Ill 



amounts, may he considered as the same, allowed for 

 errors of experiment. The degree of dissociation of these 

 salts is also very nearly the same at such high dilutions as 

 are here used. Though it would seem as if the sulphate 

 has a weaker action than chloride in the concentration 

 0,1 n, yet, in the case of 0,5 n solutions, rather the op- 

 posite takes place, and the difference must then, without 

 douht, he attributed to errors of experiment. Through these 

 series we get a good idea of the possible extent of these 

 errors of experiment and how necessary it is not to con- 

 sider single figures by themselves, but to take the whole 

 series in consideration. Therefore, in the following, we 

 use the mean values for sulphate and chloride. The 

 action of very diluted solutions of salts of ammonia. 

 0,0016 n, (add. of 0,4 cc. 0,04 n salt solution) is not 

 stronger (25%) than the corresponding solutions of the 

 salts of the strong alkalies, the series with 0,04 n 

 N 2 H 8 S0 4 /2 indicates this. When we use one part of 0,1 n 

 salt of ammonia in 6 parts of blood mixture (2 cc. to 10 

 cc. blood emulsion) viz. a 0,016 n solution the hicmo- 

 lytic action is very much lessened; it is then no greater 

 than that of a 4 times smaller amount of ammonia. If 

 we add half this amount of ammonium salt (0,008 n 

 = addition of 1 cc. 0,1 n salt solution), the action is 

 decreased to about 1 / s , and when we add y 4 of this 

 amount of salt, (0,004 n = = 0,5 cc. 0,1 n salt sol.), the 

 action is decreased to about 2 / 5 . By an addition of 0.06 

 n salt (1,5 cc. 0,5 n salt solution), the action is de- 

 creased to V 7 . 



We can sum up these values as follows: 

 S signifies the added salt of ammonia. 



S. 0,0016 0,004 0,008 0,01(5 0,06 



reduction 0.75 0.4 0.3 0.25 0.14 



40 



