NO. 3655 PARALLEL EVOLUTION—FRIEDMANN 7 
ussheri, and pallidus, some of which are such intermediates as to be 
doubtfully recognizable as racial entities. Similarly, pumilio has 
been found to be geographically variable. Its data are much less 
extensive, being based merely on the fact that its easternmost example 
(Kakamega Forest, western Kenya) is different from Kivu topotypes, 
a difference that was sufficient to cause Williams and Friedmann 
(1965) to raise the possibility of the Kakamega bird being racially 
separable from the eastern Congo population. We found the single 
Kakamega specimen to be slightly larger (wing 70 mm), to have 
more uniformly grayish underparts without the streaks that are 
present in the birds from the Kivu and the Impenetrable Forest, 
southwestern Uganda, and to have the top and sides of the head 
less greenish, more grayish. If additional material from Kakamega 
should agree in these color characters, which are in the direction 
of a more grayish and paler shade—or, to an extremely slight, almost 
incipient degree, in the direction of the very much paler and grayer 
narokensis—there would be a situation faintly resembling that ex- 
isting in conirostris and riggenbachi in Upper Guinea and other races 
of minor in the open country north of the forest belt. Kakamega, 
where pumilio occurs, and Sigor, West Pokot, where narokensis 
has been taken, are barely 75 miles apart. 
Enough has been said to warrant raising the question whether or 
not pumilio is a true species or an unusually distinct race of narokensis. 
To be wholly consistent, I would have to treat the two as conspecific 
forms by the same standards that I use when considering conirostris 
and minor in this matter. The real question is whether they are repre- 
sentative races or whether they have differentiated to the point where 
they are specifically distinct. Unfortunately, this question could 
be answered only by bringing them into contact under natural con- 
ditions, which cannot be done. I, therefore, favor considering the 
two as races of a single species merely to point out their mutually 
representative nature and to suggest their closer relationship in the 
not too distant past. That they are closely allied is intimated by anal- 
ogy with minor and conirostris, and while this is certainly no proof, 
it may yet reveal a pattern or trend in these small species of Indicator. 
Without pressing the point beyond its realistic implications, I 
add the suggestion that maculatus (with stictithorar as a race) is no 
more than a west African forest representative of variegatus of eastern 
and southern Africa. It is a comparable case. 
Morphological Overlap and Sympatry in Sibling Species 
While I conclude that narokensis is a species distinct from meliphilus, 
I realize that other observers may have the experience of finding 
specimens that are difficult to place in either species. It would be 
