542 



CELL MECHANICS 



A second important difference depends on the presence or absence 

 of centrosomes. This problem has always seemed to be a difficult 

 one, but its difficulty does not lie so much in the observation of the 

 centrosomes as in the inference of their action. In the higher plants 

 no centrosomes are seen, but the behaviour of the spindle is not 

 sharply distinguishable from that in animals, in which centrosomes 

 are seen. Abnormal treatment produces multipolar spindles and 

 accessory nuclei, formed by lagging chromosomes at meiosis inde- 



'^\n 



A/v^ 



DC 



PG 2xRT 



:r,PT 



AA^ 



2xPT 



lyvo 



AAn 



LA/J 



K/^ 



\r\A 



Fig. 156. — Comparison of metaphase and anaphase arrangements 

 in pollen grains and pollen tubes of TuJipa, showing the effects 

 of chromosome number, cell size and spindle size on plate 

 formation. (Upcott, 1936.) 



pendently of the spindle-poles, fail to develop a normal spindle at 

 their second division. 



The extreme opposite of the central spindle type of mitosis is 

 found in the pollen-tubes of plants (Upcott, 1936 h). Here the 

 nuclear membrane breaks down before the chromosomes are fully 

 contracted, possibly by the premature action of the spindle. There 

 is a considerable delay in congression on the plate, owing presum- 

 ably to the confined space of deployment for the chromosomes. As 

 in other cases of delay in congression, the spindle lengthens to an 

 exceptional extent. The plate is restricted in width, and in poly- 



