206 DONALD WALTON DAVIS 



pair of directives. This order of events may be interpreted as 

 an epigenetic form- determining series, in which event A leads 

 to event B, etc. 



Influence of old hounding mesenteries 



A most striking difference between the regeneration I have 

 described for S. luciae and that of all the types given by Carlgren 

 and Gary consists in the total absence from the latter of any 

 variation ascribed to the influence of old mesenteries. In Aiptasia, 

 where old mesenteries are resorbed, and even in cases of Types 

 IV and IX in S. viduata, where little or no old tissue belonging 

 to mesenteries or column is present, the lack of influence of old 

 mesenteries upon regeneration is not surprising. But in many 

 other cases, in both S. viduata and in Metridium, well developed 

 mesenteries apparently exert no influence over regeneration. 

 The variations in arrangement of new mesenteries in the species 

 described by Carlgren are of wholly different character from the 

 variations seen in S. luciae, which are governed almost com- 

 pletely by the mesenteries on the torn edges of the old piece. 

 The influence of these bounding mesenteries in S. luciae is 

 apparent in the earliest stages of regeneration. They have no 

 obvious effect upon the directive mesenteries nor upon the first 

 non-directive mesenteries, which precede the directives. The 

 character of the bounding mesentery may, however, determine 

 the nature of the third mesentery on either side. If the bound- 

 ing mesentery is complete, this third mesentery becomes its 

 mate and a fourth mesentery, which becomes the mate of the 

 first non-directive, very soon appears. If, on the other hand, 

 the bounding mesentery is incomplete, the third mesentery is 

 followed by a fourth and a fifth. The determination of the 

 number of mesenteries is effected before these mesenteries have 

 their longitudinal muscles developed. Aly impression, here as 

 in connection with the determination of siphonoglyphs and 

 directives, is that we have to do not with one, but with a num- 

 ber of form-determining influences successively brought to bear. 

 Some of these influences have been suggested, but we are far 



