Report on the Bacteriology of Water. 301 



show) divide at different rates from the first. I am disposed to 

 regard the slower growth of the proximal segment, in part at any 

 rate, to its being more especially concerned with the absorption of 

 food-materials from the spore ; though the fact that it is still behind- 

 hand, even after escaping from the spore-membrane, may indicate a 

 deeper meaning possibly that differences between basal and apical 

 regions are more strongly defined in these organisms than we suppose. 



But another point must be considered before the curve can be 

 understood, and for this purpose it seems necessary to introduce a 

 simple nomenclature for the divisions and segments. 



We may term the first septum, which divides the whole germinal 

 filament into its first two segments, the primary septum ; thus the 

 primary septum was first visible at 11.30. dividing the filament into 

 a proximal and a distal primary segment. At 12.10 the longer 

 distal primary segment showed a further division by a secondary 

 septum into two secondary segments ; but the corresponding secondary 

 septum in the proximal primary segment was not visible until 12.38. 



At 12.56 the distal secondary segment of the primary distal seg- 

 ment had a tertiary septum plainly visible, and the filament as a 

 whole, therefore, consisted at this hour of five visible portions, two 

 belonging to the primary proximal segment, and three belonging to 

 the primary distal segment. 



On turning to the curve of growth (fig. 27), it may now be 

 possible to understand its principal features if we first accept as a 

 fact that the period when a septum is first distinctly visible in these 

 brilliant living cells is some time after thj moment of actual cell- 

 division. This, I think, must be accepted, because I find stained pre- 

 parations of such filaments show many more septa to be actually 

 present than can be seen in the living filaments, owing to the 

 extremely high ryfrangibility of the protoplasm obscuring the view 

 of the most recently formed and still tenuous walls. 



In the curve referred to, as I understand it, its general form, with 

 a higher and higher rate of ascent as time goes on, is due to the total 

 increasing elongation of all the segments simultaneously aided in 

 this case by the slight continuous rise of temperature. 



But, although the general sweep of the curve is such, there are 

 clear indications of smaller curves, convex also to the abscissae, on 

 which points of maximum rate of growth are seen at 11.30, 12.10, 

 12.38, and 12.56, and points of minimum rate in each case about 

 mid-way between these times. 



Now, it was at just these periods the intervals between which are 

 approximately 40, 30, and 20 minutes respectively that the new 

 cell-walls were first perceptible, and it seems almost certain that 

 the intervening periods of slowest rate of growth, viz., about 11.15, 

 11.50, 12.20, and 12.40 respectively, were the approximate moments 



