176 CTENOPHOR^, Part IT. 



to con.sidor the method, and not the vahie, of the characters assigned to tlie groups 

 themselves. Ina.smuch as Clegenbaur has introduced the element of form in the 

 characteristics of his families, instead of alluding solely to the features of their 

 structural complication, as Eschscholtz had done, he has made a decided advance 

 over the classification of the latter. This is the more aj^parent when it is re- 

 membered that these structural complications are not at all overlooked by him ; 

 but, on the contrar}^, are made use of in grouping the families under three more 

 comprehensive heads, to which, however, he has assigned no names, but which 

 correspond very closely to the groups called families by Eschscholtz. It is proper, 

 therefore, that we should inquire into the meaning of these groups, and at the 

 same time remember also that Leuckart likewise admits divisions among the Cte- 

 nophora) superior to the families, which he calls orders, and of Avhicli he admits 

 two, — the EuKYSTOMATA and Stenostomata, the former corresponding to Eschscholtz's 

 family Beroidtv, the latter to Eschscholtz's united families Callianirida^ and Mne- 

 miida^ ; admitting, further, that the (Jtenophora3 are a class and not an order. We 

 have thus, as subdivisions of Ctenophorre, three families distinguished by Eschscholtz, 

 corresponding to three higher divisions, iiKludiruj fire families admitted by Gegenbaur, 

 and iii'o orders admitted by Leuckart. 



With these facts before us it cannot be difficult to untie the knot of these 

 conflicting views, only leaving, for the present, the question of the natural limits 

 of all these groups out of consideration.^ Leuckart and Gegenbaur have evidently 

 both felt that the natural families of the Ctenophoraj are linked together l)y features 

 of a more comprehensive value than family characters ; but, placing only a sub- 

 ordinate importance in questions of classification, one of them has given no names 

 to the groups based upon those features, while the other has called them orders. 

 If, as I have urged again and again, families are Ijased upon peculiar patterns 

 of form, and orders, natural groups founded upon the degree of complication 

 of the structure, the characters assigned by Gegenbaur and Leuckart to these 

 divisions are truly of the kind ui)on Avhich orders are founded ; Imt, the char- 

 acters of orders resting upon the sum of structural complications whicli determine 

 their relative standing in the class, it is plain that special points in that compli- 

 cation which do not extend to all the memljers of the order can only lead to 

 the recognition of secondary natural groups sharing the characteristics of orders, 

 but not themselves orders, and for which I have proposed the name of suh-orders. 



^ In ;ill iliscussions like the present, a iierf'eet there is an end to every critical imiuiry into the 



familiarily with llie objects themselves is a iiccc's- importance and relative vahie of the characters 



sary requirement ; for if the natural features of these assigned to the divisions adopted by ditfcrent authors, 



objects are first to be ascertained during the dis- and the meaning and rank of the divisions them- 



cussion and by the study of a given classification, selves. 



