THE MITOTIC CYCLE 



any tendency for the constituent material to aggregate, both Hnes of 

 orientation will bend round and arrange themselves parallel. Such an 

 effect necessarily produces a spindle shape. At more than 45° from the 

 axis of the centrioles, the lines of orientation would be expected to bend 

 the opposite way, and in the sea-urchin egg, at least, there is clear 

 evidence that this actually happens (Plate XIII (22) ). The case of 

 a spindle which is generated without centrioles is rather more complex, 

 and may have to be explained on the lines of tactoids (page 130), where 

 the spindle shape is thought to be due to a compromise between 

 crystallization forces which promote elongation and those of surface 

 tension which tend to cause rounding-up. 



Spindle fibres normally include both chromosomal fibres, running 

 from the centrioles to the centromeres, and continuous fibres, running 

 from one centriole to the other. The former are usually more marked 

 than the latter, and this is probably due to the fact, mentioned earlier, 

 that some regions of the spindle are either more orientated, or contain 

 greater amounts of orientated material (e.g., fibrils) than others. It 

 remains to discuss how these differentiated regions of the spindle are 

 formed. Since the essential structure of the spindle is uncertain, it is 

 perhaps unduly speculative to discuss this problem in detail. There 

 seem, however, to be two possibilities, depending on which of the two 

 structures mentioned earlier is the correct one. If the spindle is a solidly 

 orientated body, with centrioles at either pole, and a number of centro- 

 meres at the equator, we should expect to find enhanced orientation in 

 the neighbourhood of each centromere, but there would seem to be no 

 reason why this orientation should not melt into the general orientation 

 of the spindle. After fixation there should therefore be a marked fibrilla- 

 tion in the neighbourhood of each chromosome, but no clear evidence 

 of fibres running from the centrioles to the centromeres. If, on the other 

 hand, the centrioles and centromeres both generate sub-microscopic 

 fibrils, those which emanate from the centromeres will fuse with those 

 from the centrioles to give typical chromosomal fibres. Fibrils from the 

 centrioles that fail to fuse with those from the centromeres will meet 

 fibrils from the opposite centriole and form continuous fibres. In this 

 instance there should thus be less tendency for the orientation produced 

 by the centromeres to melt into that of the spindle. It is not possible to 

 say with certainty which of these two schemes is the correct one, but 

 even if we allow for a certain tidying-up of spindle fibres in published , 

 diagrams, it appears that continuous fibres, and more particularly 

 chromosomal fibres are usually well differentiated. The evidence from 

 fixed cells would seem therefore to favour the idea of sub-microscopic 

 fibrils as the basic structure of the spindle. 



124 



