1892.] MICROSCOPICAL JOURNAL. 5 



ment of these appendages on the individual bacteria. In all of the prep- 

 arations that I have examined, that had been stained by this method, 

 there were a large number of bacilli which exhibited no flagella, while 

 on the others the number was variable ; but lying between the bacilli 

 were a greater or less number of flagella that had become detached from 

 the bodies of the germs, presumably during the process of preparation. 

 This is especially prominent in preparations of bacteria that are pro- 

 vided with a considerable number of these appendages, such, for ex- 

 ample, as the hog cholera and typhoid bacilli. This fact renders it dif- 

 ficult, if not impossible, to determine the number of flagella with wiiich 

 the individual bacteria of a given species are provided. 



Notwithstanding this difficulty which I have met, the method has 

 given such satisfactory results in the hands of A. Messea, an Italian 

 investigator, that he has proposed a systematic classification of bacteria 

 based upon the number and arrangement of the flagella. This classifi- 

 cation is as follows : 



I. Gymnobacteria. 



jT '-n ) I- Monotricha. t.. Amphitricha. 



II. Triciiobacteria t 1 * • I D -V • I 



i 2. Lophotncha. 4. Feritricha. 



The Mo7iotricha have one flagelhnn at one pole of the bacillus (^Ba- 

 cillus pyocyamiis). The LopJiotricha have a tuft or bunch of flagella 

 at one pole of the bacillus {Bacilhis of blue milk). The Amphitricha 

 have a flagellum at each pole {Spirilh(7?i vohitans). The Peritricha 

 are provided with rows of flagella {^Bacillus typhosus). 



Kruse,* in a review of Messea's article, says that this classification 

 can have only a secondary value. It is evident that it would conflict 

 verv seriously with the natural grouping of the Schizomycetes, as, for 

 example, the inonofricha would include bacilli, spirilla and at least one 

 micrococcus. (The motile micrococcus described by Ali-Cohen.) 



In order to find some process by which I could determine more defi- 

 nitely the minuter details respecting the number, size, and arrangement 

 of the flagella on especially the hog cholera and typhoid bacilli, I have 

 made a considerable number of tests with Loefflcr's and other methods f 

 and with various modifications in both the preparation of the solutions 

 used and in the technique of their application. The result of this ex- 

 perimental work has been very largely negative, but the careful testing 

 of each slip in the various processes, especially that of Loefller, has been 

 productive in revealing a few facts w'hich are deemed worthy of notice. 

 .Some of these have suggested certain slight modifications in the tech- 

 nique of Loefiier's process which promise to be of considerable value in 

 the further study of these structures. As his method has already been 

 quoted, I shall refer only to those sections of it for which modifications 

 are suggested or which, for other reasons, are deemed worthy of special 

 remark. These are as follows: 



(i) The distribution of the bacteria oft the cover-glass. — It is of 

 the utmost importance that the bacteria are properly isolated in the 

 preparation. This can be accomplished very satisfactorily by the fol- 

 lowing process : The cover-glasses, after being thoroughly cleaned, are 



* Centralblatt f. Bakteriology u. Parasitenkunde, Bd. ix (1891), p. 107. 



■f- Frenkmann's arc the only methods, other than LoefFler's, that I have fi)iind to be of any special 

 value. They involve, however, the same principles as those given by LoefTier, and consequently need 

 not be discussed here. For reference and titles of the various articles on the demonstration of the 

 flagella on motile bacteria see Bibliography. 



