No. 3655 PARALLEL EVOLUTION—FRIEDMANN 5 
last three examples, it became obvious that there were indeed two 
very similar species living side by side in eastern equatorial Africa. 
This caused me to assemble as complete a series as possible in order 
to review the situation again and to seek new conclusions. 
Although very similar to each other, narokensis and meliphilus 
should be treated as specifically distinct. In narokensis, the wing 
length varies from 65 to 70 mm in four females, 67 and 69 mm in 
two males; in meliphilus, 18 females have wing lengths from 69 to 
78.5 mm, 18 males, from 73.5 to 85.5 mm. The bill is ‘“stubbier’’ 
to the eye in narokensis, more so than actual measurements reveal; 
length from distal end of nostril to tip of culmen is 4.2 to 4.8 mm 
(4.7 to 5.9 mm in meliphilus), the exposed culmen is 5.7 to 7 mm 
(7.1 to 8.5 mm in meliphilus). Actually, the bills of narokensis are 
very similar to those of pumilio. This supports Chapin’s comment, 
cited above, that pumilio may be related not merely, as he put it, 
to meliphilus, but even more closely to that part of meliphilus (olim) 
now separated as narokensis. 
In my 1958 paper I explored the question of the status of the 
western race of meliphilus, which Monard (1934, p. 55) had des- 
ignated angolensis as a subspecies of minor! I still believe that 
there are two recognizable races of meliphilus, although they are 
based on average differences and not on more trenchant characters; 
angolensis is usually paler gray, especially on the throat, breast, 
and upper abdomen than nominate meliphilus. Chapin (1954, p. 633; 
1962, p. 44) considered angolensis indistinguishable from meliphilus, 
but he also put narokensis in this category. More recently, Traylor 
(1963, p. 108) has supported the belief that angolensis is the south- 
western race of meliphilus, and I may add that the additional speci- 
mens I have seen since my 1958 paper also have supported this 
division. The removal of narokensis from meliphilus increases the 
average duskiness of the residual nominate meliphilus since the 
palest members of that taxon, as formerly treated, are all narokensis. 
The narokensis-pumilio Relationship 
The relationship of narokensis and pumilio parallels very closely 
that of meliphilus and exilis and that of minor and conirostris. Inas- 
much as meliphilus and exilis have been found to overlap sympatri- 
cally in the Mwinilunga district, Zambia, and in the Upemba Park, 
in the Congo (Verheyen, 1953, pp. 406-407), their specific distinctness 
has to be granted. The situation between minor and conirostris, 
however does not include this convenient sympatry, and they usually 
are still considered to be races of one species. While conirostris is 
a dark-colored forest-inhabiting form with a discontinuous range 
from western Kenya (Kakamega and Kapenguria) west to southern 
282-405—68——2 
