THE SPKRMATOGENESIS OF UMBRA LIM 1 . 137 



tetrads are of the open ring type, later becoming closed, at 

 which time all evidence of their tetrad nature is lost. In Umbra 

 during diakenesis, the chromatin segments into blocks of varying 

 thickness which first pair up and form open rings, but later 

 contract and form closed ones (Fig. 61). A further contraction 

 at the time of the metaphase shows these tetrads condensed 

 until all appearance of their tetrad nature has disappeared. In 

 fact, unless carefully scrutinized, they may be mistaken for 

 dyads. Upon destaining, however, it can be readily seen that 

 they are arranged at the metaphase as typical tetrads (Figs. 

 35 and 36). Fig. 37 shows such tetrads arranged on the spindle 

 for the first maturation division. The condition in Umbra is 

 very similar to that shown in Figs. 45 to 48 by Moore ('95) in 

 Scyllium canicula. In Umbra, as in Scyllium, the fiber attach- 

 ment is median (Fig. 37), and the tetrads separate into dyads, 

 the separation occurring along a longitudinal axis similar in 

 manner to the opening of two V's which have been placed in 

 apposition to each other. Separation begins at the apex of the 

 V's and ends at the basal region. Thus dyads are formed which 

 go to the secondary spermatocytes. 



The behavior of the chromosomes in the secondary spermato- 

 cytes is not as easily followed as in the primary because of the 

 greater irregularity of the chromosomes in lining up at the 

 metaphase and of their subsequent merging together (Fig. 42) 

 soon after they assume definite form. Furthermore, the chromo- 

 somes are much smaller than in the preceding division. In very 

 early metaphase (Figs. 39 to 41), however, they can be readily 

 made out as dyads and can be seen to pull apart (Fig. 43), eleven 

 univalent chromosomes going to each spermatid. Another factor 

 complicating the study of the chromosomes of the secondary 

 spermatocytes is their proneness to rearrange immediately into 

 the metaphase of the secondary spermatocyte from the telophase 

 of the primary spermatocyte. 



Winge's ('22) Figs. 8a to 8c, on diakenesis in the oocyte of 

 Lebistes, indicates a method of tetrad formation similar to that 

 found in Scyllium, Gambusia, and Umbra. Agar ('n) found in 

 Lepidosiren that at the conclusion of synizesis the chromosomes 

 appear in the form of rings which later separate at their ends 



10 



