228 SUMMARY OF CURRENT RESEARCHES RELATING TO 



In Salamandra they form double loops arranged in a wreath. The 

 nuclear wall then disappears, and the cytoplasm enters the nuclear 

 cavity ; the loop of the filament becoming usually compressed in the 

 middle, thus commencing the formation of the spindle-fibres. The 

 form of these fibres now causes the adjacent Y-shaped pieces of the 

 filament to arrange themselves into a nuclear plate in such a way that 

 each half of a filament coincides with a side of the plate, and the free 

 ends face the poles. The connection around the fragments of the 

 filament is then broken, and the plate then splits into two halves. In 

 Salamandra the double loops open, and each piece divides into two 

 simple loops. 



In those cases in which the segmentation of the filament takes 

 place at a later period, it is preceded by the formation of the nuclear 

 plate, several modifications of the process occurring. The nuclear 

 filament always divides eventually into distinct pieces, and each of 

 these again breaks up into two segments, which are distributed one to 

 each side of the nuclear plate. The elements of the still undivided 

 plate vary in form according to the object, from the roundish granular 

 form to that of J- or U-shaped loops. The spindle-fibres approach 

 more or less closely at the poles, where they sometimes coalesce into 

 a " polar corpuscle." 



The separation of the two halves of the nuclear plate is preceded 

 by its elements passing from their J- or U- through a C- or S- to a f- 

 or £l-form. This is effected either directly by the curvature of the 

 polar and elongation of the equatorial end, or by the curvature of the 

 equatorial end extending like a wave to the polar end. During this 

 process a moment occurs in which the spindle has a barrel-shaped 

 form. The author holds that Flemming's description, both of these 

 cases and of Salamandra, is in some points erroneous, from his 

 having mistaken the polar for the equatorial view of the nuclear 

 spindle. 



The bending of the fi.laments is followed by the separation of the 

 two halves of the nuclear plate. At the poles the elements first of all 

 approach by their polar ends, resulting in an inbending at the equa- 

 torial side, and a coalescence of the separate pieces at both ends. At 

 the same time the whole figure contracts, and a nuclear membrane is 

 formed out of the surrounding protoplasm. The coils of the thread 

 then separate, and nuclear sap is formed between them out of the 

 cytoplasm. The substance of the filament becomes fine-grained, it 

 af'ain becomes longer, the nucleoli again make their aj)pearance 

 as lateral accumulations at the coils, and the nucleus once more 

 enters the resting condition. In the equator of the spindle-fibres 

 which still remain as uniting-threads, and which continue to in- 

 crease in number by fresh-formed threads, the cell-plate makes its 

 appearance formed of microsomes, and from this is formed the 

 cellulose-wall. 



As regards the relation between division of nucleus and of cell, 

 Strasburger somewhat modifies his previous view as to the nature of 

 direct division of the nucleus or fragmentation, that it is fundamentally 

 distinct from indirect division. He now inclines to the opinion that 



