The Coccidae of South Africa. 
83 
plentiful to allow such to be done. The measurements are then plotted in 
the centre of each antennal column and the curve completed. Thus it 
appears that where any segment is constant the curve approaches n 
simple line, and a wide band represents a wide range of variation. 
Similarity of antennal curve, whilst indicating similarity of antennal 
formula, does not of necessity indicate identity of species, but it does give 
a clue to work upon ; and, possibly at times, indicates relationship. 
Other characters — the legs, setae of anal lobes and anal ring, and dermal 
■characteristics — can then be compared ; if the similarity still persists, the 
-description can be referred to for other details. 
It should be mentioned, while dealing with the antennae, that the 
measurement of the first segment is often unsatisfactory. This is due to 
the fact that the outer edge is much narrower than the inner, the side 
being hollowed out to allow lateral bending and the insertion of segment II. 
An intermediate length is indicated in the chart given, but slight variation 
in segment I may be disregarded. 
The hairs and spines on the antennal segments are similar in the 
majority of cases, but a few striking variations occur. 
In Bhizoecus spp. there are always 8 or 4 stout falciform spines on the 
terminal segment as illustrated in Fig. lOcl, and in a few examples of 
Pseudococcus there are stouter spines which seem to indicate an inter- 
mediate stage between ordinary spines and the falciform type of Bhizoemts. 
(g) Legs. — The remarks concerning the uselessness of an antennal 
formula as a specific character apply equally well to general remarks on 
the legs, which are commonly made in descriptions of species of Pseudo- 
-coccini. In order to be of use, the measurements should be carefully 
made and tabulated, and use should be made of the three pairs of the same 
specimen. If only one set of measurements is to be given I would 
suggest that the mesothoracic leg be made use of, as this is usually a more 
or less reliable clue to the other two pairs. 
The plan adopted by the writer (1912), gives the measurements in ju, in 
the following order : — 
1. Length of coxa. 
2. Breadth of coxa across base. 
3. Length of trochanter plus femur. 
4. Breadth of femur. 
5. Length of tibia. 
6. Breadth of tibia. 
7. Length of tarsus plus claw. 
This was the outcome of an endeavour to obtain measurements in direct 
lines from points which remain fixed although the legs are folded in 
different ways in mounting, and is generally used in the present paper. 
