NO. 1788. NORTH AMERICAN ERGA8ILIDJE— WILSON. 309 



and has been followed by other volumes embracing some of the 

 remaining Copepoda, the publication being not yet finished. He 

 separated the Copepoda into seven divisions : (1 ) The Calanoida, free- 

 living and pelagic; (2) the Harpacticoida, free-living but demersal; 

 (3) the Cyclopoida, partly free-living and fresh-water species, partly 

 commensals and messmates with other animals; (4) the Notodel- 

 phyoida, semiparasitic and living upon ascidians; (5) the Monstril- 

 loida, partly parasitic and partly free; (6) the Caligoida, parasitic 

 upon fishes, moderately degenerate, and with some freedom of motion; 

 (7) the Lernseoida, fish parasites, strongly degenerate, fixed in posi- 

 tion, and with marked sexual dimorphism. 



In this scheme the Ergasilidas are in the third division, along with 

 the Lichomolgidse and Corycseida;, while the Chondracanthidas are in 

 the seventh division. This is a scheme based upon habits and mode 

 of life, though of course substantiated by morphological differences, 

 and is by far the best one yet devised along those lines. It differs 

 from that adopted by Brady in his Monograph of the Free and 

 Semiparasitic Copepoda of the British Islands (1880), only in 

 the arrangement of the Cyclopidas and Notodelphydse, and has 

 been adopted by the distinguished Scottish investigators, T. and 

 A. Scott. 



The two most recent works upon the Crustacea, that in the Cam- 

 bridge Natural History, by Geoffrey Smith, and that in Lankester's 

 Treatise on Zoology, by W. T. Colman, both bearing the date of 

 1909, adopt Giesbrecht's classification. 



These different schemes have been fully discussed elsewhere by the 

 present author." It is sufficient here to state that there are very 

 serious objections to all of them except that proposed by Sars. 



But from a careful comparison of these various schemes of classi- 

 fication we are enabled to draw certain definite conclusions: 



1. The free-swimming copepods have mouth-parts suited for biting 

 and chewing, and are to be grouped by themselves at the head of the 

 classification scheme, as is done by every one of the authors quoted. 



2. The parasitic copepods have mouth-parts suited for piercing or 

 tearing, the mandibles being inclosed in a tube or siphon, and are to 

 be grouped by themselves at the opposite end of the classification 

 scheme, as is also done by every one of the authors quoted. 



3. Between these two groups is a large middle class, including free- 

 swimmers, messmates, commensals, semiparasites, and complete para- 

 sites. In these forms the mouth-parts are in the process of transition, 

 but there is no mouth-tube or siphon. Here belongs the family now 

 under discussion, and nearly every author has placed it in a different 

 position with relation to its immediate neighbors. This is the 

 debatable ground, and the conflict is chiefly due to the fact that the 



a Zool. Anzeiger, vol. 35, no. 20, pp. 609-620. 



