60 EDMUND B. WILSON 



In the mean time the small central triad breaks up into its 

 separate components, which then pass to the poles in a very inter- 

 esting fashion. This process always begins before the division of 

 the large chromosomes, and is subject to some variation. Most 

 frequently the three components draw apart in such a way as to 

 leave the middle one lagging near the equator of the spindle while 

 the others are proceeding towards the poles (flgs. 3/i, i). Often, 

 however, one component first separates from the other two (figs. 

 3j, k) ; but even in this case it seems probable that one of the latter 

 is afterwards left lagging on the spindle, since later in the anaphases 

 this arrangement is almost invariable. In these stages the middle 

 component frequently becomes drawn out along the spindle to 

 form a rod which finally passes to one pole to enter the telophase 

 group (figs. 4e,/) . Half the secondary spermatocy tes thus receive 

 two small chromosomes and half but one, the respective numbers 

 being 12 and 11. 



Two observed anomalies may briefly be mentioned. In two or 

 three cases the middle component seems to be degenerating on 

 the spindle (fig. 40) ; but if this be really the case it must be of 

 rare occurrence, as is shown by the second division. Another 

 interesting anomaly is shown in fig. 4/i. Here there are appar- 

 ently five small chromosomes, two of which are smaller than the 

 others and are connected by a fiber as if they had recently divided. 

 I am uncertain how to interpret this case, for one of the larger 

 chromosomes (stippled in the figure) is paler than the others and 

 lies at a lower level. This may be a fragment of the original 

 plasmosome. If this be the case we have before us a case in which 

 the central small chromosome has divided precociously. If all 

 the five bodies, on the other hand be chromosomes, one of them 

 would seem to be an extra or adventitious body, comparable to 

 those described and figured by Paulmier in Anasa ('99, fig. 28a). 



c. The second spermatocy te-division 



As is to be expected from the asymmetrical distribution of the 

 three small chromosomes in the first division the secondary sper- 

 matocvtes are of two classes. These divisions are very numerous 



