CYTOLOGY OF ANISOLABIS MARITIMA BON. 241 



spermatogonial division. Here there is seen a double, rather 

 angular body (xx), and a deeply staining sphere (y). With the 

 establishment of the leptotene threads (Plate XXV, fig. 18), the 

 xx-and the y-element come into close apposition and remain 

 connected, sometimes merely by a small strand (Plate XXV, 

 figs. 19, 20), and also during syndesis they continue this connec- 

 tion. When the zygotene threads are well formed, the xx- 

 element separates from the y-element. Often they come to lie 

 at some distance from each other in the nucleus, but apparently 

 not under control of the centrosome at the positive pole of the 

 nucleus (Plate XXV, figs. 21, 22). 



With the beginning of the strepsinema the xx-element and the 

 y-element again approach each other and are again connected 

 by a narrow strand (Plate XXV, figs. 23, 24). The xx-element 

 reveals its two-fold constitution, when viewed at a favorable 

 angle in well-decolorized haematoxylin preparations, and es- 

 pecially well in crystal violet + alizarin slides. At the establish- 

 ment of the strepsistene stage we notice the coalescence of the 

 xx-and the y-element. In iron-haematoxylin the y-element has 

 a lighter cortical zone, and a more deeply stained center. The 

 doubleness of the x-portion is no longer so distinct (Plate XXV, 

 figs. 25-26). 



When the autosomal tetrads are formed, but are still very 

 granular, there appears on the xxy-complex a highly refractive 

 spherule (fig. 27, nl), which separates from its parent mass at 

 about the time the smooth, deeply staining tetrads are estab- 

 lished. This nucleolar body may then lie anywhere in the nucleus, 

 even in close appostion to one of the autosomal tetrads. As 

 the centrosome divides and preparation is made for the first matu- 

 ration spindle, this nucleolar body diminishes in size and .finally 

 disappears, leaving eleven autosomal tetrads and the xxy-hexad 

 to enter the metaphase plate. 



Not only form but also differential reaction to stains enables 

 one to trace the evolution of the hexad. In safrpira + lichtgriin, 

 the y-element is not so deep a red as the xx-coi&onjent, in ^tages 

 corresponding to those shown in figures 17 2o$ The y-elejtnent 

 also appears vacuolated, and has a greenish tifige in 



