576 FERNANDUS PAYNE 



described by Sutton ('02) in Brachyostola and Davis ('08) for 

 several species of Orthoptera. Both these authors describe an 

 end to end conjugation. The second method has been described 

 by the Schreiners ('OG) in Tomopteris, by Morse ('09) in Peri- 

 planeta and by others. In this case the ring is formed by the 

 opening or moving apart of the two parallel parts of the split 

 thread in the middle region, the ends of the two parts remaining 

 in contact. These authors assume a side by side conjugation of 

 maternal and paternal elements. 



In Forficula, Zweiger describes ring formation as arising by 

 the bending and coming together of the two ends of the chro- 

 matin thread, or what I may designate briefly as the bending 

 process. This ring in later stages condenses into a dumb-bell- 

 shaped chromosome. As he believed that the paternal and 

 maternal chromosomes united end to end in synapsis, each part 

 of the dumb-bell represented a single spermatogonial chromosome. 

 Further, as the two parts separate, the first division, according 

 to this scheme, would be reductional. Stevens ('10) describes 

 no complete rings but finds only U's and Vs. She believed the 

 two ends of the U never come together, but straighten out again 

 as condensation takes place. She also believed with Zweiger 

 that telosynapsis takes place and that the first division is reduc- 

 tional. 



My preparations of Forficula show clearly both methods of 

 ring formation. Further, the rings are complete and not merely 

 U's as described by Stevens. In fact, Zweiger's figures indicate 

 that both methods are present. In his figure 60, which is sup- 

 posed to show the various stages of ring formation and the con- 

 densation of this ring into a bivalent, it is a long jump from ^d' 

 to 'e.' In fact, I think there is no connection between them. If 

 we begin at 'd' and read in the opposite direction, 'd, c, b, a,' it 

 will be seen that the split segment is opening up, as if in the 

 initial stages of ring formation by splitting. I have seen no such 

 irregular chromosomes as his figures 'f, g, h,'and hardly believe 

 they occur in ]:)r()perly fixed material, although the two parts 

 of the bivalent do IxmkI around us he figures. While we have 

 had both methods of I'ing foi-ination described, Forficula, so far 



