DIPHTHERIA BACILLUS. 369 
matter with new and similar substance; after the stimuli 
following several repeated losses and replacements of 
this substance the cells produce it in excess. This sub- 
stance, whether originally in the normal cell or repro- 
duced there, and whether remaining in the cell or 
thrown out into the circulation, is antitoxin. 
The above summary merely gives an outline of 
some of the points in Ehrlich’s most interesting article. 
To become fully acquainted with the reason for his 
theories the article itself must be carefully read. 
Interest in both his theoretical reasoning and in his 
practical conclusions led us to subject both to a series 
of tests which have, I believe, added some interesting 
facts to those already published by Ehrlich as well as 
cast doubts on some of his conclusions. 
The results of these experiments of Atkinson and 
myself! were fully in accord with those published by 
Ehrlich as to the varying neutralizing value of a 
minimal fatal dose of ‘‘ toxin’’; they, however, also 
indicate roughly a general law in accordance with which 
these changes occur. 
The neutralizing value of a fatal dose of toxin is at 
its lowest in the culture fluid when the first consider- 
able amounts of toxin have been produced. After a 
short period, during which the quantity of toxin in the 
fluid is increasing, the neutralizing value of the fatal dose 
begins to increase, at first rapidly, then more slowly. 
While the culture is still in vigorous growth and new 
toxin is being produced, the neutralizing value of the 
fatal dose fluctuates somewhat, but with a generally 
upward tendency. After the cessation of toxin pro- 
1 Journal of Experimental Medicine, vol. iii., No. 4. 
24 
