14 DURATION OF THE SEVERAL MITOTIC STAGES 



greatly for many previous generations, so that this otherwise sudden 

 doubling effect is entirely lost — scattered over long time-intervals — 

 in the average. The fact that there is a mitotic cycle is, in the kind 

 of study here made, of biological import only. Mathematically, the 

 resting stage and the ten arbitrarily marked subsequent divisions of its 

 mitotic course might just as well have been eleven successive sections 

 from the middle of an indefinitely long process. 



(2) There is always a chance that a cell permanently — so far as 

 mitosis is concerned — set aside in the root-structure may be included 

 in the counting. Such inclusion, in the statistical method here fol- 

 lowed, would tend to lengthen the average duration of the resting 

 stage, as indeed it should (but would not make it indefinitely long, as 

 would actually timing each cell by the direct observation method) ; but 

 since this study is primarily one on mitotically active cells, it was 

 sought to eliminate this factor by (a) confining the cell-count to cells 

 within two root-tip diameters of the extreme tip, and thus to avoid the 

 region where many non-dividing cells are being left behind; and (b) 

 by basing the calculations first upon the ten mitotically active stages 

 and later upon the cycle as a whole. 



ADEQUACY OF THE PROCESSION INDEX. 



The adequacy of the procession indices and the inadequacy of the 

 actual counts and of the stage indices, to trace mitotic waves which are 

 plotted graphically in Diagram A of the Method Chart, are shown in 

 Table B of the same chart. The solid line through Table B traces an 

 attempt to follow a mitotic wave through successive time-intervals and 

 stages by connecting the high points in the actual count. It can be 

 seen at a glance that by this method, in the situation here plotted, one 

 wave is early confused with the other, and that thereafter the whole is 

 incapable of further analysis. 



The line of dashes indicates a similar attempt to trace the same 

 mitotic wave by connecting the highest points of the stage indices. In 

 this case the correction is made for difference in (a) size of the sample, 

 and (b) variation in the mitotic index. If several successive stages 

 were of approximately the same length, this indeed would suffice to 

 trace the wave (as would in fact the actual cell-count, if also the sam- 

 ples consisted of the same number of counts) ; but in the stage index a 

 processional correction is not made for variation of length of successive 

 stages of the same cell. One sees, by examining the Method Chart, that 

 tracing by count or by stage index is satisfactory until one comes to 

 stage 4, a very short stage compared with the previous ones. Neither 

 the actual count nor the stage index can, in tracing a wave, cross such a 

 stage — the bridge is shorter (1.2 min. to 4.7 min.) than the width of 

 the chasm (10 min.). Thus, not only the stage-count but also the 

 stage-index method of wave tracing fails. 



