— 192 — 



when they have not already suffered more than once from 

 this disease. If the experiment took place with healthy 

 birds, it might last much too long; and besides, the trouble- 

 some factor of individual liability to this disease could not 

 sufficiently be eliminated. Neither is it desirable to take 

 birds which are too ill, because these often, of their own accord, 

 take too little food, and for this reason also need more time 

 to recover than do animals in the first stadia of the disease. 



These birds are now given an average sample of the rice 

 which is to be investigated, and this in such a manner that 

 not more than 5 a 6 gram is provided for each bird, as 

 this is sufficient for a day's ration. When a larger quantity 

 is given, those grains are picked out upon which the most 

 silver-skin has remained, in consequence of which the test 

 might lead to wrong conclusions. For the same reason it 

 is advisable to gently rub the rice which is to be given, to 

 free it from any polishings (dedek) which might have 

 remained clinging to it. 



The demeanour of the test-animals in consequence of this 

 diet depends entirely upon the difference which the rice 

 given shows with regard to what might be considered as 

 their minimum-normal diet. According to the degree of this 

 difference recovery will take place sooner or later, or else 

 the symptoms of disease will be aggravated. 



As to forming a judgement in consequence of the course 

 of this experiment, it has appeared more and more clearly, 

 that, when all the sick test-animals recover in one or at 

 the most two days, this must be taken as a proof that the 

 kind of rice in question is of sufficiently good quality to be 

 used as food for man. 



Should recovery not take place within this time, the 

 condition remain stationary, or the symptoms become aggra- 

 vated, then the rice must be rejected, because it does not 

 afford sufficient protection against beri-beri. 



As it is possible that one of the three test-animals may 

 act differently from the two others, it is advisable to subject 



