Chap. I. SUB-ORDERS OF CTENOPHOR^. 179 



a distinct pattern of form determined by the even thickness of the spherosome, 

 or, more accurately, by the even arrangement of the motory cells of the sphero- 

 meres. I hold, therefore, that the Beroids proper constitute a distinct sub-order, 

 and not merely a fomily nor a distinct order, since their distinguishing characters, 

 though of the kind of those upon which orders may lie founded, do not extend to 

 their whole organization. We shall see presently whether these Acalephs constitute 

 only one single familj^, or not. 



It would l)e tedious, were I to analyze at the same length the value of all 

 the characters upon which are founded the other subdivisions of the Ctenophora' 

 proposed by different naturalists. Suffice it for me to say, the division proposed 

 by Leuckart under the name of Stenostomata is neither an order, as he takes it, 

 nor a natural sub-order, — not simply because the characters which he assigns to 

 these Ctenophora) are not of the kind upon which orders may be distinguished, 

 but because it is in itself a heterogeneous assemblage of different types. The 

 groups designated by Eschscholtz as two distinct families, and the corresponding 

 groups proposed by Gegenbaur, under which he imites respectively the families of 

 the Callianirida3 and Calymnida), and the Cestida3 and Cydippida^ approach much 

 nearer to the idea of sub-orders. At all events, they include the representatives 

 of two distinct sub-orders ; but, from want of personal acquaintance with the Calli- 

 aniridaj and Cestid*, I am unable to state Avith certainty how closely these may be 

 allied to the types of the two sul>orders which I have investigated. It may also 

 be noticed, that Eschscholtz and Gegenbaur disagree as to the affinities of Cestum 

 and Callianira, — Eschscholtz uniting both with Cydippe, while Gegenbaur separates 

 them as types of distinct families. So, leaving for the present the question of the 

 closer affinities of the genera Cestum and Callianira out of consideration, I am 

 prepared to .show that Eschscholtz's Mnemiida^ and his Callianirada^ do not constitute 

 two natural families, l)ut two distinct sub-orders equivalent to that of the Beroids 

 proper. On account of the questionable affinities of the genera Calhanira and 

 Cestum, I am, however, compelled for the present to designate one of these sub- 

 orders by a different name from that applied to it by Eschscholtz, and shall adopt 

 for it that of Cydippida), instead of Callianirida*. 



The anatomical peculiarities of the Cydippidce, which show them to be a sub- 

 order, and not merely a family, correspond exactly to those which distinguish the 

 Beroidn?. The eight ambulacral tubes are equal or nearly so, as in Beroida*, but 

 they terminate, on l)oth sides of the main axis, at a considerable distance from 

 the poles of the spherosome, and do not open into an oral tube ; for there is no 

 such circular cliymiferous tube around the actinostome in this group of Ctenophora?. 

 The coeliac tubes are very large, and terminate as blind sacks upon the sides of 

 the actinostome. Neither the ambulacral nor the coeliac tubes give out conspicuous 



