198 Edmund B. fFilson. 



The first of these studies {Joiirn. Exp. Zoology, I, i, 1904) 

 was especially concerned with the question of cytoplasmic ^re- 

 localization in the unsegmented molluscan egg, and gave only 

 an incidental account of experiments on the cleavage. In that 

 paper both cytological and experimental evidence was presented 

 to show that the Dentaliiim egg contains from the beginning defi- 

 nitely specified regions, consisting of visibly different materials, 

 which stand in such a relation to the morphogenic process that 

 the removal of particular areas of the unsegmented egg produces 

 corresponding definite defects in the resulting larva. It was 

 shown, further, that during the cleavage process these materials 

 are definitely distributed to the blastomeres of the early embryo, 

 and that when these blastomeres are isolated they give rise al- 

 ways to defective larvae, showing the same general character as 

 those derived from the corresponding regions of the unsegmented 

 egg. I therefore concluded that the development of these eggs 

 sustains His's theory of germinal prelocalization ("Organbil- 

 dende Keimbezirke") as applied to the unsegmented egg, and 

 Roux's mosaic theory as applied to the cleavage process, and is in 

 harmony with the theory of formative stuffs. 



In that paper, the evidence for the mosaic character of the 

 cleavage was given only in part, including only a brief account 

 of the general development, in Dentaliiim, of isolated blastomeres 

 from the 2-cell and 4-cell embryos, and of isolated micromeres 

 of the first quartet. The present paper offers more detailed evi- 

 dence in the same direction, derived mainly from experiments on 

 Patella ca^nilea. The comparative ease and certainty with 

 which blastomeres of any desired stage may be obtained by means 

 of Herbst's calcium-free sea-water led me to hope that a fairly 

 complete experimental analysis of the potencies of the cleavage 

 cells might be carried out; and I do not doubt that in time such 

 an analysis can be effected. Various practical diflRculties, how- 

 ever, have rendered the analysis here offered incomplete in sev- 

 eral directions. Nevertheless the positive results attained form 

 the most detailed and, as I think, convincing evidence of mosaic 

 development thus far produced, and in my judgment clearly dem- 

 onstrate this general principle in the molluscan egg. Despite the 



