303] PROTEOCEPHALIDAE—LA RUE 303 



cies is alike in some or all of its elements tho the proportions of these 

 elements may differ widely. Since the infection of the host is only 

 accomplished through the food eaten a general statement may be made 

 thus: The 'parasitic infestation of the host is determined hy the food 

 eaten. The character of the food is determined by the environment 

 and largely by the habitat preferred by the host. The apparent excep- 

 tions to the above conclusion are found in those cases where a group of 

 species acting as hosts to a single species of parasite feed on very 

 different elements of food. In each of these cases the probability has 

 been pointed out that there has been a confusion of species under one 

 name. If such a confusion does not exist then there is evidence of the 

 remarkable adaptability of the larval stages of the parasite to the life 

 conditions of a large series of invertebrates and fish. Both of these 

 possibilities should be investigated. 



The evidence as presented in this section is not in favor of a strict 

 adherence to the idea of the specificity of parasites. This idea cer- 

 tainly does not hold for any of the better known species of Proteoceph- 

 alus for which quite a number of collection records are available, but 

 it seems to hold for a number of species for which there are but one or 

 two collection records and for species which are imperfectly known. 

 To apply the idea to these cases is not justified. To break away en- 

 tirely, however, from the idea of specificity of parasites would be a 

 rash step, indeed, for certain cases have been well established which 

 show conclusively that for certain hosts, other than those considered in 

 this paper, certain species of parasites are specific. Such cases can 

 only be determined by many records of collection and then are to be 

 relied upon only after careful infection experiments have been worked 

 out. To apply the idea of specificity of parasites without having taken 

 these precautions is not justifiable. In working through the data on 

 distribution of Proteocephalus species which have multiple hosts one 

 notes that in many cases the range of the parasite may be greater than 

 the range of some of its hosts. This is well illustrated by the range of 

 P. longicollis which has a far wider distribution than some of the spe- 

 cies of Coregonus in which it occurs. Some of these species occur in 

 just a few lakes while P. longicollis is known from many localities of 

 northern and central Europe. So also the range of P. macrocephalus 

 is greater than that of either of its hosts. 



An explanation for this wide distribution of the parasite may be 

 stated thus: The parasite may have a better means of dispersal than 

 the host. The distribution of fish Proteocephalids may be widened in 

 the following ways: They are carried by the primary host which 

 voiding feces as it wanders about scatters the eggs of the parasite that 



