670 THE NODULE 0Kr4ANISM OF THE LEGUMINOS.E, 



Aftei' from 1 to 2 minutes, the eovei-glass was taken out of the 

 solution, rinsed thoroughly in tap water and then in distilled 

 water. .Staining was effected Ijy immersing the cover-glass, film- 

 side downwards, for 5 minutes in carbol-fuchsin, which had been 

 filtered cold and then warmed. The stained tilm was washed, 

 air-dried and mounted in balsam. When successfully stained by 

 this method the appendages of the cell are revealed. An empty 

 tiibular capsule can sometimes Ije seen attached to the organism; 

 the width of the tube, as well as the frayed end, show clearly 

 what it is. The cell has sometimes a relatively wide diffuse 

 terminal thread, whicli is in all probability a mucilage thread 

 and accidental, since it is too wide and transparent either for a 

 flagellum or for the capsular tube. A few cells have stronger 

 threads varying up to twice the length of the organism. These 

 are exceedingly like the flagella of the bacteria. They may he 

 flagella — it is more probable that the}^ are not, since they are but 

 seldom found. For example, in a 40 hours' culture at 18° C, 

 most of the organisms were actively motile, and a film of this 

 culture showed when mordanted and stained only two cells with 

 these pronounced terminal threads. Had they been flagella there 

 would have been in the same tilm many more cells endowed with 

 these appendages. The culture, however, showed that practically 

 every cell bore an exceedingly thin terminal thread A^arying up 

 to 2 ^ in length, and bearing upon the distal end a tuft like the 

 tuft upon a lion's tail or the lash upon a whip. This is undoubt- 

 edly the flagellum by means of which the cell moves. The thread 

 is so thin that even when mordanted and stained it is seen with 

 ditficulty. The tei^minal tuft, however, is easil}^ made out, and 

 assists in the discernment of the thread. The tufted flagella 

 appear singl}^ and at one end of the simple organisms. 



While th(! coccus form of other investigators is undoubtedly 

 the bud, the spirillum and slightl}^ bent forms are caused by the 

 bending of two or more cells while still enclosed in the parent 

 niembi'ane, and the collection of individual organisms appearing 

 or staining as one bacterium produces the curvature of the 

 supposed simple rod. It must not ])e forgotten, however, that in 



