XIV, 5 Johnston: The Cholera-carrier Problem 461 
vibrio was isolated; March 3 he was again positive; March 6 
and 8, respectively, a vibrio was found; March 15 he was nega- 
tive and remained so to December, 1917. 
Prisoner 8472 was positive September 28, 1914, and then 
had eight negative examinations at intervals of four days; he 
was found positive again July 13, 1916. 
No. 9638 was found positive January 17, 1915, followed by 
eight negative examinations at intervals of four days; he was 
again positive July 18, September 138, and December 1, 1916, 
and has not been positive again. 
Is it not reasonable to suppose that these men harbored the 
cholera vibrio for periods of one and two years? If this be 
true, why not for longer periods of time? 
No. 11040 was positive for the first time August 2, 1916; 
negative until September 22; again negative until October 19; 
negative again until November 25; negative, November 29; 
positive, December 3; negative, December 15 and 16; December 
20 a nonagglutinable vibrio was isolated; negative, December 
23 and 29; on January 8 and 11, 1917, respectively, a non- 
agglutinable vibrio; January 18, positive; January 20, 24, 25, 
and 27, respectively, a nonagglutinable vibrio; January 30, 
negative; February 3, a nonagglutinable vibrio; February 7, 
positive; February 10, 15, 17, and 27, respectively, a nonagglu- 
tinable vibrio; March 1, negative; March 3, positive; March 6 
and 8, negative; April 3 and 4, positive; and since then, negative 
to December. 
I have records of many other cases that I could cite, but these 
are sufficient to show the irregularity with which positive re- 
sults occur. I believe this irregularity is more apparent than 
real; for I believe that, if a cholera carrier were kept under 
exactly the same conditions of temperature, humidity, and sur- 
roundings; given the same food and the same work; and if the 
specimens were taken at the same time of day, in the same 
way, and transmitted to the laboratory under exactly the same 
conditions, the number of positive results would be more than 
doubled. 
The researches of Schébl(7) and of Schébl and Paf%aniban (8) 
on cholera carriers in guinea pigs have shown that food plays 
an important réle in the appearance and the disappearance of 
the vibrios in the feces. The normal period for recovery of the 
cholera vibrios from artificially infected guinea pigs has been 
definitely established to be approximately fourteen days, after 
> 
