26 PROTOZOA 



When cell-division is about to take place the linin, or at least 

 the greater part of it, assumes the character of a number of 

 distinct threads, and the whole of the chromatin granules are 

 distributed at even distances along these (Fig. 6, A, B, C), 

 so as to appear like so many strings of beads. Each such 

 thread is called a "chromosome." Then each head divides 

 longitudinally into two. The thread flattens into a ribbon, edged 

 by the two lines of chromatin beads. Finally, the ribbon splits 

 longitudinally into two single threads of beads (Fig. 6, E). 

 During these changes the nucleole or nucleoles diminish, or even 

 disappear, as if they had contributed their matter to the growth of 

 the chromatin proper. In Higher Animals and Plants the nuclear 

 wall next disappears, and certain structures become obvious, 

 especially in the cytoplasm of Metazoa. Two minute spheres of 

 plasm (themselves often showing a concentric structure), the 

 " centrosomes," ^ which hitherto lay close together at the side of 

 the nuclear wall, now separate ; but they remain connected by a 

 spindle of clear plasmic threads (Fig. 6, B-E) which, as the 

 centres diverge, comes to lie across the spot the nucleus occupied, 

 and now the chromosomes lie about the equator of this spindle 

 (Fig. 6, F). Moreover, the surrounding cytoplasm shows a radiat- 

 ing structure, diverging from the centrosome, so that spindle and 

 external radiations together make up a " strain-figure," like that 

 of the " lines of force " in relation to the poles of a magnet. Such 

 we can demonstrate in a plane by spreading or shaking iron 

 filings on a piece of paper above the poles of a magnet, or in 

 space by suspending finely divided iron in a thick liquid, such as 

 mucilage or glycerin, and bringing the vessel with the mixture 

 into a strong magnetic field ; ^ the latter mode has the advantage 



1 The origin of the centrosomes is a problem not yet certainly solved, if indeed 

 it be susceptible of any universal solution. They are certainly absent in many 

 plants ; and, on the other hand, structures which correspond to them often appear 

 in mitotic divisions of Protista. In some cases the centrosomes are undoubtedly- 

 of nuclear origin, and pass out through the nuclear wall into the cytoplasm. 



2 Though the forces at work in the dividing cell are similar in their effects to 

 such physical forces as magnetism, static electricity, and even capillarity, and 

 models utilising such physical forces have been devised to represent the strain- 

 figures of the cell, the cell forces are distinct from any known physical foi'ce. For 

 discussions of the nature of the forces at work, with bibliographies, see Augel 

 Gallardo, Interpretacidn Diiidmica de la Dimsi&n Gelular, 1902 ; Rhumbler, in 

 Arch. Entw. xvi. 1903, p. 476 ; Hartog, O.R. cxxxviii. 1904, p. 1525, and "On the 

 Dual Force of the Dividing-cell," \>t. i. Proc. Roy. Soc. 1905 B, Ixxvi. p. 548. 



