The Actiiiiae of the Plate Collectiou. 221 



cation of the mesenteries, a remarkable modification of the tentacles 

 at their bases and a teiitacular sphincter, and in Ftychodadis pafida 

 with the reduction of the stomatodaenm and with the multiplication 

 of the mesenteries, all of which characters must, I believe, be re- 

 garded as progressive, those associated with the tentacles highly so. 

 Again it will probably be admitted that the stichodactj^line arrange- 

 ment of the tentacles is a progressive character, and it is associated 

 with the presence of a considerable number of mesenteries, also a 

 progressive character. To my mind, the association of forms possessing 

 such characters with others which lack them is comparable, if I may 

 refer again to the Mollusca for an Illustration, to an association of 

 Argonaufa with Haliotis. 



And not only does the acceptance of Carlgren's primary division 

 lead to an association of forms presenting widely diiferent grades of 

 progressive differentiation, but it also separates forms of approxi- 

 mately the same grade. Thus to separate by a great and unpassable 

 gulf Gonadinia from Omdis is unjustiliable. Having regard to the 

 progressive differentiation these forms stand more closel}" related 

 than do Gonadinia and Boloccroides and the same is true for Gon- 

 adinia and the Edwardsias, even although the latter present pro- 

 gressive modifications in the absence of the pedal disc and the 

 development of strong muscle pennons on the mesenteries, tliis last 

 character, as Appellöf (1893) has suggested, possibly accounting in 

 part at least for the absence of the ectodermal column musculature. 



The subdivision of the Actininae into Athenaria and Thenaria 

 also, it seems to me, tends to the confusion of unrelated forms and 

 the Separation of others which are nearly related. The subdivision 

 is based upon a persistent character, it is true, and by associating 

 with this the absence or presence of a basilar muscle it seems to 

 have weight. But the relations between the pedal disc and the 

 muscle are so intimate that modifications of the one can hardly be 

 expected without modification of the other. Furthermore when Ave 

 speak of the absence of the pedal disc we are speaking rather of 

 Its modification than its absence in many cases, since its homologue 

 exists in the physa, and in such cases a homologue of the basilar muscle 

 may be seen in the parietal muscle, so well marked in Edwardsias 

 and Halcampids. 



Caelgren's Athenaria corresponds essentially with Gosse's family 

 Uyanthiäae and associates such forms as Echvardsia and HaJcampa with 

 Uijantlms parfhenopeus with its twelve pairs of perfect mesenteries, 



