[105] EMBKYOGRAPITY OF OSSEOUS FISHES. 559 



of the nuclear plates developed during the second cleavage, Eauber 

 seems to lay stress u})on a more or less extensive swinging round of the 

 axis of the cleavage spindles, after the second and especially the third 

 series of cleavages begin in the frog's ovum, by which he very clearly 

 and beautifully accounts for the variations of segmentation in that 

 form. This he thinks may be due to what he calls sajmental attraction 

 between blastomere and blastomere. The contractions of the protoplasm 

 of the blastomeres during segmenratiou, by which certain ones are dis- 

 placed, repelled or attracted by others, is also considered. He likewise 

 thinks that the poles of nuclear spindles or new centers of adjacent 

 cells may have an attraction for each other. The cleavage planes or 

 the furrows between homologous cells of similar ova may differ i'rom one 

 another in direction as much as 90°; ordinariFy the variation is much 

 less, for a variation of 3° to 5° would be sufficient to produce the most 

 strikiug variations. 



For my own part I would be inclined to ascribe the dislocation of the 

 cleavage planes in such germs as are shown in Figs. 33a and b mainly 

 to the contractions of the protoi)lasmic mass of the blastomeres during 

 segmentation. It is quite certain that at the beginning of cleavage the 

 germs of fish ova are discoidal, and that during the development of the 

 first cleavage furrow the disk elongates remarkably in a direction at 

 right angles to the first furrow, as may be seen in Fig. 9. JSTot only does 

 this occur, but the two new segments are also usually produced on the 

 upper surface into pretty acute obconical points when viewed from the 

 side. This occurs when the cleavage furrow is developing, and im- 

 plies a heaping up of the protoplasm of the blastomere by its own pow- 

 ers of movement. Sometimes after the two blastomeres have assumed 

 the conical form spoken of they slowly subside, in consequence of which 

 they again assume a depressed form with the cleavage furrow very 

 much less distinct. All of these phenomena signify an active move- 

 ment of the substance of the blastomeres, in which tlie nuclear spindles 

 undoubtedly have nn important office to perform. That contractions of 

 the outer substance of the blastomeres do occur we have evidence in 

 the development of the superficial wrinkles in the cleavage furrows 

 shown in Figs. 33a, 35, and 44. 



The eftects of the dislocation of the cleavage furrows by the contract- 

 ility, or perhaps reciprocal attractions and repulsions of the blastomeres, 

 is further shown when we glance at Figs. 12 and 13. The heavy dotted 

 lines of Fig. 33h also show how the third series of cleavage furrows 

 may be displaced so as not to meet each other exactly where they join 

 those of the second i, i. This displacement of the cleavage furrows 

 seems to be the rule rather than the exception in almost all forms of 

 development. This disjointing or dislocation of furrows, too, does not 

 seem to arise from mutual pressure during the early stages so much as 

 from actual contractions, as already pointed out, but the pressure of the 

 cells upon each other, which takes the place of the contractions after 



