ECOLOGICAL 149 



the problem has been solved in another way, namely by a relapse 

 into parthenogenesis. (/) It is also possible that parasitism has in 

 some cases evolved from shelter-associations and from commensalism. 

 Indeed, it is here often difficult to draw the line, {g) It is likely that 

 parasitism has often arisen from another form of parasitism! That 

 is to say, the parasite of a freshwater crustacean may become adapted 

 to become secondarily parasitic in a freshwater fish which habitually 

 feeds on the first host. An elaboration of this extension of range 

 comes about when different phases of life are restricted to each of 

 the two hosts, a common punctuation being that an asexual phase 

 occurs in the one host and a sexual phase in the other. Leuckart, 

 who discussed this problem very carefully, in connection with tape- 

 worms, came to the conclusion that the "intermediate host", 

 (e.g. mouse), which now contains the young non-reproductive 

 phase, was the original host, and that the "secondary or definitive 

 host" (e.g. cat) has been found later. Yet it is conceivable that in 

 some cases the intermediate hosts of the immature stages have been 

 intercalated. When there are two hosts, there is usually, though not 

 invariably, this relation between them, that the definitive host 

 eats the intermediate host as part of its normal diet. Thus the cat's 

 tapeworm has its bladderworm stage in the mouse; but it is not by 

 eating water-snails that the sheep is infected with liver-fluke. 

 A sometimes serious parasite of man (Echinococcus) , too common in 

 Iceland and Northern countries, was found by a parasitologist to 

 be derived from the dog as intermediate host. Thus in most cases 

 it can be eliminated by insistence on kennels outside the house, 

 and better scrubbing and sweeping out of dust from within. Perhaps 

 more than Icelanders would profit by knowing these facts and apply- 

 ing them. 



DAMAGE DONE TO HOST.— It is usual to emphasise the fact that 

 it is not in the parasite's interest to destroy its host; but this is 

 verging on the anthropomorphic. Parasites cannot be credited with 

 a policy. In many cases, e.g. follicle mites in man, or Monocystis 

 in earthworms, the parasite is small compared with its host, and the 

 damage done may be unimportant. Yet a heavy infection with 

 threadworms may be fatal to a horse; and if ichneumon-grubs are 

 regarded as parasites (an interpretation above criticised) they are 

 obviously fatal to their caterpillar hosts. Much often depends on the 

 numerical strength of the infection; thus a few gall-larvse, each 

 imprisoned in its gall, may be regarded as trivial, but the multiple 

 infection of a black-currant bush with "big bud" mites may soon 

 be fatal. Then a distinction must be drawn between parasites that 

 multiply in their host, as Nematode worms often do, and those, 

 like tapeworms, that cannot increase in numbers within the same 

 animal. It may be that rapidly destructive parasites have been 



