Annals of the Transvaal Museum 
217 
3. Tail very deeply forked, for more than an inch: Caffrapus gen. nov., type 
Cypselus caffer Lcht. 
Tail slightly forked, for about -5 to -8 inch: Epicypselus gen. nov., type 
C. horus Heugl. 
Tail not forked, or for less than -25 inch: Colletoptera gen. nov., type 
C. affinis Gray. 
With regard to species, C. Grant (Ihis, 1915, p. 315) records the 
occurrence of Micropus murinus hrehmorum Hartert (Naum., Vog. Naturg. 
Deutschl. IV. 233, 1901) in Damaraland, and states that M. kalaharicus 
Reichenow {Orn. Monatsb. 1908, p. 81) is a synonym of M. apus pekinensis. 
M. aequatorialis (v. Mull.) has been recorded from Gazaland by Swyn- 
nerton (Ihis, 1908, p. 393). There is a specimen of Colletoptera horus 
(Heuglin) in the Transvaal Museum collection from Beira, taken by P. A. 
Sheppard, and I have found it breeding at Koster, in April, 1922. 
Mathews (B. Australia, vii. 264, 1918) has created three new genera 
for the African Spinetailed Swifts, namely Telecanthura for Chaetura 
ussheri, Neafrapus for Chaetura cassini and Alter apus for Chaetura sabinei. 
Two species of “Chaetura” have been recorded from within our limits, 
namely, C. stictilaema (presumably referable to Telecanthura) recorded by 
Alexander, from the Zambesi (Ihis, 1900, p. 93) and C. anchietae Souza 
by C. Grant (Bull. Brit. Orn. Cl. xxi. 66, 1908 and also Ihis, 1911, p. 698). 
Specimens in the Transvaal Museum collection recorded by Gunning and 
Haagner in their Check List (1910) as Chaetura hohmi Schal., following 
Reichenow, who placed anchietae in the synonymy of hohmi, evidently 
refer to the same species as that recorded by Grant. I find, however, 
that these specimens are not referable to either hohmi or anchietae on the 
length of the wing, and possibly on the colour. The descriptions of both 
hohmi and anchietae are not too clear. Souza and Bocage both distinguish 
anchietae from hohmi by the absence of black shaft stripes to the feathers 
of the throat and the black instead of white lores. In our series of speci- 
mens, eight from Beira and three from Machile River, the base of the 
feathers of the lores is pure, the external half black, partly hiding the 
white base. Reichenow compares hohmi with cassini, not making it clear 
that the under tailcoverts in cassini are black. The specimens from Beira 
and Machile River have the under tailcoverts white, the upper tail- 
coverts, black, but with the middle ones white at the base like the rump. 
In regard to dimensions, in the original description of C. hohmi (Orn. 
Centralh. 1882, p. 183) and in subsequent repetitions, the tail length is 
given as 60 mm., which must be wrong, as the total length is given as only 
90 mm. Reichenow gives the tail length as 30 mm., and in anchietae it is 
said to be about 27 to 28 mm. In our specimens, the eight from Beira give 
a wing length of 116-120 mm., tail 20-22 mm., and the three from Machile 
River, wing 118-123, tail 20-22. As regards generic allocation of these 
birds, Mathews has diagnosed them on the difference in size, the colour 
of the tailcoverts and the proportion of the tail to the wing. In Neafrapus 
the under tailcoverts are black, size large and the tail about one-fifth of 
the length of the wing and the outermost primary is as long as the second. 
In the Beira and Machile River specimens the middle of the breast, 
al^domen and under tailcoverts white, size small and the outermost 
primary decidedly shorter than the second. Having regard to these 
