618 



d. The unsatisfactory nature of this final attempt by Giesbrecht, 

 one of the ablest investigators of the pelagic copepods, following upon 

 the failures of such men as Gerstaecker, Glaus and Thorell, seems 

 to warrant the conclusion that it will be impossible to find anyone basis of 

 primary division in the group that will include all the forms, free- 

 swimmers, semiparasites and parasites. It thus furnishes a very strong 

 negative reason for accepting the scheme proposed by S ars of making 

 the first division in the group upon the same basis as obtains in the class 

 Crustacea as a whole. 



S ars selects certain types, distinguished from one another not by 

 any single character, nor even by two or three combined, but by an 

 aggregation of all the diagnostic characters, morphological, ontogenetical 

 and ecological. What has such a primary division to still further 

 recommend it? 



e. Around each of the types thus selected can be gathered its near 

 relatives, be they free-swimmers or parasites, possessed of a gnathostome 

 or a sucking mouth, and without reference to where in their anatomy 

 the boundary may fall between the fore and the hind part of the body. 

 There will thus be formed a thoroughly natural and homogeneous sub- 

 group, and all necessity of separating close relatives will be obviated. 

 This is but the natural outcome of a tendency that has been increasingly 

 manifested with the growth of our knowledge respecting the copepods, 

 particularly the semi-parasites and parasites 



The genus Ratania was found to possess a gnathostome, and yet 

 it was classed with the typically siphonostomatous family, Asterocheridae. 

 Similarly the close relationship of theErgasilidae and Chondracanthidae, 

 which are true fish parasites, to the semiparasitic Corycaeidae and 

 Lichomolgidae has been steadily gaining acceptance. Such an affinity 

 can be openly expressed in S ars' system, together with many others 

 pertinently suggested by the phylogenetic table published by Ger- 

 staecker (1881, p. 714). 



f. S ars' scheme will include every known species of Copepod from 

 the very beginning, and there will no longer be the embarrasment of a 

 large number of "remaining" or undetermined families. If occasion 

 warrants, a single family may constitute a subgroup by itself until it 

 can be more definitely located, or until new forms are discovered which 

 may be grouped with it. 



g. Furthermore if it becomes necessary, as our knowledge advances, 

 to divide any of the subgroups or to introduce newly discovered types, 

 this can be easily done without the necessity of wholly destroying or 

 rearranging the previous classification. It possesses in this respect 



