448 



Profs. Percy Frankland and Marshall Ward. 



Curve 153, p. 442. 



Normal doubling periods : 



45 90 fj. = 3.0 P.M. 3.31 P.M. = 30 minutes at 30 '2 31 '0 C. 

 90180 = 3.31 4.6 = 35 31 '0 31 -25 0. 

 180 360 ,, = 4.6 5.3 =57 31 -25 32' 9 C. 



This filament was then left for over half an hour, the temperature 

 continuing the slow but steady rise already seen to have set in, and 

 on resuming the observations it had broken up in a manner very 

 characteristic of such cultures. Since I have not as yet described 

 any of these cases which occur quite commonly, however I pro- 

 pose to enter into the details of this one, as an instructive and 

 typical example. 



When the filament was measured at 4.6 it showed a slight angular 



bend near the middle, like this 



the segment a 



being 85'5 fn long and b 94'5 fi long total length 180, as registered 

 in the table ; when left at 4.15 a had grown to 95 /* and 6 to 107'5 /* 

 total 202'5 ft as seen. 



On resuming the measurements at 4.50 several interesting changes 

 in length and position were observable. The segments a and b had 

 separated, and grown across one another as seen in the accompanying 

 diagram. 



Not only so, but a had bent at nearly a right angle into two seg- 

 ments a short one, a', and a long one, $ measuring 45 /JL and 

 101'25 /JL respectively, while 6 had similarly segmented into a long 

 portion, 7, measuring 153 p, and a short one, 8, measuring 22'5 /t 

 altogether, but really consisting of a piece of empty sheath not 

 represented in the diagram and a piece of solid and normal- 

 looking filament, the latter measured 14 u, and the connecting 



