LABOKATOllY AND GREENHOUSE STUDIES. 127 



The inoculations were examined from time to time, and finally on 

 March 7, IS days after the injection, the material was collected and 

 attem})ts were made to isolate the orjijanisms inoculated. The 

 methods of procedure for identifying tlie coconut organism and 

 Bacillus coli were tlie same and were based on some of the char- 

 acteristic reactions of these organisms. Dolt's synthetic medium 

 No. 1 (a litmus-lactose-glycerin agar, p. 79), litmus milk (p. 94), 

 nitrate bouillon (p. 71), fermentation tubes containing peptone and 

 dextrose with neutral red (p. 80), and, in some instances, gelatin, 

 were used. These media have been recommended by various 

 investigators who have carried out extensive work with Bacillus 

 coll in connection with their ''board of health" investigations and 

 full discussions of them are given on the pages cited. Following are 

 the results of this experiment: 



Inoculation No. 1, with ammonium oxalate: The inoculation point was about the 

 same as with Bacillus coli, only drier. 



Inoculation No. 2, with Bacillus coli: Discoloration for only a short distance from 

 the inoculation hole; a water-soaked discoloration but not appearing like a soft rot. 



Inoculation No. 3, with coconut No. 5: Discoloration extended a distance of 4 

 centimeters from the hole and the tissues appeared under the microscope to be full 

 of bacteria. 



Inoculation No. 4, with coconut No. 5: Discoloration appeared for only a short 

 distance about the inoculation hole; discolored tissues appeared under the micro- 

 scope to be full of bacteria. 



On the agar plates poured in the usual way from these diseased 

 tissues there appeared round, white colonies, typical of the coconut 

 organism in the case of the coconut plates; but round, thin, white 

 colonies, some with dentate margins, both typical and atypical forms 

 of Bacillus coli, in the case of the Bacillus coli plates. 



Transfers were made of selected colonies from these plates to 

 litmus milk, and after five days all of the tubes had produced red 

 surface rings, but in only one case (from coconut inoculation No. 3) 

 had the medium turned entirely red. 



These cultures w^ere also transferred to agar containing neutral 

 red, to Dolt's synthetic medium, and again to litmus milk. In each 

 case negative results were obtained for both the coconut organism 

 and Bacillus coli. 



Plates were again poured from dilutions of the original bouillon 

 tubes containing the diseased material, and this time Dolt's litmus- 

 lactose-glycerin agar was used. In five days pink colonies typical 

 of both the coconut organism and of Bacillus coli were formed on 

 their respective plates. Twelve out of the fourteen plates poured 

 showed these colonies. 



Transfers were made from these pink colonies to nitrate bouillon, 

 and two days afterwards test for the reduction of nitrate to nitrite 



