E. M. Doidge 21 



bubble where they remain in feeble motion; the remainder gradually 

 come to rest. 



The flagella stain fairly easily by van Ermengen's method. They 

 are three to eight in number, peritrichous, slender and 5 — 7 times the 

 length of the rod (Plate XIV, fig. e). A more satisfactory and simple 

 method of staining was found in one described in detail by Ellis (7). 

 This is a modification of Loffler's flagella stain ; it is a very simple method 

 and gives a clear stain on a ground almost free from precipitate. I have 

 failed to stain the flagella by the original method of Loffler, or by 

 Pitfield's method. 



Involution forms. 



The long threads found in cultures containing large percentages of 

 NaCl may possibly be termed involution forms. Some of these are 

 curved and others are swollen irregularly (Plate XIII, fig. a). A drop of 

 + 15 beef broth was placed on a cover glass, inoculated with rods from 

 a beef-broth culture containing 8-5 % NaCl (Plate XIII) and inverted 

 over a moist chamber. The drops contained a couple of dozen rods of 

 varying length and form ; the rod shown to the right of fig. a was 

 selected for observation. 



In 30 minutes two septae had formed and ten minutes later the rod 

 separated into three distinct lengths (fig. c). These two stages were 

 observed with the T \jth oil immersion (Zeiss) and No. 12 compensating 

 ocular and were drawn with the camera lucida. At this point, however, 

 the rods moved into a deeper part of the drop and it was not possible to 

 get them into focus with the oil immersion. The observation was 

 therefore continued with the Zeiss objective D and compensation 

 ocular No. 18. The drawings were done freehand as near as possible 

 to the scale of the first two. 



The portions of the original filament now divided fairly rapidly 

 into segments of very unequal length, and after three hours they 

 became approximately the normal size of the bacillus (figs, d — g). 

 At this time, the rods being small and rather numerous, they 

 scattered, and it was impossible to trace them further as they 

 became intermingled with bacilli resulting from the division of other 

 filaments. 



Some of the rods divided into much more equal segments than the 

 one selected for observation, and the segments remained in contact 

 longer. Fig. h shows an exceptionally long one and two shorter ones 

 which had divided in this way. 



