FACTORS IN ONTOGENY 273 



segments (chromosomes) , and each shows the precocious longi- 

 tudinal splitting. These segments shorten up into the double 

 rods of B, which in C are being arranged in the developing 

 spindle. A comparison of these three figures will show clearly 

 that each chromatin segment has divided both longitudinally 

 and transversely, its parts shortening and arranging them- 

 selves in tetrad formation of D. The first division following 

 separates the tetrad along the longitudinal plane of its former 

 splitting (E) and the second division along the transverse 

 plane (F). 



In Cyclops then, the tetrads are formed by the chromatin 

 thread of the resting nucleus breaking up into one half the 

 usual number of segments, and each of these in turn dividing 

 longitudinally and transversely. A tetrad here is made up of 

 two chromosomes slightly united end to end and split longi- 

 tudinally. Thus if abcdef . . . n represent the unsegmented 

 filament of the resting nucleus, a-b-c-d-e-f would show its 

 breaking up into the normal number of chromosomes which 



split lengthwise, forming -> T> -> -v - v in the equatorial plate. 



a o c d e j 



In the Cyclops nucleus of Fig. A the filament has separated 

 into the segments ab-cd-ef . . . n, each of which has split longi- 

 tudinally into > ;> *> etc., and its transverse division, sub- 

 ab cd ef 



sequently becoming more apparent, gives to each tetrad the 



a 



b cd ef 



composition --T-J--V-V etc. By the first division in the 

 ab cd ef 



longitudinal plane, each daughter cell receives a half of each 

 chromosome; in the second, however, in the vertical plane, 

 this is not the case, as can be readily seen. This is clearly a 

 qualitative division, and the daughter cells receive unlike 

 chromosomes. This forms the "reducing division" in Weis- 

 mann's sense, and as such is a most beautiful demonstration 

 of his postulated reduction of the ancestral plasm. 



In Ascaris, however, the evidence is just as clear that no 

 reducing division in Weismann's sense takes place, though 

 the actual number of the chromosomes is also reduced. 



Boveri has shown for the egg and -Brauer for the sperm 

 that the tetrads arise by a double, longitudinal, splitting of 

 the chromatin filament which later breaks into two segments. 

 Thus abed would again represent the unsegmented filament, 



