526 H. J. CONN, CHAIRMAN 



the extremely small amount of nitrite present in D it is not likely to 

 have destroyed the nitrite in nitrate agar as fast as it produced it. 



The organisms for which inconclusive results must be recorded 

 are those which produce ammonia in B, which fail to grow on C, 

 and which destroy the nitrite in D. The number of organisms 

 in this group is probably very small. 



INDOL PRODUCTION 



The indol test has always been one of the most unsatisfactory 

 of those that have been used for characterizing bacteria. This 

 has been true, partly because of the variable composition of the 

 media used, and partly because of the inaccuracy of the tests 

 used for detecting the presence of indol. 



The first of these objections (i.e., the variable composition 

 of the media) Zipfel (1912) has tried to avoid by using trypto- 

 phane in place of peptone. As this medium is not satisfactory 

 for routine use other investigators have tried to accomplish the 

 same purpose by using a solution of peptone treated with tryp- 

 sin. Frieber (1921) for instance used a medium of this kind 

 which he prepares as follows: 



To 1 liter of ordinary peptone bouillon he adds 0.2 gram of 

 trypsin, then adds chloroform and toluol to prevent bacterial 

 growth and incubates for twenty-four hours to forty-eight 

 hours at 37°, subsequently filtering and diluting with three 

 parts of physiological salt solution. In the absence of 

 conclusive data as to the advantages of this medium, the com- 

 mittee recommends that organisms be tested both in ordinary 

 peptone solution and in the trypsinized bouillon of Frieber. 



The second objection to indol tests, namely, the inaccuracy 

 of methods for detecting indol has been discussed in some length 

 by Frieber. Indol has generally been detected by the Salkowski 

 method, that is through the use of sodium nitrite and sulphuric 

 acid. This test has been realized for some time to be an inac- 

 curate one and recently Frieber has sho^^•n quite conclusively 

 that this reaction is positive not only with indol and with the 

 methyl-indols but with indol-acetic-acid as well. The most 

 satisfactory test for indol is now regarded to be EhrUch's test 

 which gives a positive reaction only with indol itself and with 



