48o CELL MECHANICS 



In this case there is a way out : we can call in the aid of a different 

 scientific method, hitherto successfully applied in geology and 

 astronomy, to help out controlled experiment with natural experi- 

 ment. In this method we compare conditions in nature which 

 differ, so far as we can judge, in respect of the fewest variable 

 factors. For such a purpose the study of the chromosomes is very 

 well suited. When we take the whole of living organisms together, 

 the chromosomes show a more uniform behaviour both mechanically 

 and physiologically than any other structures. This is not surprising, 

 since they lie at the root of all living processes, and on their accurate 

 mechanical and physiological co-ordination the success of these 

 processes depends. And, as we have seen in regard to the alterna- 

 tion of generations, co-ordination in function enforces stability in 

 evolution. Moreover, the technique of genetics puts ready to our 

 hands the means of analysing mutations which act physiologically, 

 for the most part, in breaking down the genetic system. These 

 mutations are not directly controlled by the observer, but are 

 related in cause and effect to a large body of similar observations, 

 similarly analy sable. 



In attempting to study cell mechanics we must take the further 

 elementary (Baconian) precaution of beginning our induction from 

 the most certain and most exact evidence and proceeding to the 

 less certain and less exact. We must therefore begin with the 

 chromosomes, the cell-components whose behaviour is most constant 

 and most readily verifiable, and then on this foundation (and we 

 are as yet little above the foundations) attempt to construct a 

 general theory of cell mechanics. 



The data for our study consist in observations of mitosis and 

 meiosis in all living organisms. Our first task is to arrange the 

 data in a suitable form for comparison, and to do this we must 

 find out to what extent there exists a uniformity in each of the 

 two important types of nuclear division, mitosis and meiosis. A 

 uniformity in the general character of mitosis has long been under- 

 stood. Its demonstration was completed with the masterly review 

 of nuclear behaviour in the Protista by Belar (1926) . The uniformity 

 of meiosis has been more slowly realised. Its demonstration was 

 completed by the evidence leading to the chia;sma theory of meta- 



