OF THE FAMILY COCCIDAE 145 



while not very close to these, share between themselves a large number of characters 

 (almost 90). The three remaining genera, i.e. Ericerns, Genus A and Sphaerole- 

 canium are not very close to each other or to any of the other above-mentioned 

 subgroups (although Ericems comes close to Rhodococcus). It is therefore clear 

 that the EULECANIUM group, in which 11 out of 19 genera studied have been 

 included, is very heterogeneous and as more material becomes available it will 

 undoubtedly be further subdivided. 



B. From Tables 2 and 2 A it can be seen that the ERIOPELTJS and EULE- 

 CANIUM groups agree in a large number of characters (14), of which 6 are common 

 to them alone (i.e. exclusive). Conversely in the COCCUS and INGLISIA groups 

 the alternative condition of these 6 characters obtains and they share 15 characters 

 amongst themselves. Any other way of coupling pairs of groups results in a 

 comparatively small number of characters being shared or exclusive. It would 

 appear therefore that there are two major groups, ERIOPELTIS and EULE- 

 CANIUM on the one hand and INGLISIA and COCCUS on the other, which can 

 be separated mainly by the absence or presence of setae in various regions of the 

 body, as follows : 



INGLISIA and COCCUS groups ERIOPELTJS and EULECANIUM 



groups 



1. Fleshy dorsal head setae present. Fleshy dorsal bead setae absent. 



2. Setae between and behind ventral Setae between and behind ventral 

 eyes present. eyes absent. 



3. Genal setae present. Genal setae absent. 



4. Fleshy postmesospiracular setae Fleshy postmesospiracular setae 



present. absent. 



5. Antemetaspiracular setae present. Vntemetaspiracular setae absent. 



6. Fleshy dorsospiracular setae present. Fleshy dorsospiracular setae absent 



The fact that these two groups can only be separated by differences in the 

 chaetotaxy indicates that this family consists of a morphologically rather homo- 

 geneous group of species. This is particularly true in comparison with the 

 Diaspididae where Ghauri (1962) found sharp and very distinct morphological 

 differences between major groups. 



At present no rank is suggested for these groups of genera indicated by male 

 characters, but they may be considered to be equivalent to each other and repre- 

 senting subfamilies, while the further subdivisions in the EULECANIUM group, 

 and possibly the COCCUS group could represent tribes. 



The diagnostic characters of each group can be found in the detailed key which 

 follows later. 



C. The classification proposed above unfortunately differs considerably from 

 the existing classification suggested by Borchsenius (1957) and a few examples are 

 discussed as an illustration of these differences. 



