The Tsavo man-eaters: 28 counted in 1907, 135 by 1925

The number most people know comes from a pamphlet sold in Chicago. In 1925 the Field Museum of Natural History put two stuffed lions on show and printed a 40-page leaflet about them, Zoology Leaflet 7, written by the man who had shot them.⁸³ Its first sentence reads:

When the visitor to the Field Museum pauses before the life-like forms of the Tsavo man-eaters, it will be hard for him to realize that these two ferocious brutes killed and devoured, under the most appalling circumstances, one hundred and thirty-five Indian and African artisans and laborers employed in the construction of the Uganda Railway.³

The leaflet doesn’t say where 135 comes from. Eighteen years earlier, in The Man-Eaters of Tsavo, the same author had given a different count, and only half of it was a number:⁴

Well had the two man-eaters earned all this fame; they had devoured between them no less than twenty-eight Indian coolies, in addition to scores of unfortunate African natives of whom no official record was kept.⁴

The opening page of the leaflet is transcribed on its own page, with notes on what changed since 1907.³⁹

Patterson’s papers, 1900 to 1925

Every count of the Tsavo dead from the time that we have read goes back to one man. John Henry Patterson, an engineer, was in charge of the bridge and the works around Tsavo, and it was his rifle that killed both lions, on 9 and 29 December 1898.¹ His first account appeared in The Field on 17 and 24 February 1900. We have not read those two issues.¹³ What we have is a summary in The Spectator of 3 March 1900, which gives the earliest count we have found:⁴

In this time they killed and ate twenty-eight Indians, and it is believed at least twice this number of natives, Swahilis and the like; besides wounding and attacking others.⁴

Twenty-eight and at least twice as many again makes at least 84, our sum. The Spectator’s paragraph is on its own page.⁴

Patterson’s book followed in October 1907, with a foreword by the hunter Frederick Selous, who had first read the story in The Field.¹³ It keeps the 28 Indians and turns the Africans into “scores”, with no official record kept.⁴ On 29 November 1924 Patterson lectured at the Field Museum; he told its president, Stanley Field, that he still had the skins, and Field bought them for the museum.⁸ The skins had been kept for years as trophy rugs, never prepared for display.⁸ The museum’s taxidermist, Julius Friesser, mounted them, and the 1925 leaflet was printed for the new exhibit.⁸

The first victim Patterson names is one of his jemadars, a Sikh called Ungan Singh, seized from the tent he shared with other workmen about three weeks after Patterson reached Tsavo.¹⁴ Most of the others have no names in any of these texts.

135, the Chicago figure

The figure of 135 has one author, writing in 1925, 27 years after the attacks.³ Kerbis Peterhans and Gnoske, who reviewed the Tsavo case for the Journal of East African Natural History in 2001, trace it to the 1925 leaflet, which they read as adding 107 Africans to the 28 Indians.³ The pages of the leaflet we read give no breakdown, only the total.

The museum took the number over at once. Its annual report for 1925 says the two lions “killed and, in most cases, devoured more than 130 human beings”.¹⁵ Its pages today still repeat 135 as Patterson’s figure, and add that later research brought it down to 35.¹⁵ DeSantis and Patterson (2017) and Yeakel and colleagues (2009) cite 135 to the 1925 leaflet.³ No scholar we read defends it.

Twenty-eight Indians, uncounted Africans

The 28 is the only part of any count that stays the same from 1900 to 1907.⁴ Kerbis Peterhans and Gnoske treat it as the documented minimum: between March and December 1898, at least 28 people were taken.¹⁶ They accept it because, they write, “the British maintained precise accounting details”, and they add three deaths from other writers, two railhead workers recorded by the engineer R. O. Preston before Patterson arrived and one man from a caravan in 1896, for a possible 31.¹⁶ Yeakel and colleagues call 28 “the Uganda Railway’s documented value”, but the reference they give for it is Patterson’s book.¹⁶

The African dead are the gap. Patterson’s own phrase, “scores … of whom no official record was kept”, admits there was no count.⁴ The 1900 summary’s “at least twice this number” is an estimate, which The Spectator offers as a belief, not a count.⁴ Nothing we read replaces either with a number.

About 35, from isotopes

In 2009 Justin Yeakel and seven colleagues, among them Bruce Patterson of the Field Museum (no relation to the colonel), measured carbon and nitrogen isotopes in the lions’ bone collagen and in the hair of their tail tufts. Bone records diet over years; the tail hair covers roughly the last three months of life.² To stand in for human prey they used the skulls of Taita people collected in 1929, set against Tsavo’s grazing and browsing animals.²

The first lion shot, FMNH 23970, got about 30 percent of its food from people in its final months; the second, FMNH 23969, about 13 percent.² From an intake of about 6 kilograms a day and about 20 kilograms eaten per person, they reckon 8.1 and 3.5 people in those three months, and 24.2 and 10.5 over the nine months of recorded attacks.² That makes 34.7 in all, with a possible range of 4 to 72.²

The model counts people eaten, not people killed, and it stretches three months of hair across the nine months of attacks.² Its median lands above the 28 and far below the 135.²¹⁶

Nine feet eight inches

How big the lions were rests on one man’s tape. In The Man-Eaters of Tsavo, Patterson gives the first lion, shot on 9 December, as nine feet eight inches “from tip of nose to tip of tail” and three feet nine inches high, and says it took eight men to carry it back to camp.¹¹⁷ The second, shot on 29 December, measured nine feet six inches and stood three feet eleven and a half inches high.¹⁷ He does not say how he measured, or whether “high” means at the shoulder.

The same book offers a comparison. Later, on the Athi Plains, Patterson shot a lion he calls “a young, but full-grown lion in fine condition” and measured it at nine feet eight and a half inches, half an inch longer than the first man-eater.¹⁸ By his own tape, then, the man-eaters were no longer than a lion he met in the ordinary course of railway work; the comparison is ours. What the museum stressed in 1925 was not their size but their type: “two large males of the short-haired maneless type”.¹⁹

The mounts in Chicago can’t settle the question. The skins were trophy rugs until 1925,⁸ and the museum’s annual report says they “had been preserved for a number of years and were not prepared originally with a view to museum exhibition”, which gave its taxidermist “unusual difficulty”.²⁰

A broken canine

The first lion’s skull is the reason the teeth get so much attention. Its lower right canine was broken off with the pulp exposed, the root had an abscess, lower incisors were missing, and the jaws no longer met properly.²¹ Kerbis Peterhans and Gnoske think the injury happened early in its life and may have kept it from subduing its normal prey.²¹⁵ Patterson had his own explanation, which appears in the 1925 leaflet and not in the 1907 book: on examining the head he found that “a .303 bullet had smashed out one of his tusks, for the track of the bullet was left in the tooth stump”, and he “must have given him a bad toothache” the night the lion attacked him and Dr Brock in a freight car.²² Yeakel and colleagues found that this lion was also the one that ate the most people, which they say “agrees well with the infirmity hypothesis”.⁵²

The second lion’s case is weaker. It had a broken upper carnassial (Kerbis Peterhans and Gnoske also list a broken lower canine), an injury DeSantis and Patterson call fairly common in lions, and they suggest it “may have simply shared meals” with the first.²¹⁵ Patterson, Neiburger and Kasiki (2003), who looked at lion skulls across museum collections, found tooth breakage normal and rarely serious; they allow the infirmity theory for Tsavo but explain most attacks on people and livestock by prey depletion.⁵

A 2017 study of microscopic wear on the man-eaters’ molars adds something the skulls alone can’t. The wear looks like that of ordinary wild lions and close to that of zoo lions fed soft meat.²³ Larisa DeSantis and Bruce Patterson conclude that the lions were not crunching and eating whole carcasses in their last weeks, and that Patterson’s accounts of bone-crunching outside the camps are not supported.²³ The Field Museum now sums up the current view in one line: people were easier to catch and chew.⁵

Cattle plague

Kerbis Peterhans and Gnoske also look past the teeth, at the land. Rinderpest, the cattle plague, reached Somalia in 1889 and had run through Kenya by the early 1890s, with another outbreak in 1897.⁶ Buffalo and cattle, the main prey of Tsavo’s lions, were still scarce in the late 1890s.⁶ Yeakel and colleagues add the Mwakisenge drought of 1897 to 1900 and read their isotope results as support for the idea that scarce prey pushes individual animals toward unusual food.⁶

Bodies on the caravan road

The railway followed a caravan route that had been used for decades. Kerbis Peterhans and Gnoske collect travellers’ accounts of porters and slaves left dead or dying along such routes, and of local burial customs. The Wataita dug up their dead and put the skulls in rock shelters, and the Wakamba were said to leave some bodies in the bush.⁷ The Taita skulls that Yeakel’s team used came from exactly such a practice: after about a year in the ground, skulls were exhumed and placed in ancestral caves.⁷

The argument is that lions at Tsavo had long had access to human bodies. The same review lists lions taking people near Tsavo before 1898, as far back as 1886 by the authors’ dating, and after: an engineer named O’Hara near Voi in March 1899, and the police superintendent C. H. Ryall, dragged from a railway carriage at Kima in June 1900.²⁴ Patterson tells both of those stories himself.²⁴ On the review’s list, the pair of 1898 were neither the first nor the last lions to take people near Tsavo.²⁴

Hair in the teeth

In the early 1990s Thomas Gnoske, a Field Museum collections manager, noticed thousands of hairs packed into the lions’ broken teeth.²⁵ A first look, reported in 2001, found hair from zebra, porcupine, warthog, impala, eland and oryx, and none from people.²⁶ In 2024 Alida de Flamingh and colleagues read the DNA. They identified giraffe, human, oryx, waterbuck, wildebeest and zebra, as well as the lions’ own hair; the giraffe hair came from at least two Masai giraffes.²⁵ Wildebeest are the odd one out: the nearest population in the late 1890s was about 50 miles away.²⁵

Human hair in the teeth confirms what nobody doubted. The rest of the list matters more. It fits the first lion shot over a dead donkey, the second taking goats, and what the isotopes say: people were part of the lions’ food, not all of it.²⁶²⁷

Checking the museum’s version

Quotations that changed

The number was not the only thing that changed between the book and the leaflet. In 1907 Patterson quoted Lord Salisbury, the prime minister, telling the House of Lords that “the whole of the works were put a stop to for three weeks because a party of man-eating lions” had “a most unfortunate taste for our porters”.²⁸ In the 1925 leaflet the same quotation has no three weeks, a “pair” instead of a party, and “workmen” instead of porters.²⁸ The Spectator in March 1900 says Salisbury spoke of the lions in the Lords.²⁸ In the Lords debate on the Uganda Railway Bill of 21 June 1900, the Earl of Kimberley remarked that the planners had not reckoned with lions eating a number of the coolies, and Salisbury answered that the lions did not feed only on coolies and had recently taken a medical man. Neither version of Patterson’s quotation is in that debate.²⁸

Theodore Roosevelt’s praise moved the same way. Selous’s foreword of 1907 quotes a letter in which Roosevelt calls “the incident of the Uganda man-eating lions, described in those two articles you sent me” the most remarkable account on record.⁹ In 1925 the leaflet has Roosevelt saying “the story of the Man-Eaters of Tsavo is by far the most remarkable account of which we have any record”.⁹

After Chicago the story went to Hollywood. The Ghost and the Darkness, in 1996, was the third Hollywood film based on Patterson’s Tsavo, and DeSantis and Patterson credit it with making the two lions famous.²⁹ Kerbis Peterhans and Gnoske summed up the whole century in one line of their 2001 paper: “Over the years however, these figures have changed and the legend has grown.”¹⁰

Twenty-eight to 135

The Indian dead were counted, by Patterson, and his count is 28 in both of his accounts from before 1925 that we have read.⁴ No count of the African dead survives in anything we read.⁴ The isotope model, which cannot tell one group from the other, puts the people eaten at about 35 and allows anything from 4 to 72.² Patterson’s last figure, printed when the skins went on show in Chicago, is 135.³ Between 28 and 135 is as far as the documents go.¹⁶