28 DURATION OF THE SEVERAL MITOTIC STAGES 



upon the probable error of the determinations are known. (See pp. 

 19 and 29.) 



One thousand counts per observation having proven satisfactory, the 

 plan of making similar counts was decided upon for the subsequent study. 



The task of working out a coefficient (see pp. 13 and 30) of nodtotic 

 homogeneity, or synchronization in the mitotic area, was not under- 

 taken, because the preliminary investigation showed in the Procession 

 Index Tables an orderly succession of high points in mitotic waves 

 through successive mitotic stages and time-intervals that would not 

 have appeared had there not been a high degree of parallelism in the 

 mitotic processes in the several samples taken. Judgment, therefore, 

 dictated that it was necessary, in order to make for adequate accuracy, 

 to include in the actual temperature-studies as many cell-counts as 

 possible. Against this one possible handicap of having to use different 

 cells to restore the sequence series, instead of being able to trace the 

 succession of stages in the same cell — that is, in case the index of 

 mitotic homogeneity or synchronization proved to be low — one must 

 balance the fact that many hundreds of stained dead cells can be classed 

 by the statistical method during the time that would be consumed by 

 directly observing and definitely timing, even if it were possible, only 

 a few cells actually moving through their mitotic stages. Remembering 

 that numbers make for accuracy or, to be exact, that accuracy is a 

 function of the square root of the population of the sample, we have only 

 to increase the number of samples counted in order to increase the true- 

 ness of our statistical picture. In addition, as was stated earlier (see 

 p. 5), the statistical method has the advantage of taking fresh and 

 naturally developing tissue and killing it almost instantaneously, thus 

 insuring relatively untampered-with normal samples. 



On Saturday, September 9, 1916, the samples were taken. The 

 root-tips were 5 to 10 mm. in length and varied but little in this respect 

 in the three different constant-temperature chambers; but it must be 

 remembered that growth and mitosis are different processes. The 

 sampling began at 10 a. m. and, as was planned, continued at 10- 

 minute intervals until 1 p. m., 19 observations in all. There was one 

 person at each temperature-box and at the given signal an onion was 

 lifted out and the root-tip quickly snipped with a pair of scissors and 

 dropped immediately into Fleming's fluid. The temperature in the 

 growing compartments did not vary so much as 0.5° C. during the 3 

 hours of sampUng, although each chamber was opened 19 times; 

 doubtless the volume of water in which the onions were sprouted aided 

 in maintaining the constancy. The root-tips were embedded in paraffin 

 and cut in longitudinal sections 6 microns thick, and were stained with 

 Heidenhain's hematoxylin, due precautions having been taken, as in 

 the preliminary work, carefully to label the vials in which the specimens 

 were prepared, and finally to label the slides upon which the series were 

 mounted. 



