CELLS IN DIVISION 



spindle, having previously been unable to take in new material, is 

 suddenly enabled to do so. An exactly similar state of affairs was 

 obsei'ved by Swann^^^ in the asters of the sea-urchin egg. Having been 

 constant in size throughout prophase and metaphase, they start growing 

 rapidly in anaphase. 



Anaphase evidently involves a drastic change in the capacity of the 

 cytoplasm to aggregate and orient. Swann^^^ has related this to the 

 release by the chromosomes of a second active substance, with rather 

 different properties from the one responsible for the decline in bire- 

 fringence of the spindle. The argument for postulating a second sub- 

 stance, however, is not easily summarized, and the reader is referred 

 to the original paper for details. 



For whatever reason, the spindle does take in new material at ana- 

 phase; it remains to decide how it does so, for growth of the spindle 

 would be expected to lead mainly to an increase in diameter, whereas 

 in fact there is only an increase in length. This, like many other 

 properties of the spindle, can probably be explained in terms of 

 a hydrated system of asymmetrical particles. Such particles tend to 

 aggregate linearly, but for reasons discussed earlier, a certain curva- 

 ture is imposed on them in the spindle. If extra material is then 

 taken in, as the result of some change in the conditions governing 

 aggregation, the spindle will tend to acquire a greater curvature. This, 

 however, will be opposed by the tendency towards linear arrangement 

 of particles. Providing that internal rearrangement can take place (and 

 for a highly hydrated system this is almost certainly the case) a new 

 equihbrium will be established, and the spindle will elongate. It might 

 be expected that the original curvature would be restored, though in 

 fact the spindle at the end of anaphase is usually less curved than at 

 metaphase (p 132). Presumably therefore there must also be an in- 

 crease in the forces tending towards linear arrangement; in view of 

 the increased tendency to aggregate new material this is not unexpected. 



Many different factors must be involved in the growth and elongation 

 of spindles. The strength of the various forces involved in aggregation, 

 the nature of the centrioles, the size, shape and degree of hydration of 

 the molecules and micelles taking part may all be important. It is 

 possible, however, to make one or two rough guesses at the behaviour 

 of elongating spindles. It is likely, for instance, that strongly curved 

 spindles will elongate more than those that are long and thin. In the 

 case of four spindles for which data are available, this seems to be 

 true (Table V). 



The relative sizes of the metaphase spindle and of the whole cell must 

 also influence the degree of spindle elongation. In most somatic cells, 

 for instance, the metaphase spindle more than half fills the cell, so that 

 the amount of new material that can be incorporated inevitably limits 



129 



