STUDIES OF CHOLERA. 109 



as usual, but none of the guinea pigs died within twenty-four hours, 

 although the doses inoculated varied from 0.001 to 0.00001 cubic centi- 

 meter. In other words, the virulence had suddenly fallen at least to 

 one-fiftieth of its former strength, and how much more is not known. 

 It is important to- determine the nature of the biological processes asso- 

 ciated with this sudden abatement of virulence, but the only clue fur- 

 nished by this series is that 1 cubic centimeter of a 1 to 1,000,000 dilution 

 of the exudate from the next to the last guinea pig of the series, when' 

 plated in agar, contained colonies too numerous to count, so that at 

 least it is clear that the bacteria had not lost their viability. We may 

 account for the sudden loss of virulence either by supposing that the 

 power of toxin formation suddenly disappeared, or that the ability of 

 the bacteria to resist bactericidal action suddenly failed them. It seems 

 clear that there was a sufficient number of surviving bacteria carried 

 over from the next to last guinea pig ; this was shown by the results 

 of plating. The results of this series of inoculations are shown in Series 

 I, and the sudden rise and fall in virulence is charted in fig. 1. 



The strain of cholera which has just been described was obtained from 

 an autopsy during the height of a small epidemic. Several other at- 

 tempts were made to repeat this result subsequently, when the epidemic 

 was on the wane, both with strains of cholera inoculated into guinea 

 pigs directly from the intestine and with laboratory stock cultures. The 

 results are shown in Series II, III, IV, V, and VI and in figs. 2-5. 

 It is seen that no strain was obtained, the virulence of which could be 

 increased so successfully and regularly as that of the first. 



Turning to Series II, we find the results are less regular, although 

 here again there was a gradual rise in virulence up to a fatal dose of 

 0.0001 cubic centimeter. The irregularities are difficult to explain. For 

 instance, we find on November 11 that 0.0002 cubic centimeter is fatal, 

 while 0.001 is not so. Again, on November 13 neither 0.01 nor 0.0002 

 cubic centimeter killed the. animal, while 0.0001 and 0.001 did so. It is 

 possible that this difference can be explained upon a physical basis, 

 for the exudate was tenacious and possibly did not spread evenly through- 

 out the diluting salt solution. 



The virulence was irregular in Series III also. An observation made 

 in this series throws some light upon the mechanism of the fall of an 

 epidemic. In the fifteenth inoculation made on November 28, the only 

 animal which died within twenty-four hours received 0.01 cubic centi- 

 meter of exudate. This animal did not die until 10.15 on the morning of 

 November 29. The autopsy was performed at once and 0.25 cubic centi- 

 meter of tenacious, yellowish, cloudy exudate was recovered from the 

 peritoneal cavity. As the table shows, the animals inoculated with this 

 exudate did not die within twenty-four hours. At the same time 1 

 cubic centimeter of a 1 to 80 dilution of this exudate was inoculated 



