296 Profs. Percy Frankland and Marshall Ward. 



Moreover, the same conclusion forced itself on one if other curves are 

 compared : the general slope in each case is undoubtedly largely due 

 to the lower temperature, but the decided rise in some cases about 

 6 P.M., when the daylight failed, took place in spite of the falling 

 temperature, and the generally steeper ascent of others at this low 

 temperature might be due to the cessation of light-action at least I 

 could see no other way of explaining it at the time. 



It is clear from the foregoing results that the growth of the rods 

 and filaments of this bacillus can be measured by the methods 

 devised, and that the undertaking presents no particular difficulties 

 so far, beyond those incident to all close and patient microscopic 

 investigation. 



But, beyond the fact that growth occurs at various rates, and 

 can be thus measured, the results thus far give us far too little 

 information to be of the value I anticipated from the application 

 of the method. They simply raise a number of questions as to the 

 action of various factors in influencing the course of growth and 

 inducing the variations in the rate of growth which undoubtedly 

 occur. 



Such factors are (1) internal factors, such as the age of the fila- 

 ment, the process of cell-division, and possibly the vigour of the 

 spore itself; and (2) external factors, such as temperature, light of 

 various kinds, and the food-materials, &c. 



Growth and Cell- division. 



I made the following observations with a view to obtain more 

 information concerning the question raised as regards the con- 

 nection between growth and cell-division. The chief difficulty 

 connected with them was, as might be expected, that very high 

 powers have necessarily to be employed, and all the troubles of 

 thin cover-slips, minute and not too deep drops, and so forth, arise. 



I selected the tip of a filament which had been growing vigor- 

 ously all the afternoon and evening, and which was in all about 

 600 u long, and traced its behaviour under the l/12th oil immer- 

 sion. The piece I chose was the terminal portion from the tip to 

 the first visible septum. Its length was measured as exactly as 

 possible, and when the septum was under the division of the scale 

 the tip was exactly under the 14th division. 



I had previously determined that six divisions on the scale used, 

 each representing 4'5 /JL with my ordinary measuring combination, are 

 equal to fifteen divisions by this power, and therefore I was measuring 

 nearly 2 /a, per division (more exactly = 1'82/t). 



At 8.20 P.M. the segment in question measured (under the l/12th 

 immersion) fourteen divisions, i.e., 14 x 1'82 = 25'48 /*, and I started 



