GERM CELLS IN PEDICELLINA AMERICANA 19 



what I believe to be the final one, difficult as it is to prove it to 

 be such by indirect evidence. Owing to the rapidity with 

 which this division is passed through, I am able to describe 

 only the metaphase and the late telophase. Yet these two 

 stages are sufficient, I believe, to show that here again, it is in 

 the process of the last division that the true reduction in the 

 number of chromosomes takes place. In the cross-section of 

 the metaphase. Fig. 65, the chromosomes are almost without 

 exception no longer the characteristic V's of the preceding 

 oogonial generation but rods. Although I am unable to 

 describe the nature of this transformation in the egg save by 

 analogy of what was seen in the sperm, there can be no doubt 

 that this marked transformation has actually occurred. The 

 number of chromosomes is still unreduced, 22, and from the 

 comparatively small size of the cell, the stage most probably 

 represents the last generation. It is in the telophase of this 

 generation, however, that the contrast with the like phase of 

 the earlier generations is most striking (Fig. 66). The chro- 

 mosomes, in this preparation, are very fortunately not crowded 

 into one dense mass and may therefore be studied individually. 

 Such a study shows that in both of the daughter cells the 

 number of the V-shaped chromosomes is no longer that of 

 the foregoing metaphase, but is reduced to about eleven. It 

 is important to observe also, that these figures differ from those 

 of the preceding generations, showing in some cases, deeply 

 staining knobs at the apices. From both these latter consid- 

 erations, it appears with much probability that these eleven V's 

 are each bivalvent and represent single chromosomes united end 

 to end. 



This evidence in Pedicellina is strengthened by a comparison 

 of the somatic divisions. The chromosomes are in all cell- 

 generations V-shaped, sometimes appearing in the form of 

 temporary rings in the prophases and metaphase. Anticipating 

 the fuller description of the somatic divisions, Figs. 107- 

 109, PL III, show clearly the presence of V's, rings and 

 double V's in the metaphase, and single V's in the anaphase of 

 the first somatic division. This same type of cleavage has been 



