35<3 CALIFORNIA ACADEMY OF SCIENCES. [Proc. 3 d Ser. 



could be derived from fig. 3c by the completion of the 

 directives and the interposition of two pairs of non-directives 

 between the directives. Although these assumptions can- 

 not at present be fully tested, they yet receive considerable 

 support from the facts that a single Metridium is never 

 found with an incomplete pair of directives; and that the 

 two mesenteries of a pair frequently vary, independently of 

 each other, in breadth (pair i, fig. 3). In a case of bud- 

 ding, to be mentioned again later, the bud, having received 

 but one directive mesentery from the parent, had grown a 

 second to complete the pair. It is probable, then, that 

 complete mesenteries have arisen, after the division of the 

 siphonoglyphs, between the pairs of directives (figs. 1,7,8). 



Further investigation only can establish the relative fre- 

 quency of those divisions which include the siphonoglyph 

 to the total number. But it seems fair to conclude that 

 since the plane of division of 40 per cent, of the total num- 

 ber of dividing polyps examined has passed through at least 

 one siphonoglyph, there exists in Metridium a predisposi- 

 tion to divide in this manner. 



Division, then, in Metridium, is probably initiated by a 

 longitudinal splitting of a siphonoglyph. A constriction of 

 the mouth follows, resulting in the formation of two mouths 

 in the place of the one, each mouth obtaining a portion of 

 the original siphonoglyph. The division plane usually pro- 

 ceeds aborally, and may almost completely sever the two 

 portions of the oesophagus before any constriction becomes 

 apparent in the oral disk. An irregular line of tentacles 

 grows in between the two mouths; the constriction of the 

 oral disk, proceeding inward, gradually separates these into 

 two series, dividing the oral disk into two parts. The para- 

 pet next divides by means of a similar constriction. A lon- 

 gitudinal groove on each side of the column now marks the 

 course of the division plane; and at about this stage the 

 mesenteries of the foot disk (which is semitransparent) are 

 seen to be arranged about two centers of radiation. I have 

 not been able to follow this rearrangement. The furrow 

 between the newly formed oral disks gradually deepens. 



