222 MR. J. W. GEEGOBT ON THE 



Three groups of the Cheilostomata may be con\eniently based on the character of 

 the front wall. In one, including the Membraniporidan series, this structure is absent 

 or only imperfectly developed: the name of " Athyriata" (from a and Svpeoe, an oblong 

 shield) is therefore suggested for it. In a second group the wall is well developed, but 

 there is an additional communication between the exterior and the polypide by means 

 of a pore (trypa) on the front wall or by a sinus on the lower margin of the orifice. 

 The exact homology of these two structures has never, so far as I am aware, been 

 clearly demonstrated, but it has been generally accepted, for example, by Hincks, 

 Waters, and Macgillivray. For this group the name of " Schizothyriata " is proposed. 

 Finally, there is the group in which the calcification of the front wall is complete ; it 

 may therefore be called the " Holothyriata." 



In addition to these there is a series of forms whose afl[inities seen very doubtful. 

 AVith one or two exceptions they are rarely or never found fossil, and my opportunities 

 of studying them have been but limited. They may be divided into two divisions, one 

 of which may be a natural group. This includes the CEtiidse, Chlidoniidae, and 

 Eucratiidse ; the terminal or subterminal apertures and simple tubular or pyriform zooecia 

 of these families suggest that tliey are among the most primitive of living Cheilostomata. 

 They are here left grouped together, and Busk's name, the Stolonata, is accepted. For 

 the other division Smitt's name of " Cellularina" is adopted ; but this is certainly not a 

 natural group. Thus some, such as the Cellulariidse, Bicellariidse, and Epistomiidae 

 (Notamiidse of Hincks), seem clearly allied by their large membranous areas and aperture 

 to the Membraniporidan group ; the Catenariidse may include representatives of both 

 the Holothyriata (e. g. CafeniceUa utriculus, Macgill.) and the Schizothyriata. Among 

 the latter there may be divisions corresponding to both of the great families ; thus 

 Catenicella amphora. Busk, is analogous to the Microporellidse, and C.pulchella (Maple- 

 stone) to the Schizoporellidae. It is, however, not improbable that the Catenicellidse 

 branched off independently from the main Cheilostomatous stem at a very early 

 period. 



Without more detailed information upon the anatomical structure of the polypides 

 of the families in this " carpet-bag " group it seems unadvisable to attempt to place 

 them definitely. In the Catenariidse we have both holostomatous and schizostomatous 

 (e. g. Claviporella) genera, but until we know more of the anatomy of the polypides it 

 seems very uncertain as to whether this character possesses the same significance as in 

 those higher Cheilostomata where the skeleton is of a specialized and complex type. 

 Amongst these the hard parts certainly seem to offer reliable classificatory characters. 



Through each of the three suborders an evolutionary series can be traced. Thus 

 among the Athyriata the Membranoporidse seem to be the most primitive, and this 

 family passes up into the Cribrilinidse and Hiantoporidae in the manner suggested by 

 Mr. Hincks [No. 2, pp. 199-200, and No. 5 [pt. 3], pp. 471-472 and 479-480] and 

 Mr. Kirkpatrick [No. 2, pp. 616-617]. 



