302 ILLINOIS BIOLOGICAL MONOGRAPHS [302 



How can the infection of so many species of fish with this parasite 

 be explained? For the sake of the argument it is granted that all 

 these species are suitable hosts for this parasite. If all these species are 

 to be infected it is necessary that some of the invertebrates included 

 among plankton forms, bottom fauna, and molluscs must be able to serve 

 as intermediate hosts of this parasite. The fish which become infected 

 from eating these invertebrate hosts may likewise become self-infected 

 and so serve as intermediate hosts. If such fish are then eaten by the 

 predaceous fish of the list then the latter may become infected. Thus 

 it is possible to offer explanations that will cover the whole question of 

 the source of infection for these hosts which have such different food 

 habits. The analysis, however, is not very convincing. The more plau- 

 sible explanation is that several species of parasite are being dealt with 

 under this one name. It seems not improbable that the parasites of 

 the cyprinids may belong to at least two or probably three species while 

 the jion-cyprinids may or may not harbor any of these. Further inves- 

 tigations on the parasites of the hosts of this list should be undertaken 

 with the view of settling some of these questions. 



CONCLUSIONS 



A careful analysis of the extensive data presented in this section 

 shows that species of Proteocephalus may occur in multiple hosts in the 

 following combinations: 



1) A species of Proteocephalus may occur in different host species 

 of the same genus. Five species are limited exclusively to various species 

 within the same host genus. 



,2) A species may occur in the different genera of the same family. 

 One case. 



3) A species may occur in the members of closely allied genera, 

 i. e. of the same order. Four cases are known. 



4) A species may occur in families of very wide relationships, i. e. 

 of different orders. There are two cases of which one is doubtful. 



It is further shown in this section that when a species of Proteo- 

 cephalus occurs in multiple hosts under any of the combinations above 

 suggested the host species have a continuous distribution. The only 

 apparent exception to this rule is in the case of AnguUla vulgaris of 

 Europe and A. chrysypa of North America, both harboring P. macro- 

 cephalus. In this case, however, the exception is apparent and not real 

 for these two forms are perhaps no more than varieties of a cosmopoli- 

 tan species. Analysis of the data presented in this section also shows 

 that in many cases of multiple hosts the food of the different host spe- 



