ON THE SEA-ANEMONE, SAGARTIA LUCLE. 141 



In order to settle this point it was necessary first to observe actual cases of divi- 

 sion and apportionment of stripes. Attempts were first made to observe division 

 by stocking a large aquarium with many individuals. Into the aquarium a quan- 

 tity of mussels covered with anemones was thrown. The anemones soon crept from 

 the mussels and covered the glass sides of the tank. Here a great many cases of 

 division were followed part way, but, owing to the fact that the individuals may move 

 several inches a day and the length of time spent in division often consumes a night as 

 well as a part of a day, the resulting individuals were not recognizable with absolute 

 certainty. The aquarium, however, taught me that longitudinal division is very 

 frequent in this anemone, and that it begins at the basal end and may easily be over- 

 looked from the expanded oral end until practically completed. 



A few individuals were then isolated in small dishes. By this method the oppor- 

 tunity to find dividing individuals came necessarily more slowly. In this way, how- 

 ever, several cases of longitudinal division were followed through all stages, and in each 

 instance division began at the basal end and was aboral-oral in direction as described 

 by McCrady ('59, p. 275) for Actinia cavernosa and by Carlgren ('93, p. 3) for Pro- 

 tothea simplex and as observed by Torrey ('98) in some cases of the Californian 

 Metridium. The division, in all cases observed by me, passed through the short axis 

 of the mouth so that one pair of directives was given to each of the resulting indi- 

 viduals. Thus a monoglyphic S. lucise is the result of longitudinal division of a 

 diglyphic form. This fact, as Torrey has already pointed out from his observations 

 on the Californian Metridium, precludes the possibility that either the monoglyphic 

 or diglyphic types can have the values of "varieties" as suggested by Parker ('97). 

 I have never observed the plane of division to separate the components of a pair of 

 directives as Torrey states for Metridium, nor have I certainly observed a case of 

 division of a monoglyphic individual, although the rapidity with which division 

 follows division, judging from the great reduction in the number of stripes to one 

 (Fig. 12) or even to zero, inclines me to believe that such division may occur. I have 

 found individuals dividing and others almost sexually mature on one and the same 

 pebble. Multiplication by division and by sexual methods may go on at one and 

 the same time among radiates, as Lang ('86) has shown for Gastroblasta raffaelei. 

 Parker ('99) suggests that the diglyphic type may be the result of sexual reproduc- 

 tion. I have never positively observed the metamorphosis of a monoglyphic to a 

 diglyphic type, although I had one individual under observation which I thought to 

 be producing new directives. Unfortunately the dish in which I kept it was appro- 

 priated by another. However, I have observed that, at the period of greatest activity, 

 in longitudinal division monoglyphic types are the more numerous, while at other 



