KOFOID. — LAWS OF CLEAVAGE. 187 



in the case of odd generations. Macromeres when present may- 

 be designated by capital letters, or any special quartet, as the 

 " primary micromeres " may be distinguished by special forms of 

 type without change of letter or exponents, or subordinate di- 

 chotomous systems may be introduced for protoblasts and their 

 progeny. 



My work upon Limax is as yet incomplete, but it is probable 

 that at least through the thirty-six-cell stage the cleavage of 

 Limax is identical, blastomere for blastomere, spiral for spiral, 

 with that of Nereis. Beyond that stage I have not followed 

 the cleavage. In the eggs of Limax agrestis the yolk is almost 

 equally distributed, and cleavage is almost equal. Macromeres 

 in the etymological sense do not exist after the eight-cell stage, 

 and it is with difficulty that the poles of the egg can be dis- 

 tinguished by the size of the blastomeres after the sixteen-cell 

 stage. 



The discussion in Limax will be limited to the generations from 

 the third to the sixth, inclusive. 



Third Generation. The cells (a, b, c, d) of the four-cell stage 

 present the typical arrangement of the furrows at the vegetative 

 and animal poles, and the spiral, as indicated by the obliquity 

 of the spindles, must be called according to my nomenclature a 

 left spiral, as Heymons ('93, p. 249) has called the similar spiral 

 in Umbrella. 



The Fourth Generation is reached in the eight-cell stage by the 

 formation of four macromeres (a 41 , b*\ c 41 , d 41 ), and four micro- 

 meres (a 42 , b 4 -\ c 4 - 2 , d 42 ), Plate I. Fig. 1. Each micromere as a 

 result of the obliquity of the spindle lies above and to the right 

 of the macromere which has a cognate origin with it, and the 

 spiral is therefore a right spiral. 



Fifth Generation. The two quartets of the preceding genera- 

 tion divide at about the same time, giving rise to the sixteen-cell 

 stage composed of the four quartets a 51 - d 51 , a 5 - 2 - d 52 , a 5 - 3 - d 53 , and 

 a sA_ ^5.4^ ^ n inspection of Plate I. Fig. 1, shows that the nuclear 

 conditions of the first quartet are slightly in advance of those of 

 the second quartet. Consequently in Limax the twelve-cell stage 

 is abbreviated almost to obliteration, the egg passing from the 

 eight- to the sixteen-cell stage without the intervention of a pro- 

 nounced twelve-cell stage. This lateral view of the egg also 

 shows that all of the spindles stand in a similar position with 



