146 



THE AMERICAN NATURALIST 



[Vol. LV 



Subfamilies 



Evaniin^ 

 Ichneutinse 



Cockroaches and their oothecsB. 

 Saw-fly larvae. 



Genera 



Polygnotus 



Cecidomiid larvae. 



Coccophagus Soft scales. 



If we should reverse the order of the above list and 

 attempt to tabulate groups of related hosts that are af- 

 fected only by certain groups of parasites we should have 

 great difficulty in finding examples. This, of course, is 

 to be expected on account of the passive condition of the 

 host and the active role of the parasite, whereby it first 

 came to infest some certain kind of host. Inheritance of 

 such specific instincts over long periods of time, during 

 which groups were becoming differentiated, will lead 

 naturally to the evolution of groups of parasites attached 

 to groups of hosts which have meanwhile been developed. 

 Such reasoning appears to be sound and may explain 

 some of the conditions tabulated above. 



I think, however, that there is a deeper basis than this, 

 and that we can not fully understand such combinations 

 without inquiring into the actual physiological relations 

 between host and parasite. 



It has been customary among entomologists to place 

 great emphasis upon the maternal instinct of invariable 

 selection as determining and restricting the range of 

 hosts affected by specific parasites. Among zoologists 

 who deal with other parasites, particularly Protozoa and 

 lower invertebrates, no such idea has ever been enter- 

 tained, as the parasite plays a passive role in attaining 

 its host. The malarial parasite is ingested by all insects 

 that suck human blood, but is able to continue its para- 

 sitic life only in certain particular mosquitoes. Similarly, 

 a certain Cestode worm parasitic in birds has as interme- 

 diate host, the garden slug, from which the definitive host 

 obtains it by eating the slug. That this Cestode does not 

 occur in other hosts that may eat infected slugs is a phys- 



