactus concolor (Fischer) 269 

AparaUactus uluguruensis Barbour & Loveridge 270 

ELAPINAE 

Elapsoidea guntherii Bocage 271 

Naja melanoleuca Hallowell 271 

Naja nigricollis nigricollis Reinhardt 272 

Dendraspis jamesoni kaimosae Loveridge 273 

Dendraspis angusticeps (Smith) 273 



loveridge: African reptiles 215 

VIPERIDAE Page 

Causus resimus (Peters) 277 

Causus dcfilippii (Jan) 278 

Causus lichtensteinii (Jan) 278 

Bitis arietans (Merrem) 278 

Bitis nasicornis (Shaw) 279 

Atheris squamigcra (Hallowell) 280 

Atractaspis bibronii Smith 280 

Atractaspis microlcpidota Giinther 281 

GEKKONIDAE 

Cnemaspis africanus africanus (Werner) 281 

Cnemaspis africaiius elgonensis Loveridge 282 

Cnemaspis quattuorseriatus (Sternfeld) 283 

Hemidactylus brookii Gray 284 

Hemidactylus mandanus Loveridge 285 

Hemidactylus mabouia (Moreau de Jonnes) 285 

Hemidactylus persimilis Barbour & Loveridge 286 

Hemidactylus frenatus Dumeril & Bibron 286 

Hemidactylus werneri werneri Tornier 286 

Hemidactylus tropidolepis squamulatus Tornier 287 

Bwwcnemis modestus Giinther 287 

Lygodactylus fischeri schcffleri Sternfeld 288 

Lygodactylus picturatus picturatus (Peters) 289 

Lygodactylus picturatus mombasicus Loveridge 289 

Lygodactylus picturatus gutturalis (Bocage) 290 

AGAMIDAE 

Agama rueppelli septentrionalis Parker 291 

Agama agama agama (Linnaeus) 291 

Agama agama elgonis Lonnberg 292 

Agama agama lionotus Boulenger 293 

Agama planiceps caudospinosa Meek 294 

Agama atricollis Smith 295 

ZONURIDAE 

{Zonurus tropidostemum Cope) 296 

Chamacsaura tenuior Giinther 297 

VARANIDAE 

Varanus ocellatus Riippell 298 

Varanus niloticus (Linnaeus) 299 

AMPHISBAENIDAE 

Geocalamus acutus Sternfeld 300 



216 bulletin: museum of comparative zoology 

LACERTIDAE Page 

Laccrta jackson i Boulenger 300 

Algiroides alleni Barbour 302 

Latastia longicaudata rcvoili (Vaillant) 304 

Eremias neumanni Tornier 305 

Eremias speJcii spekii Gunther 306 

Eremias spekii sextaeniata Stejneger 307 

GERRHOSAURIDAE 

Gerrhosaurus major major Dumeril 308 

Gerrhosaurus flavigidaris fiavigularis Wiegmann 309 

SCINCIDAE 

Mabuya maculilabris (Gray) 310 

Mabuya planifrons (Peters) 311 

Mabuya brevicollis (Wiegmann) 313 

Mabuya megalura (Peters) 315 

Mabuya quinquetaeniata obsti Werner 315 

Mabuya varia varia (Peters) 316 

Mabuya striata (Peters) 318 

Mabuya irregularis Lonnberg 319 

Riopa mabuiiformis Loveridge 320 

Riopa tanae Loveridge 320 

Riopa sundevallii (Smith) 321 

Riopa modestum modestum (Gunther) 322 

(Riopa pembanum (Boettger) ) 323 

Riopa anchietae (Bocage) 323 

Siaphos kilimensis (Stejneger) 324 

Ablepharus boutonii africanus Sternfeld 326 

Ablepharus icahlbcrgii (Smith) 327 

Acontias percivali Loveridge 328 

CHAMAELEONTIDAE 

Chamacleon senegalensis Daudin 329 

Chamaeleou gracilis gracilis Hallowell 329 

Chamaeleon dilepis ropcri Boulenger 330 

Chamaeleou dilepis quilensis Bocage 331 

Chamaeleoit dilepis dilepis Leach 331 

Chamaeleon bitaeniatus bitaeniatus Fischer 333 

Chamaeleon bitaeniatus hohnelii Steindachner 333 

Chamaeleon bitaeniatus altaeelgonis Loveridge 334 

Chamaeleon fischeri tavetensis Steindachner 335 

Brookesia kerstenii kerstenii (Peters) 335 



LOVERIDGE: AFRICAN REPTILES 217 

Systematic List of Species Collected 

CROCODYLIDAE 

Crocodylus niloticus Laurenti 

Crocodylus niloticus Laurenti (part), 1768, Syn. Rept., p. 53: "Habitat in 
India orientali, et Aegypto." 

2 (M. C. Z. 40001-2) between Kau and Kipini, K. C. 7.V.34. 

Native name. Ngwena (Kipokomo). 

Measurements. Snout to anus 550 and 510 mm., tail 565 and 540 
mm., hind limb 207 and 202 mm., weight 10 and 8 lbs. A third ex- 
ample weighed 13 lbs. 

Diet. These young crocodiles had been kept in captivity for some 
years and were fed upon fish and meat in a tank at Lamu. I was asked 
to shoot them as they were snapping at their native custodian who had 
become afraid of them. A shot with a .22 bullet at close quarters killed 
them instantly and did not injure the slate bottom of their tank. 

The Wapokomo told me that when the Tana is in flood the croco- 
diles get plenty of fish and consequently do not carry off so many 
humans as at other seasons. Nevertheless, just before my arrival at 
Ngau, a native, who had been standing six feet from the water's edge 
up-river, was seized by the leg by a crocodile which dashed out of the 
water and dragged him back into the river. Fortunately for the man, 
he retained hold of his spear, and by stabbing the reptile three or four 
times persuaded it to let go its hold. The man, with his leg badly 
mangled, was still in the mission hospital at Ngau. I was also told 
that the previous year a crocodile had risen from the water and 
snapped at the arm of an Englishman who w T as descending the river 
in a dugout; the reptile only succeeded in tearing away the man's 
sleeve. 

Parasites. None. 

Enemies. An Mpokomo informed me that when the river subsides 
his people organize hunts and spear many scores of crocodiles in the 
course of a day, usually a few monsters but the majority of small 
size. In former years a Mpokomo youth had to kill a crocodile before 
he could marry. They claim to eat all of the reptiles except the bones 
and skin of the back. I expressed doubts as to their being able to eat 
the skin of the belly, but was told that after boiling it for a long time 
they chewed off pieces and swallowed them. 

Six vultures {Necrosyrtes m. monachus) gathered about the carcasses 



218 bulletin: museum of comparative zoology 

of the three crocodiles skinned at Lamu, but for some unfathomed 
reason refused to touch them. The boys suggested that it was because 
they smelled like fish. 

TESTUDINIDAE 

Kinixys spekii Gray 

Kinixys Spekii Gray, 1863, Ann. Mag. Nat. Hist. (3), 12, p. 381: Central 

Africa (i.e. Tanganyika Territory east of the lakes). 
Kinixys j or dani Hewitt, 1931, Ann. Natal Mus., 6, p. 482, pi. xxxvii, figs. 7-9 

(not figs. 1-3 as stated in text) : Isoka, Northern Rhodesia. 

& & Young (M. C. Z. 40007-8) Kibwezi, K. C. 23-28.iii.34. 
2 tf 1 6 9 (M. C. Z. 40009-16) Voi, K. C. 9-13.iv.34. 
<? (M. C. Z. 40017) Golbanti, K. C. 22.vi.34. 

Native names. Nguru (Kisagalla and Kitaita) ; fudi (Kipokomo). 

Variation. We have in the Museum of Comparative Zoology a cf 
and 9 box tortoise from Simbo a few miles north of Tabora (Un- 
yanyembe), Unyamwezi, Tanganyika Territory. These may well be 
considered as topotypes of spekii coming as they do from a village on 
the probable route which Speke took when he first descried Lake Victoria. 

When encountered, I was struck by their depressed appearance and 
planned describing them as distinct from belliana with which I was 
familiar. I took the precaution, however, of first submitting them to 
the British Museum. I was then informed by the late Miss J. Procter 
that they were within the range of variation of belliana to the synony- 
my of which Boulenger had already referred the injured type of 
spekii. I therefore discussed them (Loveridge, 1923, p. 924) under the 
name of belliana as a depressed form inhabiting the arid thorn-bush 
country as opposed to belliana with a more vaulted carapace living in 
the grassy steppes. 

More recently Hewitt (1931, loc. cit.) described five new members 
of the genus from South Africa. Two as races of belliana, three as 
species of the depressed (i.e. spekii) group. He very kindly sent me a 
paratype of jordani which so closely resembles the c? topotype of 
spekii that I have no hesitation in referring it to the synonymy of 
that form. Hewitt mentions a specimen from "Gatta Plain, British 
East Africa" (i.e. Yatta Plain, Kenya Colony), which he says resembles 
jordani and australis, but differs in certain ways which he enumerates. 
His later (1935, p. 347) remarks are rendered nugatory by the second 
Simbo specimen. 



loveridge: African reptiles 219 

First let us reduce the vague terms "slightly flattened," "more 
depressed" to actual percentages of greatest height contained in 
greatest length. In most of the series it was possible to take these 
by placing the tortoise between two blocks. Arranging our material 
from south to north we find : — 

M. C. Z. 33450 paratype shell of jordani. Height 2.6 times in the length 

M. C. Z. 18154-5 &, 9 topotypes of spekii. " 2.4 

M. C. Z. 40009-16 being 2 cf , 6 9 from Voi. " 2.3-2.5 

M. C. Z. 40017 a young <? from Golbanti. " 2.2 

M. C. Z. 40007-8 young and <? from Kibwezi." 2.2-2.4 

M. C. Z. 8158 a juvenile 9 from Ithanga Hills." 2.0 

Sex has no bearing on this variation in height in relation to length 
but age is very obviously a factor for all the four instances of 2 to 
2.2 times the data is based on juveniles which are under 100 mm. in 
length while all the rest of our series are 100 mm. or over and furnish 
a ratio of 2.3 to 2.6 times for height included in length. 

The width of the nuchal may be included in the width of the 
adjacent marginal from 3^ (M.C.Z. 40010) to 7 (M.C.Z. 40015) 
times in the Voi series alone. 

Great extremes of variation in the amount of projection of the 
gulars in females is shown. 

Marginal VI is narrowly in contact with the inguinal in the para- 
type of jordani (M.C.Z. 33450), topotypes of spekii (M.C.Z. 18154-5), 
Ithanga Hills specimen (M.C.Z. 8158) but only in one of the Voi series 
(M.C.Z. 40014). In all others it is broadly in contact. 

The anal suture may be only half the length of the femoral suture 
(M.C.Z. 40017) or three times as long (M.C.Z. 40013). In the topo- 
typic pair of spekii alone it is much longer or much shorter! 

My conclusion, therefore, is that the characters on which the 
alleged species jordani, australis and youngi were based are subject 
to infinite variation. I definitely refer jordani to the synonymy of 
spekii and imagine that when a good series of topotypic australis 
and youngi are available it will be found that they cannot be main- 
tained as distinct. 

Coloration. The plastrons of the spekii topotypes as well as that of 
the jordani paratype exhibit well-defined hollow squares of black 
pigmentation corresponding to the contours of the shields, evidently 
of little consequence as differing somewhat from the types of jordani 
as described by Hewitt. In the Kibwezi, Voi and Golbanti tortoises 
this definite marking is obsolescent though indicated by fragmentary 



220 bulletin: museum of comparative zoology 

markings of a deep black tone. The juvenile female from the Ithanga 
Hills has traces of a stellar pattern in the black markings of its plas- 
tron. The rich pattern of the carapace is so variable in the females, 
often only dull olive in the males, that it can serve no useful purpose 
to describe it here. 

Measurements. Largest cf (M.C.Z. 40007) measures 150 mm. in 
total length, 62 mm. in height; largest 9 (M.C.Z. 40011) measures 
180 mm. in length, 72 mm. in height. Smallest tortoise (M.C.Z. 
40008) measures 53 mm. and 24 mm. respectively. 

Sex. Males may be readily distinguished by their slightly concave 
plastrons and their very long tails. 

Diet. A male was observed feeding beside the road at 10 a.m. in 
bright sunshine. On being approached, he turned and made off into 
the bush with surprising agility. 

Testudo pardalis babcocki Loveridge 

Plate 1, figs. 1 and 2 

Testudo pardalis babcocki Loveridge, 1935, Bull. Mus. Comp. Zool., 79, p. 4: 
Mount Debasien, Karamojo, Uganda. 

Type 9 (M. C. Z. 40003) Mt. Debasien, U. 23.xi.33. 
9 & young (M. C. Z. 40004-5) Mt. Mbololo, K. C. 17.iv.34. 
8 eggs (M. C. Z. 40018) Mt. Mbololo, K. C. 17.iv.34. 
cT (M. C. Z. 40006) Wema, Ngatana, K. C. 17.vi.34. 

Distribution. This species has also been taken at Kaimosi by Heller. 

Affinities. In 1933 Hewitt proposed separating this species from 
Testudo under the name Megachersine. 

Native names. Akuma (Karamojong) ; ilcudu (Lugishu); Likudu 
(Luragoli and Lutereki); nguru (Kitaita, but not specific). 

Measurements. Greatest length of c? (M.C.Z. 40006) 238 mm., of 
9 (M.C.Z. 40004) 385 mm., and of young (M.C.Z. 40005) 58 mm. 

Breeding. From the oviducts of the female taken on Mt. Mbololo, 
April 17, 1934, I removed and blew eight eggs. These eggs measured 
38 x 36 mm. in diameter and were noticeably larger than the seven 
eggs, measuring 35 x 32.5 mm., laid on May 21, 1922 by a tortoise 
from Pwaga, Tanganyika Territory (vide Loveridge, 1923, p. 927). 
In addition to the eggs removed there were the usual developing ova 
of various sizes in the ovaries. 

Parasites. Ticks (Amblyomma exornatum) were removed from the 
neck and limbs of the Mt. Mbololo and Wema tortoises. 



loveridge: African reptiles 221 

Enemies. In addition the latter had very many heads of siafu or 
soldier ants (Dorylus nigricans subsp.) attached to it. 

From the shell of the big Mbololo female, it would appear as if a 
hyena or other carnivore had attempted to crack the shell by biting 
it when the reptile was of smaller size. The carapace has an injured 
area on its summit while the plastron has a corresponding mark 
where it has been perforated (by a canine?). It is interesting to note 
how the hole has been repaired from the inside by a bony growth, 
parallelling the repairs carried out on an oyster shell by its occupant. 

Habitat. The Debasien tortoise was taken among rocks on an arid 
ridge or spur of the mountain at an altitude of 5,000 feet. The young 
Mbololo specimen was found at 4,000 feet, the adult rather lower down 
the mountain. The Wema reptile was inhabiting the plains fringing 
the north bank of the Tana River which could not have been more 
than 1,000 feet above sea level. The name of Leopard Tortoise, 
therefore, seems preferable to Mountain Tortoise by which it is some- 
times known in South Africa. 

Testudo tornieri Siebenrock 

Testudo tornieri Siebenrock, 1903, Ak. Wiss. Wien, Math.-nat. Klasse, 24, p. 
185: "Bussisia" i.e. Busisi, Tanganyika Territory. 

1 young (Nairobi Mus.) Sokoki Forest, nr. Malindi, K. C. (H.J.A.T.) 1932 

Distribution. This specimen, which I examined when passing 
through Nairobi, constitutes a most interesting extension of range 
for the soft-shelled land tortoise. On October 14, 1934, Mr. Allen 
Turner wrote to me stating that this specimen "was caught by a boy 
while cleaning round his rice plants about a hundred yards from 
my camp at the top of Mida Creek about ten miles south of Malindi." 
This unusual habitat for a species usually associated with rocky hills 
in arid regions, is also quite a surprise. 

CHELONIIDAE 

Chelonia mydas (Linnaeus) 

Testudo mydas Linnaeus, 1758, Syst. Nat. ed. 10, p. 197: Ascension Island, 

Atlantic Ocean. 
Testudo japonica Thunberg, 1787, Sven. Vet.-Akad. Handl., 8, p. 178, pi. vii: 

Japan. 

J* (M. C. Z. 40019) Lamu, Lamu Id., K. C. 7.V.34. 



222 bulletin: museum of comparative zoology 

Native name. Kasa (Kiamu.) 

Measurements. Length over all of carapace 710 ram. 

Diet. Stomach full of sea grass but free of parasites. 



PELOMEDUSIDAE 

Pelusios sinuatus (Smith) 

Sternothaerus sinuatus A. Smith, 1838, Illus. Zool. S. Africa, 3, pi. i: South 
Africa, "in rivers to the north of 25° S. latitude." (i.e. region of the head- 
waters of the Limpopo River.) 

Pelusios sinuatus zuluensis Hewitt, 1927, Rec. Albany Mus., p. 360, pi. xx, 
figs. 1-3, text fig. Id: near Umsinene River, Zululand. 

Pelusios sinuatus leptus Hewitt, 1933, Occ. Pap. Rhodesian Mus., p. 1, pi. ix, 
fig. 1 : Isoka, northeast Northern Rhodesia. 

9 9 (M. C. Z. 40020, 40050) Tsavo River, K. C. 2-3.iv.34. 

Variation. These two large terrapin, taken from the river at almost 
the same spot, exhibit striking variation in almost every character 
which Hewitt thought to be of importance in differentiating the 
race zuluensis. They bear out my previous observations (Loveridge, 
1929, p. 15 and 1933, p. 208) based on material from Ujiji, Tangan- 
yika Territory and Mount Chirinda (Selinda), Southern Rhodesia. 

The more important variations of the Tsavo females are as follows. 
Vertebrals I to IV exhibit a protruberance (M.C.Z. 40020) or are 
perfectly smooth (40050) ; vertebrals II to IV are much longer than 
broad (40020) or as long as broad or very slightly longer (40050); 
outer border of the pectoral is shorter than that of the humeral 
(40020) or the outer border of the pectoral is longer than that of the 
humeral (40050) ; the intergular is not twice as long as broad (40020, 
whose intergular is 37 mm. long, 26 mm. broad) or the intergular is 
more than twice as long as broad (40050, whose intergular is 37 mm. 
long, 15 mm. broad); the lateral gulars very small, the anterior edge 
of one (9 mm.) being included at least 2}/% times in that (24 mm.) 
of the intergular (40020) or the lateral gulars relatively large, the 
anterior edge of one (11 mm.) being included less than 1}4 times in that 
(15 mm.) of the intergular (40050); the sides may be steeply sloping 
(40020) or more gradually so (40050); the relative lengths of the 
claws of the hind feet, being subject to growth and wear, is not 
comparable on the right and left feet of either terrapin. 

Coloration. The plastron of the smaller has the characteristic yellow 



loveridge: African reptiles 223 

centre sharply defined from the black border; in the larger specimen 
the juncture of the two colors is blurred. 

Measurements. M.C.Z. 40020 measures 285 mm. in outside length 
of shell, 95 mm. in greatest depth. The smaller terrapin measures 
227 mm. in length and 80 mm. in depth. This gives a depth into length 
of 3 and 2.8 times respectively, the four young specimens from Tan- 
ganyika and Southern Rhodesia ranged from 80 to 88 mm. in length 
in which the depth was included from 2.4 to 2.5 times. 

Pelusios nigricans nigricans (Dondorff) 
Plate 2, figs. 1 and 2 

Testudo nigricans Dondorff, 1798, Zool. Beytr. des Linn, natur., 3, p. 34: Type 
locality unknown. 

34 (M. C. Z. 40021-40049) Kaimosi, K. C. 13.ii-10.iii.34. 

Native names. Likudu (Luragoli); lihodu (Lutereki). 

Variation. This long series was secured to see if any embraced the 
proportions of P. n. castanoides Hewitt which was based on a single 
individual from Richard's Bay, Zululand. This type measured 225 
mm. in length and 98 mm. in height, giving a height into length of 
2.2. times. While the Kaimosi series actually ranges from 2.17 to 
3.30 times this is inconclusive for it is only a 100 mm. terrapin that 
furnishes the low figure and a 248 mm. specimen that is 3.30 times; 
a 225 mm. example (No. 40025) has a height of only 75 mm. giving 
a height into length of 3 times as against 2.2 times in Hewitt's type 
of castanoides. In passing, a few further remarks on the results of 
measuring these 34 terrapin might be added. Taking the eight 
largest males (151 to 240 mm.) and eight largest females (168 to 252 
mm.) no appreciable difference is found in the ratio of height into 
length or breadth into length. On the other hand there is a very 
marked age difference in these proportions, thus 

8 young 8 males 8 females 
Height is included in the length 2.1-2.4 2.5-3.0 2.7-3.1 
Breadth is included in the length 1.2-1.4 1.4-1.5 1.3-1.5 

The breadth of castanoides was 146 mm., giving a ratio of 1.54 times 
which can be matched by several of our series of a comparable length. 
Hewitt (1931, p. 466) is correct in stating that the intergular shield 
is pear-shaped and longer than the inner border of the humeral in 
East African nigricans. I notice that the intergular is flat in our 



224 bulletin: museum of comparative zoology 

examples of P. n. rhodesianus while it is rounded or embossed in these 
Kaimosi specimens, especially noticeable in the younger individuals. 
The intergular is always longer than the inner border of the humeral, 
in 14 reptiles it equals the united lengths of the inner border of humeral 
and pectoral, in 16 it is even longer, in only 3 is it shorter, 1 terrapin 
is damaged. 

On the other hand, I fail to see that "outer border of femoral 
strongly arched" is of any diagnostic significance as against "moder- 
ately arched" in castanoides, or "slightly arched" in castaneus (which 
I treat as a synonym of nigricans). The range of variation in this 
character in the Kaimosi series is enormous and must be seen to be 
appreciated. They do differ in this character from our examples of 
rhodesianus where they are "not definitely arched." 

The outer border of the pectoral is shorter than the outer border 
of the humeral in 24 specimens, is equal to it in 10, so that, as I have 
remarked elsewhere, (Loveridge, 1933, p. 209) this key character 
for distinguishing sinuatus and nigricans is useless. 

Since the foregoing was written, Hewitt (1935, p. 345) has decided 
to make rhodesianus a full species, because he has seen an example 
of it from Entebbe, Uganda, where nigricans is common. The greater 
probability is that rhodesianus is a synonym of nigricans. 

Coloration in life. Underside of a 30 mm. young one. Throat spotted 
with pink. Plastron peripherally blotched with red alternating with 
the general black ground color; marginals red and black. 

Measurements. Largest & measures 240 mm. in length of shell; 
biggest 9 measures 252 mm. ; smallest specimen 30 mm. 

Breeding. The youngest terrapin alluded to above, was taken on 
March 1 and appeared to have recently hatched as its abdominal 
shields were still unhealed. 

Diet. Crab's (Potamon sp.) claws and grass occurred in the faeces. 

Parasites. Leaches (some have been preserved) were commonly 
found upon these terrapin. 

Enemies. Two terrapin (M.C.Z. 40033, 40035) appeared to have 
been bitten by hyenas or some other carnivore when younger (both 
measure 167 mm. in length), one having a piece apparently taken 
out of its side. 

Habits. On February 13 the first rain for months fell between 4.30 
and 5.30 p.m. The following morning the first of these reptiles was 
brought in by a native. Its back caked with mud from which it 
seemed obvious that the creature had been aestivating. Several 
others were brought in subsequent to other showers. 



loveridge: African reptiles 225 

When hunting frogs at night with an electric torch, I discovered 
several of the largest in the series lying in shallow water at the edge 
of the mill pond, these were successfully netted. Attempts to capture 
them by setting a turtle trap in the outlet of the mill pond, proved 
abortive. 

Pelomedusa galeata (Schoepff) 

Testudo galeata Schoepff, 1792, Hist. Testud., p. 12, pi. iii, fig. 1: "Habitat in 
India orientale, Carolina." 

3 (M. C. Z. 40051-3) Kirui's village, Kitosh, K. C. 25.L34. 
6 (M. C. Z. 40054-9) Voi, Seyidie Province, K. C. 22.iv.34. 

Native name. Nguru (Kitosh, but not specific). 

Measurements. Largest specimen (M.C.Z. 40055), apparently a 
female, measures 142 mm., the smallest (M.C.Z. 40053) only 60 mm. 

Habitat. The Voi series were taken in a watercourse after a heavy 
downpour, one of several erratic showers inaugurating the rainy 
season after a prolonged drought. 



TYPHLOPIDAE 
Typhlops punctatus punctatus (Leach) 

Acontias punctatus Leach, 1819, in Bowdich, Miss. Ashantee, p. 493: Fantee, 

Gold Coast. 
Typhlops Boulengeri Bocage, 1893, Jorn. Sci. Lisboa (2), 3, p. 117: Quindumbo, 

interior of Benguela, Angola. 

Synonymy. As I failed to secure any Typhlops on Mount Elgon, 
Dr. H. Rendahl of the Royal Swedish Museum, Stockholm, was 
kind enough to send me for study, the specimen No. 2580, taken on 
the eastern slopes at about 6,500 feet, by Dr. Hugo Granvik, and 
referred to as " Typhlops boulengeri Bocage?" by Lonnberg (1922, p. 7). 

This snake has 29 (not 28) midbody scale-rows; its prefrontal is 
not more than twice as large as its supraocular, the latter forming 
a very broad suture with the nasal. Its length is 264 (257 -f- 7) mm., 
and diameter 11 mm., the latter being included in the former 24 
times. All these are characters of punctatus. 

I consider, however, that boulengeri was based on young (ISO to 
260 mm.) specimens of punctatus in which the prefrontal was slightly 
developed at the expense of the supraocular. As has been shown 
elsewhere (Loveridge, 1933, p. 218) in regard to another species, a 



226 bulletin: museum of comparative zoology 

more rounded snout is characteristic of the young, the obtusely 
horizontal edge being a secondary character developed to facilitate 
burrowing. 

Typhlops kaimosae Loveridge 

Typhlops kaimosae Loveridge, 1935, Bull. Mus. Comp. Zool., 79, p. 5: Kaimosi, 
Nyanza Province, Kenya Colony. 

Holotype (M. C. Z. 40060) Kaimosi, K. C. 7.iii.34. 

Native names. Inanyanza (Luragoli); mutumbo (Lutereki). 

Diagnosis. Distinguished by the preocular being separated from 
the upper labials by the ocular forming a broad suture with the nasal 
beneath it. No other East African Typhlops presents such an ar- 
rangement. 

Typhlops schlegelii mucruso (Peters) 

Onychocephalus mucruso Peters (part), 1854, Monatsb. Akad. Wiss. Berlin, p. 

621: Maganga (i.e. Makanga), Mozambique. 
Typhlops mandensis Stejneger, 1893, Proc. U. S. Nat. Mus., 16, p. 725: Wange, 

on mainland opposite Manda Island, Kenya Colony. 

1 (M. C. Z. 40061) Peccatoni, K. C. 24.V.34. 

2 (M. C. Z. 40062-3) Mkonumbi, K. C. 28.V.34. 

2 (M. C. Z. 40064-5) Near Witu, K. C. 31.V.34. 

3 (M. C. Z. 40066-8) Ngatana, K. C. 14.vi.34. 
1 (M. C. Z. 40069) Gongoni, K. C. 27.vi.34. 

Distribution. Also 3 from Sokoki Forest (H.J.A.T.). 

Native name. Nyoka vishwa viwili (Kipokomo). 

Synonymy. I suspect that Gunther's (1894, p. 87) record of T. 
punctatus from Mkonumbi was based on a juvenile mucruso with 
rounded snout. 

Variation. T. mandensis was based on an individual about to slough 
resulting in the eyes being hidden, this is the case with the Peccatoni 
and Gongoni snakes. Lake Peccatoni lies about forty miles south 
of Wange. A detailed discussion on the synonymy of mucruso has 
recently been given (Loveridge, 1933, pp. 216-219). 

The nine snakes listed above have: Midbody scale-rows 30-34 
(average 32.6); the rostral as seen from above is a little longer than 
broad in both adult and young; the eye, when visible, is beneath the 
ocular against the suture between ocular and preocular; the nasals 
are separated behind the rostral; the nasal suture rests on the first 



loveridge: African reptiles 



227 



labial, and not on the second labial which is the distinguishing char- 
acter of T. brevis Scortecci from Kismayu, two hundred miles to 
the north. 

Coloration in life. Young. Above, blue gray, each scale edged lat- 
erally with black which results in a lineolate appearance. Below, 
pinkish white. Adult. Above, black, the centre of each scale with 
a small buff spot. Below, buff. Numbers 40061 and 40069, being 
about to slough, are : Above, pale silvery gray or white. Below, pinkish 
white. In life the Gongoni reptile was noted as being white, or slightly 
bluish white. 

Measurements. Largest snake (M.C.Z. 40064) measures 447 
(440 + 7) mm. Diameter at midbody 15 mm. Smallest snake (M.C.Z. 
40062) measures 135 (133 + 2) mm. Diameter 5 mm. Diameters 
included in lengths from 27 to 33 times. 

Typhlops unitaeniatus unitaeniatus Peters 

Typhlops (Letheobia) unitaeniatus Peters, 1878, Monatsb. Akad. Wiss. Berlin, 
p. 205, pi. ii, fig. 5: Taita, Kenya Colony. 

2 (M. C. Z. 40079-80) Voi, Taita, K. C. 12.iv.34. 
1 (M. C. Z. 40081) Malindi, K. C. 28.vi.34. 

Distribution. Voi being at the foot of the Taita Mountains, the 
specimens from there may well be considered topotypes. Another 
snake from Kibwezi, K.C. (M.C.Z. 18175) has been utilized for 
these remarks. 

Native name. Ngomo (Kisagalla). 

Variation. Midbody scale-rows 22-24; eye, when visible, is beneath 
the nasal; diameter is included in the length from 61 to 76 times. 

Coloration in life. The dorsal band is a beautiful mustard yellow, 
in sharp contrast to the jet black of the rest of the body. 

Measurements. The largest snake (M.C.Z. 18175) measures 380 
(376.5 + 3.5) mm. Diameter at midbody 5 mm. 

Diet. Termites in the Kibwezi snake, apparently termites in the 
Malindi specimen. 

Typhlops pallidus (Cope) 

Letheobia pallida Cope, 1868, Proc. Acad. Nat. Sci. Philad., p. 322: Zanzibar. 

4 (M. C. Z. 40075-8) Ngatana, K. C. 14.vi.34. 

Distribution. Originally described from Zanzibar, later reported 
from Pemba Island, the above series from the Tana River constitute 



228 bulletin: museum of comparative zoology 

the first record from the mainland. They have been carefully com- 
pared with a topotype collected at the same time as the type. 

Variation. Midbody scale-rows 22; diameter included in total 
length 51-73 times (50 in Pemba, 58-60 in Zanzibar specimens); 
tail into total length 58-98 times, one might suggest possibly 58-64 
in males and 92-98 for females but the material is unsexed. 

Measurements. The largest (M.C.Z. 40075) measures 185 (183 + 2) 
mm. It is interesting to note that a snake measuring 152 mm. when 
just killed, now measures only 147 mm., others measured in the 
field have contracted in the same way. 

Habitat. Under vegetable debris heaped along banks of a rice swamp 
as described under Leptotyphlops longicauda. 

Typhlops braminus (Daudin) 

hfryx braminus Daudin, 1803, Hist. Nat. Rept., 7, p. 279: Bengal, India. 
Glauconia braueri Sternfeld, 1910, Mitt. Zool. Mus. Berlin, 5, p. 69: Bagamoyo, 
Tanganyika Territory. 

Sternfeld based his description of Glauconia braueri on a very 
small snake of 83 mm., which he later amended to 85 mm. The 
original description was exceedingly brief and did not even state the 
number of midbody scale-rows. Later the same year, however, in 
his "Die Schlangen Deutsch-Ostafrikas" (1910, p. 13) he gave the 
number as 14 which definitely placed the reptile among the Lepto- 
typhlops (Glauconia auct.). 

Its short tail separated it from all East African Leptotyphlops 
and agreed rather with the genus Typhlops. This aroused my sus- 
picions so I asked my friend, Mr. Karl P. Schmidt, who was visiting 
Berlin at the time, if he would kindly reexamine the type. He replied 
that the tail measurements were correct, but that the midbody 
scale-rows numbered 20! There is therefore no difference between 
braueri and braminus, the latter long known from the East African 
coast, though rare. It was long ago described from South Africa by 
Sir A. Smith under the name of Onychocephalus capensis. 

Typhlops lumbriciformis (Peters) 

Onychocephalus (Letheobia) lumbriciformis Peters, 1874, Monatsb. Akad. Wiss. 

Berlin, p. 377: Zanzibar coast. 
Typhlops kleebergi Werner, 1904, Zool. Anz., 27, p. 464: Usambara, Tanganyika 

Territory. 



loveridge: African reptiles 229 

1 (M. C. Z. 39951) Sokoki Forest, K. C. vi.32. H.J.A.T. 
1 (M. C. Z. 40070) Mkonumbi, K. C. 30.V.34. 
1 (M. C. Z. 40071) Gongoni, K. C. 27.vi.34. 
3 (M. C. Z. 40072-4) Malindi, K. C. 29.vi.34. 

Distribution. AJ1 these specimens were taken on the coast between 
Lamu and Mombasa and are therefore topotypes in the sense that 
this region formed part of the Zanzibar coast. As Hildebrandt col- 
lected the types it is more probable that they came from nearer 
Mombasa. 

Variation. Midbody scale-rows 18; eyes hidden, but the pigment 
may be detected beneath the nasal in some; preocular in contact 
with the 2nd and 3rd upper labials; ocular very small; snout with 
sharp horizontal edge; diameter is included in the length from 70 
to 83 times. 

If one regards the preocular as the ocular scale, the most natural 
thing to do as the eye is beneath the nasal, and the real ocular is 
scarcely enlarged, then the above series agree with the description 
of kleebergi in every detail except that the diameter of the type was 
included 56 times in the total length. 

Boulenger (1915, p. 615) referred kleebergi to the synonymy of 
lumbriciformis. Barbour & Loveridge (1928, p. 104) revived it on 
account of the very obvious differences between the description of 
kleebergi and that given for lumbriciformis by Boulenger (1893, p. 55). 
The original description of Peters is too vague to be of much help. 

On comparing the above series with Boulenger's redescription 
(1893, p. 55) which was based on a single snake from Zanzibar, and 
using his nomenclature for the scales, I find that ours differs in the 
preocular being in contact with the 2nd and 3rd labials only, not 
2nd, 3rd and 4th; ocular not in contact with nasal; and I should be 
inclined to add not separated from the 4th labial by a subocular but 
in contact with the 4th labial; this, however, depends on which of 
two shields is considered the homologue of the ocular. 

It would appear that Boulenger's description is inaccurate in 
some respects, and his naming of certain scales, though correct, 
misleading and likely to result in the description of synonyms. 

Coloration in life. Flesh pink. 

Measurements. Largest snake (M.C.Z. 40070) measures 500 (493 
+ 7) mm. Diameter at midbody 6 mm. 

Breeding. Four eggs measuring 16 x 5 mm. in 9 taken 29 June, 
1934. 

Diet. Termites in one of the Malindi snakes. 



230 bulletin: museum of comparative zoology 

Habitat. The first specimen listed is one of three collected by 
Mr. H. J. Allen Turner in the Sokoki Forest near Malindi. I found 
the Mkonumbi snake lying dead upon the road five miles south of the 
village. It had probably been drowned out of its burrow by the 
torrential rains that had fallen that morning. Thf Gongoni reptile 
was in sandy soil in a native garden devoted to maize and mahoga. 
The Malindi series in the reddish soil of a cotton plantation. All 
were secured after heavy rain. 

LEPTOTYPHLOPIDAE 

Leptotyphlops boulengeri (Boettger) 

Glauconia boulengeri Boettger, 1913, in Voeltzkow, Reise in Ostafrika, 3, p. 
354, pi. xxv, fig. 1 : Manda Island, Kenya Colony. 

4 (M. C. Z. 40084-7) Lamu, Lamu Island, K. C. 7.V.34. 
1 (M. C. Z. 40088) Kitau, Manda Island, K. C. 18.V.34. 

Variation. This topotype and the series from the neighbouring 
island of Lamu are, I believe, the first examples of this rare snake to 
be taken since it was described over twenty years ago. The original 
description, being based on a single specimen, may be expanded as 
follows. Midbody scale-rows 14; diameter is included in the length 
33 to 35 times (30 in type) ; tail is included in the total length 1 1 to 
12 times (11 in type); supraocular scarcely once and a half times as 
long as broad; rostral about equal in width to (slightly wider than) 
the nasal, reaches backwards to an imaginary line connecting the 
anterior borders of the eyes; upper portion of the nasal almost in 
contact with its fellow behind the rostral. 

Coloration in life. Flesh pink. 

Measurements. The largest snake (M.C.Z. 40085) measures 201 
(185 + 16) mm. Diameter at midbody 6 mm. 

Diet. Termites in largest specimen. 

Habitat. I caught the Kitau reptile among termites beneath a sheaf 
of rotting grass lying on red soil at the edge of one of the native 
gardens. Persistent search failed to reveal another example and it is 
possible that the type came from Manda, an abandoned settlement 
at the other end of the little island. 

The species was almost certainly introduced to Lamu Island with 
dhow cargoes of the red soil, which in Arab times was extensively 
taken over to Lamu for building purposes, possibly imported with 
sand. It may seem strange that sand should be imported to Lamu 



loveridge: African reptiles 231 

which is nothing but a sandbar. Nevertheless even during our brief 
stay a dhow came over and loaded up with sand; from the owner I 
learned that the coarse sand from Manda Island made better cement 
than the fine kind found on Lamu. 

Leptotyphlops longicauda (Peters) 

Stenostoma longicauda Peters, 1854, Monatsb. Akad. Wiss. Berlin, p. 621: Tete, 
Mozambique. 

1 (M. C. Z. 40089) Peccatoni, K. C. 24.iv.34. 
33 (M. C. Z. 40090-119) Ngatana, K. C. 14-21.vi.34. 

Distribution. These specimens constitute the first record of the 
occurrence of this species in Kenya Colony and come from an area 
almost exactly half-way between the most northerly record for 
longicauda, that of Tanga, Tanganyika Territory (Angel, 1925, p. 30) 
and the type locality of the related ficchteri in Italian Somaliland. 

Variation. It is often difficult to measure these small snakes twice 
with equal precision, a trifling fluctuation in the diameter, a slight 
stretching of the body may result in very different proportionate 
readings. Especial care has been taken therefore and the extremes 
checked and rechecked. The following figures are based on the 
Ngatana series only so as to give an idea of the range of variation 
in one locality. 

Total lengths range from 61-108.5 mm., midbody diameters 1-2 
mm., diameter is included in the length 46 to 81 times, average 59.5; 
tails included in the total length 9.5 to 12.5 times, average 10.4. The 
long-tailed specimens are males, the short-tailed are females. 

These specimens have been compared with Mozambique material 
which I collected at Lumbo. No examples of fiechteri have been seen 
for it is known only from the type which had a diameter into length 
of 68.5, and a tail into length of 12.5 times, which is precisely that of 
our Peccatoni snake. 

Coloration in life. Flesh pink and somewhat transparent. 

Measurements. The largest snake (M.C.Z. 40092) measures 108.5 
(98 + 10.5) mm., the smallest (M.C.Z. 40019) measures 61 (55 
+ 6) mm. 

Habitat. Ten of the Ngatana series were taken from a huge termite 
hill at the edge of a rice swamp and less than a quarter of a mile 
from the north bank of the Tana River. The rest were taken within 
six inches of the surface of black cotton soil forming part of the banks 
surrounding, and running through, flooded rice fields. In some 



232 bulletin: museum of comparative zoology 

instances these banks were heaped with weeds, in others old maize 
stalks had been thrown upon them and formed some protection from 
the powerful sun. It appears probable, that, on the flooding of the 
rice fields these tiny snakes are driven to seek safety in the raised 
banks, resulting in a concentration of them until the water subsides. 
I might add that these rice fields are on the site of the old village of 
Ngatana, a mile or so west and north from the new village of Wema. 

Leptotyphlops conjuncta (Jan) 

Stenostoma conjuncta Jan. 1861, Arch. Zool. Anat. Fisiol., 1, p. 189: South 
Africa. 

2 (M. C. Z. 40082-3) Kibwezi, K. C. 24.iii.34. 

Corrigenda. Sternfeld's (1910, p. 13) record of Glaucoma signata 
from Kibwezi is undoubtedly referable to this species. 

Variation. Midbody scale-rows 14; diameter is included in the 
length 53 to 56 times; tail is included in the total length 10 times. 
It is interesting to note that when measured immediately after 
death the result was 51 (instead of 56) times for the larger snake. 

Measurements. The larger snake (M.C.Z. 40082) measures 127 
(115 + 12) mm. Diameter at midbody 2.25 mm. (in life 2.5 mm.). 

Habitat. Taken in camp on freshly-cleared ground beneath a fig 
tree, and within two hundred yards of the station, the morning after 
a shower during the night. 

BOIDAE 

Python sebae (Gmelin) 

Coluber sebae Gmelin, 1788, Syst. Nat., 1, p. 1118: No type locality. 

1 (M. C. Z. 40301) Mkonumbi, K. C. 28.V.34. 
1 (M. C. Z. 40302) Ngatana, K. C. 14.vi.34. 

Distribution. No attempt was made to secure a series of this well- 
known reptile. On the western slopes of Mount Debasien a broad 
trail was found in close proximity to an abandoned Karamoja village 
and traced through the bush for a hundred yards. (20. xi. 33). At 
Tsavo a seven-foot python was seen basking under a rock on a kopje 
not far from the river. When I stalked it, the snake withdrew into a 
crevice and disappeared. At Voi I was shown a large skin of a locally- 
killed python. It is said to occur in a swamp on a remote part of Lamu 
Island. I examined specimens from the Sokoki Forest (Turner coll.) 



loveridge: African reptiles 233 

and Naivasha (L. S. B. Leakey coll.). Sternfeld has recorded an 
example from Kibwezi. 

Native names. Emorotot (Karamojong) ; nzatu (Lugishu); ivaga 
(Luragoli); ivaka (Lutereki); are (Kitaita); satu (Kipokomo and 
Kiswahili). 

Variation. Midbody scale-rows 88-90; ventrals 275-282; anal 
entire; subcaudals 68-73. 

Diet. There were the remains of a bird in the young python from 
Ngatana. 

Eryx colubrinus loveridgei Stull 

Eryx thebaicus loveridgei Stull, 1932, Occ. Papers Boston Soc. Nat. Hist., 8, 
p. 29, pi. ii, fig. B: Mbunyi, Kenya Colony. 

14 (M. C. Z. 40303-16) Voi, K. C. 9-24.iv.34. 

Native name. Ngwao (Kisagalla). 

Affinities. Recently, Flower (1933, pp. 804-5) has given cogent 
reasons for placing thebaicus in the synonymy of colubrinus (Linne) 
which he has removed from the synonymy of Eryx jaculus (Linne) 
where it had been placed by Boulenger (1893, p. 125). 

Overlooking Flower's paper, Ahl (1933, p. 325) referred loveridgei 
to the synonymy of thebaicus (i.e. colubrinus). In this action I was 
inclined to agree, but on going into the status of the East African 
form more fully, concluded that it may be recognized on the average 
lower number of ventrals as set forth in the key below under E. c. 
rufescens Ahl. 

Variation. In 1932, Dr. O. G. Stull separated East African speci- 
mens under the name of loveridgei, distinguishing them from their 
northern allies on the following grounds. "This subspecies differs 
from the allied E. t. thebaicus (Reuss) of North Africa in the higher 
number of scale rows (53-59 instead of 47-49), the lower average 
number of ventrals (168-182, average 173.2) instead of 175-192 
(average 184.8), and the immaculate belly and sides." 

Some years ago, (1916, p. 82) I published scale counts of five 
East African specimens taken from Kismayu, Italian Somaliland to 
Taveta, Kenya Colony. More recently, when passing through Nairobi, 
I took the opportunity of rechecking these counts which are as 
follows: — Midbody scale-rows 46-53; ventrals 162-182; subcaudals 
21-27. 

The fourteen examples from Voi, listed above, give: — Midbody 
scale-rows 46-56; ventrals 162-182; anal single; subcaudals 20-27. 



234 bulletin: museum of comparative zoology 

The latter are normally single though rarely as many as 5 (M.C.Z. 
40316) may be paired. 

If one collates the data of the 24 East African boas now available 
we have: — Midbody scale-rows 46-59, average 49; ventrals 162-182, 
average 170; subcaudals 20-27, average 24. 

This effectually disposes of the suggestion that there is a difference 
in the number of midbody scale-rows between North and East African 
examples, neither is there any difference in subcaudal range, nor in 
color pattern, which is discussed below. Dr. Stull informs me that 
there are differences in the mandibular and pterygoid teeth as well 
as in the hemipenes of the two forms, this evidence will be published 
in her forthcoming Monograph of the Boidae. 

Coloration in life. The Voi snakes were rich orange heavily mottled 
with black or chocolate brown. Thus they present a totally different 
appearance from the colored plate of an Egyptian colubrinus fur- 
nished by Anderson (1898, pi. xxxii, fig. 2) who describes their general 
color as "yellowish, with large, irregular, more or less transverse, 
purplish-brown markings," Major Flower writes me that he has no 
notes on the coloration in life of Egyptian colubrinus. In the absence 
of such data I tentatively assume that Anderson's plate and descrip- 
tion were based on alcoholic material. Justification for this assump- 
tion may be found in the fact that Egyptian and Kenya material are 
almost indistinguishable after a few years in alcohol, though the 
Egyptian snakes may be slightly more rufous and less chocolate brown. 

As both Boulenger (1893, p. 125) and Anderson (1898, p. 237) 
describe 'thcbaicus' (i.e. colubrinus) as having "lower parts uniform 
white," or "Under surface yellowish, immaculate." I am at a loss 
why Stull should have cited the "immaculate belly and sides" of 
East African specimens as a distinguishing character. 

Measurements. The largest snake, a female (M.C.Z. 40303) meas- 
ures 634 (584 + 47) mm., and is therefore the record for the southern 
form, the smallest (M.C.Z. 40311) measures 176 (160 + 16) mm. 

Breeding. On April 24, 1934, a native brought in a female and her 
seven young which he had found altogether in a hole. The length of 
the mother is given above, the young ranged from 176 to 189 mm. 

Diet. A gerbil (Dipodillus pusillus) was recovered from one snake. 
Another boa, taken on April 9, towards the close of a prolonged 
drought, had nothing in its stomach, but was well nourished and 
exhibited large deposits of fat. Such was also the condition of two 
females taken on the 12th. 

Defence. The cloacal glands contain a very evil-smelling secretion, 



loveridge: African reptiles 235 

those of the mother of the seven young were full and, on being pressed, 
shot out fine threads of the viscid fluid to a distance of one foot away. 
Habitat. I captured the two adult females on the 12th beneath 
the debris of a collapsed native hut about a mile southeast of Voi 
Station. Associating with them in this situation were two female 
burrowing vipers (Atractaspis microlepidotus) . 

Eryx colubrinus rufescens Ahl 

Eryx rufescens Ahl, 1933, Sitz. Ges. naturf. Freunde Berlin, p. 324, figs.: 
Dadab, 1 South Ethiopia. 

1 (M. C. Z. 39110) Bulbar, British Somaliland (Brockman). 

Affinities. Ahl recently described as a full species, a color form 
of colubrinus, which has long been confused with thebaicus (now a 
synonym of colubrinus). While it is not usual to recognize color 
races among the ophidia, our knowledge of the herpetofauna of the 
northeast corner of Africa leads one to suspect that it may be recog- 
nizable even though none of the scale characters cited by its author 
differentiate it from the typical form. 

As might be expected, the specimen from British Somaliland listed 
above, agrees with Ahl's Ethiopian holotype in the uniform nature 
of its dorsal pigmentation. At mathematical midbody, however, it 
possesses 50 (instead of 44-46) midbody scale-rows; it has 17 (in- 
stead of 14) infralabials, an extremely variable character in this genus. 
I fail to see any difference such as 'somewhat broader head' or 'larger 
scales on the underside of the head' between this Somaliland snake 
and a topotypic colubrinus from Luxor, Egypt. 

The following tentative key may aid in distinguishing the forms 
until more material is available. 

Dorsal coloring uniform; ventrals 181-194, average (2 ex.) 187 

E. c. rufescens 
Dorsal coloring consisting of heavy blotches. 

Ventrals 175-197, average (36 ex.) 186.3. . . E. c. colubrinus 
Ventrals 162-182, average (24 ex.) 170.8. .E. c. loveridgei 
It might be added that the uniting of ventral and subcaudal counts 
for these two last forms, merely reflect the same variation of 16 
as is shown in the ventral counts in the key. I am indebted to Dr. 
O. G. Stull for furnishing me with the ventral counts, based on 
literature and her own researches, for E. c. colubrinus. 

!? Dudub, roughly 7° N., 43° E. 



236 bulletin: museum of comparative zoology- 

COLUBRIDAE 

Natrix olivacea olivacea (Peters) 

Coronella olivacea Peters, 1854, Monatsb. Akad. Wiss. Berlin, p. 622: Tete, 
Mozambique. 

6 (M. C. Z. 40316-20) Kaimosi, K. C. ll-20.ii.34. 
1 (M. C. Z. 40321) Mkonumbi, K. C. 22.V.34. 

1 (M. C. Z. 40322) Near Witu, K. C. 31.V.34. 

2 (M. C. Z. 40324-5) Belazoni, K. C. 5.vi.34. 
1 (M. C. Z. 40326) Laini, K. C. 6.vi.34. 

18 (M. C. Z. 40327-44) Ngatana, K. C. ll-20.vi.34. 

7 (M. C. Z. 40345-50) Golbanti, K. C. 22.vi.34. 

Distribution. Already recorded from Ngatana by Giinther (1894), 
p. 87) under the name of Coronella olivacea var. dumerilii Giinther, 
a race described from the Gold Coast. 

Native names. Kigoyogoyo (Luragoli); shigoyogoyo (Lutereki); tine 
(Kipokomo). Some Wapokomo, however, apply tine also to Crota- 
phopeltis h. hotamboeia which, next to olivacea is the most abundant 
snake along the lower Tana River. They have a specific name for 
hotamboeia. 

Variation. Midbody scale-rows 18-19, the three snakes with 18 
have 19 a short distance anterior to midbody; ventrals 131-147; anal 
divided; subcaudals 44-61; labials 8 (rarely 7 or 9), the 4th and 5th 
entering the orbit, M. C. Z. 40319 has 7 with 4th and 5th entering the 
orbit, M. C. Z. 40322 has 8 with 3rd, 4th and 5th, M. C, Z. 40331 has 
9 with 4th, 5th and 6th, M. C. Z. 40325 has 9 with 5th and 6th, all 
these aberrations, however, occur on the right side only; preoculars 1, 
except for M. C. Z. 40333 which has 2; postoculars 3 in twenty-nine 
snakes, 2 in five, 1 on left side only of M. C. Z. 40336; temporals 
1 + 2 excepting in four snakes where, by vertical division of the 
anterior temporal, we find 1 + 1 + 2, at least on one side of the head. 

Though so short, the Kaimosi series displays a noticeably higher 
ventral count (132-147) and a lower subcaudal (44-61) as contrasted 
with the larger coastal series which possess 131-138 ventrals, and 
52-69 subcaudals. However, a check-up with all the available data of 
West African olivacea, to which one might expect the Kaimosi snakes 
to belong, shows the range to be substantially the same as in the East. 

Coloration. The Kaimosi series alone exhibit gray, olive and mauve 
types of coloring. All, however, possess the dark olive band or dorsal 



loveridge: African reptiles 



237 



stripe characteristic of snakes from the Central African Lake Region. 
This stripe does occur, though rarely, in the coastal series. 

Measurements. The largest cf (M. C. Z. 40329) measures 464 
(332 -+- 132) mm., females in the Kaimosi series easily surpass those 
from the coast, the largest (M.C. Z. 40318) measuring 491 (400 + 91) mm. 

Breeding. The two largest females taken at Kaimosi on February 14 
and 19 respectively, each held 6 well-developed eggs which measured 
approximately 23 x 7 mm. At Ngatana between June 11 and 20 the 
following were found: (1) 6 small eggs, (2) 6 eggs measuring 15x6 mm., 
(3) 6 eggs 19 x 7 mm., (4) 6 eggs 19 x 11 mm., (5) 6 eggs 22 x 10 mm., 
(6) 6 eggs 23 x 11 mm. These latter quite ready for deposition. 

Compare these with the smaller number (2 to 4) produced by the 
montane race N. o. uluguruensis from Tanganyika Territory. 

Diet. Bufo stein dachnerii was recovered on three occasions from 
snakes taken at Laini and Ngatana; Arthroleptis minutm from a 
Kaimosi specimen; Hyperolius milnei at Golbanti; Rana sp. at Witu, 
and a fish from another Ngatana reptile. 

Habitat. One beneath a board by the millpond at Kaimosi, another 
in a tussock of grass in midstream. One was observed to swim the 
Tana River in full flood at a point just above Golbanti where the river 
was at least fifty yards broad. On reaching the north bank it paused 
to rest, thus enabling me to catch it. 

To do so, I placed one foot on the muddy bank, the other remaining 
on the side of the dugout in which we had given chase. The Pokomo 
boatmen, however, on observing me seize the snake, divined my inten- 
tion of returning to the canoe with it, incontinently dropped their 
paddles and scrambled back along their unsteady craft. Without their 
paddles counteracting the current, the dugout was swirled away from 
shore. For a fleeting moment my legs spreadeagled, then I was 
plunged into the muddy torrent without touching bottom. As I went 
down I tossed the snake into the dugout where it disappeared beneath 
the baggage. Later it was recovered when we were unloading the 
craft at Golbanti. 

Both at Golbanti and in the vicinity of Wema, Ngatana, the Olive 
Water Snake was exceedingly common among the drying grass and 
weeds heaped in lines to demarcate the natives' gardens. 

Boaedon lineatus Dumeril & Bibron 
(Plate 3, fig. 1) 
Boaedon Lineatum Dumeril & Bibron, 1854, Erpet. Gen., 7, p. 363: Gold Coast. 



238 bulletin: museum of comparative zoology 

20 (M. C. Z. 40351-70) Sipi, U. 14-22.xii.33. 

5 (M. C. Z. 40371-5) Butandiga, U. 5-1U.34. 
33 (M. C. Z. 40376-407) Kaimosi, K. C. 7-28.ii.34. 

3 (M. C. Z. 40408-10) Voi, K. C. 9.iv.34. 

1 (M. C. Z. 40411) Mt. Mbololo, K. C. 17.iv.34. 
32 (M. C. Z. 40412-43) Lamu Id., K. C. 7-14.V.34. 

1 (M. C. Z. 40444) Mkonumbi, K. C. 23.V.34. 

2 (M. C. Z. 40445-6) Peccatoni, K. C. 26.V.34. 

3 (M. C. Z. 40447-9) Near Witu, K. C. 31.V.34. 
2 (M. C. Z. 40450-1) Laini, K. C. 6.vi.34. 

13 (M. C. Z. 40452-64) Ngatana, K. C. 9-21.vi.34. 

1 (M. C. Z. 40618) Golbanti, K. C. 28.vi.34. 

2 (M. C. Z. 40465-6) Malindi, K. C. 30.vi.34. 

Distribution. Also 4 from Sokoki Forest. (H. J. A. T.) 

Native names. Namage (Lugishu); moya (Kisagalla); ilumangiu 
(Kitaita); nyoka kitangu (Kiamu); muatu (Kipokomo). 

Variation. Midbody scale-rows 25-33; ventrals 186-238; anal sin- 
gle; subcaudals 46-71; labials 8, the 4th and 5th entering the orbit on 
both sides of the head in 102 snakes, 8 labials with the 3rd, 4th and 5th 
on both sides in three snakes, 9 labials with the 4th and 5th on both 
sides in nine snakes, 9 labials with 5th and 6th on both sides in two 
snakes, 9 with 4th, 5th and 6th on the right side while the left is normal 
in two snakes, azygous combinations of the above variations in six 
others; preoculars 2 except M. C. Z. 40618 which has only 1 on the right 
side and sixteen snakes which have only 1 on both sides, seven of the 
latter are from Ngatana, five from Kaimosi; postoculars 2 except for 
M. C. Z. 40456 which has 3; temporals 1 + 2 in 103 snakes, 1 + 3 in 
five, an azygous combination of these in nine others and 2 + 3 on the 
right side of M. C. Z. 40397. 

On analysis of the above data we find the material falls into two very 
distinct groups, viz. : 

62 snakes from the interior plateau (2,000 to 7,000 feet) which 

have from 29-34 midbody scale-rows and 201-238 ventrals. 

56 snakes from the coastal plain (under 500 feet) which have from 

25-27 midbody scale-rows and 186-213 ventrals. 
The former of these corresponds to lineatus Dumeril & Bibron, for 
the latter at least two names are available. For the present it seems 
inadvisable to recognize the latter for my records of several hundred 
scale-counts from South and West Africa show that the ranges are not 
too well defined. In brief it may be said that the nominate form 
inhabits West and central Africa while the counts for Angola, South 
Africa and the East African coast are distinctly lower, individuals with 



loveridge: African reptiles 239 

higher counts occur in this area, however, and conversely those with 
low counts in the general region allocated to typical lineatus. 

Measurements. The largest & (M. C. Z. 40363) measures 820 (680 
+ 140) mm., the largest 9 (M. C. Z. 40381) measures 944 (837 + 
107) mm. The smallest snake (M. C. Z. 40435) measures 122 (85 + 
37) mm. 

I was very much struck by the smaller size of all, except two large 
females, in the Lamu series (see remarks on diet below). The average 
total length of thirteen males was 480 mm. as against an average total 
length of thirteen unselected males from the more tropical central 
region (Sipi, Butandiga, Kaimosi) of 657 mm. 

The largest c? in the Lamu series measured 554 mm., the largest 9 , 
918 mm. It will be seen that the latter is not far short of the biggest 
9 from Kaimosi, but together with one other 9 it was far larger than 
the average for the island. 

Breeding. The undermentioned records of developing eggs were 
made. 

Sipi, December 16, 1933. 9 held 1 egg measuring 37 x 18 mm. 

Butandiga, January 11, 1934. 9 held 1 egg measuring 28 x 13 mm. 
Kaimosi, March 1, 1934. 9 held 7 eggs measuring 29 x 15 mm. 

Kaimosi, March 1, 1934. 9 held 10 eggs "ready for deposition." 

Lamu Id., May 10, 1934. 9 held 5 eggs measuring 28 x 12 mm. 

The first two are difficult of explanation for the snakes were large 
examples, the Butandiga reptile held a second egg only half the size 
of the one measured. The Kaimosi snake with seven eggs actually 
measured 38 inches. 

Diet. Rodent fur was present in three snakes from Sipi and one 
from Malindi. The shrew {Crocidura t. zaodon) was recovered from a 
Sipi and a Butandiga snake; a roof rat (Rattus r. kijabius) measuring 
7j/2 inches from snout to anus was in the stomach of a 38 ' female 
House Snake at Kaimosi (see plate 3, fig. 1), a smaller rat in a second 
snake; a family of pigmy mice (Leggada t. triton) in one Sipi snake, a 
single one in another reptile from Kaimosi; at this locality two snakes 
had swallowed mice {Leggada g. grata.) 

A young House Snake was taken in the act of swallowing a gecko 
(Hemidactylus mabouia) at Ngatana; another species of gecko (Lygo- 
dactylus p. mombasicus) had been swallowed by one at Voi; a lizard 
(Eremias neumanni) was recovered from another Ngatana snake; a 
skink (Siaphos kilimensis) from an Mbololo reptile. 

1 Misprinted in mammal report as 32". 



240 bulletin: museum of comparative zoology 

On Lamu Island, for the first time in my experience of this reptile, 
amphibian bones were found in one, a frog {Rana o. oxyrhynchus) in 
another, the burrowing frog (Hemisus m. marmoratum) in a third. It 
might be added that mammals are very scarce on the island, this factor 
may have forced the snakes to an amphibian diet; possibly we have 
here the explanation of the small size attained by Boaedon lincatus on 
Lamu (see above) though this is probably correlated with the reduced 
number of midbody and ventral scale counts. 

Parasites. Two Sipi snakes had the larval stages of Protocephalid 
tapeworms encysted on their mesenteries, their viscera being all snarled 
up with adhesions resulting from the encysted worms. A Kaimosi 
snake was similarly affected. 

Enemies. One House Snake was found in the stomach of a Banded 
Harrier-Eagle (Circaetus fasciolatus) at Ngatana. 

Aestivation? On February 10, 1934, at Kaimosi, a native brought in 
four House Snakes, all of the same size and nearly full-grown. He 
stated that while engaged in digging a pit, he had come upon them all 
together in a hole. As the rains had not broken as yet, it would appear 
that they were aestivating in company. 

Folklore. On Mount Mbololo several Wataita told me that the 
House Snakes which I had collected, were the mountain form of the 
Mamba (Dcndraspis angusticeps) which was found lower down! 
Whether this idea is widespread, or the invention of my informant, I 
cannot say. 

Boaedon olivaceus (Dumeril) 
Holurophis olivaceus A. Dumeril, 1856, Rev. Mag. Zool., p. 466: Gaboon. 
& (M. C. Z. 39965) Mabira Forest, U. 28.viii.33. 

Distribution. This snake was collected and presented by Captain 
C. R. S. Pitman. The species has been recorded from Uganda pre- 
viously. 

Variation. Midbody scale-rows 25; ventrals 192; anal entire; sub- 
caudals 52; labials 8, the 4th and 5th entering the orbit; temporals 
1 + 2. 

Boulenger (1893, p. 335) with only Cameroon material, states: 
"frontal once and a half as long as broad, as long as its distance from 
the end of the snout." In a series from Metet, Cameroon (M. C. Z. 
10307-13) the frontal is 134 to V/% as long as broad. In the Uganda 
snake, which otherwise conforms to Boulenger's description except in 



loveridge: African reptiles 241 

coloration, the frontal is scarcely longer than broad and only as long 
as the prefrontals and internasals together, i.e. shorter than its dis- 
tance from the end of the snout. 

Coloration in alcohol. Above, uniformly plumbeous. Below, yellow, 
the dorso-lateral pigmentation strongly impinging upon the ventrals 
which also are spotted or blotched with plumbeous sparsely and irreg- 
ularly. 

Measurements. This cf measures 594 (495 + 99) mm. 

Habitat. Captain Pitman expressed surprise on hearing that this 
reptile was a relative of the Brown House Snake for in habits he be- 
lieves it to be a waterside or semi-aquatic snake. 



Lycophidion capense capense (Smith) 

(Plate 3, fig. 2) 

Lycodon capense A. Smith, 1831, S. Africa Quart. Journ., (1) No. 5, p. 18: 
Kurrichane, i.e. Rustenberg district, Transvaal. 

1 (M. C. Z. 39966) Kigezi district, U. C.R.S.Pitman. 
1 (M. C. Z. 40467) Sabei, Mt. Elgon, U. 9.xii.33. 

3 (M. C. Z. 40468-70) Sipi, Mt. Elgon, U. 10-27.xii.33. 

4 (M. C. Z. 40471-3) Kaimosi, K. C. 25.ii-2.iii.34. 
1 (M. C. Z. 40474) Mkonumbi, K. C. 28.V.34. 

3 (M. C. Z. 40475-7) Ngatana, K. C. 18.vi.34. 

Native name. Wahobi (Lugishu); Kifuya (Kipokomo). 

Variation. Midbody scale-rows 17; ventrals 180-214; anal entire; 
subcaudals 29-54; upper labials 8, the 3rd, 4th and 5th entering the 
orbit; distance from the frontal to the end of the snout shorter, or 
much shorter, than the length of the parietals. 

Coloration. Agreeing with the typical form in the throat being more 
or less white. In the field the Uganda and Kaimosi (i.e. western Kenya) 
snakes were noticeably more spotted than the coastal specimens. 
This difference is not correlated with any scale characters enabling 
one to separate them. 

Measurements. The largest cT (M. C. Z. 40469) measures 443 (377 
+ 66) mm., the largest 9 (M. C. Z. 40468) measures 558 (495 + 63) 
mm. though one from Sabei with an injured tail surpasses the snout to 
anus length by 10 mm. 

Breeding. The big Sabei female held six eggs measuring 17x12 mm. 
on December 9, 1933. In marked contrast was the smallest Kaimosi 



242 bulletin: museum of comparative zoology 

reptile, measuring 267 mm. which held but a single egg measuring 
20 x 6 mm. on March 2, 1934. 

Diet. A Striped Skink (Mabuya striata) was recovered from a Sipi 
snake and an Ablepharus wahlbergii from another at Kaimosi. 



Lycophidion capense > < acutirostre Giinther 

Lycophidion intermediates between capense and acutirostre Loveridge, 1933, 
Bull. Mus. Comp. Zool., 74, p. 234: Zanzibar and Bagamoyo, Morogoro and 
Kilosa in Tanganyika Territory. 

2 (M. C. Z. 40478-9) Kibwezi, K. C. 23.iii.34. 
1 (M. C. Z. 40480) Mt. Mbololo, K. C. 26.iv.34. 
1 (M. C. Z. 40481) Malindi, K. C. 30.vi.34. 
1 (Destroyed) Changamwe, K. C. 5.vii.34. 

Distribution. Also 4 from Sokoki Forest (H. J. A. T.). These 
records extend the area of intermediates to the northward as was to be 
expected. It seems probable that eventually it will be advisable to 
increase the range of ventrals in acutirostre so as to include them with 
that form. For further discussion see citation given above. What is 
needed is a large series of Zanzibar material so that the ventral range 
on that island may be ascertained. 

Variation. Midbody scale-rows 17; ventrals 154-170; anal entire; 
subcaudals 28-37; upper labials 8, the 3rd, 4th and 5th entering the 
orbit; parietal length slightly or much longer than the distance from 
the frontal to the end of the snout. 

Coloration. All the above agree in having the throat and lower 
surface uniformly blackish or blackish-brown. 

Measurements. The largest c? (M. C. Z. 40478) measures 294 
(260 + 34) mm., the largest 9 (M. C, Z. 40481) measures 399 (350 + 
49) mm. 

Breeding. The largest female held eight eggs measuring 13 x 7 mm. 
on June 30, 1934. 

Diet. A lizard (Eremias s. spelcii) and a skink (Riopa sundevallii) 
were recovered from the two Kibwezi snakes. 

Parasites. A single nematode (0 phidascaris sp.) was preserved from 
the stomach of the Malindi female. 

Enemies. A male was disgorged by another snake {Calamelaps 
unicolor) taken at Changamwe. The wolf snake was only 20 mm. 
shorter than its captor. It was so far decomposed that, after taking its 
scale counts, I destroyed it. Its first three subcaudals were unpaired. 



loveridge: African reptiles 243 



Mehelya nyassae (Giinther) 

Simocephalus nyassae Giinther, 1888, Ann. Mag. Nat. Hist., (6), 1, p. 328: 
Lake Nyasa, Nyasaland. 

9 (M. C. Z. 40482) Wema, Ngatana, K. C. 13.vi.34. 

Distribution. Besides the type, in 1893, the British Museum only 
possessed a second specimen from Zanzibar. In 1918 I collected a 
female at Lumbo, Mozambique. The above specimen from the north 
bank of the Tana River constitutes a northward extension of its 
range by some 250 miles and is the first record for Kenya. 

Variation. Midbody scale-rows 15; ventrals 175; anal entire; sub- 
caudals 55 (as against 62-63) ; labials 7, the 3rd and 4th entering the 
orbit; preocular 1; postocular 1; temporals 1 + 2. It differs from the 
description in the Catalogue of Snakes (1893, p. 347, pi. xxiii, Fig. 2) 
in the internasals being two-thirds, not half, the length of the pre- 
frontals; the frontal as long as, not longer than, broad, and shorter 
than its distance from the rostral. 

Measurements. This gravid 9 measures 554 (452 + 102) mm. 

Breeding. The oviducts held three eggs measuring 10x4 mm. on 
June 13. 

Defence. On being struck with a stick by one of a gang employed in 
clearing grass immediately in front of my tent, this Nyasa File Snake 
emitted a most foul odour, far surpassing that of the European Grass- 
snake. 

Chlorophis carinatus Anderson 

Chlorophis carinatus Andersson, 1901, Svenska Vetensk.- Akad. Handl., 27, 
No. 5, p. 9: Cameroon. 

c? (M. C. Z. 40173) Sipi, U. 18.xii.33. 
<? (M. C. Z. 40483) Kaimosi, K. C. ii.34. 

Distribution. Originally described from the Cameroon, subse- 
quently recorded by Schmidt (1923, p. 74) from the Belgian Congo, the 
two specimens listed above constitute the first records for the occur- 
rence of carinatus in Uganda and Kenya Colony. 

Native name. Kangasira (Lugishu). The Bagishu recognize the 
distinctiveness of this species from C. hoplogaster which occurs along- 
side it both at Sipi and Kaimosi, but which is far more abundant. 

Affinities. Naturally I have considered the possibility of these being 



244 bulletin: museum of comparative zoology 

aberrant examples of either hoplogaster or irregularis. The combina- 
tion, however, of 13 midbody scale-rows and an undivided anal appear 
to preclude this possibility even should we negative the distinctive 
coloring of carinatus. Elsewhere I have discussed the relationships of 
carinatus with heterodermus which so frequently occurs with it. 

If the ventral and subcaudal ranges as given by Hecht (1929, p. 331) 
are correct then there is nothing to differentiate the western carinatus 
from the eastern macrops (confined to the Usambara Mountains) 
except the number of labials and their arrangement — normally 9, with 
the 4th, 5th and 6th entering the orbit in carinatus, normally 8 with the 
5th and 6th entering in macrops. From their coloration also it is un- 
doubtedly true that carinatus is the western representative of macrops 
but it appears inadvisable to treat it as a race in view of the fact that 
their ranges are still separated by some 450 miles. 

Variation. Midbody scale-rows 13; ventrals 146-152, faintly 
keeled on anterior third of body; anal entire; subcaudals 82-86; 
labials 9, the 4th, 5th and 6th entering the orbit, or 10, the 4th, 5th, 
6th and 7th entering on the left side only (M. C. Z. 40175); preocular 
1; postoculars 2; loreal 1; temporals 2 + 2 or 2 + 1. 

Measurements. The larger d" (M. C. Z. 40173) measures 238 
(152 + 86) mm. 

Coloration in life. On receiving the somewhat damaged Sipi snake, 
I immediately recognized its distinctive appearance as compared with 
hoplogaster which was so common in this locality. Fortunately I 
recorded its coloring at once, for in alcohol it is indistinguishable from 
that of hoplogaster. This coloration should be compared with that of 
macrops which I noted in the Usambaras (Barbour & Loveridge, 
1928, p. 116). 

Above, dark olive with 104 deep-black, irregular crossbands between 
head and anus, these are represented on the tail by black flecks; the 
olive scales between the crossbands are edged with pale blue on the 
anterior two-thirds of the body; upper lip brownish-olive anteriorly, 
white below the eye shading off into olive posteriorly. Below, throat 
pure white, anterior ventrals tinged with yellow, remainder of the 
under surface dark green with its anterior third heavily suffused with 
yellow; on the anterior two-thirds edged with yellow laterally, on the 
posterior third with bluish white, on the tail with dusky. 

Eye. The centre of the eye is black surrounded by a light area, then 
by a fine orange line, then by an olivaceus area flecked with black; 
outermost ring, black. 



loveridge: African reptiles 245 



Chlorophis hoplogaster (Giinther) 

Ahaetulla hoplogaster Giinther, 1863, Ann. Mag. Nat. Hist. (3), 11, p. 284: 
Port Natal, i.e. Durban, Natal. 

1 (M. C. Z. 40496) Mt. Debasien, U. 15.xi.33. 
22 (M. C. Z. 40151-72) Sipi, Mt. Elgon, U. 12-23.xii.33. 

12 (M. C. Z. 40484-95) Butandiga, U. 5-16.i.34. 

1 (M. C. Z. 40497) Kirui's Village, K. C. 25.i.34. 

13 (M. C. Z. 40498-509) Kaimosi, K. C. 19.ii-5.iii.34. 

Native name. Emun (Karamojong) ; naranyase (Lugishu), but neither 
specific. 

Variation. Midbody scale-rows 15; ventrals 153-178; anal divided; 
subcaudals 87-109; labials 8, the 4th and 5th entering the orbit on 
eighty sides, or 9, the 4th and 5th entering on two sides, or 9, the 5th 
and 6th entering on sixteen sides; preoculars 1, or 2 on three sides 
(M. C. Z. 40501, 40504); postoculars 2, or 3 on left side of M. C. Z. 
40492 only; temporals 1 + 2 on sixty-three sides, 1 + 1 on thirty- 
four sides, or 2 + 2 on one side of M. C. Z. 40489. 

Some years ago, I (1916, p. 78) invited attention to the variations 
in two large collections of Chlorophis made by Mr. H. J. Allen Turner 
at his camps at Kaimosi and on the Yala River nearby. Shortly after- 
wards a third collection was received in which the hoplogaster material 
exhibited the following astonishing labial variations : 

1 specimen with 7-7 supralabials, 3rd and 4th entering the orbit. 
1 " " 7-7 " 4th and 5th 

1 " " 7-8 " 4th and 5th " 

1 " " 8-7 " 4th and 5th & 3rd and 4th " 

10 " " 8-8 " 4th and 5th " " " 

1 " " 8-8 " 5th and 6th 

Measurements. The largest c? (M. C. Z. 40499) measures 930 (700 
+ 230) mm. but the tip of the tail is lacking; largest 9 (M. C. Z. 
40155) measures 1045 (762 + 283) mm.; the smallest snake (M. C. 
Z. 40166) measures 435 (305 + 130) mm. 

Breeding. The undermentioned records of developing eggs were 
made in the field. It should be noted, however, that several females 
were taken at Sipi between December 12 and 23, which showed no 
signs of developing ova and the snake taken on the 23rd held an eighth 
egg much smaller than the other seven. 



246 bulletin: museum of comparative zoology 

Mt. Debasien, November 11, 1933. 9 held 7 eggs measuring 21 x 8 mm. 

Sipi, December 12, 1933. " 7 " " 17 x 5 mm. 

" 4 " " 22 x 6 mm. 

" 5 " " 30 x 8 mm. 

" 4 " " 31 x 13 mm. 

23, 1933. " 7 " " 27 x 8 mm. 

Butandiga, January 5, 1934. " 6 " " 13 x 5.5 mm. 

Kaimosi, February 20, 1934. " 7 " " 30 x 12 mm. 

25, 1934. " 8 " " 31 x 13 mm. 

March 3, 1934. " 5 " " 34 x 12 mm. 

Diet. A gecko (Cnemaspis a. elgonensis), the freshly dropped tail 
of a lizard (Lacerta jacksoni), a skink (Mabuya striata), a chameleon 
(C. b. hohnelii), and a frog (Rana f. fuscigula) were found in Sipi 
snakes. Three snakes held chameleons (C. b. hohnelii) and a fourth a 
toad (Bufo r. regularis) at Butandiga. A skink (M. striata) in an old 
hoplogaster, an enormous frog (Rana o. gribinguiensis) in a quite 
moderately sized snake, a tree frog (Hyperolius rossii) in a very young 
one at Kaimosi. 

Parasites. Encysted nematodes were found in the stomach wall of 
a Kaimosi reptile and preserved. 

Chlorophis neglectus (Peters) 

Philothamnus neglectus Peters, 1866, Monatsb. Akad. Wiss. Berlin, p. 890: 
Prazo Boror, Mozambique. 

9 (M. C. Z. 40515), Nairobi, K. C. 17.iii.34. 
2 (M. C. Z. 40516-7) Peccatoni, K. C. 24.V.34. 

1 (M. C. Z. 40518) Mkonumbi, K. C. 29.V.34. 
cf (M. C. Z. 40519) Near Witu, K. C. 30.V.34. 
c? (M. C. Z. 40520) Kau, K. C. 4.vi.34. 

2 (M. C. Z. 40521) Ngatana, K. C. 12.vi.34. 

Distribution. On April 21, 1934 at 4,800 feet on Mount Mbololo, I 
captured a green snake which I believe to be this species, unfortunately 
it escaped. It was in a forest glade at the summit. 

Native name. Homboka (Kipokomo). 

Variation. Midbody scale-rows 15; ventrals 151-171; anals 2; sub- 
caudals 92-109; labials 8, the 4th and 5th entering the orbit; preocu- 
lar 1 ; postoculars 2 ; temporals 1 + 1 except for the Nairobi snake which 
had 1 + 2. 

Measurements. The largest d 71 (M. C. Z. 40519) measures 735 (500 
+ 235) mm., the largest 9 (M. C. Z. 40515) measures 723 (553 + 
170) mm. but the tip of its tail is missing. 



loveridge: African reptiles 247 

Breeding. The females taken at Peccatoni on May 24, held 5 eggs 
measuring 17 x 5 mm. and 7 eggs measuring 18 x 9 mm. respectively. 
The Ngatana 9 held 6 eggs measuring 23 x 8 mm. on June 12, 1934. 

Diet. One Peccatoni snake had swallowed two juvenile frogs (Rana 
edulis). 

Habitat. I captured one of the Peccatoni snakes far out among the 
lily pads of a swamp where it was hunting frogs, the other was asso- 
ciated with a couple of Spotted Wood Snakes (Philothamnus s. semi- 
variegatus) in a small doom palm growing in knee-deep water in this 
same flood area. 

Chlorophis irregularis (Leach) 

Coluber irregularis Leach, 1819, in Bowdich, Miss. Ashantee, p. 494: Ashanti, 

Gold Coast. 
Ahaetulla emini Giinther, 1888, Ann. Mag. Nat. Hist. (6), 1, p. 325: Mon- 

buttu, Belgian Congo. 
Chlorophis sc/iubo/zi Sternfeld, 1912, Wiss.Ergebn. Deut. Zentral-Afrika-Exped., 

1907-1908, 4, p. 269, fig.: Bwanja, near Bukoba, Tanganyika Territory. 

5 (M. C. Z. 40523-7) Mt. Debasien, U. 18-30.xi.33. 

6 (M. C. Z. 40510-4, 40529) Kaimosi, K. C. 12.ii-l.iii.34. 

Native name. Emuri (Karamojong), but not specific. 

Variation. Midbody scale-rows 15; ventrals 161-184; anal divided 
except in M. C. Z. 40525 in which it is single though this snake is 
undoubtedly correctly referred to irregularis; subcaudals 98-122; 
labials 9, the 4th, 5th and 6th entering the orbit on seventeen sides 
(8 snakes normal), or 8, the 3rd, 4th and 5th entering the orbit on six 
sides, or 7, the 3rd, 4th and 5th entering the orbit on one side; preocu- 
lars 1, or 2 in M. C. Z. 40525, in contact with, or separated from the 
frontal; postoculars 2, or 4 in M. C. Z. 40524; temporals 1 + 1 on 
nineteen sides, 1 + 2 on three sides. 

Measurements. The largest cf (M. C. Z. 40523) measures 769 (548 
+ 221) mm., the largest 9 (M. C. Z. 40514) measures 774 (547 + 
227) mm. 

Habitat. On Mount Debasien, I captured one snake as it was bask- 
ing on the bank of a dry water course at 6,000 feet, another in vegeta- 
tion on the bank of the Amaler River at 5,000 feet. 

Philothamnus semivariegatus semivariegatus Smith 

Philothamnus semivariegatus A. Smith, 1849, Illus. Zool. S. Africa, 3, pis. lix, 
lx, lxiv: Bushman's Flats and Kurrichane, S. Africa. 



248 bulletin: museum of comparative zoology 

1 (M. C. Z. 40528) Mount Debasien, U. 24.xi.33. 

1 (M. C. Z. 40530) Kibwezi, K. C. 23.iii.34. 

3 (M. C. Z. 40531-3) Mt. Mbololo, K. C. 18.iv.34. 
3 (M. C. Z. 40534-6) Lamu, Lamu Id., K. C. 8.V.34. 

3 (M. C. Z. 40537-9) Peccatoni, K. C. 26.V.34. 

4 (M. C. Z. 40540-3) Ngatana, K. C. ll.vi.34. 

2 (M. C. Z. 40544-5) Malindi, K. C. 30.vi.34. 

Distribution. Also from Sokoki Forest (H. J. A. T.). 

Native names. Ekumbu (Kitaita, but also applied to the Boom- 
slang); kongoani and ukutiwiti (Kiamu); hasowitu (Kipokomo). 

Variation. Midbody scale-rows 15, except for M. C. Z. 40530 
which has 13 though this snake exhibits the normal number an inch in 
advance of midbody; ventrals 164-196; anals 2; subcaudals 130-160; 
labials 8-10, though 8 only on the right side of M. C. Z. 40544 which 
has the 5th and 6th labials fused, and 10 on the right side of M. C. Z. 
40539, 5th and 6th labials entering the orbit except on the right side of 
M. C. Z. 40544 where it is the 5th only, and three snakes (Nos. 40533, 
40538, 40545) where the 4th, 5th and 6th enter either on both sides, 
or left only, or right only; preoculars 1, except M. C. Z. 40542 which 
has 2 ; postoculars 2, except on the right side of M. C. Z. 40528 where 
there are 3; temporals 2 + 2, except M. C. Z. 40535 which has 1 + 2 
and M. C. Z. 40534, 40536 which have 2+1, all these last three vari- 
ants being from Lamu Island. 

Measurements. The largest d 71 (M. C. Z. 40545) measures 1032 
(650 + 382) mm., the largest 9 (M. C. Z. 40534) measures 1233 
(785 + 448) mm. 

Breeding. Three eggs measuring 32 x 10 mm. were taken from a 
Lamu snake on May 8, 1934; three others measuring 25 x 6 mm. 
present in the Malindi female on June 30, 1934. 

Diet. x\ gecko (Lygodactylus p. gutturalis) in the young Spotted 
Wood Snake from Mount Debasien, two (L. p. v ombasicus) in the 
Kibwezi and one in the Mount Mbololo reptiles; two geckos (Hemi- 
dactylus persimilis) in a Lamu snake; two geckos (H. mabouia) in one 
from Malindi, a frog (Rana o. oxyrhynchus) in the other. 

Enemies. One was recovered from the stomach of the rare Banded 
Harrier Eagle (Circaetus fasciolatus) at Ngatana. 

Habitat. The Debasien snake was very cleverly ascending the trunk 
of a big fig tree; one Peccatoni snake was in a doom palm with Chloro- 
phis neglcctus, I caught another among water plants in waist-deep 
water, where it was presumably hunting frogs. Yet another was cap- 



• LOVERIDGE : AFRICAN REPTILES 249 

tured near Witu at 8 p.m., it was in a bramble bush within a foot of a 
tree frog (Chiromantis xerampelina) which it was apparently stalking. 
The bush was growing in knee-deep water and I had to reach up to 
seize the snake which I transferred to my handkerchief pocket, stuffing 
the handkerchief in on top to detain it: on reaching camp, however, 
I found that the snake had departed without my knowledge. 

Hapsidophrys lineata Fischer 

Hapsidophrys lineata Fischer, 1856, Abhand. Nat. Ver. Hamburg, 3, p. Ill, 
pi. ii, fig. 5: Elmine, West Africa, i.e. Elmina, Gold Coast. 

3 (M. C. Z. 40546-8) Kaimosi, K. C. 15-19.ii.34. 

Distribution. The Coryndon Museum has a specimen from Kedowa, 
on the Kenya-Uganda Railway, west of Kisumu. 

Variation. Midbody scale-rows 15; ventrals 158-166; anal 1; sub- 
caudals 93-98; labials 8, the 4th and 5th entering the orbit; preocular 1 ; 
postoculars-2-3; temporals 2 + 2, or 2 + 1 in M. C. Z. 40546. 

Measurements. The largest of these three females (M. C. Z. 40546) 
measures 974 (700 + 274) mm. 

Breeding. The large 9 held four eggs measuring 16x5 mm. on 
February 19. 

Diet. In her stomach was a frog (Ranaf. chapini). 

Thrasops jacksonii jacksonii Giinther 

Thrasops Jacksonii Giinther, 1895, Ann. Mag. Nat. Hist., (6), 15, p. 528: 

Kavirondo, Kenya Colony. 
Thrasops Rothschildi Mocquard, 1905, Bull. Mus. Paris, 11, p. 287: "Afrique 

orientale anglaise." 

3 (M. C. Z. 40680-2) Sipi, Mt. Elgon, U. 14-15.xii.33. 
• 5 (M. C. Z. 40551-2, 40683-5) Butandiga, U. 8-ll.i.34. 
3 (M. C. Z. 40550, 40686-7) Kaimosi, K. C. 25-28.ii.34. 

Native names. Yakobe for black adults, isilukanga for dark olive 
adults and half-grown (Lugishu). 

Synonymy. The type of jacksonii had 19 midbody scale-rows; 198 
ventrals; 138 subcaudals. The type of rothschildi, which may have 
come from either Kenya or Uganda, had 17 midbody scale-rows; 187 
ventrals; 141 subcaudals. Both types were uniformly black. 

Peracca (1909, p. 172) shows the extraordinary variation in head 
shields exhibited by a pair (d 71 9 ) obtained at Toro, Uganda. 



250 bulletin: museum of comparative zoology 

Schmidt (1923, p. 85, Fig. 6) gives by far the best account of varia- 
tion in jacksonii, his remarks being based on eighteen snakes from the 
Belgian Congo. It is important to note that of these, sixteen had 19 
midbody scale-rows, one had 17, one had 21; ventrals 192-211; sub- 
caudals 135-155. His figure shows that there is in reality no difference 
in rostral and frontal characters between jacksonii and rothschildi or 
Lonnberg's snakes from Mount Kenya. 

In each of the series listed above — Sipi, Butandiga and Kaimosi — 
there is one snake with 17 midbody scale-rows, the others having 19. 
This appears to show a tendency in the eastern part of the range for a 
reduction in the number of midbody scale-rows. One of the Kaimosi 
series (M. C. Z. 40550), coming as it does from a spot bordering on the 
type locality of jacksonii, has a scale count almost identical with that 
of the type of rothschildi. It has 17 midbody scale-rows; 190 ventrals; 
141 subcaudals. I consider, therefore, that rothschildi is a synonym of 
jacksonii. 

Variation. Midbody scale-rows 17-19; ventrals 188-202; anal 
divided; subcaudals 130-144; labials 8, the 4th and 5th entering the 
orbit, except on left side of M. C. Z. 40686 which has 9 with 4th and 
5th entering, and M. C. Z. 40552 which has 9 with 5th and 6th enter- 
ing on both sides of the head; preoculars 1, or 2 in three snakes only; 
postoculars 3; temporals 1 + 1. 

Measurements. The largest d" (M. C. Z. 40681) measures 1637 
(1140 + 497) mm., the largest 9 (M. C. Z. 40683) measures 1797 
(1275 + 522) mm., but is surpassed in length from snout to anus by 
38 mm. in a snake with mutilated tail. 

Breeding. The undermentioned records of developing eggs were 
taken. 

Sipi, December 14, 1933, a 9 held 12 eggs measuring 19 x 8 mm. 

Butandiga, January 8, 1934, " 10 " " 31 x 16 mm. 

8, 1934, " 7 " " 35 x 8 mm. 

11, 1934, " 8 " " 23 x 9 mm. 

Diet. One Sipi snake had mouse fur in its stomach, the Butandiga 
specimens (1) a tree rat (Cenomys b. editus), (2) bird, (3) Chamaeleon 
senegalensis, (4) Chamaeleon b. hbhnclii. 



CORONELLA SEMIORNATA SEMIORNATA Peters 

Coronella semiornata Peters, 1854, Monatsb. Akad. Wiss. Berlin, p. 622: Tete, 
Mozambique. 



loveridge: African reptiles 251 

Zamenis fischeri Peters, 1879, Monatsb. Akad. Wiss. Berlin, p. 777: Malindi, 

Kenya Colony. 
Coronella inornata Fischer, 1884, Jahr. Hamburg, Wiss. Anst. 1, p. 6, pi. i, 

fig. 2: Masailand, near Arusha, Tanganyika Territory. 
Coronella scheffleri Sternfeld, 1908, Sitzber. Ges. Naturf. Freunde Berlin, p. 93; 

and 1908, Mitt. Zool. Mus. Berlin, 4, p. 243, figs. 1 and 2: Kibwezi, Kenya 

Colony. 
Coronella semiornata var. mossambicae Cott, 1935 (1934), Proc. Zool. Soc. 

London, p. 967: Charre and Fambani, Mozambique. 

9 (M. C. Z. 40553) Kibwezi, K. C. 29.iii.34. 

Synonymy. C. scheffleri was based on an individual which was said 
to differ from semiornata in its more projecting snout, the preocular 
being separated from the frontal, the rostral (originally stated to be as 
deep as broad) only slightly broader than deep, the more numerous 
subcaudals. 

C. scheffleri had 91 subcaudals, the range for semiornata given by 
Boulenger (1894, p. 195) who only had three specimens, 63-88, my 
topotype of scheffleri has 88, other Kenya-Tanganyika material I have 
examined range from 71-91. The latter figure being reached by a 
snake from the Ulukenya Hills which I referred to (1929, p. 26) as an 
intermediate. The character of the preocular being separated from, 
or in contact with, the frontal is of no significance in this species where 
both conditions occur in the same locality, even in the same specimen 
(M. C. Z. 23067). From Sternf eld's figures I am inclined to think 
that it is the frontal that is shortened rather than the snout which is 
lengthened in his type, the frontal length in scheffleri equals its dis- 
tance from the rostral whereas normally in all our semiornata it is 
greater than its distance from the end of the snout. The topotype is 
normal in this respect in view of which I think that scheffleri may be 
considered a slightly abnormal individual. 

In describing mossambicae Colt was under the assumption that his 
specimens resulted in a considerable extension of the range of semior- 
nata. He overlooked the fact that the type came from Tete, some 125 
miles from Charre. His specimens agree well with the fine colored 
plate xvii, fig. 2 furnished by Peters in the Reise nach Mossamb., 3, 
(1882), but indeed the markings on which alone he bases the race are 
subject to considerable variation by the intermediate bars becoming 
obsolescent. 

Measurement. This 9 measures 484 (366 + 118) mm. 

Breeding. She held two eggs measuring 40 x 8 mm. in her oviducts. 



252 bulletin: museum of comparative zoology 

Variation. Midbody scale-rows 21; ventrals 196; anals 2; subcau- 
dals 88; labials 8, the 4th and 5th entering the orbit on the right, or 9, 
the 5th and 6th entering the orbit on the left; preoculars 2 and 1 
(right and left respectively) ; postoculars 2 ; temporals 2 + 3 (right, like 
schcffleri) or 2 + 2 (left, like semiornata). 

Kibwezi is on the edge of the red laterite country so that it is not 
surprising to find this snake presenting an anomalus scalation inter- 
mediate between the typical form, with which it agrees in coloring, 
and the recently described race C. s. fuscorosea. 

CORONELLA SEMIORNATA FUSCOROSEA Loveridge 

Coronella semiornata fuscorosea Loveridge, 1935, Bull. Mus. Comp. Zool., 79, 
p. 8: Mount Mbololo, Taita, Kenya Colony. 

3 (M. C. Z. 40555-7) Mt. Mbololo, K. C. 25.iv.34. 
1 (M. C. Z. 40554) Tsavo, K. C. 2.iv.34. 

Affinities. When describing this snake, it was with some misgivings 
that I referred it to the genus Coronella. While its nearest relative 
actually appears to be C. semiornata, the question arises as to whether 
the tropical African snakes semiornata, coronata, etc., are really con- 
generic with the palearctic austriaca, amaliae and girondica which repre- 
sent true Coronella, austriaca being the genotype. These paleartic 
species have a broad, shield-shaped frontal. 

The tropical African species, and Dr. Malcolm Smith informs me 
that the Indian species brachyura agrees with them in this respect, 
have a more elongate, concave-sided frontal, in which they agree with 
Coluber fiorulentus of North Africa and Northern Kenya as well as 
with the genotype Coluber c. constrictor of North America. 

Boulenger (1894, p. 3) separated Zamenis (i.e. Coluber for our pur- 
poses) from Coronella on the following characters: 
Head elongate, distinct from neck. Subocular present ...... Zamenis 

Head not or but slightly distinct from neck. No subocular. . Coronella 

The head of fiorulentus is unappreciably longer and not more dis- 
tinct from the neck than is that of semiornata of fuscorosea. The scale 
termed subocular in constrictor and fiorulentus by Boulenger, might 
equally well be called a lower preocular, Dr. E. R. Dunn suggests sub- 
preocular for this scale which I called a lower preocular when describing 
fuscorosea. 

The markings on the young Tsavo paratype (they disappear entirely 
in the adult) bear so strong a resemblance to those of fiorulentus that 



loveridge: African reptiles 253 

I cannot help thinking that fuscorosea is derived therefrom, forming a 
connecting link between that species and semiomata and the other 
socalled Coronellae forming the tropical African group. The cranial 
characters which Boulenger (1913, p. 47) subsequently utilized in dis- 
tinguishing Zamenis and Coronella also lead one to suspect that these 
tropical African forms have little in common with Coronella of Europe 
and North Africa. 

I am not transferring them to Coluber at the present time for such 
action should be based on a thorough revisionary study of the whole 
genus Coluber and certain closely related genera. Precipitate action 
might result in ultimate confusion. 

Habitat. The juvenile, taken on the banks of the Tsavo River near 
the station, had its head protruding from a hole at the base of a leaf- 
less, scrubby bush. It promptly withdrew into the burrow from which 
I dug it out. According to the natives, no rain had fallen; for a year, 
and though only about 9 a.m. the weather was intensely hot. 



Coronella coronata (Schlegel) 

Calamaria coronata Schlegel, 1834, Phys. Serp., 2, p. 46: Gold Coast. 
Meizodon regularis Fischer, 1856, Abhand. Nat. Ver. Hamburg, 3, p. 112, pi. 

iii, fig. 3: Peki, Gold Coast. 
Coronella regularis praeornata Angel, 1933, "Les Serpentes de l'Afrique occi- 

dentale Francaise." Paris, p. 123: West Africa and Uganda. 

1 (M. C. Z. 40323) Kau, Tana River, K. C. 4.vi.34. 

1 (M. C. Z. 40559) Ngatana, Tana River, K. C. 12.vi.34. 

1 (M. C. Z. 40558) Golbanti, Tana River, K. C. 22.vi.34. 

Distribution. These three snakes constitute the first records of the 
occurrence in Kenya of coronata (or regularis) but are in keeping with 
other elements of a West African fauna surviving along the Tana River 
close to the Indian Ocean. 

Affinities. Boulenger (1894, p. 190) was illogical in keeping regu- 
laris distinct from coronata. He separated them as follows: 
Four lower labials in contact with an anterior chin-shield ; belly yellow 

coronata 
Five lower labials in contact with an anterior chin-shield; belly blackish 

regularis 
yet of its East African ally, C. semiomata, he writes (p. 195) : 

"four or five lower labials in contact with the anterior chin-shields 
ventrals yellowish, uniform or edged with black." 



254 bulletin: museum of comparative zoology 

As previously stated (1929, p. 27), I agreed with Schmidt (1923, 
p. 87) that regularis should probably be united with coronata. The 
additional evidence furnished by the Tana River snakes makes this 
necessary for they continue the combination of the alleged key char- 
acters to which Schmidt and I have already invited attention. 

The three snakes listed above agree with regularis in having five 
lower labials in contact with an anterior chin-shield, but with coronata 
in having immaculate bellies only impinged upon laterally by the 
olive dorso-lateral coloration. 

More recently, Angel has named a striking color variant whose range, 
however, appears to be coextensive with that of regularis and therefore 
without geographical significance. I would suggest that his two speci- 
mens may be examples in which the juvenile coloring has persisted into 
later life. 

Variation. Midbody scale-rows 19; ventrals 168-176; anal divided; 
subcaudals 65-74; labials 8, the 4th and 5th entering the orbit; pre- 
ocular 1 ; postoculars 2-3 ; temporals 1 + 2. 

Coloration. So closely did these sriakes resemble Natrix o. olivacea 
that I failed to distinguish them in the field. The absence of dark pig- 
mentation on the labial sutures and the much more numerous ventrals 
serve at once to distinguish them. 

It might also be remarked that the coloring of these Tana River 
coronata is duplicated by that of a Dar es Salaam specimen of semi- 
ornata which, moreover, has 1 + 2 temporals on the right side of the 
head; on the left, however, it has the normal 2 + 2 and its midbody 
scale-rows are 21. 

Measurements. The largest c? (M. C. Z. 40558) measures 322 
(247 + 75) mm., the 9 (M. C, Z. 40323) measures 520 (410 + 110) 
mm. 

Prosymna ambigua stuhlmanni (Pfeffer) 

Ligonirostra stuhlmanni Pfeffer, 1893, Jahr. Hamb. Wiss. Anst., 10, p. 78, pi. i, 

figs. 8-10: Usambara, Tanganyika Territory. 
Prosymna variabilis Werner, 1909, Jahr. Nat. Ver. Wurttemb., 65, p. 57: Moshi, 

Tanganyika Territory. 

9 (M. C. Z. 40560) Mapenya, near Witu, K. C. 28.V.34. 
& (M. C. Z. 40561) Mkonumbi, K. C. 28.V.34. 
c? 9 (M. C. Z. 40562-3) Ngatana, K. C. 20.vi.34. 

Distribution. To the best of my belief these specimens constitute 
the first record of the genus in Kenya Colony. 



loveridge: African reptiles 255 

Affinities. Boulenger (1894, p. 248) synonymized stuhlmanni with 
ambigua from Angola. The type of the latter, however, had 17 midbody 
scale-rows while the cotypes of stuhlmanni, as well as a dozen speci- 
mens from East Africa which I have examined, have 15. Furthermore, 
Mr. H. W. Parker writes on February 26, 1935 that there are 16 speci- 
mens in the British Museum from Zanzibar; Mombasa; Eastern 
Province, Uganda; Shire Valley; Mazoe, Rhodesia; and Kosi Bay, 
Zululand all with 15 midbody scale-rows. It seems reasonable, there- 
fore, to recognize an eastern race characterized by 15 midbody scale- 
rows. Those with 15 range to Garamba, Belgian Congo (Schmidt, 
1923, p. 89). 

Barbour and Loveridge (1928, p. 121) and I (1933, p. 244) have 
referred to this matter before, suggesting that the name bocagii Bou- 
lenger from the Congo might be used. Apart from the fact that stuhl- 
manni has four years' precedence, I note that bocagii has a slightly 
more numerous ventral count than stuhlmanni and must be regarded 
as distinct. 

Variation. Midbody scale-rows 15; ventrals 134-152; anal 1; sub- 
caudals 20-32; labials 6, the 3rd and 4th entering the orbit; preocular 
1, so minute on the left side of the head in M. C. Z. 40562 as to permit 
the prefrontal entering the orbit as in bocagii; postoculars 2 or 1 in 
M. C. Z. 40563 only; temporals 1 + 2 except in M. C. Z. 40560 where* 
it is 1 + 1 as the lower temporal is fused with a labial. 

Sexual differences. Sexes of this species, as judged by the counts 
of eight males and seven females, may be readily distinguished by the 
ventral and subcaudal range, viz. 

Males have 133-140 ventrals and 30-34 subcaudals. 
Females have 140-152 ventrals and 19-24 subcaudals. 

Breeding. The Mapenya 9 held three eggs measuring 20 x 6 mm. 
on May 28, 1934. 

Measurements. The larger cf (M. C. Z. 40562) measures 238 
(200 + 38) mm., the larger 9 (M. C. Z. 40563) measures 254 (232 + 
22) mm. 

Scaphiophis albopunctatus Peters 

Scaphiophis albopunctatus Peters, 1870, Monatsb. Akad. Wiss. Berlin, p. 645, 
pi. i, fig. 4: Kita, Guinea (i.e. French West Africa). 

2 (M. C. Z. 39953-4) Sokoki Forest, K. C. 1932. 

These two juvenile snakes are from a series of ten collected by Mr. 
H. J. Allen Turner, the rest are in the Coryndon Memorial Museum, 
Nairobi. 



256 bulletin: museum of comparative zoology 

Affinities. I would suggest that the recently described S. calciatii 
Calabresi from near Cunama, Eritrea is a synonym of S. raffreyi 
Bocourt from Ethiopia, and that Boulenger (1894, p. 254) was in error 
in referring the latter to the synonymy of albopunctatus. Both Bocourt 
and Calabresi stress the same points of difference between their new 
species and albopunctatus. 

Variation. Midbody scale-rows 23; ventrals 192-207; anals 2; 
subcaudals 57-69; labials 5; suboculars 3; preocular 1 ; postoculars 2 or 
3 on right side of M. C. Z. 39954; temporals 4 + 5 or 5 + 5 on left side 
of M. C. Z. 39953; lower labials in contact with an anterior chin 
shield 3. 

Coloration. Pale gray flecked and spotted with white, this juvenile 
coloring is very different from that of the adult. 

Measurements. The larger, apparently a 9 , measures 402 (342 + 
60) mm. 

DASYPELTINAE 

Dasypeltis scaber (Linnaeus) 

(Plate 4, fig. 1) 

Coluber scaber Linnaeus, 1766, Syst. Nat., 1, p. 384: Indiis. 

10 (M. C. Z. 40564-73) Sipi, U. 13-23.xii.33. 
1 (M. C. Z. 40574) Butandiga, U. 12.1.34. 
1 (M. C. Z. 40575) Elgonyi, K. C. 30.L34. 
4 (M. C. Z. 40576-9) Kaimosi, K. C. 10-16.ii.34. 

1 (M. C. Z. 40580) Mt. Mbololo, K. C. 28.iv.34. 

2 (M. C. Z. 40581-2) Lamu Id., K. C. 12.V.34. 
1 (M. C. Z. 40583) Near Witu, K. C. 30.V.34. 

1 (M. C. Z. 40584) Laini, Tana R., K. C. 6.vi.34. 
1 (M. C. Z. 40585) Ngatana, K. C. 16.vi.34. 

Distribution. Also 1 from Sokoki Forest (H. J. A. T.). 

Native names. Namagi (Lugishu); lucha (Kipokomo). 

Variation. The Mount Mbololo snake differs from all the others in 
having 2 preoculars and 3 + 5 temporals and in its coloring which 
was a bright coral pink above, paler pink below. This coloring harmo- 
nizes with the reddish soil of the region. The data of the rest of the 
series is as follows. 

Midbody scale-rows 22-25; ventrals 208-232; anal 1; subcaudals 
50-96, this last count, which I have rechecked, surpasses the range 
given by Boulenger (1894, p. 355), is of the cf snake from Lamu; 
labials 7, the 3rd and 4th entering the orbit, or 6 due to the fusion of 



loveridge: African reptiles 257 

5th and 6th in M. C. Z. 40576 and on left side of M. C. Z. 40578; 
preocular 1 except as noted above; postoculars 1-2, five Sipi and one 
Kaimosi snake have a single postocular, twelve snakes have 2, the 
rest are azygous intermediates; temporals 1 + 3 (M. C. Z. 40579 
only), 2 + 2 on ten sides, 2 + 3 on twenty-seven sides, 2 -j- 4 on four 
sides, 3 + 5 (M. C. Z. 40580 only). 

Coloration. Wholly black specimens at Sipi and Kaimosi; uniform 
brown above at Sipi and Butandiga; bright coral pink on Mount 
Mbololo; uniform gray on Lamu Island; rhomboidal dorsal markings 
(var. B. of Boulenger) at Laini and Ngatana; rhomboidal dorsal mark- 
ings and also ventrals conspicuously edged with black at Witu ; rhom- 
boidal dorsal markings confluent with lateral bars (var. E) on the 
young Elgonyi snake. 

Measurements. The largest cf (M. C. Z. 40570), a brown one, 
measures 615 (518 + 97) mm., the largest 9 (M. C. Z. 40564), a 
black one, measures 825 (725 + 100) mm. 

Breeding. At Sipi a female held numerous eggs measuring 16x5 
mm. on December 23, 1933; at Kaimosi another held fourteen eggs 
measuring 17x8 mm. on January 16, 1934. 

Diet. One large Sipi female had her stomach and intestines distended 
with quantities of yolk, in bulk about equalling the contents of three 
hen's eggs. 

Parasites. She was heavily parasitized with encysted nematodes. 



BOIGINAE 

Crotaphopeltis hotamboeia hotamboeia (Laurenti) 

Coronella hotamboeia Laurenti, 1768, Syn. Rept., p. 85: India orientali, i.e. 
Africa. 

1 (M. C. Z. 40586) Mt. Debasien, U. 20.xi.33. 

2 (M. C. Z. 40587-8) Mt. Mbololo, K. C. 17.iv.34. 

1 (M. C. Z. 40583) Lamu Island, K. C. 14.V.34. 

2 (M. C. Z. 40590-1) Peccatoni, K. C. 26.V.34. 
2 (M. C. Z. 40592-3) Near Witu, K. C. 31.V.34. 
1 (M. C. Z. 40594) Kau, Tana R., K. C. 4.vi.34. 

18 (M. C. Z. 40595-611) Ngatana, K. C. 16-21.vi.34. 
7 (M. C. Z. 40612-7) Golbanti, K. C. 22.vi.34. 

Distribution. Seen also at Kibwezi and Malindi. 
Native names. Bafu (Kitaita); goomugalla (Kipokomo). See also 
remarks under Natrix o. olivacea. 



258 bulletin: museum of comparative zoology 

Variation. Despite the amazing amount of variation displayed by 
this coastal series of the White-lipped Snake, they are consistently of 
the typical lowland form with 19 midbody scale-rows and usually 1 
preocular and 2 postoculars. 

Midbody scale-rows 19; ventrals 157-174; anal single; subcaudals 
40-65, Boulenger (1894, p. 90) gives 32-54 as thesubcaudal range but 
half the above series are between 55-65; labials 6-9, all but nine sides, 
however, are the normal 8, the 3rd and 4th (two sides), the 3rd, 4th 
and 5th (15 sides), the 4th and 5th (42 sides), the 4th, 5th and 6th 
(5 sides), or the 5th and 6th (2 sides) entering the orbit, it will be 
observed that there is a strong tendency in this region for the 4th and 
5th only to enter the orbit, this is in marked contrast to the thirty- 
four snakes collected by me (1933, p. 247) in the Central Lake Region 
which, without exception, had the 3rd, 4th and 5th entering the orbit 
as is normal for the species according to Boulenger (1896, p. 89); 
preoculars 1-2, the latter on the right side only of a Lamu snake and 
both sides on two Ngatana and a Golbanti reptile; postoculars 2-3 
the latter on right side only of M. C. Z. 40588; frontal invariably well- 
separated from the preocular by the supraocular; temporals 1 + 1 
(6 sides), 1 + 2 (59 sides), 1 + 3 (1 side); lower labials in contact with 
an anterior chin-shield 4 (9 sides), 5 (54 sides) or 6 (3 sides). 

Coloration. These snakes do not display the white flecks so usual in 
snakes from further south. When about to cast their epidermis it 
becomes a smoky blue-gray, obscuring the usual black ground color. 

On capturing the first Mbololo snake, though at 3,800 feet in a rotten 
log at the edge of the rain forest, I observed that the iris of its eye was 
grayish olive, not red, so was quite prepared on later examination to 
find that, with the exception of three postoculars on the right side of 
its head, its scale characters were those of the typical, or lowland, 
form. 

Measurements. The largest c? (M. C. Z. 40595) measures 564 (470 
+ 94) mm., the largest 9 (M.C.Z. 40587) measures 519 (446 + 73) mm. 

Sex. This does not appear to be determinable from scale counts, 
thus fifteen males have a ventral count of 157-166 and subcaudal of 
44-65; twelve females have from 159-174 ventrals and 40-54 sub- 
caudals. 

Breeding. The Kau 9 held several eggs measuring 13 x 4 mm. on June 4. 

An Ngatana 9 " 4 " " 27x12 mm. " 16. 

9 " 6 " " 17 x 7 mm. . " 17. 

" 9 " 3 " " 32 x 8 mm. " 17. 

A Golbanti 9 " 5 " " 35x11 mm. " 26. 



loveridge: African reptiles 259 

It should be added that only two of the batch of three eggs reached 
the dimensions given, the third egg was considerably smaller. Con- 
trary to what one usually finds, the female producing the small com- 
pliment of four eggs was of large size, measuring 500 mm. over all. 

Diet. A Peccatoni snake held amphibian bones; one from Ngatana 
had swallowed both a frog (Rana m. mascareniensis) and a toad {Bufo 
stein dachnerii); toads of this species were also recovered from the 
stomachs of two Golbanti reptiles. 

Parasites. A tapeworm (Ophiotaenia crotaphopeltis) was taken from 
the alimentary tract of a Golbanti snake in whose stomach were worms. 

Habitat. The young Debasien snake was cut in half by one of my 
men engaged in cutting down grass on the bank of the i\.maler River at 
5,000 feet, another of similar size was found on the river bank, dead 
and macerated. The Lamu specimen was taken in a garden. In addi- 
tion to those captured at Peccatoni, was a young one which escaped 
me, it was lying at the base of a bussu palm in a swampy area flooded 
by the recent heavy rains. 

Chamaetortus aulicus aulicus Giinther 

Chamaetortus aulicus Giinther, 1864, Proc. Zool. Soc. London, p. 310, pi. xxvi, 
fig. 2: Zambesi. 

9 (M. C. Z. 40619) Kibwezi, K. C. 30.iii.34. 
& (M. C. Z. 40620) Ngatana, K. C. 17.vi.34. 

Distribvtio?i. This rare snake has previously been recorded from 
Kibwezi by Sternfeld (1908, p. 244), constituting, I believe, the first 
record of its occurrence in Kenya Colony. A second example, now 
in the Coryndon Memorial Museum, was obtained by Mr. H. J. Allen 
Turner at Fundi Isa, north of Malindi, in June 1932. 

Native name. Tambasi (Kipokomo). 

Affinities. Trinomials are employed on account of C. a. ellenbergeri 
Chabanaud, 1917 (1916) from Lambarene, Congo. It differs from the 
typical form in the preocular being in contact with the frontal; labials 
9, the 4th, 5th and 6th entering the orbit; ventrals 209; subcaudals 95 
and its color, a uniform greyish brown. 

Variation. Midbody scale rows 17; ventrals 178-187; anal single; 
subcaudal 93 on 9 ; labials 8, the 3rd, 4th and 5th entering the orbit; 
preocular 1 + the loreal which enters the eye; postoculars 2 ; temporals 
1 + 1 ; lower labials in contact with an anterior chin shield 4-5. 

Measurements. The d 71 has the tip of its tail lacking. The 9 meas- 
ures 607 (458 + 149) mm. 



260 bulletin: museum of comparative zoology 



Hemirhagerrhis kelleri Boettger 

Hemirhagerrhis kelleri Boettger, 1893, Zool. Anz., 16, p. 129: Webi Jhal, 
Abdallah and south of Ogaden, Ethiopia. 

Amplorhinus taeniatus Sternfeld, 1908, Mitt. Zool. Mus. Berlin, 4, p. 244, fig. 
3. and 1910, Fauna der deutschen Kolonien, 3, Die Schlangen Deutsch- 
Ostafrika, p. 27 (reprint), fig. 27: Lamu Island, Kenya Colony. 

1 (M. C. Z. 40621) Mkonumbi, K. C. 28.V.34. 

Distribution. This specimen is almost topotypic of Sternfeld's A. 
taeniatus, for Mkonumbi is on the mainland almost opposite Lamu 
Island and there is a daily dhow service between the two during 
most of the year. 

The specimen of kelleri listed by Boulenger (1896, p. 649) as collected 
by the late Sir F. J. Jackson in Uganda, is now M. C. Z. 26928. In 
view of the fact that Jackson visited Witu, eighteen miles inland from 
Mkonumbi, and that other herpetological material which he collected 
on his journey to Uganda was erroneously attributed to 'Uganda' 
(e.g. Chamaeleon jacksoni Boulenger from Kikuyu), I think that we are 
justified in believing that this snake was taken in what is now called 
Kenya Colony. I mention this here as my reason for omitting 
kelleri from the Uganda list. 

A snake in the Nairobi Museum (No. I 100) without locality, was 
erroneously referred by me (1916, pp. 80 and 85) to H. kelleri. On 
recent examination I find that it is an example of Pseudoboodon 
lemniscatus of Ethiopia. Whether it was taken on the Kenya side oft he 
Kenya-Ethiopian boundary is of course uncertain, though likely, but 
until substantiated P. lemniscatus cannot be admitted to the Kenya 
list. 

Sternfeld (1908a, loc. cit. supa. p. 243), in his introduction to the 
paper in which he described Amplorhinus taeniatus and Rhinocalamus 
meleagris makes acknowledgement to various authors for material, 
among them "Denhardt (Lamu, Pokomonie)." Though he gives 
localities for the thirty-five other species mentioned in the paper, 
strangely enough he omits mention of the type localities of the two 
mentioned above. Later in the year, however, he added "Lamu 
Island." 

It should be remembered that "Lamu" formerly was applied to the 
mainland opposite the island, later being called "Lamu district," at 
the same time it was applicable to the township of Lamu, Lamu Is- 
land. "Pokomonie" which literally means the "place of the Pokomo 



loveridge: African reptiles 261 

tribe" is the name of a creek close to Wange where Denhardt had an 
abortive coconut plantation. This is that Wange which Stejneger 
(1893, p. 712) was misled in placing on Manda Island opposite the 
creek. No Wapokomo live on the creek today, the tribe being re- 
stricted to a narrow strip bordering the Tana River from Kidori 
(1 S.) to Charra 2° 30 ' S.) near the mouth of the river. This is the 
region generally known today as Pokomoni. 

While it is true that Gustav Denhardt's headquarters were on Lamu 
Island, I visited his plantation and house, the latter in a semi-ruinous 
state, he and his brother had numerous interests on the mainland in 
addition to the Wange venture. I learned that he made several busi- 
ness trips up the Tana River. While it is entirely possible that the 
snakes which became the types of taeniatus and meleagris may have 
been accidentally introduced into Lamu, possibly with the fuel which 
constituted another of Denhardt's business undertakings, it seems 
more probable that a burrowing type like Rhinocalamus would have 
been taken at Wange, along the Pokomoni creek, or in the Pokomo 
country bordering the Tana. 

The fact that I spent a week in searching for these two species on 
Lamu Island without finding them, is no proof that they do not occur 
there, but in view of the vagueness attaching to the localities on other 
Denhardt material, it is at least possible that they were taken at 
Lamu in its mainland sense, if not Pokomonie. 

Variation. Midbody scale-rows 17; ventrals 157; anals 2; subcaudals 
78; labials 8, the 4th and 5th entering the orbit; preoculars 2 on the 
right, 1 on the left side; postoculars 2; temporals 2 + 3. 

Measurements. Total length 212 (162 + 50) mm. 

Rhamphiophis rubropunctatus (Fischer) 

Dipsina rubropunctata Fischer, 1884, Jahr. Hamburg. Wiss. Anst., 1, p. 7 f 
pi. i, fig. 3: Near Arusha at the foot of Kilimanjaro, Tanganyika Territory, 

1 (M. C. Z. 40650) Voi, K. C. 23.iv.34. 

9 (M. C. Z. 40651) Mt. Mbololo, K. C. 27.iv.34. 

Native names. The Wataita distinguish the spotted young from the 
uniform adult under the names of manganga (young) and ngunuku 
(adult). 

Variation. Midbody scale-rows 19; ventrals 235-237; anal divided; 
subcaudals 149-154; labials 8-9, the 4th and 5th or 5th and 6th 
entering the orbit; preoculars 2; postoculars 2; loreal 1; temporals 
2 + 4, 3 + 4, or 4 + 4. 



262 bulletin: museum of comparative zoology 

Measurements. The 9 measures 1608 (1092 + 516) mm., and 
appears to be the largest recorded example of this rare species. 

Rhamphiophis rostratus Peters 

Rhamphiophis rostratus Peters, 1854, Monatsb. Akad. Wiss. Berlin, p. 624; 
1882, Reise nach Mossamb., 3, p. 124, pi. xix, fig. 1: Tete; Mesuril; Quitan- 
gonha, Mozambique. 

2 (M. C. Z. 40652-3) Voi, K. C. 24-27.iv.34. 

3 (M. C. Z. 40654-6) Lamu Island, K. C. 7.V.34. 
1 (M. C. Z. 40657) Peccatoni, K. C. 26.V.34. 

1 (M. C. Z. 40658) Mkonumbi, K. C. 28.V.34. 
1 (M. C. Z. 40659) Witu, K- C. 31.V.34. 

Distribution. Also 2 from Sokoki Forest ( (H. J. A. T.) 

Native names. Kitangu (for adults) and mbono (for spotted young) 
hi Kiamu. 

Variation. Midbody scale-rows 17; ventrals 154-187; anal divided; 
subcaudals 90-118; labials 8, the 5th only entering the orbit; pre- 
oculars 3 (ten sides) or 4 (six sides); postoculars 2 except on the left 
side of M. C. Z. 40659 where there are 4; temporals 2 + 3, rarely 
2 + 4 (three sides). 

Measurements. The largest c? (M. C. Z. 40659) measures 1025 
(715 + 310) mm., the largest 9 (M. C. Z. 40652) measures 1334 
(922 + 412) mm. 

Diet. A very young snake from Peccatoni had the limbs of a frog 
(? Kassina senegalensis) in its stomach. 

Psammophis siBiLANS (Linnaeus) 

Coluber sibilans Linnaeus (part), 1766, Syst. Nat., 12th ed., 1, p. 383: "Asia". 

5 (M. C. Z. 40622-6) Sipi, U. 13-27.xii.33. 
1 (M. C. Z. 40627) Butandiga, U. 8.i.34. 

4 (M. C. Z. 40628-31) Bukori, K. C. 18.i.34. 
1 (M. C. Z. 40632) Kaimosi, K. C. l.iii.34. 

1 (M. C. Z. 40633) Kibwezi, K. C. 23.iii.34. 

1 (M. C. Z. 40634) Voi, K. C. 12.iv.34. 

2 (M. C. Z. 40635-6) Mt. Mbololo, K. C. 17.iv.34. 
1 (M. C. Z. 40637) Golbanti, K. C. 3.V.34. 

1 (M. C. Z. 40642) Near Witu, K. C. 31.V.34. 
4 (M. C. Z. 40644-7) Ngatana, K. C. 12-15.vi.34. 
1 (M. C. Z. 40648) Malindi, K. C. 30.vi.34. 

Distribution. Also 4 from Sokoki Forest (H. J. A. T.) 



loveridge: African reptiles 263 

Native names. Namasanurugi (Lugishu); aerenet (Kimasai); 
ndasiangombe (Kitaita); juaka or paa (Kipokomo). 

Variation. Midbody scale-rows 17; ventrals 158-181; anal divided; 
subcaudals 81-106; labials 8, the 4th and 5th entering the orbit 
except on left side of M. C. Z. 40636 where there are 7, the 3rd and 
4th entering the orbit; preocular 1; postoculars 2; temporals 2 + 2 
on 24 sides, 2 -f- 3 on 20 sides; rostral as broad as deep in 13 snakes, 
broader than deep in 9, both conditions occurring in snakes from the 
same locality. 

Coloration. The ventrals of the forest-edge specimens from the 
Central African material (Sipi, Butandiga, Kaimosi) exhibit well- 
defined, though dusky, lateral lines, in the coastal specimens these 
are represented by dashes or dots except a young Ngatana snake 
which is pure white below. One of the Mbololo snakes (M. C. Z. 
40636) apparently represents Var. A. of Boulenger's "Catalogue of 
Snakes" (1896, p. 161) it is pure white below while the back is brown 
and striped like that of P. subtaeniatus. 

Measurements. The largest d" (M. C. Z. 40622) measures 1095 
mm. from snout to anus, the largest 9 (M. C. Z. 40624) measures 
1303 (935 4- 368 )mm. 

Breeding. Only three snakes in the above series held developing ova. 

Sipi, December 13, 1933. 9 held 4 eggs measuring 13 x 6 mm. 

Bukori, January 18, 1934. ? " 10 " " 27 x 10 mm. 

" 9 " 7 " " 38 x 19 mm. 

These last were quite ready for laying. 

Diet. Rodent fur in a Sipi snake; a striped mouse (Lemniscomys 
s. massaieus) in the Butandiga reptile; a tree rat (Oenomys b. editus) 
swallowed by the Kaimosi specimen, while a tree lizard (Agama 
atricoUis) was recovered from one of the Bukori series. 

Parasites. All these Bukori snakes were heavily infested with 
nematodes (Physaloptera paradoxa) as was the Kaimosi specimen. 
A few in one of the Mkonumbi snakes and one only in the Malindi 
reptile. 

Psammophis subtaeniatus Peters 

Psa?nmophis sibilans var. subtaeniatus Peters, 1882, Reise nach Mossamb., 3, 
p. 121 : Boror and inland from Tete, Mozambique. 

3 (M. C. Z. 40638-40) Lamu Island, K. C. 8.V.34. 
1 (M. C. Z. 40841) Mkonumbi, K. C. 28.V.34. 
1 (M. C. Z. 40643) Near Witu, K. C. 31.V.34. 
1 (M. C. Z. 40649) Changamwe, K. C. 5.vii. 34. 



264 bulletin: museum of comparative zoology 

Native name. Mchczawanawaki (Kiamu). The literal translation 
is "Plays with the women" the explanation being that when they 
are working in their gardens the women are scared and stampeded 
by the snake's appearance. The name was widely known and not 
the concoction of a native on the spur of the moment. 

Variation. Midbody scale-rows 17; ventrals 160-167; anal divided; 
subcaudals 100-113; labials 8, the 4th and 5th entering the orbit; 
4 lower labials in contact with the anterior chin shields; preocular 
1 or 2 on right side of M. C. Z. 40641 only; postoculars 2; loreal 1; 
temporals 2 + 2 on right side, 2 + 3 on left side of every snake 
except M. C. Z. 40638 which had 2 + 2 on both sides. 

Coloration. All have the pair of black lines along the belly sharply 
defined and clear. 

Measurements. The largest cf (M. C. Z. 40641) measures 1182 
(792 + 390) mm., the largest 9 (M. C. Z. 40643) measures 835 
mm. from snout to anus, the tail is truncated. 

Diet. A young frog {Rana edulis) was recovered from the stomach 
of a very emaciated Mkonumbi snake. 

Parasites. The other Mkonumbi snake had nematodes only in its 
stomach. 



Psammophis punctulatus Dumeril & Bibron 

Psammophis punctulatus Dumeril & Bibron, 1854, Erp6t. Gen., 7, p. 897: 

Arabia. 
Psammophis punctulatus var. trivirgatus Peters, 1878, Monatsb. Akad. Wiss. 

Berlin, p. 206: Taita, Kenya Colony. 

3 (M. C. Z. 40660-2) Mt. Mbololo, K. C. 20.iv.34. 

Native name. Ndasiangombe (Kitaita). 

Variation. Midbody scale-rows 17; ventrals 186-188; anal divided; 
subcaudals 118-152; labials 8, the 4th and 5th entering the orbit or 
9, the 5th and 6th; preocular 1 in contact with or separated from the 
frontal ; postoculars 2 ; loreal 1 ; temporals 1 + 2, 2 + 2 or 2 + 3. 

Boulenger (1896, p. 159) gives the subcaudal range as 130-158, I 
might add that I have verified the new low count of 118 on an unin- 
jured tail (M. C. Z. 40660). 

Measurements. The cf (M. C. Z. 40662) measures 883 (560 + 323) 
mm., the larger 9 (M. C. Z. 40660) measures 1530 (1030 + 500) mm. 

Diet. A lizard (Latastia I. revoili) was recovered from the stomach of 
one of these Spotted Sand Snakes. 



loveridge: African reptiles 265 

Psammophis biseriatus Peters 

Psammophis biseriatus Peters, 1881, Sitzb. Ges. Naturf. Freunde Berlin, p. 88: 
Taita, Kenya Colony. 

1 (M. C. Z. 40663) Tsavo, K. C. 4.iv.34. 

3 (M. C. Z. 40664-6) Voi, K. C. 7-10.iv.34. 

9 (M. C. Z. 40667-74) Mt. Mbololo, K. C. 17-30.iv.34. 

1 (M. C. Z. 40675) Malindi, K. C. 30.vi.34. 

Distribution. Mount Mbololo being in Taita (i.e. Utaita), the series 
from that locality are topotypes. 

Native name. Mararinga (Kitaita). 

Variation. Midbody scale-rows 15; ventrals 143-152; anal divided 
except M. C. Z. 40671 which has a single anal; subcaudals 92-125; 
labials 9, the 5th and 6th entering the orbit except on the right side 
of M. C. Z. 40671 which has the 4th, 5th and 6th entering; at first sight 
it would appear that Tanganyika snakes might be separated as having 
three labials entering the orbit, unfortunately this condition crops up 
elsewhere — in Somaliland for example; preoculars 1, or 2 in M. C. Z. 
40667 and on right side of 40666; postoculars 2; temporals 1 + 2 
(two sides), 1 + 3 (one side), 2 + 2 (eighteen sides), or 2 + 3 (seven 
sides); labials in contact with anterior chin-shield 5, or 4 on left side 
of M. C. Z. 40670 only. 

Measurements. The largest cf (M. C. Z. 40668) measures 872 (541 
+ 331), the largest 9 (M. C. Z. 40667) measures 1020 (660 + 360) 
mm. 

Diet. A very large lizard (Latastia I. revoili) in the Tsavo snake, a 
skink {Mabuya planifrons) in one from Voi, chameleons (C. d. roperi) 
in two of the series from Mount Mbololo. 

Thelotornis kirtlandii (Hallowell) 

L(eptophis) Kirtlandii Hallowell, 1844, Proc. Acad. Nat. Sci. Philad., p. 62: 
Liberia. 

4 (M. C. Z. 40676-9) Mt. Mbololo, K. C. 23-29.iv.34. 

Distribution. Also 4 from Sokoki Forest (H. J. A. T.) 

Native name. Mraringa (Kitaita). 

Variation. Midbody scale-rows 19; ventrals 164-172; anal divided; 
subcaudals 148-166; labials 8, the 4th and 5th or 3rd, 4th and 5th (in 
M. C. Z. 40676 only) entering the orbit; lower labials in contact with 
the anterior chin shields 4-5; preoculars 1; postoculars 3; temporals 
1 + 2.. 



266 bulletin: museum of comparative zoology 

Measurements. The largest cf (M. C. Z. 40676) measures 1422 
(821 + 601) mm., the largest 9 (M. C. Z. 40677) 1185 (713 + 472) 
mm. 

Dispholidus typus (Smith) 

Bucephalus typus A. Smith, 1829, Zool. Journ., 4, p. 441: Old Latakoo, South 
Africa. 

1 (M. C. Z. 40688) Butandiga, U. 8.i.34. 

3 (M. C. Z. 40689-91) Bukori, K. C. 18.i.34. 

8 (M. C. Z. 40692-9) Kaimosi, K. C. 12-28.ii.34. 

5 (M. C. Z. 40700-4) Mt. Mbololo, K. C. 19-28.iv.34. 

Distribution. A battered one also seen at Kibwezi, and others 
from Matalani (H. J. A. T.), Sokoki Forest (H. J. A. T.), and Fundi 
Isa (H. J. A. T.) in Nairobi Museum. 

Native name. Ikumbu (Kitaita, but also applied to the Spotted Wood 
Snake (Philothamnus s. semivariegatus.) 

Variation. Midbody scale-rows 19-21 (the latter on M. C. Z. 40702- 
4 only); ventrals 169-191; anal divided; subcaudals 87-119; labials 7, 
the 3rd and 4th entering the orbit, except on the right side of M. C. Z. 
40695 where the 3rd, 4th and 5th enter; preoculars 1; postoculars 3; 
temporals 1 + 2, or 1 + 1 on right side M. C. Z. 40700, 1 + 3 on 
M. C. Z. 40702, 2 + 3 on M. C. Z. 40694. 

Coloration. A small boy brought in an interesting pair of these Tree 
Snakes (M. C. Z. 40703-4). cf. Top of head bright brick red like the 
red soil of the region, back nut brown but the keel of each scale partly 
white. Below, buffy white. 9 . Top of head as in the male but the 
back reddish brown, tail bright pink. Below, pink. 

Green examples were taken at Bukori, Kaimosi and Mount Mbololo. 
Brownish ones at Butandiga, Bukori, Kaimosi and Mount Mbololo. 

Measurements. The largest cf (M. C. Z. 40703) measures 1298 
(962 + 336) mm., the largest 9 (M. C. Z. 40704) measures 1391 
(1073 + 318) mm. 

Diet. Chameleons, viz. C. b. hohnelii at Butandiga; C. g. gracilis in 
all three Bukori snakes; C. b. bitacniatus, four, in three Kaimosi snakes, 
C. d. ropcri in two Mount Mbololo specimens, bird's feathers in a third. 



Calamelaps unicolor (Reinhardt) 

Calamaria unicolor Reinhardt, 1843, Dansk, Vidensk. Selsk. Skrift., 10, p. 
236, pi. i, figs. 1-3: Guinea, West Africa. 



loveridge: African reptiles 267 

$ (M. C. Z. 40705) Mt. Mbololo, K. C. 27.iv.34. 
9 (M. C. Z. 40706) Ngatana, K. C. 19.vi.34. 
d" (M. C. Z. 40707) Changamwe, K. C. 4.vii.34. 

Native name. Ngogoma (Kipokomo). 

Variation. Midbody scale-rows 17-19; ventrals 163-203; anal 
divided; subcaudals 16-27; labials 6, the 3rd and 4th entering the orbit 
except in M. C. Z. 40705 which has 5, the 2nd and 3rd entering the 
orbit; lower labials in contact with the anterior chin shields 4; pre- 
ocular 0; postocular 1; temporal 1; loreal 0. 

The rostral apparently develops with age as in Prosymna, it is not 
distinguishable from that of a Rhinocalamus of similar size so should 
be avoided as a key character. 

Measurements. The & measures 296 (265 + 31) mm., the larger 9 
(M. C. Z. 40705) measures 434 (410 + 24) mm. 

Diet. Shortly after capture, the Changamwe snake disgorged a wolf 
snake {Lycojmidion c. acutirostre) only 20 mm. shorter than itself. 
The similarity in the parallel development of these two blackish, bur- 
rowing snakes was striking. The head of the wolf snake was too digest- 
ed for a labial count, but it was a cf with midbody scale-rows 17; 
ventrals 154; anal single; subcaudals 31, having 9 ventrals less, and 4 
subcaudals more than its vanquisher. 

Micrelaps bicoloratus Sternfeld 

Micrelaps bicoloratus Sternfeld, 1908, Sitzber. Ges. Naturf. Freunde Berlin, 

p. 93: Kibwezi, Kenya Colony. 
Rhinocalamus meleagris Sternfeld, 1908, Mitt. Zool. Mus. Berlin, 4, p. 244, 

fig. 4: no locality given, later (1910) stated as Lamu Island. 

9 (M. C. Z. 40708) Tsavo Station, K. C. 5.iv.34. 

Distribution. I failed to secure this rare little burrowing snake at 
Kibwezi from which Tsavo is distant some fifty miles. 

Synonymy. I believe that Sternfeld was led to describe Rhinocalamus 
meleagris by a too slavish adherence to the key in the Catalogue of 
Snakes, vol. 3 (1896, p. 31). In his bicoloratus the postocular is barely 
in contact with a temporal, in meleagris it is well separated. Some years 
ago, I (1923, p. 889) referred a snake from Gonya, near Kilimanjaro 
to meleagris. In 1925 the Museum of Comparative Zoology received 
another Tanganyika specimen identified as meleagris by Franz Werner. 
The same year Angel (1925, p. 36) recorded meleagris from Samburu, 
near Mombasa. The Tsavo snake is undoubtedly conspecific with all 



268 bulletin: museum of comparative zoology 

these yet geographically they come from round the type locality of 
bicoloratus. 

If we arrange the available scale counts for these snakes, we find 
that those from the farthest north agree closely with those from 
the furthest south. 

Type of meleagris from Lamu Island has 251 ventrals and 22 subcaudals. 



Type of bicoloratus 


Kibwezi 


" 226 


ii 


it 


16 


M. C. Z. 40708 


Tsavo 


" 235 


it 


it 


23 


Paris Museum 


Samburu 


" 202 


(i 


it 


28 


Nairobi Museum 


Gonya 


" ? 


it 


it 


28 


M. C. Z. 20948 ex. Ta 


nganyika Terr. 


" 256 


u 


it 


23 



Sternfeld's figures show the frontal of bicoloratus to be much wider 
than the supraocular while it is equal to them in meleagris. In this 
character our two snakes agree with bicoloratus but they agree with 
meleagris in having the postocular separated from the temporal. 

Variation. Midbody scale-rows 15; ventrals 235; anal divided; sub- 
caudals 23; labials 7, the 3rd and 4th entering the orbit; preocular 0; 
postocular 1 ; loreal ; temporals 1 + 1 . 

Measurements. 9 measures 273 (256 +17) mm. 

Aparallactus turneri Loveridge 

Aparallactus turneri Loveridge, 1935, Bull. Mus. Comp. Zool., 79, p. 9: Sokoki 
Forest, near Malindi, Kenya Colony. 

1 (M. C. Z. 40120-1) Peccatoni, K. C. 24.V.34. 

2 (M. C. Z. 40121-2) Mkonumbi, K. C. 28.V.34. 
2 (M. C. Z. 40123-4) Near Witu, K. C. 31.V.34. 

Variation. Midbody scale-rows 15; ventrals 120-139; anal single; 
subcaudals 31-42; labials 6, the 2nd and the 3rd entering the orbit; 
preocular 1; postocular 2, or 1 in the type only; symphisial not in 
contact with the chin shields or almost so in M. C. Z. 40124. 

Measurements. The largest d 1 (M. C. Z. 40120) measures 202 
(167 + 35) mm. 

Aparallactus capensis Smith 

Aparallactus capensis A. Smith, 1849, Illus. Zool. S. Africa, Rept., App. p. 16: 
Kaffirland eastward of Cape Colony. 

2 (M. C. Z. 40709-10) Mt. Mbololo, K. C. 17.iv.34. 

Distribution. These constitute the second record of capensis for Kenya 
Colony to the best of my belief, they involve punctatolineatus also 
for reasons explained below. 



loveridge: African reptiles 269 

Native name. Mowa (Kitaita, but generic). 

Variation. A. punctatolineatus Boulenger only differs from capensis 
in having 6 or 7 upper labials with the 2nd and 3rd entering the 
orbit while capensis has 7 upper labials, the 3rd and 4th entering the 
orbit, apparently there is no second distinguishing character. Both 
the Mbololo snakes were taken by me on the same day yet M. C. Z. 
40709 agrees with capensis, M. C. Z*. 40710 with punctatolineatus, 
the latter having only 6 labials of which the 2nd and 3rd enter the 
orbit. 

A. punctatolineatus is known from two specimens from Angola 
and Nyasaland and is therefore within the range of capensis. I 
strongly suspect that it only occurs as an aberration of capensis. 
I imagine that Tornier's (1897, p. 79) records of nigriceps, copied by 
Sternfeld (1910a, p. 36), from Tanga and Marangu, Kilimanjaro are 
similar aberrant capensis for the latter author (1910a, p. 36) records 
capensis from Tanga and Kilimanjaro on the same page. 

It might be as well to add that in these Mbololo snakes the sym- 
phisial is in contact with the anterior chin-shields and these snakes 
have only a single postocular which is well separated from the temp- 
orals. It is clear therefore that they cannot be referred to either 
werneri of the Usambara or jacksonii of Mount Kilimanjaro which 
was more to be expected. 

Midbody scale-rows 15; ventrals 140-163; anal entire; subcaudals 
37 and mutilated; preocular 1. 

Measurements. The larger 9 (M. C. Z. 40709) measures 268 + 
(230 + 38 +) mm., its tail-tip being missing. 

Diet. The stomach of the smaller contained two centipedes of a 
species of which a tubeful were collected. 

Habitat. One was taken beneath a rotting log in the forestry 
nursery at 4,000 feet, the other beneath a stone on the eastern slope 
at 4,500 feet approximately. 



Aparallactus concolor (Fischer) 

Uriechis concolor Fischer, 1884, Jahr. Hamburg Wiss. Anst., 1, p. 4, pi. i' 
fig. 1 : Arusha, Tanganyika Territory. 

1 (M. C. Z. 40711) Voi, K. C. 17.iv.34. 

3 (M. C. Z. 40712-4) Mt. Mbololo, K. C. 23-30.iv.34. 

Native name. Motva (Kitaita, but generic). 

Corrigenda. Angel (1925, p. 36) has already recorded this species 



270 bulletin: museum of comparative zoology 

from Bura near Mbololo but under the erroneous identification of 
Elapops modestus. I have examined the snake in question. 

Variation. Midbody scale-rows 15; ventrals 143-158; anal entire; 
subcaudals 58-62; labials 7, the 3rd and 4th entering the orbit; 
infralabials in contact with the anterior chin shields 4; preocular 1; 
postocular 1; loreal 0; temporals 1 + 2. 

Measurements. The larger cT (M. C. Z. 40714) measures 342 
(280+ 62) mm., the larger 9 (M. C. Z. 40711) measures 520 (420 
+ 100) mm. 

Aparallactus uluguruensis Barbour & Loveridge 

Aparallactus uluguruensis Barbour & Loveridge, 1928, Mem. Mus. Comp. 

Zool., 50, p. 132: Nyange, Uluguru Mountains, Tanganyika Territory. 
Aparallactus concolor boulengeri Scortecci, 1931, Atti. Soc. Ital. Milano, 70, 

p. 212: Villa Duca Abruzzi and Inland from Mogadish, Italian Somaliland. 

9 (M. C. Z. 40715) Ngatana, K. C. 17.vi.34. 

Distribution. This is the first record of the occurrence of this 
species in Kenya Colony, but not wholly unexpected as other rain- 
forest forms occur at Ngatana. Its occurrence in Italian Somaliland, 
however, leads one to suppose that it cannot be regarded as a rain- 
forest form ; its distribution prevents its being treated as a race of 
concolor. 

Native name. Penge (Kipokomo). 

Variation. Midbody scale-rows 15; ventrals 155; anal single; sub- 
caudals 47; labials 7, the 3rd and 4th entering the orbit; preocular 1; 
postocular 1 ; loreal 0; temporals 1 + 2. 

This specimen agrees with the ten types in the symphisial being 
broadly in contact with the anterior chin shields and the nasal in 
contact with the preocular so that the second supralabial is separated 
from the prefrontal. These serve to separate it from concolor. The 
character of how much rostral, as seen from above, stands in relation 
to its distance from the frontal, does not seem to be of specific signifi- 
cance. While there are only three lower labials in contact with the 
anterior chin shield on the left (as in the ten types of uluguruensis), 
there are barely four on the right as in concolor. 

Measurements. 9 measures 370 (300 + 70) mm. 

Breeding. On June 17, 1934, there were 2 eggs measuring 12 x 5 
mm. in her oviduct. 



loveridge: African reptiles 271 

ELAPINAE 

Elapsoidea guntherii Bocage 

Elapsoidea Guntherii Bocage, 1866, Jorn. Sci. Lisboa, 1, p. 70, pi. i, figs. 3-3b: 
Cabinda, Portuguese Congo and Bissao, Portuguese Guinea. 

1 (M. C. Z. 40717) Nairobi, K. C. 30.X.33. 

2 (M. C. Z. 40718-9) Sipi, U. 12.xii.33. 

6 (M. C. Z. 40720-4) Kaimosi, K. C. 13-28.ii.34. 

Native name. Mugoya (Lugishu). 

Variation. Midbody scale-rows 13; ventrals 156-167; anal entire; 
subcaudals 18-26; labials 7, the 3rd and 4th entering the orbit; 
preocular 1 ; postoculars 2 ; temporals 1 + 3 ; internasals shorter than 
the prefrontals; symphisial separated from the chin shields. 

Coloration. The young Nairobi snake is of the red-and-white banded 
type (guntherii), the rest are uniformly black or black with pairs of 
light transverse lines (nigra). For a discussion of the relationship 
of these two forms see Barbour & Loveridge (1928, p. 134) based on 
a series of forty-seven snakes from the Uluguru and Usambara 
Mountains. 

Measurements. The largest c? (M. C. Z. 40722) measures 623 (575 + 
48) mm., the largest 9 (M. C. Z. 40721) measures 630 (581 + 49) mm. 

Naja melanoleuca Hallowell 

Naia haie var. melanoleuca Hallowell, 1857, Proc. Acad. Nat. Sci. Philad., p. 61: 
Gaboon, West Africa. 

3 (M. C. Z. 40725-7) Sipi, U. 19-23.xii.33. 
1 (M. C. Z. 40728) Butandiga, U. 12.i.34. 

16 (M. C. Z. 40730-41) Kaimosi, K. C. 14-28.ii.34. 

Native names. Simla (Luganda); wahobi (Lugishu). The first name 
is also commonly applied to the Egyptian Cobra (N. haje) in western 
Tanganyika Territory. 

Local European name. At Kaimosi I found the Black-and-white 
Cobra was erroneously called a "Black Mamba" by the American 
and European residents. 

Variation. Midbody scale-rows 19; ventrals 206-220; anal entire; 
subcaudals 57-68; labials 7, the 3rd and 4th entering the orbit, the 
6th constantly largest and in contact with the postoculars; rostral 
invariable broader than deep; preoculars 1, or 2 in M. C. Z. 40727 
only; postoculars 3, or 2 on eight sides only; temporals 1 + 1 on 
two sides, 1 + 2 on twelve, 1 -f- 3 on twenty-five, 2 + 3 on one. 



272 bulletin: museum of comparative zoology 

Measurements. The largest cf (M. C. Z. 40730) measures 2112 
(1790 + 322) mm., the largest 9 (M. C. Z. 40731) 1700 (1440 + 
260) mm. 

Weight. The six foot, ten and three-quarter inch male weighed 
four pounds on a spring balance when freshly killed, its stomach 
being empty. 

Breeding. On February 26, 1934, a large Kaimosi 9 (M. C. Z. 
40729) held 15 eggs, measuring 60 x 30 mm. ready for deposition. 

Diet. Rodent fur in a Sipi snake while the following mammals 
were found in three of the Kaimosi series, Dasymys h. helukus, Lemnis- 
comys s. massaicus, Arvicanthis a. nubilans. 

Parasites. Ticks (Aponomma laeve) were present on several cobras 
in the Kaimosi series but no internal parasites except a nematode 
(Physaloptera sp.), and three small nematodes (Kalicephalvs sp.) 
recovered from a Sipi male. 

Enemies. A shrivelled old hag of an Mgishu killed two big cobras 
in her garden without damaging either, the larger was about five 
and a half feet in length. 



Naja nigricollis nigricollis Reinhardt 

Naja nigricollis Reinhardt, 1843, Dansk. Vidensk. Selsk. Skrift, 10, p. 269, pi. 
iii. figs. 5 and 7: Guinea, West Africa. 

9 (M. C. Z. 40742) Kibwezi, K. C. 29.iii.34. 

Variation. Midbody scale-rows 27; the highest on record except 
Sternfeld's for 27, without locality, from Hoffman 1 ; ventrals 220 
anal entire; subcaudals 59; labials 6, the 3rd entering the orbit 
lower labials in contact with the anterior chin shields 4; preocular 1 
postoculars 3; temporals 2 + 5 and 2 + 4. 

Coloration. Bright reddish pink above with an encircling collar of 
black which covers seven ventrals on the "throat," a black blotch 
below the orbit. 

This is the rare red variety whose first mention in literature would 
appear to be that of Patterson (1907, p. 164) in "The Maneaters of 
Tsavo." where he writes: "a great red snake, about seven feet long, 
gazing at me from the side of my camp-bed." The only other ex- 
amples that I have seen are two from the base of Mount Longido, 

•As Hoffman lived at Kibwezi the snake doubtless came from there. Sternfeld refers it to 
"var. pallida," probably the reddish pink hue had faded before he saw it. 



loveridge: African reptiles 273 

Tanganyika Territory which is in the same patch of red-soil, thorn- 
bush country as Tsavo and Kibwezi. 

Measurements. 9 measures 1372 (1168 + 204) mm. 



Dendraspis jamesoni kaimosae Loveridge 

(Plate 4, fig. 2) 

Dendraspis jamesoni kaimosae Loveridge, 1936, Proc. Biol. Soc. Washington, 
49, p. 64: Kaimosi, Kakamega district, Kenya Colony. 

7 (M. C. Z. 40743-8) Kaimosi, K. C. 10-22.ii.34. 

Remarks. The above series consists of the type and paratypes of 
the eastern form characterized by a uniformly black tail and fewer 
subcaudals, viz. 94-104 instead of 103-122. 

Diet. A tree rat (Oenomys b. cditvs) in one, a swamp rat (Otomys 
t. elgonis) in another. 

Habitat, etc. The seven foot snake which had swallowed a tree rat 
was lying coiled on the horizontal, spreading branch of a tree close 
to camp. The head was protected by the coils but a charge of dust 
shot from the .410 so disabled it that the mamba was unable to make 
off and a charge of No. 8 from the other barrel brought it down. 

A native brought in another about the same size. He had been 
cycling along the road towards our camp when the snake attempted 
to dash across the track. The boy jammed on his brakes, but the 
snake was already through his front wheel and, entangled in the 
spokes, wedged into the front fork. The lad sprang from the machine 
and ran down the road. On looking back he saw the snake still caught 
in the forks at which it was striking blindly, cutting a long pole the 
native returned and belaboured the snake from a safe distance but 
without damaging it lest its value be depreciated. When stunned, 
he placed a noose round its neck and brought it alive to camp where 
it presently recovered (vide plate 4, fig. 2). 



Dendraspis angusticeps (Smith) 

Naia angusticeps A. Smith, 1849, Illus. Zool. S. Africa, 3, pi. lxx: Natal and the 
country eastwards towards Delagoa Bay. 

Dendraspis sjostedti L6nnberg, 1907, in Sjostedt, "Wiss. Ergeb. Schwed. 
Zool. Exped. Kilimandjaro, Meru umgeb. Massaisteppen." No. 4, p. 17, 
pi. i, f. 2: Kibonoto, Kilimanjaro, Tanganyika Territory. 



274 bulletin: museum of comparative zoology 

<? (M. C. Z. 40749) Kibwezi, K. C. 24.iii.34. 
skin (M. C. Z. 40750) Tsavo, K. C. iii.34. 

9 (M. C. Z. 40751) Mt. Mbololo, K. C. 16.iv.34. 
<? 9 9 (M. C. Z. 40752-3) Kitau, Manda Id., K. C. 15-19.V.34. 
c? (M. C. Z. 40754) Near Witu, K. C. 31.V.34. 
young (M. C. Z. 40755) Malindi, K. C. 29.vi.34. 

Distribution. Also one from Sokoki Forest (H. J. A. T.). 

Native names. Ilumangiu (Kitaita) ; fiha (Kiamu). 

Synonymy. D. sjostedti was based on one of seven mambas taken 
at Kibonoto, the others being referred to angusticeps. It is, as its 
author thought might be the case, only an aberrant individual whose 
irregular scalation is due to a fusing of head shields in these very 
variable reptiles. 

Variation. Midbody scale-rows 19-25; ventrals 209-250; anal 
divided; subcaudals 104-114; labials 7-9, the 4th entering the orbit; 
preoculars 3; postoculars 3-4, the lowest might be termed a subocular; 
temporals 2 + 3, 2 + 4 or rarely 2 + 5; temporals along the outer 
border of a parietal 2-4, being separated from the hindmost on the 
opposite side by from 3-7 scales. 

Measurements. The largest 9 (M. C. Z. 40752) measures 9 feet, 
i.e. 2630 (2110 + 520) mm., the tip of the tail being missing. 

Diet. One Kitau mamba held a bat {Lama f. rex) and two young 
thrushes (Turdus tephronotus) so freshly swallowed that all three 
were preserved. See below. 

Parasites. Ticks (Aponomma laeve) were preserved from Kitau 
snakes. 

Field Notes. As a native youngster and I were creeping through 
the scrub woodland, our attention occupied with searching the trees 
for hyrax, a vivid green mamba was suddenly seen by my small guide. 
He started back, colliding into me; the snake, I fancy, having been 
lying across the trail we were pursuing. When I first saw it the 
beautiful reptile was within six feet of us and rapidly ascending a 
bush to a liana, thence to a tall euphorbia. I let it attain a height of 
thirty feet before shooting it. It fell at my feet and was dispatched 
with a stick. Length 5' 8 l A". (Kibwezi, March 24, 1934). 

On arrival at Tsavo, I found the eight-foot skin of a mamba hanging 
in the cellar of the empty house which I occupied, I learned that it 
was from a snake which a fortnight before had been shot about fifty 
feet from the empty building by a prospector who had stayed in 
the house. 

One day, I was standing on a large mass of smooth, but slightly 



loveridge: African reptiles 275 

sloping, rock on a boulder-strewn hill southeast of the station. Below, 
and to the left of me, was a gunbearer searching for a hyrax which 
I had just shot. Above, and behind me, a second native was descend- 
ing after going to retrieve a lizard which I had shot. Apparently, 
in descending, he disturbed a mamba, possibly six feet in length, 
certainly not an inch less than five feet. It was so quick in its rush 
that he never saw it. I felt something bump and brush against my 
shoe, as I half-turned the snake was already in mid-air having shot 
off the rock with the impetus of its descent. It landed twenty feet 
below on a mass of scrub and thorn, never paused, slid straight 
over another huge slab in full view, then dived into a tangle of vege- 
tation beyond this rock and was seen no more. The boy on the 
rocks to the left below me, exclaimed: "Did you see that big snake 
go right between your legs?" As a matter of fact it was not actually 
between, what happened was that it had side-slipped with the velocity 
with which it arrived on the rock, then cannoned against my shoe. 
I was thankful that my back was towards it for had I been facing 
the other way I should doubtless have gone to swell the ranks of 
those who thought they had been attacked by a mamba. Though 
I had a twelve-bore shotgun in my hand there was not time to use 
it and had there been I should have hesitated to do so with the de- 
scending native just twenty feet behind the snake. (Tsavo, April 4, 
1934). 

An Mtaita brought in a fine specimen measuring eight and a 
quarter feet, he said that he shot it with an arrow as it was about to 
enter the door of his hut. Today I was summoned to catch an ex- 
ceptionally large one that was engaged in swallowing a big species 
of rat (apparently a cane rat (Thrynomys sp.) by their description); 
the mamba was on a narrow path through dense, impenetrable scrub. 
They said that this particular snake was well known as it frequented 
a tree overhanging a waterhole and scared the women coming for 
water. We passed this waterhole which was about fifty yards from 
where the snake was supposed to be swallowing the rat. On our arrival 
we found that it had departed, leaving a track as large as that of a 
small python; we lost it in the scrub. (Mount Mbololo, April 16, 1934). 

The afternoon of our arrival on Manda Island my attention was 
attracted by two thrushes fluttering about an acacia. It turned out 
that they were annoyed by a five and a half foot mamba which had 
swallowed their nestlings. A full account of the incident will be 
found in the report on the birds collected, (see page 177 of this vol.) 
under Turdus tephronotus. (Kitau, May 15, 1934). 



276 bulletin: museum of comparative zoology 

Our part of the island was largely covered by acacia trees whose 
low, wide-spreading boughs necessitated constant stooping when one 
followed the few native paths through the dense scrub. These same 
spreading boughs apparently provided ideal situations on which 
mambas could sun themselves. Today we were returning to camp 
along a narrow path through acacia forest and scrub. I, leading by 
about a hundred feet, stooped and passed beneath a spreading bough, 
as I had already done a score of times during the morning's tramp. 

Next came a native carrying a dikdik in each hand. He also stooped 
and passed on. The gunbearer followed, a collecting gun in his right 
hand, a rifle slung across his back and projecting above his left shoulder. 
A tall fellow, he did not stoop sufficiently, and the muzzle of the rifle 
became entangled in the branch. Without looking round, he im- 
patiently jerked his left shoulder; failing to free himself, he turned 
to see how best he could be extricated and found himself looking into 
the face of a mamba whose "neck" was resting against the rifle barrel 
and must therefore have been within six inches of his face. With a 
wild cry, the man sprang forward, freeing the rifle by the impetus of 
his rush. I came hurrying back, to find the snake already three trees 
away and travelling fast over foliage at a height of twenty-five feet 
from the ground. I shot it, and, on recovering the reptile, found it 
to be a nine-foot female, exclusive of the tip of the tail which was 
missing — probably lost in her youth. She was in well-nourished con- 
dition with quantities of fat though the stomach was clean and 
empty. (Kitau, May 17, 1934). 

As I was running through fairly open scrub in search of a bush- 
fowl that I had shot, I disturbed a mamba, six feet in length. Appar- 
ently it was on the ground but when first seen it was four feet from 
my face ascending a bush at a great pace, from thence to an acacia 
without pausing until it attained a height of twenty feet. There I 
shot it; in its stomach were nestling birds. (Kitau, May 19, 1934). 

The point in recounting these incidents is to show that in East 
Africa, as distinct from further South, mambas do not as a normal 
thing appear to be aggressive. Though they were evidently very 
abundant on Manda Island the natives said that they only remembered 
one of their number being bitten by this species. He inadvertently 
trod upon it, was bitten, and died; the incident occurred several 
years before. 



loveridge: African reptiles 277 

VIPERIDAE 

Causus resimus (Peters) 

Heterophis resimus Peters, 1862, Monatsb. Akad. Wiss. Berlin, p. 277, pi. - 

fig. 4: Gebel Ghule, Sennar, Sudan. 
Causus Jacksonii Giinther, 1888, Ann. Mag. Nat. Hist. (6), 1, p. 331: Lake 

Tanganyika and Lamu, Kenya Colony. 

2 (M. C. Z. 40757-8) Kaimosi, K. C. 25.ii.34. 
2 (M. C. Z. 40759-60) Peccatoni, K. C. 24.V.34. 
10 (M. C. Z. 40761-70) Ngatana, K. C. ll-20.vi.34. 

Native names. Kiukisi (Kiamu) ; lundugalla (Kipokomo) . 

Variation. Midbody scale-rows 19-21 ; ventrals 137-145; anal entire; 
subcaudals 16-23; labials 6; subocular 1 or fused with a postocular, 
the fusion reaches a climax in several snakes where the eye is sur- 
rounded by only three scales and the supraocular; preoculars 1-3, 
normally 2; postoculars 1-3, normally 2; temporals 2 -f- 3, rarely 
2 + 4. * 

Measurements. The largest cf (M. C. Z. 40762) measures 572 (517 
+ 55) mm., the largest 9 (M. C. Z. 40764) measures 552 (512 + 40) 
mm. 

Breeding. At Peccatoni on May 24, 1934, a 9 held 9 eggs measuring 
12 x 5 mm., at Ngatana on June 11, 1934, another had 4 eggs measur- 
ing 19 x 9 mm., in her oviducts. 

Two snakes, taken at Kaimosi on February 25, 1934 and at Ngatana 
about June 12, 1934, are so small as to give some indication of the 
breeding season. They measured 176 and 190 mm. respectively. 

Diet. At Peccatoni, I had just remarked that it was curious there 
were no snakes seeking the numerous frogs assembled in swamped 
grasslands near the lake, when we disturbed a Green Night Adder 
moving sluggishly along beneath a young doom palm on a little island 
in the flooded area. Shortly after capture it disgorged two young 
Rana m. mascareniensis while in its stomach was a Phrynomerus 
bifasciatus so recently taken as to be worth preservation, and a 
second which had probably been swallowed twenty-four hours before, 
as the head and anterior portion of the body were digested away. 
In an Ngatana snake were the remains of a toad (Bufo steindachnerii). 

Parasites. At Ngatana one male had a cestode {Ophiotaenia ? 
punicea), and had a Linguatulid emerging from its nostril while others 
(Porocephalus subulifer and Raillietiella boulengeri) were removed from 
its stomach; a second specimen, which like the first was killed in a 
cyanide jar, had one of these curious creatures emerging from its anus. 



278 bulletin: museum of comparative zoology 

Causus defilippii (Jan) 
H(eterodon) De Filippii Jan, 1862, Arch. Zool. Anat. Fisio., 2, p. 225: Africa. 
c? & (M. C. Z. 40771-2) Sokoki Forest, K. C. vi.32. 

Collected and presented by Mr. H. J. Allen Turner. 

Variation. Midbody scale-rows 17; ventrals 112-115; anal entire; 
subcaudals 16-17 pairs; labials 6; subocular 1; preoculars 2; post- 
oculars 2; temporals 2 + 3 and 2 + 4. 

Measurements. The larger cf (M. C. Z. 40771) measures 376 (343 
+ 33) mm. 

Parasites. Ticks (Aponomma ochraceum) of a rare species were 
found upon it. 

Causus lichtensteinii (Jan) 

H (eterodon) Lichtensteinii Jan, 1859, Rev. & Mag. Zool., p. 511: Gold Coast. 

d" (M. C. Z. 40756) Kaimosi, K. C. 28.ii.34. 

Variation. Midbody scale-rows 15; ventrals 144; anal entire; sub- 
caudals 22; labials 7; subocular 1; preoculars 3; postoculars 2; temp- 
orals 2 + 3. 

Measurements, cf measures 527 (480 + 47) mm. 

Bitis arietans (Merrem) 

(Plate 5, fig. 1) 

Vipera arietans Merrem, 1820, Vers. Syst. Amphib., p. 152: Cape of Good Hope. 

9 (M. C. Z. 40773) Aturai, Karamoja, U. ll.xi.33. 
<? (M. C. Z. 40774) Sipi, U. 22.xii.33. 
2 (M. C. Z. 40775) Bukori, K. C. 18.i.34. 
9 (M. C. Z. 40776) Kirui, K. C. 21.i.34. 
9 (M. C. Z. 40777) Elgonyi, K. C. 31.i.34. 
2 (M. C. Z. 40778) Kaimosi, K. C. 10.ii.34. 
9 (M. C. Z. 40779) Kibwezi, K. C. 23.iii.34. 
c? (M. C. Z. 40780) Mkonumbi, K. C. 28.V.34. 

Native names. Akipom (Karamojong) ; chikorviri (Lugishu). 

Variation. Midbody scale-rows 27-35; ventrals 132-143; anal 
entire; subcaudals 16-33; labials 11-15. 

Measurements. The largest cf (M. C. Z. 40780) measures 1400 
(1280 + 120) mm., i.e. 55 inches long, being the biggest d 1 I have 
seen. Though rather emaciated and its stomach empty, it weighed 
6 lbs. on a spring balance. 



loveridge: African reptiles 279 

Breeding. The Aturai 9 had ova just developing, the Elgonyi 
snake fifteen or more embryos, the exact number undeterminable 
from the pounding the snake had received from the native who 
brought it in. The Kibwezi Puff Adder is so young (213 mm.) as to 
have been but recently born. 

Diet. Rodent fur in the Sipi snake, a House Rat (Rattus r. kijabius) 
in the Kaimosi reptile. 

Distribution. Said not to occur on Lamu Island, but the natives on 
Manda Island say that it is common there and that many cattle die 
from its bite. The probability is that they die from mamba bites. 
I saw no Puff Adders during my week on the island though I searched 
specially for them one afternoon. 

Bitis nasicornis (Shaw) 

(Plate 5, fig. 2) 

Coluber Nasicornis Shaw, 1802, Nat. Miscell., 3, pi. xciv: Interior of Africa 
(from the master of a Guinea vessel). 

18 (M. C. Z. 40781-94) Kaimosi, K. C. 8-10.ii.34. 

Native name. Liheri (Luragoli). 

Variation. Midbody scale-rows 33-39; ventrals 119-129; anal 
single; subcaudals 16-30; labials 15-19. 

Measurements. The largest cf (M. C. Z. 40783) measures 760 
(705 + 55) mm., the largest 9 (M. C. Z. 40781) measures 1015 
(945 + 70) mm. 

Sexual dimorphism. Subcaudals in males are 25-30, in females 
16-19. Males up to a length of 476 mm. have the belly beautifully 
marbled and mottled as in all females, adult males, however, have 
the belly uniformly dirty white in sharp distinction to those of the 
females. On close examination markings can be vaguely discerned 
beneath the scales though this has nothing to do with sloughing. 

Breeding. Thirty-eight large embryos were present in each of two 
big females brought in on the 8th of February, 1934. 

Diet. Rodent fur was present in most stomachs, the only identi- 
fiable mammals were a shrew (Crocidura n. nyansae) and a mouse 
(Lophuromys a. aquilus). A toad (Bufo r. regularis) in a young viper. 

Parasites. One three-quarter grown snake held twenty-five large 
linguatulids (Armillifer grandis) in its intestines, stomach and in 
the viscera just behind the head. The stomach of the same animal 
held many small nematodes. 



280 bulletin: museum of comparative zoology 

Enemies. Nose-horned vipers were recovered from the stomachs 
of a civet (Civettictis c. schwarzi) and mongoose (Ichneumia a. ibeana). 

This big viper is so abundant at Kaimosi that eighteen were brought 
in by natives in three days, after which I refused to purchase more 
so that no deduction can be made as to its abundance from the num- 
ber brought back. I feel confident that I could have obtained a 
hundred during the month we were at Kaimosi. 

Atheris squamigera (Hallowell) 

Echis squamigera Hallowell, 1854, Proc. Acad. Nat. Sci. Philad., p. 193: Near 
the Gaboon River, Guinea, French West Africa. 

9 (M. C. Z. 40795) Sipi, U. 18.xii.33. 
49 (M. C. Z. 40796-841) Kaimosi, K. C. 10-28.ii.34. 

Native name. Kisigosog (Lugishu). 

Variation. Midbody scale-rows 19-23, average 20; ventrals 148-161, 
average 153; anal entire; subcaudals 40-59, average 50; labials 8-12 
though only four sides have 11 and one 12, average 9. 

Measurements. The largest cf (M. C. Z 40841) measures 595 (492 
+ 103) mm., the largest 9 (M. C. Z. 40795) measures 701 (590 +111) 
mm. 

Sex. Apart from the fact that adult females attain a larger size than 
adult males while the latter have tails proportionately longer on the 
average, the sexes cannot be distinguished either by their ventral or 
their subcaudal scale counts. 

Breeding. The 9 from Sipi, killed December 18, 1933, was bloated 
with very small embryos but so damaged as to be uncountable. None 
of the Kaimosi series was gravid. 

Diet. A tree mouse (Dendromus i. insignis), a mouse {Mastomys c. 
tinctus) in another while two held pigmy mice (Leggada g. grata). 
Unidentifiable rodent fur and a tree frog {Hyperolius rossii) completed 
the list for forty -four stomachs were empty! 

Atractaspis bibronii Smith 

Atractaspis bibronii A. Smith, 1849, Illus. Zool. S. Africa, Rept. pi. lxxi: 

Eastern districts of Cape Colony, South Africa. 
Atractaspis rostrata Giinther, 1868, Ann. Mag. Nat. Hist. (4), 1, p. 429, pi. xix, 

fig. 1: Zanzabar. 

3 (M. C. Z. 40842-4) Ngatana, K. C. 13-19. vi.34. 
2 cf (M. C. Z. 40845-6) Changamwe, K. C. 4.vii.34. 



loveridge: African reptiles 281 

Native name. Ume (Kipokomo). 

Variation. Midbody scale-rows 23; ventrals 245-258; anal entire; 
subcaudals 21-25; labials 5, the 3rd and 4th entering the orbit; lower 
labials in contact with an anterior chin shield 3; preocular 1; post- 
ocular 1; loreal 0; temporals 1 + 2. 

Measurements. The larger cf (M. C. Z. 40845) measures 448 (422 
+ 26) mm., the largest 9 (M. C. Z . 40842) measures 549 (519 + 30) 
mm. 

Diet. A skink (Riopa sundevallii) was in the stomach of one of the 
burrowing vipers from Changamwe. 

Habitat. I took the Changamwe snakes beneath a pile of rotting 
palm fronds and the palm thatch of a collapsed hut close to the 
station. 

Atractaspis microlepidota Giinther 

Atractaspis microlepidota Giinther, 1866, Ann. Mag. Nat. Hist, (3), 18, p. 29, 
pi. vii, fig. 3: Type locality unknown. "Probably West Africa." errore. 

3 9 (M. C. Z. 40847-9) Voi, K. C. 10.iv.34. 

Variation. Midbody scale-rows 32; ventrals 239-252; anal entire; 
subcaudals 27-34 (one had the first 23 subcaudals single, then 8 paired 
followed by the last 3 single); labials 6-7, the 4th entering the orbit; 
lower labials in contact with an anterior chin shield 3; preocular 1; 
postocular 1; loreal 0; temporals 2 + 4 and 3 + 4. 

Measurements. The largest 9 (M. C. Z. 40847) measures 770 
(705 + 65) mm. 

Habitat. Both the adult females were taken together under the 
rotting grass roof of a collapsed native hut about a mile southeast of 
the station. In the same spot was a large female boa (Eryx c. loveridgii) 
while among the debris of another hut fifty feet away was another boa 
and a young House Snake (Boaedon lineatus). Though taken at the 
end of a long dry season and the stomachs of all these snakes were 
empty, they possessed considerable deposits of fat. 



GEKKONIDAE 

Cnemaspis africanus africanus (Werner) 

Gymnodactylus africanus Werner, 1895, Verh. zool.-bot. Ges. Wien, 45, p. 190, 
pi. v, f. 5: Usambara Mountains, Tanganyika Territory. 

2 cf 1 9 (M. C. Z. 40877-9) Mt. Mbololo, K. C. 22.iv.34; 



282 bulletin: museum of comparative zoology 

Variation. Upper labials 6-8; lower labials 6-8; preanal pores 10, 
one male has an additional, supernumerary pore anterior to the usual 
row; back and base of tail with 10-14 irregular rows of enlarged 
tubercles. 

Measurements. The larger cf has the same head and body length 
as the 9 , but his tail is damaged; 9 measures 91 (50 + 41) mm. 

Breeding. This female was gravid, the ova being about half the 
diameter of an egg when laid. 

Diet. A cockroach and remains of small beetles. 

Habitat. Found in the remnent of rain forest capping the mountain 
at 4,800 feet. I observed one of these geckos slip beneath loose bark 
on the trunk of a huge tree at a height of six feet from the ground; by 
stripping the bark we secured this specimen and two others. Exten- 
sive search in the forest failed to produce any more. 

Cnemaspis africanus elgonensis Loveridge 
(Plate 6, fig. 1) 

Cnemaspis africanus elgonensis Loveridge, 1936 (1935), Proc.Zool.Soc. London, 
p 820: Above Sipi at 6,500 feet, Mount Elgon, Uganda. 

c? (M. C. Z. 40870) Nyenye, Mt. Elgon, U. 8.xii.33. 
2^294 eggs (M. C. Z. 40871-5) Sipi, Mt. Elgon, U. 12-14.xii.33. 
c? (M. C. Z. 40876) Kaimosi, Kakamega, K. C. 24.ii.34. 

Distribution. Eggs of this race were also found on Mount Debasien 
at 5,000 feet; a gecko was seen at Buluganya, 6,024 feet, on the 
western slopes of Mount Elgon. 

These Elgon geckos represent the type series of a central African 
form as distinct from the typical form inhabiting the montane forests 
of eastern Africa. 

Native names. Kibaragwesi (Kisabei and Lugishu, but not specific). 
Called lisiamogoma by the Maragoli of Kaimosi who believe it to be 
the young of Agama atricoUis, to which lizard this name more properly 

belongs. 

Variation. LTpper labials 5-7, average 6.1 ; lower labials 5-7, average 
6; preanal pores 6-8; back and base (only) of tail with 10-14 irregular 
rows of enlarged tubercles. 

Coloration. Varying with environment, pale olive on the olive bark 
of a wild fig tree at Nyenye. 

Adult 9 type at Sipi. Above, gray- or brown-olive, lighter on crown 
and with a pale, interrupted, vertebral line. Below, soiled white 
flecked with brown, regenerated portion of tail plumbeous. 



loveridge: African reptiles 283 

Newly-hatched young are faintly yellowish from neck to anus, tail 
pink below. Half-grown young are bright mustard yellow from chin 
to anus and this may even extend on to the base of the tail, remainder 
of tail being gray. 

Measurements. Largest & (M. C. Z. 40872) measures 109 (52 + 57) 
mm.; largest 9 (M. C. Z . 40873) measures 112 (56 + 56) ram., but 
is shorter by 5 mm. in length from snout to anus than another with a 
regenerated tail. Newly-hatched young measure 40 (19 + 21) mm. 

Breeding. On November 22, I found two fresh eggs of this species 
beneath a log on cleared land on Mount Debasien at 4,500 feet. No 
geckos were seen, however, and unfortunately these eggs broke in the 
laboratory while being measured. 

On December 12, a gravid 9 with fully developed eggs was taken 
at Sipi; on the 14th many eggs, measuring from 10 x 10 to 11 x 9 mm., 
mostly holding embryos, were found; some of the young hatched out 
during the following week. 

Diet. Three crickets in Sipi geckos examined. 

Enemies. One cf (M. C. Z . 40871) was recovered from the stomach 
of a green snake (Chlorophis hoplogaster). 

Habitat. The Kaimosi gecko was found in a bucket into which it had 
fallen and from which it could not escape. Another was seen on the 
door of a garage at the mission, showing a certain amount of adapt- 
ability in this sylvicoline species. 



Cnemaspis qtjattuorseriatus (Sternfeld) 

Gonatodes quattuorseriatus Sternfeld, 1912, Wiss. Ergebn. Deutsch-Zentral- 
Afrika-Exped. 1907-08, 4, p. 202, pi. vi, f. 1 : Kissenje; Lake Kivu; Uvira, 
etc., Belgian Ruanda. 

5 cf 9 9 (M. C. Z. 40850-9) Mt. Debasien, U. 14-22.xi.33. 
9 cf 8 9 (M. C. Z. 40860-9) Sipi, Mt. Elgon, U. 12-14.xii.33. 

Distribution. These records constitute the first for the occurrence 
of this species in Uganda, it has already been recorded from Kenya 
(Nieden) and Tanganyika (Loveridge). 

Native name. Kibaragwesi (Kisabei and Lugishu, but not specific 
as applied to the larger species which also occurs on Mount Elgon). 

Variation. Upper labials 5-7, average for sixty -two sides 5.7, only 
six sides have 7; lower labials 4-6, average 5.5, only one side has 4; 
preanal pores 8 in all fourteen males ; back, except posteriorly, devoid 
of enlarged tubercles, a dorso-lateral and a lateral series of tubercles 



284 bulletin: museum of comparative zoology 

not quite so developed as in a cotype of quattuorseriatus, occupying an 
intermediate position between that species and dickersoni of the Ituri. 

Measurements. The largest c? cf in both localities measure 83 mm., 
viz. (41 + 42 and 37 + 46) mm. respectively; the largest 9 (M. C. Z. 
40854) measures 79 (38 + 41) mm., but in length from snout to anus 
is 3 mm. shorter than another with regenerating tail. 

Breeding. On November 18, a pair of eggs, measuring 7x6 mm., 
were taken together wijli a pair of geckos in a rotting log in the dry 
rain forest at 8,000 feet. On the 21st, another pair of eggs were found 
among drifted leaves at 4,000 feet; they were broken open and found 
to be fresh. 

Diet. Remains of very small spiders, beetles and ants were recovered 
from four stomachs examined. 

Habitat. On Debasien these geckos occur in the gallery forest of 
the ravines from 4,000 feet up to the rain forest at 8,000 feet. Several 
were taken in drifts of leaves lying between the buttress roots of the 
giant mvuli trees, more frequently in rotting logs and in one case be- 
neath a stone in the dry river bed. Several were taken on tree trunks 
in camp and after sunset one was caught running across a clearing in 
gallery forest. 

Hemidactylus brookii Gray 

Hemidactylus brookii Gray, 1844, Zool. in Voyage of Erebus and Terror, pi. xv, 
fig. 2: "Australia; Borneo." (errore) 

<? (M. C. Z. 40915) Butandiga, U. 8.i.34. 
c? & yng. (M. C. Z. 40916-7) Tsavo, K. C. 30.iii.34. 
c? 9 (M. C. Z. 40918-9) Voi, K. C. 7.iv.34. 
<? 9 (M. C. Z. 40920-1) Kitau, Manda Id., K. C. 16.V.34. 

Distribution. Also one from Sokoki Forest (M.J.A.T.). 

Variation. Upper labials 6-8, average of eighteen sides 7.5, only 
one side has 6; lower labials 5-8, average 7.4, only one side has 5; 
preano-femoral pores 26-46. 

Coloration. The Voi female was a very beautiful red as were the 
three young from Tsavo, this tendency to retain and accentuate the 
juvenile coloring is probably correlated with the red volcanic soil of 
the Tsavo- Voi region. 

Measurements. The largest cf (M. C. Z. 40916) measures 137 
(67 + 70) mm., the largest 9 (M. C. Z. 40918) measures 120 (65 + 
55) mm. but the forked tail is in process of regeneration. 

Habitat. I personally took all the Kenya specimens in houses or 
the ruins (Kitau) of a house. The Butandiga gecko was brought in by 



loveridge: African reptiles 285 

a native who possibly took it lower down Mount Elgon than Butan- 
diga, which is 7,010 feet. 

Hemidactylus mandanus Loveridge 

Hemidactylus mandanus Loveridge, 1936, Proc. Biol. Soc. Washington, 49, 
p. 60: Kitau, Manda Island, Kenya Colony. 

9 (M. C. Z. 39995) Kitau, Manda Id., K. C. 15.V.34. 
Remarks. This is the holotype of the species. 

Hemidactylus mabouia (Moreau de Jonnes) 

Gecko mabouia Moreau de Jonnes, 1818, Bull. Soc. Philom. Paris, p. 138: 
Antilles and adjacent mainland. 

6 cf 4 9 (M. C. Z. 40880-4) Kibwezi, K. C. 23-24.iii.34. 

9 (M. C. Z. 40885) Near station, Tsavo, K. C. 2.iv.34. 

2 d 1 2 9 (M. C. Z. 40886-7) Voi, K. C. 10.iv.34. 

2 d* 4 9 (M. C. Z. 40888-9) Mt. Mbololo, K. C. 19.iv.34. 
4 <? 2 9 (M. C. Z. 40890-3) Golbanti, K. C. 2-3.V.34. 

9 (M. C. Z. 40894) Lamu, Lamu Id., K. C. 5.V.34. 
9 & eggs (M. C. Z. 40895) Kitau, Manda Id., K. C. 15.V.34. 
9 (M. C. Z. 40896) Witu, K. C. 30.V.34. 
9 (M. C. Z. 40897) Belazoni, K. C. 6.vi.34. 

3 J* 1 9 (M. C. Z. 40898-9) Ngatana, K. C. ll.vi.34. 
9 & yng. (M. C. Z. 40900-1) Malindi, K. C. 29.vi.34. 

2 9 (M. C. Z. 40902) Mombasa, K. C. 5-6.vii.34. 

Native names. Molukandua (Kitaita); ndikafiri (Kiamu); goria 
(Kipokomo). 

Variation. Preanal pores in seventeen males range from 32-52, 
average 41. The female from Lamu Island, while undoubtedly mabouia 
in size and scalation, is one of those exceptional individuals which in 
having 5 (instead of 7-9) pairs of subdigital lamellae under the median 
digit, agrees with persimilis which occurs in the same locality. 

Breeding. On May 15, a pair of eggs were collected from the bark of 
a baobab tree on Manda Island. 

Diet. On June 24, at Golbanti on the Tana River, I was seated at 
breakfast on the verandah of the rest house, when my attention was 
attracted by a slight commotion among the rafters supporting the 
grass thatch. A H. mabouia had seized a small gecko (Lygodactylus p. 
mombasicus) and was holding the head of the latter in its jaws. The 
prey was an adult male which twirled its body round and round in a 
vain attempt to free itself. 



286 bulletin: museum of comparative zoology 



Hemidactylus persimilis Barbour & Loveridge 

Hemidactylus -persimilis Barbour & Loveridge, 1928, Mem. Mus. Comp. Zool,, 
50, p. 140, pi. iv, figs. 1 and 3: Dar es Salaam, Tanganyika Territory. 

7 young (M. C. Z. 40903-4) Lamu, Lamu Id., K. C. 7.V.34. 
3 young (M. C. Z. 41922-3) Mombasa Id., K. C. 5.vii.34. 



Hemidactylus frenatus Dumeril & Bibron 

Hemidactylus frenatus Dumeril & Bibron, 1836, Erpet. Gen., 3, p. 366: South 
Africa, etc. 

11 young & eggs (M. C. Z. 40905-6 Lamu, Lamu Id., K. C. 10-14.V.34. 



Hemidactylus werneri werneri Tornier 

Hemidactylus werneri Tornier, 1897, Arch. Naturg., 63, p. 63: Dalalani, Tan- 
ganyika Territory. 

7 cf 5 9 & eggs (M. C. Z. 40908-14) Voi, K. C. 9.iv.34. 

9 (M. C. Z. 40000) Ngatana, K. C. 18.vi.34. 

Native name. Goria (Kipokomo, but not specific). 

Variation. Upper labials 6-8; lower labials 5-6; preanal pores 10-20 
in six males, 6 in two geckos which are possibly of either sex and very 
young. 

Though the majority of these geckos come from a locality not much 
more than twenty miles from Bura, type locality of H. w. alluaudi 
Angel, not one of them has the mental separating the chin shields 
which is the sole characteristic of that race. I am inclined to think 
that alluaudi may have been founded on an aberrant individual. 

Coloration in life. Adult 9 Ngatana. Above, olive variegated with 
flecks of black and white, the white chiefly on enlarged keeled scales 
and uniformly arranged in four longitudinal series; edges of digits 
flecked with Chinese white. Below uniform white. Pupil vertically 
elliptic, black flanked by gold. 

Measurements. None of the series is fully grown, they range in size 
from 43 (23 + 20) mm. to 78 (40 + 38) mm. 

Breeding. On April 9, four pairs of detached gecko eggs, from 8 x 8.5 
to 9 x 9.5 mm. were found in association with these young geckos in 
a spot where no other species of geckos were encountered. 

Habitat. The whole of the Voi series were taken in a single morning 



loveridge: African reptiles 287 

of intensive search among the crumbling ruins and collapsed thatching 
of some mud huts on the Msinga Estate, a few miles from Voi. The 
absence of adults, which are known to live in the burrows of insects 
or termitaria, leads one to assume that the adults had only resorted 
to this spot to deposit their eggs. 

Hemidactylus tropidolepis squamulatus Tornier 

Hemidactylus squamulatus Tornier, 1897, Die Kriechthiere D-O-Afrikas p. 10: 
Kakoma, Ugundu, Tanganyika Territory. 

9 (M. C. Z. 40907) Changamwe, K. C. 4.vii.34. 

Distribution. Also 2 from Sokoki Forest (H.J.A.T.). 

Measurements. Total length 84 (44 + 40) mm. 

Habitat. This big example of a rare species was taken beneath a pile 
of rubbish in a native garden and not more than a couple of miles dis- 
tant from where the male was taken in 1929 (Loveridge, 1933, p. 284). 

Bunocnemis modestus Giinther 

Bunocnemis modestus Giinther, 1894, Proc. Zool. Soc. London, p. 95, pi. viii: 
Ngatana, Tana River, Kenya Colony. 

1 & 3 9 (M. C. Z. 39996-9) Ngatana, K. C. 12-18. vi.34. 

Native name. Goria (Kipokomo, but applied to Hemidactylus spp. 
also). 

Affinities. It is somewhat problematical as to whether this small 
gecko should be generically separated from Hemidactylus. It is so 
closely related to the members of the H. squamulatus group, which 
have imbricate scales on the dorsum and scarcely dilated digits, that 
the only character of generic significance remaining is the undivided 
subdigital lamellae. 

Unfortunately in the young, many of the subdigital lamellae are so 
deeply grooved that without very close scrutiny it is extremely diffi- 
cult to say whether they are paired or single. It will be recalled 
that Tornier has described a second species, which he referred to this 
genus, and states that some lamellae are single, some paired, on each 
digit. The fact that this second species, matschiei comes from Togo- 
land raises doubts as to whether these two geckos have not arisen 
separately from Hemidactylus, on the other hand it is true that the 
fauna of Tanaland has distinct West African affinities. B. matschiei 



288 bulletin: museum of comparative zoology 

appears to bridge the slight gap between Hemidactylus and Bunoc- 
nemis, but until direct comparison can be made the point had better 
be left unsettled. 

Variation. Upper labials 6-8; lower labials 5-7; preanal pores 14. 
Considerable variation is displayed in the number of the subdigital 
lamellae owing to the gradual diminuition in size so that it is often 
hard to decide as to whether some of those at the base of the digits 
should be considered lamellae or only transversely-dilated scales; 
generally it may be said that there are from 6 to 8 under the fingers 
and longer toes, from 3 (Giinther) or 4 to 6 under the inner and outer- 
most toes. 

Measurements. The only & (M. C. Z. 39996) measures 73 (45 + 28) 
mm., but the tip of the tail is undoubtedly reproduced, the largest of 
the females with a perfect tail, measures 56 (29 + 27) mm., so that 
we are justified in assuming that normally an uninjured tail should 
almost equal the length of the head and body. 

Diet. Two of the stomachs were empty, the other two contained the 
abdomen of an isopod (I am indebted to Mr. N. Banks for the identi- 
fication) and a number of eggs, apparently those of a small species of 
grasshopper. 

Enemies. The hind foot and tail of one female had been lost long 
ago, the wounds healed. 

Habitat. Few reptiles gave me more trouble to secure than did this 
topotypic series of a gecko known only from the type. I had planned 
to spend one week at Ngatana but two elapsed before a native brought 
in the first example of the lizard for which I had been out hunting 
daily. Describing to me where he had found it, enabled me to catch 
three more during the following week. These geckos were found in 
the piles of rubbish and rotting vegetation which had been cleared 
from the native gardens, situated among the mango trees which mark 
the site of the vanished village of Ngatana. Their habitat, therefore, 
is exactly similar to that of H. squamulatus. 

Lygodactylus fischeri scheffleri Sternfeld 

Lygodactylus fischeri scheffleri Sternfeld, 1912, Wiss. Ergeb. der Deut. Zentral- 
Afrika-Exped. 1907-1908, 4, p. 206: Kibwezi, Kenya Colony. 

d> (M. C. Z. 40994) Voi, K. C. 10.vi.34. 

Variation. Upper labials 6-7; lower labials 6-5; supranasals in con- 
tact; preanal pores 6; regenerated tail with a single row of transversely 
enlarged subcaudals. 



loveridge: African reptiles 289 

Coloration. Unfortunately the coloration in life of this handsome 
little gecko was not recorded; it lacks the jet black patch in front of 
the fore legs mentioned by Sternfeld; it has six, instead of two, black 
vertical blotches between the fore and hind limbs, they increase in 
intensity posteriorly; the faint reddish gray band from neck to root of 
tail is just discernible. 

Measurements. Total length 44 (24 + 20) mm., but tail regenerat- 
ing; the male type was 51 (24 + 27) mm. 

Habitat. Voi is about eighty miles southeast of the type locality. I 
first caught sight of this specimen as it ran up the trunk of a mango- 
like tree, which had laurel-like leaves, in fairly dense gallery forest 
beside the (dry) Voi River. The gecko disappeared beneath a sliver of 
bark at a height of twenty feet, by throwing up a stick I dislodged the 
sliver which, together with the gecko, fell to the ground. The reptile 
dashed up the trunk again and beneath another piece of loose bark 
where I was able to catch it. 

Lygodactylus picturatus picturatus (Peters) 

Hemidaclylus picturatus Peters, 1870, Monatsb. Akad. Wiss. Berlin, p. 115: 
Zanzibar. 

5 (M. C. Z. 40940-4) Voi, K. C. 10-ll.iv.34. 

Variation. These are somewhat of intermediates between the typi- 
cal race and L. p. mombasicus but had the mustard-yellow heads of 
true picturatus. They were taken on big trees along the bed of the Voi 
River where the extreme southeasterly type of mombasicus also occurs, 
the type which is characterized by two well-defined and very distinct 
dorsolateral bands lying on either side of the vertebral line from neck 
to root of tail. 

Lygodactylus picturatus mombasicus Loveridge 

Lygodactylus picturatus mombasicus Loveridge, 1935, Proc. Biol. Soc. Wash- 
ington, 48, p. 198: Kilindini, Mombasa Island, Kenya Colony. 

32 (M. C. Z. 40924-39) Kibwezi, K. C. 23.iii.34. 
5 (M. C. Z. 40945-6) Voi River, K. C. 10-ll.iv.34. 

1 and eggs (M. C. Z. 40947) Mt. Mbololo, K. C. 16-25.iv.34. 

2 and eggs (M. C. Z. 40948-9) Lamu, Lamu Id., K. C. 8.V.34. 

1 (M. C. Z. 40950) Kitau, Manda Id., K. C. 16.V.34. 
5 (M. C. Z. 40951-3) Witu, K. C. 31.V.34. 

29 (M. C. Z. 40954-75) Ngatana, K. C. vi.34. 
3 (M. C. Z. 40976-8) Golbanti, K. C. 22.vi.34. 

2 (M. C. Z. 40979-90) Kilindini, K. C. 5.vii.34. 



290 bulletin: museum of comparative zoology 

Distribution. Also 2 from Sokoki Forest (H.J.A.T.). 

Native name. Mvuvi (Kipokomo, but not generic. It is interesting 
to note that this name is applied to the caecilian in Kikami). 

Variation. As this has been covered in the description of this pattern 
race of picturatus, it need not be repeated here. 

Breeding. On April 16 and 17, 1934, twenty-eight eggs, measuring 
circa 8x9 mm., were found under rocks lying against a fallen log at the 
very edge of the rain forest on Mbololo at 4,000 feet. As they were 
alive they were mailed to the zoological gardens but failed to hatch. 
On May 8, four pairs of eggs, measuring 8 x 9.5 mm., were collected on 
Lamu Island, a newly hatched young one appeared in my tent and 
remained till camp was struck. It measured 23 (12 +11) mm. On 
May 31, many eggs were seen attached to verandah posts at Witu. 
About June 8, a pair of eggs, measuring 9.5 x 9.5 mm., and a young 
gecko measuring 29 (15 + 14) mm., were taken at Ngatana. 

Enemies. An account of how one of these small geckos was being 
attacked by a larger species will be found under Hemidactylus mabouia. 
At Kibwezi and on Mount Mbololo, three of these geckos were re- 
covered from the stomachs of two Spotted Wood Snakes (Philotham- 
nus s. scmivariegatus), at Voi from a young House Snake (Boacdon 
lineatus). 

Habitat. At Kibwezi on the boles of the numerous baobabs; at Voi 
on the wild fig trees which fringed the river bed; common on the she- 
oaks along the front at Lamu; a male was taken on an acacia growing 
right on the seashore at Kitau ; I captured the Witu series on the walls 
of a hut — an unusual situations for geckos of this genus. 

Lygobactylus picturatus gutturalis (Bocage) 

Hemidactylus gutturalis Bocage, 1873, Jorn. Sci. Lisboa, p. 211: Bissao, Portu- 
guese Guinea. 

11 (M. C. Z. 40981-7) Mt. Debasien, U. 15-30.xi.33. 
2 (M. C. Z. 40988-9) Nabagut, U. 7.xii.33. 
4 (M. C. Z. 40990-3) Nyenye, U. 8.xii.33. 

Distribution. Specimens were also collected on the march at Lobo- 
rokojo and Kananyait, between Mts. Debasien and Elgon, but de- 
composed in the heat before they could be preserved. 

Native names. Ageragera (Karamojong) ; kibaragwesi (Kisabei). 

Variation. Upper labials 6-8; lower labials 5-7; nostril between the 
first upper labial and 3 (rarely 2) nasals, frequently separated from the 
lower postnasal by a narrow rim resulting from an upward prolonga- 



loveridge: African reptiles 291 

tion of the first labial; supranasals separated by 1-2 granules in the 
ratio of 10 to 7; mental followed by 2-3 postmentals in the ratio of 
13 to 4, preanal pores 6-8, average 7.3 for nine males. 

Measurements. The largest d 1 (M. C. Z. 40987) measures 89 (41 + 
48) mm.; the largest 9 (M. C. Z. 40981) measures 71 (37 + 34) mm. 

Breeding. On December 7 and 8, 1933, eggs were found in the de- 
cayed interiors of trees at Nabagut and Nyenye respectively. 

Enemies. One gecko was recovered from the stomach of a Spotted 
Wood Snake (Philothamus s. semivariegatus) in forest at the foot of 
Mount Debasien. 



AGAMIDAE 

Agama rueppelli septentrionalis Parker 

Agama rueppelli septentrionalis Parker, 1932, Journ. Linn. Soc. London, Zool., 
38, p. 225: Mount Nyero; Madago's village; Voi and Mbunyi, Kenya Colony i 

3 c? 3 ? (M. C. Z. 41003-8) Voi, K. C. 17.iv.34. 

Distribution. In 1932, Parker (I.e. pp. 354-5) studied the agamas 
of this group and came to the conclusion that vaillanti Boulenger is a 
synonym of typical rueppelli while for the form inhabiting central 
and southern Kenya, he proposed the name of septentrionalis. 

Variation. Scales on the vertebral line 30-33; dorsal scales in an 
oblique series of the standard length (tip of snout to ear) 11-13. 

Measurements. The largest 9 (M. C. Z. 41003) measures 148 (90 4- 
58) mm. 

Breeding. This female held nine eggs measuring from 17 x 10 mm. to 
18 x 9 mm. 

Diet. An acridian; stinging ants; head of a fossorial hymenopteran; 
wings and abdomens of hymenoptera; head and elytra of a beetle. 

Habitat. Not uncommon on the stunted thorn trees a few miles out- 
side Voi township. 

Agama agama agama (Linnaeus) 

Lacerta agama Linnaeus, 1758, Syst. Nat., ed. 10, 1, p. 207: "America." 
Agama colonorum Daudin, 1830, Hist. Nat. Rept., 3, p. 356: "l'Amerique 
meridionale," etc. 

d 1 juv. (M. C. Z. 40135) w. Mt. Debasien, U. 24.xi.33. 
4 d" 8 9 (M. C. Z. 40125-34) near Budadiri, U. 8.L34. 



292 bulletin: museum of comparative zoology 

Distribution. The series of twelve were brought up from the western 
foothills of Mount Elgon to my camp at Butandiga. I am by no means 
certain whether the Uganda form may not be subspecifically distinct 
from the typical race which is supposed to have come from the west coast. 

Native name. Ekihobo (Karamojong). 

Variation. Midbody scale-rows 74-84; dorsal scales in a standard 
length 17-24; preanal pores 8-11. 

Coloration in life, cf . Above, head, neck and a wedge-shaped patch 
from nape to middle of back, vermilion with a few small yellow flecks 
on occipital region; limbs, back, base and tip of tail, a dark, bluish- 
black, the median portion of tail bright vermilion. Below, throat and 
neck rich vermilion flecked with white and gray in such a way as to 
produce the effect of silver spotting; limbs (except palms and soles), 
chest, flanks and end of tail bluish-black but not so dark as on the back. 

9 . Above, sooty black, apple-green markings on anterior part of 
head becoming paler on occiput; a longitudinal streak on either flank 
from opposite the axilla to midbody is cream colored anteriorly shad- 
ing to, or tipped with, vermilion posteriorly, after an interspace fol- 
lowed by a blotch of vermilion ; median vertebral line a lighter greenish 
white anteriorly, grayish posteriorly; flecks of gray between it and the 
lateral streaks and also on the limbs. Below, throat white with gray- 
black vermiculations ; chest and belly white, anteriorly and laterally 
flecked with gray; limbs dirty white flecked with gray; tail white. 

cf juv. (M. C. Z. 40135). A dusky blue patch on base of throat 
followed by red. 

Measurements. The largest o" (M. C. Z. 40131) measures 298 + 
(128 + 170 +) mm., tail tip lacking; the largest 9 (M. C. Z. 40126) 
measures 250 (100 + 150) mm. 

Breeding. No development is shown in the ovules of the four adult 
females. 

Parasites. Acarine mites under the scales of M. C. Z. 40129., 
nematodes in M. C. Z. 40132, the only one of eight examined found 
to be infected. 

Habitat. I shot the young male from Mount Debasien on a miombo 
tree in long grass at an altitude of about 4,000 feet. 

Agama agama elgonis Lonnberg 

(Plate 6, fig. 2) 

Agama elgonis Lonnberg, 1921, Arkiv. for Zool., 14, No. 12, p. 2: Mount Elgon, 
Kenya Colony. 



loveridge: African reptiles 293 

Agama agama turuensis Loveridge, 1932, Bull. Mus. Comp. Zool., 72, p. 376: 
Unyanganyi, Turu, Tanganyika Territory. 

7 cf 1 9 (M. C. Z. 41009-15) Sipi, U. 14-22.xii.33. 
5 cf 2 yng (M. C. Z. 41017-23) Lukungu, U. 8.L34. 
10 cf 8 9 (M. C. Z. 41024-33) Elgonyi, K. C. 20-31.i.34. 

Distribution. All these localities are on Mount Elgon; the series 
was collected with a view to ascertaining whether the range of varia- 
tion on Elgon would be found to include turuensis as I suggested 
(Loveridge, 1933, p. 299) it might. This was found to be the case. 

Native names. Karingis (Kisabei); gimbiri (Lugishu); bakladut 
(Kimasai). 

Variation. Midbody scale-rows 74-90; preanal pores 10-16, aver- 
age for twenty-two males 13.3. It is true that the sixty types of 
turuensis average fewer midbody scale-rows, ranging from 72-82, and 
the average of preanal pores for thirty -four males is less, viz. 11.3, 
but the overlap is too considerable to make turuensis worthy of 
recognition in view of the face that the gular pattern and colour is 
the same in both elgonis and turuensis. 

Measurements. The largest cf (M. C. Z. 41025) measures 318 (135 
+ 183) mm., parallelling in dimensions the largest in the series of 
turuensis. 

Agama agama lionotus Boulenger 

Agama lionotus Boulenger, 1896, Proc. Zool. Soc. London, p. 214, pi. viii: 
southeast of Lake Rudolph, Kenya Colony. 

1 cf 2 9 (M. C. Z. 41034-6) Kacheliba, U. 8.xi.33. 

9 cf 6 9 (M. C. Z. 41037-50) Kibwezi, K. C. 23-29.iii.34. 

5 cf (M. C. Z. 41051-5) Tsavo, K. C. 30.iii.34. 

10 cf 4 9 (M. C. Z. 41056-9) Voi, K. C. 9-13.iv.34. 

11 cf 1 9 (M. C. Z. 41060-6) Mt. Mbololo, K. C. 16-28.iv.34. 

Distribution. Young ones seen among the rocks in the bed of the 
Karita River, Uganda. 

Native names. Ekibobo (Karamojong) ; mandari for adults, mitarongo 
for young (Kitaita). 

Variation. Midbody scale-rows 70-84; preanal pores 11-16, average 
for thirty-six males 13.1. 

Coloration. Top of the head mustard yellow, throat red, noted of 
the Kacheliba male; this is the normal colouring of the head in this 
race. 



294 bulletin: museum of comparative zoology 

Measurements. The largest cf (M. C. Z. 41060) measures 340+ 
(135 + 205+) mm.; the largest 9 (M. C. Z. 41049) measures 255+ 
(110 + 145+) mm., the tips of the tails being lacking in both speci- 
mens. 

Breeding. The three adult females from Kibwezi were all gravid, 
holding 10, 10 and 12 eggs which measured 15 x 10 mm., 17 x 11 mm., 
and 20 x 14 mm. respectively. One of the Voi females held 9 eggs 
measuring approximately 19 x 12 mm. 

No deductions should be made as to the relative preponderance of 
males in the material listed above. Females appear to be more 
abundant and are certainly easier to collect, a special effort was made 
to obtain males only. 

Diet. Vast numbers of ants, a few beetles, one hemipteron and a 
caterpillar were present in the stomachs examined; no vegetable 
matter was noticed. 

Parasites. Nematodes (Strongylurus brevicaudata and Saurositus 
agamae) were present in some of the Kibwezi and Voi series. 

Habitat. In the vicinity of old buildings at Tsavo it was noticed 
that the young were mostly on the walls of the ruins, the adults 
chiefly oh trees and rocks. 

Agama planiceps caudospinosa Meek 

Agania caudospinosa Meek, 1910, Field Mus. Nat. Hist. Zool. Series, 7, p. 407: 

Lake Elementeita, Kenya Colony. 
Agama agama kaimosae Loveridge, 1935, Bull. Mus. Comp. Zool., 79, p. 10: 

Near Kaimosi, Kakamega, Kenya Colony. 

8 <f 11 9 (M. C. Z. 40136-50) Near Kaimosi, K. C. 2-9.iii.34. 

Native name. Lisiamogoma (Luragoli, but generic). 

Synonymy. At the time I described kaimosae, I regarded both it 
and caudospinosa as races of A. a. agama. In view of the fact that both 
A. a. lionotus and caudospinosa occur in the same localities in several 
areas of Kenya, it seems advisable to treat the forms with depressed 
bodies as races of planiceps rather than of agama. 

Nearly twenty years elapsed between the time I collected any 
caudospinosa and the series of kaimosae. My impression of the former 
was that they were yellow in life in marked contrast to the gorgeous 
coloring of kaimosae. However that may be, after two years in al- 
cohol, the coloring of kaimosae is rapidly approaching that of a cotype 
of caudospinosa. Other races of agama which I have described do not 



loveridge: African reptiles 295 

depend on fugitive coloring but on characteristic sexual markings 
which remain (vide Loveridge, 1933, Plate 2). 

The other alleged distinguishing character which I cited, that of a 
markedly less developed spinosity of the tail, on reexamination I 
now believe to be the result of abrasion in the whole series of kaimosae, 
which were taken among rocks into whose fissures they fled when dis- 
turbed. A. a. kaimosae is relegated to the synonymy. 

Diet. Large numbers of termites and ants, a few beetles and a con- 
siderable amount of vegetable matter. 

Parasites. The digestive tracts of these lizards teem with nematodes 
(Strongylurus brevicaudatus and Physaloptera sp.) 

Folklore. The Watereki have a wonderful saying that if one is bitten 
by an agama, the site of the bite will cause recurring pain during the 
rainy season or whenever the bitten person hears thunder! 



Agama atricollis Smith 

Agama atricollis Smith, 1849, Illus. Zool. S. Africa, 3, Appendix, p. 14: Natal 

South Africa. 
Agama gregorii Gunther, 1894, Proc. Zool. Soc. London, p. 86: Mkonumbi, 

Kenya Colony. 

14 tf 6 9 6 yng. (M. C. Z. 41067-79) Bukori, K. C. 18-19.L34. 
& (M. C. Z. 41080) Elgonyi, K. C. 25.L34. 
5 & 4 9 (M. C. Z. 41081-9) Kirui's, K. C- 6-7.ii.34. 
4 cf 4 9 (M. C. Z. 41090-5) Kaimosi, K. C. 10-15.ii.34. 
3 c? 2 9 (M. C. Z. 41096-100) Lamu Id., K. C. 7.V.34. 

Distribution. I had forgotten that Mkonumbi was the type locality 
of the synonym gregorii or I should have secured topotypes during my 
short stay there, the Lamu specimens are, however, almost topotypic. 
Gunther's species was based on a single adult male which he com- 
pared with cyanogaster of Ethiopia, it does not seem possible to separate 
East from South African examples even subspecifically. 

Native names. Kockarfionda (Kitosh); cherengisia (Kimasai); 
lisiamogoma (Luragoli); isiamakom (Lutereki); kandi (Kiamu). 

Variation. Males possess two, occasionally even three, rows of 
preanal pores, posterior row has from 8-13 pores, average 11.0, the 
second 8-12, average 10.1. 

Coloration. When brought into camp, one of the Lamu females was 
brown; after being anaesthetised with cyanide potassium the head 



296 bulletin: museum of comparative zoology 

became a most brilliant blue, also the sides except for a line of red 
spots, vertebral line verdigris green. 

Measurements. The largest male (M. C. Z. 41096) measures 303+ 
(143 + 160+) mm.; the largest female (M. C. Z. 41073) measures 
255 (125 + 130) mm., a Lamu female with the same snout to anal 
measurement has a mutilated tail. 

Breeding. Between February 10 and 15, a Kaimosi female held 6 
eggs measuring 18x11 mm. ; on May 7 a Lamu female held 9 eggs 
measuring 11 x 11 mm.; another, but slightly smaller, taken at the 
same time held a single egg measuring 21 x 17 mm., it seems possible 
that the native captor had taken her when engaged in ovipositing. 

Parasites. Nematodes (Physaloptera amaniensis) were numerous in 
Kaimosi agamas, presumably it is the same species and Strongylurus 
sp. with which most of the others in the above series are infested. 

Enemies. One was recovered from the stomach of a Hissing Sand 
Snake (Psammophis sibilans) at Bukori. 

Folklore. The Watereki say that this lizard attracts lightning to a 
tree or house, the fact is proved because when lightning strikes a tree 
or house it is usual to find one or more of these lizards lying dead ! 

An English lad at Kitale told me that he had been told by another 
boy at Nairobi School, that these lizards are very poisonous and 
donkev's blood was the onlv cure for the bite! 



ZONURIDAE 

ZONURUS TROPIDOSTERNUM Cope 

Zonurus tropidosternum Cope, 1869, Proc. Amer. Philos. Soc, 11, p. 169: 

"Madagascar." (errore) 
Zonurus frenatus Pfeffer, 1889, Jahrb. Wiss. Anst. Hamburg, 6, p. 6: Mhonda, 

Tanganyika Territory. 

1 (M. C. Z. 39955) Sokoki Forest, K. C. vi.1932. H.J.A. Turner. 

Distribution. This specimen is the first recorded member of the 
genus Zonurus to be taken in Kenya Colony, and constitutes a note- 
worthy northerly extension of the range. 

Variation. It differs from the type (M. C. Z. 5742) of tropidosternum 
in that its nasals form a suture which separates the rostral from the 
frontonasal. Nieden (1913, pp. 71-74) has dealt very fully with the 
variability of this character, showing how both types occur together 
on Tendaguru Mountain, near Lindi, and crop up in erratic fashion 
throughout the Territory. 



loveridge: African reptiles 297 

For this reason I do not refer the Sokoki specimen to Z. parkeri 
Cott, a species whose nasals form a suture in the type but are separated 
in the paratype, for I think it may eventually prove to be a synonym 
of tropidostemum. Cott was mistaken in assuming that the caudal 
scales of tropidostemum are not serrated, they are as serrated in the 
type as in the type of parkeri; in a juvenile, however, the serrations are 
not noticeable, though well-developed in adults from the same lo- 
cality (Morogoro, near Mhonda) as well as in the Sokoki lizard; this 
would appear to be an age character. 

In the type of tropidostemum and the Sokoki specimen, the 2nd 
finger extends a claw and scale length beyond the 5th. In three Moro- 
goro lizards the 2nd finger only extends a claw length beyond the 5th 
while in the type of parkeri "it does not, or scarcely extends beyond 
the 5th." 

The 3rd and 4th fingers are nearly equal, or the 3rd minutely shorter 
than the 4th in the type of tropidostemum, as well as the rest of our series, 
which would not seem to differ greatly from the condition obtaining in 
parkeri where the "third finger is shorter than the fourth." 

The 2nd toe is equal to the 5th in the type of tropidostemum so that 
it was an error to assume that the 2nd extended beyond the 5th in the 
type of tropidostemum; it barely extends to the 5th in the type of 
parkeri, it is shorter than the 5th in the Sokoki lizard, shorter than, or 
equal to, in the Morogoro series (incidentally it is equal to in ukin- 
gensis). 

The 2nd toe is much shorter than the 4th both in our series of 
tropidostemum (including the type) as well as in the types of both 
parkeri and uhingensis, so that alleged difference also disappears. 

Measurements. Total length 168 + (86 + 82 +) mm.; the tip of 
the tail is missing. 

Chamaesaura tenuior Giinther 

Chamaessaura tenuior Giinther, 1895, Ann. Mag. Nat. Hist. (6), 15, p. 524, 

pi. xxi, fig. B: Kampala, Uganda. 
Chamaesaura annectans Boulenger, 1899, Proc. Zool. Soc. London, p. 97: 

Ravine Station, Mau Mountains, Kenya Colony. 7,500 feet. 

lc?5 9(M.C. Z. 41101-5) Kaimosi, K. C. 12.ii.34. 

Distribution. I was shown a dried skin, presumably referable to this 
species, at Kacheliba, northeast Uganda. This would constitute a 
northward extension of the range. 

Native names. Mugoye (Luragoli); shikoi/e (Lutereki). 



298 bulletin: museum of comparative zoology 

Variation. Midbody scale-rows 24; longitudinal rows between occi- 
put and anus 38-40; hind limbs monodactyle except M. C. Z. 41104 
where they are didactyle; femoral pores 1. 

I (1929, p. 59) have discussed the variations and synonymy of a 
much larger series collected by Heller in this same locality. 

Measurements. The largest 9 (M. C. Z. 41101) surpasses even 
Heller's largest, it measures 637 (135 + 502) mm., two others have 
also a head and body length of 135 mm., a third is 137 mm. but the 
tails are not so fine; the single & (M. C. Z. 41105) measures 97 mm. 
from snout to anus. 

Breeding. All five females are gravid, two examined hold 9 and 10 
embryos respectively, one of these latter was measured and found to 
be 117 (35 + 82) mm. 

VARANIDAE 

Varanus ocellatus Riippell 

Varanus ocellatus Riippell, 1827, Atlas Reise nord. Afrika, p. 21, pi. vi: Kordo- 
fan, Anglo-Egyptian Sudan. 

Egg & 2 (M. C. Z. 41106-8) Mt. Mbololo, K. C. 25-28.iv.34. 
3 (M. C. Z. 41109) Lamu, Lamu Id., K. C. 7.V.34. 
1 (M. C. Z. 41110) Kitau, Manda Id., K. C. 15.V.34. 
1 (M. C. Z. 41111) Malindi, K. C. 30.vi.34. 
1 (M. C. Z. 41112) Sokoki Forest, K. C. vi.1932. H.J.A.T. 

Native names. Mongagi (Kitaita); uru (Kiamu, Kipokomo and 
Kiswahili). 

Measurements. The largest 9 (M. C. Z. 41107) measured 1150 
(510 + 640) mm., the second 9 from the same locality had a similar 
snout to anus measurement but the tip of its tail was lacking. The 
smallest, one of a series of four young collected by Mr. H. J. Allen 
Turner, only measures 300 (140 + 160) mm. 

Breeding. On April 25, at Mbololo, this large female held 34 eggs, 
each measuring about 55 x 31 mm., ready for deposition. 

Diet. Her stomach was empty except for parasites. Stomach con- 
tents of others in the series were as follows: (1) five huge, wingless, 
stridulating, spiky grasshoppers, a hard-shelled beetle, two millipedes, 
two operculae of large snails. Mbololo; (2) nine large cockchafers, 
four hawk-moth larvae, a green caterpillar, numerous small insects. 
Lamu; (3) many ants, a few termites, a green cetonid and other beetle 
elytra. Lamu; (4) many cockchafers and smaller beetles, three large 



loveridge: African reptiles 299 

caterpillars, two snails; the intestines also were well-filled with the 
elytra of beetles. Lamu; (5) many of the hard-shelled, black beetles 
which are common above high-tide lines, a millipede. Kitau. 

Parasites. Ticks (Amblyomma marmoreum and A. exornatum) were 
collected on the Mbololo monitors, the second species only on Lamu 
specimens. Doubtless the ticks occurring on the other monitors in 
the series are referable to one or other of these species. 

Nematodes (Physaloptera paradoxa) were collected from the 
stomachs of the Mbololo and Lamu monitors, a huge mass of them in 
the gravid female from the first locality. At Kitau a native brought in 
a fairly large specimen in a terribly emaciated condition, there was 
little left of the tail but skin and bone. On opening it, I found, as an- 
ticipated, a heavy infestation of worms (Polydelphis sp.) though the 
stomach was well-supplied with a good assortment of food as recorded 
above. 

Enemies. Marketed as food by some of the Lamu natives, presum- 
ably non-Moslem elements of the population. 

Habitat. I was constantly disturbing these huge lizards in the dry 
underbrush of certain restricted areas on Lamu Island where they must 
be exceedingly common. I shot the Malindi specimen as it raised its 
head from the edge of a black, coral-rag cliff as I passed along the reef 
below; under these conditions the big head reminded me strongly of 
photographs of Galapagos marine iguanas in similar situations. Its 
tail was truncated just posterior to the anus, the stump long-since 
healed. 

Varanus niloticus (Linnaeus) 
Lacerta nilotica Linnaeus, 1766, Syst. Nat., ed. 12, p. 369: Egypt. 
1 (M. C. Z. 41113) Ngatana, Tana River, K. C. ll.vi.34. 

Distribution. Also 7 from Sokoki Forest (H.J.A.T.) examined. I 
especially refrained from collecting more than one of these common 
and useful lizards. Others were seen at Karita River camp (9.xi.33), 
Amaler River camp (xi.33), Greeki River camp (4.xii.33), Bukori 
(lS.i.34), Kaimosi (19.ii.34), Tsavo (3.iv.34), Mkonumbi (29.V.34), 
Belazoni (5.vi.34) and in many spots during my canoe journey up the 
Tana River. 

Native names. Anakana (Karamojong) ; imbidu (Luragoli and 
Lutereki); gedo (Kipokomo). 

Enemies. Eaten by the Wapokomo of the Tana River. 



300 bulletin: museum of comparative zoology 

AMPHISBAENIDAE 

Geocalamus acutus Sternfeld 

Geocalamus acutus Sternfeld, 1912, Wiss. Ergeb. Deut, Zentral-Afrika-Exped. 
1907-1908, 4, p. 209: Voi, Kenya Colony. 

16 (M. C. Z. 41114-23) Voi, K. C. 7-13. iv.34. 

Distribution. Sternfeld based his description on two specimens of 
which the second was attributed to "Deutsch Ostafrika" collected 
by Huebner. We know, however, that Huebner lived at Kibwezi, west 
of Voi, for many years and most of the species obtained by him I 
found at Kibwezi. It seems likely that the second specimen came from 
that locality. I take this opportunity of restricting the type locality 
to Voi in view of the possibility of error respecting the place of origin 
of the second example cited in the original description. At Voi I found 
it on the flats at Msinga Estate and near the northwest foot of Mount 
Mbololo. 

Native names. Kilimagonde (Kisagalla and Kitaita); moore (Kitaita, 
but this name was also applied to the local caecilian). 

Variation. Midbody scale-rows 38-42 ; transverse rows on body and 
tail 231-245, of which about 21-23 are on the tail; upper labials 3; 
lower labials 2, except No. 41119 which has 3 like modestus; no post- 
frontal; temporals 2-4 being very subject to subdivision. 

Coloration. In life a delicate flesh pink. In alcohol each dorsal scale 
violet brown edged with lighter, immaculate white below or scales 
beneath the tail irregularly mottled with violet brown. 

Measurements. The largest known specimen, a 9 (M. C. Z. 41115), 
measures 281 (248 + 33) mm., the smallest (M. C. Z. 41116) measures 
105 (93 + 12) mm. 

Breeding. No signs of gestation observed. 

Diet . A large individual held what was apparently a young worm or 
caecilian, another some skin of what may have been a caterpillar, 
in all there was much soil and grit possibly indicating that they 
swallow it like earthworms to obtain such nutriment as it may contain. 

LACERTIDAE 
Lacerta jacksoni Boulenger 

Lacerta jacksoni Boulenger, 1899, Proc. Zool. Soc. London, p. 96, pi. x: Ravine 
Station, Mau Mountains, Kenya Colony. 7,500 feet. 



LOVERIDGE: AFRICAN REPTILES 301 

Lacerta jacksoni kibonotensis Lonnberg, 1907, in Sjostedt, Kilimandjaro, Meru 
Exped., 4, Rept. & Batr., p. 5: Kibonoto, Mount Kilimanjaro, Tanganyika 
Territory. 

55 (M. C. Z. 41124-50) Sipi, U. 12-24.xii.33. 

1 (M. C. Z. 41151) Buluganya, U. 12.i.34. 

2 (M. C. Z. 41152-3) Elgonyi, K. C. 20.i.34. 

17 (M. C. Z. 41154-63) Kaimosi, K. C. 10-28.ii.34. 
6 (M. C. Z. 41164-9) Mt. Mbololo, K. C. 18.iv.34. 

Distribution. Boulenger's locality "Kegamaia" is a corruption of 
Kakamega, the specimens actually came from Kaimosi, Kakamega as 
did those listed above. As the species has been recorded from Mombo 
it may descend to below a 1,500 foot level though more abundant in 
montane forests at 6,500 feet. Lonnberg (1922, p. 3) had already re- 
corded it at that altitude on the eastern slopes of Mount Elgon, the 
first three localities cited above are at approximately the same alti- 
tude on the western and southern aspects. 

Native Jiames. Mabusiiba (Lugishu); kelondangombe (Luragoli); 
shinakombero (Lutereki); malasagasa ya murtu (Kitaita). 

Variation. Based on fifty specimens only. Midbody dorsolateral 
scale-rows 37-49, except for one lizard (M. C. Z. 41134) with 54, 
average 43.1; the outer rows of the ventrals while normally enlarged 
sufficiently to be counted, making 8 longitudinal rows, are quite fre- 
quently so reduced that they should not be reckoned as such ; normally 
4 labials anterior to the subocular but 5 occurring on one side of the 
head, or both, in every locality; femoral pores on right leg 15-21, aver- 
age 18. 

Measurements. The largest c? (M. C. Z. 41129) measures 204 (81 + 
123) mm. 

Breeding. Two females from Sipi hold ova of 4 and 7 mm. diameter 
respectively, two others from Kaimosi, approximately two months 
later, each held four eggs, the lots measuring 14 x 7 and 15 x 7 mm. 
respectively. 

Diet. Ten stomachs examined, held the following: (1) moth, (2) 
moth, (3) moth and small beetle, (4) beetles, (5) beetles, (6) big beetle 
and weevil, (7) many beetles including weevil and spiders, (8) numer- 
ous hard-shelled beetles, two spiders and a cockroach, (9) two crickets 
with long ovipositors, (10) spiders. 

Parasites. Larval tapeworms of the Dithridium group (which com- 
plete their development in carnivores) infested the peritoneal surfaces 
of both abdomen and liver in several of the Sipi lizards. 



302 bulletin: museum of comparative zoology 

Enemies. The tail of one of these lizards was recovered from the 
stomach of a green snake (Chloropkis hoplogaster) at Sipi, the lizard 
itself had evidently escaped. 

Habitat. In a deforested area at Buluganya, I disturbed a Jackson's 
lizard under debris of plantains in a banana plantation by the river. 
While writing in my tent in a forest clearing at Elgonyi, I saw and 
caught one of these lizards as it was running over my boxes, the 
second specimen from that locality, where they appeared to be scarce, 
I shot as it was basking beside a hole in the tree in which it lived. 
I shot the Mbololo series on big trees on the eastern, northern and 
southern edges of the forest cap at 4,800 feet. 

Folklore. The Luragoli name for this arboreal lizard refers to their 
belief that it 'follows the cows', it would seem as if they had con- 
fused it with some other species of terrestrial lizard but no other 
was taken in the vicinity. It is probable that the belief was imported 
into the district when the people migrated there. 

Algiroides alleni Barbour 
(Plate 7, fig. 1) 

Algiroides alleni Barbour, 1914, Proc. New England Zool. Club, 4, p. 97: 
northeast slope Mount Kenya, Kenya Colony. 

28 (M. C. Z. 41170-89) Kaburomi, U. 27-30.xii.33. 
1 (M. C. Z. 41190) near Madangi, U. 2.L34. 

Distribution. Both these localities are on the western slopes of 
Mount Elgon in the alpine zone at 10,500 and 11,500 feet respectively. 
They constitute the first records of the occurrence of this species on 
Mount Elgon though Mr. H. W. Parker tells me that the British 
Museum has a series from Majuwa (? Mujur of map) on the moun- 
tain. 

Corrigenda. Two important corrections are necessary to the key 
of the genus Algiroides given by Boulenger (1920, p. 339). He in- 
cludes alleni in the section of the key with "temporal scales keeled" 
whereas the types, as well as all in the above series, have smooth 
temporals as in the European forms. The word 'dorsals' appears to 
have been inadvertently dropped from the last line of the key (p. 
340) which was intended to read "the laterals and anterior dorsals 
entirely smooth." 

Variation. Nostril between two nasals, first labial, and frequently 
the rostral also; a supernumary loreal split off from the postnasal 



loveridge: African reptiles 303 

in one specimen; supraoculars in contact with supraciliaries in all; 
parietals in contact with upper postocular in 24 lizards, separated 
on right side in 1 (M. C. Z. 41175), on both sides in 4 (M. C. Z. 
41171, 41173^1); temporals smooth in all; enlarged plates in collar 
4-6, usually flanked on either side by a smaller one; transverse rows 
of ventrals from collar backwards 22-29, average 25.3; longitudinal 
rows of dorsolaterals at midbody 18-22, average 20.7; lamellae under 
fourth toe 16-21, average 18.3; femoral pores 10-13, average 11.3. 

Coloration in life, cf . Above, crown of head and broad dorsal 
band ochraceous brown, each head scale with one spot of sepia brown, 
inner edge of supraoculars deep black; a narrow, black, vertebral 
line from occipital scale to tip of tail; a lateral band (three scales 
wide) deep brown narrowly bordered by a black and then a pale 
yellow line both above and below, these tend to converge as they 
approach the tail; sides of body and limbs flecked with deep black; 
upper lip white, each of the posterior labials with a black spot. Be- 
low, throat to gular fold iridescent, slightly greenish, white; rest of 
under surface of body, hind limbs, and tail, deep orange which be- 
comes paler towards the sides. 

9 . Above, as in male but much darker. Below, throat to gular 
fold dull white, gular fold to in front of anus slightly greenish white, 
around anus faintly tinged with orange; below tail salmon pink. 

It is interesting to observe the close similarity in pattern and 
coloring of many of the Mabuya v. varia which occur in similar situa- 
tions at Kaburomi and exhibit similar habits. 

Measurements. The largest cf (M. C. Z. 41177) with perfect tail, 
measures 147 (52 + 95) mm., a 9 (M. C. Z. 41171) measures 58 
mm. from snout to anus but has a regenerated tail. More than half 
the series have regenerated tails, indeed it is most unusual for an 
adult to have a perfect tail. 

Breeding. The testes of the males are large but none of the females 
has indications of developing ova. 

Diet. In addition to more minute insects, the following stomach 
contents were noted: (1) caterpillar, (2) caterpillar, (3) caterpillar 
and ant, (4) caterpillar and winged ant, (5) apparently a full-fed 
lycaenid caterpillar, (6) caterpillar and spider, (7) spider and orthop- 
teran, (8) spider and praying mantid, (9) cockroach and beetle, 
(10) three small beetles. 

Habitat. These lizards occur in the alpine meadows above the 
timber line. They sleep in the dense tussocks of grass growing about 
clumps of the cactus-like Acanthus from which they emerge to bask 



304 bulletin: museum of comparative zoology 

in the sunshine during the middle of the day (circa 10 a.m. to 3 p.m.). 
If disturbed while basking they are quick to seek refuge under the 
base of the plant but are not very difficult to dislodge with a stick. 
When captured they make some show of biting but their tiny teeth 
have no effect on the human skin. 

None was seen in the vicinity of the rest camp at Madangi, though 
they probably occur there as M. v. varia is common enough in the 
neighborhood, the specimen listed was actually taken six or seven 
miles north of Madangi. 

Latastia longicaudata revoili (Vaillant) 

Eremias revoili Vaillant, 1882, Miss. Revoil Pays Comal., Rept., p. 20, pi. iii, 
fig. 2: Somaliland. 

16 (M. C. Z. 41191-200) Kibwezi, K. C. 23-24.iii.34. 
1 (M. C. Z. 41201) Tsavo, K. C. 4.iv.34. 
3 (M. C. Z. 41202-4) Voi, K. C. 10.iv.34. 
3 (M. C. Z. 41205-7) Mbololo Mt., K. C. 20.iv.34. 
1 (M. C. Z. 41208) Malindi, K. C. 29.vi.34. 

Distribution. Also 1 from Sokoki Forest (H. J. A. T.). 

Native name. Ngozo (Kitaita, but not specific). 

Variation. Midbody dorso-lateral scale-rows 58-67, average 63.1; 
gular scales between symphsial and collar 32-43, average 36.2; trans- 
verse rows of ventrals 26-30, average 28.2; lamellae beneath fourth 
toe 22-29, average 24.4; femoral pores 8-11, average 9.5. 

Coloration. A pair taken in coitu at Kibwezi clearly displayed the 
dimorphic coloring. The cf showing much darker transverse barring 
than the female, in the latter such bars being only faintly indicated. 
In the 9 the longitudinal lines are either much more pink than in 
the d 1 or show to better advantage. 

Measurements. Both the largest & (M. C. Z. 41201) and 9 
(M. C. Z. 41198) possess regenerated tails, from snout to anus they 
measure 100 mm. and 82 mm. respectively. 

Breeding. Three females taken at Kibwezi each held 4 eggs, these 
ranged from 6x6 mm. to 8 x 5 mm.; a Voi female held 3 eggs measur- 
ing 10 x 7 mm.; two Mbololo females each held 3 eggs measuring 
11 x 7 mm. 

Diet. Of the ten stomachs examined, seven held grasshoppers, one 
being distended with numerous newly-hatched grasshoppers besides 
an adult; of those remaining, two were distended with termites and 
one held a large smooth-skinned caterpillar. 



loveridge: African reptiles 305 

Parasites. Small larval tapeworms (Dithyridium) were present in 
the body cavity of a Kibwezi lizard. 

Enemies. The largest male was recovered from the stomach of a 
Two-striped Sand Snake (Psammophis biseriatus) at Tsavo, possibly 
it fell an easy prey because it was so gorged on termites, also it had 
no tail to discard as it was regenerating from the basal stump. At 
Mbololo another was found in the stomach of a Spotted Sand Snake 
(Psammophis punctulatus). 

Habitat. I ran one to earth in a rat hole in a cotton plantation 
from which I had previously seen it emerge, and at the entrance of 
which it was basking. Though both nest and food of the rodent were 
in the burrow the owner was absent. 



Eremias Neumann i Tornier 

Eremias neumanni Tornier, 1905, Zool. Jahrb. Syst., 22, p. 376: Brussa Valley, 
north of Lake Stephanie, Ethiopia. 

66 (M. C. Z. 40201-50) Ngatana, Tana River, K. C. ll-16.vi.34. 

Distribution. These constitute the first record of the occurrence of 
this species in Kenya Colony, and are the first examples taken since 
the type was described thirty years ago. Ngatana is rather more than 
500 miles southeast of the type locality. 

Native name. Mvuvi (Kipokomo, but not specific. This is the word 
for a caecilian in Kiuluguru). 

Affinities. Boulenger (1921, p. 20) suggested that this species was 
identical with Eremias siebenrocki Tornier of Porto Novo, Dahomey, 
which he transferred to the genus Latastia on the strength of Nieden's 
remarks (1913, p. 77). Elsewhere (1933, p. 306) I have shown the 
falsity of Nieden's conclusions resulting from his identifying Latastia 
johnstoni with E. siebenrocki. 

E. neumanni is certain to be distinct from siebenrocki and is un- 
doubtedly an Eremias though possessing many characters common to 
Latastia and in Boulenger's key (1921, p. 227) heading the list by 
falling into Section I, A. 

Variation. The following addition to our knowledge of variation 
in this species, is based on the individual examination of the fifty 
tagged examples from Ngatana. 

Midbody dorsolateral scale-rows 40-48, average 43.8 (the type had 
46); transverse rows of ventrals 22-30, extremes rare, average 25.6 
(the type had 26); longitudinal rows of ventrals 6 (the type had 8, 



306 bulletin: museum of comparative zoology 

Tornier presumably counted as ventrals the outer row of very small 
smooth scales which Boulenger treats as laterals); nostril between 
three nasals (except on the right side of M. C. Z. 40205 where fusion 
of the upper and lower postnasals has taken place) separated from 
the first upper labial by a narrow rim; 5 labials anterior to the sub- 
ocular as in the type, except for five specimens where there are 4 and 
three where there are 6; femoral pores 8-12, average 10 (the type had 
10). For the last two items, labials and pores, only the right jaw and 
right leg of each specimen were examined. 

Measurements. The largest specimen, a 9 (M. C. Z. 40248) mea- 
sures 146 (53 -f- 93) mm., but is 12 mm. larger than the next of 
which there are half-a-dozen with a head and body length of 47 mm. 
The smallest example (M. C. Z. 40234) measures 66 (25 + 41) mm. 

Breeding. It seems possible that the main breeding season was over 
though half-a-dozen females were gravid. One held 3 eggs measuring 
8x6 mm., another 3 measuring 9x5 mm., a third 2 eggs measuring 
11 x 6 mm. 

Diet. Spiders, as many as half-a-dozen in one stomach, would 
appear to constitute the principle article of diet with orthoptera, 
chiefly young grasshoppers, a close second. The egg capsule of a 
cockroach, but no cockroach, was present in one stomach. It should 
be borne in mind, however, that small insects which are masticated 
beyond recognition, may figure in the menu quite considerably. 

Parasites. Two larval Spiuroidea present in one stomach. 

Enemies. One Neumann's lizard was recovered from the stomach 
of a snake (Boaedon lineatus). 



Eremias spekii spekii Giinther 

Eremias spekii Giinther, 1872, Ann. Mag. Nat. Hist. (4), 9, p. 381 : Unyamwezi, 

Tanganyika Territory. 
Eremias rugiceps Peters, 1878, Monatsb. Akad. Wiss. Berlin, p. 202, pi. ii, 

fig. 1 : Taita, Kenya Colony. 

38 (M. C. Z. 41209-41) Kibwezi, K. C. 23-30.iii.34. 
9 (M. C. Z. 41242-9) Mt. Mbololo, K. C. 23-28.iv.34. 
1 (M. C. Z. 41250) Sokoki Forest, K. C. vi.32. H. J. A. Turner. 

Distribution. The series from Mount Mbololo, Taita, are topotypes 
of E. rugiceps Peters, which has the characteristics of the typical race 
and has been referred to its synonymy by Boulenger (1921, p. 235). 
The Sokoki Forest lies but a few miles south of Malindi and thus marks 



loveridge: African reptiles 307 

the northern limits of the typical form upon the coast, it was one of a 
series collected there in 1932, now in Nairobi Museum. 

Native name. Ngozo (Kitaita, but not specific). 

Variation. Femoral pores 12-18, average for thirty-one specimens 
15. All agree with the type in having the subocular bordering the lip, 
but the Sokoki lizard approaches the northern race in having six, 
instead of five, light longitudinal lines on the posterior half of the body. 

Measurements. The largest cf (M. C. Z. 41237) measures 184 (60 + 
124) mm., the largest 9 (M. C. Z. 41210) has the same length (60 
mm.) from snout to anus but has a regenerated tail. These surpass 
any in Boulenger's fine series. 

Breeding. Four gravid females in the Kibwezi series were found to 
be carrying ova, 4 in three instances, 5 in one. These measured 5x5 
mm., 9x6 mm., 11x6 mm., and 11x6 mm., these last two lots were 
clearly ready for laying. At the same time it was noticed that consid- 
erable numbers of young spekii were running about the paths, such 
juveniles measured about 85 (30 + 55) mm. 

Diet. Termites were present in each of ten stomachs examined, in 
addition one held an antlion, another a relatively large spider. 

Parasites. One of the Kibwezi series was found to be affected with 
the same larval tapeworm (Dithyridium) already recorded for Latastia 
I. revoili. 

Enemies. A Speke's Lizard was recovered from the stomach of a 
Cape Wolf Snake (Lycophidion c. capense). 

Eremias spekii sextaeniata Stejneger 

Eremias sextaeniata Stejneger, 1883, Proc. U. S. Nat. Mus., 16, p. 718: Tana 
River, Kenya Colony. 

13 (M. C. Z. 41251-9) Karita River Camp, U. 9.xi.33. 
7 (M. C. Z. 41260-6) Mkonumbi, nr. Witu, K. C. 28.V.34. 
3 (M. C. Z. 41267-8) Golbanti, Tana R., K. C. 23.vi.34. 
3 (M. C. Z. 41269-70) Karawa, K. C. 26.vi.34. 
2 (M. C. Z. 41271-2) Malindi, K. C. 28.vi.34. 

Distribution. The Golbanti specimens may be considered topotypes. 
The series from Karita River appear to constitute the first records for 
Uganda for as I have shown elsewhere (1929, p. 64), Boulenger's in- 
clusion of Uganda in the distribution was based on a specimen from 
Ndi, Uganda Railway, Kenya. This fresh material bears out the 
views already expressed (1929, p. 65) as to the meeting points of the 
two races, Takaungu being just south of the Malindi-Sokoki area. 



308 bulletin: museum of comparative zoology 

Native names. Agerigeri (Karamojong, probably for lacertids in 
general). 

Variation. With the exception of a Malindi Lizard (M. C. Z. 41271) 
where the arrangement on the right side of the head is that of the 
typical form, all the series agree with the types of sextaeniata in having 
the subocular excluded from the lip by labials. 

Labials anterior to middle of orbit 4-5, in equal proportions; dorso- 
lateral scales at midbody 58-74, average 63.8; longitudinal rows of 
ventrals 6; transverse rows of ventrals 23-29 (but only nine specimens 
examined); plates in collar 6-9, average 7; lamellae under fourth toe 
20-25, average 21.7; femoral pores 10-17, average 13.6. 

Coloration in life. The color of Mkonumbi specimens was much 
brighter than those from Karita River which presented an appearance 
very similar to the typical form though possessing the six lines of 
sextaeniata. 

At Mkonumbi the dorsal pair of lines are cream colored ; the dorso- 
lateral china white anteriorly, cream faintly tinged with yellow pos- 
teriorly; the outer lateral 'lines' are broken up into a series of pale 
yellow dashes below which is a series of pale green dots that touch the 
outermost row of ventrals. 

At Golbanti the brick-red ground color was brighter in some speci- 
mens than in others, the lines pure white, the 'line' of lateral spots 
white, the lower series grass-green. 

Measurements. The largest d 71 (M. C. Z. 41265) measures 158 (48 
+ 110) mm., the largest 9 (M. C. Z. 41259) measures 148 (48 + 
100) mm. 

Breeding. At Karita River, most, if not all, of the females were 
gravid with developing ova of 2.5x3.5 mm. to eggs of 11x6 mm. 
An Mkonumbi 9 held 4 eggs measuring 9x6 mm. At Karawa a young 
lizard was captured which measured only 63 (23 + 40) mm. 

Diet. As in the typical form, termites predominated, being present 
in the half-dozen Karita lizards examined, a large caterpillar was 
present in one. A caterpillar was also found in the stomach of an 
Mkonumbi specimen. 

GERRHOSAURIDAE 

Gerrhosaurus major major Dumeril 

Gerrhosaurus major Dumeril, 1851, Cat. Method, coll. Rept., Paris, p. 139: 

Zanzibar. 
Gerrhosaurus bergi Werner, 1906, Zool. Anz., 30, p. 54, figs. 1-3: Usambara 

Mountains, Tanganyika Territory. 



loveridge: African reptiles 309 

cf (M. C. Z. 41273) Voi, Kenya Colony. 10.iv.34. 
a* (M. C. Z. 41274) Mt. Mbololo, K. C. 22.iv.34. 

Distribution. Also 3 from Sokoki Forest (H. J. A. T.). 

Variation. Longitudinal rows of dorsals 19-20; longitudinal rows 
of ventrals 10; frontonasals paired as in bergi, in contact with the 
rostral in the 9 , separated from the rostral by the supranasals which 
form a broad suture in the d*. Fairly recently, I (1929, p. 66) discussed 
the instability of the head shields in this species. 

Measurements. The cf measures 440 (180 + 260) mm., the 9 
measures 518 (206 + 312) mm. 

Breeding. The latter holds a single developing, 114 mm., ovum in 
addition to numerous smaller ova. 

Parasites. Nematodes (Physaloptera sp.) were removed from the Voi 
specimen. 

Habitat. I dislodged her from among drifted leaves in a rock fissure 
in dense thorn scrub, after disturbing her basking in the vicinity. 

Gerrhosaurus flavigularis flavigularis Wiegmann 

Gerrhosaurus flavigularis Wiegmann, 1828, Isis, p. 379: "Africa merid. Krebs." 
Gerrhosaurus flavigularis forma intermedia Lonnberg, 1907, in Sjostedt, Kili- 
mandjaro, Meru Exped., 4, Rept. & Batr., p. 7, pi. figs. 1 a-b: "on the 
steppe near the Natron lakes, Kibonoto," Tanganyika Territory. 

12 (M. C. Z. 41275-9) Voi, Kenya Colony. 7-10.iv.34. 
4 (M. C. Z. 41280-2) Mt. Mbololo, K. C. 19.iv.34. 
1 (M. C. Z. 41283) Ngatana, K. C. 17.vi.34. 
1 (M. C. Z. 41284) Golbanti, K. C. 22.vi.34. 

Distribution. Also 7 from Sokoki Forest (H. J. A. T.). 

Native names. Malombo (Kitaita); nakavara (Kipokomo). 

Variation. Longitudinal rows of dorsals 20-22; longitudinal rows 
of ventrals 8; femoral pores 13-15, average 13.5; prefrontals broadly, 
or barely, in contact except in two Voi lizards where they are sep- 
arated; laterals keeled or smooth in Mbololo series. 

Measurements. All are considerably smaller than the largest 
Tanganyika records, the largest (M. C. Z. 41280) is a 9 measuring 
475 (142 + 333) mm., three others have a head and body length of 
142 mm. 

Breeding. In both Voi and Mbololo females are ova in all stages of 
development, the best developed consist of four eggs each almost 
ready for deposition; eggs in these three lots measure approximately 
27 x 15 mm., 24 x 11 mm., and 21x11 mm. respectively. 



310 bulletin: museum of comparative zoology 

Diet. Stomach contents consisted of: (1) grasshopper, (2) grass- 
hopper, (3) three grasshoppers, (4) grasshopper and grass, (5) a locust 
about four inches in length, (6) several crickets of two species, (7) two 
cockroaches and what appeared to be beetle remains, (8) scales of a 
large Gerrhosaurus, apparently part of the cast skin of the lizard in 
whose stomach they were found. 



SCINCIDAE 

Mabuya maculilabris (Gray) 
Euprepis maculilabris Gray, 1845, Cat. Liz. Brit. Mus., p. 114: West Africa. 

9 (M. C. Z. 41285-90) Mt. Debasien, U. 14-29.xi.33. 

1 (M. C. Z. 41291) Below Sipi, U. 15.xii.33. 

3 (M. C. Z. 41292-3) Kau, Tana R., K. C. 4.vi.34. 

1 (M. C. Z. 41294) Golbanti, K. C. 23.vi.34. 

5 (M. C. Z. 41295-7) Ngatana, K. C. 17-20.vi.34. 

1 (M. C. Z. 41298) Changamwe, K. C. 4.vii.34. 

Distribution. Also 1 from Sokoki Forest (H. J. A. T.). 

Native names. Lamatwan (Karamojong) ; kikomiakao (Lugishu); 
mvuvi (Kipokomo, but not specific). 

Variation. Midbody scale-rows 30-34, this range holds good for 
Mount Debasien alone, from whence came the only specimen with 30 
scales; keels on dorsal scales 5-9; supraoculars 4; supraciliaries 4-6, 
normally 5; prefrontals in contact in 4 skinks, separated in 16; su- 
pranasals in contact in 3 skinks, separated in 17. 

Coloration in life. A male from Mount Debasien had the sides of 
its head and throat lemon yellow shading to a brick orange on the 
flanks. The male from Sipi had the lower lip, anterior half of the body 
and the flanks suffused with red ljke the laterite soil of the vicinity 
where it was taken. 

My general impression in the field was that the Debasien series 
were more uniformly brown and rather more slender in habit than those 
taken at the coast and at Sipi on Mount Elgon where the rainfall, 
resulting in greater abundance of insect life, is more plentiful. The 
Sipi male and coastal series approximate to Sternfeld's race of major. 

Measurements. The largest c? (M. C. Z. 41293) measures 200 (86 
+ 124) mm., the largest 9 (M. C. Z. 41294) measures 218 (95 + 123) 
mm., but are surpassed in tail length by a cf with a 142 mm. tail and 
a 9 whose tail measures 146 mm. 

Breeding. Females were gravid both inland and on the coast, thus 



loveridge: African reptiles 311 

a 9 taken on Mount Debasien, November 29, 1933, held 5 eggs 
measuring 11x7 mm., an Ngatana 9 , taken June 17-20, 1934, held 
13 eggs (which must constitute a record for the species ?) measuring 
12 x 8 mm. 

Habitat. Between 5,000 and 7,000 feet on Mount Debasien where one 
was captured among undergrowth on an earthy bank, another on rocky 
ledges in a stream bed, a third basking high on a tree trunk, a fourth 
on a pile of vegetation and yet another beneath sacks in a tent. 
The Ngatana series were on the walls of a deserted building by the 
river bank. The Changamwe male in the matting sheath of a young 
coconut palm at a height of six feet from the ground. 

Mabuya planifrons (Peters) 

Euprepes (Euprepis) planifrons Peters, 1878, Monatsb. Akad. Wiss. Berlin, 

p. 203, pi. ii, fig. 2: Taita, Kenya Colony. 
Euprepes (Euprepis) taitanus Peters, 1878, Monatsb. Akad. Wiss. Berlin, p. 

203, pi. ii, fig. 3: Taita, Kenya Colony. 
Mabuia diesneri Sternfeld, 1911, Sitzber. Ges. Naturf. Freunde Berlin, p. 248: 

Kibwezi, Kenya Colony. 

1 (M. C. Z. 41311) Kibwezi, K. C. 23.iii.34. 

3 (M. C. Z. 41312) Voi, K. C. 9.iv.34. 

10 (M. C. Z. 41313-20) Mt. Mbololo, K. C. 23.iv.34. 
5 (M. C. Z. 41323-6) Lamu, Lamu Id., K. C. 7.V.34. 

2 (M. C. Z. 41327-8) Kitau, Manda Id., K. C. 15.V.34. 

1 (M. C. Z. 41329) Peccatoni, K. C. 26.V.34. 

2 (M. C. Z. 41321-2) Golbanti, K. C. 22.vi.34. 

4 (M. C. Z. 41330-3) Malindi, K. C. 28.vi.34. 

Distribution. Seen also at Witu and Mkonumbi, and 3 from Sokoki 
Forest (H. J. A. T.) examined. These records show that this species 
extends right up the coast and serves to link Lonnberg's record for 
Kismayu, Italian Somaliland and Parker's for Nogul Valley, British 
Somaliland with the type locality. 

It will be noted that a good series of topotypes of planifrons and 
taitanus were obtained on Mbololo in Utaita, in addition to a single 
topotype of diesneri during the week spent at Kibwezi. 

Native name. Jumbakoka (Kiamu). 

Synonymy. The topotype of diesneri served to confirm me in my 
action of referring this species to the synonymy of planifrons (1923, p. 
956). I now add taitanus which was based on a younger specimen than 
the type of planifrons. With one exception there is nothing in the 



312 bulletin: museum of comparative zoology 

descriptions which is not within the range of age and variation in 
planiceps. 

At the time of publication of the Catalogue of Lizards, vol. 3, only 
the types were known. They had, however, been seen by Boulenger 
(1887, pp. 167, 171) who differentiated them thus: 

Ear lobules, if present, short planifrons 

Ear opening partly concealed under the rounded scales of its anterior 
border taitana 

In one of my topotypes which has received a blow on one side of the 
head, the ear opening corresponds to the above description for taitana 
on that side but is like planifrons on the other. As, to the best of my 
belief, no second specimen of taitana has been taken since it was de- 
scribed nearly sixty years ago, I respectfully suggest that a careful 
reexamination of the type will show that the head has either been 
slightly damaged, or else forced back against the neck prior to fixation, 
which would result in the rounded scales of the anterior border of the 
ear-opening partly concealing the aperture. 

Variation. Midbody scale-rows 26-30, only one of the former and 
two of the latter, normally 28, three with 29 ; scales tricarinate except 
for three specimens which have quinquecarinate scales though the 
outermost pair are faint; supranasals in contact; prefrontals separated 1 
except for three individuals in which the frontonasal and frontal form 
a X suture; supraoculars 4; supraciliaries 4-7, normally 5; supralabials 
anterior to subocular 4-5, more frequently the former. 

Relative limb length is an age character, thus in a young skink 
measuring 30 mm. from snout to anus, the adpressed hind limb reaches 
to the elbow of the backward-pressed fore limb while in a 113 mm. 
adult the tips of the toes barely meet those of the fingers. 

Coloration. The great majority of these Kenya specimens differ 
from those of Tanganyika by the conspicuous lines of black spots 
down the length of the back, the nut-brown instead of khaki ground 
colour, and more vivid coloring generally. 

In this respect the Tanganyika skinks serve to link these typical 
planifrons with binotata Bocage of Angola (a valid species erroneously 
referred to the synonymy of quinquetaeniata by Boulenger (1887, p. 
198) ). As one proceeds northward planifrons appears to give rise to 
somalicus Calabresi distinguished by having 32 midbody scale-rows as 
a normal condition instead of a rare one as is the case with Kenya and 

'The condition in all the Tanganyika series listed in Bull. Mus. Comp. Zool., 74, p. 315 
where a misprint read "prefrontals in contact" instead of "prefrontals not in contact." 



loveridge: African reptiles 313 

Tanganyika planifrons. Ultimately both planifrons and somalicus may 
prove to be races of binotata which was described in 1867. 

Measurements. The largest cf (M. C Z. 41330) measures 253 (113 
+ 140) mm. but is surpassed in tail length by others with tails of 208 
and 205 mm., the largest 9 (M. C. Z. 41321) measures 339 (107 + 
232) mm. 

It will be seen that the tail may constitute two-thirds of the total 
length, a condition only met with in megalura so far as East African 
skinks are concerned. So long a tail, however, is exceptional for there 
seems to be a good deal of individual variation in this respect. 

Breeding. Testes large in several of the males examined, ova small 
in the largest 9 (M. C. Z. 41321) taken on June 22, 1934, when sun- 
ning with another on a hollow tree in scrub country. 

Diet. Stomachs examined held: (1) locusts, (2) 3 grasshoppers, 2 
large caterpillars, (3) 2 grasshoppers, 2 large caterpillars, (4) cricket 
and hard-shelled beetle, (5) a great many small beetles, (6) small 
beetles. 

Habitat. Captured in crumbling ruins of native huts at Voi. On 
Lamu Island they were fond of sunning on logs or piles of coconut- 
palm fronds, when seriously disturbed they would dash for the nearest 
palm and, ascending the stem with great celerity, soon be lost among 
the dense foliage above. On Manda Island one was basking on a stump 
at 9 a.m., the other on the stem of an acacia in the late afternoon. 
Perhaps the majority were found sunning on fallen logs for which they 
appeared to display a preference. 



Mabuya brevicollis (Wiegmann) 

Euprepes brevicollis Wiegmann, 1837, Arch, fur Natur., p. 133: Ethiopia (as 

Abyssinia) . 
Mabuya chanleri Stejneger, 1893, Proc. U. S. Nat. Mus., 16, p. 721: Tana River, 

Kenya Colony. 

1 (M. C. Z. 41299) Karita Camp, U. 9.xi.34. 

4 (M. C. Z. 41300-3) Kibwezi, K. C. 23-25.hi.34. 

3 (M. C. Z. 41304-6) Tsavo, K. C. 4.iv.34. 

3 (M. C. Z. 41307-9) Voi, K. C. 9.iv.34. 

1 (M. C. Z. 41310) Mt. Mbololo, K. C. 22.iv.34. 

Variation. Midbody scale-rows 32, except for M. C. Z. 41305 
which has 30; supranasals in contact in 6 specimens, separated in 6; 
prefrontals in contact in 4 specimens, separated in 8; frontal in con- 



314 bulletin: museum of comparative zoology 

tact with 1st and 2nd, or 1st, 2nd and 3rd supraoculars; supraoculars 
4; supraciliaries 4-6, average exactly 5. 

Coloration. Two young, measuring 50 and 54 mm. from snout to 
anus, were taken at Kibwezi and Mount Mbololo respectively. They 
are black with white spots and are so distinct from the adult coloring 
that it is small wonder that Stejneger made one the type of a sup- 
posedly new species, chanleri. Such young are rare in collections. 

Measurements. The largest specimens are all males, the maximum 
length from snout to anus is 150 mm. (M. C. Z. 41306), the longest 
tail measures 170 mm. (M. C. Z. 41301). 

Breeding. The testes of four males examined were small. 

Diet. Stomachs examined held: (1) locust, (2) grasshopper, (3) 
many small grasshoppers, (4) termites. 

Parasites. Nematodes of a new genus and species and 9 9 of 
an oxyuroid (Pharyyngodon sp.) were present in two of four skinks 
examined from Kibwezi and Tsavo. 

Enemies. One Voi specimen was recovered from the stomach of a 
sand snake (Psammophis biseriatus). 

Habitat. At Kibwezi these skinks occur among the piles of lava 
which cover so much of the country east and west of Kibwezi station. 
On being approached they slip quietly away so that it is more usual 
to obtain a glimpse of a vanishing tail than of a skink. The only 
way to obtain them was by shooting with dust shot from a .22 gun. 
Owing to the exceptionally hot weather, these skinks only appeared 
to bask for an hour or two in the early morning so that one had to 
restrict hunting to this limited period. I never saw more than two 
in an hour and considered myself fortunate if one of these was 
secured. 

The day after our arrival at Tsavo, I caught sight of one of these 
skinks disappearing among the ruins of a native hut; it had been 
basking on the crumbling lumps of mud of which the walls had been 
composed. Though I revisited the place time and again I never got 
another glimpse of the reptile. A native was set to work with a pick 
and shovel to remove the wall, a dusty and unsavoury task. After 
I had watched him remove half-a-cartload, we gave up, and I ordered 
him to set two snap-back rat traps baited with cheese. On previous 
occasions this method had proved successful, the skink being attract- 
ed by the insects which gathered on the bait. The following day we 
took the skink from the trap and shortly afterwards secured a second 
at the same spot, in the same trap. 



loveridge: African reptiles 315 

Mabuya megalura (Peters) 

Euprepes (Mabuia) megalura Peters, 1878, Monatsb. Akad. Wiss. Berlin, p. 
204, pi. ii, fig. 4: Taita, Kenya Colony. 

3 (M. C. Z. 41334-6) Kaimosi, K. C. 25.ii.34. 
1 (M. C. Z. 41337) Malindi, K. C. 29.vi.34. 

Distribution. A very fine specimen escaped me at Ngatana, it was 
clambering about in long grass at a height of three feet from the 
ground. 

Native name. Lubuyu (Luragoli). It is surprising that the Mara- 
goli do not distinguish between this lizard and the serpentiform 
Riopa anchietae, when chided about it they replied "Why should we? 
We don't eat them." 

Variation. Midbody scale-rows 26 in the plateau specimens, 22 in 
the skink from the coastal plain. This is a new low number, the range 
heretofore being 24-28, our other coastal material is normal. 

Coloration. Two of the three Kaimosi females have a lineolate ap- 
pearance below, this results from the longitudinal sutures of the breast 
and belly scales being plumbeous, the third female is uniformly white 
beneath. 

Breeding. These two largest Kaimosi females, measuring 73 mm. 
from snout to anus, are gravid, holding ten and eleven embryos re- 
spectively, these were fully formed and with very little yolk remaining. 

Diet. Stomachs examined, held (1) cricket, (2) cricket and spider, 
(3) spider, (4) two isopods in the Malindi skink. 

Mabuya quinquetaeniata obsti Werner 

Mabuia obsti Werner, 1913, Mitt. Nat. Mus. Hamburg, 30, p. 43: Kwa Mtoro, 

Central Province, Tanganyika Territory. 
Mabuia quinquetaeniata hildebrandti Sternfeld (not of Peters), 1917, Wiss. 

Ergebn. Zwei, Deutsch. Zent.-Afrika-Exped. 1910-1911, 1, p. 438, pi. xxiv, 

fig. 3: Taita, Kenya Colony. 
Intermediates 

1 (M. C. Z. 41338) Kacheliba, U. 8.xi.33. 
4 (M. C. Z. 41339-42) Karita River, U. 9.xi.33. 
1 (M. C. Z. 41343) Nabagut, Greeki R., U. 6.xii.33. 
1 (M. C. Z. 41344) Elgonyi, K. C. 22.L34. 

Typical obsti 

3 (M. C. Z. 41345-7) Tsavo, K. C. 30.iii.34. 
45 (M. C. Z. 41348-79) Mt. Mbololo, K. C. 16-26.iv.34. 



316 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY 

Distribution. Skinks from Ethiopia, northern Uganda, and northern 
Kenya are intermediate between typical quinquetaeniata of Egypt and 
the Sudan and the larger tropical race obsti which I now recognize 
for the first time as a result of studying the long series of topotypic 
hildebrandii from Mbololo in Taita. These intermediates agree with 
typical quinquetaeniata in the number of midbody scale-rows, but 
with obsti in the immaculate throat of the male. 

Native names. Longisia (Kimasai); malasagasa (Kitaita, but not 
specific). 

Variation. (Intermediates) Midbody scale-rows 36-40, average 
38.1; prefrontals in contact in eight skinks, separated in two; supra- 
ciliaries 5-6, average 5.6; supralabials anterior to subocular 4. 

(Typical obsti) Midbody scale-rows 42-48, average 44.2; prefrontals 
in contact in twenty-four skinks forming a X with frontal and fronto- 
nasal in two, separated in twenty-two; supraciliaries 5-7, average 6.2; 
supralabials anterior to subocular 4, except on one side in three skinks 
which have 5; the subocular of one skink is not larger than the ad- 
jacent labials. 

It might be remarked that in the fifty-eight Egyptian skinks counted 
by Anderson (1898, pp. 187-193) the range of midbody scale-rows is 
35-42, average 37.6. 

Measurements. The largest cf (M. C. Z. 41357) measures 252 (101 
+ 151) mm., the largest 9 (M. C. Z. 41379) measures 239 (97 + 
142) mm. 

Breeding. The only gravid 9 was one from Mount Mbololo which 
held 7 eggs measuring 17 x 9 mm. 

Diet. Stomachs examined, held: (1) termites and ants, (2) termites 
and ants, (3) termites, ants and grasshopper, (4) termites, grasshopper, 
beetle and millipede, (5) termites and caterpillar, (6) grasshopper, (7) 
large orthopteran, (8) cricket, (9) cricket, (10) cricket. 

Parasites. Nematodes (gen. et sp. n.) were present in skinks from 
Karita River and Mount Mbololo. 

Habitat. Like the hyrax, this skink depends on rocky outcrops 
which restrict its distribution. At Kacheliba, Elgonyi and Mbololo 
they were on kopjes, along the Karita, Greeki and Tsavo Rivers 
among exposed rocks. 

Mabuya varia varia (Peters) 

(Plate 8, fig. 1) 

Euprepes (Euprepis) varius Peters, 1867, Monatsber. Akad. Wiss. Berlin, p. 20: 
Tete, Mozambique. 



loveridge: African reptiles 317 

6 (M. C. Z. 41380-3) Above Sipi, U. 7,000 ft. 14.xii.33. 

77 (M. C. Z. 41384-449) Kaburomi, U. 10,000 ft. 28-31. xii.33. 

7 (M. C. Z. 41450-3) Madangi, U.11,500 ft. 3-4.i.34. 
2 (M. C. Z. 41454) Kibwezi, K. C. 2,985 ft. 23.iii.34. 

6 (M. C. Z. 41455-8) Mt. Mbololo, K. C. 4,000 ft. 16.iv.34. 
1 (M. C. Z. 41459) Gongoni, K. C. 200 ft. 27.vi.34. 

Natives names. Kihunya (Lugishu); malasagase (Kitaita, but not 
specific). 

Variation. The long series from the alpine zone of Mount Elgon, 
were collected as they seemed to me to be different from those with 
which I was familiar at lower levels. Apart from smaller size, however, 
critical examination in the laboratory has failed to reveal any charac- 
ter which may not be matched sometime by a skink from lower levels. 
In this connection see also the remarks on coloration. The following 
statistics while revealing the variation in the whole series, apply equally 
well both in range and average number of scale-rows to the montane 
(7,000 to 11,500 feet) group alone, nor do the latter exhibit any varia- 
tion in these characters, except for carination as noted, which is not 
shared by the small series from lower levels. 

Midbody scale-rows 30-34, average 32.1, except for the coastal 
(Gongoni) specimen which has only 28 rows, the first record of so low 
a number; scales tricarinate in all except three Mbololo skinks in which 
a few scales are quattuor- or quinquecarinate ; prefrontals in contact 
in 11 skinks, forming an X with frontal and frontonasal in two, 
separated in seventy-seven; frontoparietals distinct; subocular border- 
ing lip in all. 

Coloration. That of adult (not young) males from Sipi is suffused 
with rufous on the flanks both above and below the lateral line; 
below they are bright mustard yellow from nape to anus, almost all 
the gular, and in some specimens the ventral and subcaudal scales as 
well, are edged with black. 

At Kaburomi, I was impressed by the striking resemblance in dorsal 
markings and coloration which these skinks bore to Algiroides alleni 
whose habitat and habits seemed similar, below, however, alleni was 
bright reddish orange. 

The dorsolateral line in the montane specimens was reduced to 
hair-like proportions, in typical specimens it is ribbon-like, forming a 
white band almost a scale wide. At Kibwezi, however, the band is 
lacking and I noted in the field that the coloration was "quite black," 
the skink being shot while basking on the black lava which covers so 
much of the countryside in the vicinity of the station. 



318 bulletin: museum of comparative zoology 

Measurements. Of the montane lizards, the largest cf (M. C. Z. 
41412) measures 133 (52 + 81) mm., and largest 9 (M. C. Z. 41398) 
141 (63 + 78) mm., being about 10 mm. less in length from snout to 
anus than large lowland skinks. 

Breeding. Very many of the Kaburomi females are gravid; of 
half-a-dozen dissected one held 8 ova measuring about 8 mm. in 
circumference, the rest had from 4 to 6 embryos in every stage of 
development. A Kibwezi skink held embryos almost ready for par- 
turition. One female in the Mbololo series was gravid, its ova being in 
about the same stage of development as those of the Kaburomi skink 
with 8 eggs. 

Diet. Stomachs examined, held: (1) beetle, (2) beetle, hemipteron, 
caterpillar, (3) cricket, spider, (4) cricket and own sloughed scarf-skin. 

Parasites. Nematodes (Thubunaea asymmetrica) , and a new genus 
and species, were common in the Kaburomi skinks. 

Enemies. Augur Buzzards, harriers and kestrels were seen, both at 
Kaburomi and Madangi, quartering over the alpine meadows where 
these skinks appeared to be the dominant form of vertebrate life. 
There were five skinks in the stomachs of one buzzard (Buteo r. augur), 
and one in another, shot at Kaburomi; one skink in a kestrel (Falco t. 
tinnunculus) shot near Madangi. 

It was interesting to note that only two of these seven skinks had 
dropped their tails when captured, that the little creatures live a 
hazardous life in these uplands is obvious from the fact that the 
great majority of the ninety collected had reproduced tails: the latter 
may be distinguished by their having single transversely enlarged 
scales below, instead of the numerous small scales present on the 
original portion. 

Habitat. Astonishingly abundant in the alpine meadows, these skinks 
behave in much the same way as does the lizard (Algiroides alleni), 
which see for further details. 

Mabuya striata (Peters) 

Tropidolepisma striatum Peters, 1844, Monatsber. Akad. Wiss. Berlin, p. 36: 
Mozambique. 

3 (M. C. Z. 41461-2) Mt. Debasien, U. 16.xi.33. 
1 (M. C. Z. 41463) Sabei, Elgon, U. 9.xii.33. 

1 (M. C. Z. 41464) Sipi, Elgon, U. 27.xii.33. 

2 (M. C. Z. 41465) Butandiga, Elgon, U. 8.i.34. 
5 (M. C. Z. 41466-8) Kirui's, K. C. 19.i.34. 

5 (M. C. Z. 41469-71) Kaimosi, K. C. 10.ii.34. 
2 (M. C. Z. 41472) Mt. Mbololo, K. C. 17.iv.34. 



loveridge: African reptiles 319 

2 (M. C. Z. 41473) Lamu Island, K. C. 7.V.34. 
1 (M. C. Z. 41474) Mombasa Id., K. C. 3.vii.34. 

Distribution. Seen also at Kitale, Nabagut and Molo. 

Native names. Lamatwan (Karamojong) ; matundi (Lugishu); 
longisia (Kimasai); lisiagali (Luragoli); lisiakali (Lutereki); malasa- 
gasa (Kitaita, but not specific); jumbakoka (Kiamu). 

Variation. Midbody scale-rows 34-40 for Debasien series alone, or 
exclusive of Debasien series ; keels 3-7, an age character as young have 
only three, subsidiary keels, usually faint but sometimes strongly de- 
veloped, appear in older skinks ; prefrontals in contact in three skinks, 
separated in nineteen ; frontoparietals distinct except in a Lamu skink 
where fusion has almost certainly resulted from an injury when young. 

Measurements. The largest skink, a d" (Lamu), measures 208 (93 
+ 115) mm., but the tip of its tail is missing; the largest 9 (M. C. Z. 
41466) measures 93 mm. from snout to anus but the tail is regenerating. 

Breeding. On January 8, a 9 from Butandiga held 6 spherical ova 
about 7 mm. in diameter; on January 19, one 9 from Kirui's village 
held 6 ova about 8 mm. diameter, while another had 6 very young 
embryos; on February 10, a 9 from Kaimosi had 6 large, but unpig- 
mented, embryos. 

Diet. Two skinks held beetles, two termites, two a grasshopper each, 
while one held a cockroach and snail, the shell of the latter gone. 

Parasites. A nematode was recovered from only one stomach out 
of ten examined. 

Enemies. Striped skinks were recovered from the stomachs of a hawk 
(Melierax m. metabates) on Debasien, a wolf snake (Lycophidion c. 
capense) and a green snake (Chlorophis hoplogaster) at Sipi, and from 
the latter species only at Kaimosi. 

Habitat. In miombo bush on Debasien, where it ascends trees and 
behaves like maculilabris , all three were shot on trees; elsewhere 
chiefly obtained on the walls of rest camps or native huts. The Sipi 
skink lived for several days in the tent, apparently sleeping in the 
tent pockets. The tent was struck at night. Several days after our 
arrival at Kaburomi, where the species does not occur, its flattened 
tailless body was found in one of the tent pockets. 

Mabuya irregularis Lonnberg 
(Plate 8, fig. 2) 
Mabuia (striata ? var.) irregularis sp. n. ? Lonnberg, 1922, Arkiv for Zool., 14, 
p. 4: Soy, Kenya Colony. 

9 (M. C. Z. 41460) Kaburomi, 10,500 ft., U. 28.xii.33. 



320 bulletin: museum of comparative zoology 

Distribution. Though only about forty miles northwest of the type 
locality, this record not only constitutes the first occurrence of this 
montane form on Mount Elgon, but an addition to the herpetofauna 
of Uganda. Formerly known from Mount Kenya and the Aberdare 
Range. 

Variation. Midbody scale-rows 34; supraciliaries 4; parietals 
separated posteriorly by the interparietal, each parietal shield bordered 
by 4 scales only 1 of which is enlarged so that there is but a single pair 
of nuchals enlarged. 

Measurements. This 9 measures 150 (75 + 75) mm. 

Breeding. Five scaled and pigmented embryos, within about a week 
of parturition, display the most astonishing variation in development. 
Of three measured, one is 64 (31 -f- 33) mm., another 66 (29 + 37) 
mm., and the third 47 (23 + 24) mm. 

Diet. The stomach held two species of weevil, an ant (probably 
Phcidole sp.), two parasitic hymenoptera (Proctotrypidae), two cater- 
pillars and two species of spiders. I am indebted to Dr. Joseph 
Bequaert for these identifications. 

Parasites. Nematodes (Thubunaea asymmetrica) present in the 
stomach. 



Riopa mabuiiformis Loveridge 

Riopa mabuiiformis Loveridge, 1935, Bull. Mus. Comp. Zool., 79, p. 12: 
Ngatana, Tana River, Kenya Colony. 

11 (M. C. Z. 40266-71) Ngatana, Tana River, K. C. 14-19.vi.34. 

These are the type and paratypes of a smooth-scaled species with 
28 to 30 midbody scale-rows. 

Native name. Lukumvivi (Kipokomo for all three species of Riopa 
occurring in this locality). 

Diet. Stomachs examined, held: (1) large black cricket and own 
slough, (2) both items as in one, (3) grasshopper. 

Habitat. Taken beneath vegetable debris heaped on raised banks 
separating the flooded rice-fields just north of the grove of mango 
trees. 



Riopa tanae Loveridge 

Riopa tanae Loveridge, 1935, Bull. Mus. Comp. Zool., 79, p. 11: Kau, near the 
mouth of the Tana River, Kenya Colony. 



loveridge: African reptiles 321 

14 (M. C. Z. 40251-9) Kau, Tana River, K. C. 4.vi.34. 

1 (M. C. Z. 40261) near Witu, north of Kau, K. C. 31.V.34. 

2 (M. C. Z. 40264-5) Ngatana, Tana R., K. C. 19.vi.34. 
2 (M. C. Z. 40262-3) Golbanti, Tana R., K. C. 23.vi.34. 

These are the type and paratypes of a smooth-scaled species with 
22 to 24 midbody scale-rows. Native name as for last species. 

Breeding. Most, or all, of the adult females in the series from Kau 
were gravid, holding from 2 to 3 ova or embryos, some of the latter 
were fully formed, furnishing evidence that the species is viviparous. 
On June 23, a Golbanti female held 3 eggs measuring 9x5 mm. 

Diet. Crickets were definitely present in two skinks but the food 
was so well masticated in the other specimens examined that it was 
indeterminable. 

Riopa sundevallii (Smith) 

Eumices (Riopa) sunderallii A. Smith, 1849, Illus. Zool. South Africa, 3, App., 
p. 11: Natal. 

2 (M. C. Z. 41475-6) Mt. Debasien, U. 22.xi.33. 

1 (M. C. Z. 41477) Sipi, Mt. Elgon, U. 12.xii.33. 

2 (M. C. Z. 41478-9) Kibwezi, K. C. 23.iii.34. 
29 (M. C. Z. 41480-99) Voi, K. C. 7-12.iv.34. 

2 (M. C. Z. 41500) Mt. Mbololo, K. C. 17.iv.34. 
9 (M. C. Z. 41501-7) Lamu Id., K. C. 5.V.34. 

1 (M. C. Z. 41508) Mkonumbi, K. C. 28.V.34. 

2 (M. C. Z. 40260, 41509) Witu, K. C. 31.V.34. 

20 (M. C. Z. 41510-23) Ngatana, K. C. 12-19.vi.34. 

1 (M. C. Z. 41524) Golbanti, K. C. 22.vi.34. 

2 (M. C. Z. 41526) Malindi, K. C. 29.vi.34. 

2 (M. C. Z. 39967-8) Sokoki Forest, K. C. vi.32. H.J.A.T. 

Native navies. Kilimagonde (Kitaita, but not specific); kiumambusi 
(Kiamu); lukummvi (Kipokomo). 

Variation. Midbody scale-rows 24-28, only one Debasien and two 
Mbololo skinks with 28, average 25; rostral separated from fronto- 
nasal except in three Ngatana skinks which agree with production 
Boulenger in this respect; frontonasal broader than long except in 
three skinks (M. C. Z. 41498-9, 41526) where it is longitudinally 
divided; prefrontals and frontoparietals distinct; nostril between 
supranasal and two smaller nasals except in one Ngatana skink where 
the two nasals are fused into a single scale on the right side only; 
second toe well in advance of fifth. 



322 bulletin: museum of comparative zoology 

Coloration in life. Halfgrown specimens at Voi were noted as being 
plumbeous, or black, above; light below. The throat of an adult 
Ngatana male was a beautiful lemon-yellow color. 

Measurements. The largest d 71 (M. C. Z. 41481) measures 193 
(122 + 71) mm., but his tail is regenerated; the largest 9 (M. C. Z. 
41480) measures 225 (118 + 107) mm. 

Breeding. In mid-June, at Ngatana, one skink held 4 eggs measur- 
ing approximately 13 x 7 mm., another held eggs 18 x 9 mm. On 
June 19, I came upon six eggs measuring 19 x 11 mm. in a cavity in 
a large termitarium, situated at the edge of a rice-swamp ; many adult 
skinks were disturbed and captured during the process of demolishing 
the termites' hill. On July 4, four of these eggs were found to have 
hatched, the emerged young measuring 62 (24 + 38) mm. 

Diet. Stomachs examined held: (1) caterpillar, (2) grasshopper, 
(3) grasshopper and hemipteron of the Thyreocoridae, (4) beetle 
larvae, (5) termites, (6) minute snails, (7) sandhopper at Lamu. 

Enemies. Sundevall's skinks were recovered from the stomach of 
a wolf snake (Lycophidion c. capense) at Kibwezi, and a burrowing 
viper (Atractaspis bibronii) at Chamgamwe. One Ngatana skink 
had no digits on the hind limbs, the right fore limb had a bud, while 
stumps of digits remained on the left fore limb. 

Riopa modestum modestum (Giinther) 

Sepacontias modestus Giinther, 1880, Ann. Mag. Nat. Hist. (5), 6, p. 235: 
Mpwapwa, Ugogo, Tanganyika Territory. 

5 (M. C. Z. 41527-30) Tsavo, K. C. 3.iv.34. 
19 (M. C. Z. 41531-45) Voi, K. C. 7-13.iv.34. 
5 (M. C. Z. 41546-50) Mt. Mbololo, K. C. 23-29.iv.34. 

Relationships. Recently, Parker (1932, pp. 357-361) has thrown 
considerable light on the tangled status and synonymy of the north- 
east African skinks of this group, further he has provided a really 
excellent and workable key. 

At the time of its appearance, I had some manuscript in hand in 
which I adduced reasons for regarding modestum as a race of sunde- 
vallii, and let this go to press (1933, p. 322) rather than reinvestigate 
the whole business. I am now entirely convinced that Parker's ar- 
rangement is the correct one and that I was wrong in laying such 
emphasis on the abnormal individuals which I regarded as inter- 
mediates. Further proof is now furnished by the fact that both 
sundevaUii and modestum occur together at Voi and Mbololo. 



loveridge: African reptiles 323 

Variation. Midbody scale-rows 24-27, average 24.6, definitely less 
than in Tanganyika Territory material ; rostral separated from fronto- 
nasal; prefrontals distinct; nostril between a supranasal and postnasal 
only; second toe well in advance of the fifth. 

Coloration in life. The males at Voi were a beautiful lemon-yellow 
shade below, similar to the throat of a male sundevallii. 

Measurements. None exceeded measurements previously given for 
this species, several had a snout to anal length of 80 mm., their tails 
being regenerated. 

Breeding. In early June most of the Voi females held 2, rarely 3, 
developing ova; of these one pair measured 5 mm. in diameter, two 
pairs 7 mm., and a fourth lot of three eggs 13 x 6 mm. 

Diet. Stomachs examined held: (1) caterpillar, (2) grasshopper, 
(3) grasshopper and earwig, (4) cricket, (5) termites. 

Habitat. As Tsavo is the type locality of Lygosoma gromeri, I made 
an especial attempt to secure skinks of this group. I captured a female 
modest urn beneath a heap of tanning bark that had been lying aban- 
doned for years. Putting it in a test-tube, I sent it round to the few 
natives attached to this isolated station. An Mtaita brought in three 
which he had hoed up in his garden, uufortunately all had lost their 
tails as he was afraid of them. I do not think that gromeri is an 
aberrant modestum, it seems to be a Siaphos with a divided fronto 
rostral, otherwise near kilimensis. It should be looked for on the 
highest kopjes near Tsavo, where forest may have existed in past 
centuries. 

Riopa pembanum (Boettger) 

Lygosoma (Riopa) pembanum Boettger, 1913, in Voeltzkow, Reise in Ost- 
Afrika, 3, p. 350, pi. xxiv, figs. 4^5: Pemba Island. 

1 (M. C. Z. 39969) Sokoki Forest, K. C. (H.J.A.T.) vi.32. 

Distribution. Parker (1932, p. 361) was the first to record this 
skink from the African mainland, this addition to the Kenya fauna 
being based on two skinks in the British Museum from Takaungu 
which lies between the Sokoki Forest and Pemba Island. It occurs 
with sundevallii at Sokoki, has 24 midbody scale-rows and the pre- 
frontals lacking, being fused with the frontonasal. 

Riopa anchietae (Bocage) 
(Plate 7, fig. 2) 

Eumecesa anchietae Bocage, 1870, Jorn. Sci. Lisboa, 3, p. 67, pi. i: Huilla Pla- 
teau, Mossamedes, Angola. 



324 bulletin: museum of comparative zoology 

1 (M. C. Z. 41551) Kirui's Village, K. C. 23.L34. 
27 (M. C. Z. 41552-71) Kaimosi, K. C. 10-28.ii.34. 

Native name. Luhuyu (Luragoli and Lutereki). 

Variation. Midbody scale-rows 22-26, only M. C. Z. 41558 with 
22 and M. C. Z. 41563 with 26; fingers 2, toes 3, except for a few 
skinks which appeared to have lost their digits azygously by accident, 
for example M. C. Z. lacks the right fore limb while its right hind 
limb is devoid of digits. 

Measurements. The largest perfect c? (M. C. Z. 41551) measures 
492 (171 + 321) though surpassed in body length by many in the 
Kaimosi series which have regenerating tails. 

Breeding. In mid-February four Kaimosi females held from 4 to 8 
embryos, the highest number being in the largest female, which 
ranged from very minute embryos to fully formed and pigmented 
skinks measuring 128 (52 + 76) mm. 

Diet. Stomachs examined held: (1) to (5) grasshoppers, (6) grass- 
hopper, caterpillar, termites and a millipede, (7) cockroach and its 
egg-case, caterpillar, weevil, (8) beetles, moth, (9) grasshopper, wasp, 
(10) large spider. In addition two of these skinks had a mass of their 
own sloughed scales in their stomachs. 



Siaphos kilimensis (Stejneger) 

Lygosoma kilimensis Stejneger, 1891, Proc. U. S. Nat. Mus., 14, p. 405: Kili- 
manjaro, Tanganyika Territory. 

Lygosoma clathrotis Boulenger, 1900, Ann. Mag. Nat. Hist. (7), 6, p. 194: 
Foot of Mount Kenya, Kenya Colony. 

3 (M. C. Z. 41572-4) Voi, K. C. 10.iv.34. 
15 (M. C. Z. 41575-85) Mt. Mbololo, K. C. 16-26.iv.34. 

Distribution. I am a little doubtful of the data of the Voi material, 
it is possible that I collected them along the Voi River and confused 
them in the field with Rio pa m. modestum to which they bear so strong 
a resemblance in coloration and habit, though not in habitat. Alter- 
natively they may have been brought in to my Voi camp by a native 
who may possibly have secured them on one of the neighbouring hills. 

Native name. Kilimagonde (Kitaita, but not specific). 

Affinities. Boulenger referred his clathrotis to the section Liolepisma. 
I have reexamined the specimen from Nairobi River (U. S. N. M. 
42517) which I referred to clathrotis (1929, p. 78) and find it indistin- 
guishable from kilimensis. It is true that the type of clathrotis was 



loveridge: African reptiles 325 

said to have 22 midbody scale-rows but, though rare, skinks with this 
number occur alongside those with 24 on Mbololo. Moreover Dr. 
Malcolm Smith informs me that he has reexamined the type of 
clathrotis and finds it had 24 (not 22) midbody scale-rows. He confirms 
my action in referring clathrotis to the synonymy of kilimensis. 

Variation. Midbody scale-rows 22-24, all but two with 24; lamellae 
under fourth toe 12-15; limbs pentadactyle. 

Measurements. The largest c? (M. C. Z. 41577) measures 173 (72 
+ 101) mm., the largest 9 is 2 mm. less in length from snout to anus, 
tail regenerated. 

Breeding. Only two of the females are gravid. On x\pril 16, I found 
24 eggs, each measuring about 14 x 8 mm., beneath a single log in 
the forestry nursery at the edge of the rain forest. The eggs contained 
young skinks some of which were found to measure 49 (23 + 26) mm. 
It was difficult to say how many eggs constituted a clutch as most 
were scattered about, one group, however, well separated from the 
rest, comprised four eggs which agrees with the number previously 
recorded by Barbour & Loveridge (1928, p. 163) as found in a gravid 
female in the Uluguru Mountains. In addition to the 24 fresh eggs, 
there were many others which had hatched out and appeared to have 
belonged to a previous breeding season. 

Returning to the forestry nursery on the following day, April 17, 
I uncovered 68 more fresh eggs of which about 50 were found under 
one log. This log was favourably situated on the east side of the 
nursery close to the western edge of the rain forest where the sun 
would not reach it until noonday. In the loose soil adjacent to the 
eggs, five adult skinks were taken of which two were gravid and had 
presumably come to this spot to lay. It should be remarked that 
there were an abundance of logs lying along the forest edge yet the 
skinks had concentrated their laying under a few. The finding of these 
eggs appears to point to a definite breeding season. Hatched-out 
shells from a previous season were present beneath this log also. 

On April 23, two of the eggs hatched out skinks which measured 58 
(26 + 32) and 56 (26 + 30) mm. respectively. On April 26, two more 
hatched. I then packed all the eggs in appropriate soil and mailed 
them to the London Zoological Gardens, but they did not hatch. 

Diet. Stomachs examined, held: (1) spider, (2) spider and cater- 
pillar, (3) spider and hymenopteran, (4) spider and woodlouse, 
(5) caterpillar. 

Parasites. Nematodes (gen. et sp. n.) were present in the stomach 
of one lizard. 



326 bulletin: museum of comparative zoology 

Enemies. One of these skinks was recovered from the stomach of a 
house snake (Boaedon lineatus) which I found in the hollow log be- 
neath which so many eggs were located on April 17. The house snake 
was only halfgrown, the skink was a very large one. 

Habitat. These skinks were in the habit of basking on sunlit patches 
beside the paths which penetrated the forest, others were captured at 
the base of wild banana plants in the forest. 

Ablepharus boutonii africanus Sternfeld 

Ablepharus boutoni africanus Sternfeld, 1918, Abh. Senckenb. Nat. Ges., 36, 
p. 423: Manda Id., Malindi and Pemba Id. 

3 (M. C. Z. 41586-7) Lamu Id., K. C. 12.V.34. 
10 (M. C. Z. 41588-95) Kitau, Manda Id., K. C. 15.V.34. 
6 (M. C. Z. 41596-600) Malindi, K. C. 30.vi.34. 

Distribution. The Kitau and Malindi series are topotypes of this 
race. 

Variation. Midbody scale-rows 22-24, average 22.7. 

Measurements. The largest cf (M. C. Z. 41593) measures 107 (44 
+ 63) mm., the largest 9 (M. C. Z. 41595) measures 111 (45 + 66) 
mm. 

Breeding. In mid-May, three females on Lamu and Manda Islands 
each held two eggs measuring 12 x 5 mm., at the latter locality one 
held a pair of eggs measuring 7x4 mm., another had small ova. At 
Malindi breeding appeared to be over on June 30. 

Diet. The stomachs of ten specimens examined held remains of 
minute insects (nymphal orthopterans ?) and Crustacea too masticated 
to be identifiable except for a cockroach and beetle. 

Habitat. These littoral skinks were very common on the rocks 
forming the breakwater along the sea front by the wireless station on 
Lamu Island. They were presumably introduced with the rock as all 
rock on Lamu has been imported by dhow. 

Abundant on the rock masses scattered along the shore near Ras 
Kitau, where they appear to be feeding more on insects than is the 
case at Mombasa. These specimens were killed with dust shot from .22. 

At Malindi they were plentiful on the coral cliffs at the extreme south 
end of the bay, about a mile south of the township. Naturally they 
do not occur elsewhere in the sandy bay or near the town. My am- 
munition being exhausted, I armed myself with a towel and pair of 
long forceps. The towel should be rolled up lengthwise and the two 



loveridge: African reptiles 327 

ends held in the hand. Having approached within striking distance, 
a sudden blow is given with the soft towel, this usually only discon- 
certs the skink, at most stunning it momentarily so that it is necessary 
to pounce upon the reptile with all possible speed. If it has fallen, or 
slipped into one of the numerous fissures, the forceps will be found 
invaluable for recovering the little reptile without injuring its tail. 



Ablepharus wahlbergii (Smith) 

Cryptoblepharus wahlbergii Smith, 1849, Illus. Zool. S. Africa, 3, App. p. 10 : 

Natal. 
Ablepharus carsonii Boulenger, 1894, Proc. Zool. Soc. London, p. 735, pi. xlix, 

figs. 4-4a: Fwamba, Northern Rhodesia. 
Ablepharus massaiensis Angel, 1924, Bull. Mus. Hist. Nat. Paris, p. 52: Masai 

Plains near Nairobi, Kenya Colony. 

23 (M. C. Z. 41601-10, 41614-7) Kaimosi, K. C. ii.34. 
1 (M. C. Z. 41611) Voi, Coast Province, K. C. 9.iv.34. 
1 (M. C. Z. 41612) Golbanti, Tana R., K. C. 2.V.34. 
1 (M. C. Z. 41613) Lamu, Lamu Island, K. C. 5.V.34. 

Native name. Lisiakali (Lutereki, but not specific). The Watereki 
natives who gave me this name believed this little skink to be the 
young of Mabuya striata. When the difference was pointed out, they 
still maintained that lisiakali was the only name that they had for 
it, and asked what reason was there that they should have another 
seeing that it was not good to eat. 

Synonymy. A. carsonii was differentiated from wahlbergii because 
of the fusion of the frontoparietals with the interparietal to form 
a single large shield. Four of the skinks in the Kaimosi series agree 
with carsonii in this respect and others from this locality show inter- 
mediate conditions of parietal fusion. A. carsonii was based on a 
single specimen which I must assume was an aberrant individual. 

The aberration is, however, of major importance as it throws such 
individuals into the first section of Boulenger's (1887, p. 344) key. 
So important a variation should not be lost sight of. It will be observed 
that the head shields of this species are subject to considerable fusion. 

Variation. Midbody scale-rows 22-26, average 23.8; frontoparietals 
and interparietal fused to form a single shield (as in carsonii and 
boutonii) in four skinks (M. C. Z. 41614-7), the frontoparietals are 
fused but interparietal distinct (i.e. like wahlbergii) in the rest of 
the series except for a few where the separation is ill-defined; pre- 
frontals separated except in M. C. Z. 41601 in which the left pre- 



328 bulletin: museum of comparative zoology 

frontal is fused with the frontal, and M. C. Z. 41616 where they are 
in contact; supraoculars 3 (wahlbcrgii) except in two skinks which 
have 3-2 and two others which have only 2 (massaiensis) as the result 
of fusion. 

Measurements. The largest is a 9 (M. C. Z. 41608) measuring 131 
(53 + 78) mm. 

Breeding. Between February 17 and 28, most of the Kaimosi 
females appear to be gravid, three examined held 4, 5, and 6 eggs 
respectively, the lots measuring approximately 3 x 5, 4 x 4 and 4x8 
mm. A young skink, presumable nearly a year old, measured 44 
(19 + 25) mm. 

On May 5, at Lamu, several skinks were seen among the fallen 
leaves which blanketed the sand about the base of an old mango tree. 
One egg was found beneath the leaves, it contained an embryo mea- 
suring 13 mm. from snout to anus, the tail tip now missing. 

Diet. Stomachs examined, held: (1) termites, (2) termites, (3) 
three termites and four spiders, (4) a beetle and apparently a beetle 
larva. 

Enemies. A Wahlberg's Skink was recovered from the stomach of 
a Cape Wolf Snake (Lycophidion c. capense) at Kaimosi. 

Habitat. The Voi skink was taken among the crumbling ruins of 
a native hut on the Msinga Estate. 



Acontias percivali Loveridge 

Acontias percivali Loveridge, 1935, Bull. Mus. Comp. Zool., 79, p. 13: Foot of 
Mount Mbololo, Taita Mountains, Kenya Colony. 

39 (M. C. Z. 40175-200) Mt. Mbololo, K. C. 26-30.iv.34. 

Distribution. The majority of these, the type series, came from 
Kitibu at the west, or northwest foot of the mountain. 

Breeding. No signs of it in fifteen specimens examined. 

Diet. Stomachs examined, held: (1) termites, (2) termites, (3) 
termites and ants, (4) ants' heads, (5) beetle, (6) beetle, (7) beetle 
larva, (8) beetle larva, (9) hairy caterpillar, (10) caterpillars and small 
cricket, (11) millipede. In addition two skinks had their own sloughs 
in stomach or intestines and many had much sandy soil of the dis- 
trict, apparently ingested with food. 

Parasites. Nematodes (Thelandros sp. n.) were present in the intes- 
tines of three of the largest specimens examined. 



loveridge: African reptiles 329 

CHAMAELEONTIDAE 

Chamaeleon senegalensis Daudin 

Chamaeleo senegalensis Daudin, 1802, Hist. Nat. Rept., 4, p. 203: Region 
watered by the Senegal and Niger Rivers; Gambia and Guinea. 

Chamaeleon laevigatus Gray, 1863, Proc. Zool. Soc. London, p. 95: 500 miles 
south of Khartoum, Anglo-Egyptian Sudan. 

1 (M. C. Z. 41618) w. foot of Mt. Debasien, U. 16.xi.33. 
4 (M. C. Z. 41619-22) Lukungu, w. foot Mt. Elgon, U. 8.L34. 
13 (M. C. Z. 41623-32) Bukori, s. foot Mt. Elgon, K. C. 18.i.34. 

Native name. Agiar (Karamojong, but not specific). 

Synonymy. Dr. H. Heckenbleikner, in his forthcoming monographic 
revision of the genus, can find no grounds for regarding laevigatus 
(hitherto applied in a subspecific sense to East African specimens) 
as distinct from senegalensis. 

Measurements. The largest c? (M. C. Z. 41619) measures 184 
(108 + 76) mm., the largest 9 (M. C. Z. 41623) measures 170 
(105 + 65) mm. 

Breeding. On January 8, the two females from Lukungu were 
gravid with 14 eggs measuring 6 mm. diameter, and 7 eggs measuring 
7 mm. diameter respectively. On January 18, however, half-a-dozen 
females from Bukori exhibit undeveloped ova. The difference may be 
explainable by prolonged drought at Bukori? 

Diet. Stomachs examined, held: (1) beetles, (2) flies, (3) ants?, 
(4) ants?, (5) grasshopper. 

Habitat. The Debasien specimen was found clinging to some stout 
grass in an area swept over by grass fires the previous day though 
somewhat patchily burnt. 

Chamaeleon gracilis gracilis Hallowell 

Chamaeleo gracilis Hallowell, 1842, Journ. Acad. Nat. Sci. Philad., p. 324, pi. 
xviii: Monrovia, Liberia. 

1 (M. C. Z. 41633) Mt. Debasien, 5,000 ft., U. 24.xi.33. 
3 (M. C. Z. 41634-6) Lukungu, w. foot Mt. Elgon, U. 8.i.34. 
26 (M. C. Z. 41637-69) Bukori, s. foot Mt. Elgon, K. C. 18.i.34. 

Native name. Agiar (Karamojong, but not specific). 

Measurements. The largest cf (M. C. Z. 41637) measures 209 
(105 + 104) mm., the largest 9 (M. C. Z. 41638) measures 249 
(135+ 114) mm. 



330 bulletin: museum of comparative zoology 

Sexual dimorphism. Before dissection, I sorted the material into 
two groups, viz. those with a spur and those without. Dissection 
demonstrated that each of the 10 spurred specimens was a male, each 
of the 30 devoid of spurs proved to be a female. In no instance had 
the original sorting to be amended or reversed. In C. g. etiennei 
Schmidt, of Banana, Belgian Congo, the male is spurless like the 
female. 

Breeding. The testes of the males appeared to be enlarged, the ova 
in all thirty females was not, or only just beginning to be, enlarged. 

Diet. Stomachs examined, held: (1) four half-grown praying man- 
tids, katydid, cockroach, Camponotus ant, beetle, (2) three agrionid 
dragonflies, numerous beetles, spider, (3) remains of several agrionid 
dragonflies. 



Chamaeleon dilepis roperi Boulenger 

Charnaeleon roperi Boulenger, 1890, Proc. Zool. Soc. London, p. 85, pi. viii, 
fig. 4: Kilifi, north of Mombasa, Kenya Colony. 

29 <M. C. Z. 41670-98) Kibwezi, K. C. 25-30.iii.34. 
2 (M. C. Z. 41699-700) Voi, K. C. 7-10.iv.34. 
7 (M. C. Z. 41701-0) Mt. Mbololo, K. C. 17-27.iv.34. 

Distribution. Also 14 from Sokoki Forest (H. J. A. T.) seen at 
Nairobi. 

Native name. Malunge (Kitaita, but only generic). 

Variation. All the males, and there are some from each locality, 
agree with the type in being devoid of tarsal spurs. See also remarks 
under dilepis. 

Measurements. The largest cf (M. C. Z. 41670) measures 215 (118 
+ 97) mm.; the largest 9 (M. C. Z. 41702) measures 220 (128 + 92) 
mm. 

Breeding. Towards the end of March all the Kibwezi females appear 
to be gravid, one examined held 26 eggs measuring 10 mm. in diameter. 
A month later on Mount Mbololo, a female held 23 eggs measuring 
14 x 7 mm. 

Diet. Stomachs examined, held: (1) praying mantis, (2) grass- 
hoppers. 

Enemies. Roper's Chameleons were recovered from the stomach of 
a Lizard Buzzard (Asturinula m. monogrammica) at Voi, a sand snake 
(Psammphis biseriatus) and two Boomslangs {Dispholidus typus) on 
Mount Mbololo. 



loveridge: African reptiles 331 

Chamaeleon dilepis quilensis Bocage 

Chamaeleo dilepis var. quilensis Bocage, 1866, Jorn. Sci. Math. Phys. Nat. 
Lisboa, 1, p. 59: Rio Quillo, Angola. 

d 1 d 1 (M. C. Z. 41707-8) Lamu Island, K. C. 7.V.34. 
c? 9 (M. C. Z. 41709-10) Peccatoni, K. C. 24.V.34. 

9 (M. C. Z. 41711) Mapenya, K. C. 28.V.34. 
d" 9 (M. C. Z. 41713-4) Mkonumbi, K. C. 28.V.34. 

Distribution. It was a most disconcerting discovery to me to find 
these undoubted quilensis occurring in the small area about Lamu 
Island, for a suggested explanation see remarks under dilepis. 

Native name. Gegeuka (Kiamu). 

Variation. All the males agree with the type in having tarsal spurs 
and minute occipital lobes. 

Measurements. The largest d 1 (M. C. Z. 41707) measures 200 (104 
+ 96) mm.; the largest 9 (M. C. Z. 41711) measures 262 (135 + 127) 
mm. 

Breeding. On May 28, the Mapenya female held 36 eggs measuring 
9x8 mm., but about the same date the ova in both Peccatoni and 
Mkonumbi females was only slightly developed. 

Diet. Stomachs examined, held: (1) grasshopper, (2) three grass- 
hoppers, (3) three grasshoppers and a spider. 

Chameleon dilepis dilepis Leach 

Chamaeleo dilepis Leach, 1819, in Bowdich, Miss. Ashantee, App. p. 493: 
Gaboon. 

c? (M. C. Z. 41715) Ngatana, Tana River, K. C. 13.vi.34. 

Native name. Lumvivi (Kipokomo, but not specific). 

Variation. This male has tarsal spurs and relatively large occipital 
lobes. 

Measurements. <? measures 194 (94 + 100) mm. 

Discussion on status of the forms. The determination and distribution 
of the members of this group have given considerable trouble to 
taxonomists so that the records are in a chaotic state and the distri- 
bution appears even more inexplicable than is actually the case. 

The key furnished by Boulenger (1887, p. 440) cannot be improved 
upon so far as it goes, but his species parvilobus has long been recognized 
as a synonym of quilensis; and roperi was undescribed at that time, 
having been long confounded with quilensis or with dilepsi in the 
literature. 



332 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY 

Two points have to be borne in mind. Firstly, that it is usually 
impossible to distinguish very young specimens, for their lobes are 
insufficiently developed. Secondly, the females of quilensis are indis- 
tinguishable from those of roperi so that it is essential to collect males. 

The following key will serve to separate the males : 

Males without a tarsal process or spur on hind foot, occipital dermal 

lobes only slightly developed C. d. roperi 

Males with a tarsal process or spur on hind foot 

Adults (and the young) with occipital dermal lobes only just 

movable C. d. quilensis 

Adults (but not young) with occipital dermal lobes well developed 
flaps C. d. dilepis 

In view of the fact that Madagascar and East Africa form the centre 
of distribution of the Chamaeleontidae, it is suggested as a working 
hypothesis to explain the somewhat anomalous distribution of these 
forms, that C. d. roperi formed the parent stock. 

Its distribution, so far as we know, extends from Kilifi (its type 
locality) and Mombasa on the coast, inland to Kilimanjaro and north 
through Voi, Mount Mbololo, Mtito Andei and Kibwezi to Fort Hall, 
Meru and Mount Jombeni, just north of Mount Kenya. In addition, 
however, we must admit the occurrence of a few spurless males in a 
series of spurred males from Ifakara, near Mahenge, southeast Tanga- 
nyika Territory. The several Somaliland records should be rechecked. 

Assuming that quilensis was an offshoot of roperi, the former must 
have been the dominant form over the greater part of Africa till dilepis 
developed, possibly in the Mozambique-Southwest Tanganyika- 
Rhodesia region where we find chameleons today with occipital lobes 
of intermediate size. Formerly I (1933, p. 331) referred these to 
quilensis but now consider them to be nearer to typical dilepis though 
admittedly occupying an intermediate position. 

The typical form of dilepis appears to reach its maximum size and 
best development of occipital lobes in Central Tanganyika but it has 
pushed out westwards till it occupies the greater part of Central 
Africa while quilensis occurs on the periphery of its range. 

With typical dilepis occurring on the Tana River between the ranges 
of roperi and quilensis, and with quilensis and roperi persisting in the 
Iringa-Ifakara country, it seems quite possible that the study of more 
extensive material may show the position to be even more involved 
than appears at present. At some not too distant date I hope to 
investigate and revise all the East African records in the literature. 



loveridge: African reptiles 333 

Chamaeleon bitaeniatus bitaeniatus Fischer 

(Plate 9, fig. 1) 

Chamaeleo bitaeniatus Fischer, 1884, Jahrb. Hamb. Wiss. Anst., 1, p. 23, pi. ii, 
fig. 7: Masailand, East Africa. 

7 (M. C. Z. 41715-21) Bukori, K. C. 18.L34. 
52 (M. C. Z. 41722-50) Kaimosi, K. C. 7-28.ii.34. 

Native name. Invambu (Luragoli and Lutereki). 

Measurements. The largest d* (M. C. Z. 41744) measures 143 (78 + 
65) mm., the largest 9 (M. C. Z. 41724) measures 148 (92 + 56) mm. 

Breeding. On January 18, at Bukori, all six females were gravid 
with spherical eggs, these numbered 11, 12, 13, 16, 17 and 20 re- 
spectively and all were about 7 mm. in diameter. In mid-February 
at Kaimosi, most females appeared gravid with eggs bearing well- 
formed embryos, one lot numbered 13 eggs measuring 9 x 10 mm. 

Diet. Stomachs examined held: (1) flies and a stick insect, (2) flies 
and a caterpillar. 

Enemies. Four of these chameleons were recovered from the 
stomachs of three Boomslangs (Dispholidus typus) at Kaimosi. On 
February 25, as I was standing beneath a very tall tree in camp, 
two adult male chameleons fell fighting at my feet. They were green 
as they landed, turned very dark, separated, then hastily strode to 
the trunk of the tree and reascended its rough bark. Two chameleons 
from this locality had truncated tails, one a mere stump. 

Chamaeleon bitaeniatus hohnelii Steindachner 
(Plate 9, fig. 2) 

Chamaeleon hohnelii Steindachner, 1891, Sitzber. Akad. Wiss. Wien, 100, 
part 1, p. 309, pi. i, fig. 2: Laikipia, Kenya Colony. 

52 (M. C. Z. 41751-69) Sipi, U. 12-24.xii.33. 

1 (M. C. Z. 41770) Bulambuli, U. 4.i.34. 

59 (M. C. Z. 41771-95) Butandiga, U. 5-15.L34. 
4 (M. C. Z. 41796-9) Budadiri, U. 17.i.34. 

2 (M. C. Z. 41800) Elgonyi, K. C. 25.i.34. 

Distribution. The first four of these localities are on western Elgon 
between 4,000 and 9,000 feet, the last on the southern slope at 
about 6,000 feet. 

Native name. Ikanyafu (Lugishu). 



334 bulletin: museum of comparative zoology 

Measurements. The largest <? (M. C. Z. 41754) measures 199 (100 
+ 99; mm., the largest 9 (M. C. Z. 41751) measures 190 (100 + 
90) mm. ; the smallest only 86 (48 + 38) mm., except for one mentioned 
below. 

Breeding. From December 12 to January 17, almost all females 
held 9 to 22 eggs, or embryos, according to the size of the mother. 
A few held eggs only 8 mm. in diameter, but the majority consisted 
of large, irregular-shaped membranes containing embryos many of 
which had little yolk left. On January 25, at Elgonyi, a native brought 
in the first newly born chameleon which we had seen, it measured 
only 45 (25 + 20) mm. 

Diet. Stomachs examined held: (1) ants, beetle, hemipteron, (2) 
ants, beetles, cockroach, flies, (3) flies, (4) flies and a large piece of its 
own scarf epidermis. 

Enemies. Hohnel's Chameleons were recovered from the stomachs 
of four green snakes (Chlorophis hoplogaster) at Sipi and Butandiga, 
also from a Boomslang (Dispholidus typus) at Sipi. 

Chamaeleon bitaeniatus altaeelgonis Loveridge 

Chamaeleon bitaeniatus altaeelgonis Loveridge, 1935, Bull. Mus. Comp. Zool., 
79, p. 15: Kaburomi, 10,500 feet, Mt. Elgon, Uganda. 

52 (M. C. Z. 40274-40300) Kaburomi, U. 27-28.xii.33. 
2 (M. C. Z. 41801-2) Madangi, U. 3-4.i.34. 

Remarks. As might be expected, this dwarf, montane race from the 
alpine zone of Mount Elgon, produces fewer eggs than does its larger 
relative, C. b. hdhnelii, which occurs lower down the same mountain 
from 5,000 to 7,000 feet. 

Breeding. On December 28, ten of the thirty females in the type 
series were examined and found to be gravid with 5, 6, 6, 7, 7, 8, 9, 
9, and 10 ova of which the smallest measured 7 mm. diameter, in 
nine females the embryos were in various stages of development, one 
lot quite ready for birth. 

Diet. Ten stomachs examined all held finely masticated little flies 
and beetles, other identifiable material was: (1) cricket, (2) cricket, 
grasshopper, froghopper, and (3) caterpillar. 

Parasites. Nematodes (Strongyluris ? media) were abundant in the 
intestines of the type series. 

Enemies. Elgon Chameleons were twice recovered from the stomachs 
of Augur Buzzards (Buteo r. augur) at Kaburomi, one bird held four. 



loveridge: African reptiles 335 

Chamaeleon fischeri tavetensis Steindachner 

Chamaeleon tavetensis Steindachner, 1891, Sitzber. Akad. Wiss. Wien, 100, 
part 1, p. 310, pi. i, figs. 3-3a: Taveta Forest, south foot of Kilimanjaro, 
Tanganyika Territory. 

Chamaeleo abbotti Stejneger, 1891, Proc. U. S. Nat. Mus., 14, p. 353, text fig.: 
At 4,600 feet, Kilimanjaro, Tanganyika Territory. 

& ? (M. C. Z. 41863-4) Mt. Mbololo, K. C. 17.iv.34. 

Distribution. To the best of my knowledge, this species has not 
previously been taken in the Taita Mountains, the name "taitensis 
Steindachner" used by Stejneger (1893, p. 724) and again by Tornier 
(1896, p. 57) is a nomen nudum substituted for tavetensis. 

Affinities. From its geographical position between the Nguru 
mountains (type locality of fischeri) and Meru, Moi nt Kenya (type 
locality of C. f. excubitor), this small form snoUd be regarded as 
a race of fischeri. 

Native name. Malunge (Kitaita, but generic). 

Measurements. The d" measures 143 (4 + 62 + 77) mm. from end 
of the 4 mm. horns; the 9 measures 121 (56 + 65) mm. 

Breeding. On April 17, the female held 4 spherical ova 6 mm. in 
diameter. 

Diet. Stomachs examined, held: (1) beetle and small grasshopper, 
(2) weevil, membracid, hemipteron, muscid fly and an ant. 

Brookesia kerstenii kerstenii (Peters) 

Chamaeleo kerstenii Peters, 1868, Monatsb. Akad. Wiss. Berlin, p. 449: Wanga 
{i.e. Vanga), near Mombasa, Kenya Colony. 

9 (M. C. Z. 41805) Voi, K. C. 10.iv.34. 

9 (M. C. Z. 41806) Mt. Mbololo, K. C. 24.iv.34. 

c? (M. C. Z. 41807) Witu, K. C. 31.V.34. 

9 (M. C. Z. 41808) Ngatana, K. C. 12.vi.34. 

Native name. Lumvivi (Kipokomo, but not specific). 

Coloration in life. The broad rim round the inner aspect of both upper 
and lower jaws was bright orange in the large female from Ngatana. 

Measurements. The cf measures 59 (37 + 22) mm. ; the largest 9 
(M. C. Z. 41808) measures 79 (55 + 24) mm. 

Breeding. On April 24, on Mbololo, a female held 3 eggs about 3.5 
mm. in diameter; on June 12, at Ngatana, about 6 of the ova were 
slightly developed to 2 mm. diameter. 

Diet. Stomachs examined, held: (1) two grasshoppers, (2) grass- 
hopper and spider, (3) two ants. 



.336 bulletin: museum of comparative zoology 



BIBLIOGRAPHY 
Ahl, Ernst 

1933. "Zur Kenntnis der afrikanischen Wuhlschlangen der Gattung 
Eryx." Sitz. Ges. naturf. Freunde Berlin, pp. 324-326, figs. 

Anderson, John 

1898. "Zoology of Egypt, 1, Reptilia and Batrachia." London. 

Barbour, Thomas and Loveridge, Arthur 

1928. "A Comparative Study of the Herpetological Fauna of the Ulu- 
guru and Usambara Mountains, Tanganyika Territory, with De- 
scriptions of new Species." Mem. Mus. Comp. Zool., 50, pp. 
87-265, pis. i-iv. 

BOULENGER, G. A. 

1887. "Catalogue of Lizards in the British Museum (Natural History)." 
Ed. 2, 3, London. 

1893. "Catalogue of Snakes in the British Museum (Natural History)." 

1. London. 

1894. "Catalogue of Snakes in the British Museum (Natural History)." 

2. London. 

1896. "Catalogue of Snakes in the British Museum (Natural History)." 

3. London. 

1921. "Monograph of the Lacertidae." 2. London. 

Flower, S. G. 

1933. "Notes on the recent Reptiles and Amphibians of Egypt, with a 
list of the species recorded from that Kingdom." Proc. Zool. Soc. 
London, pp. 735-851, map. 

Hewitt, John 

1935. "Some new forms of Batrachians and Reptiles from South Africa." 
Rec. Albany Museum. Grahamstown. pp. 283-357, pis. xxvii- 
xxx vi. 

LoNNBERG, ElNAR 

1922. "Sammlungen der schwedischen Elgon-Expedition im Jahre 
1920. 6. Reptiles." Arkiv Zool., Stockholm, 14, No. 12, pp. 1-8. 

Loveridge, Arthur 

1916. "Report on the Collection of Ophidia in the Society's Museum." 
Journ. East Africa and Uganda Nat. Hist. Soc, 5, pp. 77-87. 

1923. "Notes on East African Lizards collected 1920-1923, with the 
description of two new races of Agama lionotus Boulenger." 
Proc. Zool. Soc. London, pp. 935-969. 



loveridge: African reptiles 3-37 

1929. "East African Reptiles and Amphibians in the United States 
National Museum." Bull. U. S. Nat. Mus., No. 151, pp. 1-135, 
pl.i. 

1933. "Reports on the Scientific Results of an Expedition to the South- 
western Highlands of Tanganyika Territory. VII. Herpetology." 
Bull. Mus. Comp. ZooL, 74, pp. 197-416, pis. i-iii. 

Nieden, Fritz 

1913. "Neues Verzeichnis der Kriechtiere (ausser den Schlangen) von 
Deutsch-Ostafrika. I. Reptilia." Mitt. Zool. Mus. Berlin, 7, 
pp. 51-100. 

Peracca, M. G. 

1909. in Abruzzi, "Spedizione al Ruwenzori di S. A. R. il Principe L. 
Amedeo di Savoia. Rettili ed Anfibi", 1, pp. 165-180. Milan. 

Schmidt, K. P. 

1923. "Contributions to the Herpetology of the Belgian Congo based on 
the Collection of the American Museum Congo Expedition, 1909- 
1915. Part II.— Snakes." Bull. Amer. Mus. Nat. Hist., 49, pp. 
1-146, pis. i-xxii. 

Stejneger, Leonard 

1893. "On some Collections of Reptiles and Batrachians from East 
Africa and the adjacent Islands, recently received from Dr. W. L. 
Abbott and Mr. William Astor Chanler, with Descriptions of New 
Species." Proc. U. S. Nat. Mus., 16, pp. 711-741. 

Tornier, Gttstav 

1896. Die Reptilien und Amphibien Ost-Afrikas. In Deutsch-Ost- 
Afrika, 3, Lief, iii-iv. 



EXPLANATION OF PLATES 



PLATE 1 



Loveridge — African Reptiles. 



PLATE 1 

Fig. 1. Babcock's Leopard Tortoise (Testudo pardalis babcocki). 

Ventral aspect of Type 9 (M. C. Z. 40,003) from Mount Debasien. On 
ordinary ground these fine tortoises soon right themselves by setting up an 
oscillatory motion. Occasionally, however, one finds the remains of an in- 
dividual which has lost its balance among rocks, and in consequence met with 
a miserable death. 

Fig. 2. Babcock's Leopard Tortoise (Testudo pardalis babcocki). 

This species is often called the mountain tortoise on account of the prefer- 
ence which it exhibits for rocky, hilly country. The race is characterized by the 
highly vaulted carapace when contrasted with the more depressed shell of the 
typical form from southwest Africa. 



BULL. MUS. COMP. ZOOL. 



Loveridge. African Reptiles. Plate 1 





PLATE 2 



Loveridge — African Reptiles. 



PLATE 2 

Fig. 1. Black Terrapin (Pelusios nigricans nigricans). 

One of several from Kaimosi, Kakamega, Kenya Colony, whose carapace 
and plastron exhibit severe injuries, presumably resulting from the unsuccess- 
ful attack of a powerful-jawed carnivore, such as a hyena. Apparently, how- 
ever, the injury had not impaired the health of the water tortoise. 

Fig. 2. Ventral view of a Black Terrapin from Kaimosi. 

In Tropical Africa, during the prolonged dry season, many of the semi- 
aquatic tortoises bury, and become imprisoned, in mud. They are released by 
the first torrential downpour of the 'big rains' and set forth still plastered with 
mud, as is the figured specimen, in search of suitable stagnant water. 



BULL. MUS. COMP. ZOOL. 



Loveridge. African Reptiles. Plate 2. 





PLATE 3 



Loveridge — African Reptiles. 



PLATE 3 

Fig. 1. A Brown House Snake (Boaedon lineatus) at Kaimosi. 

More often black when adult, this species owes its name in part to the 
frequency with which it is found about dwellings, where it is often killed under 
the impression that it is a "young black mamba." Actually it visits houses in 
search of rodents; the thirty-eight inch specimen figured, had swallowed a 
black rat which measured seven and a half inches from the tip of its snout 
to the root of its tail. 

Fig. 2. Cape Wolf Snake (Lycophidion capense capense). 

The wolf snakes have been so-named on account of their teeth, but, like the 
house snakes, they are non-poisonous and deserve protection on account of 
their diet, which, in the more slender wolf snakes, consists very largely of mice, 
though lizards and even other snakes are sometimes eaten. 



BULL. MUS. COMP. ZOOL. 



Loveridge. African Reptiles. Plate 3. 





PLATE 4 



Loveridge — African Reptiles. 



PLATE 4 

Fig. 1. An Egg-eating Snake (Dasypeltis scaber) at Sipi. 

This reptile subsists entirely on eggs, chiefly those of small birds, but also 
of game birds. A hen's egg has been placed in the mouth of this specimen to 
demonstrate the remarkable, though normal, distension of the gape; As its 
teeth are vestigial, the snake effects the breaking of the egg only when it 
reaches the gullet, at which point certain specialized hypophyses of the 
vertebrae fracture the shell; in this way none of the contents are lost. 

Fig. 2. Eastern Jameson's Mamba (Dendraspis j. kaimosae). 

One of the type series from Kaimosi. This new form is characterized by its 
uniformly black tail and lower subcaudal count. The specimen photographed 
shot through the spokes of a native's cycle, became wedged in the forks, and 
was brought into camp alive. Mambas are the most poisonous of all African 
snakes. 



BULL. MUS. COMP. ZOOL. 



Loveridge. African Reptiles. Plate 4 . 



fwmmm 











' l3F- : ■ flmS 














v\ 


*t> ammmm * 




V 






'*! 




1* ' ij^^^ • i '• i[inv' ; riM— i 




■*. 


■*~'-" "£e 


> 


^ 


J 






*'''■:- ... .■ ■■ . ;,:■ 


V* 



PLATE 5 



Loveridoe — African Reptiles. 



PLATE 6 

Fig. 1. Puff Adder (Bitis arietans) striking at the photographer. 

This photograph is chiefly interesting as showing how very little of even so 
heavy-bodied a species as a Puff Adder, remains on the ground when a snake 
strikes. The camera used, a Baby Rolleiflex, necessitated the operator crouch- 
ing within three feet of the subject. After several successful exposures had 
been made, the snake, which had been growing more and more irate, launched 
itself at the photographer's face at a moment when she was preoccupied with 
the view finder. Sudden and unexpected as was the attack, the author's stick, 
to be seen in the photo, was just in time to intercept the snake when only a 
few inches from her face. 

Fig. 2. Rhinoceros Viper (Bitis nasicornis) at Kaimosi, Kakamega. 

These beautiful though deadly reptiles are excessively abundant in this 
locality, where they attain a length of forty inches. They feed upon shrews, 
mice and toads, and are themselves preyed upon by civets and the white-tailed 
mongoose. As many as forty-five young are produced at a birth. 



BULL. MUS. COMP. ZOOL. 



Loveridge. African Reptiles. Plate 5. 










if&^pr^NF "W 1 ^ 







PLATE 6 



Loveridge — African Reptiles. 



PLATE 6 

Fig. 1. A Rain-forest Gecko (Cneniaspis africanus elgonensis). 

This photograph is of the Kaimosi specimen which was not wholly typical 
of the new race discovered on Mount Elgon. Each female lays two hard- 
shelled eggs in cavities of rotting wood, hollow trees, or beneath logs. The 
young emerge in mid-December. These geckos feed upon crickets as well as 
smaller insects, while they themselves are preyed upon by green snakes 
(Chlorophis hoplogaster) . 

Fig. 2. The Rock-dwelling Elgon Agama (Agama agama elgonis). 

These gorgeous, red-headed lizards, with backs of dark blue spotted with 
lighter and tails alternately banded with the same shades, bask upon, or dash 
about, the rocky outcrops of the mountain. The photograph, taken at Sipi, 
is of an adult male. Ants and termites furnish them with their principal food. 
The race acts as host to numerous nematodes. 






BULL. MUS. COMP. ZOOL. 



Loveridge. African Reptiles. Plate 6 





PLATE 7 



Loveridge — African Reptiles. 



PLATE 7 

Fig. 1. Allen's Montane Lizard (Algiroides alleni Barbour). 

This interesting inhabitant of the alpine zone of Mount Elgon, from whence 
it is now recorded for the first time, was discovered by Dr. G. M. Allen on 
Mount Kenya, later it was found on Mount Kinangop. The large series ob- 
tained at Kaburomi, 10,500 feet, provides useful data as to variation, diet and 
parasites. 

Fig. 2. A Skink {Riopa anchietae) with rudimentary limbs. 

This Angolan lizard reaches its northeasterly limits at Kaimosi in Kenya, 
where this photograph was taken. The young are apparently born in March 
when from four to eight are produced. These defenceless creatures, frequently 
killed in mistake for snakes, feed upon grasshoppers, weevils, caterpillars and 
the like. 



BULL. MUS. COMP. ZOOL. 



Loveridge. African Reptiles. Plate 7. 





PLATE 8 



Loveridge — African Reptiles. 



PLATE 8 

Fig. 1. A Variable Skink (Mabuya varia varia) from Elqon. 

Probably the only African lizard which ranges from sea level to 12,000 feet, 
where it is abundant among the heather-like tussocks of the alpine zone. 
Many fall victims to birds of prey; no fewer than five of these skinks were re- 
covered from the stomach of a single Augur Buzzard (Buteo r. augur), another 
from a migrant Kestrel (Falco t. tinnunculus) . Photographed at Kaburomi. 

Fig. 2. Irregular Skink (Mabuya irregularis) on Mount Elgon. 

Another inhabitant of the alpine zone at Kaburomi, now recorded from this 
mountain, as well as from Uganda, for the first time. Taken in December, this 
gravid female held five young almost ready for birth. She had recently eaten 
a weevil, caterpillar and two spiders, which shared her stomach with parasitic 
nematodes (Thunbunaea asymmetrica) . 



BULL. MUS. COMP. ZOOL. 



Loveridge. African Reptiles. Plate 8. 




Mm* 



KI *f .' 4 fc 




PLATE 9 



Loveridge — African Reptiles. 



PLATE 9 

Fig. 1. A new-born Chameleon (Chamaeleon b. bitaeniatus). 

Unlike many other species of the genus which are hatched from eggs, the 
young of bitaeniatus are born alive in batches of from eleven to twenty. At 
Kaimosi this takes place apparently in March when the rainy season com- 
mences. The little reptile is grasping a coin the size of a quarter. 

Fig. 2. Von Hohnel's Chameleon (Chamaeleon b. hohnelii). 

These bearded chameleons were photographed at Butandiga, 7,000 feet, on 
western Elgon, Uganda. At this locality and nearby Sipi, four green snakes 
(Chlorophis hoplogaster) and Boomslang (Dispholidus typus) were found to 
have fed upon these creatures. It is obvious that something is instrumental 
in reducing their numbers for they are very prolific, from nine to twenty-two 
being produced at a birth. 



BULL. MUS. COMP. ZOOL. 



Loveridge. African Reptiles. Plate 9. 









\ 



-4 «->, 



Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 
Vol. LXXIX, No. 6 



SCIENTIFIC RESULTS OF AN EXPEDITION 
TO RAIN FOREST REGIONS IN EASTERN AFRICA 

VI 
NEMATODA 



By J. H. Sandgkound 



CAMBRIDGE, MASS., U. S. A. 
PRINTED FOR THE MUSEUM 
November, 1936 



No. 6. — Reports on the Scientific Results of an Expedition to the 
Rain Forest Regions in Eastern Africa 

VI 

Nematoda 

By J. H. Sandground 

In the course of his 1933-1934 expedition to Kenya Colony and 
Uganda, as a Fellow of the Guggenheim Memorial Foundation, Mr. 
Arthur Loveridge collected a small quantity of helminthological ma- 
terial. Subsequently in the course of making a systematic study of the 
food habits of the many reptiles collected, Mr. Loveridge has en- 
countered many other instances of parasitism. Where parasites un- 
covered in this way have proved of special interest, it has been possible 
to examine the alimentary tracts of a large series of alcohol or formalin 
preserved hosts and in this way to secure a much more useful collection 
than could be obtained in the field. Although some of the worms col- 
lected many months after their death are not preserved in as effective a 
manner as they might have been if secured alive, most of them are still 
suitable for study. This method of gathering nematodes has the special 
advantage of permitting a very thorough inspection of the gut contents. 
As a result, the taxonomically all-important males, elusive because 
of their minute size and relative rarity, can usually be detected. 

A new genus and several new species of nematodes are described in 
this paper. The collection also contained a number of forms that have 
been previously observed and described. Only in one instance (see p. 361 
of this paper) is their occurrence in a new host or locality deemed of 
sufficient helminthological interest to warrant a special account; how- 
ever, these new records are incorporated in Mr. Loveridge's herpeto- 
logical papers, under the caption "parasites." 

The type material of the new species described in this paper are de- 
posited in the Helminthological Collection of the Museum. 



.342 



bulletin: museum of comparative zoology 



SPIRUROIDEA 

Paraspirura mabuyae gen. et sp. nov. 

Host: Mabuya brevicollis Wiegmann. 
Location: Stomach (free in lumen). 
Locality: Kibwezi or Tsavo, Kenya Colony. 

This worm is represented by 9 male and 12 female specimens. The 
translucent white body is fairly straight in the preserved state, taper- 
ing gradually towards the extremities, and relatively slender. Cuticu- 
lar striae conspicuous, 4 /jl apart. The head, about 60/z wide, is not set 
off from the rest of the body by a constriction or inflation of the cuticle. 
Two small, simple, lateral lips or pseudolabia, each bearing towards its 




Fig. 1. Paraspirura mabuyae. Frontal view of head. 



base two pairs of papillae of which the inner are minute, the outer 
fairly large and prominent; in addition, somewhat anterior to the 
papillae in the median field of each lip is a relatively large circular 
amphid. Cuticular elevations in the form of narrow membranous 
shields (? interlabia) form a bridge between the lips dorsally and ven- 
trally. On the internal face of each lip near its apex is a minute tooth- 
like projection, and behind this is a serrated edge that recalls the 
denticular ridges of some species of Abreviata. From the frontal view 
(Fig. 1) the oral aperture is seen to be dorso-ventrally elongated with 
a pronounced constriction in the middle. The lips are not superficially 
lobed though lateral and dorsal views of the head (Figs. 2 and 3) 
show grooves and cuticular flanges on their internal faces. 

Buccal cavity cylindrical, 45 to 55 n long and about 30 fj. wide, with 
delicate, lightly cuticularized walls. 



sandground: African nematoda 



343 



Cervical papillae acicular, symmetrical, located in front of nerve 
ring which encircles the middle of the muscular part of oesophagus. 
Excretory pore conspicuous, about 0.10 mm. behind nerve ring. 




Fig. 2. Paraspirura mabuyae. Dorso-ventral optical section. 

Male. Length 12.6 to 11 mm., maximum width near equator 0.33 
to 0.37 mm. Total length of oesophagus 4.03 to 5.2 mm., of which the 
anterior muscular part measures 0.23 to 0.27 mm. Caudal alae wide, 




Fig. 3. Paraspirura mabuyae. Lateral optical section. 



becoming confluent with body contour about 0.2 mm. in front of 
cloaca and extending posteriorly to tip of tail. Cuticle on venter orna- 
mented with a delicate pattern of linear and small tessellated markings. 



344 



bulletin: museum of comparative zoology 



Cloacal aperture salient, and large. Tail 0.23 to 0.25 mm. long. There 
are fqur pairs of lateral stalked papillae equidistantly spaced in front 
of anus. A median papilla with relatively large end-organ is found on 
the anterior rim of the cloaca. Post-anally there are five pairs of papil- 
lae: the first pair are sessile and situated towards the median line; a 
larger pair of stalked papillae near the middle of the tail ; between the 




Fig. 4. Paraspirura mabuyae. Tail of male; ventral view. The cuticular 
ornamentation is not represented in the figure. 



last mentioned and the tip of the tail is a pair of very small lateral 
sessile papillae; adorning the tip of the tail are two pairs of lateral 
papillae with short peduncles in addition to the minute pores of the 
caudal glands. Spicules unequal and dissimilar; left spicule is the 
more robust and measures 0.23 to 0.24 mm. and about 0.035 mm. wide 
in the middle of the shaft; it is superficially marked with strong irregu- 
lar lines. Right spicule thinner, more lightly cuticularized, measures 



sandground: African nematoda 



345 



0.16 to 0.18 mm. by 0.015 mm. broad. Gubernaculum, apparently 
v shaped or triangular in ventral view, measures about 0.060 mm. but 
lightly cuticularized and consequently difficult to define. 

Female. Length 17 to 21 mm.; maximum width 0.40 mm. Muscular 
oesophagus 0.35 to 0.37 mm.; glandular oesophagus 4.3 to 5.4 mm. 
Body tapers posteriorly to form a gracefully curved tail with rounded 
end provided with two minute spike-like papaillae. Anus slightly 




Fig. 5. Paraspirura mabuyae. Female genitalia from dissection. 

salient, 0.37 to 0.41 mm. from extremity. Vulva with slightly tumid 
lips situated at the beginning of the posterior third or fourth of the 
body, 4.1 to 6.3 mm. from the extremity. Vagina directed posteriorly. 
Ovejector thick-walled, 0.4 mm. long, continues into a short (0.25 
mm.) so-called common trunk formed from the union of the two uteri 
which soon bend anteriad then diverge to run in opposite directions. 
Eggs rounded oval, with thick smooth shells averaging 42 /x by 33 ju; 
embryonated when discharged. 



SYSTEMATIC POSITION 

If, in spite of the practical difficulties that may be connected there- 
with, we regard the nature of the buccal cavity and the structure of the 



346 bulletin: museum of comparative zoology 

lips with their associated sense organs as representing the characters of 
primary taxonomic significance, and we employ the key to the families 
of Spiruroidea which Chitwood and Wehr (1934, p. 312) have con- 
structed chiefly on the basis of cephalic structures, the species de- 
scribed above must be classified among the Spiruridae. On further in- 
vestigation it is found that the species shows closest affinity with 
either the Spirurinae (Railliet, 1916) or the Habronematinae Chitwood 
and Wehr, 1934, depending upon whether we are to interpret the 
narrow flange-like membranes that connect the lips in the dorsal and 
ventral fields as interlabia. 

With the exception of Hcdruris, whose relationship with the present 
form is evidently a distant one, there is no genus in the Spiruridae with 
representatives recorded in reptiles. Leaving out of consideration the 
difficult question of the interlabia, I am inclined to attribute our 
species to the Spirurinae chiefly on account of several similarities that 
it has with certain species of Spirura. Thus, the constriction of the 
oral rim is comparable with that of S. rytipleurites (cf. Chitwood and 
Wehr, 1934, p. 297) and S. michiganensis Sandground, 1934, and the 
structure of the female genitalia and male caudal extremity is very 
similar to that described for S. talpae by Stefanski (1934). 

Our species must, however, be generically differentiated from 
Spirura on account of the absence of labial lobes, the presence of labial 
teeth and the shape of the buccal cavity. 



Paraspirura gen. no v. 

Spiruridae. Two small lateral lips each bearing two pairs of 
papillae submedially towards the base and a larger circular amphid in 
the median line. Minute apical tooth and dentigerous ridges on inner 
labial face; membranous flanges (? interlabia) between the lips. 
Mouth dorso-ventrally elongated. Buccal vestibule cylindrical with 
lightly cuticularized walls. Oesophagus plainly divided into muscular 
and glandular parts. Male with wide caudal alae and cuticular orna- 
mentations on ventral surface. Four pairs pedunculate and a single 
median preanal papillae; five pairs postanal papillae. Spicules unequal 
and dissimilar. Gubernaculum present. Vulva post-equatorial; 
Oviparous. 

Type species P. mabuyae in Lacertilia. 



sandground: African nematoda 



347 



Abreviata (Polydelphyoptera) poicilometra sp. nov. 

Host: Cercopithecus mitis kibonotensis Lonnberg. 
Locality: Ngatana, Kenya Colony. 

The material consists of four females and a single male, robust in 
form with slight attenuation towards the truncated head and a 
more sharply pointed tail. Delicate, transversely striated cuticle loosely 
wrinkled in cephalic region where it may be partially or wholly re- 
flected over the lips. Labial armature of denticles similar to that de- 
scribed by Ortlepp (1926) for Physaloptcra caucasica, consisting of a 




0. 2mm. 
Fig. 6. Abreviata (P.) poicilometra. Inner face of a lip. 



prominent triangular apical tooth with a single small membranous 
structure projecting from its internal base, two split or chela-like teeth 
on submedian edge, and several denticular ridges on the inner face 
visible only on separating the lips. (Fig. 6.) 



Measurements of A. poicilometra in millimeters. 

Male Female 

35 38-63. 

1.4 1.6 

0.58 0.7 

0.4 0.4 

4.4 6.0 

1.9 1.5 

10.3 
0.049 x 0.038 average 



Length 

Equatorial width 

Cervical papillae from antr. end 

Glandular oesophagus 

Muscular oesophagus 

Tail 

Vulva to antr. end. 

Eggs 



FEMALE GENITAL SYSTEM 

Vulva inconspicuous externally. There is a long, narrow vagina 
which gradually dilates into a spindle-shaped egg-chamber which, at 
its broadest point, measures 0.37 mm. The posterior end of the egg 
chamber again narrows to form the so-called common trunk, which 
receives the uteri. In three specimens that were dissected, the number 



:\\s 



bulletin: museum of comparative zoology 



of uteri as well as their mode of union with the common trunk showed 
the following variations: In the longest female (63 mm.) there were 
ten uteri arising by irregular dichotomy, six uterine branches uniting 
in pairs to form one arm of the common trunk, and four uterine 
branches pairing up to form the second arm. (Fig. 7B.) In the second 




Fig. 7. Abreviata (P.) poicilometra. Dissections of the genitalia of three 
females showing variation in the number and mode of union of the uteri with 
the common trunk. 



specimen (Fig. 7A) the genitalia consisted of 13 uteri, the common 
trunk first giving off an arm that divides into 3 branches and further 
back the other arm splits into three very short branches, of which the 
inner divides dichotomously into 4 uteri while the remaining two each 
give rise to three uteri. In the smallest example of the species (38 mm.) 
the individual uteri were intertwined in a manner which did not per- 
mit determination of their exact number. There were certainly more 
than 14 uterine tubes uniting by twos and threes to form seven arms 
which individually emptied into the base of the common trunk as 
shown in figure 7C. 



sandground: African nematoda 



349 



MALE GENITAL SYSTEM 

The "bursa" is lanceolate in shape, measuring 2.3 mm. in length 
and 1.2 mm. at its widest point. Its ventral surface shows the usual 
cuticular ornamentation. The long stalked lateral papillae are ar- 
ranged in pairs in front of and behind the anus. Three sessile papillae 




Fig. 8. Abreviata (P.) poicilometra. Caudal extremity of male from ventral 
aspect. 

arranged at the points of a triangle mesially in front of the anus. Two 
median, sessile pairs of papillae immediately post-anal. Four pairs of 
lateral papillae with short peduncles behind the anus in positions illus- 
trated in figure 8. Spicules unequal and dissimilar: left thin and flexible 
measures 2.8 mm. in length; the right is of more substantial structure 
but as it was partially extruded it was not in a favorable position for 
measurement; approximately its length is around 0.3 mm. 



SYSTEMATIC POSITION 

As it was defined and treated in the comprehensive revision by 
Ortlepp (1922), the genus Physaloptera Rudolphi was one of the largest 
and taxonomically one of the most unwieldy of the genera of parasitic 



350 bulletin: museum of comparative zoology 

nematodes. Many of the better known species in the genus are known 
to display great ranges of variation effecting many of the criteria that 
have been utilized for specific differentiation. Even though Ortlepp 
reduced many of the species to synonomy, there can be no doubt that 
many ill-differentiated species are still retained in the catalogue. A 
critical revision of the group, undertaken as soon as information on 
developmental morphology secured from cross-infection life-history 
experiments is available, will probably reveal that many species 
erected mainly on the basis of host occurrence should be eliminated. 
Although it was not undertaken with a view to eliminating the spuri- 
ous species that have accumulated, the most valuable of recent con- 
tributions to the taxonomy of the Physalopteridae is that of Schulz 
(1927) which, being published in Russian in a not often available 
"Festschrift," has not received the attention it merits. To subdivide 
and regroup the species of Physaloptera (s.l) Schuh has employed a 
so-called Principle of Taxonomic Co-efficients, a device which ap- 
parently loses nothing because of its purely pragmatic purpose. The 
genus is divided primarily on the dentation of the lips into three 
genera, each of which is secondarily separated into subgenera on the 
basis of the uterine branching in the female. 

Classified on the basis presented by Schulz, the species described in 
the present paper will be assigned to the genus Abreviata (Travassos) 
emend Schulz, which is characterized by the presence of an apical 
tooth, dorsal and ventral submedian teeth, and with or without denti- 
cular ridges on the internal face of the lips. It will further be placed in 
the subgenus Polydelphyoptera on account of its multiple uteri. 
Schulz assigned only one species, A.(P-)- capensis Ortlepp, 1922 from 
the South African rodent, Xerus capensis, to this subgenus. A ques- 
tionable second species of the subgenus is Physaloptera joy euxi Gendre, 
1928 from Phacochoerus africanus. Our species is also to be assigned 
here. The outstanding differential feature in the male is its sixth pair 
of post-anal papillae, other species usually carrying only five pairs. 
The most interesting feature of the species is the surprising variation 
displayed in the number and mode of origin of the uteri. In the first 
specimen dissected (Fig. 7 A) the uteri are identical with that described 
for A.(P). capensis; the uterine complex in the second and third of our 
dissected specimens appears to be unique among polydelphous forms 
thus far described among species of the group. 



sandground: African nematoda 



351 



STRONG YLOIDE A 

OSWALDOCRUZIA LOVERIDGEI Spec. IIOV. 

Host: Siaphos kilimensis Stejneger (Scincidae). 
Locality: Mt. Mbololo, Taita, Kenya. 

The range of measurements of the 8 male and 4 female worms, 
taken from the stomachs of 3 of 5 specimens of the host examined, are 
as follows: 



Length 

Maximum breadth 
Antr. end to nerve ring 
Antr. end to excret. pore 
Length of oesophagus 
Length of tail 
Length of spicules 
Head to vulva. 
Size of eggs in uteri 



Male 


Female 


3.5 -4.1 


5.9 -7.5 


0.08-0.09 


0.1 -0.15 


0.15-0.16 


0.19-0.21 


0.19-0.25 


0.26-0.28 


0.27-0.30 


0.36-0.38 




0.23-0.25 


0.19 






4.84-4.9 




76-84 fj. x 42-48 /* 



The specimens are well extended but the state of preservation of 
some specimens is not all that can be desired. Perhaps for this reason 
considerable differences are observed in the cuticular inflation of the 




Fig. 9. Oswaldocruzia loveridgei. Anterior extremity of a male with 
maximum development of cephalic inflation. 



cephalic region. In some worms this inflation is hardly perceptible 
while in others it is conspicuous and extends for 65 p. to 70 p. in both 
sexes. In several specimens the cuticle, including that of the cephalic 
inflation, is devoid of cross striations or other markings; in others a 
faint striation is to be seen. Occasionally the cephalic inflation is uni- 
formly marked with fine or coarse granulations (Fig. 9). Variations 



352 



bulletin: museum of comparative zoology 



in the appearance of this structure consequently appear to be of little 
taxonomic significance. 

Narrow lateral alae may be found in the anterior portions of the 
body in both sexes. Cervical papillae, and in the male, prebursal 
papillae were not observed. 




Fig. 10. Oswaldocruzia loveridgei. Lateral view of caudal region of male. 

As illustrated in Fig. 10, the arrangement and form of the bursal 
rays are characteristic of the genus, but the terminal digitations of 
the dorsal ray, identical in all specimens examined, appears to be dis- 
tinctive of the species. The golden yellow spicules are about 15/u. wide 
in the middle and are split distally into 4 splinter-like processes which 
spread out like the ribs of an open fan (Fig. 13) when the spicule is 
partially extruded from the cloaca. 

No special peculiarities were found in the females. 



Although the records do not indicate that members of this genus dis- 
play any high degree of host specificity, few of the fourteen or more 
presumably distinguishable species of Oswaldocruzia have been de- 
scribed from reptiles. Among these are 0. agamae Sandground, 1929, 
0. malayana Baylis 1933 and 0. brasiliensis Lent and Freitas, 1935. 
The present species seems to show the closest resemblance to 0. bra- 
siliensis, recently described from a Brazilian snake, Drymobius bifos- 
satus. This is especially true when size is considered, but since con- 



sandground: African nematoda 



353 



siderable differences in size may be expected in the same species 
associated with different hosts (Harwood, 1932) little stress is to be 
attributed to the fact that the species here described is among the 




Fig. 11. Os- 
waldocruzia lov- 
eridgei. Termi- 
nal digitation 
of dorsal ray; 
highly magni- 
fied. 



ss 




Fig. 13. Os- 

waldocruzia lov- 
eridgei. Extrem- 
ity of exserted 
spicule; highly 
magnified. 



Fig. 12. Oswaldocruzia loveridgei. Spicule 
in its sheath drawn from dorsal aspect. 

smallest thus far recorded in the genus. The view that the species is 
new is based more on a consideration of the form of the spicules and 
the terminal branching of the dorsal ray. 



OXYUROIDEA 

Pharyngodon mabuyae sp. nov. 

Host: Mabuya varia varia. 

Location: rectum. 

Locality: Mt. Elgon, Uganda. 

This species, comprising some thirty male and female specimens, 
was found in two of three examinations of the host. As the following 



354 



bulletin: museum of comparative zoology 



table of measurements will show, the species is somewhat larger than 
those previously described in the genus. In all features, other than 
those mentioned below, our form closely resembles Pharyngodon 
laevicauda as described by Seurat (1914), but after having been com- 



Fig. 14. Pharyngodon 
mabuyae. Male viewed 
from ventral aspect. 




Fig. 15. Pharyngodon mabuyae. Caudal 
extremity of male from ventral aspect under 
higher magnification. 



pared with the descriptions of all the species recognized by Spaul 
(1926) and the four species that have subsequently been added to the 
genus, it is believed to represent a distinctly new species. 



Measurements in millimeters 

Female 
Length 6.6 -7.1 

Maximum width 0.465 

Length of oesophagus including bulb. 0.65-0.69 

Antr. end to nerve ring 0.46-0.5 



Male 

2.0 -2.3 
0.25 

0.43-0.46 
0.34 



sandground: African nematoda 355 

Measurements in millimeters 

Antr. end to excret. pore 

Antr. end to vulva 

Anus to postr. extremity 

Intervals between striae 

Eggs; length x breadth 170-178 ix x 45-50 n 

Lateral alae are to be found in both males and females arising from 
the mid-oesophageal region. In the female the alae are visible in the 
anterior part of the body as narrow bands, only slightly elevated 
above the cuticular surface. In the male the alae broaden gradually 



Female 


Male 


0.63-0.85 


0.63 


0.69-0.87 




0.65-0.71 


0.18-0.21 


lljU 


7m 




Fig. 16. Pharyngodon mabayae. Egg drawn to show bi-polar opercula. 

from before backwards, but they only become conspicuous in the 
posterior sixth of the body where they flare out into broad wings ex- 
tending to the level of the cloaca. There are the usual three pairs of 
papillae. The preanal pair are smallest; the paranal and basal-caudal 
pairs are larger. Neither from the ventral nor the lateral view of the 
body can one see any cuticular expansion that can be regarded as a 
caudal ala or bursa. 

Between the paranal papillae and projecting over the ventral border 
of the cloaca there arises a bluntly conical elevation which is flanked on 
either side by a serrated cuticular ridge or row of irregular cuticular 
spines. This structure is not described for other species of Pharyngo- 
don but it is found in a number of species of the related genus The- 
landros. 

A slender, poorly chitinized spicule, measuring about 85 to 90 
microns in length, is apparently present. However, it is so poorly 
definable that one cannot be sure of its existence. 

For a discussion of the systematic relations of P. mabuyae see below. 



356 



bulletin: museum of comparative zoology 



Thelandros seurati sp. nov. 

Host: Acontias percivali Loveridge (Scincidae). 

Location: Rectum. 

Locality: Taita, Kenya Colony. 

About a dozen specimens, of which only two were males, were found 
in the rectum of the above named host on two occasions. The females 
are all fully mature and contain several hundreds of eggs. The chief 
measurements of the species are as follows: 





Female 


Male 


Length 


4.4 -4.75 


2.8 -2.85 


Maximum width (rather flattened) 


0.60 




0.35 


Nerve ring to antr. end. 


0.18 




0.17 


Excret. pore to antr. end. 


'i 


► 


1.1 


Length of oesoph. including bulb 


1.2 - 


1.27 


0.75-0.80 


Length of tail 


0.25- 


-3.20 


0.10 


Length of spicule 






0.12 


Antr. end to vulva 


1.95- 


-2.5 




Eggs (length x width) 


0.076 x 0.040 





The worms are usually well preserved, turgid forms with conspicuous 
annulations which in the oesophageal region are about 25 /x apart in 
the males and 30ju apart in females. Head set off from body by a well 
defined constriction. The three lips are plainly bilobed. Amphids are 
represented by fine tubules penetrating the latero-ventral lips towards 
the dorsal angle. Cephalic papillae are too small to be accurately de- 
scribed. There is a distinct buccal cavity in the form of a shallow 
saucer-like depression with a conical base. The broad lateral fields are 
composed of a small number of relatively large quadrate cells whose 
spherical nuclei are about 25/* in diameter. Lateral alae are not seen 
in the females but in the male (Fig. 17) the lateral alae flare out as 
conspicuous vanes which commence at a point, 0.31 mm. from the 
cloaca, to attain a width of 0.06 mm. and terminate in a broad curve 
at the point where the tail springs from the body. The tail of the male 
has the shape of an elongate cone, that is constricted and bent in the 
middle and is displaced dorsally by the development of a cushion-like 
cuticular prominence capped by the cloaca. Two pairs of slightly 
elevated papillae adorn this prominence; one pair on its anterior border 
and the other at the side, on a level with the cloacal aperture. The 
posterior lip of the cloaca is guarded by a conspicuous projection that 
resembles a large median papilla in appearance, but its tip bears a 



sandground: African nematoda 



357 



series of exceedingly minute bosses or blunt spines, making it im- 
probable tbat the structure is actually a papilla. It is, perhaps, 
homologous with the structure that has been interpreted as a guber- 
naculum in Thelandros sexlabiate Ortlepp, 1933, and as a median post- 
cloacal papilla in Parapharyngodon maplestoni Chatterji, 1933. In our 



f 



Fig. 17. Thelandros seurati. Male in ventral view. 



specimens there is no trace of the cuticular ctenoid structure that 
adorns the superior border of the cloaca of several .species of Thelandros 
and which has been described in Pharyngodon mabuyac. The third pair 
of caudal papillae is the only one of the three pairs that could be de- 
scribed as pedunculate. It occupies the usual position in the proximal 
third of the caudal appendage. The spicule is a delicate structure, at 
least 0.12 mm. in length and about 7/i broad. It is poorly chitinized 



358 



bulletin: museum of comparative zoology 



and was only detected in one specimen where the rounded distal ex- 
tremity of the spicule was found protruding from the cloaca. 

Most of our female specimens exhibit the phenomenon referred to by 
Suerat (1914) as "endotokie matricide" wherein the eggs accumulate 





Fig. 18. Thelandros 
seurali. Caudal extrem- 
ity of male from lateral 
view. 



Fig. 19. Thelandros 
seurati. Caudal extrem- 
ity of male from ventral 
view. 



in the uteri in such numbers that other organs seem to degenerate and 
the body becomes little more than an egg sac. The excretory vesical 

and its pore can no longer be made out, and 

the position of the non-salient 

vulva, near the middle of the 

body, can only be detected in a 

few favorable specimens. There 

is a glandular vagina, about 0.12 

mm. in length followed by a 

more muscular walled part, the 

ovejector, of about three times 

this length. The coils of the 

uteri extend into the turgid 

post-anal region of the body; 

anteriorly the two ovaries are 
coiled around the prebulbar stem of the oesphagus. The eggs (Fig. 21) 
are ellipsoidal in shape, with the thick outer shell finely striated and 
provided with an operculum at one pole. 





003, 



Fig. 20. The- 
landros seurati. 
Caudal extremity 
of female. 



Fig. 21. The- 
landros seurati. 
Egg with opercu- 
lum at one pole. 



sandground: African nematoda 359 

SYSTEMATIC POSITION OF Pharyngodon mabuyae and 

Thelandros seurati 

The oxyurid parasites of reptiles are a primitive, or at least very 
specialized, group whose numerous representatives, now separated into 
about eight genera, show many features in common. Compact as the 
group is, none of the genera appear to be so closely related as are 
Pharyngodon and Thelandros, which are always placed in juxtaposi- 
tion in comprehensive analytical surveys of the group. That real diffi- 
culties are experienced in keeping their constituent species apart is 
shown by the fact that, when not defined in exactly the same terms, 
different authors stress different morphological features in diagnosing 
the genera and separate them by different key characters. Perhaps the 
most elaborate definitions of these genera are those presented by 
Seurat (1917) who in a key separates the genera by the following 
characters : 

(a). Vulva situated immediately behind excretory pore; no vagina; 

eggs very large; tail of male greatly prolonged. . . . Pharyngodon 

(b). Vulva considerably behind excretory pore; distinct vagina 

present; tail of male in dorsal position Thelandros 

Seurat envisages Thelandros being derived from Pharyngodon by the 
loss of caudal alae and the concomitant diminution of the peduncle of 
the first pair of post-anal papillae and of the tail ("point caudale dor- 
sale") in the male. These features, together with the differentiation of 
a vagina, are to Seurat indicative of the genus Thelandros being more 
highly evolved than Pharyndogon. 

While the retention of these two genera could be advocated on the 
score of convenience, our investigation into the matter of the specific 
relationships of the two species described in this paper indicates that 
the various described species of Thelandros and Pharyngodon present 
a mosaic of characters that cannot be satisfactorily separated into ele- 
ments sufficiently constant in their association to satisfy the require- 
ments of separate genera. The purposes of taxonomy could probably 
be better served by treating the group as a single genus, Pharyngodon, 
constituted by two sub-genera: 

Pharyngodon Pharyngodon (type; P.P. spinicauda (Dujardin)) and 

Pharyngodon Thelandros (type; P.T. alatus (Wedl)). 

The adoption of this procedure would probably check the creation 

of new monotypic genera that there is a tendency to propose for slightly 

aberrant species which cannot be fitted into either genus as diagnosed 

at present. 



360 bulletin: museum of comparative zoology 

In P. spinicauda, the type species of Pharyngodon, as well as in 
several other species of the genus, we may recognize the following 
features as representing the most distinctive generic characters: the 
presence in the male of a fairly long subulate tail with relatively broad 
cuticular flanges or caudal alae arising just in front of the cloaca and 
usually extending far enough backwards to be supported by the post- 
anal pair of papillae. In the female, the tail is also long and sometimes 
spinose (P. extenuatus Rud.) but the most constant features are: (1) 
the conspicuous excretory pore, with the vulva contiguous with it and 
situated anteriorly in the region of the oesophageal bulb; (2) the eggs 
are large, elongated and provided with opercula at each pole. 

Contrasting with this, in Theldndros alatus, the type, and in many if 
not all of the eleven species 1 which have been described in this genus, 
we find that there are no caudal alae in the male and that the tail 
proper (point caudale of Seurat) usually takes the form of a short, 
often spike-like, process that has been shifted dorsally by the massive 
development of a cuticular protuberance on which the cloacal aperture 
is situated. In the female the excretory pore is not so conspicuous, 
and the vulva, though slightly variable in position, is usually situated 
towards the middle of the body. 

From the descriptions of the thirteen species assigned to Pharyngo- 
don, little variation is found affecting the above mentioned generic 
characters of the female. However, there is a group of species, includ- 
ing P. laevicauda (Seurat), P. extenuatus (Rud.) P. teetipenis Gedoelst 
and P. tarentolae Spaul, in which the caudal alae do not extend suffi- 
ciently posteriad to embrace the last pair of papillae, and according to 
Spaul 's key to the genus Pharyngodon, caudal alae are absent in the 
species megalocerca (Skrjabin). This feature is not mentioned in 
Skrjabin's description though reference to his illustrations supports 
Spaul's inference. 

It therefore appears that among the species of Pharyngodon we can 
follow the gradual reduction of the caudal alae and we may interpret 
it as an evolutionary tendency towards the condition found in the 
various species of Thelandros. The special characteristics of the 
female, more particularly the form of the eggs, are more stable and 
consequently of greater taxonomic significance. It is mainly for this 
reason that we have assigned our parasite from Mabuya varia to the 
genus Pharyngodon; otherwise it could be as well accommodated in 

1 I have been unable to confirm these features in T. oswaldocruzi Travassos, 1925 and T. 
micruris Patwardhan, 1935, descriptions of which are not available to me,. From Seurat's 
description it appears that post-cloacal alae are a differential feature of T. numidicus. 



sandground: African nematoda 361 

Thelandros. P. mabuyae, is evidently closely related to Skrjabin's 
megalocerca, which also was described from an East African represen- 
tative of the Geckonidae. From this species, however, it may be dis- 
tinguished by its general size, the length of the tail in the male and 
the shape of the lateral alae, etc. 

Our species seurati, from the skink, Acontias percivali, is more typi- 
cally a member of Thelandros. In the absence "of the denticulate, 
cuticular processes on the anterior border of the cloaca, seurati ap- 
pears to be very closely allied to T. bulbosus (v. Linstow). It is also close 
to maplestoni, a species for which Chatterji (1933) created the new, 
and to my mind, unnecessary, genus Parapharyngodon . From both of 
these species it may be distinguished by the size of its smaller eggs and 
relatively longer oesophagus as well as on other dimensions. 



Strongyluris brevicaudata Miiller, 1894 

Host: Heliosciurus rufobrachiatus nyansae (Neumann). 
Locality: Sipi, Mt. Elgon, Uganda. 

In all structural details our specimens, 15 females and 5 males, were 
indistinguishable from worms in our collection bearing the identifica- 
tion S. brevicaudata that were collected in East Africa from species of 
Agama and Chameleon. The head, with its three prominent lips is 
set off by a constriction from the main body and presents the same 
arrangement of cephalic papillae and amphids as is illustrated by Hsu 
(1932) for S. brevicaudata and by Harwood (1935) for S. omata. The 
tapering flange of cuticle that projects anteriorly from the inner side 
of each lip is more prominent in some individuals than in others but 
this is apparently dependent upon the state of functional contraction 
at the time of fixation. The cuticle is delicately marked by fine stria- 
tions, upon which is superimposed a coarser annulation or wrinkling. 
Commencing just behind the post-cephalic constriction, but not con- 
spicuously forming a cervical ring, and arranged in a double row es- 
pecially restricted to the dorsal and ventral cuticular fields, is a series 
of small but prominent papillae with narrow peduncles. These somatic 
papillae are present in both sexes, and extend at fairly regular inter- 
vals to within a short distance of the anus. In the female, the body 
tapers rapidly behind the anus to form a conical tail that is provided 
with a pair of prominent caudal papillae and a spike-like terminal 
appendage. In the male, the posterior end is very abruptly truncated 
and, obscured from view by the narrow bursa, the sucker, anus and 



362 



bulletin: museum of comparative zoology 



some of the caudal papillae are difficult to delineate. These organs are 
best seen by flattening the worm under a cover-glass or examining the 
cut-off tail in the end-on position. The presence of the usual ten pairs 
of caudal papillae can then be detected (Fig. 22). Of the three pairs of 




Fig. 22. Strongyluris brevicaudata from Sciurus rufobrachiatus. Caudal ex- 
tremity of male; ventral view. 



parasuctorial papillae, the first are small and narrow compared with 
the large, fusiform second and third pairs. The three terminal pairs of 
papillae are closely set, the largest pair tending to hide one or other 
of the remaining two pairs so that it may easily be overlooked. 

The chief measurements (in millimeters) of the worms are given in 
the following table: 







Male 


Female 


Length 




8.9 -9.8 


9.4 -12.6 


Breadth at equator 




0.65-0.74 


0.7 -0.92 


Breadth at cervical constriction 


0.77 


0.92 


Excret. pore from antr. 


end. 


1.46-1.56 


1.77-1.84 


Length of pharynx 




0.25-0.28 


0.27-0.29 


Length of eosophagus including bulb 


1.8 -2.0 


2.1 -2.2 


Anus to postr. end. 




0.084-0.11 


0.2 -0.23 


Caudal spine 




0.040 


0.056-.075 


Diameter of sucker 




0.11-0.13 




Sucker to anus 




0.08-0.09 




Spicules (Equal) 




1.3 -1.52 




Vulva to postr. end 






3.8 -4.7 


Eggs 






0.072-075 x 0.040-.042 



sandground: African nematoda 363 

DISCUSSION 

Prior to the description of Strongyluris paradoxus Sandground, 1933, 
from a glossy ibis, Hagedashia h. niloticus from Tanganyika, some ten 
species of Strongyluris had been described. All had been found in 
lacertilian hosts and consequently the writer held it necessary to con- 
sider the possibility of there having been a mix-up of host labels or 
alternatively that the finding represented an instance of spurious 
parasitism such as may follow the preying of one animal upon another. 
However, the evidence surrounding the finding offered no support for 
such suspicions. 

In the present paper we actually extend the host range of the genus 
Strongyluris further. No doubt can here be entertained concerning 
the authenticity of the host since the writer himself collected the 
worms from the caecum of the squirrel whose viscera were preserved 
in the field and brought back in a separate jar. Two additional speci- 
mens (cf & 9 ) of a closely related if not identical species of this genus 
bearing the label "Rat gris, Lac Albert, Congo Beige; Leg. Dr. For- 
nara," were later given to me by Dr. L. van den Berghe of the Prince 
Leopold Institute of Tropical Medicine, Antwerp. Hence, there can 
be no doubt of the genus Strongyluris being found in certain African 
rodents. Concerning the specific identification of the material some 
question may be raised. On first seeing the worms our first idea was 
that we were confronted with a new species, but, as subsequent study 
showed, it is difficult to support this contention. S. brevicaudata, the 
type of the genus, was first recorded from the West African Agama 
agama agama ( — A. colonorum) and has since been reported from other 
species of East African Agama. Of the thirteen species that are now 
listed in the genus only two, namely S. gigas Spaul and S. loveridgei 
Spaul may be recognized by their significantly larger size. In the ab- 
sence of other specific criteria, the great variation in length and the 
proportions of various parts of the body makes the identification of 
most of the remaining species rather uncertain. Baylis and Daubney 
(1922) and Spaul (1923) have questioned the validity of several species, 
while Taylor (1925) contended that S. ornata Gendre,. and S. strep- 
toesophagus Connal and probably also S. chamaeleonis Baylis and 
Daubney and S. calotis Baylis and Daubney are indistinguishable from 
the genotype. Recently, however, Harwood (1935) after a scrutiny of 
the literature and the examination of some of the material in our 
museum's helminthological collection, has constructed a key to the 
genus wherein he tentatively reestablishes the validity of all the species 



364 bulletin: museum of comparative zoology 

whose standing has come into disrepute. Whether the characters 
utilized by Harwood are specifically significant need not concern us 
in considering the status of the species here involved, for not only do 
its structural features and dimensions coincide for those described 
for S. brevicaudata but prolonged comparison of the specimens with 
others from various species of Agama has failed to reveal a single 
distinguishing feature. Though one might expect that a lacertilian 
parasite could not become adapted to a mammal without some ac- 
companying somatic changes, these presumptive morphological differ- 
ences are so elusive that an identification of S. brevicaudata is the only 
one warranted. 



sandground: African nematoda 365 



BIBLIOGRAPHY 

Baylis, H. A. and Daubney, R. 

1922. Report on the parasitic nematodes in the collection of the Zoo- 
logical Survey of India. Mem. Indian Mus. 7, p. 263. 

Baylis, H. A. 

1933. On a collection of nematodes from Malayan reptiles. Ann. & 
Mag. Nat. Hist. ser. 10. 11, p. 615. 

Chatterji, R. C. 

1933. On a new nematode, Parapharyngodon maplestoni, from a Burmese 
lizard. Ann. Trop. Med. and Par. 27, p. 131. 

Chitwood, B. G. and Wehr, E. E. 

1934. The value of cephalic structures as characters in nematode classi- 
fication, with special reference to the superfamily Spiruroidea. 
Ztschr. f. Parasitenk. 7, p. 273. 

Gedoelst, L. 

1919. Une espece nouvelle de Pharyngodon. C. R. Soc. Biol. 82, p. 869. 

Harwood, P. D. 

1932. The helminths parasitic in the Amphibia and Reptilia of Houston, 
Texas and vicinity. Proc. U. S. Nat. Mus. 81, Art. 17, pp. 1-71. 

Harwood, P. D. 

1935. Two new species of Strongyluris (Nematoda) and notes on the 
genus. Jl. Tennessee Acad. Sci. 10, p. 131. 

Lent, H. and Freitas, J. F. T. de 

1935. Sob re uma nova especie do genero Oswaldocruzia Travassos, 
1917. Mem. Inst. Osw. Cruz. 30, fasc. 3, p. 379. 

Ortlepp, R. J. 

1922. The nematode genus Physaloptera Rud. Proc. Zool. Soc. London. 

Dec. 1922, p. 999. 
1926. On the identity of Physaloptera caucasica v. Linstow, 1902 and 
Physaloptera mordens Leiper, 1908. Jl. Helminthology 4, p. 199. 

Ortlepp, R. J. , 

1933. On some South African Reptilian Oxyurids. Onderstepoort Jl. 
Vet. Sci. & Animal Ind. 1, p. 99. 

Sandgrotjnd, J. H. 

1933. Parasitic nematodes from East Africa and Southern Rhodesia. 
Bull. Mus. Comp. Zool. Harvard. 75, No. 2, p. 274. 



366 bulletin: museum of comparative zoology 

Schulz, R. E. 

1927. Die Familie Physalopteridae Leiper, 1908, und die Prinzipien ihrer 
Klassifikation. (Russian with German summary.) Samml. 
Helminth. Arb. Prof. K. I. Skrjabin gewidmet. Moscau. p. 221. 

Seurat, L. G. 

1914. Sur un cas d'endotokie matricide chez un Oxyure. C. R. Soc. 

Biol. 76, p. 850. 
1914a. Sur un nouvel oxyure des reptiles. Ibid. 77, p. 96. 
1917. Sur les Oxyures des Sauriens du Nord-Africain. Arch. Zool. 

Exp. et. Gen. 56, p. 401. 

Spaul, E. A. 

1923. Nematodes of the genus Strongyluris from Agama. Ann. & Mag. 
Nat. Hist. ser. 9, 12, p. 149. 

1926. On a new species of the nematode genus Pharyngodon. Ibid. ser. 
9, 17, p. 585. 

Skrjabin, K. I. 

1916. Parasitic trematodes and nematodes collected by the expedition 
of Prof. V. Dogiel and I. Sokolov. in East Africa. 1, p. 99. 

Stefanski, W. 

1934. Sur le developpement et les characteres specifiques de Spirura 
rytipleurites (Deslongchamp). Ann. de Parasitologic 12, p. 203. 

Taylor, E. L. 

1924. Notes on some nematodes in the Museum of the Liverpool School 
of Tropical Medicine. Ann. Trop. Med. & Parasit. 18, p. 608. 



ri-v- 



Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 

Vol. LXXIX, No. 7 



SCIENTIFIC RESULTS OF AN EXPEDITION 
TO RAIN FOREST REGIONS IN EASTERN AFRICA 

VII 
AMPHIBIANS 



By Arthur Loveridge 



With Three Plates 



CAMBRIDGE, MASS., U.S.A. 

PRINTEDFORTHEMUSEUM 

December, 1936 



No. 7. — Reports on the Scientific Results of an Expedition to Rain Forest 

Regions in Eastern Africa 

VII 

Amphibians 

By Arthur Loveridge 

CONTENTS 

Introduction Page 

Material 369 

Acknowledgments 370 

Summary of novelties and changes 370 

List of species collected 373 

Systematic Discussion 

Caecilians 375 

Toads 380 

Tree and sedge frogs 383 

Terrestrial frogs 408 

Bibliography 429 



INTRODUCTION 

The collection on which this report is based was made by the author, 
as a Fellow of the John Simon Guggenheim Foundation, with a view 
to elucidating the present-day distribution of the montane, sylvicoline 
fauna of certain mountains in eastern Uganda and Kenya Colony. 

This subject will be dealt with in the concluding contribution to 
the series of reports, in a paper which will also contain the itinerary 
and full information regarding localities and altitudes. 

MATERIAL 

The period of collecting was from November 8, 1933, to July 9, 
1934, during which time, 2,528 amphibians, representing 46 species, 



370 bulletin: museum of comparative zoology 

were preserved. A dozen additional species donated by friends, though 
dealt with in this paper, are not included in the total, of which 2 species 
of caecilian and 8 of toads and frogs are new to the collections of the 
Museum of Comparative Zoology. 

One might single out for special mention, topotypes of Dermophis 
gregorii, Boulengerula denhardti and BdeUophis unicolor. Each of 
these has been known from the holotype only, the two latter are now 
relegated to the synonymy of the former. Also a good series of the 
rare Bnfo steindachnerii and topotypes of Leptopelis concolor and Rana 
wittei both known only from their respective holotypes. 



ACKNOWLEDGEMENTS 

The material generously donated by Messrs R. E. Moreau, C. A. 
du Toit, and H. J. Allen Turner is distinguished in the following pages 
by the initials of the collector. The Museum is deeply grateful to 
these gentlemen for their welcome gifts. 

I am indebted to Herr. P. de Grijs of the Zoological Museum at 
Hamburg for allowing me to examine certain of the Pfeffer types 
carefully preserved in that institution. My thanks also go to Mr. 
H. W. Parker of the British Museum for answering questions relative 
to specimens in his charge. 

Mr. R. D. Milne of Witu, took the delightful photograph of frogs' 
nests which is reproduced on plate 2, fig. 1. I am indebted to my 
wife for the other illustrations accompanying this paper. They are 
from photographs taken with the small Rolleiflex camera. 



SUMMARY OF NOVELTIES AND CHANGES 

The only novelties in the collection were: 

Boulengerula taitanus, Mount Mbololo, Taita Mountains, Kenya Colony. 
Hyperolius milnei, Witu, and other coastal localities in Kenya. 
Hyperolius ahli sp. n., Lake Peccatoni, Coastal Province, Kenya Colony. 

In addition to these new forms and Arthroleptides dutoiti Loveridge, 
collected by Mr. C. A. du Toit at Koitobos River, Mount Elgon, on 
the Kenya side, the undermentioned species are recorded from Uganda 
or Kenya for the first time. 



loveridge: African amphibians 371 

New for Uganda 

Rana fuscigula chapini Noble of the Belgian Congo. 

Rana oxyrhynchus gribinguiensis Angel of the French Congo. 

Rana ansorgii Boulenger of Angola. 

Phrynobatrachus graueri (Nieden) of Belgian Ruanda-Urundi. 

New for Kenya Colony 

Xenopus laevis victorianus Ahl of Tanganyika Territory. 

Bufo steindachnerii Pfeffer of Tanganyika Territory. 

Leptopelis johnstoni (Boulenger) of Nyasaland. 

Hylambates maculatus Dumeril of Zanzibar. 

Megalixalus brachynemis Boulenger of Nyasaland. 

Hyperolius rossii (Calabresi) of the Belgian Congo. 

Hyperolius undulatus (Boulenger) of the Belgian Congo. 

Hyperolius ? vermicularis Ahl of Zanzibar. 

Hyperolius sansibaricus (Pfeffer) of Zanzibar. 

Hyperolius parkeri Loveridge of Tanganyika Territory. 

Rana edulis (Peters) of Mozambique. 

Rana fuscigula chapini Noble of the Belgian Congo. 

Rana fuscigula angolensis Bocage of Angola. 

Rana oxyrhynchus gribinguiensis Angel of the French Congo. 

Rana mascareniensis venusta Werner of Uganda, etc. 

Arthroleptis adolfifriederici Nieden of Belgian Ruanda-Urundi. 

All these species are known from Tanganyika Territory, and several 
have been recorded from Kenya though under other names, as for 
example Rana edulis, hitherto confused with the South African aolspersa. 

The genera Hylambates sensu strictu and Arthroleptides have their 
range extended northwards into Kenya. 

The undermentioned are considered to be synonyms: 

Boulengerula denhardti Nieden = Dermophis gregorii Boulenger 

Bdellophis unicolor Boettger = Dermophis gregorii Boulenger 

Cassina angeli Witte = Kassina senegalensis Dumeril & Bibron 

Hyperolius brevipalmatus Ahl = Megalixalus fulvovittatus (Cope) 

1 Hyperolius pygmaeus Ahl = Megalixalus brachynemis Boulenger 

1 As explained in the text, I synonymized this with M. fornasini through a confusion of 
material. 



372 



bulletin: museum of comparative zoology 



Hyperolius phrynoderma Ahl 
Hyperolius mohasicus Ahl 
Hyperolius oculatus AM 
IHyperolius irregularis Ahl 
Rappia symetrica Mocquard 
Rappia platyrhinus Proctor 
Hyperolius asper Ahl 
Hyperolius nyassae Ahl 
^Hyperolius pulchromarmoratus Ahl 
Hyperolius scheffleri Ahl 
Hyperolius melanopthalmus Ahl 
Hyperolius renschi Ahl 
Hyperolius rubripes Ahl 
Hyperolius noblei Ahl 
Hyperolius callichromus Ahl (part) 
Hyperolius flavoviridis Peters 
Hyperolius tettensis Peters 
Rappia platycephala Pfeffer 
Rappia granulata Boulenger 
Hyperolius microps Giinther 
Hyperolius usaramoae Loveridge 
Hyperolius translucens Power 
Phrynopsis boulengeri Pfeffer ■ 

Pyxicephalus flavigula Calabresi 
Phrynopsis usambarae Ahl 
Rana aberdariensis Angel 
Rana erlangeri Ahl 
Arthroleptis stenodactylus 
uluguruensis Loveridge 



= Hyperolius rossii (Calabresi) 

= Hyperolius rossi (Calabresi) 

= Hyperolius rossii (Calabresi) 

= Hyperolius rossii (Calabresi) 

= Hyperolius undulatus (Boulenger) 

= Hyperolius undulatus (Boulenger) 

= Hyperolius undulatus (Boulenger) 

= Hyperolius undulatus (Boulenger) 

= Hyperolius striolatus Peters 

= Hyperolius striolatus Peters 

= Hyperolius mariae Barbour & Loveridge 

= Hyperolius mariae Barbour & Loveridge 

= Hyperolius sansibaricus (Pfeffer) 

= Hyperolius puncticulatus (Pfeffer) 

= Hyperolius argentoviltis Ahl 

= Hyperolius argus Peters 

= Hyperolius argus Peters 

= Hyperolius argus Peters 

= Hyperolius nasutus Giinther 

= Hyperolius pusillus (Cope) 

= Hyperolius pusillus (Cope) 

= Hyperolius pusillus (Cope) 

= Rana edulis (Peters) 

= Rana edulis (Peters) 

= Rana edulis (Peters) 

= Rana wittei (Angel) 

= Rana floweri Boulenger 



= ^4.s. lonnbergi Nieden 



loveridge: African amphibians 373 



LIST OF SPECIES COLLECTED 1 

CAECILIIDAE Page 

Dermophis gregorii Boulenger 375 

Boulengcrula taitanus Loveridge 378 

Boulenger ula changamwensis Loveridge 378 

PIPIDAE 

Xenopus laevis victorianus Ahl 379 

(Xenopus muelleri (Peters)) 379 

BUFONIDAE 

Bufo regular is regularis Reuss 380 

Bufo stcindachnerii Pfeffer 381 

(Bufo lonnbergi nairobiensis Loveridge) 383 

POLYPEDATIDAE 

Chiromantis xerampelina Peters 383 

Leptopelis concolor Ahl 385 

Leptopelis johnstoni (Boulenger) 387 

Hylambates maculatus Dumeril 387 

Kassina senegalensis (Dumeril & Bibron) 388 

Megalixalus fornasinii (Bianconi) 390 

Megalixalus fulwvittatus (Cope) 391 

Megalixalus brachynemis Boulenger 391 

Hyperolius rossii (Calabresi) 393 

Hyperolius picturatus Peters 394 

(Hyperolius montanus (Angel)) 395 

Hyperolius undulatus (Boulenger) 395 

Hyperolius flavoguttatus Ahl 397 

Hyperolius striolatus Peters 397 

(Hyperolius mariae Barbour & Loveridge) 398 

(Hyperolius ? vermicularis Ahl) 399 

Hyperolius sansabaricus (Pfeffer) 400 

Hyperolius puncticulatus (Pfeffer) 401 

(Hyperolius argentovittis AH) 401 

1 Specie3 in parenthesis were collected and presented by others. 



374 bulletin: museum of comparative zoology 

POL YPEDA TIDAE continued Page 

Hyperolius ahli sp. nov 402 

Hyperolius parkeri Loveridge 404 

Hyperolius nasutus Gunther 405 

Hyperolius milnei Loveridge 406 

Hyperolius pusillus (Cope) 407 

RANIDAE 

Rana edulis (Peters) 408 

Rana occipitalis Gunther 409 

Rana fuscigula chapini Noble 409 

Rana fuscigula angolensis Bocage 410 

Rana wittei (Angel) 411 

Rana galamensis bravana (Peters) 413 

Rana floweri Boulenger 414 

Rana oxyrhynchus oxyrhynchus Smith 415 

Rana oxyrhynchus gribinguiensis Angel 416 

Rana mascareniensis mascareniensis Dumeril & Bibron . . .417 

(Rana mascareniensis uzungwensis Loveridge) 418 

Rana mascareniensis venusta Werner 419 

(Rana ansorgii Boulenger) 419 

(Arthroleptides dutoiti Loveridge) 420 

(Arthroleptis stcnodactylus lonnbergi Nieden) 420 

Arthroleptis adolfifriederici Nieden 420 

Arthroleptis minutus Boulenger 421 

Phyrnobatrachus keniensis Barbour & Loveridge 422 

Phrynobatrachus graueri (Neiden) 424 

Phrynobatrachus kinangopensis Angel 424 

Phrynobatrachus acridoides (Cope) 425 

Phrynobatrachus natalensis (Smith) 426 

Hemisus marmoratum marmoratum (Peters) 426 

BREVICIPITIDAE 

(Hoplophryne rogersi Barbour & Loveridge) 427 

Phrynomerus bifasciatus (Smith) 4-8 



loveridge: African amphibians 375 

CAECILIIDAE 

Dermophis gregorii Boulenger 

Dermophis gregorii Boulenger, 1894, Proc. Zool. Soc. London, p. 646, pi. xl, 

fig. 4: Ngatana, Tana River, Kenya Colony. 
Boulengerula denhardti Nieden, 1912, Sitz. Ges. naturf. Freunde Berlin, p. 199: 

Tana Region, Kenya Colony. 
Bdellophis unicolor Boettger, 1913, in Voeltzkow, Reise in Ostafrika, 3, p. 353, 

pi. xxiii, fig. 18: Peccatoni, Kenya Colony. 

130 (M.C.Z. 20101-50) Peccatoni, near Witu, K.C. 24-26.V.34. 
4 (M.C.Z. 20151-4) Mkonumbi, near Witu, K.C. 28.V.34. 

1 (M.C.Z. 20054) Mombosasa, near Witu, K.C. 31.V.34. 
48 (M.C.Z. 20055-92) Kau, n. bank Tana River, K.C. 4.vi.34. 

3 (M.C.Z. 20093-5) Laini, s. bank Tana River, K.C. 6.vi.34. 

2 (M.C.Z. 20096-7) Ngatana, n. b. Tana River, K.C. 20.vi.34. 

Native names. Nyoha mai (Kiamu, literally 'water snake') at 
Mkonumbi; sango (Kipokomo) at Kau; ntzango (Kipokomo) at 
Ngatana. 

Synonymy. The Ngatana specimens are topotypes of gregorii, the 
Kau and Laini of Bonlengerida denhardti, the Peccatoni of Bdellophis 
unicolor. Peccatoni is only fifty miles from Ngatana, the other local- 
ities lie between or a little to the south. 

These three species, gregorii, denhardti and unicolor, are only known 
from the types so these localities were visited with the object of 
obtaining adequate material for comparative studies. 

The coloring of the types was as follows : 

gregorii. "Dark brown above, paler beneath." 
denhardti. "Light yellowish brown." 
unicolor. "Uniform glossy black resulting in a 
lacquered appearance." 

The last description aptly described my material when freshly caught, 
now they are plumbeous, slightly darker above than below. I consider 
the first two descriptions were based on specimens in various stages 
of fading. 



376 bulletin: museum of comparative zoology 

The number of primary and secondary annuli were for 

gregorii 160 according to Boulenger. 
denhardti 165 according to Nieden. 
unicolor 145 according to Boettger. 

giving a range of 20 which is less variation than is known to occur 
in Scolecomorphus uluguruensis of which an adequate series was 
available. 

Generic status. I have compared our material with topotypes of the 
genotype (mexicanus) of Dermophis as well as with topotypes of the 
genotype (boulengeri) of Boulengerula and fail to see any external 
characters on which these genera could be separated. A careful exam- 
ination discloses what are apparently cycloid scales in the skin of 
Boulengerula (notwithstanding my key of 1930, p. 8) though Tornier 
made no mention of them and they are not so well developed as in 
gregorii. 

Dermophis gregorii differs from the four species of Boulengerula 
in its eye being distinct and the great number of secondary annuli 
on the posterior portion of the body. In these characters it agrees 
with mexicanus. There is a possibility that Dermophis itself is a 
synonym of Gymnophis, and as my friend Professor de Villiers is 
engaged on anatomical investigations as to the status of these genera, 
the present name may stand pending the publication of his results. 

Variation. Dunn (1928, p. 71) has emphasized the advisability of 
confining counts to the primary annuli owing to the varying degree 
of distinctness exhibited by the secondary rings. For comparative 
purposes I have treated the material by dividing it into two lots. 
The 53 Tana River caecilians have from 107 to 118 primary annuli, 
average 112.4, while 53 of the Lake Peccatoni series have from 111 
to 119 primary annuli with an average of 114.8 

Measurements. The 53 Tana River caecilians range from 147 to 
300 mm. (M.C.Z. 20055) in length with midbody diameters of from 
5 to 10 mm. giving a variation of diameter into length of 26.2 to 35.8 
with an average of 29.7. 

The 53 Lake Peccatoni specimens used range from 127 to 360 mm. 
(M.C.Z. 20114) in length with midbody diameters of from 4.5 to 
14.5 mm. giving a variation of diameter into length of 23.4 to 37.2 
with an average of 30.2. 

Breeding. Ova in those examined were minute, nevertheless the 
natives at Peccatoni stated that these caecilians lay great numbers 
of eggs during the rainy season, and that they lay them under water! 



loveridge: African amphibians 377 

Diet. Stomachs of Tana River specimens held termites, those of 
Peccatoni caecilians, earth worms. 

Habitat. In view of the rarity of this creature, of which only three 
specimens have been taken in forty years, the following notes on 
their habitat may be of interest. 

The 'big rains' had broken nearly a month before, and torrential 
downpours were occurring daily. On May 24, I captured the first 
through seeing its tail projecting from a small tussock of earth and 
grass which projected above the surface of a sheet of water in a flooded 
extension of Lake Peccatoni. 

This was shown to the natives in the village who stated that, 
though they dig up these caecilians in their gardens a quarter-of-a- 
mile from the shore of the lake, they are most abundant at the water's 
edge. During the course of the next two days, a score of men secured 
the series listed above, then told me that if I had only come in the 
dry season they would have procured a thousand with the same 
amount of exertion that they now dug up a hundred. Some said that 
in the dry season they could be found in writhing masses by digging 
in shallow water, from which it may be deduced that there is a tend- 
ency to concentrate in moist situations during the dry weather. One 
and all agreed that they live under water. I examined the spot from 
whence the series had been obtained and found that they had been 
dug from black mud, which had a slight admixture of sand, beneath 
the waters of the lake. 

The Kau series was obtained during the course of an afternoon by 
native children digging in the mud close to the village. The whole 
Tana delta in the vicinity of the village was waterlogged, and storm 
after storm of rain swept over the countryside until about two hours 
before sunset. 

Despite the willingness of the Wema villagers at Ngatana to bring 
in specimens, and urgent requests for caecilians, only two were pro- 
cured. One was taken in mud under sedges on the river bank, the 
other in dryish earth at the base of some bananas less than a hundred 
yards from the river, but a mile and a half below Wema. 

This creature, I found, was well-known to the natives at Golbanti. 
They said it occurred in the mud where they plant rice on the edge 
of an extensive swamp. I visited the place and found conditions closely 
paralleling those at Lake Peccatoni. One Mpokomo told me that 
very early in the month of June he had killed great numbers of caecil- 
ians while cutting his rice crop; he thought that they were a kind of 
snake. We spent a morning digging in the man's rice plot and at suit- 



378 bulletin: museum of comparative zoology 

able spots in the vicinity but without result. Though half the popu- 
lation of Golbanti was at work in these rice fields during the two 
days I camped in the village, not a caecilian was brought in. 

It will be noted that the time of their abundance at Golbanti 
coincides with that of Kau. It seems probable therefore, that these 
caecilians approach the surface during the heaviest part of the 'big 
rains' when the subsoil is flooded, and retire to greater depths as the 
rains slacken. However, on June 22-23, at Golbanti, water welled up 
and stood in many places when we dug more than eight inches below 
the surface. 

Boulengerula taitanus Loveridge 

Boulengerula taitanus Loveridge, 1935, Bull. Mus. Comp. Zool., 79, p. 16: 
Mount Mbololo, Taita Mountains, Kenya Colony. 

30 (M.C.Z. 20001-24) Mt. Mbololo, K.C. 14-24.iv.34. 

Native name. Murwe (Kitaita). 

Parasites. Nematodes (Oxyurus sp.) present in the intestines. 

Habitat. Taken in the rich leaf mould by digging beneath rotting 
tree trunks getween 4,000 and 4,800 feet. The stomach contents of 
those examined was so finely masticated as to be indeterminate. No 
signs of breeding were noted. 



Boulengerula changamwensis Loveridge 

Boulengerula changamensis Loveridge, 1932, Bull. Mus. Comp. Zool., 72, 
p. 381 : Changamwe, near Mombasa, Kenya Colony. 

3 (M.C.Z. 20098-100) Changamwe, K.C. 4.vii.34. 

Variation. Annuli 146-152. 

Measurements. Now the largest known example (M.C.Z. 20098) 
measures 235 mm. in total length; midbody diameter 5 mm.; the 
diameter being included in the length 5 times as against 4.5 times in 
the type. 

Breeding. No signs of it in either of two examined. 

Diet. Numerous termites in one of two stomachs examined. 

Habitat. These topotypes were also taken under a mango tree, 
possibly the same one from whence the type series came. Though 
on several successive days, energetic search was made in many situa- 
ations which appeared promising, no others were forthcoming. 



loveridge: African amphibians 379 

PIPIDAE 

Xenopus laevis victorianus Ahl 

Xenopus victorianus Ahl, 1924, Zool. Anz., 60, p. 270: Busisi, Lake Victoria, 
Tanganyika Territory. 

19 (M.C.Z. 20155-64) Kaimosi, K.C. 19.ii.34. 

Native name. Likele (Luragoli, but also applied to toads). 

Coloration in life. Above, uniformly black. Below, throat, chest 
and abdomen white, vermiculated with gray ; posterior portion of body 
and underside of limbs bright yellow thickly covered with large, 
black spots. Based on a breeding adult, immature examples more or 
less immaculate as stated in my (1933, p. 351) key. 

Measurements. A female measuring 70 mm. is the largest known 
example of this race. 

Breeding. Many pairs were taken in embrace, all females examined 
were distended with ova resulting in the greatest width of head being 
included twice in the width of the body. 

Habitat. Taken in a corner of the mill pond (pi. i, fig. 1) where 
covered by a dense matting of broken and rotting sedges. With care 
one could walk on this floating vegetation and by depressing it in 
a given spot my wife found that the Xenopus popped up. The series 
was obtained by a systematic application of this method; a retrial 
of it on successive days, however, failed to produce any more, 
so it is possible that they had assembled at this spot for breeding 
(many being in coitu), or alternatively that we had captured the 
majority. 

Diet. Large quantities of larval and freshly-emerged neuroptera, 
a water boatman and other hemipteron, limbs of a spider. 



Xenopus muelleri (Peters) 

Dadylelhra muelleri Peters, 1844, Monatsb. Akad. Wiss. Berlin, p. 37: Mozam- 
bique. 

2 (M.C.Z. 19865-6) Eldoret, K.C. (C.A. du T.) 25.L34. 

Distribution. Part of a series obtained on the Doinyo Lessos Estate, 
Uasin Gishu Plateau, by Mr. du Toit. 



380 bulletin: museum of comparative zoology 

BUFONIDAE 

BUFO REGULARIS REGULARIS ReUSS 
Bufo regularis Reuss, 1834, Mus. Senckenberg, 1, p. 60: Egypt. 

(M.C.Z. 20205-6) Mt. Kinangop, K.C. (H.J.A.T.) 1930. 



2 
1 
13 
7 
1 
4 
1 
1 
1 
1 
2 
1 
3 



M.C.Z. 20207) Mt. Debasien, U. 24.xi.33. 
M.C.Z. 20208) Sipi, Mt. Elgon, U. 12.xii.33. 
M.C.Z. 20209) Butandiga, Mt. Elgon, U. 8.i.34. 
M.C.Z. 20210) Elgonyi, Mt. Elgon, K.C. 31.1.34. 
M.C.Z. 20211) Kaimosi, K.C. 10.ii.34. 
M.C.Z. 20212) Kibwezi, K.C. 23.iii.34. 
M.C.Z. 20213) Voi, K.C. 10.iv.34. 
M.C.Z. 20214) Mt. Mbololo, K.C. 22.iv.34. 
M.C.Z. 20215) Lamu, Lamu Id., K.C. 7.V.34. 
M.C.Z. 20216) Ngatana, Tana R., K.C. ll.vi.34. 
M.C.Z. 20217) Karawa, K.C. 26.vi.34. 
M.C.Z. 20218) Malindi, K.C. 29.vi.34. 



Distribution. Also captured and released at Budadiri and Nairobi. 

Native names. Akidodok (Karamojong for all toads and frogs); 
likeru (Lugishu); likele (Luragoli, but also for Xenopus); lisheri 
(Lutereki); chula (Kitaita, but not specific); hovehove (Kipokomo, 
but not specific). 

Coloration in life. The Lamu specimens were a rather handsome 
olive color with cream -colored markings; it was their breeding season. 
The male from Ngatana had red on its flanks extending forwards to 
midbody, in addition to the more usual red on the hinder aspect of 
the thighs. 

Measurements. The largest cf (M.C.Z. 20218) measures 85 mm.; 
the largest 9 (M.C.Z. 20209), as well as Sipi females, also 85 mm. 

Breeding. On February 10, at Kaimosi, calling, pairing (pi. i, fig. 2), 
and spawning in shallow pools outside the forest. On May 7, at Lamu, 
they were to be heard calling everywhere, and many jumped into 
deep wells in the ecstasies induced by the first big rains. On June 29, 
at Malindi, a male was taken calling, in company with others, in a 
swamp. 

Parasites. A leech was attached to the rump of a breeding male at 
Kaimosi, on detaching itself it left a raw circular patch. Leeches were 
seen on others in this pool. 

Enemies. A young toad in the stomach of a Green Snake (Chlorophis 
hoplogaster) at Butandiga, an adult female in a Nosehorned Viper 
(Bitis nasicornis) at Kaimosi. 



loveridge: African amphibians 381 

Habitat. From 100 to 7,000 feet. The Debasien toad was ensconced 
beneath a log in very dry forest. The Elgonyi specimen was also 
taken in a clearing in dry forest but near a stream. 



Bufo steindachnerii Pfeffer 

Bufo Steindachnerii Pfeffer, 1893 (1892), Jahr. Hamburg Wiss. Anst. 10, p. 
103, pi. ii, fig. 8: Kihengo, Tanganyika Territory. 

1 (M.C.Z. 20165) Lamu, Lamu Id., K.C. 9.V.34. 

16 (M.C.Z. 20166-9) Mkonumbi, K.C. 22.V.34. 

17 (M.C.Z. 20170-4) Peccatoni, K.C. 24.V.34. 

6 (M.C.Z. 20175-9) Near Witu, K.C. 30.V.34. 

7 (M.C.Z. 20180-4) Laini, Tana R., K.C. 6.vi.34. 
31 (M.C.Z. 20185-9) Ngatana, K.C. 9.vi.34. 

89 (M.C.Z. 20190-9) Golbanti, K.C. 22.vi.34. 
1 (M.C.Z. 20200) Malindi, K.C. 28.vi.34. 

Distribution. The finding of this, hitherto the rarest of East African 
toads, in Kenya, not only adds a species to the fauna of that Colony 
but clears up an anomalous distribution. Since the taking of the 
type near Morogoro in central Tanganyika, the species has not been 
collected in that Territory, and there has been nothing to link that 
record with the one from between Badditu and Dime in Gallaland. 
It now appears that the species is a coastal form that happened to 
be described from the most southerly limit of its range. 

Variation. Superficially, particularly as regards color pattern, this 
species closely resembles a third-grown B. r. regularis. They may be 
distinguished by the key characters employed by Noble (1924, p. 167) 
with the exception that the first finger, though normally shorter 
than the second as in the right hand of M.C.Z. 20190, may be equal 
to it as in the left hand of the same specimen. The absence of a tarsal 
fold, more acuminate snout, greater wartiness, all assist in separating 
steindachnerii from regularis. 

Coloration in life, d" . Malindi. Above, yellow ochre with numerous 
markings in burnt umber, viz. a streak from the nostril downwards 
to the buccal border and posteriorly to the eye; a pair of] light-centered 
streaks from the upper eyelids directed inwards and backwards so 
as to almost form a V but interrupted on the vertebral line, this 
marking is followed by three or four similar pairs of blotches; dorsal 
tubercles ochraceous with umber spines; limbs regularly barred and 
blotched; on groin, hinder aspect of thigh, and sometimes on the 



382 bulletin: museum of comparative zoology 

tibia as well, are patches of rose-carmine or madder. Below, dirty 
white, region of gular sac gray; soles of fore and hind feet dark. 

Measurements. The largest cf (M.C.Z. 20175) measures 52 mm.; 
largest 9 9 (M.C.Z. 20190-1) measure 54 mm. 

Breeding. Males only were met with at Lamu and Mkonumbi, in 
the last locality they were calling vociferously as in all localities, 
including Kau, to Ngatana, that is to say from May 22 to June 20. 
They may have been calling at Golbanti and Malindi but no notes 
were made. The change in the proportion of the sexes was most 
instructive. At Peccatoni 7 males and 2 females, at Witu 5 males 
and 1 female, at Laini 1 male and 6 females, at Ngatana 5 males and 
26 females, at Golbanti 18 males and 71 females. The females were 
gravid in all localities. A puzzling fact, however, was the collecting 
of eight 18 mm. young at Peccatoni on May 24, though as the lake exists 
throughout the dry season, and the rains had already been coming 
down for a month, it seems possible that some toads had spawned 
at the very commencement of the first rains. 

Diet. Ants predominated in the first eight stomachs examined, 
these were of toads from all localities, except Lamu, to Ngatana, 
the only other creatures present were a carabid beetle, spider and 
woodlouse. A different picture was presented by the two Golbanti 
stomach contents, these consisted of (1) an ant, cicindelid beetle, 
beetle larva, 5 woodlice; (2) two beetle larvae, two millipedes, ten 
woodlice. 

Parasites. Nematodes present in a Golbanti toad were not preserved. 

Enemies. Recovered from the stomachs of Olive Water Snakes 
(Natrix o. olivacea) at Laini and Ngatana, also from those of a White- 
lipped Snake (Crotaphopeltis h. hotamboeia) and a Night Adder 
(Causus resimus) at Ngatana. 

Habitat. The call of this species, which attracted my attention to 
it, is quite distinct from that of regularis though with a certain simi- 
larity in a minor key. In the flooded areas where rains had formed 
lakes, the males sit calling on lily pads or weeds floating on the surface, 
they also hide among the roots of tussocks of grass which are just 
showing above the water. In the last situation they are exceptionally 
difficult to detect as their coloring blends and harmonizes with that 
of the peaty tussocks. However cautiously one approaches them, 
they usually cease calling before you can get within six feet. If you 
still persist in moving in their direction, they slip quietly into the 
water and dive to the bottom. 

The best method of capturing them under these conditions is to 



loveridge: African amphibians 383 

stalk them with the greatest caution, standing still whenever the 
toad stops calling — often involving a long and trying wait — moving 
on again only when he recommences, and, after locating him — no 
easy matter when he is resting on submerged vegetation with just 
the head showing above the surface — pounce upon him swiftly by 
hand. A net was found to be almost useless as it so frequently became 
fouled in unseen grass beneath the surface. This was the way in 
which we obtained the toads at Mkonumbi, Peccatoni and Witu. 
At Laini I captured them at night in the rice fields by the aid of a 
flashlight. 

At Ngatana and Golbanti, these toads were seen hopping about in 
swamps and marshy spots where they were easier to secure. I set 
the children to work at catching them in these two places, and the 
fact that they brought in the Golbanti series in one day shows that 
the species is by no means so scarce at the right season as its rarity 
in collections had led me to suppose was the case. I imagine that it 
burrows during the dry season. The young were taken on paths 
where they run, seldom hopping, like B. calamita which inhabits 
similar sandy country in Europe. 

Bufo lonnbergi nairobiensis Loveridge 

Bufo lonnbergi nairobiensis Loveridge, 1932, Occ. Papers Boston Soc. Nat. 
Hist., 8, p. 48: Nairobi, Kenya Colony. 

3 c? 1 9 (M.C.Z. 20201^) Mt. Kinangop, K.C. (H.J.A.T.) 1931. 

Variation. I should have expected these to have been referable to 
the typical form but in practically all measurements they answer to 
the key (loc. cit., p. 50) characters of the Nairobi race. Thus, third 
finger included 6.7 to 7.8 times in the length from snout to anus; 
fourth toe 4.6 to 4.8 times; tibia 3 to 3.2 times. 

Measurements. The largest cf (M.C.Z. 20202) only measures 28 
mm., 9 (M.C.Z. 20204), largest known, 39 mm. 

POLYPEDATIDAE 

Chiromantis xerampelina Peters 

Chiromantis xerampelina Peters, 1855, Arch. Naturg., 21, part 1, p. 56: Tete 
and Sena, Mozambique. 

Nest and 12 (M.C.Z. 20537-41) Witu, K.C. 30.v.— 2.vi.34. 
1 (M.C.Z. 20542) Sokoki Forest, K.C. (H.J.A.T.) vi.32. 



384 bulletin: museum of comparative zoology 

Variation. The tibiotarsal articulation of the adpressed hind limb 
marks the eye or just beyond. Males with dermal spinosities on 
dorsum. 

Measurements. The largest c? (M.C.Z. 20538) measures 56 mm., 
the only 9 (M.C.Z. 20539) 67 mm. 

Breeding. On May 30, two nests, containing small tadpoles, were 
found attached to the lower side of a branch of a tree standing near 
the centre of a large pool. The nests were about five feet above the 
surface of the water. 

On May 31, three very small nests were found attached to the 
underside of the branches of a large tree which was growing at the 
edge of another pond. The nests were far out over the water, how- 
ever, and at heights of from seven to eight feet above it. A fourth, 
and larger, nest was attached to a dead limb at a height of three feet 
above the water. A fifth nest was in long grass only six inches above 
the water. No frogs were seen. 

In a third pond near the forest, a small doom palm bore traces of 
half-a-dozen nests which may have been deposited a month ago, at 
the commencement of the rains. Six frogs were taken in this bush-like 
palm. Their capture was a comparatively simple matter for they 
clung to the sloping, spiny fronds, sidled round them as one ap- 
proached, but did not leap away until the last moment. If their 
jump landed them in the pool, as was frequently the case, they soon 
reappeared, clambered up a grass stem, and were then easily caught. 
In the rank grass which fringed the pool two large nests, evidently 
formed the preceding night, were discovered at a height of six inches 
from the ground. 

On June 2, nearly a score of nests in very diverse situations (pi. 1, 
fig. 2), their height above the water ranging from six inches to five 
feet, were found in yet another pool formed by the torrential rains. A 
few frogs were sitting on bare sprays of thorn in the full glare of a 
noontime sun; this apparently in preference to the shade afforded by 
the foliage on nearby shrubs and trees. In some instances these frogs 
were very cryptically colored. On returning to this pool at night, I 
captured three more males which were sitting on horizontal sprays of 
bramble and uttering small, rather bird-like, chirruping cries. Within 
a foot of one was a Spotted Wood Snake (Philothamnus s. semivarie- 
gatus) which may have been stalking the frog. 

Diet. Stomachs examined, held: (1) to (4) grasshoppers; (5) grass- 
hopper and two wingless crickets; (6) cockroach; (7) a 42 mm. cater- 
pillar; (8) caterpillar and beetle; (9) small beetles. 



LOVERIDGE: AFRICAN AMPHIBIANS 385 

Leptopelis concolor Ahl 

Leptopelis concolor Ahl, 1929, Sitz. Ges. Naturf. Freunde Berlin, p. 192: Witu, 
Kenya Colony. 

11 (M.C.Z. 20549-53) Witu, K.C. 3 & 31.V.34. 

1 (M.C.Z. 20554) Peccatoni, K.C. 24.V.34. 
20 (M.C.Z. 20555-9) Ngatana, K.C. 9-19.vi.34. 

7 (M.C.Z. 20560^) Malindi, K.C. 28.vi.34. 

Distribution. Known from the holotype male only. Heard calling 
at Mkowe, Mkonumbi, Laini, Golbanti, Karawa and Marareni. 

Affinities. Among East African Leptopelis, this distinctive little 
species appears to occupy a position intermediate between bocagii 
and johnstoni, having the shovel-shaped metatarsal tubercle as well 
as color and dorsal pattern common to both; it has the reduced 
webbing of the former, but the well-developed digital disks of the 
latter. 

Coloration in life. Malindi. Above, pale brown or ochre, lighter on 
sides; a burnt-umber band from nostril, through eye, over tympanum, 
beyond which it may sometimes be continued as a series of flecks for 
a very short distance; a triangular marking of paler umber between 
the orbits, its apex directed posteriorly, and, in occipital region, 
sometimes giving off diverging lines which may disappear about 
midbody, but usually persist as a series of flecks which converge 
upon the anus; similar flecks present on the limbs but most con- 
spicuous on the tibia. Below, white, or colorless except where tinted 
by the internal organs showing through. 

Measurements. The largest cf cf in each locality measure 39 mm., 
the twenty-three males range from 31 to 39 with an average of 36 mm. ; 
largest 9 (M.C.Z. 20560) measures 42 mm., the nine females range 
from 36 to 42 with an average of nearly 40 mm. 

Breeding. Though males were heard calling from May 3 to June 28, 
in fact during the whole period of my stay in their district, breeding 
took place in early May for only one female is gravid, she was taken 
May 3. Moreover seven young ranging from 19 to 24 mm. were 
captured at Ngatana in the middle of June, the two smallest had 
slight rudiments of tails. After the business of breeding is over, the 
frogs accumulate stores of fat, this is very noticeable in the Ngatana 
and Malindi series. 

Diet. Of ten stomachs examined, six held finely chewed insect 
remains, consisting, I imagine, of midges and mosquitoes which were 
abundant in the haunts of these frogs; one held a small moth; another 



386 bulletin: museum of comparative zoology 

a sphingid caterpillar 44 mm. in length, its horn alone measuring 
8 mm., a remarkable meal for so small a frog. 

Habitat. I purposed making a stay at Witu with a view to collecting 
this species. Two miles out of Witu, however, our truck stuck in 
tenacious black cotton soil concealed beneath a gleaming sheet of 
water which appeared, in the moonlight, to extend indefinitely over 
the water-logged, palm-studded, savanna. Fortunately an island of 
sandy ground was not far off as it was obvious that we must stay 
the night. Frogs were calling on every side and amongst the calls 
was one entirely new to me, it was a very distinctive cry formed of 
three separate notes. The calls came not only from the swamped 
lands, but from trees, and at a height of ten feet. I surmised that this 
songster was the tree frog of which I had come in search. While the 
boys removed a few essential loads from the truck, I set off in search 
of the frog. I captured the first within ten feet of the truck, a second 
in some grass, and a third in a bush, where it was resting at a height 
of four feet from the ground. 

On May 4, at Mkowe, I could hear one calling throughout a tor- 
rential downpour. On May 25, at Lake Peccatoni, I set out with a 
torch and eventually located a male calling from the dead branch of 
a doom palm where he was sitting seven feet above the ground. With 
his vocal sac inflated like a great bubble he looked very much like a 
Hyla. The bagginess of the throat serves to differentiate the males 
at this season, but I imagine that it subsides at the close of the 
breeding time. 

On May 31, back at Witu, I had great sport hunting them down 
with a torch which was shone in their eyes as they sat in the doom 
palms. It was always a gamble whether, on reaching the frog which 
you might have come fifty yards to get, you would find him within 
reach or high up on some frond. 

On June 11, at Ngatana, I took six in half-an-hour. They were 
calling as they clung to the tall sedges at heights of from four to 
eight feet. 

On June 28, at Malindi, the series was found on aloes at the edge 
of a swamp. 

The question arises as to why this frog, seeing it is so common in 
the Kenya coastlands, should have escaped collectors and description 
for so long? I suggest that it will be found that it remains in its 
burrow (note the shovel-shaped metatarsal tubercle) during much of 
the year, emerging only when the rains break, a season when Europeans 
shun this region. The abundant insect life furnishes it with a suffi- 



loveridge: African amphibians 387 

ciency of food to enable it to accumulate stores of fat in preparation 
for a further period of aestivation. This suggestion is frankly specu- 
lative. 

Leptopelis johnstoni (Boulenger) 

Hylambates johnstoni Boulenger, 1897, Proc. Zool. Soc. London, p. 803, pi. 
xlvi, fig. 4: Kondowe to Karonga and Nyika Plateau, Nyasaland. 

cf 9 young (M.C.Z. 20546-8) Ngatana, K.C. 12 & 20.vi.34. 

Distribution. These constitute the first records of the occurrence 
of this species in Kenya Colony. 

Variation. The tibiotarsal articulation of the adpressed hind limb 
reaches the eye in the male, scarcely so far in the female. In the field 
it was noted that both dorsal and ventral surfaces of the male were 
noticeably granular in life, they remain so in alcohol. 

Coloration in life, cf . Above, on all surfaces exposed when at rest, 
olive, concealed surfaces paler; upper lip white, an indistinct dark 
band, edged with lighter above, from end of snout through nostril, 
across upper eyelid, to above the tympanum; a dark green area both 
before and behind tympanum; three dark green cross-bands, irregular 
in outline, on fore arm and on tibia. Below, white. 

9 . Some rich green patches on the dorsolateral stripe. The young 
one was bright enamel-green above. 

I have never noticed any sign of green on Tanganyika frogs, but 
specimens from Mwaya, which are nearly topotypic with johnstoni, 
agree well with the coloring of the Ngatana series now that they have 
been in alcohol for two years. The male alone seems rather darker 
than the male from Mwaya. 

Measurements. The d" measures 46 mm.; the 9 , 62 mm.; the young 
one 21 mm. 

Breeding. The gonads of both adults are small, while the presence 
of the young one indicates that the breeding season was past. 

Habitat. I captured the male when it was sitting on a shrub in 
the forest. 

Hylambates maculatus Dumeril 

Hylambates maculatus A. Dumeril, 1853, Ann. Sci. Nat. (3), 19, p. 165, pi. 
vii, figs. 1-lb and 4: Zanzibar. 

9 (M.C.Z. 20543) Witu, K.C. 2.vi.34. 
d 1 <? (M.C.Z. 20544-5) Malindi, K.C. 28.vi.34. 

Distribution. These constitute the first records of the occurrence 
of this species in Kenya Colony. 



388 bulletin: museum of comparative zoology 

Coloration in life. 9 • Above, black, which on closer inspection will 
be found to be mottled with an even deeper black, each spot faintly 
outlined in silver; patches of scarlet on armpit, groin, fore and hind 
aspects of thigh, inner and upper surfaces of foot. Below, white, so 
nearly obscured by brown that the sides appear brown finely spotted 
with white. 

Measurements. The cf cf measure 72 mm., the 9 only 68 mm. 

Breeding. On June 28, at Malindi, I observed a male calling with a 
bubble-bursting note like that of Kassina senegalensis. He was half- 
submerged, resting on vegetation beneath the shelter of a doom palm 
in a swamp. The same night I saw the head of a second male just 
showing above the water in a similar situation; on seizing him I found 
that he was clasping a female, the latter escaped. 

Habitat. Having lifted the dry-grass nest and eggs of a rail from a 
clump of sedges growing in deep water, I was wading ashore with it 
resting on the palm of my hand when I felt something moving in the 
damp sedges which formed the bottom of the nest. On examination 
it proved to be the female maculatus listed above. The nocturnal 
habitat is described in the breeding note. 



Kassina senegalensis (Dumeril & Bibron) 

Cystignathus Senegalensis Dumeril & Bibron, 1841, Erpet. G£n., 8, p. 418: 

Galam, Senegal. 
Cassina Angeli Witte, 1933, Rev. Zool. Bot. Afr., 23, p. 172: Lukafu, Kende- 

lungu, Belgian Congo. 

55 (M.C.Z. 20768-75) Kaimosi, K.C. 15.ii & 6.iii.34. 
1 (M.C.Z. 20776) Kibwezi, K.C. 26.iii.34. 
1 (M.C.Z. 20777) Peccatoni, K.C. 24.V.34. 
125 (M.C.Z. 20778-82) Malindi, K.C. 29.vi.34. 
1 (M.C.Z. 20783) Opp. Kilindini, K.C. 6.vii.34. 

Synonymy. A 9 cotype of angeli (M.C.Z. 21658) is indistinguish- 
able from some of the females in the Kaimosi series. Its author sepa- 
rated it from senegalensis on shorter hind limbs and rugose and strongly 
denticulated anal flaps. The first of these characters is variable 
throughout East Africa, both short and long limbed examples of both 
sexes occurring in the same locality when adequate series are collected. 
The second character I believe to be a breeding season one, the smooth 
and scarcely denticulate flaps becoming rough and strongly papillose 



loveridge: African amphibians 389 

in females assembled for breeding. Specifically: 9 (M.C.Z. 20772) 
15.ii.24 (dry season) has smooth anal flaps and the tibio-tarsal articula- 
tion reaching her armpit. 9 9 (M.C.Z. 20773^) and all others taken 
6.iii.34 (in pond after rains had broken) are gravid and have rugose 
anal flaps while the tibio-tarsal articulation reaches the elbow in 3, 
the armpit in 2. 

Variation. One would expect that an eastern race of so widespread 
a species could be demonstrated; unfortunately we lack topotypic 
material, our nearest approach to Senegal being Belgian Congo speci- 
mens. The name somalica Scortecci might be applicable to such an 
eastern race, but its status seems doubtful unless almost the whole area 
of Kenya and Tanganyika is a region of intermediates. 

K. somalica was differentiated from senegalensis on the following 
grounds: (1) because the whole of its belly is granular; (2) its inter- 
orbital space is 1% the width of an upper eyelid; (3) its tympanum 
is % the width of the orbit. 

The corresponding conditions in K. senegalensis are: (1) the antero- 
central portion of the belly is smooth; (2) the interorbital space is 
said to equal an upper eyelid; (3) the tympanum is 2 /z the width of 
the orbit. 

If we take the forty-nine males in the Kaimosi series alone, we 
find: (1) many have the belly entirely granular while others have 
the antero-central area smooth as is the case with all females both 
from Kaimosi and the coast; (2) the interorbital space occupies 
an intermediate position being about V/i to l}/£ times the width of 
an upper eyelid; (3) the character is too close to be appreciated when 
applied to the material. 

Coloration in life. d 1 . Kaimosi. Above, pale bronze, a dark brown 
streak from nostril to eye where it is continued along edge of upper 
eyelid then backwards and downwards; a streak commences between 
eyes and extends backwards to the region of the coccyx to be followed 
after an interspace by an azygous blotch slightly to the right and a 
pair of blotches still further back; on the sides a pair of blotches, one 
anterior, the other posterior; still lower on flanks numerous irregu- 
larly shaped blotches; limbs spotted and blotched; outermost fingers 
and toes barred. Below, throat black, rest of undersurfaces pure white. 

Remarkable variation, however, is exhibited, ranging from those 
with an almost complete vertebral and dorso-lateral lines to others 
in which they are broken up into spots. While the latter condition 
(M.C.Z. 20770) is rare at Kibwezi, it is normal at Malindi but not 
at other coastal localities to the north or south of Malindi. 



390 bulletin: museum of comparative zoology 

Measurements. The largest <? <? (M.C.Z. 20769) and 9 9 (M.C.Z. 
20772), measure 43 mm. 

Breeding. On March 3, at Kaimosi, following the breaking of the 
rains, great numbers of males assembled at the millpond and started 
calling, previous to this only a single specimen had been obtained 
(February 15). Not only were five Kaimosi females gravid, but 
all other adults consisting of one on May 24 at Peccatoni, and two on 
June 29, at Malindi, emphasizing the difference between the rainy 
seasons at the coast and in the interior. 

Young, ranging in length from 24 to 30 mm., were swarming at 
Malindi on June 29. 

Diet. Stomachs examined, held: (1) many diptera including green- 
bottle; (2) flies, ants, three caterpillars. 

Habitat. The distinctive bubbling notes of these frogs could be 
heard a mile from the millpond at Kaimosi where the males were often 
half-submerged as they called. At Peccatoni, not only were they 
calling from two pools, but some called from thickets and one or two 
from long grass. There had been heavy rain for sometime past at 
Malindi, as well as on the very morning when so many young ones 
were found beneath logs and vegetable debris in a cotton plantation. 



Megalixalus fornasinii (Bianconi) 

Euchnemis Fornasinii Bianconi, 1848 (not 1850), Spec. Zool. Mosamb., Rept., 
pi. v, fig. 1: Mozambique. 

4 (M.C.Z. 20577-80) Peccatoni, K.C. 24.V.34. 
2 (M.C.Z. 20565) Near Witu, K.C. 30.V.34. 

5 (M.C.Z. 20566-70) Ngatana, K.C. 19.vi.34. 
64 (M.C.Z. 20571-5) Golbanti, K.C. 22.vi.34. 

1 (M.C.Z. 20576) Bulfagi Waterhole, K.C. (H.J.A.T.) v.32. 

Synonymy. Note correction under 31. brachynemis. See remarks 
under the heading of measurements below as to whether an inland 
race can be recognized. 

Coloration. Forty-five of the Golbanti series present a uniform 
dorsum, in thirty-nine a vertebral streak is present. 

Measurements. The figure of the type of fornasinii measures 30 mm. 
On collecting the coastal series listed above, I was struck by their 
smaller size as contrasted with the extensive collections of montane 
material available for comparison. 

The fourteen c? c? from the coast range from 26 to 31 mm. (only 



loveridge: African amphibians 391 

one is 31, and one 30 mm.) with an average of 27 mm.; the fourteen 
largest 9 9 range from 26 to 30 mm. (only two attain 30 mm.) with 
an average of 28 mm. 

Contrast this with the published range and averages of a hundred 
males and a hundred females from the Uluguru and Usambara Moun- 
tains and it will be seen that the coastal frogs average nearly 7 mm. 
shorter for both males and females respectively, quite an appreciable 
amount when it is a fifth of the total length. 

That it is not a question of montane versus coastal plain, however, 
seems probable because frogs from Mwaya (on the shores of Lake 
Nyasa), Kilosa, and Morogogo, attain the proportions of the montane 
frogs. It is possibly explainable as a coastal plain versus an upland- 
inland form for which latter the name loveridgii Procter is available 
in a subspecific sense. 

Breeding. On May 24, the single adult female taken at Peccatoni, 
was gravid. Even so late as June 24, some of the Golbanti females 
had not spawned. 

Habitat. All taken on sedges in or around swamps. 



Megalixalus fulvovittatus (Cope) 

Hyperolius fulvovittatus Cope, 1860, Proc. Acad. Nat. Sci. Philadelphia, p. 

517: Liberia. 
Hyperolius brevipalmatus Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 25: Sang- 

melina, south Cameroon. 

12 (M.C.Z. 20596-600) Kaimosi, K.C. iii.34. 

Variation. These frogs have been compared with a topotypic series 
from Liberia which had previously been compared with what remains 
of the type. A comparison of Aid's description and fig. 150 of his 
brevipalmatus (Das Tierreich, 1931, p. 279) with that of fulvovittatus 
given on p. 331, fig. 204 of the same volume, will remove any doubts 
as to their identity. 

Measurements. The largest cf & and 9 9 alike measure 28 mm. 

Breeding. Nine of the frogs are males, they were found calling in 
the millpond on the evening immediately succeeding the breaking of 
the rains. 

Megalixalus brachynemis Boulenger 

Megalixalus brachynemis Boulenger, 1896, Ann. Mag. Nat. Hist. (6), 18, p; 
403, pi. xvii, fig. 2: Chiradzulu, Nyasaland. 



392 bulletin: museum of comparative zoology 

Hyperolius pygmaeus Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 22: Tanga, 
Tanganyika Territory. 

10 (M.C.Z. 20581^) Peccatoni, K.C. 24.V.34. 
10 (M.C.Z. 20585-9) Near Witu, K.C. 30.V.34. 
10 (M.C.Z. 20590-4) Golbanti, K.C. 22.vi.34. 
1 (M.C.Z. 20595) Changamwe, K.C. 4.vii.34. 

Distribution. Seen also in a pond at Kililana on May 21, 1934. 

Corrigenda. In 1933 (pp. 397-398), I compared the description of 
the type of Hyperolius pygmaeus Ahl with M.C.Z. 16805 from Moro- 
goro, and claiming they were the same, referred pygmaeus to the 
synonymy of M. fornasinii. While I still think that M.C.Z. 16805 
is specifically identical with pygmaeus, I must confess to having over- 
looked the fact that this Morogoro frog was a brachynemis, though 
taken with a long series of fornasinii in bananas growing near Moro- 
goro station. 

M. brachynemis is so exceedingly like a Hyperolius that, without 
examining their pupils, I again mistook some in the field. The species 
may, however, not only be distinguished from all other East African 
Hyperoli, but from M. fornasinii as well, by its shorter hind limb. 
In M. brachynemis the tibio-tarsal articulation of the adpressed hind 
limb reaches only to the insertion of the fore arm, or in old softened 
material, at most scarcely to the hinder corner of the orbit. 

Coloration in life. cf. Peccatoni. Above, silvery yellow with 
irregular longitudinal streaks of brown, a brown streak from tip of 
snout through nostril and orbit to the groin; tibia silvery yellow 
obliquely streaked with brown; thighs and feet colorless except for 
numerous minute black specks and a blood vessel which shows red 
along the whole length of the thigh. Below, throat yolk-yellow; 
belly silvery white, not cream. 

Measurements. The largest o 71 c? from all localities, measure 20 
mm.; the largest 9 (M.C.Z. 20590) is 24 mm. 

Breeding. On May 24, at Lake Peccatoni, only males were taken, 
good evidence that they were just assembling for breeding. On May 30, 
near Witu, 30% of those taken were females, two of which were gravid. 
On June 22, at Golbanti, about 70% of the frogs taken were females; 
most of the adults were gravid, a few presumably had spawned 
recently. 

Habitat. On May 5, a few frogs of both sexes were taken in bananas 
at Golbanti, as was the case with the male from Changamwe. All the 
others were taken in pools or swamps where they had assembled for 
breeding as recorded above. 



loveridge: African amphibians 393 



Hyperolius rossii (Calabresi) 

Rappia Rossii Calabresi, 1925, Atti. Soc. Ital. Sci. Nat. Milano, 64, p. 121, 

fig.: Upper Uele region, Belgian Congo. 
Hyperolius phrynoderma Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 71: No 

locality, but coll. Deut. Zentral-Afrika Expedition. 
Hyperolius mohasicus Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 85: Lake 

Mohasi, Belgian Ruanda-Urundi. 
Hyperolius oculatus Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 103: Balaibo, 

Duki River, w. of Lake Albert, Belgian Congo. 
^Hyperolius irregularis Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 114: Lake 

Mohasi, Belgian Ruanda-Urundi. 

1 (M.C.Z. 20601) Kirui's Village, K.C. 20.ii.34. 
118 (M.C.Z. 20602-14) Kaimosi, K.C. iii.34. 

Distribution. In addition to a cotype of phrynoderma, a name based 
on faded material, we have this species from Zambi, Belgian Congo 
(part of the series referred, to pusillus Cope by Noble, 1924, p. 256), 
and Kisumu, Kenya Colony (referred to ? pusillus by Barbour & 
Loveridge, 1930, p. 795). It is evidently a common species in the 
Central Lake Region. 

Variation. The adults are generally recognizable by the numerous 
fine red spots on the throat, flanks and belly of the females, the 
dorsum of the males. In the latter, more rarely in the former, they 
may form the centre of a raised pimple, or the pimple be present 
without the spot; as the backs of some of the males are smooth; 
however, it may be a breeding season development. It is the only 
East African species known to me which has the dorsum so studded; 
a few species have pimples or granules on the head only. 

Coloration in life, cf . Above, uniformly putty-color except for a 
very few, scattered, minute, black spines; thighs colorless except 
for subdermal red areas usually forming four streaks, a similar streak 
on the inside of the foot, toes also red; shins and outer aspect of feet 
putty-color but finely flecked with gray. Below, throat and granular 
abdominal area cream-color, the oval patch surrounded by a grayish 
area. 

9 . Above, head, back, shins and outer aspect of feet cream-color, 
speckled with numerous, well-defined, small, round, red spots; thighs 
and inner surface of feet colorless except for subdermal, blood-red 
streaks and spots. Below, as in male but the whole of the undersurface 
with scattered dark red spots which are larger than those on the 
dorsum. 



394 bulletin: museum of comparative zoology 

9 . Above, head, back, shins and outer aspect of feet green with 
numerous large, round, yellow spots each of which has a dark red 
centre, the yellow spots often coalesce so that the ground color forms 
a green network about them; shins with only a single spot; thighs 
and under surface of foot redder than in the other female. Below, as 
in other female, but throat and abdominal area bright yellowish- 
green spotted with dark red. 

Measurements. The largest cf (M.C.Z. 20603) measures 33 mm., 
the largest 9 (M.C.Z. 20606) 35 mm. 

Breeding. The Kirui and Kaimosi females are gravid, the latter 
(50) being taken with the males (40) at the millpond where they had 
assembled. 

Enemies. Two were taken from the stomachs of green snakes 
(Chlorophis hoplogaster) and one from a tree viper (Athens squamigera) 
at Kaimosi, while the Kirui specimen was removed intact from the 
stomach of a frog (Rana in. venusta) I 

Habitat. This is definitely a rain-forest form which lives in the 
tops of the tall forest trees. On March 1 there was a terrific down- 
pour in the afternoon, on the following day leaves twice came rotating 
down to my feet from a great height, on each occasion one of these 
beautiful little frogs was squatting on the leaf. 

Folklore. The Maragoli and Watereki adhere tenaciously to a belief 
that cattle are killed while grazing through inadvertently swallowing 
these frogs; the frogs are alleged to stick in their throats! 

Hyperolius picturatus Peters 

Hyperolius picturatus Peters, 1875, Monatsb. Akad. Wiss. Berlin, p. 206, pi. 
ii, fig. 2: Victoria, Cameroon. 

148 (M.C.Z. 20615-9) Kaimosi, K.C. ii-iii.34. 

Native names. Lutu (Luragoli); lungala or loingolu (Lutereki, but 
the former is also applied to ranid frogs). 

Affinities. H. simus Ahl from Usumbura, northeast of Lake Tan- 
ganyika, will probably prove to be synonymous with picturatus. In 
the absence of topotypical material, and in view of his remarks about 
the coloration of the young, I refrain from synonymizing it at present. 

Coloration in life. Most of the adults exhibited a solid, bright red 
flash on the anterior aspect of the thighs. This greatly assisted in 
distinguishing them from examples of rossii. After preservation in 
formalin nothing remains of the flash. 



loveridge: African amphibians 395 

Measurements. The largest c? (M.C.Z. 20615) measures 32 mm., 
the largest 9 (M.C.Z. 20616) 36 mm. 

Breeding. On March 3, following the breaking of the rains, males 
and females were assembling at the millpond. 

Habitat. On February 17, I noted that all the picturatus were 
found in wild bananas in the forest, whereas rossii, apart from those 
that fell from the forest canopy, were found in sedges. Later of course 
both species migrated to the millpond for breeding. 

Hyperolius montanus (Angel) 

Rappia montana Angel, 1924, Bull. Mus. Hist. Nat. Paris, 30, p. 269: Mount 
Kinangop, Aberdare Range, Kenya Colony. 

8 (M.C.Z. 16118-23) Mt. Kinangop, K.C. (H.J.A.T.) 1931. 
3 (M.C.Z. 20620-2) Uplands, K.C. (C.A. du T.) l.ii.34. 

Distribution. In addition to a cptype, the Museum of Comparative 
Zoology possesses this species from several localities around Mount 
Kenya. 

Affinities. Though subject to wide variation, the color pattern of 
the young and males show it to be an offshoot of H. picturatus. 

Hyperolius undulatus (Boulenger) 

Rappia undulata Boulenger, 1901, Ann. Mus. Congo (1), 2, fasc. 1, p. 4, pi. 

ii, fig. 2: Pweto and Lofoi, Katanga, Belgian Congo. 
Rappia symetrica Mocquard, 1902, Bull. Mus. Hist. Nat. Paris, 8, p. 408: 

Athi River, Kenya Colony. 
Rappia platyrhinus Procter, 1920, Proc. Zool. Soc. London, p. 416, fig. 3: 

Nairobi, Kenya Colony. 
Hyperolius asper Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 49; Nairobi, Kenya 

Colony. 
Hyperolius nyassae Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 66: Langenburg 

(i.e. Manda?), Tanganyika Territory. 

3 (M.C.Z. 20623-5) Uplands, K.C. (C.A. du T.) l.ii.34. 
30 (M.C.Z. 20626-31) Nairobi, K.C. 19.iii.34. 
2 (M.C.Z. 20632-3) Mt. Mbololo, K.C. 20 & 27.iv.34. 
1 (M.C.Z. 21336) Amani, Usambara Mtns., T.T. (R.E.M.) vi.35. 

Synonymy. The Museum of Comparative Zoology possesses cotypes 
of undulatus and nyassae. Ahl's figure (1931, Das Tierreich, p. 340, 
fig. 213) might almost have come from the same block as his figure 
(fig. 212) of symetrica, both give an excellent representation of the 



396 bulletin: museum of comparative zoology 

lateral pattern of the species. The dorsal pattern of a triangular 
interocular spot and vertebral series of blotches, or a line, may be 
seen in the figures of undulatus and platyrhinus. Occasional indi- 
viduals have these characteristic markings obsolescent, or even absent. 

In addition to the synonymy listed above, the Nairobi cotype of 
Ahl's pictus (= marginatus Peters) should be referred to undulatus. 
This is apparently also the case with the cotypes of his breviceps 
from Ravine Station, Kenya Colony, a faded cotype (M.C.Z. 17629) 
of which has been available. It is possible that the type, which came 
from Tschimbo, Mozambique, is something different. 

Coloration in life. tf> (M.C.Z. 20632). Above, yellowish white, a 
pigment formed from the concentration of minute brown dots produces 
a ) ( shaped mark from the upper eyelids ; a very zigzag line from the 
posterior corner of the eye to the middle of the flank where it termi- 
nates; a pair of blotches on the back in lumbar region, and also a 
cross-bar on the tibia; upper surfaces of hands and feet white flecked 
and blotched with olive; concealed surfaces red. Below, pure white 
except for the concealed red surfaces. 

Measurements. Adult cf cf from both Nairobi and Mbololo, 30 
mm. ; no adult females taken. 

Habitat. On March 19, 1934, having had the advantage of heavy 
showers during the two preceding nights which had brought to an 
end a prolonged season of drought, I took a car out on the Machakos 
Road to a spot where the river is choked with sedges and almost 
stagnant. We arrived half-an-hour before dark, but though five of 
us searched while daylight lasted, we only succeeded in securing 
half-a-dozen of these frogs. Most of these were ensconced at the base 
of the broad leaves of the plant known as Crinium kirlcii (I am in- 
debted to Miss Napier of the Coryndon Museum for the identification). 
These flat cool leaves retained a small amount of moisture from 
the previous showers, in much the same fashion as do the outer leaves 
of the wild banana plants. 

As soon as it became dark, however, these little frogs appeared 
in such numbers upon the sedges and grasses at the water's edge as 
enabled us, aided of course by electric torches, to procure the rest of 
the series in a further half-hour. It seems, therefore, that the majority 
pass the hot hours of the day concealed deep down in the base of 
tufts of grass, possibly also on the sedges at water level. Most of the 
series are young, many still carrying a caudal stump. A single adult 
H. striolatus was taken with them. 

One Mbololo male was taken on a papyrus growing in a stream on 



loveridge: African amphibians 397 

the southern slope at 3,000 feet, the other on sedges in a marsh on 
the eastern side at 4,000 feet. 

Hyperolius flavoguttatus Ahl 

Hyperolius flavoguttatus Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 96: Bukoba, 
Tanganyika Territory and Mount Kenya, Kenya Colony. 

d* (M.C.Z. 20634) Mt. Mbololo, K.C. 24.iv.34. 

Native name. Kengele (Kitaita, literally 'a bell' on account of its 
call, doubtless applied to the preceding species as well). 

Affinities. We have one of the faded Bukoba cotypes which bears 
no resemblance to Ahl's figure in Das Tierreich, we have also a topo- 
type from Mount Kenya which does, and with which this species 
compares tolerably well. Our specimen closely resembles Giinther's 
reticidatus, 1860, of no known locality, while the Kenya frog is still 
more like Megalixalus flavomacidatus Giinther, 1864, Rovuma Bay. 

If Tornier's figures, reproduced by Ahl (1931, Das Tierreich, 
p. 320) as representing H. taeniatus Peters of Mozambique do, in 
reality represent that species, then our Mbololo frog should be 
referred to taeniatus. However, Peters figure of the type from Boror, 
Mozambique, seems to me possibly to represent a different frog. 

Coloration in life. cT • Above, a very dark olive anteriorly becoming 
progressively paler towards the anus, both head and back handsomely 
spotted with yellow; thighs pink, except for a subanal area which is 
cream-colored without markings; tibia like the dorsum; fore limbs 
and tarsus tend to be cream-colored heavily vermiculated or marbled 
with brown. Below, gular disk yellowish edged with brown; chest 
and abdomen pure white; rest of undersurfaces flesh-pink. 

Measurement, cf. 34 mm. as is our Mount Kenya male. 

Breeding. Attracted by an explosive whistle, not unlike the bubble 
note of Kassina, I descended from my camp to the marsh where I 
had caught a solitary male undulatus three days before, and captured 
this lone specimen on a sedge. The only two that called during the 
week in which I camped at this marsh. 

Hyperolius striolatus Peters 

Hyperolius striolatus Peters, 1882, Sitz. Ges. naturf. Freunde Berlin, p. 9: 

Taita, Kenya Colony. 
Rappia ferniquei Mocquard, 1902, Bull. Mus. Hist. Nat. Paris, 8, p. 407: 

Athi River, Kenya Colony. 



398 bulletin: museum of comparative zoology 

Hyperolius coeruleopunctatus Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 76: 

Nairobi and Kibwezi, Kenya Colony. 
? Hyperolius pulchromarmoratus Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 92: 

Kenya Colony (Hiibner legit). 
Hyperolius scheffleri Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p.l 11: Kibwezi, 

Kenya Colony. 

1 (M.C.Z. 20635) Uplands, K.C. (C.A. du T.) l.ii.34. 
1 (M.C.Z. 20636) Nairobi, K.C. 19.iii.34. 
29 (M.C.Z. 20637-45) Kibwezi, K.C. 23.iii.34. 

Synonymy. The Kibwezi series may be considered topotypic of all 
three species described by Ahl, for Hiibner, as well as Scheffler, 
lived there and I took these frogs from a pond beside the road between 
the station and their homes. H. pulchromarmoratus appears to me to 
have been based on an individual in wdiich the minute punctatae 
have coalesced to form short dashes, it is rather more extreme than 
any in our series, so I am not too certain as to its disposition. 

Coloration. Some of the nine young attributed to this species, 
exhibit traces of the lateral markings associated with undulatus. The 
latter almost certainly occurs at Kibwezi but until definitely proven 
it seems best to refer the young to striolatus. 

Measurements . The largest c? d" (M.C.Z. 20637) measure 34 mm.; 
the largest 9 9 (M.C.Z. 20638) 36 mm.; the youngest 16 mm. 

Breeding. On March 23, at Kibwezi, a tink-tonk call was attrib- 
uted to this frog; males were assembling and greatly outnumbered 
the females, only two of which were adult and gravid. 

Hyperolius mariae Barbour & Loveridge 

Hyperolius mariae Barbour & Loveridge, 1928, Mem. Mus. Comp. Zool., 50, 
p. 217, pi. hi, fig. 1 : Derema, Usambara Mountains, T.T. 

Hyperolius melanopthalmus Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 68: 
Zanzibar. 

Hyperolius renschi Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 115: Zanzibar. 

9 9 (M.C.Z. 21334-5) Amani, Usambara Mtns., T.T. (R.E.M.) vi.35. 

Synonymy. Mr. H. W. Parker writes me that he considers melan- 
ophthalmus a synonym of mariae. I have compared a cotype of the 
former with the type of the latter and entirely agree. In addition 
I would add renschi to the synonymy. Mr. Parker considers the 
subdermal dark streak on the flank diagnostically important and I am 
inclined to agree. In this case fuelleborni from Lake Nyasa, which 
I placed in the synonymy in 1933, will have to be revived. 



loveridge: African amphibians 399 

Hyperolius vermicularis Ahl ? 

Rappia vermiculata Pfeffer, 1893 (1892), Jahrb. Hamburg Wiss. Anst., 10, 

part 1, p. 98, pi. i, fig. 12: Zanzibar. 
Hyperolius vermicularis Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 24: n.n. for 

R. vermiculata Pfeffer, 1893, preoccupied by Hyperolius vermiculatus Peters, 

1882. 

5 (M.C.Z. 20646-50) Mkonumbi, K.C. 22.V.34. 
152 (M.C.Z. 20651-9) Peccatoni, K.C. 24-26.V.34. 
5 (M.C.Z. 20660) Near Witu, K.C. 31.V.34. 

1 (M.C.Z. 20661) Golbanti, K.C. 22.vi.34. 

2 (M.C.Z. 20663-4) Malindi, K.C. 30.vi.34. 

1 (M.C.Z. 20665) Bulfagi Waterhole, K.C. (H.J.A.T.) v.32. 
1 (M.C.Z. 20666) Fundi Isa, K.C. (H.J.A.T.) vi.32. 

Identification.. This is probably the commonest species along the 
coast; an adequate series having been collected at the outset, further 
material was only taken for locality record. Its identity with vermi- 
cularis must remain in doubt, however, until compared with topotypic 
material from Zanzibar. 

Some specimens agree well with Pfeffer's rather poor figure of one 
of his two somewhat shrivelled types. As these were not males, it is 
always possible that he had the young of some closely related, but 
larger, species, such as sansibarica. Their smaller size precludes 
them from being referred to rubripes Ahl, from Kililana, just north of 
Mkonumbi. 

Coloration in life, c/ 1 . Lake Peccatoni. Above, olive green to gray 
green, sides lighter, thighs and feet red, the thighs varying from 
unpigmented to those which are heavily speckled with black. Below, 
throat cream or buffy white, necked with black, the vocal sac an 
egg-yolk yellow; belly cream; lower surface of limbs transparent 
pink, necked along the edges with minute black specks. 

A note was made at Malindi that the coloring agreed well with the 
above. At Mkonumbi it was remarked that these frogs had not red 
legs like those of sansibarica. In the laboratory it is observed that 
the females are all vermiculated, either darkly or very lightly. While 
the majority of the males appear uniformly gray in alcohol, the 
younger ones are often like the females, the pigment being grad- 
ually lost, frequently leaving a few irregularly disposed smudges in 
in the lumbar region as in the figure of petersi Ahl, from Mombasa, 
a species that may prove to be synonymous with this material. On 
sorting the 140 Peccatoni males, 28 are vermiculated, 59 uniform,, 
while 53 might be said to be intermediate. 



400 bulletin: museum of comparative zoology 

Measurements. The largest cT c? (M.C.Z. 20651-2) measure 27 
mm., the largest 9 9 (M.C.Z. 20656-7) 28 mm. 

Breeding. On May 22, only males were taken at Mkonumbi, where 
they were calling with an explosive snap; 140 males to only 12 females 
collected at Peccatoni, indicating that the males were just assembling 
at the end of May. 

Hyperolius sansibaricus (Pfeffer) 

Rappia sansibarica Pfeffer, 1893 (1892), Jahr. Hamburg Wiss. Anst., 10, p. 97, 

pi. ii, fig. 4: Zanzibar. 
Hyperolius rubripes Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 88: Kililana, 

near Lamu Island, Kenya Colony. 

6 (M.C.Z. 20667-9) Mkonumbi, K.C. 22.V.34. 
10 (M.C.Z. 20670-4) Peccatoni, K.C. 24.V.34. 
12 (M.C.Z. 20675-9) Near Witu, K.C. 30.V.34. 

1 (M.C.Z. 20680) Ngatana, K.C. 20.vi.34. 

1 (M.C.Z. 20662) Karawa, K.C. 26.vi.34. 

1 (M.C.Z. 20681) Malindi, K.C. 29.vi.34. 

Synonymy. These frogs have been compared with Pfeffer's types. 
They are specifically identical with the frogs from Bagamoyo which 
I (1933, p. 399) referred with a query to sansibaricus. Also at least 
the adult of the two frogs from Dar es Salaam which Barbour & 
Loveridge (1928, p. 225) identified as flavoviridis Peters. The latter 
name, I am now convinced, was based onad 1 argns Peters of which 
it should be considered a synonym. They agree in size and color with 
Ahl's rubripes from Kililana, near Mkonumbi, which is now made 
a synonym. It appears possible that the name sansibaricus may have 
to be synonymized with citrinus Gunther from the Zambesi. 

Coloration in life. cf. Mkonumbi. Above, pale yellowish green 
becoming yellower on the limbs; thighs and lower surface of tibia 
blood-red. Below, throat yellow; chest cream-color. 

cf . Peccatoni. Above, uniform rich green (rapidly becoming yel- 
lowish green after death) on head, back, tibiae and exposed outer 
edges of arms and feet (under a lens the tibiae are seen to be minutely 
dotted with black); thighs, extending backwards to inner anterior 
surface of tibiae, as well as the concealed upper surface of feet, blood- 
red. Below, throat yolk-yellow; belly cream-color. 

9 . Peccatoni. Above like male, but below the throat is cream-color 
instead of yellow. 



loveridge: African amphibians 401 

Measurements. The largest cf (M.C.Z. 20681) measures 31 mm.; 
the largest 9 • 9 are 32 mm. 

Breeding. On May 22, at Mkonumbi, males only were taken, their 
cry an explosive snap like that of the preceding species. Two days 
later, at Peecatoni, I found a female on her mass of eggs which were 
wrapped about the stem of a reed at a height of eight inches above the 
surface level of a pond, that was being daily augmented by the tor- 
rential rains. On May 31, near Witu, many masses of spawn were 
found in swamped grassland at a height of six inches above the 
water; it is possible that some were laid by the previous species — 
vermicularis. 

Hyperolius puncticulatus (PfefTer) 

Rappia puncticulata Pfeffer, 1892 (1893), Jahr. Hamburg Wiss. Anst., 10, 

part 1, p. 31, pi. ii, fig. 2: Zanzibar. 
Hyperolius noblei Ahl, 1931, Mitt. Zool. Mus. Berlin, 17, p. 118: Kilwa, 

Tanganyika Territory. 

9 (M.C.Z. 20701) Malindi, K.C. 28.vi.34. 

Synonymy. From the description I see no reason for regarding 
noblei as distinct. The Kilwa fauna is that of the coastal plain with 
an admixture of species such as Spaeleophryne methneri known only 
from near Kilwa and the Uluguru Mountains, where H. ■puncticulatus 
is so abundant. 

Coloration in life. Above, straw yellow, a broad, dusky canthal 
band. 

Measurements. 9 measuring 37 mm. 

Habitat. Taken the same evening, and in the same pond, as the 
males of 11. akli sp. no v., from which it differed in the absence of 
spotting and not having a green ground color. 

Hyperolius argentovittis Ahl 

Hyperolius argentovittis Ahl, 1931, Mitt. Zool. Mus. Berlin, p. 72: Ujiji, Lake 

Tanganyika, Tanganyika Territory. 
Hyperolius callichromus Ahl, 1931, Mitt. Zool. Mus. Berlin, p. 99: Western 

Bank of Rusisi River and northwest shore of Lake Tanganyika, Belgian 

Congo. 

Synonymy. The Museum of Comparative Zoology possesses a 
topotype of argentovittis and cotypes of callichromus. It was unfor- 



402 bulletin: museum of comparative zoology 

tunate that in using the name callichromus in 1933 (p. 403) I failed 
to realize that argentovittis was only an example in which the spots 
had coalesced to form stripes. It can be matched in our extensive 
series from Lake Tanganyika, though the Ujiji topotype is typically 
callichromus in its blotches. 

I might mention that the Museum possesses a poorly preserved 
frog (M.C.Z. 20700), collected by Mr. H. J. Allen Turner at Kaimosi, 
which has only the silvery, black-edged, lateral streaks, the back being 
uniform. It probably represents some different form and is not re- 
ferred to argentovittis. 

Hyperolius ahli sp. nov. 

Hyperolius argus Loveridge (part, not of Peters), 1925, Proc. Zool. Soc. Lon- 
don, p. 788: Tindiga and Kipera, near Kilosa, Tanganyika Territory. 

Hyperolius callichromus Ahl (part), 1931, Mitt. Zool. Mus. Berlin, 17, p. 99: 
Kililana, Kenya Colony; Dar es Salaam, etc., Tanganyika Territory. 

9 (M.C.Z. 10435) Tindiga, T.T. 23.vii.21. 
9 (M.C.Z. 10438) Kipera, T.T. 5.V.23. 
c? (M.C.Z. 20901) Dar es Salaam, T.T. 9.xi.26. 
1 of 3 9 (M.C.Z. 20682-4) Peccatoni, K.C. 24.V.34. 
10 & 8 9 (M.C.Z. 20685-9) Near Witu, K.C. 31.V.34. 

8 9 (M.C.Z. 20690-4) Golbanti, K.C. 22.vi.34. 
5 c? 3 9 (M.C.Z. 20695-7) Malindi, K.C. 28.vi.34. 

2 9 (M.C.Z. 20698-9) Marafa, K.C. (H.J.A.T.) v.32. 

Remarks. It may be recalled that when Dr. Ernst Ahl described 
callichromus from the Rusisi River and localities northwest and 
northeast of Lake Tanganyika, he listed other paratypes from Kililana 
in Kenya, Kawenda, Bagamoyo and Dar es Salaam in Tanganyika 
Territory. He remarked that the Dar es Salaam frogs differed in 
color pattern from the Lake specimens in certain ways, but that he 
had insufficient material to say whether these individuals represented 
a distinct form or not. 

Having collected puncticulatus at Dar es Salaam, I made the mistake 
of assuming that the material would only be variants of that variable 
form and I (1933, p. 406) referred this part of callichromus to the 
synonymy of puncticulatus. With the fresh material now at my 
disposal I humbly apologize for my error and take pleasure in asso- 
ciating the name of Dr. Ahl with this new form which links argent- 
ovittis (inc. callichromus) and puncticulatus in which the sexes are alike, 
with argus which displays such remarkable sexual dichromatism. 



loveridge: African amphibians 403 

In this connection it might be interesting to remark that in the 
field I provisionally listed the spotted females as callichromus, and 
the green males as a species unknown to me. On my return to Cam- 
bridge I spent a day in unsuccessfully endeavouring to identify the 
males with some known species. A week later, when measuring the 
spotted forms, I was astonished to find that all twenty-four were 
females. Parker's interesting discovery of the sexual dichromatism 
in argns then occurred to me, while on referring to his colored plate 
(1931 (1930), Proc. Zool. Soc. London, p. 902, pi. i) I was immedi- 
ately struck by the close parallelism in coloring between the sexes 
of argus and ahli. 

I might here state that Peters founded argus on a female, while 
later in the same paper his flavoviridis and tettensis were undoubtedly 
males and so become synonyms of argus. 

Type. Museum of Comparative Zoology, No. 20,682, an adult cf 
from Lake Peccatoni, northeast of Witu, coast of Kenya Colony, 
collected by Arthur Loveridge. 

Paratypes. The rest of the material listed above. 

Diagnosis. A large species related to puncticulatus, but occupying 
an intermediate position between argentovittis Ahl and argus Peters. 
From the latter the male differs in little but size, the female, how- 
ever, differs from female argus in the limbs being usually spotted 
like the dorsum and the spots on both being white-centred instead of 
yellow. 

Neither sex of ahli have a color pattern of a white (or silvery) 
vertebral line or series of spots, while the females have large spots 
or biack-edged reddish-brown streaks on the limbs like those on the 
dorsum, instead of small black specks as is the case with argentovittis. 

Coloration in life, cf Type. Peccatoni. Above, all exposed surfaces 
green including the thighs, which, with the feet, are paler than the 
back; head, back, fore arms and tibiae dotted with black; a light 
green line, extending from orbit to two-thirds of the distance to the 
groin, is faintly indicated. Below, green, though gular disk, chest 
and belly are slightly cream-color; groin, knees, and other joints are 
bluish green; hands and feet tinged with yellow. 

9 Paratype. Peccatoni. Above, olive, following the contour of the 
snout from its end backwards to, and around, the eyes, a broad 
V-shaped mark which, with a lateral streak, two azygous spots on 
dorsum, a supra-anal spot, two spots on fore arm and three on tibia, 
are pure white edged with black; hands, thighs and feet, pale red. 
The number and arrangement of these spots, which at times coalesce 



404 bulletin: museum of comparative zoology 

to form bars, are subject to infinite variation, yet never form a verte- 
bral row or streak. 

Coloration in alcohol, cf Type. Above, flesh-eolor, the black dots 
now dusky, as is also a rather faint canthal band from nostril to eye 
which was not noticeable in life. Below, lighter flesh-color. Some of 
the paratypes are uniform, others show a light, canthal-lateral band 
which is edged above and below with a series of dusky dashes. 

9 Paratype. Above, pinkish brown, the markings pale flesh-color 
heavily edged with black. Below, light flesh-color. 

Measurements, cf Type measures 35 mm. from snout to anus, 
the sixteen paratype cf cf range from 32 to 37 mm. with an average 
of 34 mm.; the 9 paratype whose coloration is described above, meas- 
ures 31 mm., the largest of twenty-six 9 9 measures 35 mm. 

Habits. Apparently these frogs remain under water during the day 
for in half-an-hour spent wading about in a pond covered with lily 
pads at Witu, only one was seen. Continuing over the same area 
immediately after sunset, more than a dozen were disturbed. The 
males went leaping and splashing across the surface in a character- 
istic fashion. Females, despite their conspicuous coloring, are very 
wary, plunging into the water and swimming away while one is still 
far distant. 



Hyperolius parkeri Loveridge 

Hyperolius parkeri Loveridge, 1933, Bull. Mus. Comp. Zool., 74, p. 140: 
Bagamoyo, Tanganyika Territory. 

7 d" (M.C.Z. 20702-6) Peccatoni, K.C. 24.V.34. 

4 c? 5 9 (M.C.Z. 20707-8) Near Witu, K.C. 30.V.34. 

1 9 (M.C.Z. 20709) Ngatana, K.C. 13.vi.34. 

1 9 (M.C.Z. 20710) Golbanti, K.C. 22.vi.34. 

4 <? 1 9 (M.C.Z. 20711) Bulfagi Waterhole, K.C. (H.J.A.T.) v.32. 

Secondary sexual character. In addition to the color differences 
noted in the original description, it is now observed that these breed- 
ing males have a patch of black spines at the hinder part of the 
abdomen, lower surface of thighs and soles of feet. 

Coloration in life. Of the Ngatana female (possibly applicable to 
all) it was noted that it was entirely green beneath. 

Measurements. The largest cf (M.C.Z. 20702) measures 26 mm., 
range of fifteen males 20-26 mm., average 23 mm.; the largest 9 
(M.C.Z. 20707) measures 23 mm., range of eight females 18-23 mm., 



loveridge: African amphibians 405 

average 21 mm. It seems highly probable, therefore, that in this 
species the male may attain larger dimensions than the female. 

Breeding. On May 24, at Lake Peccatoni, males were assembled 
but no females encountered; a week later at Witu, gravid females 
were present in proportions at least numerically equal to the males. 



Hyperolius nasutus Giinther 

Hyperolius nasutus Giinther, 1864, Proc. Zool. Soc. London, p. 482, pi. xxxiii, 

fig. 3: Duque de Braganca, Angola. 
Rappia granulata Boulenger, 1901, Ann. Mus. Congo (1), 2, fasc. 1, p. 4, pi. 

ii, fig. 3: Pweto, Lake Mweru, Belgian Congo. 

1 d" (M.C.Z. 20725) Mkonumbi, K.C. 22.V.34. 
22 <? 9 9 (M.C.Z. 20726-33) Peccatoni, K.C. 24.V.34. 
3 9 (M.C.Z. 20734-6) Near Witu, K.C. 30.V.34. 
1 <? (M.C.Z. 20737) Golbanti, K.C. 22.vi.34. 

Synonymy. I feel convinced that the unique type of granulata was 
a nasutus, probably formalin preserved, for such, even though trans- 
ferred to alcohol at an early date, do not exhibit the silvery naso- 
orbital-lateral stripe. It appears rather as a flesh-colored stripe 
which is free from the minute specks, that, concentrated above and 
below, tend to accentuate it. In our material it is even less distinct 
in the females which have a slightly less prominent snout. See re- 
marks under granulatus in Loveridge (1933, p. 410). 

For comparative material I have a c? from Ngola, Angola, deter- 
mined as Jiasutus by Boulenger; the two 9 9 from Nyamkolo, thirty 
miles from the type locality of granulatus; a 9 from Kabengere and 
a juvenile from Elizabeth ville, both in Belgian Congo; and a 9 from 
Kibonoto, Mount Kilimanjaro which was one of the series incorrectly 
referred to puncticulatus by Lonnberg (1907, p. 25). 

Coloration in life. cf. Mkonumbi. Above, green, a white lateral 
stripe from eye only, is black-edged.- 

9 . Near Witu. Above, green, a small vermilion blotch on the 
anterior surface of each thigh. Throat green. (Eggs visible). 

Breeding. On May 24, at Peccatoni, all the females were gravid; 
a week later, at Witu, this condition prevailed. 

Measurements. The twenty-three males measure 20-24 mm., 
average 22 mm.; the twelve 9 9 range from 19-26 mm., average 
23 mm. 



406 bulletin: museum of comparative zoology 



Hyperolius milnei Loveridge 

Hyperolius milnei Loveridge, 1935, Bull. Mus. Comp. Zool., 79, p. 18: Near 
Witu, Coast Province, Kenya Colony. 

12 (M.C.Z. 20712-8) Kililana, K.C. 21.V.34. 
75 (M.C.Z. 20025-50) Near Witu, K.C. 31.V.34. 

2 (M.C.Z. 20051-2) Golbanti, K.C. 22.vi.34. 

1 (M.C.Z. 20053) Malindi, K.C. 28.vi.34. 

1 (M.C.Z. 20719) Sokoki Forest, K.C. (H.J.A.T.) vi.32. 

Affinities. Apparently closely related to albofrenatus Ahl from 
Tanganyika Territory, but differs in not being silvery white nor 
possessing a large silvery white subocular spot. On the other hand 
milnei normally possesses a dark canthal stripe which appears to be 
lacking in albofrenatus. 

Coloration. The Kililana frogs were not designated as paratypes 
as they lack both the canthal stripes and conspicuous dorsal spotting 
which characterize the Witu series, to which, perhaps, they should 
not be referred. The coloring in life, however, points to the prob- 
ability that they should be regarded as a single species. 

Breeding. On May 21, at Kililana, males outnumbered females 
two to one, the females were gravid and spawn was seen upon the 
submerged vegetation. On May 31, at Witu, dozens of gravid females 
were taken. On June 28, at Malindi, a male was heard calling and 
many of the species seen. 

Habits. At Kililana, these frogs were taken in a swamped mbugwe 
about a mile from the site of the German planter's house. Here, at 
midday, these rich green frogs with a pale yellow lateral stripe, 
were found squatting on the blades of water-logged grass in the 
larger pools. As often as not-the grass would be lying on the surface 
of knee-deep water, and, as one approached, the frog would dive off, 
seldom reappearing. 

At Witu, these small frogs were found sitting on the numerous 
lily pads and blades of grass lying on the surface of a lake a mile 
or so from Mr. Milne's house. By day they were very alert, leaping 
and swimming away long before one was within reach of them. It 
was noted that in this lake, they did not remain below the surface, but 
after swimming a short distance, scrambled out on to a lily pad with 
alacrity. The difference in their behaviour to that of those at Kililana, 
may possibly be accounted for by the presence of numerous cat fish 
at Witu; in fact, as I waded about almost waist-deep, I trod on 



loveridge: African amphibians 407 

several while others collided with my legs. As dusk descended these 
frogs began calling vociferously, and it was then found that they 
might be captured with relative ease by a swoop of the hand. 

Enemies. One was recovered from the stomach of an Olive Water 
Snake (Natrix o. olivaceus) at Golbanti on the Tana River. 

Hyperolius pusillus (Cope) 

Crumenifera pusilla Cope, 1862, Proc. Acad. Nat. Sci. Philadelphia, p. 343: 

Umvoti, Natal. 
Hyperolius microps Glinther, 1864, Proc. Zool. Soc. London, p. 311, pi. xxvii, 

fig. 3: Rovuma Bay, East Africa. 
Hyperolius usaramoae Loveridge, 1932, Proc. Biol. Soc. Washington, 45, p. 63: 

Mogogoni Swamp, south of Dar es Salaam, Tanganyika Territory. 
Hyperolius translucens Power, 1935, Proc. Zool. Soc. London, p. 339, figs. 6-7, 

pi. i, figs. 3-4; Port St. Johns, Pondoland coast, Union of South Africa. 

5 9 (M.C.Z. 20720-4) Golbanti, K.C. 22.vi.34. 

Synonymy. Recently, through the generosity of Mr. J. H. Power, 
we received a series of para types of his translucens with both sexes 
represented. I recognized them immediately as being similar to my 
usaramoae and careful comparison with the types failed so find any 
differences whatever. If the published accounts of the color in life, 
as given by Power and myself, be compared, it will be seen that they 
undoubtedly refer to the same species. 

Whereas Barbour & Loveridge (1928, p. 225) originally referred the 
types of usaramoae, together with some other frogs, to microps, it later 
became apparent that there were two species and regarding the 'other 
frogs' as microps (which they were not, they were later described under 
the name of parkeri Loveridge), I proceeded to describe the real microps 
material as usaramoae. 

As Port St. Johns is not far distant from Umvoti, Natal, before read- 
ing Mr. Power's description of translucens, I compared his para types 
with the description of pusillus and decided that they represented the 
same creature. Later, on looking up the original description, I saw that 
Power, himself, suggested that possibility. 

An important aid to the identification of this species, at least it is 
present in our Kenya, Tanganyika and Pondoland material, is the 
presence of two bright, light, transverse, subdermal streaks on the 
occiput just between, and slightly posterior to, the eyes. 

Measurements. These five females range from 20-21 mm. 

Breeding. They are not gravid. 



408 bulletin: museum of comparative zoology 

RANIDAE 

Rana edulis (Peters) 

Pyxicephalus edulis Peters, 1855, Arch. Naturg., 21, part 1, p. 56: Boror and 
Tete, Mozambique. 

Phrynopsis Boulengeri Pfeffer, 1893 (1892), Jahrb., Hamburg Wiss. Anst., 10, 
p. 101, pi., ii, figs. 5-6: Mozambique. 

Pyxicephalus flavigula Calabresi, 1916, Monit, Zool. Ital., 27, p. 34, pi. ii, 
fig. 1 : Oroflillo, Italian Somaliland. 

Phrynopsis usambarac Ahl., 1924, Zool. Anz., 60, p. 271: Usambara, Tangan- 
yika Territory. 

40 juv. (M.C.Z. 20219-23) Mkonumbi, K.C. 22.V.34. 
11 juv. (M.C.Z. 20224-8) Peccatoni, K.C. 23.V.34. 

3 (M.C.Z. 20229-31) Near Witu, K.C. 30.V.34. 

1 (M.C.Z. 20232) Golbanti, K.C. 22.vi.34. 

1 (M.C.Z. 20233) Karawa, K.C. 26.vi.34. 

1 (M.C.Z. 20234) Gongoni, K.C. 27.vi.34. 

2 (M.C.Z. 20235-6) Mombasa Id., K.C. 3.vh.34. 

5 (M.C.Z. 20237-8) Mainland opp. Kilindini, K.C. 6.vii.34. 

Synonymy. In 1882, Boulenger referred edulis to the synonymy of 
adspersa Dumeril & Bibron, thereafter all workers, including myself, 
have referred their Tanganyika and Kenya material to adspersa. 

In 1929, I began to be puzzled that no East African specimens 
exceeded 138 mm. in length, i.e. only about half the size of the South 
African adspersa. In 1934 I obtained gravid females of small size and 
found the species occurred all the way up the coast to Lamu, i.e. just 
south of the Italian Somaliland border. 

Pfeffer based his genus Phrynopsis on the cartilaginous omosternum 
and sternum. When I received the types for examination from Herr 
de Grijs I immediately recognized them as the young of edulis. The 
genus Phrynopsis therefore becomes a synonym of Rana or rather its 
subgenus Pyxicephalus. 

Peters characterized edulis as having a hind limb equal to the body 
length, Miss Calabresi also only mentions the shorter length of the 
hind limb as separating flavigula from adspersa. On reinvestigating the 
status of these frogs, I conclude that all East African records should be 
transferred to edulis, and adspersa struck off the East African list. 

Rana edulis differs from adspersa in much smaller size; slightly 
shorter hind limbs; slightly longer toes; a smoother, less rugose, inte- 
gument. Doubtless its coloration differs also, a green dorsal stripe is 
characteristic. 



loveridge: African amphibians 409 

While the upper lip is usually marmorate, it is uniformly white in 
usambarop and in some frogs from Kilosa and Dar es Salaam in the 
Museum of Comparative Zoology. 

Measurements. The largest d 71 (M. C. Z. 20229) measures 98 mm., 
the largest 9 (M. C. Z. 20230) 76 mm. 

Breeding. On May 30, near Witu, I took three of these frogs in a 
pool where they had undoubtedly gone to breed; of the pair whose 
measurements are given above, the 9 was gravid with spawn. Young, 
about 32 mm. in length were swarming at Mkonumbi, Peccatoni and 
Golbanti. The Karawa, Gongoni and Mombasa frogs were halfgrown. 
Youngest of all were the Kilindini series just out of the tadpole stage 
and about 10 mm. long. 

Diet. So voracious were these little creatures that two of them, 
despite the fairly rapid action of the potassium cyanide, when dropped 
into a killing bottle, seized, and partly swallowed, two of their fellows. 
At Lake Peccatoni, a young Rana m. mascareniensis was recovered 
from the stomach of an edulis scarcely larger than itself. The Witu 
female held a leaf frog (Hyperolius sp.), a small crab (presumably 
Potamon bottegoi), a yellow millipede, carabid and other beetles as well 
as the remains of various insects. 

Enemies. A young edulis was recovered from the stomach of a 
Stripe-bellied Sand Snake (Psammophis subtceniatus) at Mkonumbi. 

Rana occipitalis Giinther 

Rana occipitalis Giinther, 1858, Cat. Batr. Sal. Brit. Mus., p. 130, pi. xi: 
"Africa", "West Africa", "Gambia". 

1 (M.C.Z. 20238) Mt. Debasien, U. 24.xi.33. 

Habitat. This young, 62 mm., frog, was taken among long grass in 
the Amaler River at an altitude of 5,000 feet. 

Rana fuscigula chapini Noble 

Rana chapini Noble, 1924, Bull. Am. Mus. Nat. Hist., 49, p. 214, fig. 6a: 
Batama, Belgian Congo. 

10 (M.C.Z. 20261-5) Mt. Debasien, U. 20.xi.33. 
5 (M.C.Z. 20280-4) Sabei, Elgon, U. 9.xii.33. 

12 (M.C.Z. 20239-4) Butandiga, U. 8.i.34. 
Tadpoles & 4 (M.C.Z. 20245-9) Buluganya, U. 12.L34. 
1 (M.C.Z. 20250) Budadiri, U. 17.L34. 
84 (M.C.Z. 20285-9) Kaimosi, K.C. ii.34. 

11 (M.C.Z. 20251-9) Nairobi, K.C. 17.iii.34. 

1 (M.C.Z. 20260) Mt. Mbololo, K.C. 20.iv.34. 



410 bulletin: museum of comparative zoology 

Distribution. These records serve to fill the hiatus between the 
Belgian Congo and Nairobi, Usambara and Uluguru mountains from 
which this form is known. For comments on the occurrence of this race 
and R. f. angolensis at Kaimosi, see remarks under that form. 

Native navies. Akidot (Karamojong) ; lungalla (Luragoli and Lu- 
tereki); chula (Kitaita), but none specific. 

Measurements. The largest cf (M. C. Z. 20248) measures 72 mm., 
the largest 9 (M. C. Z. 20245) 98 mm., next largest 9 (M. C. Z. 
20260) 90 mm. 

Breeding. On January 12, at Buluganya, males were heard calling, 
and a series of tadpoles about 65 mm. in length were collected. In 
February, at Kaimosi, only two young frogs had rudiments of tails, 
these two frogs measured 43 (27 + 16) and 74 (43 + 31) mm. respec- 
tively. On March 17, two of the frogs in the Nairobi series exhibited 
tails in various stages of absorption, the longer tail measuring 28 mm. 

Parasites. Encysted nematodes are present on the top of the head 
of the large Buluganya female, and what appear to be ruptured cap- 
sules are present on the palms of her hands. Another female (M. C. Z. 
20239) from Butandiga, has lost the toes of one foot entirely, possibly 
from a similar infection? 

Enemies. One frog was recovered from the stomach of a tree snake 
(Hapsidophrys lin eata) . 

Habitat. In streams associated with montane forest from 7,000 feet 
(Butandiga) to 4,000 feet (Mbololo) except for those in a deep pool at 
Bulanganya. This pool was at the foot of a hundred-foot cliff over 
which a stream spilled into the pool. Six frogs were observed on the 
shallow edges of the pool; they were very wary, however, and dived 
into the deep mud when approached. Those from Mount Debasien 
were found sitting in the shallows of the Amaler River at 5 a.m. I 
captured them by shining their eyes with an electric torch. 

Rana fuscigula angolensis Bocage 

Rana angolensis Bocage, 1866, Jorn. Sci. Math. Nat. Phys. Lisboa, p. 73: 
Duque de Braganga, Angola. 

2 (M.C.Z. 20266-7) Sipi, U. 14.xii.33. 
Tadpoles & 4 (M.C.Z. 20268-72) Kaburomi, U. 28.xii.33. 
24 (M.C.Z. 20273-9) Kaimosi, K.C. ii.34. 

Distribution. A frog, almost certainly this species, sprang into a 
stream at Madangi, 11,500 feet, on Mount Elgon, Uganda. 

While the preceding race is usually associated with forest streams, 



loveridge: African amphibians 411 

angolensis is an upland, open-country form. It was surprising, therefore 
to find it at Sipi, on western Elgon, between Sabei to the north and 
Butandiga to the south where undoubted chapini occur. It may be 
assumed that it has followed the course of the Sipi River and its 
tributaries up to Kaburomi which is in the alpine-meadow zone. 

The Kaimosi specimens are also of considerable interest as both 
races occur in the vicinity. Much deforestation has taken place and 
today patches of forest abut on the grasslands. I personally collected 
some fuscigula (presumably chapini) along a stream flowing through 
dense primary forest, but most of the series of 104 frogs from Kaimosi 
were brought in by native children, so that speculation as to whether 
they were procured within or without the forest would be idle. Many in 
the chapini series are intermediate as the webbing on the fifth toe extends 
scarcely to the tip (vide key in Loveridge, 1933, p. 364), it seems highly 
probable that the two forms are meeting and interbreeding in this area 
as the savanna encroaches on this one-time forested country. 

Native names. Isodo (Lugishu). 

Measurements. The largest d 71 (M. C. Z. 20268) measures 55 mm., 
the largest 9 (M. C. Z. 20266) is 64 mm., Kaimosi females only a 
millimetre less. 

Breeding. On December 28, in an icy-cold stream at 10,500 feet at 
Kaburomi, tadpoles and young in all stages of development, together 
with four adult males, were taken. 

Diet. Stomachs of ten Kaimosi frogs held: (1) many moths; (2) 
moths; (3) moths, ant, and many flies; (4) moths, muscid fly, green- 
bottle fly; (5) five greenbottle flies, two large black flies, two tumbo 
flies, ichneumon or slender- waisted wasp; (6) flies; (7) flies and beetles; 
(8) beetle and hemipteron; (9) cricket, beetles and millipede; (10) ants 
of a huge black species, neuropteran, millipede. 

Habitat. At Sipi in a shallow stream flowing through native gardens 
in open, though formerly forest, country. At Kaimosi as they sat on 
the edge of the millpond at night, I captured them by shining their eyes 
with an electric torch. 

Rana wittei (Angel) 

Phrynobatrachus Wittei Angel, 1924, Bull. Mus. Hist. Nat. Paris, 30, p. 130: 

Molo, Mau Escarpment, Kenya Colony. 
Rana aberdariensis Angel, 1925, in Voyage de Ch. Alluaud et R. Jeannel en 

Afrique Orientale (1911-1912), Paris, p. 42, pi. ii, figs. 1 & 2: Mount Kinan- 

gop, Aberdare Mountains, Kenya Colony. 

3 <? (M.C.Z. 20290-2) Molo, K.C. 12.iii.34. 



412 bulletin: museum of comparative zoology 

Distribution. These topotypes of wittei are from a locality little 
more than fifty miles from the type locality of aberdariensis. 

Synonymy. R. wittei was referred to the genus Phrynobatrachus by 
its author as the type lacked vomerine teeth. As this is also a condition 
common to very young frogs of the genus Rana, I (1930, p. 31, foot- 
note) suggested that wittei was a young Rana, but hesitated to refer it 
to any species on account of its dried condition. 

I find Angel's description of wittei coincides with these three males 
except that (1) the tympanum almost equals the orbital diameter in 
our material; (2) the length of the tibia is included in the length from 
snout to anus from 1% to 2}/£ times instead of 2^ times in the type, 
but 2% times in the figure of the type (vide second citation, pi. iii, 
fig. 5) ; (3) the skin is not smooth above like the smooth skin of the 
type. 

These topotypes of wittei have the characteristic pigmented gular 
sacs of aberdariensis with a cotype of which they have been compared. 
Curiously enough, after just writing the foregoing, and coming to the 
conclusion that the two species are synonymous, I turned up some 
manuscript notes written in Paris in 1927, after examining the type of 
wittei. One reads "? a young Rana, very like the young of Rana 
aberdariensis." This suggestion had entirely escaped my memory; now 
nine years later the same conclusion is arrived at when approached 
from a different angle. 

Coloration. Throats chrome yellow. 

Breeding. The dusky vocal sac, thickened forearm, swollen base of 
first digit, in conjunction with minute dermal spines on soles of hind 
feet, outer side of shanks, inner and outer aspects of thighs, abdomen 
and back, proclaim these breeding males. They were taken just at the 
end of a prolonged dry season, the rains usually breaking in April in 
this locality. 

Habitat. We arrived at the Highland Hotel, Molo, at 7 p.m., after 
an eleven-hour truck ride from Kaimosi. It was already dark, but as 
we must start soon after daybreak next morning, I went out with a 
flashlight at 8.30 p.m. and made my way in the direction of a solitary 
frog-call. Eventually I captured these specimens in a cattle-trampled, 
boggy, yet slowly-flowing stream beside the bridge immediately 
below the Hotel on the Londiani-Molo Road. Next morning my 
wife searched the same spot and secured several Phrynobatrachus 
keniensis. 



loveridge: African amphibians 413 

Rana galamensis bravana (Peters) 

Limnodytes bravanus Peters, 1882, Sitz. Ges. naturf. Freunde Berlin, p. 3: 
Barawa, i.e. Brava, Italian Somaliland. 

1 (M.C.Z. 20293) Kirimanda River, K.C. 3.V.34. 

2 (M.C.Z. 20294-5) Peccatoni, K.C. 25.V.34. 
7 (M.C.Z. 20296-7) Golbanti, K.C. 22.vi.34. 
4 (M.C.Z. 20298-300) Malindi, K.C. 28.vi.34. 

Native name. Malondi (Kipokomo). 

Variation. Reasons for recognizing bravana as the name applicable 
to the eastern form have been given in my report on the Field Museum 
African Collections (1936, p. 95). 

The tibiotarsal articulation of the adpressed hind limb reaches the 
eye in all specimens (5 cf d 71 , 8 9 9) except a juvenile male from 
Golbanti. 

Coloration in life, cf Malindi. Above, greenish yellow, obscured 
anteriorly by burnt umber which takes the form of an undulating 
vertebral line on the forward part of the back, and irregular blotches 
posteriorly ; flanks and thighs beautifully marbled with burnt umber on 
a cream-colored ground; tibia and fore limbs colored like the sacral 
region but with a few blotches on the fore arms; fingers cream color; 
toes whitish. Below, white, the gular sacs black; the pads on the upper 
arm cream-colored; limbs exhibiting subdermal mottling; soles of hands 
and feet gray. 

The 9 coloring is essentially like that of the cf but for the absence 
of vocal sacs and arm pads. Based on a pair from Malindi. 

Measurements. The largest cf (M. C. Z. 20298) and 9 (M. C. Z. 
20299) both measure 75 mm. 

Breeding. On June 28, males at Malindi were calling 'ku-wek', yet 
readily distinguishable from the call of R. m. mascareniensis, as they 
rested on submerged vegetation in deep water sheltered by a small 
doom palm. 

Diet. Stomachs examined, held: (1) grasshopper; (2) grasshopper, 
large hairy caterpillar with stiff hairs; (3) orthopteran limbs, 40 mm. 
long wings and body of agrionid dragonfly; (4) agrionid dragonfly, 
spider; (5) larval dragonfly, spider; (6) three earwigs, spider; (7) large 
black cricket; (8) rat-tailed maggot, staphylinid beetle, spider; (9) 
fossorial wasp with sting extruded, bombardier beetle, small skipjack 
and other beetles; (10) three woodlice. This frog appears to feed on a 
great variety of unpleasant creatures. I recovered a scorpion (Isomctms 
macidatus) from a Bagamoyo specimen. 



414 bulletin: museum of comparative zoology 

Habitat. The specimen from the bed of the Kirimanda River was 
disturbed by the whirring wheels of our truck which was bogged in 
the black cotton soil. I shot the first Peccatoni male as it sat on some 
floating bark five feet from the edge of the pool. The second I seized 
at night by shining its eyes as it sat on some floating reeds in knee-deep 
water. The Golbanti series were captured beneath masses of wet, 
recently-cut sedges near the water's edge. 

Ra.na floweri Boulenger 

Ranafloweri Boulenger, 1917, Ann. Mag. Nat. Hist. (8), 20, p. 417: Rosaires, 
Blue Nile, Anglo-Egyptian Sudan. 

Rana erlangeri Ahl, 1923, Mitt. Zool. Mus. Berlin, 11, p. 5: Lake Abaya, north- 
east of Lake Stephanie, Ethiopia. 

7 <? + 16 young (M.C.Z. 20398-402) Peccatoni, K.C. 24.V.34. 
2 9 (M.C.Z. 20403-4) Ngatana, K.C. 9.vi.34. 

Variation. The tibiotarsal articulation of the adpressed hind limb 
reaches the eye in the nine adults, the eye or a little beyond in the 
young; fourth toe with two phalanges free of web, remaining toes with 
one phalanx free, or almost free, of web. 

Coloration. Considerable difficulty was experienced in separating 
these short-limbed frogs from some of the mascareniensis taken in the 
same swamp. The following pattern characters proved helpful. R. 
floweri has finer vermiculations on the buttocks; though in two 
examples there is a hair-like vertebral line, none possess the broad 
vertebral band of mascareniensis, the dorsum being usually chequered 
with black. The throat may be immaculate, or with faint infuscations, 
in both species. 

Measurements. The largest c? (M. C. Z. 20398) measures 53 mm.; 
the larger 9 (M. C. Z. 20403) only 48 mm. 

Breeding. On May 24, at Peccatoni, males were assembled and call- 
ing, but numerous young were present, ranging in length from 15 mm., 
exclusive of caudal rudiments, to 28 mm. 

On June 9, at Wema, both the 48 and 34 mm. females were distended 

with ova. 

Diet. Stomachs examined, held: (1) Two grasshoppers, smooth- 
skinned caterpillar; (2) grasshopper; (3) two grasshoppers, two beetles; 
(4) two beetles, big ant; (5) similar species of ant 15 mm. long; (6) 
gravid sphingid moth; (7 and 8) empty. 



loveridge: African amphibians 415 



Rana oxyrhynchus oxyrhynchus Smith 

Rana oxyrhynchus A. Smith, 1849, Illus. Zool. S. Africa, Rept., pi. Ixxvii, 
figs. 2, 2a-c: Kafirland and region of Port Natal. 

10 (M.C.Z. 20301-5) Mt. Debasien, U. 18.xi.33. 

5 (M.C.Z. 20376-9) Kirui's Village, K.C. 20.i.34. 

2 (M.C.Z. 20396-7) Nairobi, K.C. 19.iii.34. 
37 (M.C.Z. 20306-9) Kibwezi, K.C. 23.iii.34. 
12 (M.C.Z. 20310-4) Tsavo, K.C. l.iv.34. 
15 (M.C.Z. 20315-6) Mt. Mbololo, K.C. 19.iv.34. 

1 (M.C.Z. 20317) Lamu, Lamu Id., K.C. 8.V.34. 

2 (M.C.Z. 20318-9) Peccatoni, K.C. 24.V.34. 
4 (M.C.Z. 20320-3) Ngatana, K.C. 9.vi.34. 

9 (M.C.Z. 20324-9) Malindi, K.C. 28.vi.34. 

1 (M.C.Z. 20333) Sokoki Forest, K.C. (H.J.A.T.) vi.32. 

2 (M.C.Z. 20330-1) Changamwe, K.C. 4.vii.34. 

1 (M.C.Z. 20332) Opp. Kilindini Harbour, K.C. 4.vii.34. 

Native names. Kengele (Kitaita); chuachanco (Kipokomo), but 
neither specific being applied to frogs, as distinct from toads, in general. 

Variation. The tibio-tarsal articulation of the adpressed hind limb 
reaches far beyond the tip of the snout; fourth toe with 13^2 or 2 
phalanges free of web, remaining toes webbed to the tips, or occasion- 
ally the third toe with one phalange free of web. 

Measurements. The largest <? (M. C. Z. 20378) measures 42 mm. 
the largest 9 (M. C. Z. 20324) 55 mm. 

Breeding. Not breeding in November on Mount Debasien. On 
May 8 and 24, on Lamu and at Peccatoni respectively, males were 
assembling and calling in response to the first rains; in the former 
locality, in the absence of standing water, their enthusiasm had led 
them into a deep well. 

Between March 23 and June 28, gravid females were taken at 
Kibwezi, Mbololo, Ngatana and Malindi. During the same period, 
however, young which ranged from 17 to 22 mm. in length, were en- 
countered at Kibwezi, Tsavo and Malindi. 

Diet. Stomachs examined, held: (1) Spider; (2) spider; (3) lycosid 
spider, cockroach, grasshopper, woodlouse; (4) grasshopper; (5) grass- 
hopper; (6) grasshopper, acridian, beetle; (7) elater beetle; (8) two 
beetle larvae; (9) large cricket; (10) caterpillar, froghopper; (11) hemip- 
teron and fulgorid homopteran. 

Enemies. One was recovered from the stomach of a House Snake 



416 bulletin: museum of comparative zoology 

(Boodon lineatus) on Lamu Island, another from that of a Spotted 
Wood Snake (Philothamnus s. semivariegatus) at Malindi. 

Habitat. The Debasien series were taken at 5,000 feet. During the 
day they might be found as much as twenty feet from the river for 
which they would make if disturbed; the males leaping across the 
surface, doubtless possible by reason of their strongly webbed feet. 
At night they sat at the water's edge, or in the shallows, where they 
were easy to catch by shining a torch in their eyes. 

Rana oxyrhynchus gribinguiensis Angel 

Rana (Ptychadena) Gribinguiensis Angel, 1922, Bull. Mus. Hist. Nat. Paris, 
28, p. 399, fig. : Fort Crampel, Gribingui, French Congo. 

9 (M.C.Z. 20334) Butandiga, U. 10.i.34. 

9 9 (M.C.Z. 20335-6) Kaimosi, K.C. 19.ii.34. 

9 (M.C.Z. 12706) Amani, T.T. 20.xi.26. 

Distribution. This form is added to the fauna of all three British 
Territories for the first time. In East Africa I consider gribinguiensis 
to be the rain-forest representative of the widespread oxyrhynchus, to 
which Barbour and I (1928, p. 194) indeed referred the Amani, Usam- 
bara Mountains, specimen relisted above, considering it to be an 
example of gigantism resulting from the favourable climatic conditions 
prevailing at Amani. 

Affinities. I believe that I am justified in according subspecific 
rank to gribinguiensis for we already have Rana f. angolensis of the 
lowlands and uplands represented by the larger, better webbed R. f. 
chapini in the rain forests, again R. m. mascareniensis of the lowlands 
and uplands represented by the larger R. m. venusta in the rain forests. 

Moreover the only characters which I can find to separate gribin- 
guiensis are: (1) Greater size, the five known females measuring 59 
(type), 63, 63, 63 and 67 mm. respectively as against 55 mm. for an ex- 
ceptionally large female of oxyrhynchus, females of the latter rarely ex- 
ceeding 50 mm. (2) In gribinguiensis only l-l^ phalanges of the fourth 
toe are free of web, the remaining toes are fully webbed. In oxyrhyn- 
chus 13^-2 (usually 2) phalanges of the fourth toe are free of web, the 
remaining toes are usually fully webbed or a phalange of the third may 
be free. 

Neither the diameter of the tibia into its length, nor the length of the 
tibia into that of the head and body serve to distinguish gribinguiensis 
either from South or East African oxyrhynchus. The former may 



loveridge: African amphibians 417 

prove to be a slender average character as the diameter is included 
3.6 to 4.0 times in the length as against 3.5 to 3.8 in oxyrkynchus, the 
tibia is included 1.2 to 1.3 times in the length from snout to vent in the 
four females listed above, as against 1.3 to 1.5 times for an equal num- 
ber of females from Southern Rhodesia. 

Measurements. The largest known 9 (M. C. Z. 20335) measures 
67 mm. 

Diet. Stomachs examined, held: (1) Hymenopteran of large size; 
(2) two hemiptera; (3) beetle and cockchafer. 

Parasites. Larval nematodes (Physaloptera sp.) were present in a 
Kairnosi frog. 

Enemies. One of the Kaimosi specimens was recovered from the 

stomach of a green snake (Chlorophis hoplogaster). 

I 

fc. - 

Rana mascareniensis mascareniensis Dumeril & Bibron 

Rana mascareniensis Dumeril & Bibron, 1841, Erpet. Gen., 8, p. 350: Mada- 
gascar; Mauritius; Seychelles. 

1 (M.C.Z. 20337) Kikuyu, K.C. 14.iii.34. 

2 (M.C.Z. 20338-9) Nairobi, K.C. 19.iii.34. 
21 (M.C.Z. 20340-9) Kibwezi, K.C. 23.iii.34. 

5 (M.C.Z. 20350-4) Mkonumbi, K.C. 22.V.34. 
24 (M.C.Z. 20355-9) Peccatoni, K.C. 24.V.34. 
1 (M.C.Z. 20360) Kau, Tana R., K.C. 4.vi.34. 

13 (M.C.Z. 20361-4) Laini, K.C. 6.vi.34. 

14 (M.C.Z. 20365-9) Ngatana, K.C. 9.vi.34. 
11 (M.C.Z. 20370-4) Golbanti, K.C. 22.vi.34. 

1 (M.C.Z. 20375) Karawa, K.C. 26.vi.34. 

Variation. The tibiotarsal articulation of the adpressed hind limb 
reaches the eye, nostril, or just beyond end of snout, the latter chiefly 
in halfgrown or young frogs, irrespective of sex, all three conditions 
may be found in the Kibwezi series; 2 phalanges of the fourth, and 1, 
rarely 1^ phalanges, of the remaining toes free of web. 

Measurements. The largest d 1 (M. C. Z. 20359) measures 48 mm., 
the largest 9 (M. C. Z. 20338) 59 mm. 

Breeding. On May 24, at Peccetoni, it was noted that males were 
calling. Between March 19 and June 6 gravid females were taken at 
Nairobi, Kibwezi, Peccatoni and Laini. During the same period, 
however, young, which ranged from 17 to 19 mm. at Mkonumbi, 20 to 
21 mm. at Peccatoni, 24 to 28 mm. at Laini, and 27 mm. at Nairobi, 
were encountered. 



418 bulletin: museum of comparative zoology 

Diet. Stomachs examined, held: (1) a 15 mm. ant (Megaponera); 
(2) 25 mm. moth; (3) large smooth-skinned caterpillar; (4) cricket; 
(5) cricket, grasshopper, spider; (6) cockroach, grasshopper; (7) grass- 
hopper; (8) grasshopper, fly, spider; (9) beetle, cockchafer, spider. 

Parasites. Larval nematodes (Physaloptera sp.) were numerous in 
the stomachs of a Peccatoni and Ngatana frogs. 

Enemies. One recovered from the stomach of a Cormorant (Halietor 
a. africanus) and on two occasions from White-lipped Snakes (Crota- 
phopeltis h. hotamboeia) at Wema, Ngatana. Twice from Night Adders 
(Causus resimus) and once from a young frog (Rana edulis) at Pecca- 
toni. 

Rana mascareniensis uzungwensis Loveridge 

Rana mascareniensis uzungwensis Loveridge, 1932, Bull. Mus. Comp. Zool., 
72, p. 384: Dabaga, Uzungwe Mountains, Tanganyika Territory. 

Synonymy. This race has recently been referred to the synonymy of 
R. subpunctata Bocage, by my friend, Mr. K. P. Schmidt (1936, p. 129), 
who makes subpunctata a subspecies of mascareniensis. 

That Schmidt's material from Gauca and Chitau is correctly identi- 
fied with uzungwensis is possible, but that the latter is identifiable with 
subpunctata I most heartily disagree. Before describing uzungwensis I 
went into this possibility and rejected it for several reasons. 

R. m. uzungwensis was based on a series of thirteen breeding frogs, 
with immaculately white bellies, of which the largest cf measured 42 
mm., the largest 9 only 44 mm. Now the holotype cf of subpunctata 
measured 51 mm., which, if it was a mascareniensis at all — which I 
doubt — would place it in the giant R. m. venusta Werner group. How- 
ever, unlike the forms of mascareniensis, its throat is marbled and its 
breast, abdomen and limbs are spotted. In fact its color description is 
much like that of katangae Witte as will be seen from the figures of 
that species (W'itte, 1921, Revue Zool. Afr., 9, pi. ii) to which it also 
approximates in size. Witte's types being c? cf 50-55.5 mm., 9 9 
52.5-56 mm. 

I also emphatically disagree with the disposition of anchietae Bocage 
and porosissima Steindachner by placing them in the synonymy of 
subpunctata. It is true that Boulenger, (1882, p. 53) referred all three 
to the synonymy of mascareniensis, but Bocage (1895, p. 160), though 
willing for the first two to remain there, emphasized the distinctness 
of subpunctata. 

It may be recalled that uzungwensis differed from the typical form 



loveridge: African amphibians 419 

in having 3 phalanges of the fourth toe free of web. In his description 
of anchietae, Bocage distinctly states that only the last two phalanges 
of the fourth toe are free, this places it in the synonymy of R. m. mas- 
careniensis where he was content to leave it. Steindachner makes no 
precise statement regarding the webbing on the fourth toe of porosis- 
sima, but his figure of the foot shows only a single phalange free of 
web. 

Angola is vast enough to harbour the three forms — R. m. mascare- 
niensis, R. m. uzungwensis and R. m. venusta. 

Rana mascareniensis venusta Werner 

Rana venusta Werner, 1907, Sitz. Akad. Wiss. Wien, 116, part 1, pp. 1889 
and 1892, pi. iv, fig. 11: Entebbe, Uganda; Mongalla and Lagos. 

4 (M.C.Z. 20380-3) Butandiga, U. 17.i.34. 

7 (M.C.Z. 20384-9) Kirui's Village, K.C. 20.i.34. 

8 (M.C.Z. 20390-4) Kaimosi, K.C. 10.ii.34. 

Variation. The tibiotarsal articulation of the adpressed hind limb 
reaches well beyond the end of the snout; l-l}^ phalanges of the first 
toe, 1 of the second, 1-2 of the third, 2-3 of the fourth, and 1 of the 
fifth toe free of web. 

Measurements. The largest <? <f (M. C. Z. 20385, 20392) measure 
52 mm., the largest 9 9 (M. C. Z. 20382, 20392) measure 66 and 64 
mm. 

Breeding. Adult females in all three localities were gravid, with the 
exception of one from Kirui's Village which had apparently spawned 
already. 

Diet. Stomachs examined, held: (1) a tree frog {Hyperolius rossii); 
(2) grasshopper, pentastomid bug; (3) pentastomid bug, cockroach, 
beetle; (4) two beetles, hairy caterpillar, spider. 

Habitat. Though there is no forest at Kirui's Village today, there 
was within living memory, and within a mile of the swamp where I 
captured these specimens gallery forest connected up with the forest 
of Mount Elgon. The Kaimosi series were captured at night around 
the mill pond where they were assembling for breeding. 

Rana ansorgii Boulenger 

Rana ansorgii Boulenger, 1905, Ann. Mag. Nat. Hist. (7), 16, p. 107, pi. iv, 
fig. 1 : Between Benguela and Bihe, Angola. 

9 (M.C.Z. 14628) Behungi Escarpment, U. 5.iv.27. 



420 bulletin: museum of comparative zoology 

Corrigenda. I take this opportunity of amending the identification 
of this specimen collected by Dr. J. Bequaert and referred to mas- 
careniensis by Barbour & Loveridge (1930, p. 792). It constitutes the 
first record of the occurrence of this species in Uganda, though already 
taken in Tanganyika Territory {vide Loveridge, 1933, p. 371). 

Arthroleptides dutoiti Loveridge 

Arthroleptides dutoiti Loveridge, 1935, Bull. Mus. Comp. Zool., 79, p. 17: 
Koitobos River, Mount Elgon, Kenya Colony. 

Parasites. The type of this remarkable addition to the Kenya 
Fauna discovered by Mr. C. A. du Toit of Stellenbosch, held a female 
oxyuroid which Dr. J. H. Sandground considers is probably referable 
to the genus Aplectana. 

Arthroleptis stenodactylus lonnbergi Nieden 

Arthroleptis lonnbergi Nieden, 1915, Mitt. Zool. Mus. Berlin, 7, p. 361: Mombo, 

foot of Usambara Mountains, Tanganyika Territory. 
Arthroleptis stenodactylus uluguruensis Loveridge, 1932, Proc. Biol. Soc. 

Washington, 45, p. 61: Nyingwa, 7,000 to 8,000 feet, Uluguru Mountains, 

Tanganyika Territory. 

Synonymy. The Museum of Comparative Zoology having recently 
received by exchange from the Royal Swedish Museum one of the 
topotype series on which Nieden based his lonnbergi, I have compared 
it with the type of A. s. uluguruensis and find them conspecific. It 
may be recalled that this was suspected at the time uluguruensis was 
described, but in view of the fact that I was informed at the time that 
a cotype of the latter from Nyingwa was quite distinct from the type 
of lonnbergi, which was said to be identical with the type of stenodacty- 
lus, I accepted that opinion and synonymized lonnbergi with the 
typical form. This action has now to be reversed. 

Arthroleptis adolfi-friederici Nieden 

Arthroleptis adolfi-friederici Nieden, 1910, Sitz. Ges. naturf. Freunde Berlin, 
p. 440: Rugege Forest, Belgian Ruanda-Urundi. 

cf (M.C.Z. 20405) Mt. Mbololo, K.C. 21.iv.34. 

Distribution. Taken in the forest at 4,500 feet, this solitary speci- 
men constitutes the first record of the occurrence of the species in 
Kenya Colony. The find, however, is in conformity with other records 



loveridge: African amphibians 421 

of common Usambara Mountain species inhabiting these Taita Moun- 
tains. 

Arthroleptis minutus Boulenger 

Arthroleptis minutus Boulenger, 1895, Proc. Zool. Soc. London, p. 539, pi. 

xxx, fig. 4: Duro, southwestern Ethiopia. 
Arthroleptis scheffleri Nieden, 1910, Sitz. Ges. naturf. Freunde Berlin, p. 438: 

Nairobi and Kibwezi, Kenya Colony, etc. 
78 (M.C.Z. 20406-11) Mt. Debasien, U. 14.xi.33. 
130 (M.C.Z. 20430-9) Kaimosi, K.C. ii.34. 
3 (M.C.Z. 20412-4) Kikuyu, K.C. 14.iii.34. 
12 (M.C.Z. 20415-9) Kibwezi, K.C. 23.iii.34. 
20 (M.C.Z. 20420^) Mt. Mbololo, K.C. 17 & 19.iv.34. 
6 (M.C.Z. 20425-9) Peccatoni, K.C. 24.V.34. 

Distribution. Also taken at Eldoret by Mr. C. A. du Toit. 

Native name. Lungalla (Luragoli and Lutereki, for they believe 
these little frogs to be the young of a Rana). 

Variation. The Debasien locality is only some three hundred miles 
southwest of the type locality of minutus, the holotype of which 
measured 16 mm.; the Kibwezi series are topotypes of scheffleri of 
which the largest of 169 co types measured 20 mm. 

There are certainly two size groups in the material listed above, 
those from the first three localities have adult males of 14-15 mm. and 
females of 17 mm. Kikuyu, however, is only a few miles from Nairobi 
which is one of the type localities of scheffleri and in color and markings 
these Kikuyu frogs appear indistinguishable from the Kibwezi material. 
The Kibwezi-Mbololo frogs have males measuring 18-19 mm. and 
females of 20-22 mm. 

I have carefully measured up our minutus material from twenty-five 
localities in Uganda, Kenya and Tanganyika and it appears possible 
that the name minutus might be restricted to the smaller frogs in the 
northwestern part of the range (Duro, Gondokoro, Debasien, Entebbe, 
Kisumu, Kaimosi and Eldoret) while scheffleri could be applied in a 
subspecific sense to the frogs in the east and southwest. It seems 
advisable to await more material as there may be more overlapping 
than is shown by our series. 

It might be as well to state that the small northwestern frogs, 
though agreeing in size with the parvulus material reported on (Lover- 
idge, 1933, p. 386), have not the dark throat of the males nor so spotted 
an undersurface as the frogs from the southwestern highlands of Tan- 
ganyika. 



422 bulletin: museum of comparative zoology 

I am not entirely satisfied that the Peccatoni frogs are racially 
identical with the larger Kibwezi-Mbololo specimens. The Peccatoni 
frogs have well-defined digital disks on some toes though absent on 
others. I was struck in the field by their general pallor, pale brown bar- 
ring and white throats. The latter appears to be but an adaptation to 
their sandy habitat as is the case with albifer Ahl from Usaramo, also 
in the coastal belt. Our topotypic series of albifer, which is considered a 
synonym of minutus, show no tendency towards digital disks, however. 

Coloration in life. Noted of the Kaimosi series, that the males, 
though all smaller than the material referred to minutus in 1933, appear 
to have dark throats as if adult. In alcohol these are now only dusky. 

On Mount Mbololo, eleven males, of which one was embracing a 
female, had buff throats which were sometimes suffused with pink. 
That of the solitary female was only slightly buffy. 

Measurements. The largest and smallest specimens west to east. 
Mt. Debasien, largest cf 14 mm. 9 17 mm., smallest young 10 mm. 
Kaimosi, 

Kikuyu, " 

Kibwezi, 
Mt. Mbololo, 
Peccatoni, 

Breeding. While the largest females were gravid with spawn in all 
localities I do not think that they were ready for breeding at Debasien 
or Kaimosi. On March 19, on Mount Mbololo, ten males were calling, 
hopefully optimistic of the breaking of the rains; one male was embrac- 
ing a female. The Peccatoni frogs had undoubtedly assembled for 
breeding. 

Enemies. One of these frogs was recovered from the stomach of an 
Olive Water Snake (Natrix o. olivacea) at Kaimosi. 

Habitat. On Mount Debasien, from 5,000 to 7,000 feet, abundant 
among fallen leaves on sandbars in the drying water courses where 
shaded by trees. 

On Mount Mbololo, from 3,000 to 4,800 feet, on sandbars at edge of a 
stream at lower altitude, beneath a log at edge of forest at higher level. 

On damp mud at the north end of Lake Peccatoni, northeast of Witu. 



Phrynobatrachus keniensis Barbour & Loveridge 

Phrynobatrachus keniensis Barbour & Loveridge, 1928, Proc. New Eng. Zool. 
Club, 10, p. 89: Mount Kenya, Kenya Colony. 



c? 14 mm. 


9 


17 mm., 


a 


10 mm. 


c? 15 mm. 


9 


17 mm., 


a 


14 mm. 


cf 19 mm. 


9 


22 mm., 


a 


' 11 mm. 


o 71 18 mm. 


9 


20 mm., 


u 


15 mm. 


cf 16 mm. 


9 


18 mm., 


it 


14 mm. 



LOVE RIDGE: AFRICAN AMPHIBIANS 423 

3 (M.C.Z. 20474-6) Molo, K.C. 12.iii.34. 
2 (M.C.Z. 20477-8) Kikuyu, K.C. 14.iii.34. 

Distribution. Since this very Arthroleptis-like species was described, 
the Museum of Comparative Zoology has received the undermentioned 
material : 

7 (M.C.Z. 16114) Mt. Kinangop, K.C. (H.J.A.T.) 1930-1. 
1 (M.C.Z. 19870) Kikuyu, K.C. (C.A. du Toit) 3.ii.34. 
3 (M.C.Z. 19871-3) Uplands, K.C. (C.A. du T.) l.ii.34. 

The following remarks are based on all sixteen specimens listed. 

Affinities. As the apparently closely related Cacosternum b. boettgeri 
occurs with keniensis on the Kinangop Plateau, it might be noted 
that an easy way of separating them at sight is to be found in the 
immaculate chin and throat of boettgeri. In keniensis the chin, and 
usually most of the throat, is flecked or speckled with reddish brown 
like the belly. 

Coloration. One female (M.C.Z. 16114) exhibits a very broad, light, 
vertebral band such as one sometimes sees in P. acridoides. All the 
others are more or less uniformly dark, none exhibits a hairlike verte- 
bral line. 

Measurements. The largest c? (M.C.Z. 20474) measures 21 mm., 
the largest 9 (M.C.Z. 19870) 25 mm., thus a milimetre larger than 
the type. 

Breeding. On March 12 and 14, at Molo and Kikuyu respectively, 
females were swollen with spawn while the single male was distinguish- 
able by the swollen base of the first digit. 

Diet. Stomachs examined by Mr. Nathan Banks, held: (1) two 
simulid diptera, three acalypterid diptera; (2) two acalypterid diptera, 
a fulgorid homopteran, beetle; (3) acalypterid dipteron, dipterous 
larvae, beetle, formicid ant, a winged and two wingless ants; (4) two 
fulgorid homoptera, capsid hemipteron, three springtails (Collembola) ; 
(5) weevil, beetle larva, uropodid mite (preserved). 

Habitat. At Molo these frogs were heard calling an hour or two 
after sunrise but were located with great difficulty. They were secured 
eventually by parting the tufts of grass protruding from the water 
of a slow-flowing stream on the Londiani-Molo Road just below the 
Highland Hotel, a habitat which they shared with Rana wittei. At 
Kikiyu they were found in cattle-trampled marshland bordering a 
stream, an environment essentially similar to that at Molo. 



424 bulletin: museum of comparative zoology 

Phrynobatrachus graueri (Neiden) 

Arthroleptis graueri Nieden, 1910, Sitz. Ges. naturf. Freunde Berlin, p. 441: 
Rugege Forest, Belgian Ruanda-Urundi. 

12 (M.C.Z. 20440-4) Sipi, U. 12.xii.33. 
Tadpoles + 35 (M.C.Z. 20445-9) Butandiga, U. 8.L34. 
11 (M.C.Z. 20460-4) Kaimosi, K.C. 12.L34. 

Distribution. Also clearly seen at Buluganya; these three records 
of this rain-forest species on Mount Elgon constitute the first for its 
occurrence in Uganda. I (1929, p. 106) had already noted its presence 
at Kaimosi in western Kenya Colony. 

Measurements. The largest d" (M.C.Z. 20440) measures 22 mm., 
the largest 9 (M.C.Z. 20441) 28 mm., they are, however, equalled 
by frogs from Butandiga. 

Breeding. Males were calling constantly both morning and evening, 
and on cloudy days, in a swamp close to our camp at Butandiga. 
Tadpoles and diminutive frogs, the smallest measuring 8 mm., were 
taken in this locality. On January 12, a male was heard calling from 
a small stream at Buluganya; he disappeared into the mud when 
stalked. The smallest frog in the Kaimosi series measures 11 mm. 

Phrynobatrachus kinangopensis Angel 

Phrynobatrachus Kinangopensis Angel, 1924, Bull. Mus. Hist. Nat. Paris, 30, 
p. 130: Mount Kinangop, Aberdare Mountains, Kenya Colony. 

2 (M.C.Z. 20455-6) Mt. Kinangop, K.C. (H.J.A.T.) 1931. 
12 (M.C.Z. 20457-61) Kikuyu, K.C. 14.iii.34. 
1 (M.C.Z. 20462) Nairobi, K.C. 19.iii.34. 

Distribution. We are indebted to the generosity of Mr. H. J. Allen 
Turner for the topotypes. We have also recently received this species 
(M.C.Z. 19868) from Lari Forest Station, Uplands, K.C, through the 
kindness of Mr. C. A. du Toit. 

Variation. Secondary sexual characters of the four breeding males 
from Kikuyu consist of a strongly swollen base for the first digit as 
in Rana, and nuptial asperities on the dorsum, thighs and tibiae in 
marked contrast to the smooth-skinned females. 

Measurements. The largest cf (M.C.Z. 20457) measures 21 mm., 
the largest 9 (M.C.Z. 20458) 24 mm. 

Breeding. On March 14 and 19, at Kikuyu and Nairobi, males were as- 
sembled for breeding while the females were greatly distended with ova. 

Diet. Stomachs examined, held several species of small beetles and 
a number of what were apparently homopterous nymphs. 



loveridge: African amphibians 425 

Phrynobatrachus acridoides (Cope) 

Staurois acridoides Cope, 1867, Journ. Acad. Nat. Sci. Philad., 6, p. 198: 
Zanzibar. 

9 (M.C.Z. 20463) Kaimosi, K.C. ii.34. 
4 (M.C.Z. 20464) Tsavo, K.C. 6.iv.34. 
2 (M.C.Z. 20472-3) Peccatoni, K.C. 24.V.34. 
cf (M.C.Z. 20465) Witu, K.C. 30.V.34. 
4 (M.C.Z. 20466-9) Kau, K.C. 4.vi.34. 
c? (M.C.Z. 20470) Ngatana, K.C. 9.vi.34. 
9 (M.C.Z. 20471) Golbanti, K.C. 22.vi.34. 

Distribution. Also collected in Sokoki Forest by Mr. H. J. A. 
Turner. 

Correction. I have just reexamined the material (M.C.Z. 16864-7 
and 16869-85) from six localities in Tanganyika Territory which I 
(1933, p. 374) referred to acridoides. They all lack digital disks and 
are now considered to be the young of ?iatalensis. I therefore retract 
my view that digital disks in acridoides are largely a breeding season 
character. The rest of the material have disks and were correctly 
referred to acridoides. 

Affinities. One might have expected the Kaimosi frog to be re- 
ferable to perpalmatus, the Central African representative of acridoides, 
but this does not appear to be the case. Peracca has recorded acridoides 
from Toro, Uganda and Sternfeld from Lake Chad and Parker from 
Cameroon. 

Variation. Secondary sexual characters of the five males consist 
in the granular nature of the baggy skin on the throat. 

Measurements. The largest & & d 1 (M.C.Z. 20470, 20472-3) 
measure 23 mm., the largest 9 (M.C.Z. 20471) 28 mm. 

Breeding. In February at Kaimosi, April 6 at Tsavo, June 4 at Kau, 
and June 22 at Golbanti, females were distended with ova and in the 
last locality males were heard calling. 

In addition, it might be noted that three young, ranging from 6 to 
8 mm., in length, were captured at Tsavo. As no rain had fallen in 
this locality for about a year (according to local natives), these frogs 
were taken in a water furrow which was supplying the gardens from 
the tanks at the station. 

Diet. Stomach of the Golbanti frog held a beetle, four larvae of a 
tineid moth type, and four undigested skins of a larger species of 
caterpillar. 



426 bulletin: museum of comparative zoology 



Phrynobatrachus natalensis (Smith) 

Stenorhynchus natalensis Smith, 1849, Illus. Zool. S. Africa, Rept., Appendix, 
p. 24: Port Natal. 

38 (M.C.Z. 20479-85) Mt. Debasien, U. 14.xi.33. 

1 (M.C.Z. 20486) Butandiga, U. 10.i.34. 

3 (M.C.Z. 20487-9) Nairobi, K.C. 19.iii.34. 
46 (M.C.Z. 20490-9) Kibwezi, K.C. 23.iii.34. 

1 (M.C.Z. 20500) Tsavo, K.C. 5.iv.34. 

Variation. De Witte (1919, p. 221), the last reviser of this genus, 
revived ranoides Boulenger of Pietermaritzburg, Natal, with which 
he synonymized P. n. var. gracilis Andersson of the Sudan on the 
basis of limb length. The Kibwezi series listed above, being comprised 
of adults and young of all ages, reveals that the tibia which is included 
1^4 to 2 times in the length from snout to vent, is a juvenile character, 
while those of the adults are included 2 to 2% times; none of this 
Kibwezi series reach 23^ times. 

Measurements. The largest & & (M.C.Z. 20487, 20490) measure 
32 mm., the largest 9 (M.C.Z. 20491) 38 mm., both sexes surpassing 
my previous records. 

Breeding. On March 23, at Kibwezi, a dozen calling males were 
captured, as well as double that number of gravid females, of which 
several were taken in embrace. The Debasien and Tsavo specimens 
were all subadult, the largest measuring only about 26 mm. 

Diet. Stomachs examined, held: (1) termites; (2-6) empty, the 
frogs having been taken early in the evening. 

Habitat. On Mount Debasien between 5,000 and 7,000 feet as 
described under Arthroleptis minutus. The habits of the two species 
were very different when disturbed for the Phrynobatrachus immedi- 
ately sprang into the water, then swam back to the edge where they 
would remain floating with their snouts out of the pool. 

The Nairobi frogs were found resting on submerged grasses' close 
to the edge of a small stream. At Kibwezi, on weeds in the muddy 
shallows of a very large pond a couple of miles northeast of the 
station, yet close to the line. The Tsavo frog in a water furrow 
with P. acridoides. 

Hemisus marmoratum marmoratum (Peters) 

Enystoma marmoratum Peters, 1855, Arch. Naturg., 21, part 1, p. 58: Caba^eira, 
Mozambique. 



loveridge: African amphibians 427 

2 (M.C.Z. 20501-2) Voi, K.C. 7.iv.34. 
Eggs + 48 (M.C.Z. 20503-9) Lamu Id., K.C. 8-12.V.34. 
6 (M.C.Z. 20510-5) Ngatana, K.C. 17.vi.34. 
2 (M.C.Z. 20516-7) Gongoni, K.C. 27.vi.34. 
2 (M.C.Z. 20518-9) Malindi, K.C. 29.vi.34. 
4 (M.C.Z. 20520-4) Changamwe, K.C. 4.vii.34. 

Variation. The gigantic female from Voi mentioned below, falls 
into the H. m. guineensis section of my key (1933, p. 390), but is 
coloured like the typical form and is not referable to the large western 
race. 

Measurements. Both sexes are within the measurements given in 
my report (1933, p. 388) except a 9 (M.C.Z. 20501) measuring 37 
mm., i. e. 4 mm. longer than any specimen previously taken. 

Breeding. On May 8, thirty-nine males and six females were taken 
alive from a well on Lamu Island. Escape from the well was impos- 
sible and numbers were already drowned. They shared this predica- 
ment with Bvfo r. regularis and Rana o. oxyrhynchus. On May 12, 
a male was found clasping a female which was ensconced in a hole 
in the sand beneath the stem of a fallen coconut palm; nearby, in a 
precisely similar situation, another female was found covering a mass 
of freshly-laid white eggs. The rains had only just commenced and 
it seemed probable that the area would be flooded in the course of a 
week or two. On July 4, at Changamwe, a limbed tadpole, and a small 
frog with only the rudiment of a tail, were taken. 

Diet. Stomachs of three examined, held ants. 



BREVICIPITIDAE 

Hoplophryne rogersi Barbour & Loveridge 

Hoplophryne rogersi Barbour & Loveridge, 1928, Mem. Mus. Comp. Zool., 
60, p. 258, pi. ii, fig. 5: Mount Bomoli, Amani, Usambara Mountains, 
Tanganyika Territory. 

3(^29 (M.C.Z. 21346-50) Amani, T.T. (R.E.M.) vi.35. 

Sexual dimorphism. The female of this species was unknown until 
the capture of this material by Mr. R. E, Moreau. I take this oppor- 
tunity, therefore, of stating that the two gravid females differ from 
the males in lacking nuptial excrescences. As might be expected, 
there is no dagger-like spine on the site of the first digit, which is 
represented by a tubercle-like swelling comparable to the subarticular 



428 bulletin: museum of comparative zoology 

tubercles on the three normal digits. There are five digits on the hind 
foot, of these both the first and fifth lack digital expansions. 

Phrynomerus bifasciatus (Smith) 

Brachymerus bifasciatus A. Smith, 1849, Illus. Z06I. S. Africa, Rept., pi. lxiii: 
Country to the east and northeast of Cape Colony. 

8 (M.C.Z. 20524-8) Peccatoni, K.C. 24.V.34. 
11 (M.C.Z. 20529-33) Malindi, K.C. 29.vi.34. 
3 (M.C.Z. 20534-6) Changamwe, K.C. 4.vii.34. 

Affinities. It is somewhat difficult to decide as to whether this 
genus should be included in the Brevicipitidae or relegated to a 
separate family, the Phrynomeridae. Parker (1934, p. 3) has quite 
logically adopted the latter course on account of its possession of 
intercalary cartilages. De Villiers (1933, p. 275), and more recently 
Vos (1935, p. 261), however, stress the numerous characters common 
to Phrynomerus and true Brevicipitids which they think should 
offset the single difference. For the present I accept the conservative 
view and let it remain in the Brevicipitidae. 

Sexual dimorphism. The dark throat, which I supposed was charac- 
teristic of males, is present on all Peccatoni specimens. On dissecting 
the largest, however, it was found to be a gravid female; two others 
examined were males. 

Measurements. The largest d 71 (M.C.Z. 20524) measures 51 mm., 
the largest 9 (M.C.Z. 20534) 55 mm. 

Breeding. On May 24, at Peccatoni, and on July 4, at Changamwe, 
females were gravid. The Malindi frogs are all juvenile, ranging 
from 25 to 31 mm.; two of the Changamwe frogs are also young, 
measuring 22 and 27 mm. respectively. 

Diet. Stomachs of three examined, held ants with powerful jaws. 

Enemies. Two were recovered from the stomach of a night adder 
(Causus resimus) at Peccatoni. 

Defence. These aposematic frogs have a slightly fishy smell which 
disappears shortly after death. 

Habitat. On hearing the distinctive call of these frogs after a lapse 
of four years, it was immediately recognized, and I led my 'boys' 
half-a-mile across country to the swamped ground where they were 
assembling. Many of the Malindi series were taken from a hole be- 
neath a log; the adult from Changamwe was under a pile of palm-leaf 
debris; the young were up in bananas though not far from the ground. 



loveridge: African amphibians 429 



BIBLIOGRAPHY 

4 

Ahl, Ernst 

1931. "Anura III." in Das Tierreich, 55, pp. i-xvi + 1-462, figs. 1-320. 

Barbour, Thomas and Loveridge, Arthur 

1928. "A Comparative Study of the Herpetological Fauna of the Uluguru 
and Usambara Mountains, Tanganyika Territory, with Descrip- 
tions of new Species." Mem. Mus. Comp. Zool., 50, pp. 87-265, 
pis. i-iv. 

1930. "Reptiles and Amphibians from the Central African Lake Region." 
in Strong, "Report of the Harvard- African Expedition upon the 
African Republic of Liberia and the Belgian Congo." 2, pp. 786- 
796. 

BOULENGER, G. A. 

1882. "Catalogue of the Batrachia Salientia s. Ecaudata in the Collec- 
tion of the British Museum." Ed. 2. London, pp. i-xvi + 1-305, 
figs., pis. i-xxx. 

Dunn, E. R. 

1928. "Notes on Central American Caecilians." Proc. New Eng. Zool. 
Club, 10, pp. 71-76, pi. v. 

LoNNBERG, ElNAR 

1907. "Reptilia and Batrachia." in Sjostedt, "Wissenschaftliche Ergeb- 
nisse der Schwedischen Zoologischen Expedition nach dem Kili- 
mandjaro, dem Meru und den Umgebenden Massaisteppen 1905- 
1906." Upsala, pp. 1-28, pi. i. 

Loveridge, Arthur 

1933. "Reports on the Scientific Results of an Expedition to the South- 
western Highlands of Tanganyika Territory. VII. Herpetology." 
Bull. Mus. Comp. Zool., 74, pp. 197-416, pis. i-iii. 

Noble, G. K. 

1924. "Contributions to the Herpetology of the Belgian Congo based on 
the Collection of the American Museum Congo Expedition, 1909- 
1915. Part III. Amphibia." Bull. Am. Mus. Nat. Hist., 49, pp. 
147-347, figs. 1-8, pis. xxiii-xlii. 

Parker, H. W. 

1931. "A Collection of Frogs from Portuguese East Africa." Proc. Zool. 
Soc. London, 1930, pp. 897-905, fig., pi. i. 

1934. "A Monograph of the Frogs of the Family Microhylidae." London, 
pp. i-viii + 1-208, figs. 1-67. 



430 bulletin: museum of comparative zoology 

Schmidt, K. P. 

1936. "The Amphibians of the Pulitzer Angola Expedition. ^Ann. Car- 
negie Museum, 25, pp. 127-133. 

Villiers, C. G. S. de 

1933. "Breviceps and Probreviceps : comparison of the cranial osteology 
of two closely related Anuran genera." Anat. Ans., 75, pp. 257- 
276, figs. 1-7. 

Vos, C. M. DE 

1935. "Spelaeophryne and the bearing of its Cranial Anatomy on the 
monophyletic origin of the Ethiopian and Malagasy Microhylids." 
Anat. Anz., 80, pp. 241-265, figs. 1-5. 

WlTTE, F. G. DE 

1919. "Revision de Genre Phrynobatrachus Gunth. et description 
d'une espece nouvelle." Revue Zool. Afr., 6, pp. 220-228. 



EXPLANATION OF PLATES 



Loveridge — African Amphibians. 



PLATE 1 

Fig. 1. The Millpond at Kaimosi, Kakamega, Kenya Colony. 

After the breaking of the 'big rains' on March 3, 1934, this pond became 
the rendezvous of thousands of frogs. It was by treading on the sedges in 
the foreground that the smooth-clawed frogs (Xenopus Isevis victorianus) were 
induced to pop their heads out of the water. 

Fig. 2. A pair of Square-marked Toads (Bufo r. regularis). 

Those photographed were removed from a shallow pool, where numbers 
were in embrace, at Kaimosi, on February 10, 1934. A leech, which may be 
seen on the right, detached itself from the rump of the male and left behind 
a raw circular patch. Leeches were seen on other toads in this pool. 



BULL. MUS. COMP. ZOOL. 



LovERiDGE. African Amphibians. Plate 1 





* iff 



„»jp . jfe 










■35?* - *.>' 



PLATE 2 



Loveridge — African Amphibians. 



PLATE 2 

Fig. 1. Nests of Tree Frogs (Chiromantis xerampelina) at Witu. 

When ovipositing, the female frog exudes a quantity of albumen, which, 
by energetic trampling of the hind feet, she works into a mass of foam 
resembling a meringue. Its exterior hardens after a few hours exposure to 
the atmosphere. The eggs within hatch in due course, the resulting tadpoles 
feed upon the albumen and receive their oxygen from the bubbles of im- 
prisoned air. The central portion of the nest, containing the tadpoles, soon 
resolves into a fluid condition. When little but the outer crust remains, the 
'nest' ruptures horizontally, precipitating the tadpoles into the pond beneath. 
There they continue to develop and metamorphize. Photograph by Mr. R. 
D. Milne, May 31, 1934, on the estate near Witu, K. C. 

Fig. 2. Searching for tree frogs in Wild Bananas on Mount Elgon. 

In Tanganyika, wild bananas are a favourite haunt of the Polypedatid 
frogs (Leptopelis, Megalixalus, and Hyperolius). To reach them it is neces- 
sary to strip the outer leaves from off the stem. Our efforts on Mount Elgon 
proved fruitless, while at Kaimosi only Hyperolius picturatus was found. 
The photograph was taken in a ravine near Sipi at an altitude of about 7,000 
feet. 



BULL. MUS. COMP. ZOOL. 



Loveridge. African Amphibians. Plate 2. 





PLATE 3 



LovEniDGE — African Amphibians. 



PLATE 3 

Fig. 1. Forest-dwelling Leaf Frogs (Hyperolius rossii) at Kaimosi. 

In this locality, camp was pitched in a clearing in the rain forest. Twice 
on March 2, as I was working at a table outside my tent, leaves came rotating 
down to my feet from a great height. In the centre of each leaf, one of these 
beautiful little orange-spotted frogs was squatting. We removed them to a 
nearby shrub where they were photographed. 

Fig. 2. A terrestrial, burrowing Polypedatid (Kassina senegalensis) . 

The distinctive bubbling notes of these handsomely spotted, pale bronze 
frogs, might be heard a mile from the millpond at Kaimosi, where they 
assembled on March 3, following the breaking of the rains a day ortwo before- 
Some months later, though at the coast, many young ones were found 
beneath logs and vegetable debris in a cotton plantation. Caterpillars, ants, 
and many species of flies were found to furnish them with food. 



BULL. MUS. COMP. ZOOL 



Loveridge. African Amphibians. Plate 3. 





204 



- 



Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 

Vol. LXXIX, No. 8 



SCIENTIFIC RESULTS OF AN EXPEDITION 
TO RAIN FOREST REGIONS IN EASTERN AFRICA 

VIII 
OLIGOCHAETA 



By Dr. Wilhelm Michaelsen 
Hamburg, Germany 



Trie Li 1 : ry 
Museum of ( native Zool 

i 



CAMBRIDGE, MASS., U. S. A. 

PRINTED FOR THE MUSEUM 

October, 1937 



No. 8. — Reports on the 'Scientific Results of an Expedition to Rain 
Forest Regions in Eastern Africa 

VIII 

Oligochaeta 

By Wilhelm Michaelsen 

INTRODUCTION 

The following pages contain an account of the Oligochaeta obtained 
by Mr. Arthur Loveridge on his expedition to Uganda and Kenya 
Colony in the years 1933 and 1934. In addition there are a few worms 
collected by him in the Uluguru Mountains of Tanganyika Territory 
(Mandated German East Africa) during his journey of 1926-1927, 
which were not examined by my late colleague and friend, Mr. J. 
Stephenson, in his report (1933, pp. 225-247) on the material collected 
in 1929-1930. 

My thanks are due to Dr. Thomas Barbour, Director of the Museum 
of Comparative Zoology, for entrusting me with the examination of 
this oligochaetal collection, valuable as are all such from Tropical 
Africa, a region so productive of interesting species of Oligochaeta. 
The main collection, containing the types of the new species, is in the 
Museum of Comparative Zoology, Cambridge, Massachusetts; some 
cotypes are deposited in the Zoological Museum at Hamburg. 

Family ACANTHODRILIDAE 

Subfamily OCTOCHAETINAE 
Genus Dichogaster Beddard 

DlCHOGASTER BAGILOANA Sp. nov. 

One mature specimen, internally somewhat softened, from Bagilo (about 
6°50' south lat., 37°50' east long.), 6,000 feet, Uluguru Mountains, 
Tanganyika Territory. 

External Characters. Length 60 mm., diameter 4-4^ mm. Seg- 
ments about 130. 

Colour whitish, result of posthumous bleaching? 
Body cylindrical. 



434 



bulletin: museum of comparative zoology 



Head? Prostomium drawn in, obviously very small; a longitudinal 
dorsal furrow, presumably coming from the prostomium, divides 
segment I. 

Setae very slender, very strictly paired, all of them decidedly ventral 
in position. 

First dorsal pore at the intersegmental furrow V/VI. 

Clitellum annular, occupying segments XIII-XX ( = 8), being 
only feebly developed at segment XX which is as short as the normal 
segments following, whereas the proper clitellar segments are dis- 
tinctly longer. 




Fig. 1. Dichogaster bagiloana. Male sexual field. 



Male sexual field (fig. 1) medio-ventral, between the intersegmental 
furrow XVI/XVII and XVII/XVIII, biscuit-shaped, bilaterally 
symmetrical, distinctly longer than its greatest breadth, broadest in 
segments XVII and XIX, narrowed in segment XVIII, curved later- 
ally but with straight anterior and posterior borders. The bordering 
walls are moderately broad but progressively narrowing until reach- 
ing segment XVIII. The interior of the male field bears three pairs of 
lighter, nearly whitish, straight longitudinal walls, all rather close 
together, those of each pair uniting at the ends by a round connecting 
part. The median pair is slightly longer than either of the lateral ones, 
all of them extend far into segments XVII and XIX. Between the 
longitudinal walls of each pair is a sharp, straight, longitudinal furrow. 
I presume that the furrows of the lateral pairs are seminal furrows. 
The space between these walls and the outer wall bordering the male 
field in segments XVII and XIX is somewhat depressed, especially 
at the sides. Here it is somewhat broader, the lateral pairs of longi- 
tudinal walls being shorter than the median pair. I presume that the 
pores of the prostates — 2 pairs at segments XVII and at XIX — are 
situated in the deepest depressions in the 4 corners of the male field, 



michaelsen: African oligochaeta 435 

but I could not recognise them, neither could I recognise a connection 
between these depressions and the supposed seminal furrows. Equally 
uncertain is the position of the male pores which are to be sought in 
the interior of the male field, presumably in the seminal furrows at 
segment XVIII. 

Spermathecal pores indistinct, 2 pairs in the intersegmental furrows 
VII/VIII and VIII /IX in the lines of the median pairs of setae, in ab. 

Internal Anatomy. Septa VI/VII-IX/X very thin but IX /X 
slightly thickened in the centre, while X/XI-XIII/XIV are mod- 
erately thickened. 




Fig. 2. Dichogaster bagiloana. Prostata. 

Alimentary canal. Two gizzards in segments V and VI. 3 pairs of 
almost equally large, whitish chylous pouches in segments XV, XVI 
and XVII, apparently entirely separated from one another, kidney- 
shaped with relatively smooth surfaces, but their somewhat broad, 
convex ridge is crossed by many densely-grouped, darker stripes 
which partly take the shape of furrows. The intestine bears a simple, 
large, ribbon-shaped, irregularly meandering typhlosolis. 

Nephridia (mostly destroyed in the type) are very small, sac-shaped, 
and apparently very numerous (about 6 or more in each half of a 
segment ?). 

Anterior male organs holoandric. Two pairs of testicles apparently 
depend free into the body cavity of segments X and XI; apparently 
only one pair of seminal vesicles depend from septum 11/12 into seg- 
ment XII. They are small, much broader than long, multipartite, 
with a short and narrow stalk. 

Posterior male organs. Two pairs of nearly equally large prostates 
(fig. 2) in the body cavity of segments XVII and XIX. These seg- 
ments are slightly expanded dorsally. Glandular part moderately long, 
above the intestine, touching or nearly touching that of the other side, 



436 bulletin: museum of comparative zoology 

irregularly cylindrical, rather thick, especially at their ectal part 
(where they are as much as 1 mm. thick), irregularly meandering and 
bent, the folds of the undulations closely adpressed; glandular part 
smooth externally and without noticeable muscles, mainly composed of 
glandular cells, its axial lumen very narrow. I am not quite sure 
whether the lumen is provided with an epithelium, the cells of which 
are indistinct between the densely crowded fine ducts of the glandular 
cells discharging into the lumen. The duct of the prostate is sharply 
set off from the thick ectal end of the glandular part, rather short and 
uniformly very thin, (about 0.1 mm. thick in all its length, muscular 
with a distinct epithelium covering its narrow, axial channel. Each 
prostate is provided with a penial-seta-sac. Each sac contains a single 
penial seta which resembles in shape, though not in dimensions or orna- 
mentation, that of the allied species D. kigogoana Stephenson (1933, 
p. 233, fig. 6). Stephenson's drawing of the whole seta might well pass 
for that of my new species if we do not consider the matter of size. 
A penial seta of D. bagiloana is about 2.3 mm. long and about 15/* 
thick in the middle of the shaft (for D. kigogoana, 1.34 mm. long and 
15/x thick in the middle of the shaft), distinctly thickened at the ental 
end (here about 50/x thick). In general the seta is nearly straight, being 
only noticeably bent in its ental eighth, while its ectal end is bluntly 
tipped and slightly hooked. There are a few very faint undulations on 
the ectal half of the shaft (somewhat less distinct than in the figured 
seta of D. kigogoana, but in some penial setae of this species its author 
remarks that the undulations "may be almost indistinguishable"). 
I could recognise no ornamentation of the penial seta of D. bagiloana 
such as would appear to be characteristic of D. kigogoana; in the latter, 
however, the ornamentation is so scanty that its absence would not 
constitute an important systematic difference. 

Spermathecae entirely similar to those of D. kigogoana. In fact, 
Stephenson's drawing (1933, p. 233, fig. 5) of a spermatheca could pass 
for one of D. bagiloana, but on a closer examination of the latter no 
thecocystis was found. Ampulla rather small, globular or stoutly 
pyriform, thin-walled, clearly distinct from the middle portion. The 
latter is greater than the ampulla, thicker and somewhat longer, 
globular, or nearly so, with slightly stouter walls which are smooth on 
the outer surface but with some longitudinal projections into the 
lumen. The muscular duct of the spermatheca is about as long as the 
middle portion, not sharply set off from it, conical, narrowing towards 
its ectal end, with narrow and smooth axial canal and thick muscular 
wall. Its ental end does not project noticeably into the lumen of the 



michaelsen: African oligochaeta 437 

middle portion. A single, rather large, sub-globular diverticulum, 
containing a sperm ball, depends downwards beside the ental half of 
the duct, discharging into the basal part of the middle portion of the 
spermatheca through a short, narrow stalk. 

Remarks. D. bagiloana is so nearly allied to D. kigogoana Stephenson 
(1933, p. 232) from the Uzungwe Mountains, that I was at first un- 
certain whether or not to regard it as a subspecies of that worm. 
D. kigogoana is much smaller, about half as long and half as thick. 
The difference in the size of the body as well as in the size of certain 
organs cannot be considered as a juvenile character for the type 
specimen of the smaller worm is fully mature, and already sexually 
functional, the spermatheca containing sperm. The main differences 
are as follows: the clitellum is annular in the new species, saddle- 
shaped in Stephenson's worm. In my species the prostates have a 
rather large, very thick, meandering glandular part with a very nar- 
row lumen, whilst in D. kigogoana they are said to be rather small, 
tubular, and not coiled. In D. bagiloana the penial setae are very much 
larger and show no trace of the ornamentation which characterizes 
the smaller examples of D. kigogoana. A remarkable difference may be 
found in the configuration of the male sexual field which is very com- 
plicated in D. bagiloana, whilst of D. kigogoana Stephenson mentions 
only that "the seminal grooves are straight, bordered by faint lips." 
The complex configuration of the male field in D. bagiloana resembles 
the rather more complicated figure of the male field in D. ficta Michael- 
sen (1934, p. 26, fig. 14) from the Belgian Congo. The male field of 
this species also exhibits some longitudinal furrows, one of which lies 
in the median- ventral line; in other respects, however, the species from 
the Belgian Congo is very different from D. bagiloana. 



DlCHOGASTER ELGONENSIS Sp. nOV. 

Three somewhat softened specimens, from Kaburomi (about 1°15' north lat., 
34°30' east long.), 10,500 feet, western slope of Mount Elgon, Uganda. 
28.xii.33. 

External Characters. Length about 90 mm., diameter 2-3 mm. 
segments about 140. 

Colour dark, dirty gray (possibly not the original colour). 

Head pro-epilobous, with a minute, roundish dorsal appendage in- 
vading segment I. 

Setae slender, strictly paired, decidedly ventral in position, the 



438 



bulletin: museum of comparative zoology 



median-dorsal distance equalling about three-quarters of the cir- 
cumference of the body, the median-ventral distance is slightly 
greater than the middle lateral (aa 7 be, dd=%ix, approximately). 

Clitellum occupying segments XIII-^XXI ( = 8>£), inclined against 
the intersegmental furrow XII/XIII as well as against the middle zone 
of segment XXI, annular only at segment XIV, interrupted ventrally 
in the rest, the interruption being broader at segments XIII and XXI. 

Male sexual field (fig. 3) bilaterally symmetrical. Two pairs of 
prostate pores at segments XVII and XIX in the line of a pair of 
ventral setae, ab, those of either side united by a nearly straight, 



n 



7 



w 



a 



TA 



quiz 



Fig. 3. Dichogaster elgonensis. Male sexual field. 



longitudinal seminal furrow which is slightly bent at the extremities, 
convex against the median line. Each seminal furrow is closely flanked 
by a pair of rather narrow walls which unite after bending round the 
prostate pores. The terminal ends of such a pair of walls are continued 
forwards and backwards upon the neighbouring segment (XVI and 
XX) into a somewhat less prominent, not sharply bordered glandular 
field. 

Spermathecal pores indistinct, 2 pairs in the intersegmental furrow 
VII/VIII and VIII /IX in the line of the ventral pairs of setae ab. 

Internal Anatomy. Septa IX /X-XIII /XI V thickened, XI/XII and 
XII/XIII strongly, X/XI and XIII /XIV moderately, IX /X slightly, 
the remaining septa very thin. 

Alimentary canal. Two large gizzards in segments V and VI; 3 pairs 
of chylous pouches, broadly ridged in segments XV, XVI and XVII, 
relatively long, bean-shaped, somewhat irregularly incised at the 
convex borders, occupying the entire flanks of the oesophagous within 
the range of those three segments, those of a pair nearly meeting each 
other in the median-dorsal and median-ventral line of the oesophagus. 



michaelsen: African oligochaeta 



439 



Only the chylous pouches of the first pair in segment XV discharge 
separately by means of an individual, short and narrow, tubular 
stalk. The stalks of the posterior pouches in segment XVII are bent 
forwards into segment XVI, and unite with the stalks of the middle 
pair. The intestine bears a simple, smooth, ridge-like typhlosolis. 




Fig. 4. Dichogaster elgonensis. Ectal part of a penial seta, x 140. 

Michronephridia in the anteclitellian part of the body small and 
numerous, in the postclitellian portion more or less large, flattened, 
sac-shaped, rather constantly 4 in each half of a segment. 

Anterior male organs holoandric. Two pairs of testicles ventrally in 
the anterior parts of segments X and XI, enclosed in the narrow 
basal parts of thin-walled, pyriform, testis sacs which extend as far as 
the dorsal part of their segment and are crowded with masses of 



440 bulletin: museum of comparative zoology 

developing sperm. I failed to recognise seminal vesicles but would 
hesitate to assert that such organs were absent. 

Posterior male organs. Two pairs of equally large prostates in 
segments XVII and XIX. Glandular part cylindrical, about 0.25 mm. 
thick, very long, irregularly coiled, with smooth, yellowish-white sur- 
face; duct tubular, uniformly thick (about 0.1 mm.) throughout its 
length, moderately long, sharply set off from the glandular part, dis- 
charging in a simple manner through the prostatic pore. Penial setae 
(fig. 4) uniform, about 1.9 mm. long, at the ental end 4:5 n, in the centre 
25 fx and just before the ectal end 12 /z thick, very slightly and simply 
bent; the ectal half with the exception of the terminal part shows a 
slight undulation causing only about five moderately prominent 
waves at the profile-lines of the seta; in height these waves are equal to 
about a quarter of the width of the seta. The ectal tip of the penial 
seta is simple and moderately sharply pointed; ornamentation of this 
seta is restricted to its ectal half, differing in appearance in various 
parts of the seta, commencing with a small tract beneath the ectal 
tip of a relatively large, triangular thorn, rising out of the ental part of 
a not very distinct scar, and bent off from the seta so as to project con- 
siderably above the profile line of the seta. The subsequent ornamenta- 
tion is somewhat distant from the first mentioned as well as from one 
another and is somewhat irregularly arranged. Proceeding entalwards, 
the ornaments alter in appearance successively, the thorns become 
smaller, shorter and finally quite indistinct; meanwhile the scar, from 
which the thorn rises, becomes more distinct, longer and deeper in its 
now sharply bordered ental end; these scars devoid of thorns, or at 
least without noticeable thorns, are situated at the ectal slope of the 
above-mentioned undulations of the seta. 

Spermathecae (fig. 5) two pairs of equal size. Ampulla rather small, 
inverted pyriform, much shorter than the duct and middle part 
together; middle part sharply set off from the ampulla, about half as 
thick (fig. 5a), if not somewhat swollen (fig. 56), about as long as the 
muscular duct, from which it is not at all, or at least not sharply set 
off. A single pyriform diverticulum arises beside the middle part and 
the ectal half of the ampulla. The swollen ental half of the diverti- 
culum is directed upwards, parallel to the main part of the spermatheca 
against which it is inclined. It contains a single oval sperm compart- 
ment filled with spermatozoa. The ectal half of the diverticulum is a 
narrow tubular stalk of which the basal part is bent upwards and dis- 
charges rather high up in the middle portion of the spermatheca not 
very far from the entrance into the ampulla. This relation of the 



michaelsen: African oligochaeta 



441 



diverticulum to the other parts of the spermatheca cannot be seen 
directly, as the basal part of the diverticulum is covered by a tissue 
which slopes equally upwards against the sperm compartment of the 





Fig. 5. Dichogaster elgonensis. Spermathecae of two different specimens; 
a with contracted middle portion, 6 with swollen middle portion, the first only 
in outline, the latter made transparent. 

diverticulum and downwards against the muscular duct of the sper-. 
matheca; these organs and their relative positions can only be recog- 
nised properly in cleared preparations (fig. 56) and in slides of longi- 
tudinal sections. 



Dichogaster kaburomina sp. nov. 

Several well-preserved adult and immature specimens, from Kaburomi 
(about 1°15' north lat., 34°30' east long.), 10,500 feet, western slope of 
Mount Elgon, Uganda. 28.xii.33. 

External Characters. Length about 90 mm., diameter 4*^-5 mm., 
segments about 155. 

Colour light yellowish red, a little more intensive at the clitellum, 
with dusky mottling at the middle and hinder part of the body, the 
light nephridia and the dark contents of the intestine showing through 
the semi-transparent body wall. 

Body cylindrical only in its anterior part, slightly depressed behind 
the clitellum. 



442 bulletin: museum of comparative zoology 

Head epilobous (about ^£), lateral borders of the dorsal appendage 
converging posteriorly, uniting here so as to form a narrowly rounded 
hinder end of the appendage. Segments of the anterior portion of the 
body triannulate. 

Setae very slender, strictly paired, median-dorsal distance occupying 
about two-thirds of the circumference of the body (aa=bc; dd= 2 /z u, 
approximately), median-ventral distance a little greater than the 
middle lateral distances. 

Dorsal pores only distinguishable behind the clitellum. 

Clitellum annular, occupying segments XIV-XX (=7), occasionally, 
though not in all specimens, less developed ventrally. 

Male sexual field bilaterally symmetrical, biscuit-shaped, a little 
longer than its maximum breadth median-ventrally at segments 
XVII-XIX, laterally extending a little over the lines of setae b; 
its contour, broadest at segments XVII and XIX, smallest at segment 
XVIII; it is marked laterally, though sometimes indistinctly, by a 
rather flat and broad wall. The 2 pairs of prostate pores occupy the 4 
corners of the male field at segments XVII and XIX in the line of the 
innermost pairs of setae, ah. The seminal furrows connecting the two 
pores of each side, and close to the median side of the lateral walls, are 
deeply curved in a medial direction in the range of segment XVIII, 
being adapted to the curvature of the lateral walls. I failed to dis- 
tinguish the male pores. 

Spermathecal pores, 2 pairs at the intersegmental furrows VII/VIII 
and VIII /IX in the line of the setae ah. 

Internal Anatomy. Septa V/VI-IX/X very thin, X/XI-XIII/XIV 
somewhat thickened, XIV /XV slightly thickened, those following very 
thin. 

Alimentary canal. Two large gizzards in segments V and VI. 3 pairs 
of chylous pouches in segments XV, XVI and XVII, are rather large, 
approximately equal, longish bean-shaped, occupying the whole 
length of the oesophagus in these three segments. The intestine bears 
a simple, irregularly zigzag, ridge-like typhlosol. 

Mieronephridia in the anticlitellian segments small, irregularly 
scattered rather large in the postclitellian segments, sac-shaped, 
mostly with a rounded-quadrangular contour; usually 5 or 6, rarely 
7, mieronephridia in each half of a segment, 7 only when the lowest 
nephridium is replaced by two smaller ones. 

Anterior male organs holoandric. Testicles indistinguishable; 2 
pairs of testis sacs in segments X and XI, these large pear-shaped sacs 
extending far into the dorsal region of their segment, the broad, lower 



michaelsen: African oligochaeta 443 

parts of each pair united ventrally, containing a pair of large male 
funnels in the posterior part, and presumably the testicles in the an- 
terior part. The two male funnels of one segment are not distinctly 
separated from one another; 2 pairs of moderately large seminal 
vesicles depend from the septa X/XI and XI/XII into segments XI 
and XII, those of the anterior pair in segment XI enclosed in the 
testis sacs of the anterior pair, those of the posterior pair free in the 
body cavity of segment XII; these seminal vesicles are broader than 
long, grape-shaped, the somewhat small berries being closely crowded . 
Posterior male organs. Two pairs of prostates in segments XVII 
and XIX, laterally attached to the intestine; glandular part yellowish, 
very long, tubular, about 0.5 mm. in diameter, more or less broadly 
convoluted, the adjacent bends pressed against each other; duct very 
much shorter but relatively long, very thin throughout its length, 
about 0.1 mm. in diameter; each prostate accompanied by a penial- 
seta-sac, each of which contains some penial setae, one that was 
closely examined held 5. The setae of one penial sac are of very dif- 
ferent size and shape. At first I believed that this was a case of 
dimorphism such as is known in some species of this genus, a careful 
examination, however, convinced me that the difference depends upon 
the stages of development. In many mature specimens of D. kabu- 
romina penial setae were protruded but only one at each prostate pore 
and always that one of the largest size, not 2 dimorphic ones, such as 
may be found protruding side by side in species with true dimorphic 
setae. A fully grown penial seta of D. kaburomina (fig. 6a) is about 
1.6 mm. long, in the middle about 60 /j. thick, broadened at the ental 
end to about 85 n, and gradually diminishing towards the ectal end to 
a diameter of about 36 n a short distance beneath the ectal end, 
nearly straight in the middle, moderately and equally curved in the 
ectal fourth, slightly curved in the same direction at the extreme ectal 
end. The ectal end is moderately sharply and simply pointed. The 
ectal half of this seta shows a characteristic ornamentation consisting 
of some short, triangular teeth which are broad at their bases. These 
teeth occupy the ental end of small scars, and project distinctly above 
the profile lines of the seta; they are arranged irregularly and well 
separated from each other; in number there may be as many as 
10 or 12; towards the middle of the seta they become very small and 
indistinct. Besides this external ornamentation the setae show a 
certain interior structure consisting of a very delicate and dense an- 
nulation which in no way alters the smoothness of the surface of the 
profile lines of the setae (fig. 66). One of the smaller setae of the same 



444 



bulletin: museum of comparative zoology 



seta-sac (fig. 6c), the only one which I could examine in an uninjured 
state, is 0.95 mm. long and in the middle 35 n thick, straight in nearly 
all its length, only slightly bent at its ectal tip. It is quite smooth with- 
out any ornamentation but internally exhibiting the same annulation. 
Female organs. A pair of very large ovaries depend from the ventral 
margin of septum XII/XIII into segment XIII. 




'A 



C 



h 



u 




Fig. 6. Dichogaster kaburomina. Penial setae, a a fully developed one, x 60, 
b middle part of the same, x 120, c an unfinished one of the same bundle, x 60. 



Spermathecae (fig. 7). Ampulla rather small, pear-shaped or nearly 
cylindrical, with a thin smooth wall, ampulla connected with the 
middle portion of the spermatheca by a short and very narrow neck, 
it always seems to be bent aside; median portion is longer than the 
ampulla and either distinctly thicker than the latter, just as thick, or 
even a little thinner, according to whether it holds a thecacystis or 
not; wall of the middle portion moderately thick, mainly smooth, but 
at its ental part near the ampulla it is provided with some ring-shaped 



michaelsen: African oligochaeta 



445 



edges narrowing the lumen; the muscles of this middle portion are not 
noticeable; the muscular duct of the spermatheca is somewhat shorter 
than the middle portion and somewhat, if not much, thinner, mod- 
erately sharply set off from it with narrow, smooth, axial canal. 
A diverticulum, about half as long as the middle portion, discharges 
into the ectal part of the middle portion against which it is inclined. I 
have examined the diverticula of two specimens, only two of these eight 
diverticula (one from each specimen) are simple pear-shaped with a 




Fig. 7. Dichogaster kaburomina. Spermatheca made transparent. 



short narrow duct and containing a simple sperm ball; in five of the 
other diverticula the broad blind end was more or less deeply cleft, 
the contour of the diverticulum became heart-shaped. The sperm 
space also is paired, the two parts of it being separated from each 
other over a more or less lengthy tract, frequently for nearly their 
entire length. In one of the diverticula examined there is a third 
sperm space in addition to the two already mentioned. This, however, 
is not placed in the plane of the others but is situated somewhat lower 
and nearer to the common stalk of the diverticulum; in the 4 sper- 
matheca of one specimen which had a narrow middle portion, the 
latter were empty, in the spermathecae of another worm with a 
broader middle portion, the latter held a thecacystis with a large, 
oval, almost globular head and a short, narrow tail extending into 
the ectal part of the narrow axial canal of the muscular duct. 



440 bulletin: museum of comparative zoology 

Subfamily OCNERODRILINAE 

Genus Gordiodrilus 

Gordiodrilus wemanus sp. nov. 

Many well-preserved adults and young, from Wema (about 2°30' south lat., 
40° east long.), Ngatana district, Tana River, Kenya Colony. 

External Characters. Length 35^5 mm., diameter 1.0-1.2 mm., 
segments about 120. 

Colour uniformly brown, scarcely darker above than below. 

Body cylindrical with rather smooth surface; intersegmental fur- 
rows in general rather indistinct, but somewhat better denned on the 
anterior portion of the body. 




Fig. 8. Gordiodrilus wemanus. Dorsal view of the head. 

Head (fig. 8) epilobous (about %). Prostomium broad and rather 
short, uniformly rounded; dorsal appendage of the prostomium broad, 
narrowing posteriorly, with two narrow, transverse furrows near 
each other, and near the hinder edge of the appendage, the converging 
lateral borders of the appendage reaching only a very little distance 
(sometimes not at all ?) over the hindmost furrow. 

Setae rather strictly paired, the lateral perhaps a little more so than 
the ventral, the difference scarcely noticeable; median-ventral distance 
a little smaller than the middle lateral ones, median-dorsal distance in 
general equalling half the circumference of the body (approximately 
aa: ab: be: cd: dd:=8: 3: 9: 3: 37; dd=Y 2 v)\ towards the clitellar 
region the setae appear to be very slightly dislocated in a ventral 
direction, not so much so, however, as that the lateral ones could be 
called ventral. The setae are moderately large, the ventral very little 
larger than the lateral, but this difference is slight. 

Clitellum annular, occupying segments XIV-XIX ( = 6), but be- 
coming indistinct towards the intersegmental furrows XIII /XIV and 

XIX /xx. 



michaelsen: African oligochaeta 447 

Male sexual field nearly quadrate extending medio-ventrally 
between the intersegmental furrows XVI/XVII and XVIII /XIX, 
laterally extending very little over the lines of setae b, in general 
somewhat prominent, laterally distinctly bordered by a pair of straight, 
smooth, longitudinal walls which sometimes are marked by a lighter 
colour. 2 pairs of prostate pores in the middle zone of segments 
XVII and XVIII in the lines of setae b in the centre of more or less 
distinct circular porophores which occupy nearly the entire length 
of their segment. The prostate pores of each side are connected 
with one another by a straight, longitudinal, seminal furrow running 
exactly midway between the bordering lateral walls. At the points 
where the seminal furrows cross the intersegmental furrow XVII / 
XVIII, or very near this point, the male pores lie in the bottom of the 
furrows ; they are very indistinct and could be recognised only in very 
thin horizontal slides. 

Spermathecal pores, 2 pairs ventrally at the intersegmental furrows 
VII/VIII and VIII /IX, invisible exact in the lines of setae b. 

Internal Anatomy. Septa V/VI-X/XI thickened, X/XI slightly, 
V/VI and IX /X a little more, VI/VII-VIII/IX rather strongly. 

Alimentary canal. Oesophagus narrow, without a gizzard or any 
muscular thickenings, the masses of chromophil cells (pharyngeal 
glands) extend backwards to segment VIII; an unpaired chylous 
pouch depends ventrally from the oesophagus in the posterior part of 
segment IX. It is ovate, unstalked, only slightly narrowed at the 
base, composed of very numerous, parallel, thin, chylous tubes; a 
rather narrow axial lumen leads from the lumen of the oesophagus 
almost into the middle of the chylous pouch. In the anterior part of 
segment XIII the narrow oesophagus suddenly swells to form the wide 
intestine which has no typhlosole. 

Last heart in segment XL 

Nephridia beginning in segment V; in the anteclitellar segments the 
nephridia are very small and slender, in the middle and posterior part 
of the body they are extensively covered by large, opaque, and very 
granular cells. 

Anterior male organs. Two pairs of testicles depend free from the 
ventral border of septa IX /X and X /XI into the coelom of segments 
X and XL Opposite to them and anterior to septa X/XI and XI/XII 
are two pairs of rather small male funnels; the male ducts, not dis- 
tinctly seen, seem to be very tenuous, those of one side obviously are 
united before discharging through their male pore, their extreme ectal 
end, in the body wall above the male pore, is seen in horizontal sec- 



448 bulletin: museum of comparative zoology 

tions as a very narrow, quite simple tube. 2 pairs of seminal vesicles 
depend from septum IX /X into segment IX and from septum XI/XII 
into segment XII; those of the anterior pair are rather large and 
multiple, consisting of a few irregularly pyriform sacs, those of the 
posterior pair in segment XII are rather small, normal sacs. 

Posterior male organs. Two pairs of very long, slender prostates 
occupy the ventral parts of some segments from XVII backwards. 
The glandular part of the prostates is loosely coiled, occupying the 
ventral part of two or three segments ; it is about 0.07 mm. thick and 
consists mainly of a thick glandular epithelium; the axial lumen is a 
very narrow channel. There are no remarkable muscles at this part; 
ectally the glandular part gradually narrows to the muscular duct 
which is irregularly bent, rather short and about 0.04 mm. in diameter 
at the centre; it becomes still narrower where it enters the body wall 
and discharges in a quite normal manner at the top of its porophore. 

Spermathecae (fig. 9). Ampulla small, elongate oval, or inverted 
pyriform, its walls smooth and moderately thick, duct rather sharply 
set off from the ampulla, twice to thrice as long. In the ental two 
thirds it is about half as thick as the ampulla, in the ectal third it 
gradually narrows towards its ectal end. The lumen is a simple, nar- 
row, straight channel only in the narrower ectal third of the duct, in 
the ental two-thirds of the duct is narrowly meandering or forms ir- 
regular spiral turns ; it is invested by a moderately thick, but not quite 
uniformly thick epithelium which follows the meanderings and con- 
volutions of the lumen in a lesser degree; on its exterior side the 
epithelium is covered by a moderately thick muscular layer which is 
smooth on the outside, this again bears on the outside a layer of very 
small, nearly globular glandular cells, a layer of unequal thickness, in 
places scarcely noticeable, especially thin at the ectal end of the duct, 
elsewhere moderately thick, imparting a certain roughness to the sur- 
face of the duct. There are no diverticula leading off from the lumen 
of the duct. 

Remarks. G. wemanus belongs to the group G. zanzibaricus Beddard 
(1894, p. 252), G. habessinicus Michaelsen (1913, p.5,pl. ii, figs. 30-31) 
and G. basin Stephenson (1928, p. 1, text fig.) characterised by the 
situation of the prostatic pores — 2 pairs at segments XVII and XVIII. 
G. zanzibaricus differs from all the others by the situation of the setae, 
which "are all of them decidedly ventral in position," whilst in the 
others the median-dorsal interval equals half the circumference of the 
body, it is not even a little less, as in G. habessinicus. The rest of the 
description of G. zanzibaricus, especially in respect to the structure of 



michaelsen: African oligochaeta 



449 



the spermathecae, is so incomplete that one cannot state what is its 
relationship to other species. G. wemanus agrees with G. habessinicus 
in having the median-ventral interval between the setae smaller than 
the middle lateral ones, in G. baski it is greater than the latter, in G. 
zanzibaricus equalling it. A character found only in G. wemanus is 
the possession of a second pair of seminal vesicles depending from sep- 




Fig. 9. Gordiodrilus wemanus. Spermatheca, made transparent. 



turn IX /X into segment IX; in the other species of the group there' is 
only one pair depending from septum XI/XII into segment XII. As 
for the structure of the spermathecae, G. wemanus differs from G. 
baski and G. habessinicus in the absence of diverticula leading off from 
the lumen of the duct. 

It is possible that G. elegans Beddard (1892, p. 84), with prostatic 
pores at segments XVIII and XIX, should be placed near to G. zanzi- 
baricus group, as it is the only other Gordiodrilus provided with diver- 
ticula in the spermathecal duct. 



450 bulletin: museum of comparative zoology 

Family EUDRILIDAE 

Genus Bettonia 
Bettonia monticola sp. now 

Two adult and an immature specimen, from Kaburomi (about 1°15' north 

lat., 34°30' east long.), 10,500 feet, western slope of Mount Elgon, Uganda. 

28.xii.33. 
Paratypes from Amaler River (about 1°50' north lat., 34°40' east long.), 

5,000 feet, western slope of Mount Debasien, Karamojo, Uganda. 14.xi. 

1933. 

External Characters. Length of adults 73-85 mm., diameters 3/4-4 
and 4-4^ mm -> 1 15 and 145 segments. (Contrasted with B. lagariensis 
Beddard, (see Remarks) which was 90 mm. long with a diameter of 
from 4-5 mm., it is slightly smaller). 

Colour bluish violet dorsally, the remainder yellowish gray. 

Body cylindrical anteriorly and somewhat depressed in the middle 
and posterior portion. 

Head epilobous (about ]^). 

Setae in general moderately large; those of segment XII and of some 
preceding it, especially the ventral ones, somewhat enlarged, the 
ventral setae of segments XIII, XIV and XVII smaller than the nor- 
mal ones. In general the median-ventral distance is only slightly larger 
than that between the setae of the ventral pair and about as large as 
the middle lateral distances, the dorsal pairs are about 3 /5 as wide as 
the ventral ones, the median-dorsal distance seems to be slightly 
smaller than half the circumference of the body (impossible to measure 
accurately as the body wall is somewhat shrivelled into longitudinal 
folds); against segment XVIII the ventral setae are distinctly dis- 
located medially at the expense of the middle lateral distances, so 
much so that at segment XVIII the median-ventral distance equals 
the width of the ventral pairs, and is smaller than half the median 
lateral distances (in general aa: ab: be: ed=8: 6: 8: 4; at segment 
XVIII aa: ab: be: cd=4: 4: 9: 3; M=Y^iX). Of B. lagariensis Beddard 
only remarks that "the setae are wider apart in the case of the ventral 
couples than in the case of the lateral." He does not state that the 
setae of the ventral couples are remarkably distant from one another, 
and in fig. 36 of B. lagariensis at segment XVI they are indeed shown 
as hardly half as far from one another as the median-ventral distance. 
More anteriorly the ventral couples are even narrower, at segment VI 
for instance ab is hardly 1 /3 as large as aa, as the setae are, for the most 



MICHAELSEN : AFRICAN OLIGOCHAETA 451 

part, rather large in this region of the body, I dare not assume that 
their very characteristic arrangement in B. monticola could have es- 
caped notice in B. lagariensis. 

Clitellum apparently annular but ventrally showing a somewhat 
different glandular modification. It occupies segments XIV-XVIII 
but is only weakly developed in segment XVIII. 

Secondary male pore unpaired, situated medio-ventrally at inter- 
segmental furrow XVII/XVIII, a great hole with notched margin 
expanding between the lines of setae a, and with a rounded triangular 
contour, the anterior angle extending over the posterior part of seg- 
ment XVII. 

Spermathecal pores, 1 pair at the intersegmental furrow XII/XIII, 
their middle part just in the lines of setae h. 

External organs of puberty : A lighter coloured, slightly prominent, 
not sharply edged, glandular cushion around the male pore, more dis- 
tinct in its hinder semicircular part at segment XVIII than in its 
rounded square anterior part at segment XVII ; a somewhat more prom- 
inent, transverse cushion ventrally on the forepart of segment XIII 
behind the spermathecal pores. 

Internal Anatomy. Septa VIII/IX-X/XI greatly thickened, VII / 
VIII and XI/XII moderately, VI/VII and XII/XIII slightly thick- 
ened, those following, as well as septum V/VI, very thin. 

Alimentary canal. A large barrel-shaped gizzard in segment V, 
3 oval, unpaired, short-stemmed chylous pouches, tubule pouches, 
depend ventrally from the oesophagus in the posterior parts of seg- 
ments IX, X and XI, a pair of lateral chylous pouches which are 
broadly ridged, discharge into the oesophagus in segment XIII. 
Intestine without a typhlosole. 

Anterior male organs holoandric. Testicles unrecognizable. The 
sperm reservoirs are thin tubes closely and irregularly coiled, empty 
and apparently collapsed in the specimen examined; two pairs of large 
broadly sac-shaped multilocular seminal vesicles depend from the 
septa X/XI and XI/XII into segments XI and XII. 

Posterior male organs. Glandular part of the euprostates thickly 
tubular, about 4^ mm. long and 0.9 mm. thick, the ectal two-thirds 
running straight backwards, the ental third bent forward, closely 
attached to the ectal part. Glandular part smooth externally, covered 
by a thin muscular mantle about 50 ij. thick; glandular epithelium ir- 
regularly thickened, |in some places forming ridges as thick as 250/z, 
in other parts much thinner, in consequence the lumen is corres- 
pondingly irregular, tending to form zigzag bends which are, in general, 



452 bulletin: museum of comparative zoology 

rather narrow. At its ectal end the glandular part of the euprostate is 
continued into a short muscular duct, about 0.45 mm. in diameter, 
which enters the hind pole of a large, oval, cushion-like, copulatory 
pouch about 2^2 mm. long and 2 mm. broad. The longer axis of the 
two pouches converge against the median secondary male pore through 
which they discharge, their anterior poles being united. The wall of a 
copulatory pouch is very muscular, irregularly thick, the moderately 
wide lumen being narrowed by large ridges and walls. At the hinder 
pole of the copulatory pouch; a moderately large and conical penis 
projects into the lumen; this penis has a narrow, smooth, axial canal 
which is a continuation of the euprostate canal and discharges through 
one of the primary male pores at the top of the penis. The vasa 
deferentia coming from the anterior male organs enter side by side 
the euprostate at the ental end of its muscular duct, then bend and 
turn ectalwards to the muscular layer of the euprostate duct and 
penis. Here the two vasa deferentia of the corresponding side of the 
body obviously unite, for in the length of the penis I could recognise 
only one narrow channel near the axial channel of the penis (from the 
external view of the posterior male organs of B. lagariensis as figured 
by Beddard (1903, fig. 37) we may assume that their internal structure 
is essentially similar to that of B. monticola) ; in the act of copulation 
doubtless the two copulatory pouches will be pushed out with the 
result that the two extended penises will diverge from each other suf- 
ficiently to enable them to enter the two spermathecal pores so far 
distant from one another. 

Female organs and spermathecae in general symmetrically paired, 
the two connected with each other by a supraoesophagial coelomic 
tube. The apparatus of one side exhibits the following structure. 
The spermathecal pore leads into a stoutly pyriform, though slightly 
narrowed in its ectal portion, spermatheca with a rather narrow, ir- 
regular lumen and a very thick muscular wall. The spermatheca lies 
not quite freely in the body cavity of segment XIII, at least its 
broader ental part is enveloped by a very delicate membrane, which, 
I suppose, is part of an ovarian bladder which in the meantime en- 
velops the female funnel. (I could not recognise this envelopment 
at the basal part of the spermatheca). The lumen of the spermatheca 
seems to end blindly, no definite opening into the ovarian bladder 
being recognisable, but in the thick wall of the ental pole of the 
spermatheca irregular narrow fissures were detected some of which 
certainly opened into what is presumably the ovarian bladder. (The 
passage of the spermatozoa from the spermatheca into the female 



michaelsen: afkican oligochaeta 453 

apparatus obviously resembles that recognised by Michaelsen (1905, 
p. 343) and by Cognetti (1910, p. 4 and fig.) in certain species of 
the genus Pareudrilus) . A very slender, club-shaped, thin-based 
ovarium arises from the hinder side of the ventral part of septum 
XII/XIII slightly above its ventral margin and somewhat medially 
from the base of the spermatheca. It extends parallel with, and as far 
as the ental pole of the adjacent spermatheca; it is closely enveloped by 
a thin membrane, forming the tubular part of an ovarian bladder. 
This bladder, on reaching the ental end of the spermatheca, bends 
sideways inclining closely against the ental pole of the spermatheca 
with whose membranous covering it apparently coalesces (I was un- 
able definitely to determine the relationships of these delicate mem- 
branes). Meanwhile the ovarian bladder, if I recognised it rightly, 
widens and envelops the female funnel also, the latter being closely 
attached to the spermatheca; there is then given off from the lateral 
portion of the ovarian bladder, a thin tubular continuation which, 
rising beside the oesophagus to which it is closely attached, enters into 
the dorsal part of segment XIII; on reaching the median-dorsal line of 
the oesophagus, this tube unites with that arising on the other side, 
the two together forming an unpaired coelomic tube which, encircling 
the oesophagus, connects both the female apparatuses. At the median- 
dorsal middle of this coelomic tube, it gives rise to a very small, blind 
sac which extends backwards for a short distance beside the dorsal 
vessel to form a small, unpaired coelomic sac; the pyriform, closed 
female funnel is rather closely attached to the ental part of the sper- 
matheca and is apparently enclosed in the ovarian bladder; its broad, 
median pole appears to open by a narrow slit into the ovarian bladder. 
Its lumen is rather narrow and not quite simple. A narrow channel 
leaves its median part obliquely in a latero-posterior direction into the 
narrow and short stalk of a rather large, kidney-shaped egg sac whose 
base covers the posterior part of the female funnel. The narrower, 
lateral pole of the female funnel is continued into a slender female duct, 
which first proceeds sideways then bends backwards to reach the female 
pore at the intersegmental furrow XIV /XV. At the ental end of the 
female duct approximately at the ectal end of the female funnel, arises 
a moderately large, pyriform appendage which contains a moderately 
large, apparently not quite simple sperm chamber (Respecting B. 
lagariensis, Beddard's solitary noteworthy detail regarding these 
organs is that the muscular, basal part of the spermatheca is enveloped 
in a coelomic sac). 

Remarks. It is with some hesitation that I describe these worms 



454 bulletin: museum of comparative zoology 

as a new species, for in some respects they show a remarkably close 
resemblance to Beddard's B. lagariensis (1903, p. 213, figs. 36-37) from 
Lagari, Mau District, south of El Burgon Range, Kenya Colony, a 
locality about 150 miles southeast of the type locality of monticola. 
It might be better to consider my specimens as representing a race of 
lagariensis, but unfortunately the description of B. lagariensis is so 
incomplete that one cannot tell whether great differences exist in the 
organs not mentioned by Beddard. In any event there is a considerable 
difference in the arrangement of the setae between my specimens of 
monticola and those of lagariensis as shown in Beddard's fig. 36, which 
I dare not ascribe to errors of the English draftsman. In the foregoing 
description of monticola, other important differences between the two 
worms are noted in addition to omissions in Beddard's description. 

B. monticola differs from the nearly related B. adol phi-friderici 
Michaelsen (1910, p. 62) principally by the structure of posterior male 
organs and of the female apparatus. In B. adolphi-friderici the copula- 
tory pouches are distinctly separated, each discharging by means of a 
proper duct through the common secondary male pore, the glandular 
part of the euprostate is irregularly bent and does not form a single 
loop as in all other known species of this genus. The spermatheca in 
B. adolphi-friderici is shorter and broader than in B. monticola, all 
the female organs of the former are situated nearer to the base of the 
spermatheca than in the latter, while the ovarium with its ovarian 
bladder has quite a different shape to that of the new species. 

B. budduensis Michaelsen (1910, p. 66) differs from all other known 
species of Bettonia in having the spermathecal pores situated in the 
lines of setae b. 

Genus Polytoreutus 

Polytoreutus loveridgei sp. nov. 

Two specimens, of which the larger is designated as the type, from Golbanti 
(2°27' south lat., 40°7' east long.) on the Tana River, Kenya Colony. 
22.vi.1934. 

External Characters. Length of type 140 mm., diameter throughout 
about 43^ mm., segments about 250. Length of paratype 90 mm., 
diameter 2^-4, segments about 175. Both specimens seem to be 
entire. 

Colour of type in general smokey gray, of the paratype dark gray 
but chamois or approximately brown at the clitellum. 



michaelsen: African oligochaeta 455 

Body cylindrical. 

Head prolobic, segments more or less distinctly triannulate. 

Setae rather slender, especially anteriorly; at midbody aa= 1}^ ab = 
\}/2'- bc = ca 5 cd; dd= x /2\i.. 

Nephridiopores in the lines of setae cd. 

Clitellum at segments XIV-XVII, annular; intersegmental furrow 
somewhat more tenuous though still distinct at the clitellum; nephri- 
diopores rather more distinct at this point. 

Secondary male pore situated medio-ventrally at intersegmental 
furrow XVII/XVIII at the summit of a transversely oval cushion 
which extends as far as the middle zones of segments XVII and XVIII. 
In the type a stump-like, cylindrical penis, somewhat longer than thick, 
and bearing the primary male pore at its summit, projects from the 
secondary male pore; in the paratype the penis is retracted so far that 
its primary male pore is on a level with the surface of the porophore. 

Spermathecal pore concealed in a very slender transverse slit medio- 
ventrally at intersegmental furrow XVIII /XIX. 

Internal Anatomy. In view of the fact that the anatomical character- 
istics of the anterior end are quite uniform in this genus, and in an 
endeavour to avoid mutilating the type, I have not dissected the region 
of the head, but confine my observations to the following. Septum 
XII /XIII is the last one somewhat thickened ; a pair of stout, kidney- 
shaped, chylous pouches (ridged pouches with accretions of calcium 
carbonate) depend from the oesophagus in segment XIII. Intestine 
without a typhlosole. 

Anterior male organs. A pair of seminal vesicles extend backwards 
through many segments posterior to septum XI/XII. Throughout 
their length they are closely attached to the alimentary tract, in the 
type as far as segments XXVI and XXVII, in the paratype as far as 
XXVII. In the type these seminal vesicles are closely attached to 
one another throughout their length except anteriorly; in the paratype 
they are separated except for some short tracts. Where separated, the 
vesicles have the shape of flattened rosaries, being swollen in the seg- 
mental areas and constricted sharply at the intersegments. Each 
individual bead of this rosary represents a flattened oval, being some- 
what shorter than broad. In the type where the two seminal vesicles 
are closely attached to one another the swellings of the joints are only 
developed on the free side, at such a spot the double organ has the 
appearance of a flattened rosary divided longitudinally by a median 
cut. 



456 bulletin: museum of comparative zoology 

Posterior male organs (fig. 10). A pair of euprostates pass from the 
medio-ventral male pore first sideways, then upwards and backwards 
through 3 or more (till 7) segments. Their glandular, ental part (eu) 
is sausage-shaped, about 11 mm. in length and at the ectal end ca 0.9 
mm. thick, entalwards slightly diminishing in thickness, simply and 
more or less irregularly bent, closely attached to the intestine. Its 
wall is rather thick, consisting principally of long, slender, glandular 
cells placed vertically against the outer surface which is provided 
sparsely with muscle fibres ; lumen moderately wide, somewhat reduced 
by 4 or 5 irregular, longitudinal ridges which project into it. In a 




Fig. 10. Polytoreutus loveridgei. Ectal part of an euprostata with the 
ectal end of the relative vasa deferentia and the muscular euprostata duct. 

transverse section its contour is very irregularly stellate. The ectal end 
of the glandular part is modified in a very characteristic manner; in 
the organ viewed "in toto" it looks like a somewhat flattened bulb 
which is nearly twice as thick as the unmodified euprostate tube from 
which it is sharply distinguished, at least posteriorly, less distinctly 
or not at all anteriorly. As viewed in a series of sections it appears 
somewhat different in the two specimens. In the cotype (fig. 10) this 
bulb appears as a somewhat widened continuation of the euprostatic 
tube which is curved to form a short, narrow, S-shaped, double loop 
only marked externally by a very slight furrow; walls and ridges of the 
main portion of the euprostate tube are continued into this modified 
terminal part where their regular longitudinal arrangement gives place 
to a very irregular one. The vas deferens (vd) reaches the bulb on the 
anterior side of its first or ental turn, its axial channel, piercing the 
wall of the bulb in a straight vertical line, discharges into the lumen 
in a quite simple manner. At the ectal pole of the ectal turn of the 
double loop arises the short, narrow, euprostate duct (eud) which is 



michaelsen: African oligochaeta 



457 



almost entirely embedded in the body wall. In this region the internal 
surface of the rather compact and muscular body wall is quite plain 
without any internal thickenings, no copulatory pouch is formed. At 
this point the euprostate duct unites with its partner from the other 
side to form a single, unpaired, penial duct which contracts into some 
narrow, short loops and terminates in a short, free, conical penis sur- 
rounded by a narrow lumen. The rather compact male porophore 




Fig. 11. Polytoreutus loveridgei. Female organs and spermatheca. 



forms, in its exterior part, a small penis pouch that is almost entirely 
filled by the penis. The primary male pore at the top of the penis is 
just level with the surface of the male porophore. In the type, however, 
the ectal part of the euprostatic apparatus has a somewhat different 
appearance: the ectal tract of the glandular part looks more like a 
simple, nearly elliptical bladder, the S-shaped double loop of the cotype 
is rendered indistinct, presumably by inflation, its lumen is rather wide 
at this spot and completely filled with fine granular secretions while 
the corresponding part of the presumably contracted paratype is 
quite empty. 

Female organs and spermathecae (figs. 11-12). I could not recognise 
ovaries or ovarian bladders. The spermathecal pore at intersegmental 
furrow XVIII /XIX leads into an unpaired flattened tube (sp) which 
is anteriorly very narrow, then broadens to about 0.65 mm., is closely 
attached to the body wall and leads straight forwards to bifurcate 
at segment XVI. The two branches (sb) which are about as broad as 



458 



bulletin: museum of comparative zoology 



the unpaired posterior tube are at first closely attached to one another 
but later diverge widely. On reaching segment XIII the two branches 
bend sideways, upwards, and finally mediad, embracing the alimentary 
canal in the region of the paired chylous pouches ; on the dorsal side of 
the oesophagus they meet without coalescing; the branches are 
stoutly club-shaped. The whole spermatheca and its branches have 




Fig. 12. Polytoreutus loveridgei. Female apparatus of the right side, a and 
b seen from different sides. 



a rather thin wall and wide lumen; this lumen is filled with fine and 
evenly granular secretions, coloured light red in haematoxylin-eosin. 
In these secretions many slender, cylindrical, irregularly bent or 
curved, rather long (some fragments being about 0.3 mm. long), in 
places densely crowded spermozeugmas are embedded. Their struc- 
tureless, light red axial portion is 18 /jl thick, totally covered by the 
spermatozoan head-ends which are dark violet, almost black, and form 
a dense layer of about 6 /i thickness; the entirely pale caudal ends of 
the spermatozoa project freely into the mass of secretions. The indis- 



michaelsen: African oligochaeta 459 

tinct female pores each lead into a slender, straight oviduct (od) which, 
passing mediad, enter an irregularly pyriform, rolled or closed female 
funnel (ff) whose narrow lumen is continued into three channels. One 
of these channels leads immediately into a large sperm chamber (sch) 
which occupies the strongly swollen, medial pole of the female funnel. 
The lumen of this sperm chamber is very irregular, being reduced by 
large ridges and walls projecting into it from the inner surface; the 
latter is festooned all over with the darkly coloured anterior ends of 
spermatozoa whose pale caudal ends project into the lumen. Another 
channel leads from the central lumen of the female funnel in a medio- 
dorso-ventral direction into the very short and narrow stalk of an 
irregularly kidney-shaped egg sac (es). This sac closely covers the 
whole dorsal side of the female funnel. The third channel leads from 
the central lumen of the female funnel into a long, slender connecting 
duct (cd). At the outset this duct forms two short and narrow loops 
which are closely pressed against the anterior side of the female funnel. 
Further on it narrows a little, passes freely mediad, and then joins the 
posterior wall of the spermathecal branch of its side. Finally becoming 
very thin, so that its axial lumen is hardly visible, it enters the sper- 
mathecal branch a short distance above its first lateral bend. 

Remarks. This new species is closely related to P. baralypton Cog- 
netti (1911, p. 507, figs. Aa and B), presumably from Nairobi, Kenya 
Colony, but differs from it in the following more or less important 
points. The seminal-vesicles (sperm-sacs) of P. baralypton are short, 
extending only into segments XII or XVIII, and each terminates 
posteriorly in a tubular appendix. In P. loveridgei these organs show 
large, moniliform swellings and extend as far as segments XXVI or 
XXVII. The question arises as to whether these organs are fully 
developed in the type of P. baralypton or whether their shape results 
from reduction with consequent systematic value. 

The ectal part of the glandular tract of the euprostates is not modi- 
fied in P. baralypton as in P. loveridgei, for in the former it is not dis- 
tinctly set off from the euprostate middle tube, at most it is only 
slightly widened. 

The female funnel and its appendages differ markedly in the two 
species; in the description of P. baralypton nothing is said of a sperm 
chamber, which is so distinct in P. loveridgei; in P. baralypton the con- 
necting duct bears a large globular appendix near its medial end while 
I have been unable to find any trace of such an appendix in any one 
of the four females whose organs I have examined. I must confess 
that I do not altogether understand the author's description and figure. 



460 bulletin: museum of comparative zoology 

Is that portion which he designates an egg capsule (fig. B.e.c.) homol- 
ogous with the organ which I call ovarian bladder (Eitriehterblase), 
or is it part of the female funnel? 

In my opinion the most striking differences between these two 
related species is to be seen in the shape of the medial part of the sper- 
matheca, and in the presence or absence of a globular appendix at the 
connecting duct. The shape of the main portion of the spermatheca 
is narrow and only bifurcating about its middle in P. loveridgei whereas 
it is broadly swollen in segments XIV, XV and XVI, and only bifur- 
cates in front of the base of the diverticula in P. baralypton. 

POLYTOREUTUS MALINDINUS sp. nOV. 

One rather soft specimen, from Malindi (3°13' south lat., 40°8' east lat.), 
50 feet, Kenya Colony. 30.vi.1934. 

External Characters. Length 58 mm., diameter \%-2% mm., seg- 
ments about 145. 

Colour yellowish gray to yellowish brown. 

Setae rather widely paired below, strictly paired laterally (approxi- 
mately aa: ab: be: cd=8. 6. 8. 3), dorsal distance approximately equal- 
ling half the circumference of the body. 

Clitellum occupying segments XIV-XVII ( = 4), apparently saddle- 
shaped, or possibly annular, though modified ventrally in a different 
manner and pale glandular, while laterally and dorsally yellowish 
brown. 

Copulatory pores (fig. 13) unpaired, medio-ventrally. Male pore 
marked by a broadly triangular depression whose base is medio- 
ventrally placed at intersegmental furrow XVII/XVIII; the blunt 
anterior angle meets a small, slender transverse furrow somewhat 
behind the middle zone of segment XVII. I was unable to determine 
whether the male pore was actually situated in this transverse furrow 
or in intersegmental furrow XVII/XVIII, though I imagine that the 
latter is correct; the depression of the male pore is at segment XVII 
surrounded by a somewhat lighter, not sharp edged, scarcely prominent 
glandular modification; spermathecal pore rather indistinct, in a 
transverse furrow behind the middle zone of segment XVIII, appar- 
ently corresponding to the similar furrow at segment XVII, if not in 
the intersegmental furrow XVIII /XIX. A lighter, not sharp edged, 
scarcely prominent glandular modification surrounds the transverse 
furrow, reaching backwards as far as intersegmental furrow XVIII / 
XIX. 



michaelsen: African oligochaeta 



461 



Female pores indistinct, laterally at segment XIV or XVI. 

External organs of puberty (fig. 13) very characteristic: A large, 
transversely oval glandular cushion situated medio-ventrally at seg- 
ment XIX of which it occupies the whole length, a second medio- 
ventrally placed cushion of different appearance at segment XXIV, not 
only occupying the whole length of this segment but encroaching on 
intersegmental furrows XXIII /XXIV and XXIV /XXV and almost 
reaching the centres of segments XXIII and XXV; in segment XXIV 




Fig. 13. Polytoreutus malindinus. Ventral view of the fore-end, schemati- 
cally. 



this cushion expands laterally to form a roundish projection on either 
side; the middle part of the cushion is flattened, if not slightly sunk, and 
bears a narrow, medio-ventral, transverse depression with spindle- 
shaped contour slightly anterior to the centre of segment XXIV. In 
addition to these postclitellar organs of puberty are two rather indis- 
tinct intraclitellar ones, two broad, medio-ventral cushions occupying 
the whole length and lower part of segments XV and XVI. These 
intraclitellian organs have somewhat the appearance of blunt-edged 
thickenings of the body wall. 

Internal Anatomy. In order to mutilate the unique type as little 
as possible I have avoided dissecting the anterior end. Septum X/XI 
(as is presumably the case with some of those preceding it) is moder- 



462 bulletin: museum of comparative zoology 

ately strongly thickened, septa Xl/XIIand XII/XIII slightly thick- 
ened, those following are thin. 

Alimentary canal. The last unpaired, ventral, chylous pouch is in 
segment XI, it is tubular and stoutly ovoid with fairly numerous 
chylous tubules; a pair of elongate, broadly ridged, chylous pouches, 
together nearly encircling the oesophagus, present in segment XIII. 

Anterior male organs: A pair of very broad testis sacs, in whose basal 
part are the testicles while the rest is filled with developing masses of 
sperms, occupy the entire length of segment XI, slightly narrowing at 
the base and somewhat more just before septum XI/XII where each 
continues into a very thin, tubular seminal vesicle of equal content. 
These tubes extend backwards through many segments and broaden 
to form the seminal vesicles proper; after reaching the posterior end 
of the long euprostates, the seminal vesicles proper extend into segment 
XXXVIII. The sperm reservoirs are moderately stout, cylindrical, 
but not smooth being irregularly crenulated and incised by the septa. 
The sperm reservoir in the posterior part of segment XI appears to be 
thickly tubular with rather stout walls and an irregularly meandering 
lumen. There the ental end narrows and then enters the posterior part 
of the appropriate testis sac, after which it immediately widens to 
form the male funnel. 

Posterior male organs somewhat severely damaged in sectioning as 
poorly fixed and consequently soft. Glandular part of the euprostates 
apparently rather long and moderately stout ; there does not appear to 
be a common copulatory pouch, certainly not a conspicuous large one. 

Female organs and spermathecal apparatus (fig. 14): The main part 
of the spermatheca is a simple, smooth, moderately long, about 0.25 
mm. diameter tube (sp) which is loosely attached to the body wall in 
the medio-ventral line and extends backwards for a short distance 
behind the spermathecal pore region (this posterior part was damaged 
in preparation). This median spermathecal tube bears 6 or 7 pairs of 
pyriform diverticula (dv) whose rather narrow stems are distinctly 
differentiated from the median tube ; their main portion is rather large, 
bent upwards, loosely attached to the alimentary canal which is almost 
completely encircled by each pair and almost extending to the dorsal 
vessel. It is possible, though improbable, that a posterior seventh pair of 
diverticula may have been destroyed. The wide lumen of the thin-walled 
diverticula contains more or less thick masses of densely and irregu- 
larly coiled, filaceous spermoneugmas while the median tube, as well 
as the anterior branches, is empty in the present specimen; the fore- 
part of the median spermathecal tube in segment XIII is regularly 



michaelsen: African oligochaeta 



463 



rounded, and gives rise to a pair of sharply set off, narrow stemmed 
anterior branches (sb) whose main part is irregularly cylindrical, 
usually about 0.35 mm. diameter. The stem, rather sharply set off 
from the main portion is moderately long and about 0.06 mm. diameter. 
The female apparatus lies posterior to the ental ends of the anterior 
branches of the spermatheca. No trace of ovaries or ovarian-bladders; 
the closed female funnels (ff) are regularly pyrif orm ; their broad me- 
dian pole is continued into a distinct connecting duct (cd) which enters 




Fig. 14. Polytoreutus malindinus. Female apparatus of the right side, and 
adjacent parts of the spermatheca, the hinder part of the latter, as well as the 
diverticles and the ectal part of the spermathecal branches being cut off. 



the stem of the appropriate anterior branch of the spermatheca. The 
connecting ducts are about 0.5 mm. diameter throughout their length, 
moderately long, irregularly bent. The lumen of the female funnels is 
rather wide but not quite simple, somewhat conchoid. I failed to 
recognise any opening of any ovarian bladder into the body cavity of 
segment XIII. It gives rise to a narrow channel extending backwards 
into the short, narrow stem of a large, kidney-shaped egg sac (es) 
which lies just behind the female funnel and partly covers it. The thin 
lateral poles of the female funnels are continued each into a long, very 
slender (approximately 0.05 mm. diameter) oviduct. A pyriform 
sperm chamber (sr), with an apparently not quite simple lumen, is 
situated in the wall of the female funnel, or near its transition into the 
oviduct, into which the laterally directed, axial lumen of the sperm 



464 bulletin: museum of comparative zoology 

chamber discharges the sperm chamber does not project above the 
general, smooth surface of the female funnel. 

Remarks. P. malindinus belongs to the group of P. violaceus Beddard 
(1894, p. 230 and Michaelsen, 1897, p. 51); for further comments see 
under P. bagiloanus sp. n. on page 472, following. 



POLYTOREUTUS ASKARORUM sp. nOV. 

One specimen, somewhat softened and with shrivelled cuticle, from Bagilo 
(about 6°50' south lat., 37°50' east long.), 6,000 feet, Uluguru Mountains, 
Tanganyika Territory. 20.X.1926. 

External Characters. Length 140 mm., average diameter 43^-5 mm., 
anterior to the clitellum swollen to a diameter of about 6 mm., seg- 
ments about 220, but this number is very inexact. 

Colour in general yellowish, light ashy gray dorsally on the anterior 
part of the body. 

Setae slender, not easy to distinguish in the shrivelled cuticle, doubt- 
less arranged as in the other species of this genus. 

Clitellum at segments XIV-XX (or XXI ?), indistinctly developed 
at segment XIV (and also XXI ?), anteriorly annular. In the clitellar 
region the intersegmental furrows are shallower, but not entirely 
eliminated. 

Male pore (fig. 15) situated medio-ventrally at intersegmental fur- 
row XVII/XVIII in the centre of a circular, whitish, glandular field 
which reaches nearly as far as intersegmental furrow XVI /X VI I and 
XVIII /XIX ; this male field does not project distinctly above the level 
of the body wall and can scarcely be called a male porophore. 

Spermathecal pore indistinctly medio-ventrally at intersegmental 
furrow XVIII /XIX. 

External organs of puberty (fig. 15) very characteristic, but rather 
indistinct in the softened and shrivelled type. In the centre of the 
anticlitellar part of the body there is a medio-ventral, regularly 
elliptical copulatory field, nearly three times as long as broad, just 
beginning at intersegmental furrow VI /VII and extending backwards 
as far as furrow XI/XII, or possibly even beyond segment XII. In 
this present specimen the copulatory field is marked by a narrow, 
though rather deep, furrow at its border line, but apparently not by a 
glandular modification of the body wall. Intersegmental furrows are 
not obliterated in this area; at its posterior end the copulatory field 
appears to be open, the furrows bordering it are apparently continued 



michaelsen: African oligochaeta 



465 



over segment XII in lines parallel to the axis of the body. Obviously 
there are postclitellar organs of puberty correlated with this ante- 
clitellar organ, but they are rather indistinct in this specimen, in fact 
I did not recognise them until I specially examined the related region 
of the body as far behind the copulatory pores (the male and sper- 
mathecal pores) as the copulatory field lies before them. Then I 
recognised a series of medio-ventral, transversely-oval patches at seg- 
ments XXIII, XXIV and XXV, even possibly at segments XX and 




Fig. 15. Polytoreutus askarorum. Ventral view of the fore-end schematically. 

XXVI though scarcely recognisable. These patches are nearly as 
broad as the anteclitellar copulatory field, i.e. somewhat broader than 
long, occupying the entire length of their segment. They are distin- 
guished only by a somewhat lighter colour and perhaps by a scarcely 
noticeable thickening of the body wall. 

Internal Anatomy. Septa V/VI-X/XI very much thickened, XI / 
XII moderately thick, those following thin. 

Alimentary canal. A rather small, elongate gizzard in segment V, 
3 unpaired, chylous pouches depend ventrally from the oesophagus 
in segments IX-XI; paired chylous pouches, unrecognisable as the 
alimentary canal, is injured in this region. 

Anterior male organs metandric. A pair of testicles depend from 
the ventral margin of septum X /XI into segment XI, each is enclosed 



466 bulletin: museum of comparative zoology 

in a broad, cylindrical, testis sac which, arising obliquely, passes 
to the dorsal part of septum XI/XII; presumably these sacs pene- 
trate this septum to be continued beyond as a pair of slender, tubular 
seminal vesicles. Actually, I did not see these tubular portions which 
are presumably collapsed, but I found, far behind the region of the 
sexual pores, two moderately long, stout, tubular seminal vesicles, 
one closely attached to the intestine, the other to the body wall, 
irregularly meandering through 5 or 6 segments. They are densely 
packed with spermogems, but I could not recognise any connection 
with the male organs of segment XI; doubtless the connecting tubes 
were destroyed. A pair of tubular sperm reservoirs lie in segment XI 
close in front of septum XI/XII; their ental part is thin, coiled, and 
partly (?) enclosed in the posterior part of the testis sacs, into which 
they doubtless open by means of their funnels ; their ectal part, thick, 
and with a moderately wide lumen, descends in a straight line, and 
narrowing, pierces septum XI/XII, each being continued as a vas 
deferens. 

Posterior male organs asymmetrically developed; only a single 
euprostate, that of the right side persisting; it is long, cylindrical 
though somewhat depressed, white, forming a double loop, attached to 
the intestine almost throughout its length, its rounded ental end 
lying in segment XXVI from where it passes forwards as far as into 
segment XIV, then turns backwards as far as segment XX, then 
finally forwards again; in segment XVIII it narrows and turns mediad 
enters the body wall, and apparently discharges in a simple manner 
through the median male pore at intersegmental furrow XVII /XVIII. 
The euprostate seems to be chiefly glandular without sign of muscular 
tissue either in its main portion or at its narrowed, ectal end, i.e. its 
duct. It is questionable whether the asymmetry of the euprostatic 
apparatus is normal, or should be regarded as abnormal. 

Female organs and spermatheca (fig. 16) : I could not detect ovaries, 
ovarian bladders, or ovarian channels ; perhaps they were already mac- 
erated in the present specimen. The main part of the spermatheca (sp) is 
a moderately thick, unpaired median tube, closely attached to the body 
wall, extending from segment XIII into segment XX, where it ends in 
a small, nearly globular knob, thickest at segments XVII and XVIII. 
The underside of the organ, when dissected from the body wall, shows 
furrows corresponding to the intersegmental septa; these furrows are 
not continued over the upper side of the main spermathecal tube; the 
indistinct opening of the spermatheca lies a short distance in front of 
the posterior end. The forepart of the median spermathecal tube 



michaelsen: African oligochaeta 



467 



shows no bifurcation, but the sides of its terminal tract in segment 
XIII are continued into a pair of transverse branches (sb) which are 
just as thick as the median part and not distinguished from it by any 
constriction. These cylindrical spermathecal branches have a rather 
long, rounded free end, bent upwards, and almost encircling the 
oesophagus in segment XIII. Between these branches and the terminal 
knob of the spermathecal tube in segment XX, 6 pairs of lateral 



—cUy 




Fig. 16. Polytoreutus askarorum. Spermatheca and female organs. 



diverticula (dv) arise out of the median tube, one pair in each of 
the six segments from XIV to XIX inclusive. These diverticula have 
nearly the same shape and size as the anterior branches of the main 
spermathecal tube, from which they are not sharply distinct; at the 
medial end they are not in the least narrowed, actually being so 
stout as to be in contact with one another. Like the spermathecal 
branches they are bent upwards, encircling the alimentary canal ; they 
are not shaped as regularly as those branches, however; in places they 
are swollen, especially the 2 hinder pairs, or reduced while frequently 
the free end is bent off. The female apparatus is closely attached to 
the anterior branches of the spermatheca; to the posterior side of the 



468 bulletin: museum of comparative zoology 

branches — a little distance to the side of where they spring from the 
median tube and rather far from the free ends of the branches — a 
closed female funnel is attached to each side. This funnel (fa), with a 
partly wide and partly narrow lumen, describes a double loop in its 
narrower part, bears a stout kidney-shaped, short-stemmed egg-sac at 
its upper side. Medially it is continued into a narrow, moderately 
long, connecting tube, which, becoming even more slender, enters the 
branch of the spermatheca near the point of its origin from the median 
spermathecal tube. The lateral pole of the female funnel is continued 
into a slender female duct ; at the point of junction a pyrif orm sperm 
chamber discharges through a normal lumen which is closely draped 
by the dark-coloured, anterior ends of spermatozoa. This sperm 
chamber is almost wholly embedded in the stout wall of the female 
funnel, projecting but slightly over the outer surface of the funnel. 

Remarks. P. askororum is closely related to P. violaceus Beddard 
(1894, p. 230) and Michaelsen (1897, p. 51), P. malindinus sp. n. 
described above, and P. bagiloanus sp. n. following. For further com- 
ments see Remarks under this last species. 

POLYTOREUTUS BAGILOANUS Sp. nOV. 

Two well preserved, adult specimens, from Bagilo (about 6°50' south lat., 
37°50' east long.), 6,000 feet, Uluguru Mountains, Tanganyika Territory. 
20.ix.1926. 

External Characters. Length 50 and 60 mm., diameter about 2 mm., 
segments about 72 to 82. 

Colour whitish, apparently not pigmented. 




Fig. 17. Polytoreutus bagiloanus. Dorsal view of the head. 

Head (fig. 17) tanylobous, if not prolobous; prostomium calotte- 
shaped; segment I is crossed by some longitudinal furrows, of these 
two, one on either side of the medio-dorsal line, appear to be somewhat 
sharper than the others, forming lateral borders of a dorsal appendage 
of the prostomium, making the head tanylobous. At their anterior end, 
however, is a slender transverse furrow looking like the posterior 



michaelsen: African oligochaeta 469 

border of the prostomium and making the head prolobous, if we do not 
admit as an appendage of the prostomium the two sharper furrows 
already mentioned. 

Setae moderately large, medio-ventral distance only slightly 
greater than the distance between the setae of a ventral pair, ap- 
proximately as large as the medio-lateral distances; the distance 
between the setae of a dorsal pair about half as large as the medio- 
ventral distance; medio-dorsal distance approximately equals half 
the circumference of the body (approximately aa: ab: be: cd=Q: 
5: 6: 3; dd=c&. % n). 

Clitellum chief annular, occupying the ventral portions of segments 
XIV-XVI ( = 3), dorsally segments l/n XIII-l/n XVII ( = 3 2/n), 
inclined against intersegmental furrow XII/XIII and XVII/XVIII. 

Male pore distinct, marked by a brownish spot, medio-ventrally at 
intersegmental furrow XVII/XVIII in the centre of a transversely 
oval, almost circular, whitish, glandular field. 

Female pores indistinct, laterally near, or in, intersegmental furrow 
XIV /XV. 

Spermathecal pore indistinct, medio-ventrally at intersegmental 
furrow XVIII /XIX. 

External accessory organs of puberty similar in both specimens, 3 
medio-ventral, transversely oval, nearly circular glandular cushions 
occupying the whole length of their segment, 1 postclitellar very 
prominent cushion at segment XXII and 2 clitellar cushions, some- 
what less prominent though at the same time somewhat larger as the 
real clitellar segment is longer than the ordinary segments, at seg- 
ments XIV and XV. 

Internal Anatomy. Septa V/VI and VI/VII very thin, VII/VIII- 
XI/XII moderately though distinctly thickened, IX/X-X/XI 
stoutest, XI /XII only slightly thickened. 

Alimentary canal. A moderately large, cylindrical gizzard in 
segment V; 3 unpaired, chylous pouches depend ventrally, by a short 
and slender stem, from the oesophagus in segments IX-XI they are 
ovate, tubule pouches without a central Jumen; a pair of lateral, 
chylous pouches ridged pouches, in segment XIII. 

Anterior male organs. Testicles not recognised; a pair of testis 
sacs, in segment XI, rise from a narrower base in the antero-ventral 
part of their segment; they successively accrue, their broader part 
resting against the stout sperm reservoir in the posterior part of the 
segment where a tenuous seminal tube proceeds from them. This, 
going backwards through many segments, forms the anterior, tubular 



470 



bulletin: museum of comparative zoology 



part of a seminal vesicle and extends to segment XXXI just behind 
the posterior end of the euprostates. The tubular part of the seminal 
vesicles suddenly broadens to form the seminal vesicles proper which 
extends backwards through many segments, in a carefully examined 
specimen as far as segment LVIII. A true seminal vesicle is a broad, 
flat, irregularly lobed band, closely attached to the dorsal side of the 
intestine, and deeply grooved at the intersegments. A pair of large, 
ovate sperm reservoirs are situated ventrally in the posterior part of 




Fig. 18. Polytoreutus bagiloanus. Ventral view of the fore-end. 



segment XI; out of their upper pole arises a sharply distinct, very 
narrow, short tube which bends sharply downwards and soon after 
enters the anterior end of the tubular portion of the seminal vesicles 
where it doubtless ends in a small male funnel (not distinctly seen). 
At the lower pole of the sperm reservoirs a narrow tube arises in a 
similar manner and proceeding backwards forms a single male duct. 

Posterior male organs. The euprostates have a long, tubular, 
colourless glandular part extending backwards into segment XXX. 
Its muscular coat is scanty and irregular, in places scarcely recognis- 
able. Its lumen is moderately wide, irregularly constricted, reduced 
by some longitudinal walls composed of more or less thick glandular 
epithelium. The structure of the glandular part is similar to that of 
P. violaceus Beddard (1894, pi. xvi, fig. 3), but the muscular coat of 



michaelsen: African oligochaeta 



471 



the latter seems to be thicker and more regular. The anterior ends of 
the glandular part are bent abruptly towards the middle and, without 
diminishing in diameter, unite in a short, transverse, common median 
part which, though of a similar diameter, differs by its glistening 
muscles. The male ducts enter the ectal end of the glandular part in an 
apparently normal, direct manner (not quite clearly seen). 




Fig. 19. Polytoreutus bagiloanus. Spermatheca and female organs. 

Female organs and spermatheca (fig. 19): The main portion of the 
spermatheca is a median, smooth tube (sp) about 6 mm. long and 0.6 
mm. broad, closely attached to the ventral body wall and reaching 
from segment XV into segment XX. A short distance before its 
posterior end it discharges through a minute conical duct. It bears 6 
pairs of lateral diverticula (dv) which are arranged transversely as a 
rule. These diverticula are pyriform or elongatedly oval, somewhat 
narrowed at their ectal end, and not densely crowded but distinctly 
separated from one another, mostly somewhat bent upwards but by no 
means encircling the alimentary canal as they are far too short. The 



472 bulletin: museum of comparative zoology 

diverticula of the posterior pair arise close beside each other from 
the posterior end of the common tube and are bent backwards. The 
anterior part of the main tube, reaches septum XIV/XV, and divides 
to form two branches (sb) which are about half as stout as the main 
tube. These branches diverge gradually at first, then more distinctly, 
bending sideways and upwards as they somewhat increase in size until 
they terminate in a globose blind end. The female apparatus (fa) is 
attached to the posterior side of the central and more basal part of the 
anterior branches of the spermatheca. The closed pyriform, female 
funnel bears posteriorly a short and slender stemmed, large, sub- 
globular egg-sac. Laterally the narrower pole of the female funnel is 
continued into a slender oviduct, the wall of the ental end of the ovi- 
duct, as well as of the anterior side of the female funnel, contains some 
pyriform sperm chambers whose broad, blind ends project slightly 
beyond the outer surface of the organ. In part these sperm chambers 
seem to discharge through a common duct. Out of the broad medial 
pole of the female funnel arises a rather long, moderately and equally 
stout throughout its length, connecting tube which forms two wide 
loops before becoming attached to a spermathecal branch. Descending 
beside the latter, it finally enters the spermathecal branch at a point 
only moderately distant from the beginning of the median, unpaired, 
main portion of the spermatheca. I could recognise neither the 
ovaries nor ovarian bladders. 

Remarks. P. bagiloanus, as well as P. malindinus and P. askarorum 
described above, together with P. violaceus Beddard (1894, p. 230, pi. 
xvi, figs. 3 and 7; Michaelsen, 1897, p. 51) and its variety variabilis 
Michaelsen (1897, p. 51) form a group of closely related species. As 
far as external characters are concerned these species may be readily 
distinguished by the number, arrangement, and shape of the external 
organs of puberty. P. violaceus and its variety have only a single 
postclitellar cushion. P. malindinus has two postclitellar cushions of 
different shapes, while the remaining species each have three glandular 
cushions similarly shaped, P. askarorum only postclitellar ones, 
P. bagiloanus a single postclitellar and two intraclitellar ones. As for 
the internal organization, the four species are best characterised by 
the shape of the spermatheca, whose differences are easily seen in the 
corresponding figures (Beddard, 1894, pi. xvi, fig. 7 for P. violaceus; 
and this paper, fig. 14 for malindinus, fig. 16 for askarorum, fig. 19 
for bagiloanus). 



michaelsen: African oligochaeta 473 



Polytoreutus minutus Michaelsen 

Polytoreutus minutus Michaelsen, 1912, Arkiv. Zool., 7, No. 32, p. 2: Kenya 

district, Kenya Colony. 
One specimen in good condition, from Mount Mbololo (about 3°20' south lat., 

38°30' east long.), 4,800 feet, Teita, Kenya Colony. 13-18.iv.1934. 

Locality. This additional material is especially valuable on account 
of its exact locality data, which was vague in the type. While Kenya 
Province covered a huge area extending from Thika in the south to 
the Northern Guaso Nyiro in the north, I am uncertain of the limits 
implied by Kenya district in 1911, presumably a small area around 
the mountain in Kenya Province, Kenya Colony. 

Remarks. The specimen listed above is slightly larger than the 
type, being 42 mm. long and from 1.5 to 1.65 mm. in diameter with 
about 135 segments, whereas the type was only 32 mm. long and had 
119 segments. 

The male pore in the type "ziemlich gross, augenformig" is marked 
in the Mbololo example by a rather deep, moderately broad hole 
which is inclined forwards against the middle zone of segment XVII. 
It is situated in the centre of an oval, whitish, glandular field which is 
somewhat longer than broad, and occupies the entire length of seg- 
ments XVII-XVIII, even encroaching slightly on segment XVI. 



Polytoreutus chaloneri Smith and Green 

Polytoreutus chaloneri Fr. Smith and B. Green, 1919 (1920), Proc. U. S. Nat. 

Mus., 55, p. 156, figs. 10-12: Mkonumbi, near Lamu, Kenya Colony. 
One specimen, from Mkonumbi (2°16' south lat., 40°42' east long.), 50 feet, 

near Lamu, Kenya Colony. 21. v. 1934. 
One specimen, from Mombosasa (2°20' south lat., 40°30' east long.), near 

Witu, Kenya Colony. 31. v. 1934. 

Remarks. In the type from Mkonumbi, the copulatory pores (i.e. 
the male pore and the spermathecal pore) were surrounded by only 
slight thickenings of the body wall. In this fresh material, which in- 
cludes a topotype and a worm from Mombosasa less than twenty 
miles west of the type locality, the body wall shows much more prom- 
inent modifications (fig. 20) in this sexual area. The entire ventral 
part of segments XVI-XIX has a somewhat glandular appearance 
without being sharply bordered. The male pore, marked by a small 
brownish pit, lies at the apex of a very prominent, nearly hemispherical 



474 



bulletin: museum of comparative zoology 



porophore. The circular base of this porophore occupies the whole 
length of segments XVI and XVII; the laterally distinct interseg- 
mental furrow XVI/XVII is obliterated within the limits of the 
porophore with the result that the relation of the male pore to this 
furrow is not clearly recognisable in these specimens, I should have 
described it as "in or very near XVI/XVII" but it may have been 
displaced while the male pore was projecting. It is stated of the type 
that the male pore is "slightly anterior to the middle of segment 
XVII." Segments XVI-XIX are, at least laterally, distinctly divided 
into 3 annuli, each being separated by two fine secondary annular 
furrows. While the secondary furrows adjacent to furrow XVI / 




Fig. 20. Polytoreutxis chaloneri. 
copulatory pores. 



Ventral view of the region with the 



XVII vanish, like the latter, in the limits of the male porophore, the 
secondary furrows of XVI and XVII further from furrow XVI /XVII 
cross the male porophore. The area surrounding the spermathecal 
pore is also differently modified ; the whole ventral portion of the body 
wall of segments XVIII and XIX is glandularly thickened, inter- 
segmental furrow XVIII /XIX is very distinct in the region of this 
glandular thickening, but the secondary furrows, which are only 
laterally distinct, disappear here, and are replaced by deep, trans- 
verse, medio-ventral furrows in the middle zones of segments XVIII 
and XIX, dividing the ventral parts of these segments into two 
swollen, transverse fields; the fields adjacent to intersegmental furrow 
XVIII /XIX are somewhat more prominent than those bordered by 
furrows XVII/XVIII and XIX/XX. The spermathecal pore, a 
rather inconspicuous slit situated medio-ventrally at XVIII /XIX, is 
surrounded by a slight modification of the body wall, which forms a 
transversely oval field only distinguished by its lighter colouring and 
not very prominent. 



michaelsen: African oligochaeta 475 



Polytoreutus multiporus Smith and Green 

Polytoreutus multiporus Fr. Smith and B. Green, 1919 (1920), Proc. U. S. Nat. 

Mus., 55, p. 161, figs. 13-18: Mkonumbi, near Lamu, Kenya Colony. 
Two specimens, from Mkonumbi (2°16' south lat., 40°42' east long.), 50 feet, 

near Lamu, Kenya Colony. 21.V.1934. 
Four specimens, from Gongoni (3°5' south lat., 40°10' east long.), 75 feet, 10 

miles north of Malindi, Kenya Colony. 27. v. 1934. 
One specimen, from Malindi (3°13' south lat., 40°8' east long.), 100 feet, 65 

miles north of Mombasa, Kenya Colony. 30.vi.1934. 
One specimen, from Changamwe (4°1' south lat., 39°37' east long.), 192 feet, 

3 miles west of Mombasa, Kenya Colony. 4.vii.l934. 

Remarks. The topotype from Mkonumbi, the only complete one, 
differs most noticeably from the type and cotype in its dimensions. 
It is exceptionally slender, about 270 mm. long and averaging about 
3 mm. in diameter. The much greater number of segments, viz. ca. 
670 is in conformity to its greater length. 

The external organs of puberty exhibit some variation in number 
and character, most of the specimens from the new localities agree 
with the types in regard to their postclitellar setal papillae, having 9 
or 10 papilla at each side of segments XIX-XXVII or XIX-XXVIII 
(in the types, 8 or 9 pairs at segments XIX-XXVI or XIX-XXVII). 

The Gongoni series differ remarkably from the rest in having, with- 
out exception, 5 postclitellar segments with genital setal papillae, 
normally at segments XIX-XXIII except for one individual, and 
then only on its right side, where the papillae of the last pair are 
lacking in segment XXIII being displaced to segment XXIV. 

In all eight specimens there occurs besides the posterior organs of 
puberty, some anterior ones, presumably lacking (at least not men- 
tioned) in the types from Mkonumbi. These anterior organs have a 
two-fold character, firstly there are setal papillae, like the posterior 
ones, usually 2 pairs on either side of segments XV-XVI, rarely (in 
two examples from Gongoni) indistinguishable in part, once a pair of 
supernumary papillae on the left side of segment XIV; secondly in 
some specimens (from Gongoni and one from Mkonumbi) 2 large, 
medio-ventral, transversely oval, glandular cushions are present at 
segments XV-XVI, occupying the entire length of these segments and 
bearing on their sides the setal papillae, or rather the sexual setae, 
the papillae of these setae being more or less sunken in the glandular 
mass of the cushions and in consequence rather indistinct. 



476 bulletin: museum of comparative zoology 

One of the Gongoni specimens bears a protruding penis (fig. 21) 
similar to that figured by Smith and Green (loc. cit. fig. 13) with which 
it agrees in its annulations but differs in the shape of its ectal end. 
The latter is somewhat broadened and truncated to form a generally 




Fig. 21. Polytoreutus multiporus. Projected copulatory pouch of the right 
side. 

plain surface which is not quite even but shows the external edges of 
some folds formed by the wall of the penis. The edges of these folds 
have the form of a 'W above the middle line of which lies the apical 
plane of a transversely oval bladder. It should be remembered that a 
different degree of protrusion of the penis would presumably result in 
its ectal end taking on a somewhat different form. 



michaelsen: African oligochaeta 477 

BIBLIOGRAPHY 
Beddard, F. E. 

1892. "On a new Genus of Oligochaeta, comprising Five new Species 
belonging to the Family Ocnerodrilidae." Ann. Mag. Nat. Hist. 
(6), 10, pp. 74-97, pis. vi-vii. 

1894. "A Contribution to our Knowledge of the Oligochaeta of Tropical 
Eastern Africa." Quart. Journ. Microsc. Sci., N.S., 36, pp. 201-269, 
pis. xvi-xvii. 

1903. "On a new Genus and two New Species of Earthworms of the 
Family Eudrilidae, with some Notes upon other African Oligo- 
chaeta." Proc. Zool. Soc. London, 1903, 1, pp. 210-222, figs. 
35-38. 

Martiis, L. Cognetti de 

1910. "Contributo alia Conoscenza della Fecondarione negli Oligo- 
chaeti." Atti. Reale Ace. Sci. Torino, 45, pp. 737-750, pi. xiv. 

1911. "Description of a new Species of the Genus Polytoreutus." Ann. 
Mag. Nat. Hist. (8), 7, pp. 507-513, figs. A-B. 

Michaelsen, W. 

1897 (1896). "Neue und wenig bekannte afrikanische Terricolen." 

Mitt. Mus. Hamburg, 14, pp. 1-71, pi., figs. 1-17. 
1905. "Die Oligochaten Deutsch-Ostafrikas." Zeitschr. Wiss. Zool., 

82, pp. 288-367, pis. xix-xx. 
1910. "Die terrestrischen Oligochaten des tropischen Afrikas und ihre 

geographischen Beziehungen." Wiss. Erg. deutsch-Zentral- 

Afrika-Exped. 1907-1908, 3, pp. 1-90, figs. 1-17, pis. i-ii. 
1912 (1913). "Oligochaten vom Kenya-Distrikt in Britisch-Ostafrika 

gesammelt von der Schwedischen Zoologischen Expedition 1911." 

Ark. Zool., 7, No. 32, pp. 1-5, pi. i. 
1913. "Oligochaten vom tropischen und siidlich-subtropischen Afrika." 

II Teil. Zoologica, 67, pp. 1-63, figs. 1-11, pis. i-ii. 
1934. "Opisthopore Oligochaten des Koniglichen Naturhistorischen 

Museums von Belgien." Meded. Koninkl. Natuurk. Mus. Belgie, 

10, No. 25, pp. 1-29, figs. 1-14. 
1934a. "Die opisthoporen Oligochaten Westindiens." Mitt. Mus. Ham- 
burg, 45, pp. 51-64, figs. 1-7. 

Smith, Fr. and Green, B. R. 

1919. (1920). "Description of new African Earthworms, including a new 
Genus of Moniligastridae." Proc. U. S. Nat. Mus., 55, pp. 145-166, 
figs. 1-18. 

Stephenson, T. 

1928. "Oligochaeta from Lake Tanganyika (Dr. C. Christy's Expedi- 
tion 1926)." Ann. Mag. Nat. Hist. (10), 1, pp. 1-17, figs. 1-4. 

1933. "Reports on the Scientific Results of an Expedition to the South- 
western Highlands of Tanganyika Territory. IV. Oligochaeta." 
Bull. Mus. Comp. Zool., 75, pp. 225-247, figs. 1-13. 



Y\.C.Z-L 



Bulletin of the Museum of Comparative Zoology 

AT HARVARD COLLEGE 
Vol. LXXIX, No. 9 



SCIENTIFIC RESULTS OF AN EXPEDITION 
TO RAIN FOREST REGIONS IN EASTERN AFRICA 

IX 

ZOOGEOGRAPHY AND ITINERARY 



By Arthur Loveridge 



With Fouh Plates 



The Libra y 

Museum of Comparative Zoology 

Harvard University 



CAMBRIDGE, MASS., U.S.A. 
PRINTEDFORT HE MUSEUM 
October, j 19371 



No. 9. — Reports on the Scientific Results of an Expedition to Rain 
Forest Regions in Eastern Africa 

IX 

Zoogeography and Itinerary 

By Arthur Loveridge 

INTRODUCTION 
GENERAL REMARKS 

The Expedition, upon the results of which this paper forms the 
concluding report, was undertaken by the author as a Fellow of the 
John Simon Guggenheim Memorial Foundation. 

The object in view was a clarification and extension of our knowl- 
edge regarding the distribution of the isolated, sylvicoline forms of life 
associated with the forest 'islands' of East Africa. In addition, an 
attempt has been made towards further elucidation of their relation- 
ship with the fauna of the great West African forests, particularly that 
of the Cameroon Mountains. 

In this respect the work must be regarded as an extension of a 
programme outlined in 1924. This resulted in a visit to the forests of 
the Usambara and Uluguru Mountains of eastern Tanganyika Terri- 
tory in 1926-1927, and those of the Uzungwe, Ukinga, Rungwe and 
Poroto Mountains in southwestern Tanganyika in 1929-1930. 

The principal conclusions resulting from these trips were embodied 
in various reports such as "A Comparative Study of the Herpetological 
Faunae of the Uluguru and Usambara Mountains. 1 " and the "Intro- 
duction and Zoogeography" to "Reports on the Scientific Results of an 
Expedition to the Southwestern Highlands of Tanganyika Territory. 2 " 

The latter paper, in particular, outlines (p. 27) the main points 
connected with the discontinuity of the distribution of wild life per- 
sisting in the remnants of primeval forest surviving on East African 
mountains, gives percentages of genera and species regarded as com- 
mon to the Cameroon Mountains, and contains other pertinent matter 
which need not, therefore, be repeated here. 

The Expedition of 1933-1934 consisted of the author, his gunbearer, 
and two native skinners. On this occasion forested areas lying to the 

1 Barbour and Loveridge, 1928, Mem. Mus. Comp. Zool., 50, pp. 87-265, pis. i-iv. 

2 Loveridge, 1933, Bull., Mus. Comp. Zool., 75, pp.1-43, pis. i-iii. 



482 bulletin: museum of comparative zoology 

northwest and north of the Usambara Mountains, were selected. The 
principal ones being Mounts Debasien (Kadam) and Elgon (Masaba) 
in Uganda, the Kakamega Forest at Kaimosi, and Mount Mbololo in 
Teita, 1 Kenya Colony. In addition many relic patches were visited 
as set forth in the itinerary given on pages 506 539 of the present paper. 

In my 1933 report it was postulated that the amphibia, being of 
more sedentary habits and restricted in their migrations by their 
ecological necessities, were the most promising group of vertebrates 
from which to collect data that might throw light on this problem. 
In equatorial Africa, however, where there is an alternation of dry 
and wet seasons, the probability of one's finding frogs in the middle 
of the dry season is but slightly better than the likelikhood of encoun- 
tering them in New England in midwinter. With this in mind, the trip 
was planned so as to include and take advantage of both the "small" 
October-November rains of eastern Uganda, as well as the "big" 
monsoon rains of May and June on the Kenya coast. 

Unfortunately Central Africa did not escape the extensive drought 
which, in 1933, assumed almost world-wide proportions. This drought 
adversely affected amphibian emergence during the first four months 
of the expedition. For example, at Kaimosi, which is in the region of 
the heaviest rainfall in Kenya, averaging 74.16 inches, and where rec- 
ords have been kept for thirty years, the millpond was lower than at 
any time during the three decades. Had not some showers occurred 
towards the end of February one wonders if our amphibian record 
would not have been as poor as that for Mounts Debasien and Elgon. 

So severe was this East African drought that it formed a subject of 
discussion in the House of Commons. Sir Philip Cunliffe-Lister then 
stated that between January and April the Government spent £6,500 
for famine relief in the Coast Province of Kenya alone. ("East Africa," 
17. v. 34.) 

When the big rains broke at last at the coast, they were exceptionally 
heavy, twenty inches above the normal falling during the months of 
April, May and June. Nor was this merely a local phenomenon of the 
Tana region, where I was at the time, for in the south the Zambezi 
was higher than within living memory, rising 77 feet in the gorge to a 
total depth of 135 feet. In the north the Nile flood reached its highest 
in forty years, inundating many small villages and towns as well as 
the northern and southern suburbs of Cairo. ("Times Weekly," 
London, 13. ix. 34.) 

1 The alternative of Taita, long in use by Zoologists, was employed in the taxonomic reports, 
but should give way to Teita as it is the form adopted by the Government of Kenya. 



loveridge: African zoogeography and itinerary 483 

DEFORESTATION AND EROSION 

Rainfall, deforestation and erosion are so inseparably bound up with 
the whole problem of the survival of the sylvicoline fauna that it would 
appear appropriate to make some remarks on the subject at this point. 

The tragic results of uncontrolled agricultural development and 
overstocking, recently culminating in the creation of the so-called 
'dust bowl' in the middle western United States, is an object lesson 
which should not go unheeded in Eastern Africa. It must be remem- 
bered that these dire consequences have supervened after less than a 
century of exploitation of the land. 

Owing to the unparalleled developments which have been taking 
place in East Africa during the last thirty years, extensive areas have 
undergone deforestation with the consequent disappearance of the 
specialized fauna inhabiting them. Only recently a conchologist of 
wide African experience told me how he had landed on a certain 
island in Lake Kivu in search of a species of snail known only from 
there. The virgin forest, which once clothed the island, had been 
destroyed to make way for coffee plantations. The slump in coffee 
prices had resulted in the virtual abandonment of the plantations. 
Not a living snail could be found, however, nothing but a few dead 
shells — the new conditions were inimical to the survival of the 
species. 

The Teita porters who carried my loads up the almost precipitous 
ascent of Mount Mbololo, told me that they could remember when 
forest clothed the mountain side. Today only about a thousand acres 
of it survive as a narrow strip, two or three miles in length, running 
along the hog-backed ridge at 4,800 feet. This relic patch appeared to 
vary from one to two hundred yards in width. It is now under the 
protection of the Forestry Department of Kenya. 

Though the plains below lay baking in sunshine by day, almost 
nightly during our stay fog or mist collected about the magnificent, 
though pitiful, remnant of forest. There it condensed in so heavy a 
precipitation that the drops fell like rain upon the tent, continuing to 
do so until several hours after sunrise the following morning. Before 
8 a.m., a walk through the grass fringing the forest would leave one 
soaked. When digging for caecilians in the rich leaf mould of the forest 
floor, we always found the ground sodden. In the cool depths of the 
forest it was damp and pleasant throughout the day, however trying 
the heat and glare might be a few hundred yards outside. 

Below the forest, the steeper slopes were strewn with rocks which 
protruded like the bare bones of the mountain through the scanty 



484 bulletin: museum of comparative zoology 

covering of gravel which remained. Lower down, stunted scrub, a few 
scattered trees and sparse grass testified to the poverty of what little 
soil remained. Nevertheless, here and there were ravines where little 
marshes were fed by a trickle of water coming from a spring among the 
inhospitable rocks. Undoubtedly this water had percolated through 
from the forest above. 

Recently the United States Forest Service has published data, ob- 
tained by actual measurement, bearing upon the protection from 
floods which is afforded by the action of forests in conserving the run 
off. There it is shown that forest soil in the Ohio Valley absorbs 50 
times more water than does exposed earth, being 15 to 30 times more 
porous than field soil. Pasture land retains only a third to a twentieth 
of the amount which would be absorbed by an equal area of forest 
floor. 

Thirty years ago Robin Kemp made a wonderful collection of small 
mammals on the southeast face of Mount Elgon. When I visited the 
locality in 1934, I found conditions very different from those which I 
anticipated in view of the species which he had encountered there in 
1909. I therefore wrote to Mr. Kemp as follows: "... last year I 
camped half a mile above the cave which the natives still remember 
your having occupied. Could you tell me what the slopes were clothed 
with at the time of your visit? Today, as the result of annual burning, 
there is only thorn scrub and thicket with a few scattered baobabs, 
except along the upper reaches of the stream below the falls where a 
little rain forest still persists." I might have added that great gulleys 
had been cut by the run off through the already semi-arid foothills 
lying between the escarpment and Bukori. In his most interesting 
reply to my letter, Mr. Kemp says: "The stream in those days was 
very much hidden. The whole ravine in front of the cave — from the 
Kirui plain upwards to as far as I climbed in the ravine — was a 
dense mass of green timbered virgin forest. I do not remember any sign 
of fire damage either along the bottom or along the tops of the trees." 

One feels that the changes which the region is undergoing impress 
themselves more forcefully on those who, like myself, revisit places 
after lengthy intervals. That certain settlers are alive to the dangers 
threatening, is evidenced by the formation of such organizations as 
the Soil Erosion Sub-Committee of the Subukia Farmers' Association. 
Mr. D. N. Stafford, in an address to the Uganda Planters' Association, 
remarked that: "The Kampala-Entebbe Road was once the pride of 
the country, the admiration of every visitor. Today most of the forest 
has been ruthlessly destroyed, and the beauty patches have dis- 



loveridge: African zoogeography and itinerary 485 

appeared. What has happened on this road is going on in every road 
in Buganda." ("East Africa," 2. viii. 34.) 

The regional governments have all sponsored investigations into 
one phase or another of the subject. In Tanganyika, Dr. E. O. Teale 
and Mr. C. Gillman recently issued their report on the menace to the 
future water supplies in the Northern Province. These authors state 
that some of the areas which they visited were already beyond hope of 
reclamation, and they urge prompt action as vitally necessary to save 
others which are threatened. Expensive as preventive measures often 
are, yet it is infinitely more economical than moving populations 
which may ultimately be the only alternative. 

Of Nyasaland, Mr. A. W. J. Hornby writes: "Owing to the denuda- 
tion of woodland areas, the gradual destruction of forest, and sudden 
floods, the country can be said now to be incapable of supporting half 
the population it did a hundred years ago." (1934, Nyasaland Bulletin, 
No. 11.) 

Mr. E. J. Wayland, Director of Geological Survey in Uganda, 
states that the country, in common with many other parts of Africa, 
is drying up. A few years ago the influence of forests on climate and 
water supply in Uganda and Kenya was discussed at length by Mr. 
J. W. Nicholson, who visited East Africa at the request of the respec- 
tive governments. In the report published by the Kenya Forest De- 
partment, this investigator states (1929, p. 35): "There is a great deal 
of evidence hydrological, geological and botanical, to show that parts 
of Africa are undergoing progressive desiccation. . . . There is some 
evidence to show that parts of Kenya and Uganda (e.g. Karamoja — 
Lake Rudolf) are in course of drying up." More pertinent, however, is 
"Our third conclusion is that under favourable circumstances moun- 
tain forests in East Africa can induce occult precipitation up to at 
least 25 per cent of the total annual rainfall." (p. 17.) 

As a result of Mr. Nicholson's findings, the remaining patch of 
forest on Mount Debasien, in arid Karamoja, is being very strictly 
preserved. The measures taken included the moving of several native 
families who had squatted in the forest. On the western slopes of 
Elgon above Sipi, however, we found that the Bagishu were making 
heavy inroads into the forest, whether legally or illegally I did not 
ascertain; at least it was being done. 

In so far as natives are concerned, the governments face a difficult 
problem with an expanding population. With a view to enlightening 
and instructing them, the Bantu Educational Committee has prepared 
a film on soil erosion, its causes and prevention. Native conservatism 



486 bulletin: museum of comparative zoology 

will be difficult to combat and not many Governors speak so frankly 
as did Sir Harold MacMichael when addressing a gathering of 8,000 
natives at Lushoto in the Usambara Mountains, in September, 1935. 
He charged them with the haphazard destruction of timber and the 
reckless wastage of a valuable heritage upon which their descendents 
would be dependent for a living. 

Deforestation continues at this moment, and is likely to do so until 
an enlightened public, both native and European, supports the efforts 
of the Forest Departments in their arduous task of conservation. In 
so far as saving the sylvicoline fauna is involved, replanting appears 
useless, for, with a few exceptions, the forest-dwelling lower verte- 
brates do not adapt themselves to the types of trees (gums, wattle, 
etc.) which are so largely employed in replanting operations. x 



ZOOGEOGRAPHY 

ECOLOGICAL LIFE ZONES 

It would appear that sufficient collecting has been done in East 
Africa to justify us in attempting to assign the herpetofauna to life 
zones with such a reasonable degree of accuracy that it should leave 
little room for criticism or the necessity for much future adjustment. 

The most important contribution to an understanding of the dis- 
tribution of African reptiles has been given by Schmidt (1923, pp. 
4-45), though an increase in our knowledge in the years that have 
elapsed since the appearance of this paper, inevitably necessitates some 
amendment, and even difference of opinion. For an exhaustively de- 
tailed, first-hand account of the ecological habitats of the amphibia, 
one could not wish for a more interesting one than that of Sanderson 
(1936, pp. 165-187) dealing with an area in the Cameroons. In 
respect to certain collections from British Somaliland, northern Kenya, 
Angola and South West Africa, Parker (1932a, 1932b, 1936a, 1936b) 
has allocated his material to certain groupings which, though suitable 
for the regions in question, are neither wholly applicable nor adequate 
for the present area and purpose. 

This area, embracing Kenya, Uganda, Tanganyika and Zanzibar, 
is referable to the Eastern and Southern Subprovince of the Ethiopian 

1 Since this paper was written a year ago, much has happened. Of particular interest was the 
announcement in July, 1937 by His Excellency, the Governor of Kenya, that the Government 
was planning to initiate an extensive scheme of preventive measures aiming to check further 
erosion ("East Africa and Rhodesia," 29, vii. 37.) 



loveridge: African zoogeography and itinerary 487 

Province as denned for botany by Engler. The herpetofauna of the 
area conveniently falls into nine ecological life zones of which the 
coastal merges into the savanna and the latter into the upland, though 
to a less extent. 

Most writers have been embarrassed by the ubiquitous, widespread 
species that occur in several habitats, the usual practice has been to 
assign them to a separate group (Parker, 1932a, p. 213). This practice, 
however, has the disadvantage of giving an incomplete picture of the 
fauna of the separate zones. In the present paper, therefore, I have 
adopted the course of repeating immediately after the name of each 
species, the numbers of the other zones in which it occurs should it 
inhabit several. 

Even so, exceptions were made in the case of the Northern Desert 
Zone which was created for a special purpose, and again for the Forest- 
edge Zone where most of the true forest species might supposedly be 
encountered at times. The inclusion of their names would result in 
unnecessary repetition. 

Perhaps it would be as well to remind readers that the life zones to 
which the species are assigned are those that they occupy in East 
Africa. It is conceivable that in other regions they may be found at 
very different elevations. As a case in point, I imagine that in South 
Africa such snakes as Pseudaspis, Duberria, and T rimer orhinus are by 
no means restricted to grassy uplands about and above 5,000 feet. 

For the benefit of those who are interested in the grouping of the 
virile, ubiquitous types referred to above, I list below as representative 
of the most adaptable species of the Ethiopian herpetofauna, most of 
those which occur in three or more of the zones here adopted. 

Crocodylus niloticus Dasypeltis scaber 

Thelotornis kirtlandii 
Pelomedusa galeata Dispholidus typus 

Pelusios n. nigricans Calamelaps unicolor 

Causus resimus 
Varanus niloticus Causus defilippi 

Mabuya maculilabris 

Mabuya megalura Bufo r. regularis 

Mabuya v. varia Rana g. bravana 

Brookesia k. kersteni Rana f. angolensis 

Rana o. oxyrhynchus 
Python sebae Rana m. mascareniensis 

Natrix o. olivacea Arthroleptis minutus 

Boaedon lineatus Phrynobatrachus acridoides 

Chlorophis neglectus Phrynobatrachus natalensis 



488 bulletin: museum of comparative zoology 

The fact that Parker (1936a, p. 118), when dealing with the Angolan 
fauna, also includes Hemidactylus mabouia, Gerrhosaurus flavigularis, 
Chamaeleo dilepis and Causus rhombeatus, is not indicative of any 
difference of opinion regarding them, but results from the difference 
of the zoning employed. In the east these reptiles are restricted to the 
coast and savanna zones, though possibly dilepis might be included 
with the forest-edge group of chameleons. The point is a very minor 
one. 

LIST OF THE LIFE ZONES ADOPTED IN THIS PAPER 

1. Marine. Sea level. 

2. Freshwater Rivers, Lakes and Swamps. circa 1- 5,000 feet. 

3. Littoral Rocks. circa 1- 10 feet. 

4. Northern Desert. circa 1,000- 3,000 feet. 

5. Coastal Plain. circa 10- 1,000 feet. 

6. Upland Savanna circa 1,000- 6,000 feet. 

7. Grassy Uplands and Alpine Meadows. circa 5,000-12,000 feet. 

8. Forest-edge. circa 3,000-10,000 feet. 

9. Rain Forest, usually Montane. circa 3,000-10,000 feet. 

Zone 1. Marine 

The wholly marine reptiles resident in the Indian Ocean off the 
East African coast, consist of four turtles (with the possible addition 
of the recently described Caretta gigas Deraniyagala, if recognizable) 
and a sea snake. I am unaware of any definite, specifically-localized 
records of the taking of the luth (D. coriacea) or brown and yellow 
sea snake (P. platurus) on the coast, though both are known to occur 
not far off. 

Dermochelys coriacea Caretta caretta olivacea 

Eretmochelys imbricata 

Chelonia mydas Pelamis platurus 

Zone 2. Freshwater Rivers, Lakes and Swamps 

In this group we have a number of species whose life is conditioned 
wholly, or in part, by the presence of water, at least during a portion 
of the year. Among those reptiles for whose semi-aquatic habits water 
is essential, are the crocodiles, terrapin, soft-shelled tortoises and two 
species of snakes. By zoning them on this basis, we immediately elim- 
inate the difficulty of the widespread distribution of many species, 
for it is the presence of water that governs their distribution, rather 



loveridge: African zoogeography and itinerary 



489 



than whether that water flows through coastal plain, savanna, grassy 
upland, or forest. 

Schmidt (1919, p. 401) has attempted to separate the African 
chelonians according as to whether they occur within or without the 
forest: the exceptions, however, form a considerable percentage of the 
whole. 

* 

The lizards and snakes listed are species which dwell on the borders 
of streams and swamps, principally on account of their diet which, in 
so far as most of the snakes are concerned, is one of frogs and fish. 
The water cobra (B. a. stormsi) spends its days in the waters of Lake 
Tanganyika. The species of Grayia are only slightly less amphibious. 
The limbless Melanoseps, which may be really a rain-forest skink, 
appears to be surviving only where conditions are sufficiently moist for 
burrowing. On the plains such an environment was found in gallery 
forest along the Mkata River; in the Uluguru Mountains beside a 
waterfall whose spray kept the site moist; near a patch of forest on 
grass-grown uplands, yet another was found in sandy soil among the 
rotting roots of an old stump on the bank of a watercourse. 

Among the amphibia, the caecilian (D. gregorii) is outstandingly 
conspicuous as a burrower in deep mud on the banks of the Tana 
River and beneath the waters of Lake Peccatoni. In this respect its 
environment is sharply distinguished from that of its rain-forest 
relatives, though a moist and permeable soil appears essential to both. 

While it is true that the majority of frogs and toads require ponds or 
swamps in which to spawn during some part of the year, many are 
capable of existing under arid savanna conditions for many months at 
a time. Below are listed only those species for whom ponds and 
swamps seem so essential an element of their habitat that they nat- 
urally fall under this heading. 



Crocodylus cataphractus in 6. 
Crocodylus niloticus in 5, 6. 

Pelomedusa galeata in 5, 6. 
Pelusios sinuatus in 5, 6. 
Pelusios n. nigricans in 5, 6, 9. 
Pelusios derbianus (doubtful) 9. 
Trionyx triunguis in 6, 9. 
Cycloderma frenatum in 6. 

Varanus niloticus in 3, 5, 6. 
Melanoseps ater in 5, 6, 7, 9. 



Python sebae in 5, 6, 9. 
Natrix o. olivacea in 5, 6. 
Glypholycus bicolor in 6. 
Chlorophis hoplogaster in 6, 9. 
Chlorophis neglectus in 5, 6, 9. 
Chlorophis irregularis in 6, 9. 
Grayia smythii in 6, 9. 
Grayia tholloni in 6, 9. 

Boulengerina a. stormsi in 6. 
Dermophis gregorii in 5. 



490 



bulletin: museum of comparative zoology 



Xenopus I. petersii in 6. 
Xenopus I. victorianus in 6. 
Xenopus I. bunyoniensis in 7. 
Xenopus I. borealis in 6. 
Xenopus muelleri in 5, 6. 
Xenopus clivii in 4. 
Megalixalus fornasinii in 5, 6. 
Megalixalus dor salts in 6. 
Megalixalus fulvovitiatus in 6. 
Megalixalus brachynemis in 5, 6. 
Megalixalus flavomaculatus in 5. 



Hyperolius spp. in 5, 6, 7, 9. 
Rana g. bravana in 5, 6. 
Rana f. angolensis in 6, 7. 
J?ana o. oxyrhynchus in 5, 6, 7. 
.Rana to. mascareniensis in 5, 6, 7. 
Rana stenocephala in 6. 
Rana occipitalis in 6. 
Arthroleptis spp. in 5, 6, 7, 9. 
Phrynobatrachus acridoides in 5, 6, 7. 
Phrynobatrachus natalensis in 5, 6, 7. 



Zone 3. Littoral Rocks 

The little skink which is listed below, enjoys the distinction of being 
exclusively an inhabitant of the littoral limestone or coral rag. In 
this situation, at times splashed by ocean breakers, it seeks its prey — 
the smaller Crustacea. 

Where the rocks are more or less overgrown with vegetation, one 
frequently encounters monitors, chiefly the eyed species, for the 
Nilotic monitor seldom strays far from the river estuaries. Both of 
these big lizards enjoy an extensive inland range. 

Varanus ocellatus also 5, 6. 
Varanus niloticus also 2, 5, 6. 
Ablepharus b. africanus 



Zone 4. Northern Desert 

This zoning coincides with Parker's (1932b, p. 214) group IV of 
"Species endemic in the Somaliland region, sometimes extending into 
Kenya." While it contains three times the number of species listed 
by him as represented in the collection with which he was dealing, I 
imagine that we would be in entire agreement as to its composition. 

Buxton (1937, p. 85) in a most interesting account of this region, 
with which I am personally unacquainted, prefers "semi-desert" for 
the area as a whole. In his article will be found an excellent detailed 
account of its physiographic, floral and faunistic aspects. 

In northern Kenya Colony, from the Ethiopian and Somaliland 
boundary southwards to the equator, and just past it at Njoro, there 
is to be found a southward extension of desert life accompanying the 
slowly encroaching desert. Only 2 of the 15 forms listed have been 
taken south of this area. The single example of H. citernii found by 
me on a dump outside the lines of the King's African Rifles at Nairobi, 



loveridge: African zoogeography and itinerary 491 

shortly after the return of a contingent from the north, was undoubt- 
edly introduced. The second species, E. neumanni, described from 
just north of Lake Stephanie, was found to be firmly established at 
Ngatana,on the north bank of the Tana River, just south of 2° latitude. 
It is possible, therefore, that this species belongs to the coastal fauna 
rather than to that of the northern desert. 

From Mount Longido in northern Tanganyika, northwards through 
Voi, we find a second desiccating area, largely on red laterite, which 
seems to be developing a somewhat similar fauna. However, the latter 
merges with that of the adjacent drier savanna areas to such an extent 
that its separation would result in extensive duplication. Species 
characteristic of this arid laterite region are listed separately and with 
the other savanna species of Zone 6 to which they belong. 

In his first paper on the herpetofauna of the East African Lake 
region, Parker (1932b, p. 213) refers Lygodactylus p. gutturalis to his 
Group I comprised of " 'Eremian' species, i. e. those found in Africa 
north of about the 10th parallel (north), Arabia, or both, and not 
extending southeast into Kenya." The latter part of this definition 
referring to Kenya would appear to be a lapsus, for on page 223 he 
records examples from Baringo and the Turkana Plains in Kenya. The 
position is somewhat clarified by his later (1936b, p. 602) describing 
of the race keniensis to which they would be referred. However, L. p. 
gutturalis, as I understand it, extends from its type locality Bissao, 
Portuguese Guinea, east to Uganda (Mbale, Karamoja, etc.) and 
southwards to Ujiji, Lake Tanganyika. I recently separated the 
Sudanese geckos under the name of L. p. sudanensis. 

The saw-scaled viper (Echis carinatus) is likewise placed in his 
group of 'Eremian' species, though on p. 221 of the same paper, Parker 
records it from the mouth of the Kaliokwel River, i. e. about the 3rd 
parallel (north), and it has been recorded from this same general region 
by both Boulenger and myself. 

Hemidactylus ruspolii Philochortus i. rudolfensis 

Hemidactylus macropholis Leptotyphlops fiechteri 

Hemidactylus citernii (see above) Coluber florulentus 

Hemidactylus isolepis Coluber smithi 

Lygodactylus p. keniensis Coluber keniensis 

Agama r. occidentalis Echis carinatus 
Eremias neumanni also 5. 

Eremias smithi Xenopus clivii 
Eremias striata 



492 



bulletin: museum of comparative zoology 



Zone 5. Coastal Plain or Belt 

The hot and humid coastal plain varies considerably in width 
though its altitude may be assessed as usually well under a 1,000 feet. 
Along the Tana River valley of Kenya its inland extension is as much 
as 150 miles, and almost as far along the line of the Central Railway 
in Tanganyika to Morogoro. On the other hand we find localities near 
to the coast, such as Amani at 3,000 feet in the Usambara Mountains, 
which indubitably belong to the forest zone. 

While there is an interesting coastal fauna restricted to the palm- 
bearing, sandy plain in close proximity to the ocean, this region is 
principally populated by a far larger group with a high percentage of 
cosmopolitan species whose headquarters are the savanna. The first 
group is represented by a score of species, 1 half of which are burrowing, 
while there are nearly five times as many forms in the second. 

From this it will be seen that ecological conditions in these two 
zones — the coast and savanna — often approximate, and coastal 
forms may find congenial conditions of life in a hot and humid mon- 
tane valley even though it may exceed the 1,000 foot level. Nyange, 
in the Uluguru Mountains, provides just such conditions though its 
surrounding forest-clad heights support a forest fauna. 

Fourteen reptiles, mostly snakes with a wide distribution, find 
conditions in patches of coastal bush and forest not too far removed 
from rain-forest conditions to prevent their occurrence in both habi- 
tats. Sternf eld's astonishing record of Miodon gabonensis from Dar es 
Salaam, is, however, omitted pending confirmation. 



Crocodylus niloticus also 2, 6. 

Testudo p. babcocki also 6. 
Testudo tornieri also 6. 
Kinixys spekii also 6. 
Pelomedusa galeata also 2, 6. 
Pelusios sinuatus also 2, 6. 
Pelusios n. nigricans also 2, 6, 9. 

*Diplodactylus wolterstorffi. 

Hemidactylus brookii also 6. 

Hemidactylus parkeri (insular) 

Hemidactylus mabouia also 6. 
* Hemidactylus persimilis 

Hemidactylus mandanus (insular) 



Hemidactylus frenatus (introduced) 

Hemidactylus w. werneri also 6. 

Hemidactylus t. squamulatus also 6. 

Bunocnemis modestus 

Lygodactylus grotei also 6. 

Lygodactylus p. mombasicus also 6. 

Lygodactylus p. picturatus also 6. 
*Platypholis f. fasciata 
*Phelsuma laticauda (introduced) 

Agama m. mossambica 

Agama a. lionotus also 6. 

Agama atricollis also 6. 

Zonurus tropidosternum also 6. 

Varanus albigularis also 6. 

Varanus ocellatus also 3, 6. 



1 Such are indicated by an asterisk in the following list. 



loveridge: African zoogeography and itinerary 



493 



Varanus niloticus also 2, 3, 6. 
*Chirindia ewerbecki 
*Amphisbaenula orientalis (? distinct) 

Eremias neumanni also 4. 

Eremias s. spekii also 6. 

Eremias s. sextaeniata also 6. 
*Gastropholis vittata 

Gerrhosaurus m. major also 6. 

Gerrhosaurus f. fiavigularis also 6. 

Mabuya maculilabris also 6, 9. 

Mabuya planifrons also 6. 

Mabuya brevicollis also 6. 

Mabuya megalura also 6, 7. 

Mabuya v. varia also 6, 7. 

Mabuya striata also 6. 

Riopa mabuiiformis 

Riopa tanae 

Riopa sundevallii also 6. 

Riopa m. modestum also 6. 

Riopa pembanum 

Ablepharus wahlbergii also 6. 

Scelotes tetradactyla also 9. 

Melanoseps ater also 2, 6, 7, 9. 
*Scolecoseps acontias 

Chamaeleon d. roperi also 6. 

Chamaeleon d. quilensis also 6. 

Chamaeleon d. dilepis also 6. 

Brookesia brevicaudata also 6, 8. 

Brookesia k. kersteni also 6, 8. 

Leptotyphlops conjuncta also 6. 
Leptotyphlops emini also 6. 

* Leptotyphlops boulengeri (insular) 

* Leptotyphlops longicauda 
*Typhlops lumbriciformis 
*Typhlops braminus (introduced) 
*Typhlops pallidus 
*Typhlops u. unitaeniatus 
*Typhlops platyrhynchus 

Typhlops s. mucruso also 6. 
Python sebae also 2, 6, 9. 
Natrix o. olivacea also 2, 6. 
Natrix o. pembanum (insular) 
Boaedon lineatus also 6, 9. 
*Lycophidion c. acutirostre (insular) 



Mehelya c. chanleri also 6. 
Mehelya nyassae also 6. 
Chlorophis neglectus also 2, 6, 9. 
Philothamnus s. semivariegatus also 

6, 9. 
Coronella coronata also 6. 
Coronella s. semiornata also 6. 
Prosymna a. stuhlmanni also 6. 
Scaphiophis albopunctatus also 6. 
Dasypeltis scaber also 6, 9. 
Tarbophis s. semiannulatus also 6. 
Tarbophis guentheri 
Crotaphopeltis h. hotamboeia also 6. 
Chamaetortus aulicus also 6. 
Hemirhagerrhis kelleri 
Rhamphiophis rostratus also 6. 
Rhamphiophis rubropunctatus also 6. 
Psammophis punctulatus also 6. 
Psammophis sibilans also 6. 
Psammophis subtaeniatus also 6. 
Psammophis biseriatus also 6. 
Psammophis angolensis also 6. 
Thelotornis kirtlandii also 6, 9. 
Dispholidus typus also 6, 9. 
Calamelaps unicolor also 6, 9. 
Micrelaps bicoloratus also 6. 
Aparallactus turneri 
Aparallactus concolor also 6. 
Aparallactus uluguruensis also 9. 
Naja nigricollis also 6. 
Dendraspis angusticeps also 6. 
Causus resimus also 6, 9. 
Causus defilippi also 6, 9. 
Vipera superciliaris also 6. 
Bitis arietans also 6. 
Atractaspis bibronii also 6. 
Atractaspis microlepidota also 6. 

Dermophis gregorii also 2. 
Hypogeophis guentheri (insular) 
Boulengerula changamwensis 

Xenopus muelleri also 2, 6. 
Bufo r. regular is also 6, 7, 9. 
*Bufo steindachnerii 



494 



bulletin: museum of comparative zoology 



Chiromantis p. petersii also 6. 

Chiromanlis xerampelina also 6. 
*Leptopelis ccncolor 
*Hylambates maculatus 

Kassina senegalensis also 6. 

Megalixalus fornasinii also 2,6. 

Megalixalus brachynemis also 2, 6. 

Megalixalus flavomaculatus also 2. 

Hyperolius spp. (about thirty) 

Rana g. bravana also 2, 6. 

Rana floweri also 6. 

Rana o. oxyrhynchus also 2, 6, 7. 

Rana m. mascareniensis also 2, 6, 7. 



.Rana edulis also 6. 
jffiana d. delalandii also 6. 
Arthroleptis s. stenodactylus also 6. 
Arthroleptis minutus also 6, 7. 
Phrynobatrachus acridoides also 2, 6, 

7. 
Phrynobatrachus natalensis also 2, 6, 

7. 
Hemisus m. marmoratum also 6. 
Breviceps mossambicus also 6. 
Spelaeophryne methneri also 6. 
Phrynomerus bifasciatus also 6. 



Zorce 6. Upland Savanna 

The term savanna is used here in the rather broad phytogeographic 
sense of a tropical grassland containing scattered trees of a xerophilous 
type, locally known as 'miombo.' In this dry woodland Brachystegia 
and Pterocarpus are dominant with here and there a giant baobab. 
During the long dry season much of this country desiccates to such 
a degree that it can best be described as desert. In its more arid 
stretches dense thickets or scattered thorn bush prevail. 

Such a type of country reaches its extreme in the Taru Desert area. 
As already mentioned when discussing the Northern Desert zone, we 
find a rather characteristic group of species inhabiting this arid 
laterite region which extends from Voi in Kenya to Kilimanjaro and 
Mount Longido in northern Tanganyika. The following are typical 
of this red soil country, to which they are apparently largely restricted. 



Agama r. septentrionalis 
Geocalamus spp. 

Eryx c. loveridgei 
Coronella s. fuscorosea 
Dasypeltis scaber (red phase) 



Rhamphiophis rubropunctatus 
Dispholidus typus (red phase) 
Rhinocalamus dimidiatus ? 
Micrelaps bicolor 
Naja n. nigricollis (red phase) 



R. dimidiatus is only known from Mpwapwa considerably to the south 
but may be confidently looked for in the northern area. All these 
species are referred to the Upland Savanna zone for the reasons 
already stated. 

A second subgroup comprising some tortoises, the agamas, zonures 
and a couple of skinks, centre round the small rocky hills, denuded of 
soil by erosion, which are quite characteristic of the savanna. 



loveridge: African zoogeography and itinerary 



495 



It will be noted that I refer Ichnotropis bivittata to this group for I 
have encountered it in the hot and dry woodland savanna east of Lake 
Tanganyika, moreover I have captured its allies I. squamulosa and 
/. capensis (this last in Mozambique) under even more arid conditions 
on open plains almost devoid of trees, but with scattered patches of 
scrub. Parker (1936a, p. 118) places bivittata with the rain-forest 
fauna of Angola; while this may be the case on the other side of the 
continent it is a point worth reconsidering. Chamaeleon etiennei is 
another species which I should imagine is primarily a resident in 
orchard savanna rather than in the rain forest. 



Crocodylus cataphractus also 2. 
Crocodylus niloticus also 2, 5. 

Testudo p. babcocki also 5. 
Testudo tornieri also 5. 
Testudo procter ae 
Kinixys spekii also 5. 
Kinixys belliana 
Pelomedusa galeata also 2, 5. 
Pelusios sinuatus also 2, 5. 
Pelusios n. nigricans also 2, 5. 
Trionyx triunguis also 2, 9. 
Cycloderma frenatum also 2. 

Hemidactylus tanganicus 
Hemidactylus brookii also 5. 
Hemidactylus mabouia also 5. 
Hemidactylus w. werneri also 5. 
Hemidactylus w. alluaudi 
Hemidactylus t. squamulatus also 5. 
Lygodactylus strigatus 
Lygodactylus capensis 
Lygodactylus scheffleri 
Lygodactylus grotei also 5. 
Lygodactylus p. gulturalis 
Lygodactylus p. ukerewensis 
Lygodactylus p. mombasicus also 5. 
Lygodactylus p. picturatus also 5. 
Pachydactylus b. turneri 
Pachydactylus boulengeri 
Agama r. septentrionalis 
Agama h. armata 
Agama m. montana 



Agama a. agama 
Agama a. lionotus also 5. 
Agama a. usambarae 
Agama a. elgonis 
Agama a. dodomae 
Agama a. ufipae 
Agama p. mwanzae 
Agama p. caudospinosa 
Agama atricollis also 5. 
Zonurus tropidosternum also 5. 
Zonurus ukingensis 
Varanus albigularis also 5. 
Varanus ocellatus also 3, 5. 
Varanus niloticus also 2, 3, 5. 
Amphisbaena phylofiniens 
Amphisbaena mpwapwaensis 
Geocalamus modestus 
Geocalamus acutus 
Nacras b. boulengeri 
Nucras b. kilosae 
Eremias s. spekii also 5. 
Eremias s. sextaeniata also 5. 
Latastia johnstonii 
Latastia I. revoili 
Ichnotropis tanganicana 
Ichnotropis bivittata 
Ichnotropis squamulosa 
Gerrhosaurus m. major also 5. 
Gerrhosaurus m. zechi 
Gerrhosaurus f. flavigularis also 5. 
Gerrhosaurus f. nigrolineatus 
Mabuya maculilabris also 5, 9. 
Mabuya planifrons also 5. 



496 



bulletin: museum of comparative zoology 



Mabuya brevicollis also 5. 
Mabuya megalura also 5, 7. 
Mabuya q. obsti 
Mabuya bayonii 
Mabuya v. varia also 5, 7. 
Mabuya striata also 5. 
Rio-pa sundevallii also 5. 
Riopa to. modestum also 5. 
Riopa anchietae also 7. 
Lygosoma gromieri (! Tsavo) 
Ablepharus wahlbergii also 5. 
Ablepharus megalurus 
Melanosepa ater also 2, 5, 7, 9. 
Acontias percivali 
Chamaeleon anchietae 
Chamaeleon senegalensis 
Chamaeleon g. gracilis 
Chamaeleon d. roperi also 5. 
Chamaeleon d. quilensis also 5. 
Chamaeleon d. dilepis also 5. 
Chamaeleon b. bitaeniatus also 8. 
Chamaeleon melleri 
Brookesia platyceps also 8. 
Brookesia temporalis also 8. 
Brookesia brevicaudata also 5, 8. 
Brookesia k. kerstenii also 5, 8. 

Leptotyphlops conjuncta also 5. 
Leptotyphlops emini also 5. 
Typhlops s. mucruso also 5. 
Typhlops s. excentricus 
Python sebae also 2, 5, 9. 
Eryx c. loveridgei 
Natrix o. olivacea also 2, 5. 
Glypholycus bicolor also 2. 
Boaedon lineatus also 5, 9. 
Lycophidion c. capense also 9. 
Mehelya c. chanleri also 5. 
Mehelya nyassae also 5. 
Chlorophis hoplogaster also 2, 9. 
Chlorophis neglectus also 2, 5, 9. 
Chlorophis irregularis also 2, 9. 
Philothamnus s. semivariegatus also 5, 

9. 
Philothamnus s. dorsalis 



Coronella coronata also 5. 
Coronella s. semiornata also 5. 
Coronella s. fuscorosea 
Grayia smythii also 2, 9. 
Grayia tholloni also 2, 9. 
Prosymna a. stuhlmanni also 5. 
Scaphiophis albopunctatus also 5. 
Dasypeltis scaber also 5, 9. 
Tarbophis s. semiannulatus also 5. 
Crotaphopeltis h. hotamboeia also 5. 
Crotaphopeltis degeni 
Chamaetortus a. aulicus also 5. 
Amplorhinus nototaenia 
Rhamphiophis acutus 
Rhamphiophis rostratus also 5. 
Rhamphiophis rubropunctatus also 5. 
Dromophis lineatus 
Psammophis punctulatus also 5. 
Psammophis sibilans also 5. 
Psammophis subtaeniatus also 5. 
Psammophis biseriatus also 5. 
Psammophis angolensis also 5. 
Thelotornis kirtlandii also 5, 9. 
Dispholidus typus also 5, 9. 
Calamelaps unicolor also 5, 9. 
Rhinocalamus dimidiatus 
Micrelaps bicolor also 5. 
Aparallactus jacksoni 
Aparallactus lunulatus 
Aparallactus guentheri 
Aparallactus capensis 
Aparallactus concolor also 5. 
Chilorhinophis gerardi 
Boulengerina a. storm-si also 2. 
Naja h. haje 
Naja melanoleuca also 9. 
Naja n. nigricollis also 5. 
Dendraspis angusticeps also 5. 
Causus rhombeatus 
Causus resimus also 5, 9. 
Causus defilippi also 5, 9. 
Vipera superciliaris also 5. 
Bitis worthingtoni 
Bitis arietans also 5. 
Atractaspis bibronii also 5. 



loveridge: African zoogeography and itinerary 



497 



Atractaspis katangae 
Atractaspis microlepidota also 5. 

Scolecomorphus kirkii also 9. 

Xenopus I. victorianus also 2. 

Xenopus I. borealis also 2, 7. 

Xenopus muelleri also 2, 5. 

Bufo car ens 

Bufo r. regularis also 6, 7, 9. 

Bufo vittatus 

Bafo parkeri 

Bufo urunguensis 

Bufo ushoranus 

Bufo t. taitanus 

Bufo micranotis also ?7. 

Chiromantis p. peter sii also 5. 

Chiromantis xerampelina also 5. 

Leptopelis bocagii 

Hylambates verrucosus ? 

Kassina senegalensis also 5. 

Megalixalus fornasinii also 2, 5. 

Megalixalus dorsalis also 2. 

Megalixalus fulvovittatus also 2. 

Megalixalus brachynemis also 2, 5. 

Hyperolius spp. 

Rana a. bravana also 2, 5. 



Rana f. angolensis also 2, 7. 

Rana floweri also 6. 

Rana o. oxyrhynchus also 2, 5, 7. 

Rana m. mascareniensis also 2, 5, 7. 

Rana ansorgii 

Rana stenocephala also 2. 

Rana occipitalis also 2. 

Rana edulis also 5. 

Rana d. delalandii also 5. 

Rana macrotympanum 

Rana ornata 

Rana pulchra 

Arthroleptis bottegi (if it occurs) 

Arthroleptis s. stenodaclylus also 5. 

Arthroleptis whytii 

Arthroleptis moorii 

Arthroleptis minutus also 5, 7. 

Phrynobatrachus acridoides also 2, 5, 

7. 
Phrynobatrachus natalensis also 2, 5, 

7. 
Hemisus m. marmoratum also 5. 
Hemisus m. guineensis 
Breviceps mossambicus also 5. 
Spelaeophryne methneri also 5. 
Phrynomerus bifasciatus also 5. 



Zone 7. Grassy Uplands and Alpine Meadows 

At high altitudes in East Africa it is usual to encounter extremes of 
temperature which, on the plateaus particularly, are correlated with 
the presence of a usually treeless grassland. Nairobi (5,452 feet) is 
situated where the savanna steppe and this type of upland meet; in 
consequence it possesses a herpetofauna which is preponderatingly 
that of the savanna with a small admixture of characteristically 
grassland species. 

On reaching the alpine zone it is only natural to find the poikilo- 
thermous fauna much reduced. Such hardy species are marked by an 
asterisk (*) on the following list. Among them are two races of cha- 
meleon, occurring on Mounts Ruwenzori and Elgon, which are somewhat 
doubtfully placed here as they are most abundant along the upper 
limits of the forest. To restrict them to zone 8, however, would be 



498 



bulletin: museum of comparative zoology 



artifically curtailing their actual habitat for they do occur on the small 
shrubs and tree heaths scattered through the alpine meadows. It is 
possible that recognizable forms occur at a similar altitude on Mounts 
Kenya and Kilimanjaro; whether the names which have been proposed 
for bitaeniatus from these mountains are really applicable