No. 637] 



PARASITISM 



147 



iological matter and is always regarded as such by hel- 

 mintliologists who encounter many instances of this kind. 

 On account of the definite requirements of such parasites, 

 Cobb ('04) suggested some years ago that they might 

 give valuable clues to the taxonomic affinities and physi- 

 ological peculiarities of their hosts, the latter particu- 

 larly in cases where there is a wide range of hosts. 



In insects, and, quite fortunately for the present dis- 

 cussion, in the parasitic Hj-menoptera, there are avail- 

 able some extreniely pertinent observations made by 

 Timberlake ('12) relating to the fate of eggs in the 

 bodies of host insects in which they do not normally 

 develop. His experiments were made with an Ichneu- 

 monid, Linnerinm validum, commonly parasitic in cater- 

 pillars of the fall web-worm. This parasite will also ovi- 

 posit in larvffi of various other moths, when persuaded to 

 do so in captivity, by depriving it of its normal host ; but 

 it can not complete its development in the experimental 

 hosts. This is due to the death of the young larvae, which 

 succumb to the reactions of the host soon after hatching, 

 or possibly in some cases even before hatching. The an- 

 tagonistic action of the tissues of the host is manifest by 

 the accumulation of amcebocytes about the unwelcome 

 objects. In one other abnormal host, the tent-caterpil- 

 lar, this Linnerimn may survive and complete its trans- 

 formations, but there is a high mortality among the par- 

 asites, for many are destroyed by the host. 



These experiments show very clearly why this parasite 

 is restricted to certain hosts and, from the nature of the 

 reaction, which is so similar to that exhibited by animals 

 in general toward microorganisms and other foreign ma- 

 terials, there is little reason to doubt that insects usually 

 react in this fashion. This also furnishes an explanation 

 for the continued restriction of parasites to specific hosts, 

 based upon natural selection, since individuals choosing 

 unsuitable hosts will suffer a very material reduction in 

 the number of their immediate progeny. This is, I think, 

 especially important, as it takes much of the burden from 



