°\0

HARVARD UNIVERSITY H

Library of the

Museum of

Comparative Zoology

Bulletin of the Museum of Comparative Zoology

AT HARVARD COLLEGE

Vol. LXXV, No. 1.

REPORTS ON THE SCIENTIFIC RESULTS OF AN

EXPEDITION TO THE SOUTHWESTERN HIGHLANDS

OF TANGANYIKA TERRITORY

INTRODUCTION AND ZOOGEOGRAPHY

By Arthur Loveridge

With Three Plates

CAMBRIDGE, MASS., U. S. A:

PRINTED FOR THE MUSEUM

January, 1933

No. 1 . Reports on the Scientific Results of an Expedition to the Southwestern Highlands of Tanganyika Territory

I Introduction and Zoogeography

By Arthur Loveridge

The following remarks are intended to serve as an introduction to the series of reports which have been prepared on collections made by the writer during an eight months' safari in East Africa on behalf of the Museum of Comparative Zoology.1

There is much invertebrate material still awaiting study and on which it is hoped that further reports will appear. The mollusca are being utilized by Dr. Joseph Bequaert and W. J. Clench for their revisionary work on the genera of African land and freshwater shells. The paper on nematodes deals with only a portion of the parasitic worms collected. In addition to these invertebrates, and seven thousand, four hundred and eleven vertebrates, about half a ton of ethnological speci- mens were purchased, labeled and brought back for the Peabody Museum of Harvard University. The personnel of the expedition consisted of the writer and three native assistants, whom he had pre- viously trained in the preservation of mammals, birds and reptiles.

Objective of the Expedition

The purpose of the journey might be said to be threefold. The primary object was to add to our very scanty knowledge of the her- petological fauna of the chain of mountains in Tanganyika Territory, commonly referred to as the southern and southwestern highlands, with a view to throwing light on the composition of the unusually interesting fauna occurring in the Uluguru Mountains of East Central Tanganyika.

As a secondary consideration special attention was to be paid to the zoologically little known range flanking the eastern shore of Lake Nyasa and known to cartographers as the Livingstone Mountains. To the local inhabitants, however, this name is utterly unknown; they refer to its various sections by the name of the inhabiting tribe, thus the northern portion is called the Ukinga Mountains as the Wakinga dwell there.

•This expedition would have been impossible without a grant of half the expenses from the Carnegie Institution of Washington, the sum involved originally voted by the Carnegie Cor- poration of New York. It is a great pleasure to tender our grateful thanks and acknowledge- ments to these institutions and their presiding officers, Dr. John C. Merriam and Dr. F. P. Keppel. (T. Barbour, Director.)

4 bulletin: museum of comparative zoology

It seemed a pity to proceed on so expensive a safari, costly because so far removed from the railway and easy communication with the coast, without taking advantage of the opportunity to secure certain species still lacking in the collections of the Museum of Comparative Zoology; particularly species of uncertain taxonomic status whose position would be more firmly established if adequate topotypic series composed of both sexes could be secured. With this object in view a return itinerary was planned which would embrace certain type localities. This then formed the third motive of the trip.

Perhaps it would have been better to have attempted less. So great an area had to be covered, occasionally by such primitive methods of travel as walking, that inadequate time in some in- stances only two days had to be allowed in which to collect topo- types of such elusive genera as Typhlops, Leptotyphlops, and Am- phisbaenula. At the same time a certain measure of success attended these efforts and topotypes of a large number of species unrepresented in any American museum were secured.

As to the second objective referred to above, viz. a zoological recon- naisance of the whole of the Livingstone Mountains, I failed, for only three weeks were spent in the range and all of these were in the temperate rain forest section surrounding Madehani at the northern end of Lake Nyasa. My reasons for abandoning a thorough examina- tion of the range were numerous. After a couple of short journeys southwards along the range I came to the conclusion that anything like a comprehensive study of the fauna would be impracticable in the time at my disposal. The steepness of the mountain sides to be negotiated and the wearisome detours necessitated by some physio- graphical feature were such that in a day's march porters could not cover much more than five miles as the crow flies. Probably a straight- forward march down the centre of the range would occupy between one and two months. Anyone attempting a zoological survey of the Livingstones should be prepared to devote at least six months to the undertaking. Judging by the number of new forms found at Madehani, a thorough investigation would be productive of new races or species, but these would be more likely among the mammals and birds than among the reptiles.

The fauna is not rich and is unlikely to differ in any important respect from that of the adjacent Ubena Highlands through which we passed, or on the other hand from that of the Shire Highlands and Nyasaland Plateau into which it merges. Such parts as I saw were subjected to intensive cultivation, the valleys and hilltops were in

loveridge: African zoogeography o

places quite thickly populated. The Wakinga are exceptionally in- dustrious and grow European wheat so successfully that natives from all the surrounding country, even a hundred miles away, pro- ceed to Ukinga to purchase flour. Most of the uncultivated ground was rolling grassland over which flocks of cattle and goats wandered in charge of little goatherds who were unhampered by clothing. Missionaries, whose work took them on itineraries in the mountains, as well as the Forest Officer at Tukuyu, informed me that there were no areas of primary forest left except those in the vicinity of Madehani where I camped.

The southwestern and southern highlands from Rungwe east to the north end of Lake Nyasa and northeast to Iringa, have a mean altitude of about 4,000 feet, though Rungwe Volcano itself rises to 9,850 feet and the highest peaks of the Ukinga Mountains are 9,600 feet.

Except for their lower altitude and less precipitous slopes, the high- lands south of Iringa are not greatly dissimilar from parts of the Ukinga Mountains. They consist of a pleasantly undulating hilly country though in places steep-sided hills occur and the whole plateau is bounded in parts by lofty escarpments. The undulating country is largely grassland where the Wahehe graze their herds; the steeper hills are often clothed in dense scrub and bush of secondary growth while surviving patches of virgin forest are rare and widely scattered. This latter feature distinguishes them from the greater areas of rich primary forest occurring on Rungwe and the Ukinga range.

It is more than a coincidence that the heavily forested areas are those possessing the greatest rainfall; thus the mean rainfall at Iringa, based on records of a period of eleven years, is 26.57 inches; that of Tukuyu, lying near the foot of Rungwe Volcano, is 91.51 inches, this figure being the mean for seventeen years. This is the greatest rainfall for any part of the Territory and may in part be attributed to Tukuyu being situated between the great lakes of Nyasa and Tanganyika, whose heavy evaporation under a tropical sun is precipitated by the adjacent forested peaks. It is interesting to note that the next highest rainfall records are from Amani in the tropical rain forests of the Eastern Usambara range where the records kept for a period of eighteen years give an average of 80.09 inches per annum.

No figures are available for the Uluguru Mountains lying south of the Usambara and northwest of Iringa but they are not likely to be far short of the Amani average. The Uluguru support a similar forest to the Usambara, and are important as being the most southerly

6 bulletin: museum of comparative zoology

area of tropical rain forest of any extent. This forest is readily dis- tinguishable from the temperate rain forest of the mountains to the south. The highest peak of the extensive Uluguru massif reaches 8,000 feet. On their northern aspect these mountains rise relatively steeply from the plains, but to the south they pass gradually into hilly country not unlike that of the Iringa Highlands.

Itinerary

The following detailed information concerning the camps at which collecting was carried on, has been arranged in the order of the itiner- ary; the same arrangement has been observed in the listing of the material in the systematic reports dealing with the specimens collected.

After the name of the locality, the approximate altitude of the camp is given, followed by more precise data as to the position of the camp and the period during which collecting was carried out in the vicinity. This is intended to serve as a check to the dates on the labels accom- panying each specimen in case the figures become defaced or illegible with the passage of time.

The climatic conditions are of such outstanding importance in the collecting of lower vertebrates that the meteorological aspect of each camp during our stay is given in detail.

Mention is then made of some characteristic forms or rare species to enable a taxonomist to visualize the faunistic features of the neighborhood.

In the systematic papers dealing with the terrestrial vertebrates, the local names applied by the various tribes to the creatures taken in their vicinity, have been inserted. To be confident of absolute accuracy in regard to such names one really should live among the particular tribe for many years; this not being possible I have en- deavoured to take reasonable precautions to secure accuracy but it is not to be expected that some errors will not have crept in. Every native is not a zoologist but every native in his desire to be obliging is apt to call an animal by the name he thinks most applicable, if he should be unfamiliar with the correct one. To avoid such errors, specimens were submitted to groups of natives who argued or dis- cussed alternative names before submitting the final opinion to me. For example, at the conclusion of my stay at Madehani, the whole congregation visited my camp at the end of the morning church service and were shown an example of each species taken in the neighborhood. At Ilolo, through the exceeding kindness of Herr Gemusens, I was

loveridge: African zoogeography 7

able to exhibit representative specimens of the Ilolo and Rungwe collections to the whole school of the Moravian Mission and had the additional advantage of Herr Gemusen's own unrivalled knowl- edge of the Kinyakusa tongue.

When a species has been recorded previously from any of the localities visited during the course of the trip, this fact is entered under the heading Distribution, together with the name of the author who recorded it.

Dar es Salaam, Dar es Salaam District, Usaramo. Alt. 100 feet.

The capital and chief port of entry of Tanganyika Territory.

In hotel November 4th to 7th and 18th to 19th, 1929.

A little rain fell.

My time being fully occupied in unpacking crates and repacking their contents into safari boxes of portable size, getting licenses, etc., I made no attempt at collecting. Salimu, however, was sent to get large series of Lygo- dactylus p. picturatus and Cryptoblepharus b. africanus, in which he was suc- cessful.

On our return from Bagamoyo we reached Dar es Salaam at 3.30 p.m. and left at 11.30 next morning having railed all equipment in the interval.

Bagamoyo, Bagamoyo District. Alt. 100 feet.

Situated on the east coast opposite Zanzibar and forty miles north of Dar es Salaam.

Camped from November 8th to 18th, 1929.

During the last week of October two heavy showers had occurred and each morning from the 10th to the 14th one or two rainstorms swept the town but quickly passed so that an hour or two later it was difficult to believe that rain had fallen, so rapidly was it absorbed by the hot dry sand.

In consequence of this aridity the only spots which rewarded the digger were at the base of bananas in the native gardens' a little further inland. These plants seem unable to flourish at Bagamoyo unless planted in a pit. Half-a-dozen such pits were examined and yielded Hemisus m. marmoratum squatting upon its eggs or tadpoles in the moist soil at the base of the bananas. Several Typhlops s. mucruso were secured in this type of environment.

Following the Ngeringeri road after it crosses the Ruvu River, one reaches open woodland about three miles on the far side of the ferry. It was here that most of the bird collecting was done and a Pachycoccyx validus shot. Parrots, plantain-eaters and hornbills occur, though they were not abundant; on the other hand several species of bee-eaters were very common.

Bagamoyo is type locality for Leptotyphlops braueri (Sternfeld), Rhampho- leon boettgeri and Hylambates argenteus of Pfeffer; it was in the hope of securing topotypic series of these that Bagamoyo was visited, but without success.

8 bulletin: museum of comparative zoology

Morogoro, Morogoro District, Ukami. Alt. 1,628 feet.

A station on the Central Railway of Tanganyika, 126 miles west of Dar es Salaam. Principal town of the Ukami country situated at the foot of the Ulu- guru Mountains.

In hotel November 20th, 1929.

Only a shower or two had fallen during the month and everything was very parched and dry. It was the more surprising therefore to secure close to the station and with little difficulty a topotypic series of Megalixalus loveridgii (syn. of M. fornasinii) including a number of young specimens.

A couple of hours spent in turning logs, stones and debris gave no results though in digging at the base of a banana a queen soldier ant or siafu (Dorylus helvolus) was unearthed much to the annoyance of the myriads of workers and warriors. Dr. W. Morton Wheeler, who kindly made the identification, tells me that this is the first female of its species in any collection in the United States.

Mpwapwa (Mpapwa, Mpapua), Dodoma District, Ugogo. Alt. 4,000 feet.

Situated eighty-five kilometres east, slightly southeast, of Dodoma and about ten miles north of Gulwe (Igulwe) station on the Central Railway.

Camped from November 21st to 23rd, 1929.

At this place the lesser rains usually commence about November 15th but the countryside was desperately parched at the time of my arrival for only one small shower had fallen and that a week before our visit. The Vet- erinary Department very kindly placed their newly built and unoccupied office at my disposal for the two and a half days that I spent at Mpwapwa. This office faces a belt of fine trees, wild fig predominating, which fringe the water course that comes down from the 6,000 foot range behind. The river was, of course, dry, but a very small stream, whose source was a spring in the river bed, trickled down the course.

The domestic bananas opposite the office were devoid of frogs, and their roots, grounded in sandy soil, were dry. Along the edges of the stream we secured a very few young Rana f. chapini and near the head of the valley, where conditions were moister, hundreds of Arthroleptis xenodactylus hopped about among the dry leaves which formed a belt along either side of the stream. A great many dead tree trunks, situated more or less near the stream, were broken open and examined and the ground beneath them dug up but only one of the whole number produced any reptiles. The excavation of this one resulted in the removal of a good cartload of rubbish and uncovered a Paragonatodes quattuorseriatus, Amphisbaena sp. n., Melanoseps ater and Prosyrnna ambigua. Of these the lizards as well as A. xenodactylus are crea- tures associated with mountain rain forest and it is obvious that at Mpwapwa we have a remnant of virgin forest fauna in a region that is fast undergoing desiccation.

To the east of the office there is open thorn bush growing on a volcanic ash soil of rufous color; in places sand overlays the soil to a depth of six inches

loveridge: African zoogeography 9

and everything was so dry that the soil was like caked dust. A broad river bed occupied the open valley bottom and on one bank of this was the charred stump of an old tree which yielded results. All the other dead trunks and stumps in the vicinity were non-productive without a single exception. We spent two hours in digging out the decayed roots of this giant tree and re- moved a ton of soil altogether. It struck me as curious that so many species should be represented by only single specimens. The catch from this one tree was: Rhinocalamus dimidiatus (topotype), Causus defilippi, Hemidactylus w. werneri, 2 Riopa s. modestum (topotypes), 3 Ablepharus wahlbergii, 2 Bufo r. regularis, 3 Arthroleptis s. stenodactylus besides many invertebrates scor- pions, centipedes and polydesmids.

Mpwapwa is also type locality for Geocalamus modestus and Arthroleptis scheffleri in addition to the topotypic species secured.

Kilimatinde, Manyoni District, Ugogo. Alt. 3,591 feet.

Situated fifteen kilometres south of Saranda station on the Central Railway and southeast of Manyoni which is 4,160 feet.

At Mission on November 26th and 27th.

The rains had failed except for a very few showers, in consequence every- thing was parched and the fields of stubble were blanketed with dust; the areas of red volcanic soil were somewhat more compact.

Five hyrax (Heterohyrax brucei prittmtzi) were shot one evening in this, its type locality; they yielded a rich harvest of parasitic worms. Other mam- mals seen were Papio neumanni, Cercopithecus a. johnstoni, Myonax grantii, M. s. proteus, warthog, bushbuck, duiker and dikdik, the last of which were very common.

Reptiles were scarce except for H. mabouia, A. a. dodomae and M. v. varia which occur upon the rocks in the dry watercourses. A Chameleon d. dilepis and Chiromantis p. petersi (topotype of C. pictus Ahl) were taken upon a Manyara hedge. The only snakes seen were two Psammophis biserialus and a Naja nigricollis.

Saranda, Manyoni District, Ugogo. Alt. 3,511 feet.

A station on the Central Railway between Dodoma and Tabora.

Camped from November 28th to 30th and December 18th to 19th, 1929.

No rains had fallen at the time of our first visit though daily threatening and dust storms whirled across the countryside. The types of country in the vicinity of Saranda are numerous, open thorn bush in the vicinity of the station, miles of almost impenetrable scrub to the south while maiombo bush of varying density flourished on the dry and stony escarpment to the northwest.

Though a halt was only made at Saranda to change from train to motor lorry it was hoped to secure topotypic material of Guttera edwardi granti and Lygodactylus manni, two species of doubtful status. In this we were success-

10 bulletin: museum of comparative zoology

ful, with the result that granti which has been discredited, is considered valid and manni becomes a synonym of picturatus.

Unyanganyi {Kinjanganja) , Singida District. Alt. circa 4,500 feet.

Situated between Singida and Kondoa Irangi due north of Saranda.

Camped from December 3rd to 9th, 1929.

The first of the lesser rains fell on the afternoon of the 5th., but was quickly absorbed by the parched ground though some pools remained in the mbugwe. It rained steadily for a couple of hours on the night of the 5th but remained fine thereafter.

This place was visited, since "Kinjanganja am Turu" is the type locality for the smallest and supposedly one of the rarest skinks in East Africa, Able- pharus megalurus Nieden. Kinjanganja is synonymous with Unyanganyi, no definite locality but an area of the former German district of Turu, partly inhabited by the Wanyaturu and Wataturu. In the Unyanganyi country there are various scattered groups of huts usually known by the name of the petty chief or jumbe. The jumbes whose names appear on German maps have since died or been superseded and some are almost forgotten already. My camp was half-a-mile south of Jumbe Abdulla's of Kifumbu. There are a group of shops beside the main road from Singida to Kondoa Irangi, about fifteen miles east of the former and I was assured that this was the approxi- mate centre of Unyanganyi. This village is at the north end of an "island" of raised ground with a kopje occupying the centre, the raised ground is mostly red volcanic ash though so sandy in spots as to be like a desert. It is surrounded by a very extensive mbugwe (plain) of black (or gray) cotton soil which being very desiccated at the time of our visit is full of fissures. The plain is a mile or more broad and is backed to the north and west by a steep escarpment.

Drinking water was obtained from holes dug ten feet deep in this mbugwe and was like thick soup by reason of the gray mud in suspension. Each afternoon clouds collected in the east and hurricanes of wind raised "dust devils" which whirled through camp leaving one choking, and a deposit of dust, sand and straws over everything.

Masiliwa, Turu, Singida District. Alt. circa 4,500 feet.

Situated just south of Jumbe Ali's village, Kalingwa on the Singida to Kondoa Irangi main road a day's march east of the camp at Unyanganyi.

Camped on the night of November 9th, 1929.

Several showers had occurred in this dry thorn-bush and rock-strewn country. Porters arrived at 3 p.m. with the loads and a heavy downpour lasted from 4 till 5 p.m., then steady rain from 7 p.m. through most of the night.

Rhinoceros were said to be a great pest here, waiting about the water holes and menacing natives and cattle; they certainly had wrought havoc with the manyara hedge fifty feet from my tent. Dikdik were seen and signs

loveridge: African zoogeography 11

of larger game. During the night a leopard remained very close to camp, coughing three times during that period. Lions were heard in the distance.

After dinner I took a lamp and walked to and fro over a recently hoed field resulting in the capture of a large series of several burrowing amphibia Rana delalandii, Hemisus marmoratum guineensis and one Breviceps mossam- bicus. Half-a-dozen of the latter were taken the following morning in open maiombo bush a few miles away, also a Boomslang was found swallowing a Leptopelis bocagii and had another in its stomach; it was of interest to note that all four species taken in this arid country were of different genera yet provided with shovel-shaped metatarsal tubercles to enable them to "dig-in" during the dry season.

Harida, Usandawi, Kondoa Irangi District. Alt. circa 4,000 feet.

On a cross-country trail from Kalingwa to Mangasini.

Midday halt on November 10th, 1929.

An open area cleared of maiombo bush for cattle grazing. A series of deep water holes in the valley bottom are surrounded by thorny fences or bomas of piled-up thorn bush. To the south of these water holes, acacia thorn bush of considerable extent followed by open mbugwes; a promising game country but none seen.

Maji Malulu, Usandawi, Kondoa Irangi District. Alt. circa 4,000 ft.

On a cross-country trail from Handa to Mangasini.

Camped on the night of November 10th, 1929.

Slight showers had fallen and one storm during the night of our stay which was from about 4 p.m. to 7 a.m.

Tents were pitched beneath a baobab, one of several in a large area of cleared land in which were a dozen native tembes. To the south this area was surrounded by thorn bush on gray soil, to the north some maiombo forest on red soil. Water is said to be very scarce hereabouts.

Though I went out with a local native from 5 to 7 p.m. little of interest was seen and the only reptiles collected were a gecko (H. w. werneri) and an Egg-eating Snake (Dasypeltis scaber).

Mangasini, Usandawi, Kondoa Irangi District. Alt. circa 4,000 feet.

Mangasini is a native corruption of the word Magazine and was given to the village which was used as a store base by the German troops during the earlier part of the East African Campaign. A few miles distant is the German farm of Kwa Mtoro, type locality of Werner's Mabuya obsti, a syno- nym of M. quinquetaeniata. It was with the object of securing topotypic material of this skink that Mangasini was included in the itinerary but no trace was found of this skink on the kopjes about Kwa Mtoro though it un- doubtedly occurs there.

Camped from December 11th to 16th, 1929.

12 bulletin: museum of comparative zoology

Hardly any rain had fallen up to the time of our arrival and the whole countryside was parched and dusty. At 5 p.m. on the 12th it began to rain and after sunset there was a terrific thunderstorm and the rain came down in torrents continuing without cessation until noon of the following day. This storm awoke the amphibian and insect life and enabled us to secure series of things which under normal conditions would have been impossible.

A small undescribed species of Bufo was found in the mbugwe where thirty were collected, here also Rana floiveri of the Sudan was found in surprising numbers. This was also the case with Chiromantis p. petersi for these arboreal frogs assembled and started making their froth "nests" about the accumu- lations of rain water.

The large black scorpions (Pandinus cavimanus) were common, as also the Trombid mites of a plush appearance which the natives call "the child of the rain" as they make their appearance after showers.

Kikuyu, Dodoma District, Ugogo. Alt. 3,900 feet.

Kikuyu is but a mile and a half south of Dodoma whose physical character- istics I have recently described (Loveridge, 1928, Proc. U. S. Nat. Mus., 73, Art. 17, pp. 3-4). This Kikuyu should not be confused with Ikikuyu to the south of Gulwe station where I collected in 1923. Specimens collected from here are just as much topotypes of species accredited to Dodoma as if they were so labelled as they were hunted in the intervening area.

Camped from December 21st to 26th, 1929.

The weather was fine for three of the four days spent here but little collect- ing was done, the halt before proceeding south being made to enable me to purchase Wagogo ethnological material and to pack and dispatch to the coast the results of the past six weeks collecting. We arrived late in the afternoon of the 21st and left before 9 a.m. on the 26th, motoring to Iringa which was reached at 10 p.m. the same night.

Topotypes were collected of Elephantulus renatus, Pedetes dentifer and Agama a. dodomae as well as an undescribed species of Leggada found running about the road at night.

Dabaga, Uzungwe (Utschungive) Mountains, Iringa District. Alt. 6,000 feet.

Situated forty miles south of Iringa in south central Tanganyika though frequently spoken of as the southwestern highlands. Uzungwe was spelt Utschungwe by the Germans and undoubtedly Uhehe was often given as a type locality for things coming from this region where Wahehe are settled. The altimeter reading for the camp, situated half-a-mile below Houter's farm, was 6,025 feet which was in accordance with various survey calcula- tions made in the vicinity.

Camped from evening of December 27th, 1929 to January 4th, 1930.

Sunday was the only entirely fine day during the week; once it rained almost continuously for thirty-six hours. As a general rule the mornings

loveridge: African zoogeography 13

were fine with rain between noon and 3 p.m. and frequently rain again at 5 or 6 p.m.

The country is composed of rounded rolling hills covered (at the time of my visit) with freshly springing grass and sometimes with shrubs. Many of the hills have clumps of trees scattered here and there with dense thickets at their base, others are studded with shrubs for the most part under six feet in height, the dominant kind being a species of Protea. Others again are densely clothed in stunted forest, the trees twisted and gnarled by the action of lianas and rarely exceeding thirty feet in height, the undergrowth of brambles and, on the outskirts bracken, is so dense that it is quite impene- trable for collecting purposes. Though scheduled as rain forest it can hardly be considered primary tropical rain forest and the timber is useless except for fuel.

Dabaga becomes the type locality for three new reptiles which we collected, the finest being a new tree viper, Athens harbour i. The others were races of Lycophidion capense and Chamaeleon werneri. Topotypes of Ch. goetzi and Ch. tempeli tempeli were also secured.

Kigogo, Uzungwe Mountains, Iringa District. Alt. 6,000 feet.

Altimeter readings showed considerable variations attributable to rain and temperature fluctuations. The average was just under 6,000 feet but Survey Department reading for points 500 yards below and above the camp were 6,090 and 6,234 feet respectively.

Camp was actually situated a hundred yards behind the Forest Officer's house which is near the extreme southern end of the Uzungwe Mountains and a few miles from Mufindi. I was advised that this was the only large remnant of forest in the southern part of the range. During my stay much assistance and kindness was given to me by Mr. Fraser, the Forest Officer. Kigogo takes its name from the river which flows past the Forestry quarters, plantations and nurseries.

Camped from January 11th to 31st, 1930.

Only one or two days were entirely free from rain though much of it was in the form of mist or fine driving rain. There was more sunshine between showers than one experienced at the same altitude in the Uluguru Mountains; drying of skins was noticeably better.

While the country might still be called undulating, many of the hillsides were more precipitous than at Dabaga. Shrubs were also scarcer, the hillsides being covered with bracken and grass and the tops capped with forest. Over considerable areas the forest survives on the caps of the hills and is both higher and more open than the forest in the immediate vicinity of Dabaga. Bamboo was apparently much more extensive and quite large patches of it occurred on some of the hillsides. A striking feature of this temperate rain forest was the abundance of moss which clothed the trunks and branches of the trees and from which long wisps of moss depended.

As the result of our visit Kigogo becomes type locality for five new forms or species of Cryptomys hottentotus, Francolinus squamatus, Apalis thoracica

14 bulletin: museum of comparative zoology

and Chamaeleon, paratypes of three other new things were also taken there.

Only four species of reptiles were at all abundant, these were: Duberria I. shiranum, Trimerorhinus t. tritaeniatus, Mabuya v. varia and Chamaeleon tempeli, topotypes of Ch. w. werneri were also taken as well as the interesting limbless lizard Melanoseps ater.

Amphibia were scarce except for three species, viz. Arthroleptis minutus, A. parvulus and Hyperolius marginatus besides which only six other species were taken, the rarest being a new race of Bufo taitanus.

Madehani, Ukinga Mountains, Rungwe District. Alt. 7,200 feet.

Camp was made among the ruins of the German Lutheran Mission a hun- dred yards south of the village which is situated in the mountains at the north end of Lake Nyasa.

Camped from noon on February 13th to daybreak on the 27th, 1930.

The average of hours of sunshine per day during the fortnight was cer- tainly not more than two. The routine of meteorological conditions was fairly regular during our stay. The day would dawn with a clear sky; about 8 a.m. the sky would be obscured by fleecy white clouds. An hour or two later a white mist would come creeping up from the lake, wisps of it would blow past about 10 a.m. and gradually thicken until 11 a.m. by which time we would be enveloped in a blanket of fog, raw and especially unpleasant if accompanied by a cold wind. By noon all the trees would be dripping with precipitated moisture, one's clothes quite wet by precipitation. At 1 p.m. a crack of thunder, followed a few minutes later by a downpour of rain, at times the latter continued until 4 p.m. or alternatively in a series of heavy showers with intervals of sunshine lasting from ten to fifteen minutes. From 4 till 6 p.m. it would not rain but the sky would be hidden by clouds, the vegetation would be sodden and everything clammy to the touch. When darkness fell at 7 p.m. it would begin to rain softly though on a few evenings it held off till 9 p.m. ; rain would continue on and off till 3 or 4 a.m. ; sometimes, but not generally, heavy downpours occurred during the night. Naturally both collecting and preservation of specimens was difficult under these con- ditions, and but for Dr. James P. Chapin's valued suggestion that I should take a Primus stove to dry the skins, they would undoubtedly have suffered.

Four new races of mammals and one of birds were found at Madehani, some of which at least apparently owe their differentiation to heavy rainfall and moist conditions. The genera involved are Aethosciurus, Praomys, Otomys, Claviglis and Illadopsis.

The forest consisted of fine large trees set far apart so that there was a more or less dense undergrowth of shrubs and grass. The trees were often heavily laden with moss and ferns. Here and there along the forest edge, or on the sides of ravines in the forest, were large patches of bamboo. A road traversed both forest and bamboo for a couple of miles and it was along this road, or in cultivated patches of former forest land, that we secured the series of Crotaphopeltis h. tornieri and Atheris barbouri, all the other snakes came

loveridge: African zoogeography 15

from the open grasslands and gardens. Lygodactylus angularis occurred on big trees along this road and on smaller isolated trees in what was obviously cleared forest land. The new species of chameleon as well as the three other kinds were found on shrubs or trees at the forest edge.

The fern-grown and most promising looking banks of the streams within the forest were entirely unproductive of amphibian life. We failed to find Nectophrynoides vivipara of which these mountains are part type locality. Arthroleptis reichei lived in the forest and A. schubotzi at the base of wild bananas just outside the forest. Wild bananas were abundant two miles down the road from our camp but a close examination of them failed to reveal any frogs. The Hyperolius marginatus were captured along the sides of swiftly- flowing streams in the valley bottoms of the grasslands without the forest, Rana f. angolensis was in a similar habitat while R. f. merumontana was taken in the grasslands and A. parvulus in boggy areas of the same. These bogs were studded with sundew plants.

Mwaya, Lake Nyasa, Rungwe District. Alt. 1,700 feet.

Much of the material from this locality is labelled "Near Mwaya" as camp was pitched three miles west of the village and lakeshore to avoid a percentage of the mosquitoes. Mwaya is just north of Karonga, Nyasaland on the north- west shore of Lake Nyasa, actually the village is separated from the lake by about a mile of swamps.

Camped from March 1st to 11th, 1930.

Rain was fitful; at first we had several entirely fine days with heavy rain at night, this period was succeeded by one in which downpours lasting an hour or more occurred both morning and afternoon; during the last two days of our stay there was almost continuous rain. When not raining the sun shone with great force.

As one descended from Madehani in the Livingstone Mountains, the last thousand feet or so closely resembled the hills just north of Kilosa station; the same open maiombo bush, the same red soil and gravelly paths. At 2,300 feet one passed through a ravine bordered by big trees where butter- flies typical of the Kilosa fauna Euphaedra neophron and Hamanumida daedalus settled upon the leaf-strewn path. Emerging from the ravine we marched for miles through sword grass precisely like that to be found on the Kilosa flats.

Camp was made beside the Mwaya-Tukuyu road at a village named Ndora where banana plantations were very extensive. The Mbaka River, type locality for a race of waterbuck described by Matschie, flowed close by and the rank vegetation which smothered its banks might have been that of the Myombo River near Kilosa.

To the south stretched plains of which great patches were inundated at this season and on part of which rice was being cultivated. All the animals collected at Mwaya were common to Kilosa except three forms of antelope which had their counterparts in Usagara.

16 bulletin: museum of comparative zoology

In the main the birds were also those of the Kilosa region but with an admixture of southern forms.

The snakes were all common and widely distributed species with the ex- ception of Dromophis linealus, Rhamphiophis acutus and Vipera supereiliaris. The taking of the two last mentioned species provided the first records for Tanganyika Territory.

There was little of interest among the frogs except the taking of Arthro- leptis whytii, Leptopelis johnstoni and Megalixalus brachynemis, all of which had been described from near Karonga just across the border.

Tukuyu, Rungwe District. Alt. 5,000 feet.

In 1918 Neu Langenburg reverted to its native name of Tukuyu; it is the capital town of the Rungwe District but is about twenty miles southwest of the mountain which gives its name to the district. It is forty miles by road to Mwaya on Lake Nyasa.

At rest house from March 12 to 14th, and April 18 to 23rd, 1930.

Frequent showers and driving mist occurred during our stay.

These brief stays at Tukuyu on the way to and from Rungwe and the Porotos were made to replenish stores, arrange for transport, and to perma- nently pack specimens obtained at Mwaya and Rungwe. As the transport expected on the 21st failed, though hampered by uncertainty as to its arrival, we managed to get some collecting done.

The region about Tukuyu consists of rolling, hilly country with a vege- tation and climate strongly reminiscent of the Kikuyu highlands in the vicin- ity of Nairobi. There is no forest but a great deal of planting of introduced trees has been accomplished. One ravine has been laid out as a public garden by a past administrator Major Carveth Wells, and wild bananas and other local forest plants or trees flourish in profusion and form a centre of attraction for forest-living birds which would be absent otherwise.

Typhlops s. mucruso and Boaedon lineatus were the only snakes collected but Sternfeld lists six species, among them two very doubtful ones, viz. Chlorophis irregularis and Psammophis notostictus, one wonders if these should not be Philothamnus s. dorsalis and Psammophis sibilans both of which were common at Mwaya.

Of amphibia we collected X. poweri, B. r. regularis, R. f. angolensis, R. m. mascareniensis, A. whytii and saw A. schubotzi. Nieden lists three of these and adds P. bifasciata, R. oxyrhynchus and Phrynobatrachus acridoides.

Ilolo, Rungwe District. Alt. 4,600 feet.

Camp was made for a week-end beside the village just below the Rungwe Mission and three miles below my subsequent camp in the Nkuka Forest. There is an uplands fauna at Ilolo, as distinct from that of the forest, and during the earlier part of my stay in the forest children came up from the village with specimens which were duly labelled "Ilolo," thus for the period

loveridge: African zoogeography 17

from March 24th to April 17th labels may read either Ilolo or Nkuka Forest for the same date. Once or twice Salimu or I went down to Ilolo and col- lected birds in the vicinity either coming or returning.

March 15th, 16th, 24th to 31st and April 1st to 17th, 1930.

Rain daily and heavy. Empty four-gallon kerosene drums placed any- where beneath the awning of my tent were full and running over on the morning of the 16th.

While the village is concealed among dense banana plantations, the sur- rounding country largely consists of open grassland savannah with a few scattered shrubs here and there; the general appearance being very similar to types of country in the vicinity of Nairobi. In the direction of the Poroto Mountains there is a steady rise and both streams and rivers tend to cut deep ravines which become choked with shoulder-high grass, brambles and stunted trees.

Nyamwanga, Poroto Mountains, Rungwe District. Alt. 6,400 feet.

Nyamwanga is an Usafwa (Usafua) village a hard day's march north of Tukuyu. It is sometimes known as Marupindi's village after the name of the chief.

Camped on the nights of March 17th and 20th on the way up and down from Ngosi Volcano.

There was heavy rain on both the afternoons that I was at this camp, it was cloudy and dull in the intervals between downpours.

Nyamwanga is situated in rolling grasslands, rising steeply to the moun- tains which surround it on three sides. Shrubs are common but trees are scarce in the immediate vicinity though plentiful three miles away; doubtless they have been cut for fuel and timber in the neighborhood of the village.

Not a snake or a lizard was brought in by the natives; on the other hand four species of chameleons were so abundant that I bought over a hundred in two hours among which were a good series of topotypes of C. fullebomi and a few of a new kind.

Frogs also were plentiful but no great variety of species so that it was necessary to limit purchases which were at the rate of a dollar per four hundred. At this price the children considered themselves well repaid, i. e. an East African cent for each frog.

From the above it will be seen that the people, who see very few Europeans, were friendly at unusually short notice and are quite keen to get money. Perhaps rather too keen, as the chief's clerk and tax collector visited my tent about midnight with a view to investigating the cash box; failing to achieve this, however, he spent six months in jail.

Ngosi Volcano, Poroto Mountains, Rungwe District. Alt. 7,170 feet.

The Crater Lake of this well-known volcano was Nieden's type locality for Rana fullebomi and Arthroleptis reichei and was visited with a view to

18 bulletin: museum of comparative zoology

securing topotype series in which we were successful; it is a three hour's march from Nyamwanga.

Camped on the narrow lip of the crater from March 18-20th, 1930.

We arrived at noon on the 18th in driving rain which continued without intermission until 9 a.m. the following day. It rained on and off during our stay with a minimum of sunshine. The sodden condition of the forest during the rainy season probably causes many of the birds and mammals to leave it for that period.

Colobus, Blue Monkey and leopard were the only animals of which we had evidence, the former we actually saw.

Birds were so scarce that I only observed four species during the three days we spent on the volcano, viz. Corvultur albicollis, Pseudoalcippe stierlingi, Batis mixta and a coot which was swimming on the crater lake. No birds were shot as generally they would have fallen from fifty to a hundred feet and