The Actiniae of the Plate CoUectiou. 219 



suggestecl in an earlier paper a possibilitj' of tlie extension of the 

 Order Protactiniae to include the forms here referred to the family 

 Edwardsiidae, but a further consideration of the matter has led me 

 to abandon that idea and in fact to abolish the order Protactiniae 

 altogether. 



Without entering- into the history of the various classifications 

 which have been proposed for the simpler Actiniaria, it seems ne- 

 cessary to explain my reasons for not adopting the arrangement re- 

 cently proposed by Caelgren (1899, 1900). He has divided tlie 

 Actiniaria into two tribes the Protantheae and the Nj^iantheae, char- 

 acterized respectively by the presence or absence of longitudinal 

 muscle fibres in the ectoderm of the column wall and stomatodaeum, 

 and eacli of these tribes he divides again into two subtribes, the 

 Protantheae into the Protactininae and Protostichodactylinae and the 

 Nynantheae into theActininae andStichodactylinae, theActininaefinally, 

 being again subdivided into the Athenaria and the Thenaria accord- 

 ing as they lack or possess a distinct pedal disc. 



The fnndamental point in this Classification is the presence or 

 absence of an ectodermal longitudinal musculature in the column wall 

 and stomatodaeum and the Classification will stand only if this char- 

 acter can be recognized as strictly phylogenetic. I do not imagine 

 that anyoue will deny that the ancestors of the Anthozoa probably 

 possessed such a musculature, nor, indeed, that the epithelial cells 

 of all the Coelentera are potentially epithelio-muscular cells, and 

 hence it must be admitted that the occurrence of an ectodermal 

 musculature in the column wall is, in a certain sense, to be regarded 

 as an ancestral character. But even so it does not necessarily 

 follow that it is a character of prime classificatory value; it is only 

 so when it is associated with other characteristics which we believe 

 to be also primitive and, on the other hand, is not associated with 

 peculiarities which must be regarded as highly progressive ditferen- 

 tiations. The structural characteristics of any group of animals are 

 in part persistent and in part progressive characteristics. 

 The former are ancestral in their nature and serve to distinguish 

 the group from others; and while sonie of these characteristics may 

 be lacking in this or that adult member of the group, yet they must 

 be regarded as included in the embryonic potentialities of all and 

 may reappear in forms which really represent a higher degree of 

 specialization than others which lack them. The progressive char- 

 acteristics, however, are departures along new lines from the an- 



