1922] Sivezij: Mitosis in Endamocba Coli 319 



The polar masses or daiigliter eentrosomes are probably formed 

 from a part of the original karj-osome but it has not been possible 

 to trace this throughout the changes in general massing of the 

 chromatin in the central part of the nucleus. As the polar masses 

 divide and move to opposite poles, a slender thi'ead, the intradesmose, 

 is drawn out between the two portions and lies at all times within the 

 nuclear membrane. This is the same intranuclear structure as that 

 found in Councilma.nia hifleuri which we called intradesmose (Kofoid 

 and Swezy, 1921) to distinguish it from the paradcsmose of the flagel- 

 lates, the latter structure being outside of the nuclear membrane 

 (Kofoid and Swezy, 1915). 



Along with the formation of the chromosomes, the nucleus becomes 

 elongated (pi. 29, figs. 4—7). The spindle is foraied between the polar 

 masses within the nuclear membrane which persists throughout the 

 entire process of division. In the equatorial plate stage, the spindle 

 fibers, to which the chromosomes are attached, are found to lie in the 

 outer surface of a dovible cone (pi. 30, fig. 13). The intradesmose is 

 found to lie close to the spindle fibers (pi. 29, figs. 8, 9 ; pi. 30, fig. 15), 

 and, as it soon loses its intense staining reactions, it becomes difficult 

 to differentiate from the spindle fibers. It is probably retracted in 

 the late telophase along with the drawn-out nuclear membrane. 



The first division of the luiclevLS was not found in the material 

 under observation, but it is probable that the mode of procedure is 

 the same in this as in the later divisions. In cysts containing more 

 than one nucleus the time of division of each is not always synchronous 

 for the different nuclei (pi. 29, fig. 7; pi. 31, figs. 19-23), which 

 accounts for the frequent appearance of individuals of End<imoeha 

 coli with five, six or seven nuclei, or any number between eight and 

 sixteen. 



METArHASE 



There is some suggestion that the chromosomes divide early in the 

 prophase, as they appear double before the spindle is formed (pi. 29, 

 fig. 7). No conclusions regarding the relative positions of the chromo- 

 somes at this stage, or as to how they become attached to the spindle 

 fibers, could be drawn from observation on the material studied. 

 When the metaphase is reached the appearance of division in the 

 chromosomes is lost, and they are attached to the middle of the spindle 

 as twelve slightly irregular, compact, rod-like graniiles. In end view 

 these form a ring (pi. 30, fig. 13) in the middle plane of the nucleus 

 with a clear area in the center of the spindle. 



