21^ Professor Svante Arrhenius [June 9, 



In this manner, Madsen and Teruchi found that vibriolysin is 

 divided between red blood-corpuscles and the fluid, so that the con- 

 centration of vibriolysin in the cells is about 2;')0 times greater than 

 in the fluid. This was proved by treating different quantities of 

 cells from 0"05 up to 0*6 c.c. with 9 "OS resp. 9*4 c.c. of salt- 

 solution, charged with a certain quantity of the poison. In the first 

 case (0-05 c.c), the cells took away 56 per cent, of the poison ; when 

 the cells were eight times as many, 90 percent. The two determina- 

 tions give ratios of partition of 254:1 and 227:1, which may be 

 regarded as equal, if we consider the great errors of observation. 



I observed the same property in a more indirect way, namely, by 

 means of the ol)served degree of hemolysis. If this is constant, the 

 red blood-corpuscles are in the same state, independently of their 

 number, and the surrounding solution has the same concentration of 

 the poison. If, then, I investigate different quantities of red l)lood- 

 corpuscles in the same quantity of isotonic salt-solution, and add just 

 so much poison that a certain fraction — say 40 per cent. — of the 

 blood-corpuscles is hemolyzed, then the same quantity of poison is in 

 the fluid, and the l)lood-corpuscles have taken up a quantity of poison 

 proportional to their number. By changing this number, it is possible 

 to determine how great a part of the poison remains in the fluid, and 

 how much is taken up by each c.c. of the suspended blood-corpuscles. 

 Of course, it is necessary always to use the same temperature— I had 

 a bath of 87° C. — and time of reaction — I used two hours — in the 

 same series of experiments. In this manner I found that for saponin 

 the concentration in the blood-corpuscles was 120 times greater than 

 in the solution. For ammonia, caustic soda, and acetic acid, I found 

 values varying between about 600 and 900 times. I have found 

 similar high values for silver salts and mercuric chloride, which are 

 also hemolysing substances. 



This strong absorption of the l)lood-poisons in blood-corpuscles 

 indicates that they are probably chemically bound to some substances 

 in the cells. The absorption proceeds, according to Madsen's and 

 Teruchi's experiments, in agreement with common physico-chemical 

 laws, i.e. in proportion to the excess of the concentration in the fluid 

 over that concentration which would be in equilibrium with the 

 poison absorbed in the cell at the time observed. 



Those antibodies, which have been examined hitherto, behave in a 

 somewhat different manner. Eisenberg and Volk allowed certain kinds 

 of bacilli, viz. typhoid-bacilli or cholera vibrions, to be emulsified for 

 a certain time with solutions of different quantities of the correspond- 

 ing agglutinins, dissolved in a given quantity of 0*9 per cent, solution 

 of sodium chloride. The bacilUs slowly subsided, and the supernatant 

 fluid was decanted and its content of agglutinin determined. It was 

 then found tliat the concentration in the bacilli increased more slowly 

 than the concentration in the fluid, and proportional to the power two- 

 thirds of this concentration. This circumstance indicates that the 



