IV 



12 C the product k y/m~ is 5.4 (mean value for cane sugar 

 5.3, mannite 5.4, citric acid 5.7, tartari cacid 5.4, ethyl alco- 

 hol 5.0 and ammonia 5.8). 



We have therefore compared the velocity of diffusion 

 of diphtheria toxin and some other bodies with a sub- 

 stance of a well determined molecular weight and from 

 these facts we tried to get an approximate value of the 

 constant of diffusion of diphtheria toxin. 



As we supposed this to be very small we considered 

 a long time of diffusion to be necessary; as besides 

 the specific weight of the solutions does not perceptibly 

 vary with the percentage of toxin which they contain, 

 it was impossible to employ the usual method: to let 

 them diffuse in their ordinary vehicle of solution, the 

 water. We therefore have permitted them to diffuse in 

 solutions of gelatine. 



A 5 % solution of french gelatine was prepared, 

 neutralised with soda and sterilised; of this solution 

 15 cc. were poured into sterilised test lubes, whereby 

 a gelatine pillar about 10 cc. high was obtained. The 

 test tubes were put into an ice safe in a vertical posi- 

 tion and allowed to get stiff. Then 2 cc. of an aqueous 

 solution of the body to be examined was poured by 

 means of a pipette over the gelatine pillar and the tube 

 was closed with a sterilised cork. It was thereafter al- 

 lowed to stand in the ice safe for a certain time, during 

 which the diffusion was going on. After this time the lest 

 tube was taken up, the fluid covering the gelatine re- 

 moved, and the tube over the gelatine quickly washed 

 with distilled water. The tube was cautiously crushed 

 and the surrounding glass thus removed from the gela- 

 tine pillar. This was then cut to pieces of a certain 

 length, whose weight was found. The toxicily of the 

 lluid above and of the pieces was found by means of 



- 4 - 



