1924 ] On the Ant-genus Chrysapace Crawley 225 
postpetiole recurs in the East Indian Cerapachys antennatus F. 
Smith, which happens to be the type of the genus. 
Emery seems to have regarded the non-clavate antennae 
and margin ation of the petiole as more important characters 
than the number of antennal joints since he made Phyracaces 
an independent genus and separated Cerapachys into four sub- 
genera on this character, namely Cerapachys sens, str. with 12, 
Parasyscia with 11 , Oocercea with 10 and Syscia with 9 joints. 
But the finding of an intermediate form like Chrysapace brings 
us face to face with a dilemma. Either we must raise all the 
subgenera mentioned to generic rank and retain Chrysapace and 
Phyracaces as independent genera or we must reduce these two 
genera to subgeneric rank under Cerapachys. The ‘ ^splitters’ ^ 
will probably adopt the former, the ‘dumpers” the latter alter- 
native. Should the lumpers carry the day the specific name of 
Crawley’s species will have to be changed, because Forel had 
previously described a Cerapachys jacobsoni from Java (Notes 
Leyden Mus. 34, 1912, p. 103). In that case I suggest that the 
Sumatran ant be called Cerapachys {Chrysapace) crawleyi nom. 
nov. It is, however, not improbable that we shall do more 
splitting in the Cerapachyinse in the near future. The subfamily 
is proving to be more extensive than we had supposed. Mr. 
James Clark and I have recently brought to light quite a number 
of species of Eusphinctus and Phyracaces in Australia and there 
are several diverse, undescribed species of Cerapachys and Phy- 
racaces from the East Indies in my collection. The sexual phases, 
larvae and pupae of the great majority of Cerapachyinae are quite 
unknown. A knowledge of these phases and stages will probably 
aid materially in a final revision of the genera and subgenera of 
the subfamily. 
