Five years before the chickens at Weltevreden fell ill, the Japanese navy had been cured of beriberi. The cure was a change of ration, the man who made it believed it worked by adding protein, and that belief, which was wrong, was the handle by which Eijkman and the Dutch commission picked the whole result up and set it aside. This piece follows Takaki’s own account, given in three lectures at St Thomas’s Hospital in London and printed in The Lancet in May and June 1906, and Eijkman’s reading of it in 1898.

The disease in the fleet

Kanehiro Takaki entered the Japanese navy as a medical officer in 1872, was sent to England in 1875 to study at St Thomas’s, and came back in 1880 to a service in which, by his own tables, beriberi had run in 1878, 1879 and 1880 at about 349 cases for every 1,000 men a year, with about 8 deaths per 1,000. “What at once attracted my notice,” he said, “was the large number of beri-beri patients and the numerous deaths resulting from it.” In February 1882 he was made vice-director of the naval medical bureau, and that year three warships sent to Korea in a crisis “stayed there only 40 days but owing to shortness of hands caused by the prevalence of beri-beri among sailors the officers felt quite unfit for battle.” The disease was worst on long cruises, among ratings rather than officers, and in the training ships.

The training ship Riujo left Shinagawa on 19 December 1882 and returned on 15 October 1883, “after a voyage of 271 days to New Zealand, South America, and Hawaii.” Takaki’s lectures do not print the ship’s figures; the table he refers to was left out of The Lancet and was supplied in the Japanese translations of 1985. The James Lind Library’s account of 2013 gives them: 376 men, 169 cases of beriberi, 25 deaths. The disease, Takaki noted, “disappeared completely as soon as she arrived at Hawaii and was supplied with fresh articles of food.”

The experiment

On 5 October 1883 Takaki became director of the medical bureau, and at once asked the Minister of the Navy for a committee “on the cause of the occurrence of the almost unprecedented number of cases of beri-beri on board the training ship Riujo.” It sat from 12 November 1883, five times a week, seventy-nine times in all, and took 10,862 questions and 10,400 answers on patients, clothing, bedding, food, drink, housing, labour, rest, voyage, anchorage, weather and climate. Before it had reported, Takaki acted. He had been told that the training ship Tsukuba would sail shortly, “and at once arranged to utilise this opportunity of testing the new system of diet by making her go the same route as that of the Riujo.” On 26 November he demanded that the new diet be applied immediately; on 29 November the minister ordered the ration changed throughout the navy. The new scale cut rice and added meat, bread, barley, beans, condensed milk and vegetables. There was opposition “from various points,” and the cost was the last obstacle; with the minister’s knowledge Takaki went to Ito Hirobumi and to the Minister of Finance, Matsukata, and got a special allowance of 60,000 yen from the Treasury. A committee of a captain, a lieutenant, a surgeon and a paymaster was put aboard to watch.

The Tsukuba sailed on 2 February 1884 and returned to Shinagawa on 16 November. “The result obtained was good,” Takaki said, and “when the good report (‘no beri-beri’) of the experimental voyage of the Tsukuba became known the principal men in the navy for the first time began to support me in my fixed purpose. They said that they had always opposed me in their hearts and only obeyed the new regulations because they were ordered by the Minister, but they would give in now after such powerful practical proofs.” The ship’s figures, again from the James Lind Library: 333 men, 14 cases, no deaths; and Takaki’s own finding was that the fourteen had not eaten the new ration. In January 1885, going through the returns for 1884, he saw the navy’s general sickness nearly halved and beriberi “considerably decreased”; by his tables it fell from hundreds per 1,000 to a handful and stayed there.

The reason

Takaki’s theory was arithmetic. He had compared the diet of the men who got beriberi with the diet of those who did not, and found that the ratio of nitrogen to carbon in the sailors’ food, which was mostly rice, was far wider than in the officers’. “The causes of this disease,” he told London, “are due to the loss of equilibrium in the proportion of nutritive elements and also to the deficiency of a certain element—that is, the composition of food is not correct.” The deficient element was protein. That the protein came in with bread, barley and beans, all of which carry the thing the polished rice lacked, was invisible to him, as it was to everyone for another twenty years. He also knew the practical limit of his cure. Eijkman, who had read him, quoted it in Stockholm: “By last year’s experience we have found that most of the men dislike meat as well as bread and we do not know what we shall do next.”

Eijkman’s reading

The Dutch commission of 1886 knew of Takaki, and of van Leent, the Dutch naval surgeon who had argued in 1879 that the native sailors’ rice diet, poor in protein and fat, predisposed to beriberi, and had fed them European food. The commission rejected the idea on principle: malnutrition could not be supposed to destroy nerves. Eijkman, who had the chickens, could do better than principle, and did. In his paper of 1898 on beriberi and food, summarised in the Indische Gids, he set the Japanese result against the records. The decisive and lasting improvement in the Japanese navy’s sickness “had already begun in 1883, without change of diet, and in the following year the rice ration was even increased.” In 1885, “the first nearly beri-beri-free year,” the ration was still 608 grams of rice a man a day, “a still respectable quantity.” The men bought what they liked with the allowance the new regulations gave them, “and so rice too, to which they are very attached.” In the Japanese army, according to Bälz, the fall in beriberi had begun not when the richer ration came in but some years later, when the recruits’ date of entry was moved from summer to winter. And at home the Dutch navy’s European sailors, fed at least as well as the Japanese and on less rice, had had their highest beriberi rates after their ration was improved in 1891, while the native sailors’ rates, whose ration had been changed again and again, had risen and fallen in step with theirs. “In fact,” the review concluded for him, “we have thus already followed the Japanese example, but without result.” Whether the Japanese cure could be attributed to the change of diet he thought “very doubtful, indeed, all circumstances considered, almost unthinkable.”

Every objection was a real feature of the records, and all of them are answered by one question Eijkman did not yet think to ask of a navy, though he had been asking it of chickens for nine years: not how much rice, but what rice, and what else. The Japanese ration of 1884 still had plenty of rice in it; it also had barley and beans. The Dutch European sailors’ improved ration of 1891 had more meat and no silver skin. The chickens had already shown that meat delayed the disease and did not prevent it, and that a kernel without its coat made the birds ill however much of it they ate.

That is the shape of the thing. Takaki had done in 1884, on 333 men with a committee aboard, what the Stadsverband trial of 1896 and the prison surveys of 1896 and 1897 would do again: a comparison that settled what to do. He explained it with protein, and the explanation did not survive. Eijkman, who could show the explanation was wrong, took the failure of the reason as grounds to doubt the result, and found in the Dutch navy’s own records a diet that had been enriched by exactly the wrong thing. He was reading a right answer through its wrong reason, as his own result was being read in the same years through his. The Japanese navy, which had not waited for either, stayed cured.

Sources

• Baron Takaki, “On the Preservation of Health amongst the Personnel of the Japanese Navy and Army,” Lectures I, II and III, The Lancet, 19 May, 26 May and 2 June 1906. The rates of 1878 to 1880, his appointment of February 1882, the Korean crisis, the Riujo’s dates, the committee of 12 November 1883 and its seventy-nine meetings, the order of 29 November, the Tsukuba’s dates, the 60,000 yen, “the good report (‘no beri-beri’),” the opponents who “had always opposed me in their hearts,” and “the composition of food is not correct.” 19 May, 26 May, 2 June
• Yoshifumi Sugiyama and Akihiro Seita, “Kanehiro Takaki and the control of beriberi in the Japanese Navy,” James Lind Library, 2013, also in Journal of the Royal Society of Medicine 106 (2013): the crews, cases and deaths of the two ships, the fourteen who had not eaten the new ration, and the missing table. jameslindlibrary.org
• Christiaan Eijkman, “Antineuritic Vitamin and Beriberi,” Nobel Lecture, 1929: van Leent, the commission’s rejection, and Takaki’s “most of the men dislike meat as well as bread.” nobelprize.org
• “De stand van het beri-beri-vraagstuk,” Bataviaasch Nieuwsblad, 10 September 1898, from the Indische Gids, summarising C. Eijkman, “Beri-beri en voeding,” Nederlandsch Tijdschrift voor Geneeskunde, 1898: the Dutch navy’s rations of 1891, Japan in 1883 to 1885 and the 608 grams, Bälz on the army’s recruits, and “zeer twijfelachtig, ja, alle omstandigheden in aanmerking genomen, haast ondenkbaar.” Translated by the author of this piece. Delpher
• C. Eijkman, “Nogmaals beri-beri en voeding,” Geneeskundig Tijdschrift voor Nederlandsch-Indië 38 (1898), pages 277 to 284, on the Dutch navy’s European and native sailors. archive.org