20 Papers from the Marine Biological Laboratory at Tortugas. 



applied to the nuclear wall (fig. 48). Successive stages show the reticulum 

 arranging itself into the form of a long, continuous ( ?) thread, more 

 chromatic than during the immediately preceding stages. Attached to one 

 end is the much-elongate, club-shaped accessory chromosome (fig. 49). 

 Figure 50 shows two daughter-nuclei at this stage still connected by a cyto- 

 plasmic bridge of the mother-cell. Here the nuclear reticulum is still more 

 or less of a lattice-work character, but the threads are thinner and more 

 chromatic than in figure 48 and the accessory is already decidedly club- 

 shaped. Successively later stages reveal a segmentation of the continuous 

 thread and the arrangement of these segments into loops at one pole of the 

 nucleus (figs. 51, 52, and 53). The final stage probably represents the 

 synizesis stage of McClung. While these changes are taking place in the 

 general reticulum, the accessory chromosome lengthens into a heavy rod- 

 like structure, always with the pointed end attached to the chromatic spireme. 

 Presently it begins to split longitudinally (fig. 52), and at synizesis it has 

 also assumed the form of a loop, intensely chromatic, however, in contrast 

 to the lighter-staining loops at the pole. The loops now begin to straighten 

 out (figs. 54 and 56) and unite at their free ends into pairs, forming two- 

 armed larger pointed loops (figs. 54 and 55). These loops are in length 

 almost equal to the diameter of the nucleus, whereas the loops of synizesis 

 were only of the length of the radius. This is the synapsis stage. 



The point of synapsis is frequently very definite and after the consum- 

 mation of the process is marked by a more intensely staining area. In synap- 

 sis the longitudinal split of the accessory is again closed up, and this struc- 

 ture becomes again a compact, deep-staining, more or less club-shaped bodv 

 closely applied to the nuclear wall and now apparently unconnected with any 

 of the chromosomal loops. It is as though the univalent accessory chromo- 

 some had segmented for the purpose of again uniting the products in unison 

 with the synapsis of the ordinary chromosomes. If synapsis, then, means 

 the fusion of pairs of chromosomes to produce bivalents, and since the acces- 

 sory has no mate, the reduced number of chromosomes should theoretically 

 be 18, and this is just what one finds in the equatorial plates of the ensuing 

 mitoses (figs. 74, 75, 76, and 77). Figures 57 and 58 represent later stages 

 in synapsis and are probably identical with the " bouquet stage " of Eisen. 

 In figure 59 is shown a late postsynaptic stage which is very similar to the 

 early presynaptic stage (fig. 48), except that the chromatin thread is not 

 disposed in a regular lattice-work fashion and the reticulum is more highly 

 chromatic. A very brief resting-stage is again interpolated after post- 

 synapsis (figs. 60 and 61) just as before presynapsis, but almost immedi- 

 ately signs of the subsequent prophase appear. 



In figure 60 the accessory chromosome is ring-shaped (representing a 

 hollow sphere), an appearance frequently met with in methyl green prepara- 

 tions. In fig. 61 the accessory is very chromatic and applied to the nuclear 



