ASEXUAL REPRODUCTION IN SAGARTIA 215 



number of specimens showing odd numbers of stripes are speci- 

 mens which have recently divided and have either developed no 

 new orange stripes or have produced, at the time of observation, 

 less than the full set to be acquired. 



The number of the new orange stripes gives some idea as to 

 the completeness of the regeneration. From what has been said 

 of the number of mesenteries regenerated and of the relation 

 of orange stripes to mesenteries, it is evident that the number of 

 new orange stripes may vary from five to eleven. In case of 

 division in one or two directive endocoels, twelve or thirteen 

 are possible numbers. Numbers below seven or above eleven 

 are, however, distinctly uncommon for single completely re- 

 generated areas. Below seven it is likely that the stripes of the 

 new region are not fully formed. Above eleven — unless one of 

 the orange stripes on the boundary between new and old lines 

 is opposite a white line on the oral disc and therefore occupies 

 a directive endocoel — it is almost certain that the new area con- 

 sists in fact of two regenerating zones of not very different age. 



The number of mesenteries regenerated is such that repeated 

 fission and regenerations (p. 198) wou d tend toward an average 

 condition with about seventeen complete mesenteries. Since the 

 complete mesenteries are always paired and correspond, with 

 rare exceptions, with the orange stripes, we may say that the 

 tendency is toward approximately eight pairs of complete 

 mesenteries and sixteen orange stripes. Division followed by 

 complete regeneration would rarely give rise to individuals with 

 fewer than seven orange stripes, and never with fewer than 

 five. The great majority of individuals with small numbers of 

 orange stripes have not completely regenerated, while even 

 among specimens with high numbers of stripes many have not 

 yet completed the regeneration. 



As to the order in which orange stripes appear in the new 

 region, I have little information. Frequently the presence of 

 stripes may be ascertained while as yet they are so faint in 

 color and so close together that the number cannot be de- 

 termined. At a somewhat later stage there is some reason to 

 believe that stripes are present in complete endocoels when as 



