CELL DIVISION IN EGGS OF CREPIDULA. 549 



is illustrated in figs. 183, 185, 189, 190; the latter in figs. 184, 186, 188, 191. 

 When eggs are left for a long time in salt solutions, the chromosomes of daughter 

 nuclei are unable to absorb achromatin and to become vesicular but remain 

 small, densely chromatic masses, and in such cases the achromatin forms vesicles 

 which may lie close to the chromatin (figs. 184-188) or may surround the chro- 

 matin while remaining distinct from it (figs. 178-183). Indeed the entire achro- 

 matic portion of the mitotic figure, with the exception of the centrosomes and 

 spheres, which always remain distinct, may be inclosed in the achromatic vesicle 

 (figs. 178-182). Apparently the achromatic vesicle does not form until the telo- 

 phase of division, and even then it may form in the place of one spindle and not 

 of another (fig. 180), or in one of two daughter ceUs and not in the other (fig. 181). 

 This achromatic vesicle has a very thin membrane and the whole vesicle is 

 elongated in the direction of the spindle axis. Since this achromatic vesicle 

 conforms to the outlines of the spindle in telophase, it is conical in shape, its 

 base being turned toward the centrosome and its apex toward the Zwischenkorper. 

 The appearance of two such conical vesicles in daughter cells with their apices 

 turned toward each other, is strongly suggestive of the constriction and amitotic 

 division of the achromatin. It resembles still more those forms of nuclear 

 division in which the chromatin divides mitotically, but in which the nuclear 

 membrane persists throughout the process, and it and the achromatin divide 

 by constriction. However in this case there can be no doubt that the nuclear 

 membrane actually disappears in mitosis while the membrane which forms 

 around the achromatin is a new structure called forth by the action of the salt 

 solutions. Konopacki has observed and figured in echinoderm eggs subjected 

 to hypertonic solutions conditions similar to those just described for Crepidula. 

 He finds, as I do also, that this membrane is not of the usual type, but is much 

 thinner, nevertheless he seems to regard these forms as due to an elongation and 

 constriction of the nuclear vesicle, in which, presumably, the membrane persists 

 throughout division. Such is plainly not the case in my experiments, where 

 this achromatic vesicle may be present in one daughter cell and not in the other, 

 as in fig. 181. Kostanecki (1904) has described a form of "intranuclear karyo- 

 kinesis" in eggs of Mactra subjected to hypertonic solutions in which a spireme, 

 chromosomes, and a bipolar spindle form within the nuclear vesicle; the chromo- 

 somes separate and move to the two poles and the vesicle and central spindle 

 then disappear while the chromosomes form daughter nuclei. During this 

 whole process faint radiations may be present in the plasma, but there are no 

 centrosomes or astral systems. I have not observed this form of mitosis in 

 Crepidula. 



These results throw light on the constitution of the nuclear membrane. By 

 various writers this membrane has been held to be composed of an outer plas- 

 matic or achromatic layer and an inner nuclear or chromatic one. My own obser- 

 vations (1902) led me to adopt the view of Van Beneden (1887) that this mem- 

 brane was formed from the peripheral walls of the chromosomal vesicles, which 



