NO. 3558 MALLOPHAGA—ELBEL 45 
this plate has fewer setae in both sexes of B. productus than in cor- 
responding sexes of B. africanus. Each lateral margin of abdominal 
tergites I1I-VI between the spiracle and postspiracular seta has more 
short setae in males and fewer in females of B. productus than in 
corresponding sexes of B. africanus. Abdominal sternite I has 
approximately one-half the number of setae in both sexes of B. 
productus as in corresponding sexes of B. africanus; abdominal ster- 
nite IT has slightly more setae in males and slightly fewer setae in 
females of B. productus than of B. africanus. The male abdominal 
sternite VIII lacks the central T-shaped plate in B. productus which 
is present in B. africanus (fig. 69d). The male genitalia has the in- 
ternal branch connecting the parameres posterior to the endomeres 
pointed medially in B. productus but rounded in B. africanus. The 
female anal fringe has fewer than 58 setae in B. productus but more 
than 58 setae in B. africanus. 
Comparison on different hosts: No morphological differences were 
found between specimens of Bucerophagus productus found on the 
two hosts, Bucorvus abyssinicus and B. leadbeateri, so standard meas- 
urements were tested against the null hypothesis that there were 
no differences in measurements (tables 4-7). 
Terminology and formulae are as follows: 
x1=mean measurement of B. productus specimens on host 1, Bucorvus abyssinicus 
x2—=Mmean measurement of B. productus specimens on host 2, Bucorvus leadbeateri 
D=difference in mean measurements, (X;—X2) or (X,.—X;) 
S(x—x)? Sx?—(Sx)2/N 
“Neva TN S01 Y 
si= variance of measurements of B. productus specimens on host 1, B. abyssinicus 
s;= variance of measurements of B. productus specimens on host 2, B. leadbeateri 
s;_Nna 
si Na 
s?=variance= 
F=the ratio of the larger variance divided by the smaller="! or 
SE=Standard Error of D= 1 ee (+H) 
CL=Confidence Limits for D=D-+ (SE) (t.05) 
Since it is not possible by measurements alone to decide from 
which host specimens came, populations from two hosts are con- 
sidered conspecific. 
Material examined: 39 males and 44 females from fresh and dried 
material collected in the Ethiopian region; from the type host: 4 
males and 2 females from Gula, Uganda, Africa, July 10, 1936, col- 
lected by G. H. E. Hopkins, GHEH; 6 females from Ethiopia, Africa, 
March 1909, BMNH 3673; 1 female from CNHM skin from Africa, 
Jan. 30, 1946, REE; 3 males and 1 female from USNM skins from 
Sirre, Ethiopia, Africa, Feb. 13, 1912, collected by Childs Frick, 
REE; 3 males from USNM skins from Uganda, Africa, January— 
221-522-675 
