Report on the Bacteriology of Water. 283 



close to the rapidly elongating left-hand one, and then an interesting 

 and instructive phenomenon ensued. 



The long, curved, right-hand filament suddenly snapped at one of 

 its septa, and shortly after the drawing (e) was made the two ends 

 began to glide one over the other, and at 2.20 had assumed the 

 positions shown in fig. 14 /. The rapid growth of the extreme light- 

 hand filament caused it to curve strongly, and this marked curvature 

 increased, until at 3 P.M. the large curve looked as if the filament 

 was pressing the S-shaped one, trapped between it and the left-hand 

 spring-like filament, elastically against the latter. (Fig. 14 g\ 



At any rate, the free end of this spring-like right-hand filament 

 slipped soon after over the lower half of the S to the left, while its 

 upper part went on, so to speak, pressing the upper limb of the S also 

 to the left, until it had trapped it close up against the back of the S 

 (fig. 14 h), and then it also snapped. 



Comparison of the figures will convince us that the series of con- 

 tiguous parallel filaments making up the middle portion of fig. 14 i are 

 brought together by these repeated doublings up of the snapped fila- 

 ments, pressing up close to one another. The series also shows pretty 

 clearly how pronounced is the intercalary growth, e.g., along the 

 curved upper and lower line of the S-shaped piece in figs, g, h, and i. 



The process may be summarised as consisting in (1) the looping of 

 filaments, which go on elongating by intercalary growth, the two 

 limbs of the loop being doubled one on to the other, possibly by pressure 

 of other filaments ; (2) the snapping of the same when the doubling 

 up becomes very sharp ; and (3) the straightening out of the broken 

 pieces side by side, and farther parallel growth in close contiguity as 

 a strand or tress. 



The case followed is a relatively simple one. In larger and older 

 colonies, and in cultures where many colonies are growing together, 

 the tresses may consist of scores and even hundreds of such parallel 

 filaments. Moreover, they are not necessarily arranged in the flat, 

 ribbon-like manner depicted in these young and small cultures grow- 

 ing in a thin layer of broth, but may be thick and of various sections. 



It will also be understood how quite apart from any question of 

 mutual attraction between filament and filament the fact of the two 

 terminal stretches of a long and otherwise free loop being relatively 

 fixed as they grow along the sides of such a strand, may force the 

 loop to make coils of various kinds daring the elongation due to 

 intercalary growth, and it is thus I explain the extremely common 

 occurrence of these coils in large cultures. 



Development of Spores. 



Perhaps the most interesting as it certainly has been one of the 

 most difficult series of observations I have made on this schizo- 



x 2 



