The Problem of Synapsis. 273 



linkage in Mendelian inheritance is exceedingly ambitious in view 

 of the widespread disagreement concerning the genesis of the 

 lieterotype chromosomes. And it is a surprising fact that many 

 chromosome workers appear still to treat the mode of synapsis as 

 a matter of little concern. McClung declines to offer any view as 

 to the genesis of the bivalents in the Orthoptera from their com- 

 ponent halves ; yet it is evident that the whole theory of synapsis 

 stands or falls with its consistency with the events of the matura- 

 tion prophases. 



As rei>ards the events which occur in animals two views are 

 commonly maintained to-day. Janssens, the Schreiners, Von 

 Winiwarter, Agar, Wenrich, Wilson find in general that the initial 

 event is the parallel conjugation in pairs of chromatin (" leptotene") 

 filaments equivalent in number to the chromosomes of the pre- 

 ceding telophase ; this is followed by the longitudinal splitting of 

 the reduced " pachytene " threads. The heterotype chromosomes 

 (" tetrads ") are formed by the drawing out of the double threads 

 of the previous stage (" diplotene ") along the line of cleavage so 

 that the transverse constriction of the tetrad corresponds to the 

 longitudinal split in the diplotene filament. Assuming that this 

 cleavage separates the original conjugants, the heterotype mitosis 

 is reductional, the homotype equational (for the autosomes). Un- 

 fortunately, to be certain that actual segregation of the bivalent 

 C(jmponents occurs, it is necessary to be able to trace the double 

 character of the bivalent at every stage of the process. 



The view stated above has been criticized by Goldschmidt and 

 his pupils, Arnold and certain earlier workers on the Orthoptera 

 (Sutton, Davis). According to tliese workers the heterotype chro- 

 mosome is found to be developed by the union of the looped 

 diplotene thread by its free ends so that the heterotype division 

 constricts the diplotene thread transversely, while the longitudinal 

 cleavage is interpreted as a precocious preparation for the homotype 

 mitosis. According to the exponents of this hypothesis synapsis, 

 if it occurs, must involve the end to end union (telosynapsis or 

 metasyndesis) of the conjugants rather than the parallel conjuga- 

 tion (parasynapsis) described by the American school. In accord- 

 ance with this view they are unable to find any stage at which the 

 full diploid number of elements is present in the meiotic stage — 

 from the first the chromatin is present as the reduced number of 

 longitudinally split loops or threads. They employ the old concep- 

 tion of the " continuous spireme " to explain how the adherence of 

 chromosomes in pairs end to end may have come about ; but no 

 cytologist claims to have actually witnessed a telosynaptic union 

 in animals. Eecently Nakahara (1919) has provided striking 

 evidence for this view in the case of the Perlidse ; but the great 

 difficulty raised is the similarity of the diplotene stage in both 

 telosynaptic and parasynaptic accounts of the meiotic phase. 



