GERM CELLS IN PEDICELLINA AMERICANA 29 



somes. Altogether, there can be httle doubt that these bodies 

 represent the knobs of the anaphase of the preceding division in 

 which the longitudinal split has reappeared, separating the two 

 sister univalent portions. It is also of interest to observe in 

 one arm of the clearest of these U's a constriction at the middle 

 point. This corresponds to the secondary constrictions along 

 the univalent arms, which give rise to the quadrupartite bodies, 

 but here the processes are somewhat dissociated. 



The spindle now rotates into a radial position (Fig. 97) with 

 the chromosomes more nearly in the equator. In this figure, 

 the chromosomes are cubical and thick rod-like bodies, and in 

 this regard agree in part with what was described by Conklin, 

 '02, for Crepidtila. It must not be inferred, however, that the 

 quadrupartite appearance of some, is an expression of the real 

 constitution of the bodies. The secondary pair of indentations 

 never cut deeply into the chromosomes and may be entirely 

 absent, leaving only the constrictions of the longitudinal split 

 and from what is known of the past history of the dyads, they 

 are probably of little significance in the interpretation of the 

 division in which they occur. The spindle now elongates con- 

 siderably and moves to the periphery (Fig. 98). The chromo- 

 somes bear the same relations to the spindle that was observed 

 in the slightly earlier stage of the preceding figure ; for here, 

 too, the longitudinal constrictions, wherever they clearly occur, 

 are to be observed in the plane of division. The division 

 process w^hich now follows passes through these points and the 

 second maturation is therefore longitudinal or equational. In 

 the middle anaphase (Fig. 99) the monads appear as small, 

 deeply staining spherules. As far as can be determined, they 

 are smaller than the chromosomes of the late prophase and in 

 the great number of instances are about one half as large as the 

 latter. There is, however, a considerable amount of variation 

 in the size of the individual chromosomes which makes such 

 an estimate of size relations very difficult. In the telophase, 

 the chromosomes (Fig. 1 00), show no change in form of any 

 significance, and are now aggregated around their respective 

 poles, the centers of which are almost entirely faded out. As 



