4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I45 



attached by virtue of a "very complicated form of mimicry. . . ." 

 The great difficulty in accepting this appraisal of the host-parasite 

 situation lies in the fact that the several species of Vidua are not each 

 rigidly restricted to single species of hosts. Of some of these birds 

 our knowledge is still very scanty (or even wanting), but of others, 

 such as V. macroura with i8 recorded kinds of hosts, V. regia with 7, 

 V. chalyheata with 2, and Steganiira with 9, the available data cer- 

 tainly contradict any postulated rigid host specificity. To account for 

 the development of nestling mouth markings similar to those of the 

 host species would necessitate, as Steiner himself outlined, a strictly 

 limited host-parasite specificity, and this we do not find to be the 

 case. It is true that each of the species of viduines does appear to 

 have a single most-favored host, but the percentage of deviates from 

 it is too great to ignore. Thus, of the best known species. Vidua 

 macroura, I was able to assemble data on yy records with 18 species 

 of hosts, and of these more than three-quarters were of 10 species 

 of waxbills of the genus Estrilda and more than half were of the 

 races of a single species, Estrilda astrild. However, the different 

 species of waxbills differ as much in their mouth markings among 

 themselves as do the species of Vidua. If, as Steiner implies, the 

 mimetic similarity of buccal patterns of each species of parasite and 

 its normal host can only be looked upon as having an importantly 

 selective survival value, we would expect a considerably higher ad- 

 herence to the specific host relationship it is supposed to serve. 



It might be considered that there may have been such a rigid host 

 selection originally and that subsequently the parasites broadened 

 their range of fosterers, but this would imply a subsequent denial 

 of an original, and ostensibly a continuing, selective force. In view 

 of the inconstant nature of the differences tabulated in support of 

 familial rank for the waxbills, and in view of the great difficulties 

 such an arrangement would make in interpreting the breeding biology 

 of the widowbirds, I still think it better to keep them all in one 

 .systematic family group. 



It has occurred to me that the above argument may make it seem 

 that the conclusions arrived at may imply something akin to a manipu- 

 lation of classification to simplify or to eliminate what would other- 

 wise be a perplexing problem, rather than to maintain a systematic 

 arrangement based purely on traditional characters, and to let the tan- 

 gential problem continue to perplex us if need be. This is not the 

 case, as the characters advanced by the proponents of familial rank 

 for the waxbills are not constant, on the one hand, and the mouth 



