through the directive endocoels as representing the dorsal and ventral lines. Furthermore, in 

 cases in which but a single siphonoglyph occurred it has been customary to regard it as occ- 

 upying a ventral position. On this basis the surface of a Cerianthid occupied by the protocnemes 

 would be the ventral one, while that at which the formation of new deuterocnemes occurs 

 would be designated as dorsal. 



Objections have been raised against this terminology inasmuch as it implies an arbitrary 

 homology of the surface at which the single siphonoglyph occurs throughout the entire group 

 of the Anthozoa, and Carlgren (1893), from his observations on the arrangement of the mus- 

 culature in the mesenteries of adult Cerianthids, has maintained that in reality the siphonoglyph 

 (sulcular) surface in these forms is probably equivalent to the sulcular surface of the Alcyonaria, 

 for instance, and if we term this dorsal, then the protocnemes of the Cerianthids are dorsal 

 and the deuterocnemes form in the ventral chamber. 



In opposition to Carlgren's observations stands Boveri's account of the arrangement 

 of the musculature in Arachnactis, which shows an orientation of the longitudinal muscle fibres 

 of the protocnemes with reference to the siphonoglyph identical with that occurring in such 

 monoglyphic forms as the Zoantheae and Oractis, assuming, as I believe there is every reason 

 for doing, that the sequence which I have assigned (1891) to the protocnemes of these forms 

 is correct. But apart from this and disregarding entirely the evidence derivable from the arran- 

 gement of the musculature, there is the more pertinent evidence derived from the homoloo-v 

 of the mesenteries as determined by their sequence. If this be accepted as a criterion, and the 

 sequence of the Cerianthid mesenteries be as I have maintained, it follows that if we regard 

 the single siphonoglyph of the Zoanthids as ventral we must apply the same designation to 

 that of the Cerianthids. In what follows, accordingly, I shall employ that orientation and speak 

 of that surface of the Ceriantheae which is occupied by the protocnemes as ventral and of 

 that occupied by the deuterocnemes as dorsal l ). 



The Arrangement of the Mesenteries. The first definite information which we 

 possess concerning the arrangement of the mesenteries of the Ceriantheae was furnished by 

 Haime (1854), who observed that in C. »icmbra7iacms the mesenteries were not paired and 

 were alternately longer and shorter; further, he also noted the occurrence of a couple which 

 extended the entire length of the body and bounded a groove which he termed "la gouttière 

 interlamellaire impaire". Von Heider (1879) made an important addition to Haime's obser- 

 vations by discovering that short sterile mesenteries regularly alternated with the fertile ones, 

 which were the only ones that Haime had observed, and he further noted the existemce of a 

 couple of short sterile mesenteries, those which we now call the directives, hing between the 

 long ones which bounded the gouttière impaire. As a result of these observations he recognized 

 three grades of mesenteries as occurring in Cerianthus, namely, (1) short sterile mesenteries 

 which he called Filamentsepta, (2) the longer fertile mesenteries which he termed Genitalsepta, 

 and (3) the long couple bounding the gouttière impaire which he termed the "continuirliche 



1) I do not propose to discuss the use of the terms anterior and posterior suggested by van Beneden as indicating more 

 correctly the relationship of the Cerianthea: to the metameric animals. The theory upon which the application of these terms rests seems 

 to me too insecure to warrant so radical a change in our ideas of the tectology of the Anthozoa. 



SIEOGA-EXPEDITIE XV tl. 



