ETIOLOGY OF BERIBERI. 253 



slightly greater than the quantity of potassium in five grams of polish- 

 ings, which latter quantity is known to be sufficient to maintain fowls 

 in health. In order to facilitate administration 7 grams of each of 

 these salts except the mag-nesium phosphate were dissolved in 100 cubic 

 centimeters of distilled water. The magnesium phosphate being in- 

 soluble was simply suspended in water in the proportion 7 grams to 

 100 cubic centimeters. One cubic centimeter of these solutions and 

 the suspension was administered daily to each fowl with a pipette. 

 Therefore there can be no doubt as to what each fowl actually received. 

 We might state at this point that this method of administration has been 

 employed in all of our experiments. 



The result of this series of experiments is briefly as follows : 



Group A (receiving potassium phosphate) : One fowl developed neuritis 

 in 22 days and one in 24 days. 



Group B (receiving potassium citrate) : One fowl developed neuritis 

 in 22 days and one in 24 days. 



Group C (receiving potassium carbonate) : One fowl developed neuritis 

 in 21 days and one in 23 days. 



Group D (receiving magnesium phosphate) : Two fowls developed 

 neuritis on the 24th day. 



The experiment was discontinued on the 28th day since it was con- 

 sidered amply demonstrated that none of these salts conferred any 

 protection. 



We may draw several conclusions from this experiment. 



1. Lack of potassium as a cause of polyneuritis can probably he 

 excluded, because we have tried four salts of this element, one of 

 them known to be present in rice polishings and one of them being 

 the salt of an organic acid, and all of these salts of j)otassium signally 

 failed to confer protection. 



2. Magnesium phosphate, a salt present in considerable quantity in 

 the filtrate of Fraser and Stanton, is also shown to be of no value. 

 This probably excludes the element magnesium from further considera- 

 tion. 



3. By a rough calculation the phosphorus contained in the magnesium 

 and potassium phosphate of the filtrate of Fraser and Stanton which we 

 prepared is equivalent to about one-fifth of the 15 per cent of total 

 phosj^horus remaining according to their analyses. Since we have shown 

 that these salts are unimpoi-tant we can state at this point that 88 

 per cent of the total phosphorus of rice polishings is negligible. We 

 have not given all the figures for the sake of brevity and because our 

 results by the method which follows are so conclusive as to the unimpor- 

 tance of phosphorus compounds. 



Fraser and Stanton! 3) have shown that the neuritis-preventing princi- 

 ple of rice is soluble in hot alcohol and Hulshoff -Pol ( 4 ) proved that a 

 decoction of katjang idjo {Pliaseolus radiatu^) has the same curative 



