Evidence soon accumulated showing that rice was not the only food due to which beriberi might develop. Vedder and Clark24 and Ohler 25 showed that typical polyneuritis developed in fowls fed on a diet of white bread or bread and rice.26 Little 27 described outbreaks of the disease in Labrador and Newfoundland which he attributed to the diet of white bread, molasses, and fish upon which the inhabitants lived. While the only flour available was the product of the rough mills of the country, beriberi was unknown, but when the bolted flour of the modern mills became common, outbreaks of beriberi followed. It is told that in 1910 a ship ran ashore, laden with a cargo of whole meal wheat flour, and in order to lighten her a considerable portion of her load was removed and was subsequently consumed by the people in the adjacent districts. The result was that no case of beriberi was reported in that region for a period of one year following this event.

24 Phil. J. Sci. 7, 423, 1912.

25 Ohler, J. Med. Res. 31, 239, 1914.

26 See also Wellman and Baas, "Polyneuritis gallfnanim caused by Different Foodstuffs," Am. J. Trop. Dis. and Prev. Med. 1913, i. 129.

27 Little, J. Am. Med Ass. 58, 2029, 1912; 63, 1287, 1914.

Parker 28 has reported an outbreak among the prisoners in a jail at Elisabeth, N. J., where white bread had been used as the main staple of the diet.

More recently cases of beriberi have been reported among the troops at Kut-el-Amara during the siege from December 4th, 1915, to April 29th, 1916. During a large part of the time the British garrison were restricted to white flour or biscuits and tinned meat or horseflesh, and many fell a victim to the disease. The Indian troops, on the other hand, ate either barley or "atta " flour (containing the aleurone layer of the wheat grain), which served to protect them against beriberi, although owing to their refusal to eat meat they suffered from scurvy from which the British were free.29

Two forms of beriberi are described, the "wet" beriberi in which there is much swelling of the tissues, and the " dry " beriberi in which all the muscles of the body are atrophied and the patient becomes entirely helpless. In both forms the earliest symptoms are loss of appetite, fatigue, depression, and slight fever. Vomiting and diarrhoea are frequently reported, and Barbe 30 found a severe duodenitis in a number of his fatal cases. Riddell, Smith and Igaravidez,31 reporting on 60 cases of beriberi at an army base hospital at San Juan state, however, that gastric disturbances were uncommon, only seven cases of nausea being reported and these at the onset and for a few days only. In their experience the symptoms began in almost every case with a numbness over the legs, usually over the shins first and gradually spreading over the calves and upward to the abdomen.

28 Parker, Pub. Health Repts. 29, 339, 1914.

29 For further details of epidemics reported from 1891 to 1917 see De Mello, Loundro and Rabello, Anala. Scient. da Facul. de Med. do Porto, 4, 6, 1917.

30 Barbe, Arch. Med. Pharm. Nav. June, 1918, quoted by Jelliffe, J. Nenr. Ment. Dis. 49, 522.

31 Riddell, Smith, and Igaravidez, J. Am. Med. Aas. Feb. 22, 1919.

Pain was invariably observed in the calf muscles, these being usually the first to be affected. Only eleven patients showed any fever, and in these cases it was slight and lasted only a few days.

In the later stages there is great exhaustion, difficulty in breathing, multiple neuritis, paralytic symptoms, frequently followed by death.

The symptoms observed in fowls and pigeons are very similar. Within three weeks birds on an exclusive diet of polished rice usually show signs of weakness, followed very quickly by paralysis. Their necks are retracted, and they roll helplessly round their cages. Without treatment death follows very shortly, but with proper treatment their recovery is dramatic in its suddenness. Microscopic investigation of the tissues of a polyneuritic bird shows extensive peripheral degeneration of the nerve cells, even in the case of fowls which are killed before the symptoms of neuritis have manifested themselves." McCarrison33 finds that the oedema (swelling) characteristic of "wet" beriberi appears in the avian polyneuritis also, and is associated with pronounced increase in the size of the adrenal glands and greatly increased production of adrenalin, although the amount of adrenalin per gram of gland is actually slightly less in wet beriberi than in dry beriberi or in normal health. In 100 per cent of the cases of wet beriberi the quantity of adrenalin greatly exceeded that found in health, and in 83 per cent of these cases exceeded that found in dry beriberi.

32 Vedder and Clark, Phil. J. Sci. 7B, 423.

33 McCarrison, Proc. Roy. Soc. 91b, 103, 1020.