ORIGIN OF THE SPINDLE 543 



ploid species it is forced to lie oblique or it may even be buckled. 

 The variations in position of the anaphase chromosomes correspond 

 with those of the metaphase plate. Here we see a combination of 

 spatial and timing abnormalities, the second required by the first 

 if a regular division is to ensue. 



The distinction, therefore, between these three types of spindle 

 depends on timing differences and is co-ordinated with spatial 

 limitations. The distinction, on the other hand, between these 

 types and those which develop before any poles are estabhshed 

 is very important. At meiosis in the Coccidae and in some other 

 animals each bivalent forms its own spindle orientated indepen- 

 dently of the others, and only later is a uniform orientation pro- 

 duced, presumably by an external agency (F. Schrader, 1932). 

 Already we have seen how the centromeres influence the shape of a 

 spindle which has been developed by the centrosomes or pole- 

 determinants. Presumably, independent spindles are determined 

 by the centromeres coming into action before the centrosomes. In 

 these cases and in others, in which the poles begin to act as soon as 

 the centromeres, the pole-determinants are not located in any 

 organised bodies, but consist presumably in a diffuse charge on 

 opposite faces of the cell. Such a diffuse charge may act as the 

 opposite pole in an abnormal unipolar (or monaster) spindle (e.g., 

 Urechis, Belar, 1933 ; Triton, Fankhauser, 1934 a and b ; and 

 Sciara, Metz et at., 1926). Here the chromosomes range themselves 

 in a spherical plate around the defined pole and their daughter 

 halves are divided between it and the undefined polar surface. 



In order to see the significance of the centromere spindle we must 

 consider a fourth type of spindle formation, viz. the intra-nuclear 

 spindle. In the Protista generally, in Fungi, and occasionally in 

 the higher organisms the spindle is formed inside the nucleus. 

 It may be controlled by definite centrosome-like bodies, as in 

 Monocystis, or it may be determined by a less localised charge on 

 opposite sides of the nucleus interacting with the centromeres as in 

 Cryptomonas or Pamphagus [cf. Belar, 1926). The question then 

 arises as to why the spindles can be formed inside the nucleus and 

 by the centromeres in some conditions, but are never formed in 

 this way in the greater numbers of higher organisms where the 



