Report on the Bacteriology of Water. 279 



tenuous membrane, pushes its way out and grows as a blunt rod, 

 about 1'75/t broad, with, rounded apex, in the direction of the longer 

 axis of the spore. (Figs. 9 c, 10 c, 11 c.) 



In about four or five hours from the beginning of germination 

 this straight rodlet has attained a length equal to twice that of the 

 spore, and two or three hours later it has a length of approximately 

 four times that of the spore (figs. 9 e, 10 d, 11 e), the membrane of 

 which is still observable usually as a cap at the proximal rounded 

 end of the rod. (Same figures and fig, 12 6, d, e.) 



The above is by far the commonest mode of germination, but in 

 some cases this normal condition of affairs is so far modified that 

 both ends of the spore are softened, and each gives rise to a germ- 

 inal rodlet (fig. 12 c and g) in which case the remains of the spore- 

 membrane may be found either encircling the germinal rodlet, much 

 as a napkin ring does a rolled up serviette (fig. 12 c), or ruptured at 

 one side and merely adhering to the rodlet as in fig. 12 g. Occa- 

 sionally, rodlets which have germinated out in the normal mode are 

 found with the collapsed membrane lying loosely at one end, evidently 

 having been thrown off, as in fig. 12 f\ this seems to occur rather 

 frequently in the later stages of germination in broth-cultures. All 

 these phenomena point to the elastic nature of the thin, but tough, 

 spore membrane. 



When the germinal rod or filament has attained a length equal to 

 about four or five times that of the swollen spore, the first division 

 wall is usually seen in the centre (figs. 11 e, 12 d, and e). Whether 

 the case illustrated in fig. 10 d is really an exception to this rule, or 

 whether the apparent septum closer to the spore was really the rim 

 of the burst spore (c/. fig 12 e), I cannot be certain ; from the fact 

 that I could not trace it in the next stage (fig. 10 e) it seems likely 

 that the latter supposition is the correct one, and in any case the 

 rule is that the first transverse septum divides the whole germinal 

 filament into two cells approximately, but not necessarily exactly, 

 equal in length, and measuring about 3 to 5 /t long by 1-75 p. broad. 



The germination now rapidly proceeds by the growth in length of 

 the stiff and nearly straight filament along its whole course, and in 

 about six to eight hours from the commencement of the swelling of 

 the spore, the filament is from 8 to 10 times as long as the spore 

 (figs. 9 e, 10 e, 11 /), and each of the two cells into which it was 

 segmented by the first transverse septum has been again bisected by 

 a septum, thus cutting the filament into four segments (figs, as 

 before). 



The growth is therefore not merely terminal, but intercalary along 

 the whole filament. I have spoken of the latter as being stiff and 

 nearly straight ; the qualifying word is necessary because close 

 observation of broth cultures shows that the rounded apex performs 





