406 EDWIN G. CONKLIN. 



conditions just described those which obtain when mitosis is 

 not followed by division of the cell-body. In a former paper 

 (Conklin, 1912) I have described such cases at some length and 

 need not here go into details. In brief, if the daughter nuclei 

 and centrosomes lie so far apart in the undivided cell body 

 that they do not interfere, in subsequent mitoses every one of 

 these mitoses may be entirely normal and development may 

 be typical except that no differentiation ever appears between 

 the halves of the undivided cell. On the other hand, if the 

 daughter centrosomes lie near together in the undivided cell 

 body so that they interfere we get tripolar or tetrapolar figures 

 with irregular distribution of chromosomes and usually with the 

 formation of several karyomeres of varying sizes depending upon 

 the number of chromosomes entering into them. Such multi- 

 polar mitoses in Crepidula are rarely followed by division of the 

 cell body so that at every succeeding mitotic period the number 

 of centrosomes and chromosomes in this undivided cell body are 

 approximately doubled (Figs. 26-29). Of course such cells with 

 abnormal numbers of chromosomes and centrosomes never 

 develop normally. Normal differentiation depends upon the 

 regular distribution into separate cells of daughter centrosomes 

 and chromosomes as well as of different cytoplasmic substances. 



III. SEPARATION OF CHROMATIN AND ACHROMATIN AND FORMA- 

 TION OF CYTASTERS. 



The behavior of the chromatic and achromatic parts of the 

 nucleus in hypertonic and in hypotonic media throws a certain 

 amount of light on the constitution of the normal nucleus and 

 on the behavior of these nuclear constituents during normal 

 mitosis. When resting nuclei are subjected to hypertonic solu- 

 tions (e. g., 2-4 per cent. NaCl in sea-water) the chromatic 

 portion of the nucleus contracts into a small dense mass leaving 

 the achromatic portion as large as ever (Figs. 32, 43, 48). It 

 looks as if the chromatin had undergone complete "plasmolysis" 

 whereas the volume of the achromatin had not been affected 

 at all. The membrane or boundary of this achromatin remains 

 full and unshrunken, which would presumably not be the case 

 if this outline represented a real plasma membrane. The 



