552 Chas. W. Hargitt. 



indicated above. In keeping with this view we may naturally 

 proceed upon our analysis and comparisons under two heads: 



(i) The marginal system; and (2) the nutritive and repro- 

 ductive, or vegetative system. 



Concerning the canal system nothing will be said in connection 

 with the study of the ephyrae, since during the early larval history 

 this system is but slightly developed and therefore of but small 

 consequence in relation to variation. 



Furthermore, since as already suggested the purpose is in part 

 a comparison of the aspects of variation presented by the ephyra 

 and adult, we have again a two-fold division of the subject. 

 And since in the order of nature the ephyra precedes the adult 

 this may as well be taken as the order of research. 



Variation in the Ephyra. 



Aside from the investigations of Haeckel (op. cit.), so far as I 

 am aware Browne is the only investigator who has taken up this 

 phase of the subject in detail. And furthermore, since hereto- 

 fore investigation has been directed almost wholly to European 

 species it has seemed to me highly desirable that similar observa- 

 tions be made upon those of American waters, in order to have 

 some broader basis of comparison and deduction. 



Marginal Organs. 



The ephyrae studied are of two series, first those collected in 

 1901-02, and second, those collected in 1904. I choose to consider 

 them in this detached way chiefly from the fact that the latter were 

 just in process of metamorphosis into young Medusae, and it 

 seemed better to study them with a view to securing possible 

 evidence as to any ratio of selection which might be detected as 

 occurring during this process. The ephyrae were all of Aurelia 

 flavidula, except possibly a stray specimen of Cyanea which 

 might have drifted among them. But these were exceedingly 

 rare, if occurring at all, since an examination of several hundred 

 of the series of 1904 in process of metamorphosis failed to reveal 

 the presence of a single specimen. Moreover, since in an earlier 

 study of the development of Cyanea I found abundant evidence of 

 similar variation, the presence of an occasional specimen in the 

 estimation of percentage variation could hardly affect the results. 



