BIGELOW : NUCLEAR CYCLE OF GONIONEMUS MURBACIIII. 359 



numerous, being present in the full somatic number. The nuclei, as 

 well as the cells, are now of course much smaller than during the second 

 cleavage, but just as in that case, the resting nucleus contains no nucle- 

 olus. 1 have not attempted to follow the finer details of the prophase, 

 the available material being unsuitable for that purpose. 



In the metaphase the chromosomes, which, as in adult somatic cells, 

 are dumb-bell-shaped, arrange themselves with their long axes parallel to 

 the future plane of division, so that in polar views their rod-like outlines 

 are clearly evident, while in side views they present the appearance of 

 minute spherical masses (Fig. 159), a condition similar to that seen in 

 adult somatic cells, as well as in spermatogonia and oogonia. The split- 

 ting of the chromosomes is longitudinal, resulting in pairs of rod-like 

 daughter chromosomes ; since normally these are all visible in polar 

 views of the late metaphase, it is then difficult, or impossible, to esti- 

 mate the number of chromosomes actually taking part in the formation 

 of the equatorial plate. Counts, however, of about fifteen such cases 

 show numbers always in excess of twenty-four (27 to 38), so that it is 

 altogether probable that they are present in the full somatic number. 

 During the anaphase the chromosomes rotate so that their long axes co- 

 incide with the spindle axis. End views of such spindles (Fig. 160) 

 are fairly common, and on them the counting of the chromosomes is 

 possible with a fair degree of accuracy, the number being apparently 

 twenty-four, exactly as in the adult somatic cells. It is also evident that 

 the chromosomes are much smaller than at the corresponding stage of 

 the second cleavage spindle (compare Fig. 160 with Fig. 157). 



The later cleavage mitoses agree very closely in all their details with 

 the fourth, and all exhibit the full somatic number of chromosomes. 

 With the decrease in the size of the cells, however, the asters become 

 less and less prominent, until finally, in larvae at about the stage of the 

 first appearance of the primary tentacle, dividing cells no longer show 

 any trace of them, and therefore resemble the ordinary somatic cells of 

 adult tissues. 



In view of Hacker's ('95) well-known discovery of the reduced number 

 of chromosomes in the early segmentation stages of Cyclops, the fact that 

 in Gonionemus, too, these structures are united, and even fused, in pairs 

 in the early cleavage spindles, is of great interest. But the feature of 

 the case, which, if normal, is the most remarkable, is the appearance 

 of bivalent cln-omosomes of about half the somatic number in the second 

 cleavage spindle, whereas in the first, where Hacker discovered the re- 

 duced number, the chromosomes are present in the full somatic number. 



