NO. 2194. NORTH AMERICAN PARASITIC COPEPODS—WI LSON. 51 



close together and biramose, third pair removed a little distance and 

 uniramose, fourth pair still farther back and destitute of rami; all 

 the rami present are two-jointed and heavily armed with setae. 



Internal (jcnerlc characters of female. — Esophagus at right angles 

 to the axis of the cephalothorax; stomach without convolutions; 

 intestine passing to the dorsal surface of the trunk; rectum short 

 and inclined toward the dorsal surface; ovaries situated near the 

 anterior end of the trunk, dorsal to the intestine; oviducts straight 

 and flattened laterally; cement glands straight, narrow, and paral- 

 lel, glandular portion shorter than the ducts; chitinogcn layer of 

 body wall especially well developed in the anterior thorax and neck, 

 where the outer layer shows numerous pore canals. 



Genus halitat. — This genus penetrates from the outer surface of 

 the isthmus of the host through the intervening tissues and buries its 

 head Avith the anterior processes in the bulbus arteriosus of the fish's 

 heart. 



Type of the genus. — Cardlodectes medusaeus (Wilson), first 

 species. 



{Cardlodectes, Kaph'ia, the heart, and 577^7175, a biter or torturer.) 



Remarks. — This new genus is established to include the species 

 originally described by the present author as Lernaeenicus medusaeus 

 in 190D, and noted by Brian in 191-2, and a species described by 

 liichiardi in 1SS2 as Perodcrma hellotti and noted by Jungersen in 

 1911. 



For the original description of medusaeus there was but a single 

 specimen available, and it could not be injured. Hence, investiga- 

 tion was confined to what could be seen from the outside without 

 removing the branched processes. In the present instance these 

 processes were removed and the cephalothorax was cleaned of every- 

 thing except the lateral horns. In this way the antennae and mouth 

 parts were laid bare. 



This genus differs from Lernaeenicus in that it has these processes 

 in place of hard chitin horns, it shows no body torsion, the proboscis 

 is not extensile, and there are but three pairs of legs with rami in- 

 stead of four. 



The two species mentioned above do not belong to the genus 

 Perodcrma^ as suggested by Jungersen, first because that genus does 

 not possess the soft and laminate lateral horns, which here appear 

 behind the processes. Again, in Perodcrma the processes are tubu- 

 lar, and their interior communicates with the general body cavity; 

 they also cover the entire head and conceal the antennae and mouth 

 parts. Here they are solid and the swimming legs and antennae are 

 left free. 



In Perodcrma the neck is at right angles to the trunk and opens 

 out of the side of the latter, as in the new genus ColUpravus^ but 



