Report on the Bacteriology of }\'trr. 



289 



It is obvious that the growth is moderately rapid (the length would 

 be doubled in about 2 hours at same rate), but it seems to vary 

 from time to time. An elongation of 63 /t in 120 minutes would give 

 nearly 0*3 /a, per minute at constant rate, -whereas the rate varies 

 considerably on either side of that. 



These variations could not obviously be attributed to variations in 

 temperature, for the thermometer was steadily rising the whole time, 

 nor did I think they could be due to the measurements, for although 

 the slight nutations do occasionally interfere with the strict accuracy 

 of these, the disturbance can hardly be imagined to be so great as 

 these variations imply, and repeated experience convinces me that 

 this is not the explanation at all. 



As already stated, the microscope was placed at a south window, 

 under a bell-jar surrounded with black paper, and, except during the 

 short periods necessary for drawing and recording, very little light 

 could reach the object. The day was dull and rainy, but with some- 

 what brighter intervals. During the whole of the two hours the fila- 

 ment was describing the slight writhing movements which I regard 

 as nutation curvatures, and the extent of which can be estimated by 

 the drawings in fig. 13. 



But another idea strikes one in connection with these measure- 

 ments. If we take the germinating filaments, and draw them 

 vertically to scale, on sectional paper, at any rate as regard the 

 lengths attained during the various periods of growth, it is obvious 

 that if their lengths are arranged as vertical lines (ordinates) 

 on a base line divided into periods corresponding to the times 

 (abscisses) then the curve joining the tips of the filaments is the 

 curve of growth, and clearly we may substitute mere vertical lines 

 (ordinates) for the detailed drawing of the filaments themselves. 



For instance, the curve of growth is got by straightening out the 

 filaments of fig. 24, where all are drawn to scale at the indicated 

 hours, and joining the tips or, what amounts to the same thing, by 

 joiniug the upper ends of ordinates of equal lengths erected on a base 

 line. 



