256 INTESTINAL BACTERIA 
and then the methylene blue. The mixture is shaken and plates are 
poured. The surface of the medium is dried in the usual way before the 
plates are inoculated. We have also obtained exccllent results by sub- 
stituting for Liebig’s extract, meat infusion 1endered free from sugar 
by incubation with B. col. 
Stock solutions of 2 per cent eosin and 0.5 per cent methylene blue 
in distilled water are kept in the dark. We have not sterilized these 
solutions, as we found that they could be kept in the ice-box for weeks 
without causing contaminations of the medium. Ordinarily we do not 
heat the agar after the dyes are added, but we have demonstrated that 
the stained agar can be heated a half hour in the Arnold sterilizer without 
injury. 
Carnot and Halle’s Sand Tube Method. This method has had much 
praise. The method has been described by Gautier (1915) and Levy 
(1916) somewhat as follows: A pipette tube 33 cm. long by about 5 to 6 
mm. in diameter is softened in the middle and the ends brought up to 
form aU. One of the arms thus formed is filled to a height of 10 cm. 
with very fine sand (passed through a No. 40 sieve). The other end is 
filled with hot bouillon and tinted with neutral red, which works up 
through the sand until the other tube is filled to about the same level. 
The arm in which there is no sand is inoculated with the sample and 
incubated for eighteen hours. .The most motile bacteris will penetrate 
the sand and appear in the other arm. Gautier found that B. coli occa- 
sionally did this. Agglutination reactions should be made to confirm 
the results of the sand tube. Other adaptations of the sand-tube 
method have been reported by Borzone and Carbone (1918) and Piazza 
(1916). 
Isolation of B. typhi from Feces. (Teague and Clurman’s Method.) 
This method is especially adapted to specimens which must be kept 
for a short time before examination. The specimens are rubbed into 
a solution of 6.6 per cent sodium chloride containing 30 per cent of 
glycerol. This environment seems to be very destructive to the colon 
bacilli and other bacilli in stools but not harmful to the B. typhi. After 
this, the stool is streaked on plates made from Teague and Clurman’s 
special eosin-brilliant green agar the preparation of which has been 
mentioned in the chapter on the preparation of media. Beckler (1918) 
reports that this method has given very satisfactory results at the labora- 
tories of the Massachusetts State Board of Health. One specimen is 
reported which gave negative results on immediate examination but 
positive results after holding in the presence of glycerol as advised by 
Teague and Clurman. This work is also confirmed by Benians (1918). 
