278 ELIZABETH A. SMITH. 



threads are twisted around each other into a chiasma. Jansen 

 ('09) found that the threads of a tetrad became twisted and 

 his figures resemble closely Figs. 32 and 33. The sex-chromo- 

 some is unaltered. 



6. The Condensation of the Segments into Crosses. The curved 

 segments of stage five, now open up along the plane of the split 

 to form loops and by becoming disunited at one end V's and U's 

 occasionally occur. If the segments are extremely curved 

 when the two threads split apart, rings bent in the form of 

 eights are produced (Fig. 34). These do not correspond to 

 the eights which some workers derive from such gemini because 

 the threads are everywhere separated except at the two ends. 

 An eight corresponding to one formed from a geminus some- 

 times results when the two component threads of a segment 

 while remaining attached in the middle and at the ends, separate 

 between the middle and the ends. Fig. 35 shows a true eight 

 in the nucleus of a cell and the resemblance to the bent rings in 

 Fig. 34 is apparent. The threads of each segment are granular 

 and appear much like the leptotene threads of stage 2. In my 

 preparations of Anax junius the chromomeres are plainly dis- 

 tinguishable in the threads (Fig. 36). 



This stage is much like the prophase of the Orthopterans 

 described by McClung ('14). Among the opened out loops, 

 there are many modifications. The most striking one is the 

 signet ring. If the two threads of a segment divide along the 

 synaptic plane (a point which will be discussed more in detail 

 later) a ring is formed which is composed of two autosomes 

 united at both ends and pulled apart in the center. Now if a 

 secondary split takes place in each thread and the two halves 

 become separated at one of the synaptic ends, the signet ring 

 type is explained. When viewed from the side such a ring 

 appears as a loop with crossed ends (Figs. 37 and 38). By far 

 the larger number of segments open out into true rings without 

 this modification and this type was not found in every nucleus. 



The two autosomes after opening out to form a ring remain 

 connected at the ends. The granular threads condense into 

 more compact ones and each apparently bends out in the middle 

 in opposite directions until the connected ends which were far 



