Influence of Salts 



103 



problem at my suggestion. Their experiments agree more closely with 

 those of Eisenberg and Rostoski. They report thereon as follows : — 

 "We have quantitatively estimated the influence of salt in the following 

 way'. Tubes containing 1 in 21 dilutions of human serum with 

 gradually increasing percentages of salt were arranged in a rack, and to 

 each "1 c.c. of anti-human serum was added. We found that the 

 precipita in the tubes containing the most salt were more flocculent, 

 and owing to the increased specific gravity of the medium took longer 

 to settle (see Plate, fig. 2). 



Results of measurements showed a slight decrease to 7 °/o and later 

 an increase, probably due to the fact that the more flocculent precipitura, 

 though really less in amount, occupies a greater volume. 



Results of increasing quantities of salt in human seruin dilutions. 



Experiments with sheep and anti-sheep sera, which formed more 

 compact precipita, show the diminution in volume plainly. In this 

 experiment the tubes were all centrifugalized for the same length of 

 time in order to diminish the error due to the increasing specific gravity 

 of the solutions. 



Similar experiments to above with sheep serum. 



Salt 



17o 



2 „ 



4 ,, 



10 „ 



Saturated) 



solution \ 



. Precipitum 

 c.c. 



•0199 



•0203 



•0201 



•0140 



•0122 



Percentage of 

 precipitum 



100 7o 



102 „ 



101 „ 

 70-3 „ 



61-3 „ 



Qualitative estimations showed that when the quantity of salt was 

 increased above 5 7o the antiserum did not sink to the bottom, and 



1 The quantitative method of testing is described in Section VII. which follows. 



