4 BULLETIN 2 08, UNITED STATES NATIONAL MUSEUM 



southern Asia as to make a Miocene continuum of the two areas 

 probable. Among the genera common to both regions, Lonnberg 

 mentions Indicator. The fact that no honey-guides are known to 

 occur in Madagascar further suggests that the actual spread of the 

 group (whether from Asia to Africa or the other way round) may- 

 have taken place during the Pliocene, as at that time Madagascar 

 was completely separated from the African mainland. In fact, 

 Lonnberg reports that Pliocene adjuncts to the African fauna are 

 generally absent from Madagascar, even though they may be widely 

 distributed in Africa and in southern Asia. Lonnberg further con- 

 cludes that the present forest fauna of Africa consists partly of 

 descendants of the original endemic Miocene fauna and partly of 

 later additions which have become adapted to forest living, and that, 

 similarly, the open bushveld and grasslands of Africa have a fauna 

 that contains a good sprinkling of forms that were originally sylvan 

 in their habitat. It is among these forms that we may consider several 

 honey-guides to belong: Indicator indicator, I. minor (except for its 

 western, sylvan race, conirostris) , I. variegatus, and Prodotiscus 

 regulus. In this conclusion I am anticipated by Chapin (1932, p. 378), 

 who suggests that the family was perhaps originally a sylvan group, 

 although some of its present species are now found only in the open 

 country. Of the eleven species comprising the family, six are known 

 only in forested areas, three only in more open situations, and two 

 have races in both these types of terrain. The forest dwellers are 

 Indicator maculatus, I. archipelagicus (largely?), /. xanthonotus 

 (largely?), Melignomon zenkeri, Melichneutes robustus, and Prodotiscus 

 insignis; the species of the bushveld and grasslands are Indicator 

 indicator, I. variegatus, and Prodotiscus regulus, while Indicator minor 

 and 1. exilis have races in deep forest and also in bushveld. Indicator 

 variegatus, while not a bird of extensive evergreen forests, is somewhat 

 more sylvan than /. indicator in many parts of its range. Prodotiscus 

 insignis, in its southern race, zambesiae, is out of the true forest, and in 

 Indicator minor we have a similar ecologically intermediate race, 

 pallidus, apparently confined to the broken forest at the edge of the 

 true forest in Nigeria. 



Wliile there is fairly general agreement that the honey-guides have 

 their closest affinities with the barbets (Capitonidae), the toucans 

 (Ramphastidae), the puff birds (Bucconidae), and the woodpeckers 

 (Picidae), it should be admitted that there is some uncertainty as to 

 the degree of closeness they show towards each of these groups. Years 

 ago Garrod, Gadow, and Beddard, among others, considered the Indi- 

 catoridae only a subfamily of the Capitonidae. Ridgway (1914, p. 

 310) kept the two as separate families, but considered them as forming 

 the superfamily Capitones. More recently, Wetmore (1951, p. 20) 



