THK SKKTt'h. \KID.i:. 41 



upon (lie characters of thr margin and operculum in classifying the Sertularida'. and that is tliat 

 these characters arc inconstant, not only in sonic of the genera, but also in .some individual 

 species. 1 have already pointed out the fact that certain species of .s, ,-f i'lm-< lln lack the sup- 

 posed three or four Mapped o]ierculum. In Selagmopsis mirabilw (VeTriO) there are two Haps 

 to the operculum, while the one-flapped operculum ia characteristic of the genus as a \\hole. I 

 do not believe that any one would separate ,V. /////<//;/'/ /.v and .V. i-ijli m! rii-<i (('lark) genericallv. and 

 yet they ditl'cr in this feature upon which Levinsen bases his genera. In X, rtnl, //-,',/ <l< xniuiil, a 

 Torrcy, a for.in found on the ( 'alii'ornian coast (Plate III, fig. 1), the hydrothecal mare-ins are 

 sometimes without teeth and at others show two small teeth. The operculum is usually of a 

 single adcauline flap, but sometimes, in other parts of the same colony, is composed of two parts. 

 In this case neither the insulin nor the operculum furnishes a constant feature, even in a single 

 colony. In Abietmaria greenei (Murray) the hydrothec-e on one part of a colony will lie of the 

 typical abietinarian form, while those on another part of the same colony will have two very 

 conspicuous teeth, lioth of which are adcauline and not strictly lateral (Plate XXXVI. 

 lie's. 3 and 4). The operculum is of a single adcauline flap. Probably enough illustrations have 

 been given to show that the characters used by Professor Le\ insen are insuflicient in themselves 

 to furnish a basis for the classification of the Sertulariihv. 



It by no means follows, however, that the operculum and hydrothecal margin are characters 

 to be neglected. On the contrary, I think them most important aids in defining certain genera. 

 such as Alii' ti miriii and I >!/>liuni,i, and feel that we owe much to Professor Levinsen for his pains- 

 taking work calling general attention to these features. Careful and conscientious work such as 

 his is always valuable, whether the results are in all respects confirmed or not. 



The only remaining author whose scheme of classification we need discuss at present is l>r. 

 Karl Camillo Schneider, who published a work of interest in this connection. 1 l>r. Schneider 

 (p. r>^l) was at first much impressed with Levinsen's method of classification, but decided that a 

 review of the whole group revealed the inadequacy of the plan, and also many intergrading 

 forms. On the whole, this writer prefers the older classification of the Sertularida'. and adopts 

 the following genera, but calls them "groups" on account of their incomplete separation: X/Y//- 

 la/retta^ Dynamena^ Tlnitiiriii. Pasythea, Selaginopsis, and HydraHmania. 1 



Lest it may appear that the writer has intentionally or carelessly neglected to include the works 

 of American writers in the summary just given, attention is called to the fact that there has been 

 no general work, nor any general discussion of the family Sertularida' produced by an American 

 writer since the appearance of the classic work by the elder Agassi/, in ixt'.i'.-" 1 h, fr, the appearance 

 of Ilincks's British Hydroid Zoophytes, Istis. which I have taken as my point of departure in tin- 

 preceding discussion. Agassi/ proposed three new genera of Sertularida 1 (pp. :i.V> :-!.">(!): 

 .(nijiJiixliitiii. < 'ft iiliini. and Aiit]>li!tr<H-Jiit, which were not adequately defined, and which have not 

 been adopted by later writers, except that two of them are used by his son. l>r. Alexander 

 Agassi/. 4 



In attempting to break up the family Serf ularida- into genera, there are several principles 

 that should be clearly grasped at the outset: 



/"/AY. No one character, nor combination of two characters can be successfully used 

 throughout, as is illustrated by Levinsen's attempt based on the characters of the hydrothecal 

 margin and operculum. 



Xii-iiml. It sometimes happens that a single character will sharply differentiate a single 

 genus. For example, the unilateral arrangement of the hydrothecie in Ili/ifi-iillimniiit. 



Tlii ><!. The hydroids are an extremely plastic group, and certain characters may occur 

 sporadically in many unrelated species that occur normally and regularly in certain close! \ 

 related forms. This fact has been the cause of great confusion in the systematic treatment of 



1 lly<lrc>icl|ioly|ii M yon K"\L'no, m-l.-i I Ylx-rsiclit iilT ilas Systrtii ,ld- II\-<lniiil|iiilypeii im \IL"-iMfini-ii. 

 ogische Jahrbiicher, Systematik, \, pp. 47i > -"i">."i, .ICHM. ls:is 



Altliniijili Doctor Schneider calls these "^rmips " he t rents them :is genera, fur convenience ill lianilliiiu' 

 8 Contributions to the Natural Ilistm-y \ the I'nitc.l States of America. I \", I'liisldii. ISIiL'. 

 1 North American Aralepli;e, ISti.'i, pp. 14U-147. 



