82 Transactions of the Boyal Society of South Africa. 
If we remember, too, that most of the descriptions are made from 
mounts which are merely boiled in potash, and cleared without staining, 
the difficulty of observing the comparative lengths of the segments will be 
obvious. 
Under these conditions the dermis is colourless, and becomes very 
transparent in Canada balsam, and the exact point at which the segments 
begin, and terminate, is difficult to determine. 
Antennal formulae, constructed from stained specimens, however, are 
quite unsatisfactory for the determination of species. That this is the 
case may be best illustrated by a few measurements with the formulae 
constructed from them. 
The following are made from specimens of P. adoniclum Westw., 
collected at the Edgbaston Botanical Gardens in November, 1913. 
Antennal segments : — 
I 
II III 
IV 
V 
VI 
VII 
VIII 
65 
72 76 
38 
42 
38 
46 
104 
Formula : — E 
J 3 2 1 7 5 (6 4) 
58 
65 63 
36 
41 
38 
43 
96 
Formula : — i 
^2317564 
74 
72 74 
48 
62 
43 
43 
101 
Formula :— 8 (1 3) 2 5 4 (6 7) 
It is not an uncommon occurrence for the two antennae of the same 
insect to give different formulae, and from one collection of material, 
about 40 specimens, I constructed the given formulae of five described 
species. The measurements were made from specimens stained by the 
carbol fuchsin method, with the Zeiss microscope, obj. D, and 7*5 
measuring ocular. The results are more accurate by this method than 
by plotting from camera lucida drawings. The lengths of the antennal 
segments are, nevertheless, of great importance in the determination of 
species if they are accurately made from stained specimens, and properly 
tabulated. 
The most useful arrangement of antennal data seems to be arrived at 
by giving the range of variation in measurements of the different segments, 
with the addition perhaps of the mode of each. After working over a 
large series of slides one is impressed with the characteristic appearance 
of the different antennae. But this difference is difficult to express. The 
nearest approach is obtained by a scheme such as is used in the charts 
given with the different species, and this supplies a most useful aid for the 
preliminary location of an insect from slide specimens. 
Wherever possible the range of measurement represents at least 
10 measurements, 20 to 30 being made where material is sufficiently 
