56 FRASEK AND STANTON. 



associated with outbreaks of beriberi, there could be extracted by means 

 of various solvents any substance or substances recognizable by chemical 

 methods as poisonous in character. These researches failed of their 

 object, though it is admitted that the accuracy of the poison hypothesis 

 was not thereby disproved. 



Certain results which emerged from chemical analysis and histological 

 examination of the rices turned attention to the possibility of an explana- 

 tion of the course of events on an hypothesis of a defect of nutrition. 

 That this explanation was inadequate, if dietary constituents as estimated 

 by the ordinary analytical methods were alone considered, had been 

 shown in the preliminary investigation. 



By a series of experiments on domestic fowls, the details of which will 

 be supplied in a later publication, it was shown that these animals when 

 fed on various kinds of rice were sensitive to differences between them. 

 The fowls were confined in separate compartments and were in all respects 

 under identical conditions. The manner of arrangement of the cages is 

 shown in Plate I, fig. 1. 



By further and repeated experiments with rices known to have been 

 associated with outbreaks of beriberi, and with controls under identical 

 conditions fed on parboiled rice, it was established that a certain reaction 

 in fowls might be taken as an indicator of the beriberi-producing power 

 of a given rice when forming the staple of the diet in man. Whether 

 the disease produced in fowls be accepted or not as analogous to beriberi 

 in man, the validity of the arguments here advanced remains unimpaired. 



Bices were available that were known to have been associated with 

 outbreaks of beriberi, samples having been taken daily during the con- 

 tinuance of the preliminary inquiry in 1907 ; also through the courtesy 

 of Dr. J. D. Gimblette and Dr. G. D. Freer we were enabled to procure 

 white rice which was being consumed prior to an outbreak of beriberi 

 among Malays at the Kuala Lumpur police depot, which outbreak ceased 

 on changing the rice supplied to the parboiled variety. It was shown 

 that these rices when fed to fowls constantly produced a certain disease 

 in a large proportion of them, while parboiled rice as constantly failed 

 to produce this result in groups under identical conditions. This disease 

 is characterized by paralysis of the legs (Plate I, fig. 2, and Plate II, 

 fig. 3), followed by paralysis of the wings (Plate II, fig. 4) in the more 

 severe cases. In cases showing a moderate degree of paralysis the gait 

 resembles very closely that seen in beriberi. The nerves of fowls suffer- 

 ing from this disease show typical Wallerian degeneration (Plate III, 

 fig. 5). 



It is our belief that this disease, polyneuritis gallinarum, is truly 

 analogous to beriberi in man, similar in its etiology, in its clinical man- 

 ifestations, and we have shown them to be identical in their pathologic 



