30 DURATION OF THE SEVERAL MITOTIC STAGES 



less time and observation-instants farther removed but covering more 

 time. 



(6) To find, if possible, a theoretically perfect procession-index. — ^The 

 one used in these studies is highly practical and reliable, but, as was 

 pointed out (see p. 11) in the early part of this paper, it lacks certain 

 theoretical refinements. 



(c) To work out a coefficient of mitotic homogeneity or synchronization. — 

 This could be done by sampling a number of similar-appearing root- 

 tips from the same plant at the same instant, counting a large number 

 (say, 1,000) of cells from each, classifying their stages, and calculating 

 the percentage-frequencies of each, as was done in the study herein 

 reported for successive samples. Then one should calculate through 

 the series of samples, for each stage, the average percentage-frequencies. 

 For each calculation, because the material sampled would be homo- 

 geneous, the usual probable error of the mean would apply. Then 



%-E% 



applying the formula — ^7 = 1. H., we would have a good index of 



/o 

 mitotic homogeneity, for each stage. These values could then be 

 coordinated into a single index of mitotic homogeneity for the entire 

 cycle of mitotic stages. 



Karsten,^ in his studies, appears to have taken 4 or 5 samples at 

 about the same time and to have taken data from each of them, but 

 from each sample his cell-counts are low, generally ranging from 50 to 

 100; which being distributed over the 5 mitotic stages which he used 

 as a basis of classification, would make the calculation of their probable 

 errors valueless. But by further inspection of his tables, one finds a 

 constancy fully in accordance with expectation within the comparative 

 smallness of his samples. This would lead one to expect, in a determina- 

 tion based upon large samples, a low probable error in a coefficient of 

 homogeneity or synchronization. (See pp. 13 and 19.) 



FURTHER ANALYSIS OF THE DYNAMICS OF MITOSIS BY THE 

 STAGE-TIMING METHOD. 



It would be desirable: 



(a) To conduct experimentations similar to those here reported, 

 but in which every qualitative feature would be more precise and every 

 quantitative factor making for accuracy greatly increased. For in- 

 stance: Temperature difference of 2° C. from 8° C. to 45° C. (or from 

 the awakening to the maximum temperatures for growth in the 

 particular plant selected), all other environmental factors constant; 

 sampling at 5-minute intervals for 24 hours; 3 or 4 samples per 

 observation-instant; genotypically uniform material; possibly a revi- 

 sion of the successive stages of the mitotic cycle used in this study; at 



1 See ref. No. 3, p. 13. 



