2019-03-14

Lamprologini...part 2



An account of Lamprologini, including Neolamprologus sp. “Mwila”, a new shell-dwelling species from Mwila Island in the Kipili Archipelago, the facts about N. sp. “Eseki”, the whereabouts of Kinyamkolo, and much more (part 2 of 3)

Part 1 – Part 2 – Part 3

32. Where are Moore’s Kinyamkolo and Mbity Rocks?


Fig. 118. A colour sketch by J. E. S. Moore, in all probability illustrating his collection location to the left, Kinyamkolo, the type locality of several cichlids, including Tropheus moorii (Moore, 1903: Fig. opposite page 68).

32.1. Introduction to Kinyamkolo and Mbity Rocks
The type series of N. modestus was collected by Moore during his first Tanganyika expedition in 1895–1896, one specimen at Kinyamkolo and another at Mbity Rocks (Boulenger, 1898b: 8–9). Besides those two locations, scientific collections were made during the same expedition at Kituta Bay (Moore, 1898d: 182; Cunnington, 1899: 697), Sumbu (Boulenger, 1898b: 25), and Mleroes (Moore, 1898b: 452; 1898e: 92). Kituta is currently referred to as Chituta, Sumbu commonly Nsumbu (cf. Staeck, 2017: 17, Fig. bottom, right) and occasionally Msumbu (Hore, 1882: map; 1883: 17; Andree, 1905), and Mleroes is better known as Moliro. Moliro has also been referred to as Muriro’s Villages (Stanley, 1978b: 34; Hore, 1882: map; 1892: 226, map), Pamlero, Pamlilo (Stevenson, 1888: first map; Hore, 1889: map; Johnston, 1897: foldout map facing page 40), Pamriro (Kettler & Riemer, 1890–1895), Mriro’s (in 1892) (Luscombe, 2018a: map), Muriro (Elliot, 1896: map; Stanley, 1899c: foldout map), Pamlelos/Pamlelos (Moore, 1897a: 299 & map), Mouliro (Foà, 1899: 337, foldout map), Mlelos/Mleloes (Moore, 1901b: 110, 128), Moliros (Moore, 1903: map opposite page 8), and Moliro’s (Stewart, 1903: map opposite page 222).
In addition to these five locations (Kinyamkolo, Mbity Rocks, Chituta Bay, Sumbu/Nsumbu, and Moliro), Moore obtained samples from Lake Mweru (Boulenger, 1898c: 479; cf. Boulenger, 1899b: 84) and possibly Lake Bangweulu, west of Lake Tanganyika, as well as Lake Chilwa, south of Lake Malawi (cf. Troyer, 1991: 38). Also, one may assume that some collections were also made in Lake Malawi, which he surveyed, in company with Edmund L. Rhoades, for seven weeks in 1895 (Moore 1901a: 7; 1901b: 50; cf. Moore, 1897b: 627; 1898b: 453; 1898c: 167). For a photo of the Malawi boats that Moore presumably used, see Johnston (1897: 111). Moreover, from his camp at Kinyamkolo, Moore made several journeys by boat travelling as far north as Karema, and by land marching to the mountains east of Lake Tanganyika overlooking the northwestern end of Lake Rukwa (Nature, 1897: 258)
In June 1896, Moore was in Cameron Bay, near Sumbu, where he collected only a single fish sample, later described as Synodontis multipunctatus (Boulenger, 1898b: 25), presumably because he was interested primarily in sedimentary rocks and endemic gastropods rather than fish (Moore, 1897a: 298–299; 1898d: 182). Having restored a scrapped steamship, the Good News (Fig. 124) (§ 32.4, 32.17, 32.18), which during his second expedition would take him to Sumbu, Moore thought that it “would at any rate suffice for me to get my shells and jellyfish” (Moore 1901b: 110). Besides being a knowledgeable malacologist/conchologist (e.g. Moore, 1898c; 1898d; 1899a; 1899b; cf. Digby, 1902: 434), Moore was a devoted supporter and avid developer of the hypothesis, coined by Thomson (1882: 622–623; cf. Smith, 1881: 276) and first put forward in more detail by Günther (1894: 289–290), that Lake Tanganyika and parts of its fauna derive from the ocean (Moore, 1898g), a hypothesis that was partly based on the marine-like gastropods found in the lake. Furthermore, it may be worth pointing out that Sumbu “the British port on Cameron Bay”, which in 1900 was “a well-built station, second to none on the lake” and included a “small township” (Codrington, 1902: 601), probably was a welcome break for Moore, who spent much of his time at Niamkolo (Nature, 1897: 258), a notoriously unhealthy and unattractive place (Purves, 1898: 105; Johnston, 1890: 738; Ragsdale, 1986: 24; Horne, 1908: 268). Nevertheless, the main reason for Moore’s travel to Cameron Bay appears to have been to look for the marine-like gastropod Tiphobia horei, which he could not find in the waters off Kinyamkolo Islands (Mbita Island and Mtondwe Island) near where he was camped, or of which he could only find empty shells on the shores of Chituta Bay (Moore, 1898d: 182). Near Sumbu or somewhere in Cameron Bay, Moore was successful and caught 100 live specimens of T. horei, as well as other marine-like gastropods, some of which he described as Bythoceras iridescens (Moore, 1898b: 452; 1898d: 183; 1898e: 93), now in Paramelania.
Regarding Moore’s fish collection of 1895–1896, all samples basically originate from Kinyamkolo and Mbity Rocks (§ 32.25), except for the single specimen of S. multipunctatus from Sumbu. According to Boulenger (1906: 539), “Mr. Moore’s first collection, made in 1895–96, contained about 90 fishes, referred to 33 species, 25 of which were described as new” (see Appendix 4). Obviously a rather small collection, which Moore partly blamed on the difficulties of transport by carriers, resulting in the loss of several jars containing spirit-specimens, “through several hundred miles of often trackless, always scorching, forest” (Moore, 1898a: 26), a little more than 45 specimens of 18 cichlid species were from Kinyamkolo, while about 23 specimens of 13 cichlid species were from Mbity Rocks (Boulenger, 1898b: 2, 7–21). For photos taken in the 1890s of the better part of the road between Lakes Tanganyika and Malawi, on which Moore travelled, see Johnston (1897: 142, 143) and Moore (1901b: 49).
Contrary to the locations of Chituta, Sumbu, and Moliro, which were clearly indicated on one of Moore’s (1897a) maps, the exact locations of Mbity Rocks were never detailed by Moore, not on maps, nor in writings. Regarding Kinyamkolo, for about 100 years it was widely understood as synonymous with Niamkolo, which is now Mpulungu, the consensus view also for authors of Lake Tanganyika cichlids (§ 32.9) until Drachenfels (1995: 257–258) suggested Kinyamkolo not to be synonymous with Niamkolo/Mpulungu, but pertaining to a location or area on the west coast of Mbete Bay, possibly Cape Chaitika. Furthermore, Drachenfels (1995: 255–257) stated, erroneously, that Moore never indicated the exact location of Kinyamkolo (see more below, § 32.3).
Recently, Konings (2012; 2013a; 2013b; 2013c; 2015a: 198) presented similar ideas to those of Drachenfels (1995), suggesting Kinyamkolo is not synonymous with Mpulungu, but pertains to a greater area, possibly a district. Konings believes Moore’s view of Kinyamkolo more specifically relates to Ulungu Escarpment, the mountainous west coast of Mbete Bay, including Cape Chaitika in the north. A few arguments in favour of these suggestions have been put forward by Konings, all of which will be dealt with in this last part of the article. At the end of it, it will be concluded that Moore’s Kinyamkolo is not a district, nor synonymous with Ulungu Escarpment, but, as regarded previously, synonymous with Niamkolo and Mpulungu.

32.2. Early synonymisation of Kinyamkolo and Niamkolo
In scientific circles, the synonymisation of Kinyamkolo and Niamkolo began with Cunnington’s expedition. Upon Cunnington’s return to London, his collections of animals and plants were distributed among colleagues for systematic examination, with Moore’s Kinyamkolo assumed to be synonymous with Cunnington’s Niamkolo. Regarding Cunnington’s localities in general, Boulenger (1906: 537) wrote that they “appear here under a spelling somewhat different from that adopted by Mr. Moore”. However, Kinyamkolo and Niamkolo were not synonymised in Boulenger’s ‘Catalogue of the Fresh-Water Fishes of Africa’ (1915). Nonetheless, at the time of Cunnington’s return and until 1908, Moore was still producing scientific papers and holding lectures about his two Tanganyika expeditions (Troyer, 1991; Troelstra, 2016: 310), but never pointing out that Kinyamkolo is not synonymous with Niamkolo, likely because he knew that they were. Furthermore, in his scientific articles, he never explicitly mentioned the locations of Kinyamkolo and Mbity Rocks, probably because they were where one would expect and interpret them to be, i.e., Moore’s Kinyamkolo is synonymous with Niamkolo, currently Mpulungu, and Mbity Rocks are some rocks in shallow water near the village Mbete.

Fig. 119. Map showing William Alfred Cunnington’s route on Lake Tanganyika in 1904–1905, detailed with dates and origin of his lamprologin samples, all of which Boulenger assigned to Lamprologus (shortened ‘L.’). The abbreviation ‘Le.’ means Lepidiolamprologus, the prevailing generic name for the species currently referred to as Le. “Meeli-Boulenger”, identified by Boulenger as L. hecqui (Nos 1 and 2). Also, Cunnington’s sample of Le. profundicola was misidentified as L. elongatus (No 3).
The map is based on reports given by scientists who worked on Cunnington’s collections, including Beddard (1906), Boulenger (1906), Calman (1906), Cunnington (1906a; 1906b; 1906c; 1907; 1913; 1914; 1920), Fuhrmann and Baer (1925), Günther (1907), Halbert (1906), Kirkpatrick (1906), Laidlaw (1906), Rendle (1907), Rousselet (1907; 1910), Sars (1909; 1910; 1912), Smith (1906), Stebbing (1912), and West (1907); see also Nature (1905; 1922). According to Boulenger (1906: 539), ‘Cunnington succeeded in bringing home 300 specimens, referred to 84 species, 27 of which were new. In comparison, Moore’s first collection, made in 1895–1896, contained about 90 fishes, referred to 33 species, 25 of which were described as new. Moore’s second collection, made in 1899–1900, contained 180 samples of 48 species, 22 of which were new’.
Map clarification: red arrows to facilitate the identification of route. At locations with multiple arrows, red follows red. Date markers show arrivals at locations. Location clarification: Lofu=Lufubu, Vua=Livua, Kituta=Chituta, Kassanga=Kasanga, Chamkaluki=unknown (probably near Katili), Pembe=Wampembe, Kombe=Mkombe (adjacent to Kalila), Sumbua=Sumbwa, Kibwesi=Sibwesa, Mrumbi=Moba, Tembwi=Cape Tembwe, Toa=Mtoa, Kasawa=Kasaba, Niamkolo=Mpulungu, Niamkolo Island=Mbita/Kumbula Island.

32.3. The name of the location where Moore set up camp
Many of the subsequent publications regarding the location of Moore’s camp are based on speculations, but he provided some important and revealing details in his non-scientific narrative ‘To the Mountains of the Moon’ (Moore, 1901b). When describing fish eating birds he observed during his second expedition, he wrote: “He was a bachelor bird and lived in a crack in the rocks of an island near Kinyamkolo, where I was camped” (Moore, 1901b: 90). Regarding the location where he camped during his first expedition in relation to other nearby locations, he further wrote: “It was from Kinyamkolo on my former journey [...] that I tried to make a voyage to Cameron Bay in an old historic craft”, a rusty steel boat which Moore repaired, “then sailed her round from Kituta to Kinyamkolo”, but after “some days at Kinyamkolo she developed more pinholes” (Moore, 1901b: 93–94). The fact that Moore’s Kinyamkolo really is a location comprising a missionary station, not an area or district, is evident in the next line: “... I suggested on the verandah of the new mission at Kinyamkolo, that ...” (Moore, 1901b: 109). Furthermore, Moore treated Kinyamkolo as synonymous with Niamkolo, although he wrote ‘Nyamkolo’: “... who was staying for the sake of his health at the mission at Nyamkolo, a spot, by the way, which has as high a death rate as any place in the world. [...] In order to get to Nyamkolo, one crosses the creek to a point in the westward swamps [in the southernmost Chituta Bay] ...” (Moore, 1901b: 84).
In conclusion, based on the preceding citations, Moore camped at the missionary station, which he interchangeably referred to as Kinyamkolo and Nyamkolo (see more synonyms of Kinyamkolo/Nyamkolo/Niamkolo below, § 32.8). Also, Moore treated Kinyamkolo/Nyamkolo as a location and position, not an area or district as suggested by Drachenfels (1995) and Konings (2012; 2013a; 2013b; 2013c; 2015a: 198). Moore’s camp may have looked something like the Niamkolo camp illustrated in Thomas (1896: 113). Compare those illustrations with Moore’s (1901b: 81–82) anecdote involving a rainy and fearful night, a tent pitched on a hot sandy ground crowded with mosquitos, an appalling heat, and an incident of wondering about like a ghost in pyjamas looking for a wind while being observed in the night by two fluorescent eyes.

32.4. Niamkolo and the London Missionary Society
Niamkolo was the site where the London Missionary Society (LMS) had set up a station in 1883–1884. However, due to the outbreak of sleeping sickness in the area, as well as other fatal fever diseases, it was temporarily closed for a few years due to deaths among the missionaries and their families, reopening in 1889–1890 (Johnston, 1892: 248–251; Jones, 1897: 74; Horne, 1908: 268; Meebelo, 1971: 6, 25; Ragsdale, 1986: 19, 24; Chuba, 1995: 21–22; Orr, 1998: 98; Rotberg, 2015: 151; SOAS, 2018g). The Niamkolo site was first described by Edward Coode Hore, the LMS’s first-in-command at the lake in 1878–1888, who voyaged round the south end of Lake Tanganyika in 1880 (SOAS, 2018j). Hore (1882: 21), who referred to the site as Niumkorlo, described it to have “a fine little harbour, at the north side of which is a prominent round hill on a peninsula”. On Hore’s map (1892: 26), the location of that harbour is indicated with an anchor symbol, while on recent maps, the harbour is referred to as Musende Bay (Government of Zambia, 1970). Furthermore, the hill on the peninsula, which has a rocky shoreline, is adjacent to the present-day Mpulungu Port. That rocky shoreline is presumably the location where Moore collected all his fish labelled ‘Kinyamkolo’. However, north of the Niamkolo stone church, currently referred to as Chikola Hill and Katanka Point (Government of Zambia, 1970), is an alternative location for Moore’s collections (see below and § 32.15–32.23, 32.26). In addition to Hore, the Niamkolo station was founded by his missionary colleagues John Harris and Arthur Brooks, who, like Hore, referred to it as Niumkorlo (Hore, 1892: 252; SOAS, 2018d; 2018e). Presumably, ‘Niumkorlo’ was primarily Hore’s spelling of ‘Niamkolo’, although, maps from that time all indicate ‘Niumkorlo’, such as Kettler et al. (1886–1895), Stevenson (1888), and Stanley (1899c). However, ‘Niumkorlo’ in favour of ‘Niamkolo’ rapidly became uncommon following a series of events including (1) Harris’ death from dysentery at Niamkolo in 1885, (2) the temporary abandonment of the station and site in 1887–1888, (3) Hore’s return to England in 1888 due to illness, and (4) the murder of Brooks by coastal slave traders in 1889 (Hore, 1892: 258, 262–263, 292; Johnston, 1892: 248–250; Horne, 1908: 268; Chuba, 1995: 19–20; SOAS, 2018j). Alexander Carson, an engineer and lay missionary, who was sent to Lake Tanganyika to assemble the boiler in LMS’s steamship, the Good News (see more in Fig. 124), appears to have already referred to the site as Niamkolo in 1886 (Hore, 1892: 273; Johnston, 1892: 248; McCracken, 2012: 84; SOAS, 2018i). In addition to being LMS’s engineer, as well as one of the founders of the Fwambo station around 1887 (§ 32.12), Carson was an avid plant collector in the region between Lake Tanganyika and Malawi (cf. Johnston, 1897: 233–283), collecting pretty much everything he came across, including scorpions, spiders, and fish (Pocock, 1898: 430, 432; Boulenger, 1903: 362) (see more below, § 32.12).
Additional references to ‘Niamkolo’ during that time were made by British colonial officials (Johnston, 1890; Sharpe, 1893; Codrington, 1902) and associates of the Linnean Society and Royal Geographical Society (e.g. Brown, 1894: 138; 1895: 247).
The re-establishment of the Niamkolo station in 1889 involved Carson, Robert Stewart Wright, and Alfred James Swann (Hore, 1892: 293; Johnston, 1892: 250). Wright, primarily a missionary, was stationed at Fwambo and Kavala Island in 1887–1888, and after his stay at Niamkolo suffered severely from smallpox, returning to England in 1890 (Hore, 1892: 282, 294; SOAS, 2018l), whereas Swann, a mariner and lay missionary, who travelled and laboured in Central Africa in 1882–1909, became one of the most influential westerners on Lake Tanganyika. In 1882, Swann was a member of Hore’s second Lake Tanganyika expedition (Ragsdale, 1986: 24), but following the retirement of Hore from the lake in 1888, Swann established himself as the ‘LMS’s Admiral on Lake Tanganyika’ in “charge of all their fleet of boats and steamers” (Johnston, 1890: 738). In addition, he made friends with some of the most powerful slave traders at Ujiji, obviously not to partake in or defend their trade, but thoroughly and effectively oppose and abolish it (cf. Swann, 1910a). Following a leave of about half a year at home in England, Swann returned to the lake in 1888, accompanied by, among others, Charles Mather, an ordained missionary with a bachelor’s degree in medicine, who would spend the rest of his life at the LMS’s stations around Lake Tanganyika, including Niamkolo (Hore, 1892: 274, 287). Two years later, in 1892, Adam Purves, a Scottish lay missionary and craftsman joined the LMS at Niamkolo. Like many other missionaries, he died about ten years after his arrival in Africa, more precisely in 1901, aged 37 years. However, before his death, he made permanent imprints, besides founding a mission station in 1900 at Mbereshi 235 km southwest of Lake Tanganyika, he built LMS’s stone church in Niamkolo, which today is one of the oldest historical national monuments in Zambia (Chuba, 1995: 22, 70; Carey, 2003: 4; Hubbard, 2014: 8; Rotberg, 2015: 186; SOAS, 2018k). Erected in 1895–1896, with a tower of 15 metres, the church was said as to “act as a beacon to voyagers on the lake” (Thomas, 1896: 111). For some interesting illustrations of the church and the construction of it, see the former reference. Also, in Carlin (2018e; 2018g), there are two old photos presumably taken in 1890–1900, one illustrates ‘Niumkolo Church’, although, it appears to be a different church, while the other photo (Fig. 120) shows the lake view ‘from Niumkolo’, taken near the location for the stone church, and includes the western hill, ‘Mpulungu Hill’ (§ 32.8), with the rocky shoreline, presumably the exact location for Moore’s collections labelled Kinyamkolo, which is adjacent to the present-day Mpulungu Port (in the centre of the photo) and opposite to Makosi Hill, the southernmost part of Mbita Island (to the right). The colour sketch (Fig. 118) by Moore (1903: Fig. opposite page 68) appears to illustrate a very similar view as the former. An additional sketch (Fig. 122) with a similar view from Niamkolo made in the early 1890s, a few years prior to Moore’s visit, is available in Brown (1893: 237). Also, the photo (Fig. 121) on which the former sketch is based is found in Johnston (1897: 32). There is also an old photo (Fig. 124) taken at Mbita Island in the opposite direction, southeastward, illustrating the shores of the Niamkolo area (the background) and a small stretch of the southeastern part of Mbita Island (to the right) behind which Mpulungu Port is currently located (Swann, 1910a: 83, Fig. bottom; Carlin, 2018c). Owing to the high incidence of sleeping sickness, the Niamkolo station, including the stone church, was permanently abandoned in 1908 (Hubbard, 2014). Go to Google Maps to see the location and some recent photos of the remains of the church (coordinates: -8.757005, 31.116989).
Regardless of the allegedly unhealthy environment, the LMS station at Niamkolo must have been a rather good site for Moore to camp at as it presumably provided useful and important facilities, and it was frequently visited and supported by British officials and fellow countrymen, including Henry Hamilton (Harry) Johnston, the British colonial chief administrator, who was a friend of Moore (Johnston, 1890: 738; Moore, 1901b: 26). However, it may be worth pointing out that when Moore visited the mission station in spring 1896, it was “entirely deserted; no one lived in the neat thatched bungalows or taught in the school” and the “roof of the [stone] church was partly fallen in” (Moore, 1901b: 75, 89).
While Harris oversaw the Niamkolo station during its first two years, Swann, the superintendent of LMS’s marine department, and Mather, the doctor, took charge of the station following its re-establishment in 1889 (Hore, 1892: 258; cf. Gamwell & Gamwell, 1961: 518), by which time, a new name for the station appears to have been introduced: Kinyamkolo.

Fig. 120. Presumably taken around 1900, this old photo illustrates the rocky hill (in the centre) where Moore seemingly made his collection of fishes in 1896 and 1899, labelled Kinyamkolo (Carlin, 2018g). The rocky hill, originally referred to as Mpulungu Hill (Venning, 1960), is adjacent to the present-day Mpulungu Port and opposite to Makosi Hill, the southernmost part of Mbita/Kumbula Island (to the right). Photo reproduced from the ‘Mbala / Abercorn’ Facebook of Colin Carlin.
Fig. 121. The unhealthy location of Niamkolo and the sacrificing work of the missionaries.
Reproduced from Johnston (1897: 32), this photo apparently illustrates Moore’s Kinyamkolo, also known as Niamkolo, the present-day Mpulungu, a location which was described as “not a very alluring site; the ground is as cruelly rocky and cutting to the feet as the soil of Likoma, and the surrounding trees have a stunted, skimpy look, no doubt from the poor ground on which they grow. But the mission selected this spot in the belief that it offered the best harbour on the south end of Tanganyika” (Johnston, 1890: 738). The Niamkolo missionary station was founded by the London Missionary Society (LMS) in 1884–1885. Due to the outbreak of sleeping sickness, the station was abandoned and not reopened until 1889–1890 (§ 32.4, 32.5). According to Moore (1901b: 84), Kinyamkolo/Niamkolo “has as high a death rate as any place in the world”. When Moore walked from Chituta to Kinyamkolo (Niamkolo) during his second Tanganyika expedition, he noticed among the trees and the sun-dried rocks near the missionary station, “a number of European graves huddled together in a space of stony ground” (Moore, 1901b: 89). In 1893, when the Central Africa Mission of the LMS had been running for 16 years, 36 missionaries had been sent out by the society, of which 11 had died and 14 gave up missionary work, often after brief periods of service (Chronicle, 1897: 163; SOAS, 2018g). Most of the missionaries were continually transferred between the stations and besides preaching the word of God, they had to take the roles of doctors, farmers, mariners, engineers, builders of ships, houses, and churches, school teachers for the natives, and engage in the fight against the slave trade. Harry Johnston, the British colonial chief administrator, thought highly of “these missionaries, who are really and truly, quietly and unostentatiously, laying the foundations of civilisation in the midst of what at first appears to be hopeless savagery” (Johnston, 1890: 737).
Fig. 122. A sketch made in the early 1890s, a few years prior to Moore’s visit, illustrating a view of the rocky hill (centre) where Moore presumably made his collection of fishes, notably Tropheus moorii and Neolamprologus modestus, and the southern point of Mbita/Kumbula Island (right) (Brown, 1893: 237).

32.5. Kinyamkolo
The name Kinyamkolo was probably coined by Swann and/or Mather in 1889–1890. Following the re-establishment, there were few co-workers at the station. At the end of 1890, the staff comprised Swann, his wife, and Mather (Chuba, 1995: 61), and two years later Swann with wife, Mather with wife, James Hemans from Jamaica with wife (Carlin, 2018f), and Carson (Cousins, 1892b). Swann and Mather appear to be the only associates of the station that explicitly referred to it as Kinyamkolo, all past co-workers, such as Hore, Harris, and Brooks, adopted the name Niumkorlo, and other co-workers, including those who would join during the 1890s, such as Carson, Hemans, and Purves, did or would primarily refer to the station as Niamkolo, see more in § 32.4, 32.6, and 32.11. The reason for the introduction of the new name is not clear, but it may have represented a new start, a new administration, without the retired Hore, who was “[g]enerally disliked by his colleagues” and “seldom mentioned by them other than in criticisms of his policies and attitudes” (Wolf, 1971: 9). One may also speculate whether the new name, Kinyamkolo, referred to the very same site or another nearby site. Indeed, the construction of the Niamkolo stone church after the re-establishment was located about 1.5 km northeast of the natural harbour, Musende Bay, which Hore initially referred to. However, since that harbour, a shelter for boats in stormy weather, presumably was an important feature for the station when the site was chosen the first time, it was probably just as important the second time, hence, the two names, Kinyamkolo and Niamkolo, most likely refer to the very same site. Regarding the linguistic significance of Kinyamkolo, see below (§ 32.7). While most missionaries and non-missionaries visiting, communicating with, or reporting about the station during the 1890s adopted the name Niamkolo, a few did not, one of them was the Reverend James Johnston, who, in his narrative, ‘Missionary Landscapes’, referred to “Kinyamkolo as a suitable harbour and new marine headquarters at the south-end of the lake” and stated that the native languages were being mastered “at the marine station of Kinyamkolo where a village of 400 people had been formed” (Johnston, 1892: 250–251).
Next note about Kinyamkolo is from the diaries of William Grant Stairs, a British explorer, soldier, and adventurer, who served on the staff of the Emin Pasha Relief Expedition of 1887 under the leadership of Stanley. At Karema, en route for Katanga, currently southern DR Congo, Stairs wrote: ‘October 13th. At 9 o’clock in the morning, I received a letter from Mr. A.-J. Swann in response to the one I wrote him from Tabora. It is dated Kinyamkolo, mission station, southern end of Lake Tanganyika, October 4, 1891’ (Stairs, 1893: 126). Obviously, Swann, in charge of the station in the early 1890s, referred to it as Kinyamkolo.
Another missionary was William Harwood Nutt, who was attached to the LMS’s Lake Tanganyika Mission in 1892–1896 and stationed at Fwambo and Kambole, the latter station located on the Ulungu Plateau, south of Lake Tanganyika, where he built a church on wattle and daub (interwoven sticks covered with clay) (Chronicle, 1895: 79; Thomas, 1896: 112; SOAS, 2018h). For the location of Kambole, see Stewart (1903: map opposite page 222). Although Nutt, like many of his missionary colleagues, frequently visited the Niamkolo station, it appears that he never wrote any reports from that station, but only from Fwambo/Kawimbe (§ 32.12) and Kambole. However, in addition to building churches and teaching freed slaves to read and write, Nutt made scientific collections of various organisms. As collection location and type locality for two spider species and one subspecies, Nutt indicated Kinyamkolo (Pocock, 1897: 731–732; 1898: 433, 434, 441). It is very likely that Nutt, when indicating Kinyamkolo as the origin of his spiders, was referring to the Niamkolo station and village. At about the same time, a report on the political situation around Lake Tanganyika in the 1890s edited by Henley (1897: 239) included two European stations, referred to as Kinyamkolo and Kituta [Chituta]. In addition, it has already been described how Moore referred to the station normally as Kinyamkolo, occasionally as Nyamkolo (§ 32.3).
What most of these visitors and authors appear to have in common is that they met or had contact with Swann or Mather. Even though Swann left his work at the mission in 1894, he was still around working in that area, the British Central Africa, until 1909 (Swann, 1910a: ix; SOAS, 2017). Swann very likely met Moore, or at least knew about Moore (Moore, 1897a: 293; Swann, 1910a: 215;), and Moore definitely knew about Swann, because it was Swann who had sent samples of the marine-like jellyfish from Lake Tanganyika to England, which sparked off the theories of a former connection between Lake Tanganyika and the ocean, something which Moore regarded as utterly intriguing and exhilarating (Günther, 1893: 269; Moore, 1897b: 43; 1903: 2, 298; Swann, 1910a: 79). Most likely, it was Swann who introduced the name Kinyamkolo to Moore, hence, Moore’s view of Kinyamkolo is the same as Swann’s, i.e., it pertains to the Niamkolo station and site.
Around 1890, not everyone appears to have adopted the name Kinyamkolo. Although Swann met Johnston, by that time appointed British consul at Mozambique, his choice of name, Kinyamkolo, did not have any effect on Johnston, who referred to the station as Niamkolo (Johnston, 1890). Presumably, however, at the time they actually met, in the middle of 1889 (Johnston, 1910: vii), Swann and/or Mather had not yet invented the name Kinyamkolo.

32.6. The adventurer Edward James Glave
Researchers of the location of Kinyamkolo and whether the name pertains to a district or village would ideally need a map indicating its exact position, but few such maps exist. Swann never mentioned Kinyamkolo or Niamkolo in his narrative, ‘Fighting the Slave-Hunters in Central Africa’ (Swann, 1910a), nor did he include a map, and Moore, who published several articles following his expeditions, in which he very rarely mentioned Kinyamkolo, did not indicate Kinyamkolo on the Lake Tanganyika map that he published in his scientific volume, ‘The Tanganyika Problem’ (Moore, 1903: map opposite page 8). Nonetheless, at least two maps with Kinyamkolo exist, one from Boulenger (1901a: map opposite page x) and another from Glave (1897a: 900) (Fig. 123). While Boulenger’s map is flawed with incorrectly indicated villages, therefore rather worthless, Glave’s map does not have any such errors. Among other sites, the latter shows Moliro, Sumbu, Kinyamkolo, Kituta (Chituta), Fort Abercorn (Mbala), and Fwambo. As expected, Kinyamkolo is indicated where Mpulungu is indicated on modern maps.
Besides being an associate of Stanley, Edward James Glave was a British adventurer, author, and journalist, who exposed some of the misery and atrocity in the Congo (e.g. Russell, 1895; Glave, 1897b). Glave, who was said to have ‘travelled from Zanzibar to the Congo a hundred times’, set out on his last journey in 1893 (Hurlbut, 1895; 187; Russell, 1895: 867). “During two years of toilsome exploration he had traversed the whole breadth of the Dark Continent; with only a dozen black followers he had passed from tribe to tribe without firing a shot in defense, or even threatening a native. [...] His task was behind him; the fruits of his philanthropic mission were stored in well-filled journals and camera films; his foot almost rested on the threshold of home; the curtain was ready to rise on a stage he had trod before as an always unassuming hero of a drama of daring and fortitude; the curtain rose; but, alas! the scene was set for the familiar African tragedy. On the afternoon of Sunday, May 12, [1895] though devotedly nursed by new found friends, he succumbed to a sudden attack of fever; and on the following morning his body was laid in the soil of Africa, whose enslaved humanity it had been his highest ambition to succor, even at the risk of his life” (The Century Magazine, 1895b: 952). At the very end of what would be his last journey, Glave died aged 32 years at a missionary station in Matadi (Glave, 1897b: 712) on the Congo River, very near the Atlantic coast (Hurlbut, 1895; Glave, 1897c). Much of his diary, maps, and photos from that journey were later published in The Century Magazine in New York {link}. On 8 July 1894, Glave was the first to locate and visit “the place where Dr Livingstone’s heart is buried beneath a big tree” (Glave, 1896: 777–778), 21 years after Livingstone’s death in April 1873. For photos, see The Century Magazine (1895a).
Glave visited Kinyamkolo: “Lake Tanganyika, September 19, 1894. Reached Kinyamkolo, the London Missionary Society’s station, a splendidly situated place. [...] About one thousand people are settled within the stockade. During the last few years the general health has much improved; a few years ago there was so much sickness and so many deaths that the home society thought of abandoning the station” (Glave, 1897a: 900). “September 21. I left Kituta early this morning, and after four hours’ marching reached Fort Abercorn, a sturdy stockaded little place with houses built of white clay from the ant-hills. [...] Leaving Abercorn, I marched for three hours to Fwambo. [...] About fourteen thousand people are gathered within the palisades of the Fwambo mission. [...] Good peat is said to have been found at Kinyamkolo, also an excellent fiber for rope” (Glave, 1897a: 901). “Kinyamkolo, October 9. A miserable fever tucked me into the blankets to-day. October 10–12. The wretched fever returns every day. I cannot eat anything. I have taken any amount of quinine, which has the effect of giving me grotesque visions as soon as I close my eyes. Mr. and Mrs. Purves and Mr. Thomas have done everything for me. October 14. I am truly grateful for the kindness of the missionaries. At about ten at night we left Kinyamkolo in the Morning Star [Fig. 124], en route for Sumbu, and hoisted the sail with a favorable breeze” (Glave, 1897a: 902). “Before the expedition of Dhanis and the operations of the Antislavery Society [including the Congo Arab war, 1892–1894], slaves in chains and forks were constantly tramping through Manyema to Mtoa, and then by canoe across to Ujiji” (Glave, 1897a: 906). “I find that fevers, dysentery, etc., have lately seriously affected my nerves. After firing one shot from my rifle I tremble abominably, and cannot hold the gun steady even with a rest. I hope this is not going to last any length of time...” (Glave, 1897a: 905).
Glave, who throughout his diary referred to the missionary station as Kinyamkolo, obviously met some of the missionaries there, including Purves and William Thomas. He may also have met Swann, at least he mentioned him in his diary as if he knew him (Glave, 1897a: 901). Nonetheless, Glave’s reason for adopting the name Kinyamkolo for the station as opposed to Niamkolo was undeniably because that was the name applied by one or several of the missionaries at the station.
Regarding Glave’s map (1897a: 900), this was drawn based upon a sketch made by the explorer himself (cf. Russell, 1895: 866, text on map; Glave, 1897b: 702, text on map), and appears highly accurate with many small villages included as opposed to Boulenger’s map (1901a: map opposite page x), which was obviously not drawn by the author himself, but rather inserted by the publisher as an overview map relating to the many fishes described in that publication.
Interestingly, both Moore and Boulenger knew about Glave’s publication (1897a). Glave wrote about some peculiar fish: “October 16. Early this morning we had a strange experience. Several large fish, over four feet long and very thick like salmon, came about our slow-moving boat and grappled vigorously at the paddles, showing no fear. I shot two with my Martini, but unfortunately they sank before we could pick them up. Whether we were considered an enemy in their waters, or whether we were looked upon as food, I don’t know, but this lasted for an hour at intervals. We could have shot several, but desisted from useless slaughter, as we could not regain the creatures before they sank—very much to the disgust of the boatmen. If we had had a boat-hook or gaff we might have obtained a few. At ten this morning we reached Sumbu” (Glave, 1897a: 902). One year later, Moore commented “What the fish is that so much surprised Glaive [sic], when he crossed the lake, by attacking the paddles of his boat, is quite unknown, but I myself saw these same fishes attack the paddles of my own boat, not 20 miles from the spot where Glaive described them, on the west coast of the lake” (Moore, 1898a: 28; cf. Nature, 1897: 258; Moore, 1901b: 112). Following Moore’s second Tanganyika expedition, Boulenger stated: “Mr. Moore has satisfied himself that the ferocious fish, over four feet long, attacking the paddles of boats, to which the late E. J. Glave first drew attention (‘Century Magazine,’ liii. 1897, p. 902), is no other than the adult of the Lates [L. microlepis] discovered by him, and which was described in the first Report from young specimens” (Boulenger, 1901c: 149–150). Obviously, Moore, as well as Boulenger, knew very well about Glave’s diary and map. If Moore had not agreed upon the description and indication of Kinyamkolo, he would most likely at some point in his many reports have pointed this out, but he did not. Undoubtedly, Moore’s view of Kinyamkolo is the same as that of Glave and for Boulenger’s part, he certainly realised that Kinyamkolo is the same as Niamkolo (compare with § 32.2).
On a side note regarding prying fishes, the same year as Glave’s publication, H. H. Johnston wrote that “remarkable discoveries may yet await us, especially on Tanganyika, where numerous travellers have reported the existence of an exceedingly large fish which occasionally rushes at boats in a threatening manner” (Johnston, 1897: 360). As for an explanation, he continued: “Similar rumours of a very large fish in Nyasa are prevalent. Both Commander Cullen and Lieutenant-Commander Rhoades (of the Lake Nyasa gunboats) have reported curious circumstances tending to show that some very large fish or marine animal lives in Lake Nyasa, which amongst other things can bite off and carry away as a bait the brass log which is towed behind the vessels. It may not be more than a huge species of Bagrus, a Siluroid fish. Specimens of this creature have been already obtained which reached nearly six feet in length” (Johnston, 1897: 360). Possibly, the fish that Glave and Moore experienced were not Lates but Bagrus, or any other catfish.
On 31 December 1904, at Mpala, DR Congo, Cunnington caught a 500-mm-long catfish in a floating basket trap. The species, currently placed within the Clariidae, was classified as a silurid and described as Dinotopterus cunningtoni (Boulenger, 1906: 550–551). While Cunnington reported the observation of two additional specimens of this species with a length of 1,020 and 1,290 mm, Stanley referred to a native report when declaring the same species to reach a maximum size of 6 feet (about 1.8 metre) (Stanley, 1902: 530, 532, Fig. 1; Boulenger, 1906: 551).
As a personal anecdote, the following may be added: In October, 2008, we were diving at the small cove in southern Tanzania known as Kambwimba (about 10 km south of Kasanga). While lying flat on the sandy and rather muddy bottom, at a depth of about 35 metres, photographing Gnathochromis permaxillaris, one of us (Magnus) noticed that something was disturbing him from behind. When he turned his head to see what it was, he saw a catfish, nearly 2 metres in size, chewing on one of his diving fins. A bit startled, he instinctively began paddling with his fin as to make it stop or scare it away, but the fish would not leave. Eventually, he had to dive to another shallower spot where the catfish would not follow. Six days later, at Molwe (Fig. 125), about 8 km north of Kasanga, when photographing and filming Petrochromis sp. “Red Mpimbwe” in a depth of at least 35 metres, we observed two intrusive catfish specimens of the same size and probably same species. Once again, they could not be scared off easily. Possibly, these catfishes may have been D. cunningtoni. Furthermore, the heaviest fish Moore caught during his first Tanganyika expedition, 1895–1896, was described to be a sort of mud fish, weighing over forty kg (Nature, 1897: 258), possibly D. cunningtoni.
Recapitulating the facts so far about Kinyamkolo, (1) Moore referred to the missionary station where he was camped as Kinyamkolo, occasionally as Nyamkolo, (2) the name Kinyamkolo was not only adopted by Moore and the missionaries in Niamkolo, as suggested by Konings (2012: 10; 2013a: 20), but by many others, including soldiers and external missionaries, (3) most of the LMS’s missionaries appear to have applied the name Niamkolo in various correspondence, and (4) the name Kinyamkolo was never applied to a district (cf. Konings, 2012: 10; 2013a: 21).

Fig. 123. Map by E. J. Glave showing his journey in the Lake Tanganyika region. On 19 September 1894, he reached the London Missionary Society’s station at Kinyamkolo (today’s Mpulungu, see map). Also, Kituta (Chituta), Fort Abercorn (Mbala), and Fwambo are plotted on the map. It appears that the name Kinyamkolo was phased out about 15 years after its introduction, widely replaced by Niamkolo around 1905 and Mpulungu in the 1930s (§ 32.4–32.5, 32.8). This is one of few maps showing the location of Kinyamkolo.
Fig. 124. The steamship Good News and sailboat Morning Star of the LMS.
Reproduced from Swann (1910a: 83) [cf. Hainsworth, 1951: 27; Carlin, 2018c], this photo was taken at Mbita Island, probably in 1889–1890, and shows the eastern part of Makosi Hill, i.e., the southernmost part of the island (to the right), and the present-day Mpulungu area (in the background).
The ship to the left is the steamer Good News of the London Missionary Society (LMS). Between 1883 and 1885, the ship was transported in sections to Lake Tanganyika via the Shire River and Lake Malawi by the African Lakes Company (later the African Lakes Corporation) (Hore, 1892: 189, 194, 290; Hainsworth, 1951: 18; Bennet, 1969: xiii; Chuba, 1995: 19–20). Edward Coode Hore, the chief of the Lake Tanganyika Mission between 1878 and 1888, initially fixed upon Niamkolo, currently Mpulungu, as a site for the assemblage of the ship. However, owing to the food shortage and deserted country resulting from the slave trade, a more populous site 6.4 km upstream in the Lufubu valley was eventually chosen, where a monument of the event was erected in 1945 (Hore, 1892: 232–234, 251–252; Hainsworth, 1951: 16–17, 20; SOAS, 2018j). In charge of the assemblage was James Roxburgh, an engineer and lay missionary who died from dysentery and malaria shortly after finishing the job (Hore, 1892: 243, 260, 262; Johnston, 1892: 248; Chronicle, 1897: 163; Swann, 1910a: 105). In the following year, 1886, the LMS sent out Alexander Carson, who, assisted by Alfred Swann, the LMS’s second-in-command, assembled the steamship’s boiler at the Kavala Island station (Hore, 1892: 273; SOAS, 2018i). Ten years later, Carson died at the Fwambo station (Jones, 1896: 181; Swann, 1910a: 83). In 1889, on his journey signing treaties with local tribes in what is currently Zambia, Harry Johnston visited the Niamkolo station and commented that “the swell is often dangerously heavy on the beach. Indeed, just before my arrival the mission steamer (the Good News) had met with a serious accident. She was lifted up from her anchorage by a sudden swell coming in from the outside, and landed on the rocks in shallow water” (Johnston, 1890: 738). The photo above appears to show the repair after that accident, and the caption of the same photo in Swann (1910a: 83) reads: “The s.s. ‘Good News’ is in a dry dock, quarried out of rock, floated by pith-wood after being wrecked. Salvage operations took four months, as natives had to work under water”. In 1895, one year after Swann finished his contract with the LMS, the Good News was sold to the African Lakes Corporation (Moore, 1901b: 74; Chuba, 1995: 19–20), after which it came to play a less important role on the lake, at least for the missionaries. In 1898, Charles Lemaire and his scientific expedition to Katanga travelled with the Good News from Chituta to Moliro (Lemaire, 1902: 7). In 1899, on Moore’s second Tanganyika expedition, he found the steamer dumped in Chituta Bay, and prior to his journey north “to the mountains of the moon”, it was renovated at Kinyamkolo (Mpulungu) (Moore, 1901b: 108–110) (§ 32.17, 32.18). A few years later, in 1907–1908, the German officer Paul Graetz, who crossed Africa in a motor vehicle, travelled with the Good News from Kigoma south to Chituta, after which he had to “dismantle and hand carry the vehicle over parts of the very steep road up the escarpment to Abercorn [Mbala]” (Graetz, 2006; Carlin, 2018b) (§ 32.17). In 1915, during World War I, the German ships Hedwig von Wissmann and Kingani shelled the hulk of the Good News in Chituta Bay to ensure it would not be used against them in the war, but the Good News was already a wreck beyond repair by that time (Hainsworth, 1951: 18; Chuba, 1995: 19–20). On a side note, the British captured the Kingani at the end of 1915, renaming it HMS Fifi and in 1916, sunk the Wissman, whereupon the Germans remained with a single vessel, the Graf von Götzen, which was scuttled in Katabe Bay, Kigoma in mid-1916 and, following the salvage of the ship by the Belgians led by the Swedish engineer John Ludwig Wall in 1918, renamed SS Liemba [MV from 1952 onwards] by the British in 1927 (Coosemans, 1949b; EAR & H, 1956; Farwell, 1986; Berg, 2014).
For a 1950s or early 1960s photo of the Good News beached and abandoned in Chituta Bay, see Carlin (2018i). Parts of the remains of the Good News, “some flattened pieces of rusty metal”, were recently dug up from the Chituta Bay meadow [see the comments in Carlin (2018b)]. For additional photos, see Carlin (2019) and Atlas Obscura (2019).
The ship to the right in the above photo is LMS’s sailboat the Morning Star. It was assembled and launched by Hore and Swann at Ujiji in 1883, after its sections had been dragged overland from Zanzibar to the lake, about 1,325 km (Hore, 1892: 221–223; Swann, 1910a: 83–84; SOAS, 2018j). In 1894, the adventurer Edward James Glave travelled from Kinyamkolo (Mpulungu) in the Morning Star en route for Sumbu (Glave, 1897a: 902) (§ 32.6). Just like the sister vessel, the Morning Star was destroyed in World War I (Chuba, 1995: 19–20).
Fig. 125. The craggy ledges at Molwe are similar to Moore’s craggy ledges at Mpulungu, the type locality of several Tanganyika cichlids.

32.7. Arthur Loveridge at Niamkolo in 1930, and the significance of the Bantu/Swahili ki- prefix
During an eight-month zoological expedition between November 1929 and July 1930, Arthur Loveridge and three native assistants, whom Loveridge had previously trained in the preservation of mammals, birds, and reptiles, collected 7,411 vertebrate specimens, many invertebrate specimens, including molluscs and worms, and purchased about half a ton of ethnological specimens. The expedition passed through present-day Tanzania, Zambia, Malawi, Uganda, and Kenya. A map showing the route can be seen in Loveridge (1933a: Plate 1). On 7 May 1931, the expedition reached Niamkolo, in reference to which Loveridge wrote: “Nyamkolo has been variously spelt as Niomkolo, Kinyamkolo, and Kinyamkole. The prefix ‘Ki’ is an augmentative in the local dialect in direct contrast to its use as a diminutive in Swahili. I learned this from Mr. White, a veteran missionary of the London Missionary Society, who told me that Moore had stayed at Nyamkolo the year previous to his own arrival” (Loveridge, 1933a: 21).
Thus, according to the LMS’s veteran missionary, Mr White, who is probably the same man as Mr William White (cf. Chuba, 1995: 31), Kinyamkolo is synonymous with Nyamkolo, which in turn is synonymous with Niamkolo and a long series of additional similar names, see below (§ 32.8). Furthermore, White through Loveridge confirmed that Moore camped at the missionary station, which was indeed referred to as Kinyamkolo by Moore, see above (§, 32.3). Therefore, it appears rather safe to conclude that Moore’s Kinyamkolo is synonymous with the Niamkolo missionary station and the adjacent village, and that the name Kinyamkolo, the way Moore applied it, does not relate to a district, nor to Ulungu Escarpment, the rocky coast west of Mbete Bay, which has been suggested by Drachenfels (1995) and Konings (2012; 2013a; 2013b).
However, the significance of the name, Kinyamkolo, is interesting. As suggested by Loveridge (1933a: 21–22; 1933b: 383) and Konings (2012; 2013a; 2013b), the prefix ki- should be understood as an augmentative, relating to something bigger, better, greater, etc. This is possibly partially correct, but contrary to the statements of Loveridge, the prefix ki- likely has the same significance in both Swahili and the local Bantu language that was spoken at Niamkolo, presumably KiLungu, and it may be both augmentative and diminutive depending on the situation.
Frequently, in a certain class of nouns in the Swahili language, the prefix ki- is diminutive, it alters the meaning of a word to something minor than that which the word denotes without the prefix, such as mji (city) and kijiji (village), lugha (language) and kilugha (dialect), mlima (mountain) and kilima (hill), mtoto (child) and kitoto (infant), jiko (kitchen) or mwiko (ladle) and kijiko (spoon/teaspoon), etc. Sometimes, ki- denotes something augmentative, such as jua (sun) and kijua (scorching sun), and jasho (sweat) and kijasho (excessive sweat) (Mohamed, 2001: 42–43). From the above examples, the prefix ki- may be regarded as an emphatic marker, having an amplifying effect on a word, usually a class or adjectival noun. Compare this with the English word very, which is something of a preceding emphatic marker, almost a prefix, amplifying the main word: very small, very large, etc. In contrast to the English, the Swahili frequently merge several words of a sentence into one long word. Occasionally, the ki- prefix is neither augmentative or diminutive, as seen in the words pato and kipato, as they both have the identical meaning (income) (Mohamed, 2001: 42–43). In the latter case, the prefix ki- is seemingly alliteration (see more below, § 32.10). In summary, the prefix ki- may be augmentative, diminutive, or simply alliterative.
Therefore, the ‘Ki’ in Kinyamkolo does not necessarily mean great or greater, nor does Kinyamkolo necessarily mean Greater Nyamkolo as suggested by Konings, and partially also by Loveridge. On the contrary, Kinyamkolo may instead mean Little Nyamkolo. Indeed, Swann, the founder of Kinyamkolo station did not even mention it a single time in his narrative (Swann, 1910a), possibly due to its modest size and little importance, being much smaller and less important than, for example, the Fwambo station (see § 32.4, 32.5, 32.6, 32.11, 32.12, 32.13). Furthermore, Loveridge (1933a: 22) was surprised by the name of a village, Kitungulu, obviously with a ki- prefix, which was located in the hills east of Kasanga, “composed of less than a dozen huts”, and, as it appears, much smaller than the nearby Ntungulu village. If he had known that ki- is possibly a diminutive, pertaining to something smaller, he may not have been so surprised. Also, if kijiji (village) is a smaller version of mji (town), why would Kinyamkolo not be a smaller version of Nyamkolo, i.e., a small and less important village? Of course, both villages with their respective names do not have to exist simultaneously to apply this reasoning. As already stated, the Niamkolo mission was initially founded in 1883–1884, later closed for a few years, before reopening in 1889–1890 (§ 32.4, 32.5).
However, if ‘Ki’ in Kinyamkolo is indeed an augmentative, that does not instinctively make the name pertain to a district as Konings (2012; 2013a; 2013b) suggested, it only makes it stand out as something greater or more important. At least, this may have been the result that Swann and Mather wanted to achieve when they introduced the name following the re-establishment of the station (§ 32.4, 32.5), possibly representing a new start, a new station, with new people in charge. Instead of choosing New Niamkolo, they went with Greater Niamkolo. Of course, Kinyamkolo may very well be an alliterative version of Nyamkolo, without augmentative or diminutive significance, or ‘Ki’ in Kinyamkolo may originally have been an alliterative marker used only in a full sentence, at least by those who perfectly mastered the language (compare with Pambete/Mbete in § 32.10). Hypothetically, since Moore (1901b: 346) once indicated the term ‘Nyamkolo valley’, Nyamkolo/Niamkolo may have been the native name for the village and area, and Kinyamkolo the missionaries’ name for their station. However, additional indications by Moore and others contradict that hypothesis. Also, on one occasion, the chief of Kasanga referred to the missionaries of the Lake Tanganyika Mission as the Kizamkolo missionaries, but there is no explanation to the meaning of that name (Purves, 1895: 218). Furthermore, it should not be disregarded that Mather, with his less than two years in Africa by the time of the re-establishment, could not possibly be an expert in neither Swahili nor KiLungu, regarding the latter language which also applied to Swann, though Swann probably spoke Swahili fluently by 1890 (but see Johnston, 1892: 250). Conforming to this statement is the fact that the missionaries of the Lake Tanganyika Mission (§ 32.11) proudly reported to their employers in London whenever they managed to translate a psalm or biblical text into the local Bantu language (Chuba, 1995). Therefore, the possibility that they did not know the exact meaning of their new name should not be disregarded. All in all, the name Kinyamkolo did not catch on, but was phased out 10–15 years after its introduction. Particularly after the return home of Cunnington’s expedition in 1905, the name Kinyamkolo was increasingly deserted in scientific circles in favour of Niamkolo, including the many spelling versions of the latter, see below (§ 32.8).
In conclusion, what Swann and Mather had in mind when they coined the name Kinyamkolo is not certain. Presumably, the explanation may be found in the many letters they wrote home to the Foreign Office of LMS, which are not available online, only at the archives in London (see more below, § 32.11).
Also, although the ki- prefix can pertain to a language, it is very unlikely that Kinyamkolo relates to a Nyamkolo/Niamkolo language, because Niamkolo was not a tribe or country, rather a village and minor location. The people of Niamkolo belonged to the Lungu tribe, or the Walungu, the people of the Ulungu/Urungu, and they spoke KiLungu and KiSwahili (Swahili) (Hore, 1892: 232; Swann, 1910a: 89; Horne, 1908: 268; Orr, 1998: 90). The Walungu may also have spoken KiMambwe (ChiMambwe), the (Bantu) language of the Mambwe, the latter which was a people, head village, and country near where Niamkolo was located (Cousins, 1892b: 231; Meebelo, 1971: 6, 27–28; Chuba, 1995: 67). For maps indicating Ulungu/Urungu, and Mambwe/Mambue country and village, see Stewart and Coles (1880: map), Stewart (1903: map opposite page 222), and the map from 1892 in Luscombe (2018a: map).
To make the significance of the prefix ki- even more complicated, it may transform an adjective to an adverb, such as -fupi (short) and kifupi (shortly), and -zuri (beautiful) and kizuri (beautifully).

32.8. Additional synonyms of Niamkolo
Obviously, it is easy to get some of the local names wrong when not speaking Swahili or the local (Bantu) language and only visiting for a short while, as well as having more important priorities than obtaining the exact spelling of numerous foreign names of villages, such as the situation was for Moore, Cunnington, and many others, including more recent travellers. Furthermore, many natives and local citizens around the lake use different names and spellings of the very same place, sometimes even when such natives or citizens are from the very same village and frequently there is no clear distinction between M and N as well as R and L, as described by Spreinat (1995: 19). Similarly, in the 1990s, a fisheries officer in Mbamba Bay, Tanzania, Lake Malawi, which we (the authors) used to communicate with referred to a place called Wundu, while another representative of the local authorities called it Hundu. Obviously, both had alternative names for what is better known as Undu. In Zambia, Lake Tanganyika, the synonyms Kombe, Gombe, and Gombi, are not necessarily due to miscommunication, which is an occasional explanation, but the current most common and popular name may have been affected by a new generation and/or local newcomers; sometimes the official name of a village or district gradually alters, without opposition from the local citizens. For example, the name of the district in which Kabwe is located used to be Nkansi, but the spelling gradually altered to Nkasi in 2000–2005, which currently appears to be the official name. Also, sometimes two nearby expanding villages merge into one and take the name of one of them, such as Kabwe and Kanshenke, currently collectively referred to as Kabwe, so, it is not difficult to mix the names up.
Niamkolo, Niumkorlo, and Kinyamkolo were not the only synonyms for this site or village, 15 different versions were recorded in total. As already mentioned, it was always referred to as Nyamkolo by Loveridge (e.g. 1933a: 21) and on a few occasions by Moore (e.g. 1901b: 84). Additional renderings of the name are Niomkolo and Kinyamkole (Pocock, 1898: 430, 432; Loveridge, 1933a: 21; 1933b: 383), Nyamukolo (Meebelo, 1971: 6, 25, 27, 71, 94, 119), Niamkorlo (Chronicle, 1895: 79), Nyamkole (Brichard, 1978b; 1989: 22, Fig. map), Naimkolo (Government of Zambia, 1970; Takahashi & Hori, 2006: 189), Nimkole, Namkolo, Niumkolo (Cousins, 1892a: 47), Niamkoli (Rendle, 1907: 27), and Kinyambolo (Takahashi & Hori, 2006: 189), the latter two presumably being printing errors. Also, ‘Niumkolo Church’ and ‘From Niumkolo’ were written on two old photos illustrating ‘Niumkolo’ (Carlin, 2018e; 2018g). Regarding a possible highest-ranking synonym and supreme name/spelling, see below (§ 32.11).
While Loveridge applied the name Nyamkolo in the early 1930s, Cuthbert Christy adopted the name Mpulungu in 1926–1927 when making collections of Lake Tanganyika fish for the Natural History Museum, London. Furthermore, in the late 1920s, Cobham (2007) visited the area in a flying boat, also adopting the name Mpulungu. Originally, Mpulungu was the name of the hill west of Mpulungu Port (Venning, 1960: 290), the same hill at which Moore presumably collected his Kinyamkolo fishes. In 1926, Mpulungu Port with a customs warehouse was constructed, and during the following years, a small township emerged including a post office, police camp, stores, dispensary, and more warehouses (Venning, 1960: 290–291). Seemingly, both names, Nyamkolo (Niamkolo) and Mpulungu, were in use about the same time, possibly meaning that there were two small adjacent villages which eventually merged. On a map of Poll (1946: foldout map), two villages, Niamkolo and Mpulungu, are indicated adjacently, but on another map from the same author ten years later, Niamkolo is missing (Poll, 1956: foldout map). Alternatively, there were never two separate simultaneous villages, but rather, subsequent to the closing of the missionary station in 1908 and the presumable depopulation of the site, it was slowly repopulated under the new name Mpulungu. This repopulation took place following the constructions of the port and a road between the port and Mbala/Abercorn (Venning, 1960). As the old name faded, the new name grew, with both names equally applied for a while around 1930.
Etymologically, Mpulungu presumably derives from Ulungu, the name of the escarpment, plateau, and mountain range west of Mbete Bay (cf. Willis, 2017). Ulungu, in turn, is a widely known Bantu term for divinity, deity, the state of God, or ‘an ancestral clan’ (Frankl, 1990). In Bantu, Mulungu means God, of which Muungu or Mungu, the Swahili word for God, is a shorter version (Frankl, 1990). Historically, the mountain range stretching from Mbete to Lufubu River “is sacred ground in the lore of the ancients Urungu” (Stanley, 1978b: 36), “regarded by the natives as the abodes of spirits” (Moore, 1901b: 98), and according to Mather (1894: 235), along the Ulungu Escarpment, at the village Kapembwa, the chief god of Alungu was worshipped, and people came from far and near with their offerings. Also, the Marungu mountains along the west coast derive their name from God, possibly meaning ‘gods’ (in plural), or something holy or spiritual. Perhaps, Mpulungu means ‘give him/her divinity (or divine peace)’ [Mpeulungu: mpe (give him/her)], or ‘believer in God’, or perhaps just ‘a person from Pulungu’ [M (a person); the founder of the village], see, for example, Hore (1882: 16), who appears to have referred to Izinga Island as Pulungu Island. Also, since Mpulungu originally was the name of a hill, the linguistic meaning of Mpulungu may include a hill, a small mountain as opposed to the nearby Ulungu Escarpment. Possibly, sacred burial ceremonies were performed on the hill.

32.9. Alternative names and spelling
As stated previously (§ 32.1), many authors of Lake Tanganyika cichlids have treated Kinyamkolo synonymous with Mpulungu and/or Niamkolo, including Poll (1946: 351; 1978: 743), Staeck (1977: 243; 1985: 115), Axelrod (1978: 41), Brichard (1989: 149), Konings (1988: 26; 1998a: 54), Herrmann (1999: 28), Hanssens and Snoeks (2001: 639), Schupke (2003: 95), Schneidewind (2005: 300), and Kullander (2009). Besides Kinyamkolo/Niamkolo/Mpulungu, many additional places at the lake have altered their name or the spelling of it, usually gradually, such as the nearby Mtondwe Island (e.g. Hore, 1892: 226, map; Boulenger, 1906: 538, map; Poll, 1946: foldout map; Takahashi & Hori, 2006: 192), alternatively spelt or referred to as Motondo Island (Livingstone, 1874: foldout map), Ntondwe Island (Stanley, 1878b: 35), Mutondwe Island (Government of Zambia, 1970; Herrmann, 1987: 13, map; Konings, 1988: 11, map; 2015a: 15, map; Schupke, 2003: 35, map), Crocodile Island (Poll, 1956: foldout map; Konings, 1998a: 13, map), and Mutombwe Island (Brichard, 1989: 21, map), the latter is similar to Mtombwa, the name for a tabular mountain top of Ulungu Escarpment (Stanley, 1978b: 36). Similarly, the name of the adjacent Mbita Island (Konings, 2015a: 15, map) has been variously spelt or referred to as Murikwa Island (Stanley, 1978b: 35), Nkombola Island (Livingstone, 1874: foldout map; Hore, 1892: 226, map), Nkumbula Island (Takahashi & Hori, 2006: 175, Fig. 1), Kumbula Island (Government of Zambia, 1970; Herrmann, 1987: 13, map; Konings, 1988: 11, map; Brichard, 1989: 22, Fig. map; Carlin, 2018c), Mission Island (Poll, 1956: foldout map), and Mbete Island (Konings, 1998a: 13, map; Schupke, 2003: 35, map). The preferred name of the latter island in aquaristic circles, Mbita Island (Konings, 2015a), does not appear to be the most accurate, perhaps Nkumbula/Kumbula would have been more correct, since this is the official map name, widely adopted by the citizens of Mpulungu.
The nearby Lufubu River has been referred to as Lobu, Lofubu, Lovu, or Lofu River (Livingstone, 1874: 201, 206, foldout map; Sharpe, 1892; Boulenger, 1906: 538, map), Rufuvu River (Stanley, 1878b: 36), Lufu River (Moore, 1901b: 102; 1903: foldout map opposite page 52), and Luvu River (Poll, 1946: foldout map).
Presumably, most geographical names, including village names, occurring round the lake shores have synonyms and alternative spellings. For example, the Congolese village, Lupota, has been referred to as Mapota (Stanley 1878b: 40, 41), Mpota [in 1892, 1897, and 1902] (Roelens, 1897; Uhlig & Moisel, 1902; Luscombe, 2018a: map), Pota (Poll, 1946), Bupota [1967] (UT, 2018: map), and Upota [1969] (Luscombe, 2018e: map).
Furthermore, while most publications prior to 1950 or thereabouts may refer to the villages Kituta and Kipimbi (e.g. Poll, 1946: foldout map; Luscombe, 2018b: map from 1949), many of those later than 1950 mention them as Chituta and Chipimbi (e.g. Konings, 1988: 11). Though the French explorer and big game hunter, Édouard Foà (Coosemans, 1950a), referred to Chituta as Tchitouta in 1899 (Foà, 1899: 337, foldout map), he also referred to Kipili as Tchipiri, Kirando (Rukwa Region; Fig. 73) as Tchirando, and Ujiji as Oudjidji, spellings which may be interpreted as clues on how the names were actually pronounced in 1899, at least locally. Nonetheless, especially round the southern part of the lake, there have presumably been linguistic influences from neighbouring southern tribes of present-day Zambia and Malawi, who more commonly apply the sound/prefix chi- as opposed to ki-, hence the alteration of pronunciation/spelling of Kituta and Kipimbi to Chituta and Chipimbi.
Sometimes spelling and/or format have been deliberately altered depending on language. For example, Boulenger (1898a; 1915) referred to Tanganyika, (lake) Mweru, Mbity Rocks, and Tropheus moorii in English, but Tanganika, Moero, Mbété, and Tropheus Moorii in French (Boulenger, 1900: 148; 1901a).
Furthermore, Moore alternatingly applied Nyasa (Moore, 1897a; 1897c; 1897d; 1901a) and Nyassa (Moore, 1898a; 1898b; 1898f; 1901b; 1903), while Cunnington (1906b; 1920: 507) applied Tanganyika and Tanganika. Bujumbura was referred to as Usambura by Moore (1901b) and Usumbura by Cunnington (Boulenger, 1906: 538). Regarding the village Msamba, located adjacent to Mongwe and Kalandasi Point [see maps in Karlsson & Karlsson (2015d: 22; 2018a: 89)], Moore referred to it as Msambu and Cunnington as Msamba (Boulenger, 1901c: 138, 150; 1906: 538, 545). For certain locations, Moore and Cunnington applied the same name and spelling, prevailing at that time but mainly obsolete today, such as Kituta for Chituta, Tembwi for (Cape) Tembwe, Kibwesa or Kibwesi for Sibwesa, to name a few.
In conclusion, there are synonyms and alternative spellings for most names found round the shores of Lake Tanganyika.

32.10. Swahili grammar involving the alternative names Mbete, Pambete, Mambete, and Mombete
Regarding more relevant Swahili grammar, there are three prefixes that relate to the location of something: pa- (definite location: e.g. in, on), ku- (indefinite location: e.g. at, to), and mu- (inside location: inside). These prefixes may be significant in explaining the alternative spellings of village names, such as Mbete and Moliro (§ 32.1): Pambete (Pambeti), Mambete, Mombete (e.g. Stanley, 1878b: 36; Stewart, 1880a: 6; 1880b: 10; 1881; Sharpe, 1892) (the latter two village names possibly being misspellings of Mumbete), Pamlilo (e.g. Stevenson, 1888: first map), Pamriro (Kettler & Riemer, 1890–1895), and Pamlelos (Moore, 1897a: map). Swahili, as well as many other Bantu languages, is a language of alliteration and assonance, including repeated sounds of vowels and consonants, sometimes referred to as agreement markers, rendering the words in a sentence more compatible. Some significant prefixes may be repeated throughout the sentence. For example, the prefix pa- may be attached to several words in a sentence, possibly including a name, such as the village name Mbete. Seemingly, Pambete is part of a sentence where someone or something is stated to be located in Mbete, and the prefix pa- is partially there to make it sound more compatible with the sentence. Similarly, Mumbete (Mambete and Mombete) pertains to something being inside Mbete (mu-Mbete). Although village names are presumably not included in the current Swahili grammar, they likely were, or still are included in various Bantu languages. Furthermore, the same three prefixes as previously mentioned (pa-, ku-, mu-) repeatedly appear in the current Swahili language, always relating to location, for example: hapa [this place, here (definite)], pale [that place, there (definite)], huku [this area, hereabouts (indefinite place)], kule [that area, thereabouts (indefinite place)], and humu (in here, this inside place), mule (in there, that inside place) (Wilson, 1985: 196–205, 217). Compare the Swahili grammar with the Bantu grammar of Johnston (1919: 31), which is more or less the same. A similar explanation for the derivation of Pambete was given by Konings (2012: 8; 2013a: 18).
Alternatively, Pambete, Pamlilo, and Pamriro (e.g. Stevenson, 1888: first map; Hore, 1889: map; Johnston, 1897: foldout map facing page 40; Kettler & Riemer, 1890–1895) may be translated to Mbete’s place, Mlilo’s place, and Mriro’s place. Compare with current Swahili locative grammar: ‘-angu’ (my) and ‘pahali’ (place), but ‘pangu’ (my place). There are additional African places with names that occasionally have the prefix Pa-, such as Parumbira at Lake Malawi (Glave, 1897a: 900), normally written Rumbira/Lumbira, the former, Parumbira, which would mean Rumbira’s (or Lumbira’s) place. Obviously, Moore’s version of Moliro, ‘Pamlelo’s’ (Moore, 1897a: 299), would be tautological. However, given that Lake Malombe, south of Lake Malawi, used to be referred to as Lake Pamalombe (Livingstone, 1874: foldout map), it is more likely that ‘Pa’, at least in this case, is an alliterative marker in reference to ‘on Lake Malombe’.

32.11. The Council for World Mission Archive
Part of Konings’ conclusion (2012; 2013a; 2013b; 2013c) that Kinyamkolo is synonymous with Ulungu Escarpment was based on reports written by the co-workers of the LMS’s station at Kinyamkolo/Niamkolo during the 1890s. Statements of some of these reports are available online, though the reports themselves have not yet been digitised, but are available at the Archives & Special Collections of the School of Oriental and African Studies (SOAS), University of London.
If making an online search (6 October 2018) for Kinyamkolo and Niamkolo at the Council for World Mission Archive (https://archiveshub.jisc.ac.uk/search/), one will receive three results for Kinyamkolo and 70 for Niamkolo. The search for Kinyamkolo rendered:
(1) Report by Charles Mather, Kinyamkolo [Niamkolo, British Central Africa, later Northern Rhodesia, and Zambia], 22 Dec 1890 (SOAS, 2018a).
(2) Report by Charles Mather, Kinyamkolo [Niamkolo, British Central Africa later Northern Rhodesia, and Zambia], 19 Jan 1891 (SOAS, 2018b).
(3) Report by Alfred Swann, Kinyamkolo [Niamkolo, British Central Africa, later Northern Rhodesia and Zambia], 20 Jan 1891 (SOAS, 2018c).
Obviously, the search results for Kinyamkolo include the name Niamkolo, and Konings (2012: 10; 2013a: 21) has interpreted this as “Kinyamkolo is the name of the district in which lies the village Niamkolo where the actual mission was based”. However, this is unlikely, it is more likely that the location for the missionary station is better known as Niamkolo. In their reports, Charles Mather and Alfred Swann referred to the station as Kinyamkolo because this was the prevailing name for them. In fact, one or both of them likely introduced the name in 1889–1890, seemingly together with the re-establishment of the station, for which Swann was partly in charge (§ 32.4, 32.5, 32.7) (Hore, 1892: 293; Johnston, 1892: 250). Furthermore, none of the associates of the LMS referred to this missionary district as Kinyamkolo, but rather as Tanganyika (SOAS, 2014; 2017) or the Lake Tanganyika Mission (Cousins, 1892b; SOAS, 2018j), which was part of the larger Central African Mission, including the stations at Niamkolo, Kambole, and Fwambo/Kawimbe (Johnson, 1901: 58; Chuba, 1995: 22). For a map indicating the stations, Kambole, Fwambo, Kawimbe, and Niamkolo, see the former reference, Wallace et al. (1898), and Stewart (1903: map opposite page 222). Also, the territorial district of the British Central Africa where Niamkolo is located was referred to as Tanganyika in 1897 (Johnston, 1897: map opposite page 154).
Regarding the 70 search results for Niamkolo, 43 relate to reports, journals, letters, etc., written between 1886 and 1900. The five oldest reports (the former three are excluded) from different authors span between 1890 and 1893, and are listed below (search results for ‘Niamkolo’ sorted according to ‘Date Ascending’):
(1) Report by Alexander Carson, Niamkolo [British Central Africa, later Northern Rhodesia, and Zambia], 10 Jan 1890.
(2) Report by Alfred Swann, Niamkolo [British Central Africa, later Northern Rhodesia and Zambia], 24 Jun 1892.
(3) Report by Charles Mather, Niamkolo [British Central Africa, later Northern Rhodesia and Zambia], 9 Feb 1892.
(4) Report by James Hemans, Niamkolo [British Central Africa, later Northern Rhodesia and Zambia], 24 Jan 1893.
(5) Report by Adam Purves, Niamkolo [British Central Africa, later Northern Rhodesia and Zambia], 10 Jan 1893.
There are many additional and similar reports from the preceding five authors as well as from William Thomas (29 Dec 1894), John May (17 Dec 1897), Percy Jones (29 Oct 1898), Harry Johnson (1 Sep 1899), etc. As stated above, these 70 search results for Kinyamkolo and Niamkolo only pertain to some of the many letters that were actually sent from the Niamkolo station to the LMS’s Foreign Secretary in London. For example, not included in the online search result are two such letters, one labelled “Correspondence to Foreign Secretary from Kinyamkolo”, sent by Charles Mather on 21 March 1894, and another, “Correspondence from Niamkolo”, sent by James Hemans, five days later, on 26 March 1894 (Chuba, 1995: 333, 490, 492), both missionaries at the Niamkolo station. What can be concluded based on the online info is that most of the Niamkolo missionaries referred to the station as Niamkolo, with some of them, like Swann and Mather, referring to it as both Kinyamkolo and Niamkolo. However, it is rather likely that whatever names the station co-workers applied to their mission station in the 1890s, the current online archives do not account for them all, they primarily display names which are currently valid and proper. Subordinate synonyms are redundant and possibly confusing. Similarly, authors of species descriptions or other scientific papers may not have applied the same synonym or exact spelling as applied by some of the missionaries who occasionally made collections of plants and animals. For example, regarding collections made by Carson, Pocock (1898: 430, 432) referred to the collection location as Niomkolo, while Verdcourt (1974: 311) referred to it as Nyamkolo. However, Carson (SOAS, 2018i) himself appears to have referred to the location as Niamkolo. The proper name of the station should likely be Niamkolo, since this was the name and spelling applied in ‘The Chronicles of the London Missionary Society’ during the 1890s. However, whether Niamkolo, Nyamkolo or any other of the many synonyms is the original native name/spelling is uncertain as the natives did not have a written language and the slave traders (Brown, 1893: 149; Johnston, 1897: 155) raiding the area did not use the Latin alphabet. Also, the chronicle of the LMS never mentioned Kinyamkolo, presumably because the use of that name within the LMS was limited to some of the station associates (such as Swann and Mather) during a few years in the 1890s. Furthermore, additional publications of the LMS, such as the ‘Guide to the London Missionary Society Archive 1764–1977’ (SOAS, 2014; 2017), rarely, if ever, mention Kinyamkolo because it was a subordinate, insignificant, and short-lived synonym of Niamkolo, definitely not a district. The only native name that can be said to relate to a district where Niamkolo is located is Ulungu, as may be interpreted from the following series of citations from Hore and Swann: “Ulungu, which occupies the chief part of the south end of the lake, is a country of great importance” (Hore, 1892: 122); “The country of Ulungu occupies the whole south part of the lake, stretching, with a coast-line of at least 100 miles, from the borders of Fipa to the River Lofu” (Hore, 1892: 157); “In vain I looked for the many well-to-do villages of my old acquaintances, the prosperous and lively Walungu [...] The neighbourhood we thus examined was that of Niumkorlo and the surrounding district” (Hore, 1892: 232); “Our mission station now in Ulungu” (Hore, 1892: 238); “[P]reaching of the Word, and teaching of the young, at the two stations [including Niamkolo] in Ulungu” (Hore, 1892: 294); “We were now amongst the Walungu [the people of the Ulungu country], who owned nearly the whole of the southern end of the lake” (Swann, 1910a: 89).

32.12. Fwambo and Kawimbe
Based on further online search in the archives of the SOAS (through the Council for World Mission Archive), Konings (2012; 2013a) suggested that Fwambo is a district [southeast of Lake Tanganyika] in which the town and former missionary station Kawimbe is located, but this suggestion appears to be incorrect. A search for ‘Fwambo’ as of 6 October 2018 rendered 36 similar search results, including the following: “Report by Stewart Wright, Fwambo [Kawimbe, British Central Africa, later Northern Rhodesia, and Zambia], 05 Aug 1890”. ‘Fwambo’ outside and ‘Kawimbe’ inside the parentheses do not indicate that the former is a district and the latter a town, only that the Fwambo station was also known as Kawimbe, and vice versa (Chuba, 1995: 22, 61, 121; SOAS, 2014: 89).
Fwambo was not a district, but the original site where the mission was founded. Actually, both stations existed simultaneously for a short while until the Fwambo station was closed and its facilities moved to Kawimbe, less than 5 miles (8 km) farther north (Johnson, 1901: 58). The fact that there really was a new missionary station north of Fwambo is evident in the text by Sharpe (1957: 217), who visited it in January 1891, as well as in Gamwell and Gamwell (1961: 519). Furthermore, both Fwambo and Kawimbe are indicated on the map of Wallace et al. (1898). Owing to the unhealthy conditions that prevailed at Niamkolo, the directors of the LMS recommended their missionaries in 1886–1887 to abandon the Niamkolo station and erect a new settlement at Fwambo, a reputed healthy spot, at considerable elevation (Hore, 1892: 295; Johnston, 1892: 249–251; Jones, 1897: 74; Horne, 1908: 268, 271; Taylor & Lehman, 1961: 4; Meebelo, 1971: 6, 25, 27; Ragsdale, 1986: 19, 24; Rotberg, 2015: 151; SOAS, 2018f). The location for the Fwambo station was variously given as “some sixty miles south of the lake, on the highlands of the interior” (Cousins, 1892b: 231), “Fwambo, two days’ journey from the south end of Tanganyika” (Hore, 1892: 282), “thirty miles up in the hills [from the Stevenson Road between Lake Malawi and Lake Tanganyika]” (Brown, 1894: 138), “on the plateau between the lakes – about two days’ journey from Tanganyika” (Brown, 1895: 247), “about thirty miles from the south end of Lake Tanganyika” (Johnston, 1897: 95), and “about fifty miles from the lake” (Bennet, 1969: xxiii). Furthermore, according to Alexander Carson, one of the founders of the Fwambo station and collector of Clarias carsonii (cf. Clarias liocephalus) (Boulenger, 1903: 362), the location for his fish collection, indicated at Fwambo, is ‘21 miles east south east of the south end of Lake Tanganyika’ (NHM, 2019t). Also, from the diary of Glave (1897a: 901), it appears that Fwambo can be reached from Kituta (Chituta) by a march of, in total, seven hours. The location of Fwambo or the Fwambo station is indicated on Glave’s map (1897a: 900) (Fig. 123) and on maps in Stewart and Coles (1880), Stevenson (1888), Johnston (1890: map; 1897: foldout right map opposite page 392), Elliot (1896), Moore (1897a), Stanley (1899c), and Luscombe (2018a: map), the latter which is from 1892. For illustrations of the school house and a mission house at Fwambo, see Cousins (1892b: 232) and Glave (1897a: 902).
The construction of the Kawimbe station began in 1890 at a location some 25 or 30 miles (40–48 km) from Lake Tanganyika, in an area with a bracing, even cold climate throughout much of the year (Mather, 1897: 224; Johnson, 1901: 58; cf. Gamwell & Gamwell, 1961: 519 mentioning both Fwambo and Kawimbe). For maps indicating Kawimbe, see Wallace et al. (1898), Uhlig and Moisel (1902), Stewart (1903: map opposite page 222), and Luscombe (2018e), a map from 1969.
Following the move of the Fwambo station and subsequent merger with that at Kawimbe in 1890 (Rotberg, 2015: 151) or 1891 (Meebelo, 1971: 27), the missionary station was frequently referred to by both names, expressed as Kawimbe (Fwambo) or Fwambo (Kawimbe) (e.g. Chuba, 1995: 22, 61, 121; SOAS, 2017: 96, 98; 2018g). Occasionally, Fwambo and/or Kawimbe was referred to as the Mambwe station, the name Mambwe pertaining to the country in which the stations were located (Ragsdale, 1986: 19). Also, many missionaries died at the stations, such as Carson at Fwambo in 1896, Mather at Fwambo in 1898, Purves at Mbereshi in 1901, John May at Kawimbe in 1901, and, as already stated, Harris at Niamkolo in 1885 (Jones, 1896: 181; Chronicle, 1897: 163; Horne, 1908: 268; Swann, 1910a: 83; Goodall, 1954: 273; Rotberg, 2015: 182, 186). Regarding the sacrificing work of the missionaries and the unhealthy location of Niamkolo, see more in the caption of Fig. 121.
In conclusion, Fwambo was (and is) not a district as suggested by Konings (2012; 2013a), but a village and missionary station like Kawimbe and both villages or sites are currently located in the Mbala District.

32.13. Georges Descamps at Fwambo, 20 September 1893
Fwambo is the place where Alphonse Jacques, an officer in the Belgian antislavery campaign and the founder of the fort of Albertville (near Kalemie), met up with Georges Descamps (Fig. 126), a captain (in 1893) of the same antislavery force and the leader of an expedition transporting 100 shotguns, 30,000 pieces of ammunition, and two 47 mm Nordenfeld cannons. On 13 April 1893, Descamps’ expedition left London and after reaching Chinde at the Zambezi delta, it passed through Shire River, Lake Malawi, Karonga, Stevenson Road, Fwambo, Chituta, Moliro, eventually reaching Albertville (Kalemie) on 4 November 1893. On Stevenson Road, from Karonga and northwards, the expedition experienced “huge difficulty pulling the two canons through gorges, rivers, along paths, and on edges of valleys” (Coosemans, 1949a: 215). On 20 September 1893, Jacques and Descamps met at Fwambo, the former who made arrangement for further transportation of the two cannons to Albertville. While one of the cannons was installed at the Albertville fort, the other was used in the assaults on the Arabs, including their stronghold at Mtoa, which eventually was taken over by the Belgians and temporally named Albertville (Alexis, 1896: 106–112; Janssens & Cateaux, 1908: 337–368 (359); 1911: 56–87 (76); Engels, 1949: 501; Coosemans, 1945; 1949a: 215; cf. Marissiaux, 1961, cited in Konings, 2013d). For a map indicating Albertville at Mtoa, see Wagner and Debes (1914).

Fig. 126 (left). Georges Raoul Adolphe Descamps, 1855–1938, was an officer in the ‘Force Publique’, the private army of Leopold II, King of Belgium (Coosemans, 1949a; Janssens & Cateaux, 1911: 56–86), in what is now DR Congo (previously the Congo Free State, 1885–1908, and the Belgian Congo, 1908–1960). Upon his retirement in 1913, Descamps was appointed Major General (§ 32.13). Fig. 127 (right). Célestin Louis Marie Joseph Hecq, 1859–1910, was a captain in the same army, ‘Force Publique’, and commander of the Albertville fort at Mtoa (Coosemans, 1948; Janssens & Cateaux, 1911: 432–436) (§ 23, 31). Lake Tanganyika cichlids that bear the names of these two soldiers are Ectodus descampsii, Xenochromis hecqui, and Lepidiolamprologus hecqui.

32.14. The geological surroundings of Kinyamkolo
When Moore (1903: 46) described the country in which Kinyamkolo is located, he stated that the core of the mountains flanking the lake on the southeast coast is mainly of granitoid material, while the slopes westwards from there constitute sandstones and quartzites. These slopes dip under the water and form a deep depression, Chituta Bay, about 320 metres deep, and further westwards, at a distance of about 30 km, there is the rocky promontory of Chituta, and Mbita and Mtondwe Islands (Moore, 1903: 46). The same promontory was referred to by Hore (1882; map) as Kapata. Regarding their structure, the eastern face of the promontory or cape is a precipitous cliff of red sandstone and quartzite, rising to a height of about 180 to 245 metres; “on its western side the ridge slopes much more gradually for two or three miles, until it dips under the channel which separates the island[s] of Kinyamkolo [Mbita and Mtondwe Islands] from the mainland” (Moore, 1903: 46). Like that of Chituta Promontory, “the eastern faces of these islands are steep, precipitous cliffs, composed of somewhat metamorphosed sandstones”, while the islands slope less steeply into the lake on the western side, (Moore, 1903: 48).

32.15. Moore’s ‘craggy ledges’
The northwestern side of the rocky Chituta Promontory, more precisely near the missionary station at Kinyamkolo, is where Moore administered his dynamiting: “Nearly all the new forms which I obtained were killed by dynamite from the craggy ledges of the west coast of the lake, where the water was deep enough, about 20–35 feet [6–10 metres], but not too deep for my men to dive and procure the greater number of the fishes, which, after every shot, were invariably found on the bottom. The number of fishes in such situations is really surprising, and on several occasions, after firing a single cartridge, I obtained more than two tall negroes could well carry, when slung in a bag between them on a pole” (Moore, 1898a: 27). Regarding the phrase ‘craggy ledges’, ‘craggy’ is interpreted here to denote something stony, rugged, and uneven, while ‘ledges’ presumably pertains to layers, levels, shelves, and steps. In conclusion, Moore’s craggy ledges likely refers to the rocky, steep, and rather low walls of occasional sandstones with conspicuous layers and cracks, which are very common along the eastern side of the Chituta Promontory but decrease in both occurrence and height west towards Mpulungu. Such craggy ledges are also seen, for example, at Maswa (Fig. 131) and Molwe (Fig. 125). Incidentally, the Maswa area was described by Moore (1901b: 135) as “a wilderness of crags and rocks and trees”, a description that is somewhat reminiscent of the phrase ‘craggy ledges’ applied to the description of Kinyamkolo. Presumably, Moore partially stood on these craggy ledges to perform his dynamite fishing. Regarding the transport of heavy bags of fish obtained ‘on several occasions’, Moore and his local helpers could conveniently walk the short distance back to the camp. It is very likely, quite certainly, that the craggy ledges more specifically pertain to the small hill north of Musende Bay, referred to as Mpulungu Hill (Venning, 1960), which has a rocky shoreline and is located opposite of Mbita Island (see more in § 32.4, 32.22, 32.26). Probably not by chance, this location is illustrated in the colour sketch by Moore (1903: Fig. opposite page 68). Alternatively, Moore’s craggy ledges may be the small promontory just northeast of Niamkolo Church, currently referred to as Chikola Hill and Katanka Point on the official maps of Zambia (Government of Zambia, 1970). Particular parts of the western side of this promontory may perhaps be described as ‘craggy ledges’.
Presumably, Chikola Hill is adjacent to one of the locations where David Livingstone first saw Lake Tanganyika. On 30 April 1867, at a village he referred to as Chikula, Livingstone camped for one night before marching southwest to Mbete (Livingstone, 1874: 206, Fig. foldout map). The sketch in the previous reference appears to illustrate Kakonde Bay, east of Chikola Hill, with Chituta Promontory in the background, see also Swann (1910a: 83, Fig. bottom).
As an additional but rather unlikely option, the craggy ledges of Moore may refer to the rocky and flat mountain tops (ridges) conforming to the appearance of the Ulungu Escarpment, as suggested by Drachenfels (1995) and Konings (2012; 2013a: 22, Fig. top; 2013b; 2013c; 2015a: 198). Stanley (1878b: 36, 37), in his narrative ‘Through the Dark Continent’ described the ‘Ulungu Mountain Range’ as having ‘crags and groves’, and “cliffy heights rising in terraces”. A plain synonym of craggy ledge is rocky shelf, possibly pertaining to a mountain top or ridge. In a later edition, Stanley (1899b: 29; 1899d: 29) referred to parts of Ulungu Escarpment as “the grand cliffy walls of Kapembwa”. Furthermore, in a report by Mather (1894: 235), the “Missionary-in-charge at Niamkolo”, the Kapembwa village was referred to as “being at the base of those pillar rocks so justly considered one of the sights of Tanganyika”.
Nevertheless, Moore’s craggy ledges are unlikely in reference to Ulungu Escarpment, the latter which he specifically referred to as “the huge sandstone precipices of the west coast”, “the gigantic cliffs which stood in wild buttresses and towers of yellow sandstone thousands of feet above” (Moore, 1901b: 98, 101), “the gigantic scarps of the main western coastline of Tanganyika”, “the great western scarps”, “the main western scarps”, “the abrupt west coast”, and described to be composed of “massive quartzites, sandstones, conglomerate and shales, 2,000 feet and more being exposed along the main western coast-line of the lake” (Moore, 1903: 48). Furthermore, Moore (1898a: 27) stated that he conducted his dynamiting “from the craggy ledges of the west coast”, not ‘along or flanked by the craggy ledges’, which implies close proximity to these craggy ledges, hence, the mountain tops of the Ulungu Escarpment, which “were so far above us in the clear blue air” (Moore, 1901b: 101), appear to be a less likely alternative for Moore’s craggy ledges.

Fig. 128 (left). Craggy ledges of Singa Island. Fig. 129 (right). Craggy ledges along the coastline between Mpando Point and Chiloelo Point.
Fig. 130. Maswa.
A sketch by Moore (1901b: 65) illustrating the harbour of Maswa, currently referred to as Lubengela. Previously, the name Maswa referred to an old fortress, later a monastery located near the Malagarasi River delta (Konings, 1988: 70; 1998a: 66; 2015a: 110). However, research on Maswa does not reveal any historical fortress, but only suggests that it was a natural boat haven providing important shelter in stormy weather. Initially, Boulenger (1901c: 138) erroneously indicated the location of Maswa as about 50 miles south of Ujiji, which would be somewhere near Segunga, between Cape Kabogo and Halembe, but later (1906: 538, map) corrected the location as to right between Malagarasi River and Cape Kabogo. Moore, who referred to Maswa as Masswa, described how he and his expedition crew, after having crossed a stormy Lake Tanganyika at night from Mtoa to the east coast, “at last slipped in between two forest-clad headlands, finding ourselves at rest in the smooth water of the beautiful little harbour of Masswa” (Moore, 1901b: 65, 135). Based on Moore’s additional descriptions, including “a wilderness of crags and rocks and trees” (1901b: 135), as well as the above sketch, Maswa seems very likely to be the present-day Lubengela, located about 20 km south of the Malagarasi River delta. In the list of African names for British official use (British Official Names, 1926: 11), Maswa is synonymous with Mivozia, Mivózia, Miwosia, Maschua’s, and Masswa’s, and is indicated to be located (about) 375 km northwest of Kasanga, which is right between Malagarasi River and Lubengela. However, the publication of the British Official Names (1926) probably has a margin of error of at least 10 km. Also, Mafupa is another synonym of Maswa, for which latitude (5°25'00"S) established on Moore’s second expedition exactly conforms with Lubengela (cf. Fergusson, 1900: 391). The name Maschua’s or just Maschua presumably relates to the Swahili word ‘mashua’, which means boat. Depending on pronunciation, Maschua and mashua sound similar to Maswa. On some old maps from 1899–1925, ‘Miwosia kwa Mashua’ (Uhlig & Moisel, 1902; Moisel, 1903; Andree, 1905; Basel Mission Archives, 2018b) and Miwosia (Wagner & Debes, 1914) are indicated just south of Lugufu River, while Masoué (Foà, 1899: 337, foldout map) and ‘Miwosia [...] Maschua’  (Luscombe, 2018c) are indicated slightly farther south (nearer to Cape Kabogo). In Swahili, ‘kwa’ usually means to or for, so perhaps ‘miwosia kwa mashua’ means ‘haven for boats’ in a local Bantu language, or at least something relating to boats. Obviously, also the present-day and nearby Kirando (Kigoma Region) is a bit of a boat haven, which alternatively may have been the historical Maswa. However, at Kirando there are not “two forest-clad headlands”, only one, and Moore’s sketch does not conform as well to its coastline as it does to that of Lubengela. Furthermore, Kirando does not offer as good protection from bad weather as Lubengela does (pers. obs.), conditions which probably also prevailed 100 years ago and made (and still make) Kirando a less important natural harbour.
Perhaps the most peculiar feature of Maswa is the sandstone deposits, which are thought to be part of the old lakebed (cf. Moore, 1903: 65–66, 69, 74).
Currently, Maswa is the name of the area where the true Tropheus duboisi “Maswa” (Fig. 132) is found, which means the area between the rivers Lugufu and Msehezi, or more strictly, between the villages Kirando and Lubengela (Karlsson & Karlsson, 2016d). The rocky point just south of Kirando is perhaps the most common representative of Maswa, referred to as Maswa Point (Fig. 131).
Prior to the discovery of T. duboisi “Maswa”, Maswa was primarily known for the location of the collections made by Moore and Cunnington. For example, partly at Maswa, Moore collected the type series of Hemibates stenosoma (Boulenger, 1901c: 152), and Cunnington collected samples of algae (West, 1907: e.g. 139, 150), Polyzoa (Rousselet, 1907: 253), and parasitic copepods (Cunnington, 1913: 271) at Maswa.
Fig. 131. The craggy ledges of Maswa Point lie between the Kirando and Lubengela villages. The bay in which Lubengela is located was known to Moore as Maswa (Moore 1901b: 65, 135). See more in the caption of Fig. 130.
Fig. 132. The unique Tropheus duboisi “Maswa” is found along a 4 km stretch of shoreline between the villages Kirando and Lubengela, located between the rivers Lugufu and Msehezi (see more in the caption of Fig. 130). Many years of commercial collecting (about 45) for the ornamental fish trade has taken its toll on the population size. Today, T. duboisi “Maswa” is endangered. This photo of a pair T. duboisi “Maswa” was taken in December 2007 at the whitish craggy ledges of Maswa Point (Fig. 131) at a depth of 25 metres. For reference purposes, additional photos of Maswa Point are included in Konings (1988: 25, Fig. bottom; 1998a: 9, Fig. second from top; 2015a: 13, Fig. second from top) and Karlsson and Karlsson (2015c: Fig. 23); coordinates of Maswa Point: -5.39467, 29.76026 (5°23'40.81"S, 29°45'36.94"E). The Maswa variant of T. duboisi was discovered by Wolfgang Staeck (and his fellow travellers) in 1974 (Staeck, 1975a: 42; 1975b: 295; cf. Neergaard, 1976: 153, Tropheus-map opposite page 152), although the name Maswa was not applied until 1983, see below. Some of the earliest colour photos of T. duboisi “Maswa” are featured in Staeck (1977: 216, Fig. 162), Axelrod and Burgess (1977: 339; reproduced in, e.g., 1983: 339 and 1986: 339), Tvedegaard (1977: 324, Fig. 3), and Brichard (1978a: 304, Fig. top; reproduced in Axelrod, 1979: 30, Fig. 1). For an underwater video of T. duboisi “Maswa” and T. sp. “Crescentic” at Maswa Point, see Karlsson and Karlsson (2016d). Broad-banded variants of T. duboisi, including the Maswa variant are found from south of Lugufu River to south of Halembe Village, a distance of about 50 km. Following the discovery in 1974 and a few years ahead, Misha Fainzilber of Ocean Products Ltd, the first ornamental fish collector in Tanzania, collected broad-banded T. duboisi along its entire range (Axelrod, 1978; 1979). The variants at Miaba and Halembe in the southernmost range were of particular interest and described as being olive tinted (Miaba) or having a wide yellow band (Halembe) (Axelrod, 1978: 15, 40, 141; 1979: 29). In 1983, a broad-banded T. duboisi variant collected in the Maswa/Cape Kabogo area (Cape Kabogo is 7.5 km south of Maswa Point) and introduced as T. duboisi “Maswa” was exported by Aqua Products Ltd (Schupke, 1984: 313–314; Müller, 1985; Vaitha, 2018).

32.16. Moore’s ‘west coast’
Konings (2012: 12; 2013a: 22) claimed that ‘there is no doubt about what Moore regarded as the west coast’, implying that Ulungu Escarpment, i.e., the coast west of Mbete Bay, perfectly matches Moore’s west coast. In our view, this opinion is preconceived, what is Moore’s west coast is not that clear. For example, Moore referred to Ulungu Escarpment as “the South-West coast of Lake Tanganyika” (Moore, 1903: 67), “the great sandstone cliffs forming the west coast” (Moore, 1903: Fig. opposite page 68), or “the main western coast-line of Tanganyika” (Moore, 1903: 48).
Moore alternatingly described the collection locations, Kinyamkolo and Mbity Rocks, as the west coast and southwest coast, implying a rather unimportant description of the exact geographical position: “Nearly all the new forms which I obtained [...] from the craggy ledges of the west coast of the lake” and “The fishes which have been collected [...] from the south-western extremity of Tanganyika” (Moore, 1898a: 27).
Furthermore, in a sketch by Moore (1903: 67) comprising the coast of what is currently Tanzania and Zambia, the northern part of Chituta Promontory (mistakenly referred to as an island by Moore), and Mtondwe Island, with Kinyamkolo located west of all three localities, Moore (1903: 67) referred to the coast alone as “the East Coast” implying that Chituta Promontory, Mtondwe Island, and Kinyamkolo are not on the east coast, but possibly on the west or southwest coast. Similarly, when describing the physiographical features of the country surrounding the southern half of Lake Tanganyika, Moore (1897a: 298; cf. 1903: 48) referred to the east coast as strictly the east coast, implying the coast east of Chituta Bay.
On 3 July 1876, Stanley (1878b: 34) visited the southernmost part of Chituta Bay and declared that this was “the extreme south end of the lake”. However, Hore (1882: 18) visited the very same location on 21 April 1880, founding that Stanley was wrong and ascertained that the extreme south end of the lake is actually Mbete Bay. About 20 years later, Moore first described a photo of the Chituta Bay area as “The City (!) of Kituta and the extreme South End of Lake Tanganyika” (Moore, 1901b: 57), then two years later as “The extreme southern arm of Tanganyika from Kituta, showing the sandstone escarpment on the left” (Moore, 1903: 77, Fig.). Apparently, Moore regarded Chituta Bay as the extreme south end of the lake. If a Lake Tanganyika coast must be classified as either east coast or west coast, it is rather logic to consider the coast east of the extreme south end as the east coast, and the coast west of that end as the west coast. Presumably, that is what Moore did, therefore, if Kinyamkolo was not on the east coast (see above, this §), it must have been on the west coast. Seemingly, the coast between Chituta Bay and Mbete Bay was occasionally or frequently considered by Moore as the west coast, and although he also often regarded the coast further westwards as the west coast, he commonly referred to that coast as the main west coast: “the main western coast-line of Tanganyika” (Moore, 1903: 48). Alternatively, if Moore did not regard everything west of Chituta Bay as the west coast, at least he regarded everything west of Kasenga Point, the northernmost part of Chituta Escarpment, as the west coast, see below (§ 32.17). As a current reference conforming to Moore’s view of the west coast, in a textbook about African archaeology, both Niamkolo and Mpulungu Port are described as “on the western shore of Lake Tanganyika” (Insoll, 2015: 130; cf. Trevelyan James, 2011).
Incidentally, the description, “extreme south end of Lake Tanganyika”, was also applied by Boulenger (1901a: 138) regarding Kinyamkolo, and, similarly, Sars (1910: 736, 739) applied the description “south end of the lake” to both Chituta and Niamkolo (Kinyamkolo). Furthermore, it may also be pointed out that when Stanley visited Chituta Bay, the water level of the lake was about 11 metres higher (§ 16) than when Moore visited it, by which time the water had receded and the beach “become nearly a mile wide” (Moore, 1901b: 130), hence, Chituta Bay may indeed have been “the extreme south end of the lake” (Stanley, 1878b: 34), though, the water in Mbete Bay would, of course, also have reached farther south. Also, although Chituta Bay is not the southernmost part of the lake, it may appear as if it is, if one believes that the length of the lake follows a strict north-south direction, which it does not, it is tilted some degrees. Nonetheless, when travelling on the lake it is easy to get that impression, especially the explorers of the late 1800s may have believed so as evidenced from the following statement by the chairman of the Royal Geographical Society in 1879: “Much doubt still existed as to the exact longitudinal direction of the Lakes Tanganyika and Nyassa” (Stewart, 1881: 271). Following Moore’s second expedition, 1899–1900, the understanding of the inclination of the north-south axis was greatly improved (Anonymous, 1900: 267; Troyer, 1991: 39), but still not correct, at least regarding the longitudes, as the lake has an even greater inclination than what was obtained from the expedition’s compass survey [see the coordinates in Fergusson (1900: 391)].

32.17. Chituta Bay
In the late 1800s and early 1900s, travellers commonly reached Lake Tanganyika from Lake Malawi by the so-called Stevenson Road (Fig. 142) (Johnston, 1897: 71; Haskard, 1965; cf. Brown, 1893: 257; Moore, 1901b: 60–63), which initially was nothing more than a narrow path, partly hidden by a lush forest. On Moore’s second expedition he described the road: “Here the old native path which used to exist, worn down below the surface of the ground by the silent tread of generation after generation of Ulungus, had been converted into a broad track by the felling of trees” (Moore, 1901b: 75). From south to north, it started at Karonga, Lake Malawi, and passed through, Fwambo, Abercorn (currently Mbala), among other sites and villages, ending at Chituta, Lake Tanganyika, where travellers continued by sailboat or steamship (Moore, 1897a: map). See a map of Stevenson Road between Karonga and Chituta (Kituta) (Stevenson, 1888: foldout map opposite page 1; Wallace et al., 1898; Uhlig & Moisel, 1902; Stewart, 1903: map opposite page 222; Wagner & Debes, 1914; Carlin, 2018a).
Besides Moore and many other travellers, this route was taken by, for example, Charles Lemaire and his scientific expedition to Katanga, who left Chituta with the steamer Good News, arriving at Moliro at midnight on 3 August 1898. However, two of the expedition members, the Belgian geologist Jean De Windt (Lemaire, 1902: 21, Fig.) and the British prospector William Caisley, who travelled a week later, were hit by a storm and drowned near the southern coast of Moliro (Lemaire, 1902: 7, 18; Coosemans, 1950b: 929). For a photo of some of Lemaire’s expedition members, see Lemaire (1902: iv), for a map of the complete expedition route, see Lemaire (1906: 19), and for another map indicating the location for the accident where De Windt (aged 22) and Caisley were drowned, see Janssens (2005b); see also SGM (1901: 44–45) for geographical details.
As he was approaching Chituta, Moore (1901b: 79) described the first sight of Lake Tanganyika: “The dark red cliffs on the west and the pale purple feathery surface of the forest on the hills rising to the east, together with the nearer setting of the green marshes, form wonderful contrasts with the blue water”. According to Moore (1901b: 80, 84), Chituta used to be a place where the African Lakes Corporation [formerly called the African Lakes Company, a commercial enterprise originally with close connections to the Free Church of Scotland (Johnston, 1897: 67–79)] had established a little trading station, the African Lakes Store, selling “tools, solder, spare parts, red lead, etc.”. Upon arrival at the lake, Moore (1901b: 84) spent the first two days at Chituta, where he was busy repairing and cleaning up the Good News (Fig. 124), which had been dumped in the bay by the owners, the African Lakes Corporation, who had bought the ship from the LMS in 1895–1896 (Gamwell & Gamwell, 1961: 520). Moore was expecting to find a boat in working condition as this was the arrangement he had made with the African Lakes Corporation in London prior to his expedition, but instead he found a wreck: “The condition of the boat [Good News] was beyond all description, and it was notorious throughout the land” (Moore, 1901b: 108–109). Presumably feeling deceived, he set out to Kinyamkolo to find help mend the steam-boiler, which was completely ruined. From Chituta to Kipata (Kapata), across the crocodile infested swamps, he travelled in a dug-out with his “face a foot above the water”, and after having “tripped over a submerged root” while wading ashore and landed “head downwards in about four feet of pitchy, stinking mud”, he climbed the appearing sandstone cliffs with the scorching sun on his back, which he described as some thirty minutes of “foretaste of hell”, before coming out on top of Chituta Escarpment (Moore, 1901b: 84–86). Looking west and northwards, he described his view of the lake as constituting a great valley, referred to as the western arm, the present-day Mbete Bay, with its western shoreline, the Ulungu Escarpment, “soaring in sheer precipices two thousand feet or more in height”, and its eastern curving shoreline of yellow sand on the lake shore, partly bounded by the lake “in changing shades of ruffled blueness between the islands and the coast” (Moore, 1901b: 86). Although this curving shoreline is the eastern coast of Mbete Bay, it still constitutes a part of the ‘western arm’, hence, the shoreline between Mbete Bay and Kasenga Point, the northernmost tip of the land separating Mbete Bay and Chituta Bay was regarded by Moore as the eastern part of the west coast. Also, as stated previously (§ 32.16), alternatively, Moore’s west coast may have included the western side of Chituta Bay. However, if Chituta Bay was not regarded by Moore as a dividing point, separating the west coast from the east coast, Kasenga Point surely was.
Presumably, Chituta Bay did not have an official name by the end of the 1880s. Johnston suggested the name Rhodes Bay, as well as Port Rhodes for a new port which was planned to be constructed on the east coast of Chituta Bay (Johnston, 1890: 739 & map; 1897: foldout map opposite page 154; Sharpe, 1892; Carlin, 2018h). However, the port obviously did not materialise and the name Rhodes Bay did not catch on. Indeed, the bay was subsequently referred to as Kituta Bay (Cunnington, 1899), currently Chituta Bay.
Additional explorers who passed through Chituta are, for example, Vice-Consul Alfred Sharpe (1893: 525), Foà (1899: 337, foldout map), and the German officer Paul Graetz with a motorcar in 1907–1908 (Graetz, 2006; Carlin, 2018b).
Recapitulating the essential details of the two previous sections (§ 32.16, 32.17), while Moore’s east coast was the coast east of Chituta Bay, his west coast was that west of the same bay, alternatively west of the nearby Kasenga Point.

32.18. Moore’s second visit to Kinyamkolo (Niamkolo), 1899–1900
During Moore’s second Tanganyika expedition, 1899–1900, he revisited the missionary station at Kinyamkolo (Niamkolo) (Boulenger, 1901c: 137). Upon arrival at the station in 1899, following the short march from Chituta (§ 32.17), Moore found a Scottish craftsman which was willing to help him repair the ruined steam-boiler of the Good News, this was possibly Adam Purves (§ 32.4, 32.6, 32.12). After some briefing, Moore returned to Chituta to fetch the steamer, where after it was cleaned, its joints tightened up, and its boiler received a new cylinder top. Although Moore (1901b: 109) indicated Kalambo as the site of the steamer renovation, based on contradictory facts, it was most certainly Kinyamkolo (Niamkolo). Following the renovation, Moore and his expedition members “decided to start up the lake, first going to Kalambo, then to Sumbu, and then away to the north”, which included Moliro, crossing the lake to the islands of Msamba, and farther north to Kirando (Rukwa Region) (Moore, 1901b: 109–110, 114, 118).
After these two rather transient visits in 1899, Moore never returned to Kinyamkolo (Niamkolo). However, probably during his second stay while renovating the steamship, he managed to obtain a small collection of fish and possibly other animals. The fish collection constituted only four specimens, which were all labelled ‘Kinyamkolo’, including one specimen of Chrysichthys myriodon, two of Tylochromis polylepis, and one of a Lates species (Boulenger, 1901c: 147, 149, 154; 1911: 340; 1915: 383; NHM, 2019o; 2019p). The former two species were described by Boulenger (1900: 139, 143) based on a combination of Moore’s collection and that of Hecq, which originated from Albertville (Mtoa), DR Congo. While Moore made additional fish collections farther north, such as at Kalambo, Msamba, Sibwesa, and Maswa (Figs. 130–131), he did not collect anything, at least not fish, at the two nearby locations he visited west of Kinyamkolo, i.e., Sumbu and Moliro (cf. Boulenger 1901c: 138; Moore, 1901b: 110; NHM [Natural History Museum] (2014).
Regarding Konings’ suggestion (2012; 2013a) that Moore’s Kinyamkolo is synonymous with Ulungu Escarpment, it is extremely unlikely that Moore would travel to Ulungu Escarpment, or make a stop there en route for Sumbu, to collect three single fishes without obtaining anything else in particular to report. Moore would hardly find motivation or time to do that as his mind was rather occupied with the idea that if the fauna of Lake Tanganyika constituted “the relic of an ancient freshwater fauna, this old fauna ought to appear in some of the other lakes besides Tanganyika”, therefore, he was in a hurry to pass northward “to examine as many lakes as possible” (Moore, 1901c: 467). Consequently, it would have been more convenient, as well as much cheaper, to add a few fishes to his collection by obtaining these from the shores near where he was preparing the steamer for his big journey up north, which is highly likely what he did, hence, the labels on these four fish specimens, ‘Kinyamkolo’, most certainly relate to the shores of the immediate vicinity of the missionary station, i.e., Moore’s Kinyamkolo is not synonymous with Ulungu Escarpment but with Niamkolo. However, Moore probably did not collect the fishes himself but more likely obtained them from a fisher.
For the record, Boulenger identified the single Lates specimen from Kinyamkolo (1899–1900) as L. microlepis (1901c: 149), later as L. angustifrons (1915: 110). Currently, it is identified as L. mariae (Steindachner, 1909a: 399; NHM, 2019u), a species Boulenger (1915: 108) placed in synonymy with L. microlepis.
In addition, between September and November, 1899, Moore obtained specimens with location recorded as only Lake Tanganyika, including three specimens of Citharinus gibbosus, two each of Asprotilapia leptura [Xenotilapia (Takahashi, 2003: 375)] and Neolamprologus furcifer, and one specimen each of Alestes macrophthalmus and Julidochromis ornatus; likely, these were collected at Kinyamkolo [cf. § 32.19 (1) and (8)].

32.19. More logic of Moore
The following additional reasoning supports the view that Moore’s Kinyamkolo is synonymous with Niamkolo, contradicting the statements by Drachenfels (1995) and Konings (2012; 2013a; 2013b; 2013c; 2015a) that Moore’s Kinyamkolo pertains to a district in which Niamkolo is located and/or the coast along the Ulungu Escarpment.
(1) During Moore’s two Tanganyika expeditions, he collected fish (Boulenger, 1898b; 1901c), frogs (Boulenger, 1898c), sponges (Evans, 1899), snails/shells (Moore, 1898e; 1898f; 1901c; Digby, 1902), crabs (Cunnington, 1899), shrimps/prawns (Calman, 1899), earthworms (Beddard, 1901; 1902), lizards [e.g. Lygosoma modestum (now Mochlus sundevalli) (Günther, 1880: 235; Boulenger, 1895b: 168); NHM, 2019a], a polyzoan (Arachnoidia raylankesteri), and more, including rock samples (Moore, 1903). A few samples, which appear to have been labelled only ‘Lake Tanganyika’, were organisms that Moore presumably regarded as occurring throughout the entire lake, hence, the exact collection locations may have been regarded as less important. Alternatively, these organisms may also have been labelled more precisely, but their labels disregarded in subsequent lectures and articles in favour of ‘Lake Tanganyika’. All samples seemingly without a proper label, such as the sponges and protozoans, were presumably collected at Kinyamkolo. Regarding all other samples, Moore carefully labelled these with rather precise collection locations, which included Kituta (Chituta), Sumbu, Mleroes (Moliro), Mbity Rocks, and Kinyamkolo, see references above (§ 32.1). Moore knew very well that species or varieties of animals, including fish, may be locally endemic, therefore it was important to label them properly, including a precise location: “That the fishes differ in different portions of the lake I have myself observed” (Moore, 1898a: 27). “In the waters of the lake [Tanganyika] there are also hundreds of fishes; [...] there are innumerable rock-fishes here” (Moore, 1901b: 90). If Chituta, Sumbu, Moliro, and Mbity Rocks are rather well-defined and readily identifiable villages or sites, why would Kinyamkolo, where Moore collected most of his fish samples, not be that too? The suggestion from Drachenfels (1995) and Konings (2012; 2013a; 2013b; 2013c; 2015a) that Kinyamkolo is a district or the coast of Ulungu Escarpment offering no preciseness appears highly unlikely. Most certainly, Moore’s Kinyamkolo is synonymous with Niamkolo, currently Mpulungu.
(2) Furthermore, it would be rather improper and confusing to treat four out of five collection locations as constricted places, but the fifth (Kinyamkolo) as an area stretching along the Ulungu Escarpment, about 40 km. Moore would certainly understand that this would cause confusion and refrain from such a location labelling procedure. Logically, all five collection locations are of similar rank or extent; unless clearly indicated, one (Kinyamkolo) is not an extended area (Ulungu Escarpment), while the other five are narrow locations.
(3) If Moore collected most of his fishes along the Ulungu Escarpment, as indirectly suggested by Drachenfels (1995) and Konings (2012; 2013a; 2013b; 2013c; 2015a), why did he refer to this area as Kinyamkolo, one of the alternative names for the missionary station? There were several villages or geographical peculiarities with names that he could apply, at least there were some years earlier when Stanley (1878b: 36–37) and Hore (1882: 21; 1892: 226, map) visited the area. Also, it would have been very easy for Moore to obtain, not only some of those village names, but also the name Ulungu (Escarpment, Mountain Range, or Plateau) had he wanted. Therefore, Moore did most certainly not collect anything along the Ulungu Escarpment and his Kinyamkolo is synonymous with Niamkolo, the missionary station where he was camped and renovated steamships during his expeditions.
(4) In an almost completely unknown lake, where more or less all fish species are new to both science and a collector, hence equally interesting, why would Moore choose to conduct the major part of his fish collection at a remote location, such as along Ulungu Escarpment, instead of near to his headquarters, where he had ready access to required facilities, which he presumably did not have along Ulungu Escarpment? The answer is, of course, that he would not. Crossing the Mbete Bay would be a needlessly complicated way to get fish. Moore’s fishes labelled ‘Kinyamkolo’ were obtained very near where he was camped, the Niamkolo missionary station, where the gathering of samples was much more convenient, effective, and cheaper. Also, a collection location near the missionary station would facilitate a quick and easy return back to the camp, which may have been required since “on several occasions”, Moore “obtained more than two tall negroes could well carry, when slung in a bag between them on a pole” (Moore, 1898a: 27). Two people walking with a pole between them all the way from Ulungu Escarpment, or even Cape Chaitika (cf. Konings, 2012: 12; 2013a: 22), would really have been unnecessarily complicated. Also, the fishes may possibly have been damaged following such a long march. Seemingly and rather logically, Moore collected the largest part (two thirds) at or near his camp (Kinyamkolo), a somewhat smaller part (one third) a little farther way (Mbity Rocks), and the smallest part (a single specimen) at the most remote location (Sumbu). As already stated (§ 32.1), his intention at Sumbu was primarily to dredge for deep-living and marine-like gastropods, but while doing that, he found a strange little fish, Synodontis multipunctatus, which he decided to add to his fish collection.
Logically, one makes the bulk collection where one is camped or has one’s headquarters with important facilities. For example, during our fish collections in 2008, we first collected all lake-wide species at or near our camp (Udachi, Kabwe), later complementing the collection with locally endemic species and variants from throughout the lake.
(5) Relating to the previous argument, at Albertville in 1899, Célestin Hecq collected two specimens of Tropheus, subsequently designated as the types for T. annectens (Boulenger, 1900: 148). Previously, it was commonly believed that the collection actually occurred at Mtoto (e.g. Brichard, 1989: 149–150; Konings, 1998a: 66) or near Kalemie (e.g. Nelissen, 1979: 28; Staeck, 1985: 106), the latter previously referred to as Albertville. However, following the realisation that the Albertville of Hecq is not Kalemie, but rather Mtoa, where there was a fort named Albertville (§ 32.13) of which Hecq was in charge (Janssens & Cateaux, 1908: 152; 1911: 432–436; Coosemans, 1948), Konings (2013d) has suggested that the two type specimens originate from the Mtoa area, in which indeed a T. annectens-like population occurs. Furthermore, according to Konings (2013d: 13), “it is not with absolute certainty that the types of T. annectens were collected at Mtoa/Kavala, but it is rather unlikely that Hecq would collect his specimens during forays into other parts of the lake while he would have had a lot more time and better opportunities when at the fort”. While that statement makes perfect sense, the logic of it also applies to Moore and the missionary station, Niamkolo, where he was camped, i.e., Moore made the major part of his fish collection at or near the missionary station, which he referred to as Kinyamkolo.
(6) Furthermore, Moore’s expedition was not overly lavish, he had to make do with minimal means. While he pointed out the importance of being “heavily equipped” for “effective operations”, he did not have that much technical support, but had to work with “native dug-out boats, and with nothing better than the natives themselves as motive power” (Moore, 1898a: 26). Also, instead of fully focusing on his scientific task, prior to the use of the rented vessels, he had to locally renovate them, something which he was personally involved in [Moore 1901b: 93 (first expedition), 108 (second expedition)] (§ 32.17, 32.18). Boulenger’s remark (1901c: 137) that Moore’s expedition was “fully equipped for a complete survey of the lake” was presumably a response to his disappointment that the expedition had failed “to bring home representatives of any but the few families already known from the lake”. Of course, Boulenger’s remark was also a very deliberate statement contrary to Moore’s persistent theory that the fauna of Lake Tanganyika was partly of marine origin, something which Boulenger seemingly regarded as nonsense. However, evidence suggests that some of the lake’s fauna, such as the herring species, Limnothrissa miodon and Stolothrissa tanganicae, actually derives from the ocean and entered Lake Tanganyika by means of land-locked marine bodies of water that were transformed in size, position, and salinity over a period of about 50 million years (Wilson et al., 2008). Indeed, Moore never gave up his theory of a marine origin (Moore, 1904; 1906). Nonetheless, for Moore more than 100 years ago, or for anybody today, it is not without costs to travel on the lake and make collections. Since it was more expensive to collect fish at Ulungu Escarpment, Moore likely focused on the area near his headquarters, i.e., the Niamkolo missionary station.
(7) If Moore’s Kinyamkolo is Ulungu Escarpment that would mean that Moore did not collect anything, not a single specimen, at or near the missionary station, where he was camped, spent most of his time (Nature, 1897: 258), and had access to facilities, some which probably were necessary or at least made collection much easier. That Moore would not collect anything around the missionary station is extremely unlikely.
(8) The main reason why Moore did not explain more thoroughly or indicate on maps where he obtained his fish collections was likely because there was nothing secretive or ambiguous about his collection location, Kinyamkolo, therefore nothing to further explain. Explicitly or implicitly expressed, Kinyamkolo was and still is synonymous with Niamkolo, also for Moore in 1895–1896 and 1899–1900. If, however, Moore’s Kinyamkolo would not be synonymous with Niamkolo, but with, for example, Ulungu Escarpment, then it would have been highly expected from Moore to pinpoint his collection location and not lead his readers and the scientific community into confusion, but he did not, because his Kinyamkolo is indeed synonymous with Niamkolo. Following Cunnington’s home coming in 1905, and the commencement of systematic research on his collections, the scientific community rapidly adopted the name Niamkolo, disfavouring Kinyamkolo; as an example from current research, in a recent scientific article on frogs, Moore’s Kinyamkolo is synonymous with Niamkolo (Largen, 2001: 289). Moore did not have any objections to this; based on his response, there was no error in treating Kinyamkolo as synonymous with Niamkolo. Furthermore, presumably all samples without a more precise label than just ‘Lake Tanganyika’ were readily to be understood as ‘Kinyamkolo, Lake Tanganyika’, because Kinyamkolo/Niamkolo was where his camp was located, he spent most of his time, and certainly obtained the major part of his collection.
(9) Konings’ suggestion (2012: 9; 2013: 20) that a part of Moore’s fish collection was made along the coast of the Ulungu Escarpment is, to some degree, based on a sketch by Moore (1903: Fig. opposite page 68), captioned “View from a point on the top of the sandstone cliffs flanking the South-West Coast of Lake Tanganyika”. Konings regarded the sketch as proof that Moore visited Ulungu Escarpment at least once and hinted that while he was there, he may have obtained his fish collections. However, that sketch only suggests that Moore visited the top of the Ulungu Escarpment, not that he actually visited the shores 2,000–3,000 feet (~615–915 metres) below, and collected fish, obtaining more than two tall men could easily carry (Moore, 1898a: 27). Nonetheless, this does not mean that Moore never was near these shores, for the record, Moore travelled through this area several times, both over land and water (Moore, 1897a: 299), and he went to Sumbu by boat at least twice (Moore, 1901b: 93, 110). Possibly, Moore made the sketch while making geological and physiographical surveys of the landscape (cf. Moore, 1903: Chapter 3). Also, a little southwest, at Kambole, there was a missionary station founded in 1894–1895 (SOAS, 2018g; 2018h), which Moore may have visited while surveying the area. Furthermore, Moore possibly passed over or near the Ulungu Escarpment while en route for Lake Mweru, where he collected both frogs and fish, the latter which were not brought back home (Boulenger, 1898c: 479; 1899b: 84). Moore also marched to Sumbu several times passing over or near the escarpment. In conclusion, Moore’s sketch (1903: Fig. opposite page 68) of Ulungu Escarpment is seemingly proof that he had been there at some point, but not proof that a part of his fish collection was obtained from there.

Fig. 133. A sketch by Moore illustrating a view from the Ulungu Escarpment with Mtondwe Island and Kasenga Point of Chituta Escarpment (near the centre of the sketch), with the lake’s east coast (present-day Tanzania) in the background (Moore, 1903: 67). Possibly, Moore made the sketch while making geological and physiographical surveys of the landscape (cf. Moore, 1903: Chapter 3). Also, a little southwest, at Kambole, there was a missionary station founded in 1894–1895 (SOAS, 2018g; 2018h), which Moore may have visited while surveying the area. Furthermore, Moore possibly passed over or near the Ulungu Escarpment while en route for Lake Mweru, where he collected both frogs and fish, the latter which were not brought back home (Boulenger, 1898c: 479; 1899: 84). Moore also marched to Sumbu several times passing over or near the escarpment. See more in § 32.19 (9).

32.20. The tale that Tropheus tells
Konings’ theory (2012; 2013a; 2013b; 2013c) that Moore’s Kinyamkolo pertains to the coast along the Ulungu Escarpment, including Cape Chaitika, is strongly connected to Boulenger’s description (1898b) of Tropheus moorii. According to Konings (2012: 6–7; 2013a: 17), Moore “made color sketches of many of the fishes collected, which he gave to Boulenger so he could use them in the descriptions [...]”. As a correction, Moore probably did not make sketches of many fishes, but only of some, or, at least, Boulenger only based his descriptions on some of them: “His [Moore’s] series of sketches executed from fresh specimens, free use of which he has kindly given me, thus enabling me to represent some of the new species in their natural colours” (Boulenger 1898b: 2). Presumably, the sketches only illustrated fishes that (1) Moore collected but failed to bring home, (2) were in bad condition already prior to being placed in preservative, (3) would possibly come out as badly preserved upon arrival back home, or (4) had spectacular living colours Moore knew would fade or disappear due to the preservatives (cf. Boulenger, 1898b: 2). Also, some of the preceding categories of sketches may have included fish species which already had been described elsewhere, therefore were excluded by Boulenger in his report of new species (cf. Boulenger 1898b).
The species with colour sketches include Cyathopharynx furcifer, a Labeo species (‘Labeo, sp. inc.’), Bathybates ferox, Synodontis multipunctatus, Chrysichthys cranchii (non Leach, 1818), and Citharinus gibbosus (Boulenger, 1898b: 14, 25, plates 4 & 8; 1899a: 95; 1899b: 94; 1901c: 143; 1906: 548–549). The fact that these species were included in Moore’s series of colour sketches is seen in Boulenger’s text and drawings (1898b; 1899a; 1899b; 1901c; 1906). Possibly, a few more species were included. However, some of the collected species were obviously not, such as Ophthalmotilapia ventralis, a characteristically blue species (at the type locality), but described as “[g]rey-brown above, with or without irregular darker spots, silvery below; fins grey-brown, the ends of the ventral filaments white” (Boulenger, 1898b: 14), Cyprichromis leptosoma, always with a partly blue colouration and a blue or yellow caudal fin, but recorded as “[b]rown, lighter beneath; dorsal and anal with or without brown longitudinal streaks; caudal spotted with brown or black at the base” (Boulenger, 1898b: 15), Simochromis diagramma, typically with bright reddish or yellowish features in the south, but reported as “[o]live, whitish beneath; [...] gill-membrane sometimes with dark brown spots; fins greyish; a dark brown stripe may be present along the spinous dorsal, which has a black edge” (Boulenger, 1898b: 19), and Petrochromis polyodon, characteristically with many bright yellow features, including a yellow collar at the type locality, but described as “[o]live-brown, whitish beneath; fins grey or blackish” Boulenger 1898b: 20). Furthermore, Boulenger (1898b: 8–9) did not mention the yellow pectoral fins of N. modestus, which the species typically or frequently has, at least regarding the populations found east of Mbete Bay (§ 24, 27). Obviously, the colouration of the preceding four or five species were certainly not based on any colour sketches by Moore. Furthermore, the blue spots ascribed to Eretmodus cyanostictus (Boulenger, 1898b: 16) probably relate to successful colour preservation and not to a possible colour sketch of the fishes, which unlikely exists.
Besides his colour sketches and possibly brief notes next to the sketches, Moore did not make any additional notes on the colouration of the species like Cunnington did: “The preservation of the specimens [in Cunnington’s collection] is excellent and their value is enhanced by a careful labelling of every one of them, in most cases accompanied by either sketches or notes concerning the life-coloration, on which I have largely drawn in preparing the present account” (Boulenger, 1906: 537). Also, the Lemaire Expedition in 1898 appears to have obtained much information about the colouration of the fishes: “The Lemaire collection is further valuable for the care with which coloured sketches of most of the fishes have been taken on the spot by M. [Léon] Dardenne, the excellent artist attached to the expedition” (Boulenger 1899a: 87).
Although all thorough taxonomists of today include information on living colours in the species description along with the colours of the preserved material, this has not always been the case (Winston, 1999: 208). Presumably, most of Boulenger’s descriptions were generally based on preserved material with no information on living colours, though the collections of Cunnington and Lemaire appear to have included more information about living colours than the average collection. Apparently, the collections of Moore did mainly exclude information about living colours, so unless explicitly or implicitly stated, the colours of Moore’s fishes described by Boulenger (1898b) should be understood as relating only to preserved material. Regarding the description of T. moorii, there is nothing in Boulenger’s text (1898b: 17–19) that implies that it includes information of living colours, therefore, his colour description “[d]ark brown; a large bluish-white blotch on each side; belly reddish brown; fins blackish” pertains to preserved specimens. However, Konings (2012; 2013a; 2013b; 2013c) has suggested that these are living colours and that the only variant of Tropheus that accords with the description is the one found at Cape Chaitika. Since Moore indicated Kinyamkolo as collection location for T. moorii, Moore’s Kinyamkolo would then be Cape Chaitika. However, Konings is most likely wrong, the colours are not living but of preserved material, and the description does not match very well with the Cape Chaitika variant, but much better with any of those found along the shores near Mpulungu, including Kasakalawe. In fact, the size and placement of the large lateral blotch perfectly match with the variants of the Mpulungu area, which should not come as a surprise as Moore’s Kinyamkolo is synonymous with Mpulungu.
For comparison of the lateral blotch of the type material with that of the variants from Cape Chaitika and the Mpulungu area, see the following easily obtainable publications; for sketches and photos of the type: Boulenger (1898b: Plate 5, Fig. 2; 1915: 277), Schäfer (2003: 32, Fig. top, 33), and Konings (2013a: 17); for photos of live and mostly wild individuals of the variants of Cape Chaitika and the Mpulungu area: Konings (1988: 27, 36; 2013a: 23, 60, 61, 65, 122; 1998a: 56; 2005: 177, 178; 2015a: 79) and Schupke (2003: 154, 155).
There are five important characteristic features of the blotch of the types of T. moorii: (1) it is almost centred at the body, (2) horizontally, it covers only about half of the trunk, (3) vertically, it extends to almost the base of the dorsal fin, while it ends before the belly rendering a dark (brown) narrow abdominal area, (4) it is rather distinct, i.e., it does not gradually merge with the colours of the rest of the flank, and (5) apparently, it does not easily fade away or disappear, i.e., it is maintained despite stress, including preservation. The described blotch of the type perfectly matches with that of live specimens from the Mpulungu area. However, this is not the case when compared with the blotch of the Cape Chaitika variant, which is primarily placed on the lower half of the trunk, including the belly, where it sometimes fully extends throughout the lower half of the trunk. Frequently, there is no blotch on the Cape Chaitika variant, occasionally it is very indistinct, gradually merging with the brownish colour of the rest of the trunk. Also, the occasional blotch quickly disintegrates upon stress.
Regarding the colours of preserved specimens, when most fishes, including Tropheus, are placed in preservatives, such as formalin, the colours alter. Of course, death itself may also affect the colouration. Nonetheless, following the preservation, the somewhat dark colours usually become darker, at least in Tropheus. Green and blue may become blackish, brown changes to reddish brown, with light colours usually becoming washed-out and whitish. After some weeks or months in formalin, the yellow of any Tropheus variant changes to a whitish colour that may be described as a very pale bluish-white (Fig. 136). The latter is more or less exactly how Boulenger described the lateral blotch of T. moorii. Also, he described the abdominal area as reddish brown and the fins, including the dorsal fin, as blackish. Obviously, this description perfectly matches with a preserved Tropheus specimen from the Mpulungu area. Presumably, the blotch of the Cape Chaitika variant would readily disappear upon preservation like the frequent or occasional blotch does of similar variants, such as those between Kala Bay and Kalambo, Tanzania, including Tropheus “Red Rainbow” (Fig. 135) (cf. Karlsson & Karlsson, 2016c).
Furthermore, as mentioned above, Boulenger described the dorsal fin as blackish. If Moore had made a sketch of either the Tropheus variant from Cape Chaitika or Mpulungu, he would certainly not illustrate it as blackish but as blue mixed with some red. However, since Boulenger did not have a colour sketch of T. moorii, he described it as blackish as this was and still is the colour of the large preserved specimens (cf. Schupke, 2003: 32, Fig. top). Also, the bluish-white blotch of Boulenger does not apparently derive from living colour because both the Mpulungu and Cape Chaitika variants almost always have a yellow or yellowish blotch, not bluish-white.
Although the Tropheus variant from Kasakalawe is not identical to that from the adjacent Mpulungu, it is very similar, therefore, it relatively well represents the original T. moorii. In the absence of comparative photos of the Mpulungu variant, photos of the Kasakalawe variant work well enough.
For the record, Brichard (1989) is not the best reference for the purpose of comparing the type of T. moorii with the variants from Cape Chaitika and the Mpulungu area, as quite a few of his Tropheus captions are incorrect. For example, T. moorii of the Mpulungu area is captioned ‘T. moorii kasabae, Kalambo’ (1989: 148, Fig. right, bottom), Tropheus of the Kalambo-Chituta area is captioned ‘T. moorii kasabae, Kapemba Bay’ (1898: 161, Fig. left, third from top) and ‘T. moorii kasabae of the Kabeyeye variety’ (1989: 178), and T. moorii kasabae from Cape Chaitika, near the Lufubu outlet, is captioned ‘Tropheus sp. from Nkamba Bay’ (1989: 175, Fig. top). However, some photos which indeed illustrate the Cape Chaitika variant are found on pages 160 (right, third from top), 171, 175 (top), and 189.
In conclusion, the description of T. moorii (Boulenger, 1898b: 17–19) is exclusively based on preserved specimens without information about the living colours, and the five specimens comprising the type series were collected by Moore at Kinyamkolo, currently Mpulungu.

Fig. 134. The type locality of Tropheus moorii is Kinyamkolo, today’s Mpulungu. The London Missionary Society was based at Kinyamkolo, where Moore stayed during both his visits to the southern parts of Lake Tanganyika. The picture shows T. moorii from Kasakalawe, freshly collected by our team in 1996. The natural colours of T. moorii were first described by Wolfgang Staeck after his travels to the southern parts of the lake in 1975 (Staeck, 1977: 243–244; Schupke, 2003: 94).
Fig. 135. Tropheus “Red Rainbow” from Kambwimba, preserved specimen, deposited at the Swedish Museum of Natural History, NRM 60271. The yellow blotch of the “Red Rainbow” is not intense enough (unlike that of T. moorii “Mpulungu”) to be maintained after preservation, consequently has completely vanished.
Fig. 136. Tropheus sp. “Kaiser” from Kekese, preserved specimen, deposited at the Swedish Museum of Natural History, NRM 57976. Following fixation and preservation in formalin and alcohol, the bright yellow colour of a live T. sp. “Kaiser” diminishes and transforms into a dingy bluish-white colour, as seen in this photo. For similar reasons, Boulenger (1898b: 18) described the yellow/orange blotched T. moorii of Kinyamkolo (Mpulungu) as having “a large bluish-white blotch on each side”, referring to the colour pattern in preservative. See more in § 32.20.

32.21. Kinyamkolo: a partly swampy type locality
There are reasons to believe that Moore’s Kinyamkolo was not a purely rocky habitat, but rather one with sand and swamps nearby conforming to the immediate surroundings of the missionary station.
For example, (1) at Kinyamkolo, Moore collected a frog (Boulenger, 1898c: 479–480), which presumably favours a somewhat swampy habitat. Besides the improbability that Moore would collect such a relatively unimportant sample anywhere else than near his camp, Loveridge (1933a: 21) described where the species could be found as “among the still verdant grass of a cattle pasture near the lakeshore [at Niamkolo]”. Such a habitat is presumably much more common near Niamkolo than along the Ulungu Escarpment, where it may not even exist, or existed about 100 years ago. According to Takahashi et al. (2009: 3111), there are no shell beds under water along the shores of Ulungu Escarpment, which suggests that sandy and swampy areas on land are less common.
(2) Moore collected several catfish species at Kinyamkolo which prefer a rather swampy surrounding, including Clarias liocephalus (Boulenger 1898b: 24). Greenwood (1957: 72, cited in Seegers, 1996: 210) recorded this or a very similar species to prefer a habitat of marginal grass and water-lily swamps, while Seegers (1996: 210) described the natural diet as including vegetable matter. Presumably, the species may be rare in a typical rocky habitat.
(3) Moore collected a species of Cyathopharynx at Kinyamkolo (Boulenger, 1898b: 14). While C. furcifer, which is somewhat of a eurytopic species found in rocky, sandy, and sometimes sediment rich habitats throughout the southern half of the lake, possibly also the northern, is potentially a common species of the partially swampy surroundings of Kinyamkolo/Niamkolo, the closely related C. sp. “Neon Streak” (Fig. 137), a common species in the south and frequently referred to as C. foae (Konings, 2015a), is rarely found in such habitats, but rather exclusively in characteristically rocky and sediment poor habitats. Furthermore, at locations where both species exist, C. furcifer tends to be very rare. The fact that Moore’s collected species was C. furcifer indicates that his collection location was not typically rocky but rather rocky, sandy, and swampy, like the habitat at or near the missionary station and not, or at least to a lesser extent, like that or those along the Ulungu Escarpment.
For the record, Cunnington made collections at Niamkolo/Kinyamkolo in 1904, eight years after Moore. Among these, Boulenger identified two specimens of Paratilapia furcifer (syn. C. furcifer) (Boulenger, 1898b: 14) and three of Tilapia grandoculis (syn. C. grandoculis) (Boulenger 1899a: 94), the latter name possibly synonymous with C. sp. “Neon Streak” (Boulenger, 1900: 150; 1901a: 477; 1906: 563, 573; cf. Regan, 1920a: 43). The collection of C. grandoculis may suggest that the Niamkolo/Kinyamkolo area is not all rocky, sandy, and sediment rich, but partly characteristically rocky and sediment poor. However, it is rather uncertain if Boulenger could reliably distinguish between the different Cyathopharynx species, as they are very similar, basically only differing by male nuptial colouration and possibly female fin colouration (Konings, 2015a: 275), something which Boulenger did not really have access to. In general, Boulenger assigned many cichlid species with conical or unicuspid teeth to Paratilapia, such as C. furcifer, and those with bi and/or tricuspid teeth to Tilapia, such as C. grandoculis. Only later was it understood that the characteristics of the teeth may change ontogenetically, i.e., transform by age. Possibly, Boulenger would not have described C. grandoculis as a distinct species had he compared Cyathopharynx specimens of the same age, let alone assigned it to a separate genus, but rather regarded it as synonymous and conspecific with C. furcifer. Though, Boulenger (1895a: xi) knew that the study of skeletal rather than dental characters confers “a more rigid definition of the higher groups”, including genera, as well as “a more natural coordination of their contents”. Subsequently, Boulenger addressed the difficulty of generic classification based upon dental morphology (Boulenger, 1901a: 393), something which he became increasingly aware of declaring “some species showing bi- or tricuspid teeth when young and unicuspid teeth when adult” (Boulenger, 1907: 459) and “the dentition in certain species being subject to variation, according to age, or even of a purely individual nature” (Boulenger, 1915: 134).
In addition, the description of Ectodus foae (Vaillant, 1899: 221), which is based on a single specimen of a total length of (only) 81 mm, features such a deficient diagnosis that it is quite impossible to judge whether this species is really synonymous with C. grandoculis as suggested by Boulenger (1900: 150; 1901a: 477; 1915: 266). Although the redescription of E. foae (Vaillant, 1908: 561–565) offers more details, the most absorbing part of that may be the sketch of the type, which resembles Ophthalmotilapia sp. “Whitecap”. Interestingly, Pellegrin (1904: 345) originally introduced Ophthalmotilapia with two species: O. boops and E. foae. Also, the follow-up work of 1908 includes a colour sketch of C. furcifer, referred to as Paratilapia furcifer, but Vaillant (1908) did not conclude that E. foae may be closely related to it. In fact, although Boulenger (1901c: 154) stated that there is a considerable skeletal difference between Ectodus, including E. longianalis (cf. Enantiopus melanogenys), and Paratilapia, including C. furcifer, Vaillant (1908: 564–565) still regarded E. foae to be more closely related to E. descampsii and E. longianalis. See more in the caption of Fig. 137.
In a future article, we will include many more details about the genus Cyathopharynx, including several additional species.
In conclusion, Boulenger’s specific identification of Cyathopharynx specimens, including those collected by Cunnington at Kinyamkolo/Niamkolo, may be regarded as rather uncertain pending a taxonomic review of the genus.

Fig. 137. Commonly found in characteristically rocky and sediment poor habitats, Cyathopharynx sp. “Neon Streak” occurs throughout the southern part of the lake, except for a stretch of 100 km in the southeast where it is replaced by similar species. It is frequently referred to as C. foae (Konings, 2015a), a dubious species originally described as Ectodus foae based on a single specimen with a total length of (only) 81 mm, collected by the French explorer and big game hunter, Édouard Foà, somewhere in the southern half of Lake Tanganyika (Vaillant, 1899: 221). Based on its rather poor diagnosis, it is quite impossible to judge whether E. foae is synonymous with Tilapia grandoculis as suggested by Boulenger (1900: 150; 1901a: 477; 1915: 266; § 32.21) and whether it is distinct from C. sp. “Neon Streak”, which obviously lacks a proper taxonomic diagnosis. Boulenger (1901a: 477) examined the type of E. foae, presumably finding bi and/or tricuspid teeth, partly thereby placing it in synonymy with T. grandoculis, here referred to as C. grandoculis. However, around 1900, Boulenger was largely unaware of the ontological change of cichlid teeth, assigning specimens of unicuspid and bicuspid teeth to separate genera (§ 32.21). Furthermore, no feature excludes E. foae from being synonymous with C. furcifer. The female fin colouration, as pointed out by Konings (2015a: 275) being a distinguishing feature in Cyathopharynx, was not observed and reported by Boulenger, Vaillant, or any other taxonomist around 1900, not only because the genus was not erected until 1920, but possibly because of inattentiveness and poorly preserved type specimens. Moreover, while Vaillant (1899; 1908) primarily compared E. foae with E. descampsii and E. longianalis (cf. Enantiopus melanogenys), Pellegrin (1904: 345) grouped it with Ophthalmotilapia boops, which Boulenger (1901a: 477) stated to be a close relative, differing mainly by the smaller and more numerous lateral scales.
In addition, in the southern half of the lake additional undescribed Cyathopharynx species occur, further complicating the matter. The most characteristic feature of C. sp. “Neon Streak” is the multiple, small, upper flank blotches displayed by males in nuptial colouration. The photo illustrates a male C. sp. “Neon Streak” at Katondo, depth 15 metres. 

32.22. The specific location for Moore’s Kinyamkolo
Contrary to Konings’ statement (2012; 2013a; 2013b; 2013c), neither Moore nor the missionaries of the LMS treated Kinyamkolo as a district, but rather as a synonym for Niamkolo. More specifically, Moore’s Kinyamkolo, where he collected his fishes, is rather certainly the rocky hill with the rocky shoreline described by Moore as having craggy ledges, later referred to as Mpulungu Hill (Venning, 1960) and located immediately west of the present-day Mpulungu Port, opposite to Makosi Hill, the southern part of Mbita Island (§ 32.4, 32.15, 32.26, 32.27). This place is depicted in a colour sketch made by Moore (1903: Fig. opposite page 68) and in illustrations included in the following publications: Brown (1893: 237), Johnston (1897: 32), and Carlin (2018e; 2018g). As an alternative to that place, Moore’s Kinyamkolo could be somewhere along the western side of the peninsula located northeast of LMS’s stone church and comprising Chikola Hill and Katanka Point. For more details, see the official maps of Government of Zambia (1970), the captions (Figs. 121, 123–124), and the maps depicting Moore’s Kinyamkolo (Figs. 138 and 142).

Fig. 138. Map of the southernmost parts of Lake Tanganyika showing Moore’s Kinyamkolo (including an alternative location) and Mbity Rocks (arbitrarily plotted), type localities for several well-known cichlids, including Tropheus moorii, Neolamprologus modestus, Julidochromis ornatus, and Cyprichromis leptosoma. LMS = London Missionary Society. *Wonzye Point, type locality for Cyprichromis coloratus and Neolamprologus cancellatus. **Kasenga Point, type locality for Cyphotilapia gibberosa, Cyprichromis zonatus, and Petrochromis horii. ***Mtondwe Island, type locality for Benthochromis horii. ****Mbita Island, type locality for Haplotaxodon trifasciatus and Xenotilapia rotundiventralis. Name reference: Official maps, Government of Zambia (1970).
While Moore’s Kinyamkolo most certainly is the present-day Mpulungu, he very likely collected some or most of the rock-dwelling cichlids at the rocky Mpulungu Hill (1), alternatively Chikola Hill and Katanka Point (2), north of the Niamkolo stone church. The location of Mbity Rocks is arbitrarily plotted partly based on information of “rocks in shallow water, Mbete” from West (1907: 158, 174) and others, including Calman (1906: 202), Kirkpatrick (1906: 220), and Smith (1906: 182). Most likely, Mbity Rocks are located somewhere between Mbete and Kasakalawe, see more in § 32.25 and 32.26. The exact position is not known today, but given the fact that the current water level is about the same as in 1895–1905 when Moore and Cunnington visited Mbete, the rocks are presumably still located in shallow water and not in deep water or on dry land. Possibly, a thorough survey of Mbete Bay would reveal the exact location of Moore’s enigmatic rocks.
Furthermore, the location of Mbete1 primarily derives from Hore (1882; 1883; 1889; 1892: 226, map) and Konings (2012; 2013a; 2015a), partly Stewart and Coles (1880), as well as Moore (1897a: map), while the location of Mbete2 is based on Sharpe (1892: map), partly Livingstone (1874: map) and Stanley (1899c), as well as the website of Bauer (2019; Gregor Bauer, pers. comm.). The location of Mbete in 1867 where Livingstone first saw the lake and about 30 years later when Moore visited the same area is not necessarily the same. Not only did the lake level drop about 10 metres between 1877 and 1884 (Camus, 1965: 1244; § 16, 32.17) causing lake-shore villages to be moved, frequently villages were abandoned because of the slave trade and possibly re-established elsewhere under the same name. “Pambetè [Mbete] will be ever memorable as the spot where Livingstone first reached Lake Tanganyika. [...] At that time (April, 1867) Pambetè was a thriving and prosperous village. [...] When the East African Expedition reached it in the end of 1879 a blight seemed to have fallen upon the village, and it had almost dwindled out of existence” (Thomson, 1881, vol 2: 1). Joseph Thomson visited Mbete in April 1880, after which he reported: “We found the village almost totally deserted, and in ruins. [...] Probably there is hardly a vestige of Pambetè left by this time, and the next traveller who visits the spot will find a jungle or a field of corn where once the thriving village stood” (Thomson, 1881, vol 2: 211). What had befallen Mbete were the slave traders raiding the country. “Pambete [...] used to be the most important place in the district, but lately the Babemba raids have forced the people to live among the hills” (Stewart, 1880a: 6).

32.23. Cunnington’s and Moore’s references to Niamkolo Island and Kinyamkolo Island
Potentially, if Niamkolo Island is an island off Niamkolo, then Kinyamkolo Island is an island off Kinyamkolo, and if Niamkolo Island is synonymous with Kinyamkolo Island, then Niamkolo is synonymous with Kinyamkolo.
Mbita Island was referred to by Cunnington as Niamkolo Island and by Moore as Kinyamkolo Island, though Moore also applied the same name to the neighbouring Mtondwe Islands. Kinyamkolo Islands (Moore, 1898d: 182) was not necessarily a proper name for Moore, but rather a description of two islands near Kinyamkolo of which he did not know the native names, as evident by his comments: “which separates the island of Kinyamkolo from the mainland” and “between the islands of Kinyamkolo and the great western scarps” (Moore, 1903: 46, 48). Undoubtedly, Cunnington and Moore referred to Mbita Island as Niamkolo Island and Kinyamkolo Island, respectively, due to the fact that the island was, and still is, located immediately opposite to and in the vicinity of Niamkolo/Kinyamkolo. Although Cunnington obtained collections from Mbita Island, including specimens of Telmatochromis temporalis and various algae and plankton (Boulenger, 1906; West, 1907), he did not bother to obtain the native and proper name of the island, probably because the provisional name suited his purpose better. Similarly, Moore may have considered it unnecessary to introduce two additional names, which the readers and/or listeners would have to learn and of which Moore would have to describe the location; had they wanted, they could have chosen any names in the publications of Livingstone (1874), Stanley (1878b), Thomson (1881), and Hore (1882), which they probably had access to. Of course, they could also have asked the missionaries or the natives for the names. With the descriptive provisional name, ‘islands of Kinyamkolo’, the readers and/or listeners instantly knew that Moore referred to some islands in the Kinyamkolo vicinity, so proper names and locations were less important, partly because he did not make any collections there, or observed anything important to report.
In conclusion, Moore’s ‘Kinyamkolo Islands’ are most certainly in reference to the islands located in the surroundings near the Kinyamkolo missionary station and village, not in the Kinyamkolo district as indirectly suggested by Konings (2012; 2013a; 2013b; 2013c).

32.24. Moore, the cavalier
In his books, articles, and lectures, Moore may have incorporated mistakes and exaggerations regarding his collection locations, descriptions about these, etc., as he had a rather arrogant style for important details. “Moore was well known for the cavalier way in which he misspelt the names even of his own genera and this extended to his own name” (Verdcourt, 1994). Some of these genus names included Bathanalia and Chytra (Moore, 1898e; 1898g), occasionally referred to by Moore as Batanalia and Kytra (e.g. Moore, 1898d; 1901c). He also spelt the genus names of other taxonomists incorrectly, such as Tiphobia and Tanganyicia (Smith, 1880; Crosse, 1881), referred to by Moore as Typhobia and Tanganyikia (e.g. Moore, 1898d, 1899a). His own name was variously referred to, partly also by himself, as Salvin-Moore, Moore-Salvin, Shorec, Shorrock, plus about 15 different name-styles (Troyer, 1991: 43, Table 1). Occasionally, his second name, Edmund, was misspelled as Edmond or mistaken as Edward, and his third initial, S (Sharrock), was erroneously assumed to correspond to Salvin, his wife’s name (Troyer, 1991: 35). When Moore died in 1947, he asked to be buried without a gravestone beside his wife, who had passed away 20 years earlier and on whose gravestone, it partly said: “wife of John Edmund Shorrack Moore” (Moore, 2012b: 8). As already stated (§ 22), his full name was John Edmund Sharrock Moore. “The absolutely incredible gay abandon [in a happy and carefree way] with which Moore treated names, particularly his own, easily explains these differences” (Verdcourt, 1994). Furthermore, some of his friends, colleagues, and others, he mistakenly referred to as Boulanger (Boulenger), Rhodes (Rhoades), Johnstone (Johnston), and Glaive (Glave) (e.g. Moore, 1901b: 6, 46, 50; 1898a; 28). He alternatingly referred to the place where the LMS had a station as Kinyamkolo and Nyamkolo (e.g. Moore, 1901b: 84, 90). Presumably, he mistakenly stated Kalambo instead of Kinyamkolo as the place where he renovated the steamship, the Good News (Moore, 1901b: 109). Also, he was in error when referring to the region or protectorate where Niamkolo is located as Northern Rhodesia, as that name was not adopted until 1911 (as the name for a country, currently Zambia). In 1901, British officials referred to the region or protectorate as North-Eastern Rhodesia (cf. Moore, 1901b: e.g. 63, 64, 67, 72).
Moore interchangeably referred to Mbete (Moore, 1897a: 298, 299), Mbiti (Moore, 1903: 48), and Mbity (Moore, 1903: 164), all relating to the same place, the village Mbete. Previously (Karlsson & Karlsson, 2018a: 74), we suggested Mbity to be an English spelling-version and anglicised designation of Mbete, but Mbity was more likely (one of) Moore’s version of Mbete. Furthermore, he referred to Moliro in at least five different ways: Mleroes (Moore, 1898b: 452; 1898e: 92), Mlelos, Mleloes (Moore, 1901b: 110, 128), Pamlelo’s, and Pamlelos (Moore, 1897a: 299 & map), and he alternatingly applied Nyasa and Nyassa (§ 32.9). He interchangeably referred to the location of Kinyamkolo as the west, south west, and south coast, and on one occasion, referred to cichlids as “cyclids” (Moore, 1901b: 90).
Boulenger (1898b) reported Moore’s first Tanganyika expedition as to have taken place in 1895–1896, but Moore (1901b: 3) described it as to have started in 1895 and ended in 1897, the same years which are stated also in Troyer (1991: 32), Verdcourt (1994), and Troelstra (2016: 310). Apparently, Moore’s travel route included the Suez Canal and by December 1896, he notified his safe return to Zanzibar (Nature, 1896b: 159).
Although Moore understood the importance of a correct record of collection locations (§ 32.19), he was occasionally arrogant or ignorant towards it. For example, he recorded the collection locations for the Lake Malawi cichlids, Gephyrochromis moorii, Protomelas pleurotaenia, and Aulonocara trematocephalum (Boulenger, 1901b: 4; 1901c: 156–158; Trewavas, 1946: 244–245; Eccles & Trewavas, 1989: 46–49, 145), as ‘mouth of Rusisi River’ and/or ‘north end of Lake Tanganyika’. Contrariwise, his sample of Petrochromis nyassae (an alleged synonym of P. polyodon; Matthes & Trewavas, 1960: 350; cf. Regan, 1922: 675) was mistakenly specified to have been obtained from Lake Malawi (Boulenger, 1902: 70). Also, Moore collected the type material of P. andersonii, as it was named by Boulenger (1901d: 13), and indicated the collection location of it as Albert Edward Nyanza, currently Lake Edward. The type was reinvestigated by Trewavas (1946: 241), who concluded that it was Tilapia tanganicae (now Oreochromis tanganicae) and that the location given by Moore was in error, it should have been Lake Tanganyika.
Presumably, Moore did not know that about two thirds of his 1895–1896 fish collection were from Kinyamkolo, while about one third was from Mbity Rocks, because he wrote “Nearly all the new forms which I obtained were killed by dynamite from the craggy ledges of the west coast” (Moore, 1898a: 27). The craggy ledges are Kinyamkolo, from where he obtained about two thirds, which is not “nearly all”. Mbity Rocks, which are most likely some rocks near Mbete, alternatively, the rocky habitat at Kasakalawe (§ 32.25, 32.26), do not have craggy ledges. Perhaps Moore knew that only two thirds were from the craggy ledges but still chose to refer to these as nearly all. In any case, his description is inaccurate and causes confusion.
Occasionally, Moore provided provisional collection data that was incomplete. For example, the collection location for Bathybates fasciatus was initially communicated to Boulenger as “the west coast of Lake Tanganyika”, but later adjusted to Tembwi (Cape Tembwe, DR Congo) (Boulenger, 1901b: 3; 1901c: 153). Moore may have had additional information about his collections in his diary that was not received in time for the species descriptions. However, in a reversed fashion, sometimes Boulenger provided more accurate information on Moore’s fishes in the provisional descriptions than in the full descriptions, as seen in the description of Hemibates stenosoma, which initially was recorded to be based on three specimens (Boulenger, 1901b: 2), but later in a more detailed description, it said “several specimens” [but still only three] (Boulenger, 1901c: 152; NHM, 2019q; 2019r). Also, regarding the description of Paracyprichromis nigripinnis from Msamba, Boulenger wrote “two specimens” in his initial diagnosis (Boulenger, 1901b: 3), but “several specimens” [still only two] in his more detailed description (Boulenger, 1901c: 152; NHM, 2019s).
Moore (1898c: 162) spoke about deep-water crabs, prawns, and sponges as if they were particularly deep living, which included the sponge Nudospongilla moorei. However, if the biotope is right, they may be found in less than 15 metres. For example, there are many Nudospongilla moorei-like sponges (Figs. 38–40) in shallow water at Mwila Island, Kipili Archipelago.
Moore referred to Mpimbwe as Mpimbwi or Mpimpi, Karema as Karama, Kirando (Rukwa Region) as Kilando or Kirendo, (Lake) Mweru as Mwero, (Lake) Bangweulu as Bangweolo, Msamba as Msambu, Maswa as Masswa, Mbete as Mbity or Mbiti, Moliro as Pamlelo’s, Pamlelos, Mlelos, Mleloes, or Moliros, to name a few locations; no wonder he also applied an alternative name and spelling for Niamkolo, i.e., Nyamkolo and Kinyamkolo (Moore, 1897a: 299 & map; 1901b: 84, 110, 126, 128, 135; 1903: map opposite page 8).
In 1908, when Moore was a professor at the University of Liverpool, fellow of the Royal Geographical Society, the Zoological Society, and the Linnean Society, he suddenly left his job, resigned all fellowships, withdrew from all scientific research and professional obligations, and for about 20 years disappeared from public view. It was the same year that his father died, and rumours had it that Moore received a large sum of money from his mother, allowing Moore with wife and son to shrug off the harness of work. Perhaps he was tired of the criticism of his many colleagues regarding his theory of Lake Tanganyika as a former Jurassic sea, his refutation of imperialism and colonialism in Africa, and possibly his occasional and slightly non-patriotic attitude (e.g. Boulenger, 1901c; 1906; Moore, 2012a; 2012b; cf. Moore, 1901b: 7–8, 139, 333–334; Troyer, 1991: 40).
The abrupt way in which Moore ended his scientific career implies that he was more a passionate adventurer than a thoughtful scientist. His narrative, ‘To the Mountains of the Moon’ (Moore, 1901b), eminently readable, is an adventurous tale that does not give any other impression. Indeed, at the time of Moore’s death (15 January 1947), his occupation was listed as “explorer (retired)” (Troyer, 1991: 33). Of all Moore’s accomplishments, he was probably most satisfied and proud of being the first to reach and explore one of the summits of the Ruwenzori Range, which had been identified with Ptolemy’s ‘Mountains of the Moon’, where he ascended to the snow line at about 3,544 metres and proved the existence of permanent glaciers (Troyer, 1991: 40; Verdcourt, 1994; Troelstra, 2016: 310).
In conclusion, the details about Moore’s reports are, unlike the Ten Commandments, not written in stone. They should perhaps best be regarded as conformably adjustable, still, Moore’s Kinyamkolo is not Ulungu Escarpment but most certainly Niamkolo/Mpulungu.

Fig. 139 (140). Portrait of John Edmund Sharrock Moore (10 May 1870 – 15 January 1947), a biologist, author, and explorer of Lake Tanganyika and Central Africa. This painting once hung in the Tresco Island Hotel, the work of either J. E. S. Moore himself or his son, Osbert. The Moore family lived on Tresco, the second largest of the Isles of Scilly off the coast of Cornwall, England, for over twenty years, in the house that was later to become the Island Hotel.
In 1895, Moore embarked on the first of two scientific expeditions to central Africa, specifically the areas around Lake Tanganyika. The expeditions were mainly sponsored by the Royal Society, later the Royal Geographical Society, resulting in numerous academic papers (§ 22). In total, Moore produced 71 publications and presentations to societies.
After his second expedition in 1899–1900, Moore wrote and illustrated an intelligible account of his adventures called ‘To the Mountains of the Moon’, which was not aimed at the professional zoologist but rather the lay reader. The book received mixed reviews, some criticising Moore on his negative attitude towards colonialism in Africa.
In 1908, Moore came into a considerable inheritance (possibly from his mother) and quite suddenly, gave up all his scientific research, withdrawing from all his professional obligations and fellowships.
After a gap of nearly two decades, Moore reappeared in public in the 1920s, when he arrived on Tresco with his wife Heloise and their son Osbert (born in 1905). The exact year when the Moores arrived on Tresco is not known, but it was certainly J. E. S. Moore’s home from 1927 to his death. While living on Tresco, Moore (known on the island as ‘Old Moore’) wrote and published a book in 1934 called ‘Five Foolish Virgins’. It is a rather strange book, a blend of autobiography and facts with a playful imagination, humour, and a bit of alternative reality.
On 15 January 1947, J. E. S. Moore died of heart failure in West Cornwall Hospital, Penzance. He wished for no headstone to mark his grave but is thought to have been buried beside his wife in the island’s churchyard; the weatherworn inscription reads: “In memory of Heloise, daughter of Osbert Salvin of Hawkesfold Sussex and wife of John Edmund Shorrack Moore who died 4 Nov 1927 [sic]” (§ 32.24).
Along with the above portrait and couple of watercolours that now hang in the Racket Town Cottage, the only other evidence of the Moores’ presence on Tresco is the gravestone [illustrations and references from James R. Troyer (1991) and Alasdair Moore (2012a; 2012b)].
J. E. S. Moore is commemorated in seven fish patronyms, including Cyrtocara moorii, Gephyrochromis moorii, Tropheus moorii, and Variabilichromis moorii, see more in Appendix 1.

32.25. Mbity Rocks: Moore’s second fish collection location
Besides the specimens of Synodontis multipunctatus and Amphilius platychir, Moore collected all his fishes of the 1895–1896 expedition at Kinyamkolo and Mbity Rocks. Following the above review (§ 32.1–32.24), it appears most certain that Moore’s Kinyamkolo is synonymous with the present-day Mpulungu. As for Mbity Rocks, although Moore never explicitly indicated their exact location, it is very likely that they are some submerged rocks near the village Mbete. Regarding the collection location for the catfish, Anoplopterus platychir [Pimelodus platychir and Amphilius platychir (Günther, 1864a: 134), see Appendices 3 and 4], Moore indicated “marshes near Mbity” (Boulenger, 1898b: 24; Moore, 1903: 164), which undoubtedly refers to some marshes near Mbete, likely the “noxious swamp” formed by the nearby Izi River (Stewart, 1880a: 6). In a subsequent Belgian publication that included Moore’s fish collection, Boulenger (1901a) substituted all Mbity Rocks to Mbété. Presumably, Mbity was Moore’s rather arrogant spelling-version of Mbete (Moore, 1903: 164) and, as stated previously (§ 24), he occasionally referred to it as Mbete (Moore, 1897a: 298, 299) and Mbeti (Moore, 1903: 48). Though, one should perhaps not judge Moore too hard for his alternative names and spellings. Indeed, he had a propensity for word-coinage and came up with words like ‘halolimnic’ for the Lake Tanganyika forms appearing marine, ‘meiosis’ for a type of cell division, ‘synapsis’ and ‘gemini’ relating to chromosomes, among others (Troyer, 1991: 37–38). Alternatively, Moore rarely or never heard the natives pronouncing Mbete, but primarily his British compatriots; therefore, Mbity may have been Moore’s unbiased transcription of his verbal reference of Mbete.
Mbity Rocks have occasionally been referred to as Kumbula Island (Mbita Island) (Schupke, 2003: 84; Takahashi & Hori, 2006; 189), which is most certainly incorrect. Konings suggested Mbity Rocks to be the present-day Kombe (also known as Gombe and Gombi) on the west coast of Mbete Bay (Konings, 2012; 2013a), or Kasakalawe on the opposite coast of the same bay (2015a: 198). The latter suggestion, which appears plausible, was adopted in Karlsson and Karlsson (2018a: 74). However, the fact that there indeed appears to be some submerged rocks near Mbete makes Kombe especially and possibly also Kasakalawe less likely and Mbete more likely as the real location for Mbity Rocks. Furthermore, since Moore collected both Cyprichromis leptosoma and Julidochromis ornatus at Mbity Rocks, two species which do not exist along the west coast of Mbete Bay, i.e., the Ulungu Escarpment, all locations along that coast, including Kombe, can be ruled out as possible locations for Mbity Rocks (see more in the next section, § 32.26).
In 1905, nine years after Moore’s expedition, Cunnington made various collections near Mbete, which was referred to by others: “Mbete, south end of lake, on rocks in shallow water” (Smith, 1906: 182), “Mbete, taken on rocks, shallow water” (Calman, 1906: 202), “from rocks, shallow water, Mbete” (Kirkpatrick, 1906: 220), and “rocks in shallow water, Mbete” (West, 1907: 158, 174). Seemingly, these rocks in shallow water adjacent to Mbete might be Moore’s Mbity Rocks. Potentially, they could be the same rocks Konings referred to as Katoto (not to be confused with the site by the same name found a few km farther north) but dismissed due to the very shallow water in which they are found (Konings, 2012: 9; 2013a: 19). However, the lake level may have been slightly higher in 1895–1905, though not as high as in the 1870s and 1880s (§ 16). Indeed, Moore recorded that the lake had receded from its old front noticeably (Moore, 1901b: 138, 141, 143). Furthermore, attention needs to be paid to the water level fluctuation occurring over one or several years. Worthington and Ricardo (1937: 1065) stated the annual fluctuation to be about one metre, while Codrington (1902: 601) observed that the “level of the lake in June, 1901, was 4 or 5 feet [1.2 or 1.5 metre] higher than in the corresponding month of 1900, the Lukuga outlet having again silted up”. Also, the lake level is about three metres higher today than it was 10–15 years ago (pers. obs.) (§ 16). Regarding the habitat of Kasakalawe, although the immediate shoreline appears to be mostly reedy, there are plenty of underwater rocks of varying sizes nearby. In the 1990s, our team collected T. moorii “Kasakalawe” among those rocks (Fig. 134). Also, see the photo in Konings (1998a: 13, Fig. bottom) of the reedy shore of Kasakalawe, including a slightly visible partly submerged large rock. To a certain degree, the distribution of these underwater rocks extends south towards Mbete, the southernmost which possibly include both Moore’s Mbity Rocks and Cunnington’s “rocks in shallow water, Mbete” (West, 1907: 158, 174). Furthermore, at the time of Moore’s visit in 1896, Kasakalawe and Mbete appear to have been equally known villages (cf. Hore, 1882: 21). However, Moore never mentioned Kasakalawe, at least not in his publications, which may suggest that he never visited Kasakalawe or made any collections there, which in turn indicates that Mbity Rocks are nearer to Mbete than to Kasakalawe. For a photo of the Mbete beach in 1926, 30 years after Moore presumably was there, see Carlin (2018d).

32.26. Proof from the occurrence of Cyprichromis leptosoma and Julidochromis ornatus
Recapitulating the previous section, Moore’s Mbity Rocks most likely refer to some rocks, possibly submerged, in the southernmost part of Mbete Bay, potentially those referred to as the Katoto Rocks on the bay’s west coast (Konings, 2012; 2013a), or those located farther east somewhere between Mbete and Kasakalawe (cf. Konings, 2015a; Karlsson & Karlsson, 2018a). However, following the proceeding reasoning, it will be concluded that Mbity Rocks, as well as Moore’s Kinyamkolo, are inevitably located on the east coast of Mbete Bay. Undoubtedly, Mbity Rocks are some rocks near Mbete or Kasakalawe, and Moore’s Kinyamkolo is located farther north east, quite certainly adjacent to the present-day Mpulungu Port, where there is a hill with a rocky shoreline, alternatively, along the rocky western shore of Chikola Hill and Katanka Point.
There are two species collected by Moore which hold a key to the location of Mbity Rocks and Kinyamkolo: Cyprichromis leptosoma and Julidochromis ornatus. Five specimens of J. ornatus were collected at Mbity Rocks (Boulenger, 1898b: 12); of one of these was a skeleton produced and another sent to the Paris museum (MNHN) (Boulenger, 1915: 485; Blanc, 1962: 214). It is widely recognised that J. ornatus exists only on the east coast of Mbete Bay, not on the west coast including Ulungu Escarpment (e.g. Konings, 2015a: 197; Karlsson & Karlsson, 2018a), therefore, Mbity Rocks must be located on the east coast of Mbete Bay, which means that the Katoto Rocks of the western side of the bay, as well as the Kombe/Gombe rocky site nearby, may be ruled out as a possible location leaving the offshore eastern area between Mbete and Kasakalawe as the only realistic and pragmatic area in which Mbity Rocks can be found. The southernmost group of submerged rocks in this area is plausibly Mbity Rocks. The water level of 1896 when Moore made his collection appears to have been about the same as of today, hence, Mbity Rocks are also today located offshore and not on dry land. On a side note regarding J. ornatus, in a bay on the east side of the lake, where the striped leech was very common, Moore observed in 1896 a small fish, whose back was striped like the leech, and this seemed to protect it against the raids from the kingfishers, which avoided picking up this particular species (Nature, 1897: 258). Possibly, this species was J. ornatus.
As for C. leptosoma, Moore collected two specimens at Mbity Rocks and two specimens at Kinyamkolo (Boulenger, 1898b: 14–15). Like J. ornatus, C. leptosoma does not occur on the west coast of Mbete Bay, only on the east coast (Konings, 2015a: 197), which is why Kinyamkolo can also be ruled out as being located on the west coast (Ulungu Escarpment). If Mbity Rocks are Kasakalawe (or located near Kasakalawe), then Kinyamkolo cannot also be Kasakalawe, but must be farther northeast. In addition, there are no craggy ledges at Kasakalawe (see § 32.15, 32.22, 32.24). To some degree, craggy ledges are found at the present-day Mpulungu Port (Mpulungu Hill) and the Chikola Peninsula, 3 and 5 km north east from Kasakalawe, respectively.
For the record, there is the possibility that Moore’s two specimens of C. leptosoma from Kinyamkolo were not really C. leptosoma, but a similar species, such as C. coloratus or a C. coloratus-like species (cf. Takahashi, 2016), which indeed exists along the west coast of Mbete Bay (Ulungu Escarpment), as well as along the east coast. If so, they could technically have been collected somewhere along the west coast, thus be an argument in favour of Moore’s Kinyamkolo possibly being located along the west coast (Ulungu Escarpment). However, this loose and rather far-fetched possibility is extremely unlikely as it would not only collide with all the facts presented in this article in favour of Moore’s Kinyamkolo being synonymous with Niamkolo/Mpulungu, it would also conflict with Takahashi and Hori’s identification (2006) of Moore’s Cyprichromis specimens. Although Boulenger (1898b) might have failed to distinguish between C. leptosoma and C. coloratus (he failed to distinguish between several similar species, see Appendices 3, 4, and 5), it is extremely unlikely, practically impossible, that Takahashi and Hori (2006) failed to do so when investigating Moore’s specimens.
In conclusion, Moore’s four specimens are all C. leptosoma and Kinyamkolo is inevitably located east (northeast) of Mbete Bay.

Fig. 141. Lake Tanganyika in October 2008, sunset over Wampembe Point, home to one of the lake’s most threatened fish species, Tropheus cf. moorii “Murago Tanzania”. Despite several warnings of imminent extinction (Karlsson & Karlsson, 2012a; 2014d; 2014e; 2014f; 2015e; 2015f; 2016e; Konings, 2014; Staeck, 2015a; 2015b; 2015c), since September 2014 parts of the worldwide aquarium fish community have provided strong incentives to local collectors to continue the heartless hunt for the very last remaining individuals of the species. Lake Tanganyika fisheries basically operate under an open access regime. This is clearly untenable and will lead to overexploitation (FAO, 1997: 52; Mölsä et al., 1999: 11). The growing human population is driving the world’s wildlife and ecosystem towards a total collapse. The natural world is disappearing before our eyes, in Africa, as well as everywhere else on the planet, and collectively, we let it happen. Since the dawn of civilisation, humanity has caused the loss of 83% of all wild mammals and half of plants. Of all the mammals on Earth today, 96% are livestock and humans, only 4% are wild mammals (Carrington, 2018). David Attenborough warns that human activity has taken the world into a new era, threatening to undermine civilisation. “We can exterminate whole ecosystems without even noticing it” (Quote from David Attenborough’s speech at Davos 2019).

32.27. Summary
In summary, Moore’s Kinyamkolo is most certainly synonymous with Niamkolo and relates more specifically to the hill, Mpulungu Hill, with a rocky shoreline immediately west of Mpulungu Port, alternatively, the western shore of the Chikola Peninsula, which is partly rocky and located north of the LMS’s old stone church. Moore’s Mbity Rocks are the southernmost group of rocks, completely or partly submerged, located in the eastern part of Mbete Bay, somewhere between Kasakalawe and Mbete. Some of the statements on which these conclusions are based include: the wide use of synonyms and alternative spellings of lake shore village names, including Kinyamkolo (§ 32.1, 32.8, 32.9); the early synonymisation of Kinyamkolo and Niamkolo, which Moore appeared to have accepted (§ 32.2, 32.19.8); Moore’s explicit referral to LMS’s missionary station as Kinyamkolo, occasionally as Nyamkolo, which he regarded as a site, not district (§ 32.3, 32.6, 32.18, 32.19); the potential of the name Kinyamkolo to mean new, little, or greater Nyamkolo (§ 32.5, 32.7); a variety of people referring to the station as either Kinyamkolo or Niamkolo (§ 32.5, 32.6); Moore’s likely acquaintance with Swann, who presumably coined the name Kinyamkolo and introduced it to Moore (§ 32.5); the adventurer Glave’s explicit referral to the missionary station as Kinyamkolo, including his map indicating the exact location of Kinyamkolo, which Moore had access to (§ 32.6); Loveridge’s statement that Kinyamkolo is synonymous with Niamkolo and Nyamkolo (§ 32.7); the LMS missionaries’ exclusive referral to Kinyamkolo as their station or adjacent village, never as a district (§ 32.11); the fact that neither Kinyamkolo nor Fwambo was referred to as a district, but as a location and station (§ 32.12); Moore’s referral to craggy ledges, presumably pertaining to the partly rocky shoreline in the immediate vicinity of the missionary station (§ 32.4, 32.15, 32.22, 32.26); Moore’s view of the west coast, seemingly including the coast west of Chituta Bay or Kasenga Point (§ 32.16, 32.17); Moore’s awareness of the fishes being locally endemic and need of proper labelling (§ 32.19); the logic of making the greatest part of the total collection near one’s camp (§ 32.19); the colouration of Tropheus moorii, which conforms to the Tropheus found in the immediate vicinity of the missionary station, i.e., Kinyamkolo/Mpulungu (§ 32.20); the partly swampy surroundings of Moore’s Kinyamkolo, which conforms to the present-day Mpulungu (§ 32.21); Moore’s referral to Mbete as Mbity and the existence of rocks in shallow water near Mbete and Kasakalawe (§ 32.25); and the natural distribution of Cyprichromis leptosoma and Julidochromis ornatus only occurring along the east coast of Mbete Bay and farther north and northeast (§ 32.26).

Established facts should be questioned, but before claiming them to be inaccurate they need to be properly investigated, which is especially important for respected authors whose words carry such authority that their expressions of views are rarely questioned.


Part 1 – Part 2 – Part 3