Artificial Parthenogenesis in Thalassema Mellita 117 



employed or the duration of immersion was not such as to yield 

 the best results. Many of these abnormalities are similar to those 

 which have been described by others (c/., especially Wilson, '01) 

 and need not be spoken of in detail here. That such abnormal 

 cleavages lead to the formation of ciliated structures is clearly 

 indicated by the fact that the percentage of all eggs dividing, both 

 normally and abnormally, nearly agrees with the percentage of 

 swimming embryos which are later found in the same culture. 

 Many such counts were made, and the correspondence in per- 

 centage was found to be remarkably close. I have, moreover, 

 frequently isolated abnormally segmenting eggs and directly 

 observed them to develop into ciliated, cellular structures. It is 

 extremely doubtful, however, whether eggs that divide irregularly 

 ever undergo a later regulation and produce normal embryos, for all 

 the ciliated structures which were raised from isolated, abnormally 

 segmenting eggs departed more or less widely from normal forms. 



Previous experimenters on artificial parthenogenesis of annelids 

 have observed that ciliated structures may arise from unsegmented 

 eggs by a process of progressive cytoplasmic diff'erentiation which 

 takes place in the entire absence of cleavage [Loeb ('01) and Lillie 

 ('02) in Chaetopterus, Fischer ('02, '03) in Amphitrite and Nereis, 

 Treadwell ('02) in Podarke, and Scott ('06) in Amphitrite]. In 

 Thalassema, on the other hand, it is undoubtedly true that eggs 

 which do not divide never undergo further differentiation and 

 never form ciliated structures (c/. Bullot, '04). Without excep- 

 tion every swimming embryo observed possessed a well-marked 

 cellular structure, and eggs which had failed to segment after 

 residence in acidulated water, when afterwards isolated, in no 

 instance gave rise to a differentiated body. 



Certain common types of abnormal cleavage were constantly 

 met with, e. g., every degree of inequality in size of the blastomeres 

 formed by the first and second cleavages was found, while eggs 

 were frequently observed to fall at once into three or four cells at 

 the first division. The trefoil stage, in fact, seemed to be character- 

 istic of certain solutions, in which a great preponderance of eggs 

 exhibiting this abnormality was noted. An unequal division of 

 the eggs at the first cleavage was occasionally followed by division 



