According to the older writers, the habit of the colony was taken as the most important generic character; and there can indeed be no doubt that this feature has great importance within certain limits. Any one who has examined different species of such genera as Flustra, Cellaria, Bugula, Retepora, etc., must feel that the form of the colony goes for a good deal. But a consideration of other cases shows that there is great risk in the indiscriminate use of this method of arranging the Polyzoa. The old genus Eschara, composed of forms with an erect coral-like habit,[[584]] included species which are now placed in such different genera as Lepralia, Porella, Microporella, etc. The older works on Polyzoa include all encrusting forms of Cheilostomata, with a completely calcareous front wall, in the genus Lepralia, the members of which are now distributed in numerous widely separated genera.
As an instance of the converse arrangement—essential similarity of the zooecia with great differences of the general habit—may be mentioned the common Membranipora (Electra) pilosa.[[585]] Ordinarily growing in the form of close encrustations on seaweeds, this species may take on entirely different habits of growth. The zooecia are now dissociated, growing in single lines over the substratum; now forming erect tufts, composed of single lines of zooecia or of several rows. The erect, branching habit appears to be induced in the first instance by the character of the seaweed on which the colony begins life. Thus colonies which encrust the thin branches of Corallina may have impressed on them something of the mode of growth of the seaweed, so that when they extend beyond the tips of the branches of the Corallina, they continue to grow in delicate branches, which still retain more or less the same diameter as those which form their base. An extreme variation results in the beautiful form known as Electra verticillata, in which the zooecia are arranged with great regularity in whorls, which together form erect branches.[[586]] But with all these variations, the zooecia are so much alike that it is hardly possible to regard the extreme forms as more than varieties of a single species. A careful examination of this case would convince most observers that the characters of the zooecium are a more trustworthy guide to classification than those of the entire colony, a result which was first clearly stated by Smitt, and amply confirmed by Hincks.[[587]]
The avicularia of the Cheilostomata afford useful help in classifying this group; but while certain genera are always provided with avicularia, others include some species with these organs, and other species without them. Again, while the species of some genera (e.g. Cellepora) possess a great variety of forms of avicularia, the same pattern of avicularium may characterise several widely different genera. Further, the position of the avicularium may be very different in species which are apparently closely related. Well-developed vibracula, although constant in their occurrence in such forms as Scrupocellaria (Fig. 254) and Caberea (Fig. 242), occur here and there in species of encrusting forms which are ordinarily placed in very different families.
Now although some of these discrepancies are perhaps due to errors in classification, whereby species which are really allied have been wrongly placed in distinct genera, this explanation would not prove satisfactory in all cases. Thus in Bugula, a genus which is specially characterised by the high development of its avicularia, these organs are normally absent in B. neritina. The fact that this species was rightly placed in the genus has been confirmed by the discovery made by Waters[[588]] that avicularia occur in specimens which are believed to be identical with that species.
Fig. 254.—A, Front view, and B, back view of part of a branch of Scrupocellaria scabra, Van Ben., Durham Coast, × 43; a, lateral avicularium; a', smaller median avicularium; ap, membranous aperture; f, fornix; r, rootlet; s, seta of vibraculum; v.z, vibracular zooecium.
1. The Cyclostomata appear to fall naturally into two main groups, (A) the Articulata, including the Crisiidae (Fig. 237), distinguished by their erect branches, divided at intervals by chitinous joints; and (B) the Inarticulata, which include the remaining families, whether erect or encrusting, agreeing in the negative character of being unjointed.
2. The Cheilostomata consist of (A) the Cellularina, including the flexible, erect forms, such as Bugula (Fig. 233) and Scrupocellaria (Fig. 254); (B) the Flustrina, to which belong Flustra (Fig. 232), Membranipora (Fig. 256, A, B), Micropora (Fig. 256, C), and other forms in which the front wall of the zooecium is either membranous, or depressed and marked off by a ridge-like margin; (C) the Escharina, including the great majority of forms, in which no part of the front wall remains membranous, the wall of the zooecium being wholly calcified.
3. The Ctenostomata comprise (A) the Alcyonellea or encrusting forms; and (B) the Vesicularina or branching forms. The zooecia in the latter subdivision (Fig. 238) are given off from a tubular stem or stolon, which is usually erect and branching.
We thus have the following arrangement of recent forms. The genera mentioned are for the most part those which have already been alluded to in the preceding account:—