W. G. YanJSfame — Embryology of Eustylochiis. 279 



to points midway between these junctions, and as they draw upon 

 them the ring becomes a more and more elongated ellipse. As the 

 points where the spindle fibers are attached move apart the line join- 

 ing them becomes the long axis of the ellipse, and the points of 

 union of the segments the extremities of the short axis. The ring 

 then breaks at one of these points, and the tension of the spindle 

 fibers then tends to straighten the whole chromosome (which now 

 consists of two long segments joined at one end only) and to bring 

 it into a line parallel to the spindle fibers. The junction of the seg- 

 ments then comes at the middle point of the long slender rod thus 

 produced, which is not necessarily a straight one. Eventually it 

 breaks here also and the derived chromosomes, each having the form 

 of a shorter rod, usually bent into a U form, pass to their respective 

 poles of the spindle. As the spindle fibers are not attached to the 

 ends of the segments but near their middle, the long rod before divi- 

 sion is usually more or less curved or hooked at each end, the 

 extreme ends not being so rapidly drawn toward the spindle poles as 

 the points half way toward the middle where the spindle fibers are 

 inserted. The segments dei'ived from either this or the straighter 

 form regularly have a U form on account of the tension being applied 

 to the middle of their length instead of the end or ends. Another 

 effect of the tensions thus aj^plied, is to twist the chromosome as 

 soon as the ring has broken at one point. This, together with the 

 hooked end and a tendency which the segments have to become 

 knobbed at their ends (including those that remain joined) often 

 causes the chromosome as a whole to assume an appearance strik- 

 ingly like an italic letter/! The successive stages I have shown dia- 

 grammatically in Fig. 41. Actual examples appear in Figs. 9, 10 

 and 36. 



This process of division may be regarded as the regular one, and 

 the other methods which occur as modifications of it. The most 

 important modification is produced apparently by an increased amount 

 of cohesion of the ends of the future segments when the chromosome 

 still forms a closed ring. As before, it is drawn out into an ellipse with 

 the junctions of the end of the segments at the ends of the short axis. 

 Instead, however, of the ring oisening at one of these points the 

 segments continue to hold together, the ellipse becomes more and 

 more elongated, and the sides eventually come together and appar- 

 ently fuse, making instead of a long narrow rod consisting of the 

 future segments united to each other by one end of each, a short 

 thick rod in which each segment has the form of a loop the two 



