"^'tngii!''] PliOCEEDINGS OF THE NATIONAL MUKEUM. 133 



A detailed criticism of the various families whicii have been iirojtoscd 

 is not necessary, since this will be entered into in the descri])tivc por- 

 tion of the report, so lar as certain families of the Hexactinia' are ccm- 

 cerned. Andres ('8,'}) adchid a considerable number of families to those 

 which had previously been lecoj^nized, and the nuijority will, no 

 doubt, stand. liertwig- too has added a number of new families, and 

 at the same time h;is <^iveu an interesting criticism of Audres' classifl- 

 cation ami a comparison of it with his own. Many of the families 

 Audres recognized are more accurately defined, and atteutiou has been 

 called to the criteria ui)on wliich families should be based. One of 

 the most recent classitications, is tliat of Danielssen ('90), wliicli is 

 essentially that of Hertwig ('82), confused, and without the corrections, 

 which Hertwig' ('8S) subse(iu«Mitly introduced. In fact, it must be 

 ackuowledged that Daniclsscn's work is a great disappointment, in 

 that the descriptions ai-e given in such a. manner as to preclude con- 

 tidence in their accuracy, wliile the hgures illnstrating them are bcan- 

 tiful examples of "how not to do it.'' The tribe ..Kgirete, which Dan- 

 ielssen proposes, certainly requires further study before being accepted, 

 and the same remark api)lies to his families Sideractidse, Madoniac- 

 ti<lie, and Andvakida". 



1 shall content myself with stating the families which I believe to 

 be worthy of recognition, making some brief remarks on their linuta- 

 tious, and on certain somewhat doubtful forms. 



I think it convenient to consider the mode of arrangement of the ten- 

 tacles of classificatory im}»ortance, and to recognize two subtribes of the 

 Hexactinia* to which Andres' name maybe applied: Actininai to those 

 to which the tentacles are arrangcnl in cycles, and Stichodactylina; to 

 those in which they are arranged radially. To these two I added ('89) 

 a third, the Deudromelina', whicdi is hardly of equal value, and which it 

 will be better to reduce to the rank of a family. It includes forms which 

 possess dendritic or globular processes or arms projecting from the 

 column wall below the margin, such as are found, for instance, in Le- 

 hrunea, Ophiodisem, and Vlatrix* 



The Thalassianthina* 1 would not, however, adopt as a subtribe, since 

 they differ from the Actinina* only in the comjwund character of their 

 tentacles, and a passage to them is furnished by the member of Andres' 

 subfamily Heteractida'. This family is, however, not altogether natu- 

 ral since the genus RagaviiH must be removed Irom it and referred to the 



muscles in the latter and tlieii- abseuce in the other two groups. The abaence of 

 "addiietor" muscles in the Ceriauthei© is a question of obsewatiou, since B()veri 

 has described and figured theiu ; autl with regard to the presence of <'ct<)denii:d 

 longitudinal muscles in the Hexactinije, van Benedeu has apjiarently oxM-rlooked 

 Hertwig's account ('88) of their occurrence in Cori/nui'tis sp? and ComlliitKirphtts 

 ohtectna. In the jtreseut report I describe their occurrence in Ilalewias pUatas. 



It seems fairly certain that the actinian recently described l>y H. V. Wilson 

 ('90) as HopJophoria coral liiieiis (sic) is identical with the Jlutri.f nlohidifcra originally 

 described by Duchassaing and Michelotti ('60). 



