CELL DIVISION IN EGGS OF CREPIDULA. 535 



Not infrequently the blastomeres of the 2-cell or 4-cell stage are separated 

 from one another by the action of diluted sea water, and the further development 

 of such isolated blastomeres is similar to that which occurs when the blastomeres 

 are isolated by shaking. Thus figs. 135, 144, 151, 155 represent various stages 

 in the development of blastomeres isolated in the 2-cell or 4-cell stage by treat- 

 ment with diluted sea water ; they closely resemble figs. 33, 35 and 44, where the 

 blastomeres were isolated by shaking. In fig. 144 the macromeres A and B were 

 separated from C and D in the 4-cell stage; each macromere has given off a micro- 

 mere of the first set, which has divided normally, and the micromercs of th< 

 second set are forming in typical fashion, except that the chromoM»m< are 

 scattered along the whole length of the spindles. Fig. 151 is the half of an e«g 

 in which the second cleavage furrow was suppressed, and the m< thod of its 

 isolation is indicated by figs. 149, 150; in both of the latter the yolk cleavage was 

 suppressed after the first cleavage and the two macromeres were nearly separated 

 from each other, remaining connected only by a narrow bond; microim res have 

 formed from these half -isolated macromeres in more or less normal manner, 

 though I am unable to identify them individually. 



In a few instances, as shown in fig. 136, the polarity of the egg may be change* 

 or even reversed. In this figure a poly aster was present at the first cleavage, 

 which was also abnormal in other respects, but the most significant abnormality 

 is found in the position of the cytoplasm, nuclei and spheres at the side of the 

 cells opposite the polar bodies; the polarity of these cells is reversed as compared 

 with that of normal eggs. I am unable to explain how this condition has arisen. 



Finally we may consider the effects of diluted sea water on the mitotic 

 figures. As already noted a very common abnormality is the presence of poly- 

 asters in division stages and, as a result of this, of multiple nuclei and spheres in 

 the resting stages. Such polyasters are, in most cases, due to the suppression 

 of the cleavage furrow and to the subsequent interference of amphiasters, origi- 

 nally independent. 



Another abnormality of mitosis which is so common in these experiments as 

 to be very characteristic of them, is the scattering of chromosomes along the 

 length of the spindle, and general irregularity in the separation of chromosomes 

 toward the poles of the spindle. Even in amphiasters which are otherwise normal 

 the chromosomes show this irregular distribution (figs. 140, 144, 153) and of 

 course it appears also in polyasters (figs. 135, 137, 138, 139, 147, 152). Fre- 

 quently a part of the chromosomes are never drawn to the poles of the spindle, 

 but are scattered along the connecting fibers, and when the chromosomes which 

 have reached the poles unite to form daughter nuclei, the scattered chromosomes 

 form a chromatic strand connecting the nuclei (figs. 141, 143, 152, 154). In 

 many instances this chromatic connection between the daughter nucleus strongly 

 suggestive of amitosis (figs. 141, 143, 152, 154), although there is no doubt that 

 it is produced in the maimer stated. Similar chromatic connections be t ween 

 daughter nuclei, due to the scattering of chromosomes along the length ot tne 



