36 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I46 



Also, Clamator levaillantii has been known to lay in the hole nest of a 

 kakelaar, Phoeniculus purpureus. The tendency to utilize such nesting 

 sites has certainly not developed in jacohinus to the extent it has 

 in glandarhts, but these instances show that it is not outside the range 

 of possibilities even in the former. 



One further thought emanating from a consideration of this prob- 

 lem of host selection may be added here. Recent studies on many 

 species of self-breeding birds, particularly in North America and 

 Europe, have shown that nest-site selection is fairly rigid and fixed 

 in its major elements. Slight vegetational differences often are 

 critical to various species in the precise location of their nest sites. 

 So widespread is this tendency that it is only proper to apply it to a 

 review of the situation in brood parasites as well. In these birds 

 nest-site selection would be altered to host selection based on the 

 types of nest-sites used by the latter, and would be expressed in terms 

 of host specificity as far as the parasites are concerned. On the whole, 

 avian brood parasites fall into three main categories in this respect. 

 Some exhibit little or no such specificity ; others are specific in their 

 host choice as individuals only ("individual -host specificity," as found 

 in the European cuckoo, Cuculus canorus) ; and in still other parasites 

 the entire species is specific on one or a small group of related hosts 

 ("species-host specificity"). 



Inasmuch as nest-site selection does reflect trenchant and remark- 

 ably uniform criteria in each species of self -breeding birds (potential 

 hosts), and inasmuch as there is no reason to assume that parasitic 

 species, especially early in their evolutionary history, were necessarily 

 different from self-breeding birds in their response to familiar and 

 uniform environmental details, it may be that species-host specificity, 

 as opposed to individual-host specificity, was the original situation in 

 brood parasites and that a broader range of host choice developed 

 from it later. This would imply that the broad spectrum of hosts 

 subsequently arrived at may have evolved even as a negation of the 

 original selection pressure that operated in an earlier atmosphere of 

 species-host specificity. From this concept it would follow that rigid, 

 if not actually "obligate," parasite-host specialization is a basic rather 

 than an ultimate condition. This is quite the reverse of the often 

 assumed pattern of species-host specificity arising from individual- 

 host specificity. 



In this connection it may be pointed out that the one species of 

 brood parasite whose descent from a self -breeding form is most 

 obvious and clear, the screaming cowbird, Molothrus riifoaxillaris, 



