MUSEUM OF COMPARATIVE ZOOLOGY. 451 



gams and archisperms, makes it of considerable interest to learn the 

 conclusion which he reaches in his more recent studies (Strasburger '77) 

 on the ''Embryosack " of metasperms. The whole process within the 

 embryo-sac (studied especially in Orchis) is put in a new and unequiv- 

 ocal light. The egg cell, the two " companion cells " (Gehiilfinnen), 

 and the " antipodal cells " (Gegenfiisslerinnen) are all formed, not by 

 a free cell-formation, but by the successive divisions of the cell which 

 forms the beginning of the embryo-sac, and with each division the 

 nucleus undergoes a spindle metamorphosis. From these successive 

 divisions there result eight cells, four in each end of the embryo-sac, or 

 more properly speaking eight nuclei, only six of which (three at each 

 end), become definitely circumscribed cells, since the division is in so far 

 incomplete, that one of the four nuclei in each end of the embryo-sac 

 is left free in the protoplasm of the sac not employed to form the six 

 definite cells. The two nuclei thus left free migrate toward each other 

 and fuse (conjugate?) to form a single nucleus. The group of three 

 cells at the posterior end of the sac are the antipodal cells ; of the 

 anterior group, two are the "companion cells," whose anterior ends form 

 the " Fadenapparat " when it exists, and the remaining one is the egg 

 cell, whose sister nucleus was the anterior of the two copulating nuclei. 

 The " companion cells " cannot be considered equivalent to " canal 

 cells" (or polar globules), since they are not derived directly from the 

 egg cell. The "free" nucleus is the one last to be separated from the 

 nucleus of the egg cell, but its entirely anomalous fate prevents any 

 comparison with canal cells, or, for that matter, with any other, save 

 copulating sexual cells. 



For the present, then, the angiosperms seem to present no opportunity 

 to extend our knowledge of the possible origin of the polar globules. 

 Notwithstanding this there still remain these important facts, to which 

 Strasburger directs attention, since they show that often parts of the cells 

 which are undergoing sexual differentiation are detached at early stages, 

 and are excluded (like polar globules) from the subsequent sexual act : 

 that in ^pirogyra Heeriana a vesicular portion of the cell, which at the 

 time of copulation migrates, is excluded from the copulation (in other 

 Spirogyras, however, this is not the case) ; that in certain algse, for in- 

 stance, a part of the egg substance is simply ejected, and also that not 

 all of the substance of the antheridium is employed in the formation of 

 the spermatozoids; that in higher cryptogams the "Bauchkanalzelle" is 

 formed, and the spermatozoids carry about for a time a vesicle which 

 represents a part of the " Mutterzelle " and which is in no way con- 



