544 CELL MECHANICS 



activity of the centromeres has been made clear by their effect 

 on terminaHsation at meiosis. The answer to this question seems 

 to be provided by observations on different kinds of spindle forma- 

 tion in Artemia salina (Gross, 1935). In the cleavage divisions of 

 the egg, spindles are developed by the centrosomes and by the 

 centromeres, independently and simultaneously. These spindles 

 combine and their combination produces metaphase. At meiosis 

 on the other hand the centrosomes play no part at all in the forma- 

 tion of the spindle. It arises entirely within the nucleus, presumably 

 as a centromere spindle. But before this spindle develops the 

 nucleus contracts to a small proportion of its previous size. Now, 

 so long as a nucleo-cytoplasmic surface persists, the difference of 

 structure and behaviour on the two sides of it shows it to be acting 

 as a type of semi-permeable membrane. The sudden contraction 

 of the nucleus, therefore, probably means a loss of water and an 

 increase in the concentration of whatever large molecules are 

 present in the nuclear sap from which the spindle develops. 



The effect of water-content on spindle-formation has also been 

 shown in two ways experimentally. When the spindle is dehydrated 

 it lengthens as at meiosis with non-pairing, presumably owing to 

 an exaggeration of the polar repulsions (Belar, 1929 a). When it is 

 hydrated the orientation of the chromosomes is upset, presumably 

 owing to a decline in all repulsions (Wada, 1935). Thus the forma- 

 tion of the spindle must be supposed to depend not only on the 

 action of repulsion centres such as centrosomes and centromeres but 

 also on the presence of certain materials in the substrate and in a 

 certain concentration. Also, since the changes in the repulsion 

 centres (centrosomes and centromeres) are cyclical and independent, 

 and their effects upon the spindle are cumulative and related, they 

 must be co-ordinated or balanced in their time of action. Artemia 

 shows that this balance may be achieved in different ways. 



(vi) The Balance Theory of Mitosis. The present account has 

 shown that mitosis takes many different forms so far as its external 

 mechanics is concerned. The special mechanical systems at work 

 in many of the aberrant forms, particularly of the Protista, cannot 

 yet be analysed in detail, although they provide useful materials 

 for comparison. The commoner systems, found in the higher 



