above. Vanhöffen's observations (1895) were made, however, upon advanced larvae, and his 

 conclusions were based on the relative lengths of the different mesenteries in these. Little 

 importance can therefore be assigned to them, for the principle upon which they are based is 

 manifestly erroneous. Van Beneden (1891), however, foliowed the development of the mesenteries 

 in larvae of different ages, and, while arriving at conclusions regarding their embryonic sequence 

 identical with my own, denied that the protocnemes can be regarded as homologous with those 

 of the Hexactiniae. In proof of this position he makes a comparison between the first six 

 couples of mesenteries in the two groups, and, designating them in their order, from the ventral 

 directives dorsally, I — II — III — IV — V + VI, points out that the order of development in the 

 Actiniaceae is III— V—I— VI— II— IV, while in the Ceriantheae it is II— UI— I— IV— V— VI. 

 I have already (1893) pointed out the fallacy of such a comparison. It proves nothing to the 

 point, since it is a comparison between the protocnemes plus two couples of deuterocnemes in 

 each group, and it is the very basis of the view propounded by Boveri and myself that the 

 two groups are primarily distinguished by the mode of appearance of the deuterocnemes. If 

 these be omitted and comparison be made between the protocnemes alone then the sequence 

 is the same in both groups, namely, III — I — II — IV. Van Beneden, indeed, acknowledges this 

 identity of succession for the protocnemes, but still maintains that those of the Ceriantharia 

 cannot be regarded as identical with the so-called "Edwardsia mesenteries" of the Actiniaceae, 

 since they do not possess the adductor muscles so well developed in those mesenteries, and 

 the Ceriantharia furthermore possess a well developed ectodermal musculature which is lacking 

 in the Actiniaceae. Surely these two points must be regarded as of little moment compared 

 with the order of sequence of the mesenteries, and it must be remembered that Boveri found 

 indications of adductor muscles, identical in their arrangement with those on the Hexactinian 

 protocnemes, in the Cerianthid larvas he studied. As to the ectodermal musculature being an 

 important characteristic it is only necessary to refer to the rapidly increasing number of Acti- 

 niaceae in which it has been found, our present information on the subject seeming to indicate 

 that it must be regarded as a primitive Anthozoan character, which has been lost in the 

 majority of the Actiniaceae and retained in the Ceriantharia. 



In his later paper (1898), rich in so many important discoveries, van Beneden adheres 

 to the position taken seven years before and, while recognizing the existence of two distinct 

 phases of development in both the Actiniaceae and the Ceriantharia, declines to regard the 

 first stage as identical in the two groups. In the Actiniaceae he regards the first stage as 

 terminating with the formation of the sixth couple of mesenteries, while that of the Ceriantharia 

 terminates with the formation of the third couple; and for this first stage of the Actiniaceae 

 he suggests the name Halcampula, while for that of the Ceriantheae he employs the name 

 Cerimila. In making this delimitation of stages it is to be noted that the criterion chosen is 

 quite different in the two groups. In the Actiniaceae it is the cessation of the appearance of 

 the mesenteries in couples, while in the Ceriantharia it is an alteration in the order of sequence 

 of the couples. 



With the typical sequence of the Hexactinian protocnemes, under which term I would 

 include only the first four couples, so well determined, it is a little difficult to understand why 



7 



