1903] WHEELER:— CERAPACHYS AUGUSTAE 209 



8. The eggs are shaped like those of certain Ponerinae (Leptogenys, Pachy- 

 condyla). 



9. The larva probably spins a cocoon. 



On the other hand C. augustae exhibits unmistakable Doryline characters : — 



1. In the conformation of the head and antennae which closely resemble 

 those of Eciton. 



2. In the habit of brooding over the eggs. 



The following characters are common to both Dorylinae and Ponerinae : — 



1. The method of carrying the larvae is common to forms like Eciton and 

 Leptogenys. 



2. The larva is intermediate between that of Eciton and Stigmatomma. It is 

 covered with shorter, less flexuous, and less abundant hairs than the latter and in 

 these particulars resembles the larvae of Eciton.^ 



These facts go to show that C. augustae is a generalized form much like 

 the hypothetical ancestor from which both Dorylinae and Ponerinae are supposed 

 to have sprung. At the same time, the majority of the characters are Ponerine 

 and justify us in adopting the views of Forel who places the tribe Cerapachyi with 

 the Ponerinae. Other species of Cerapachys bear out this interpretation. Dr. 

 Hans Brauns writes me from Willowmore, Cape Colony : " Cerapachys ist doch 

 wohl sicher eine Poneride. C.peringueyi ist bei Port Elizabeth keineswegs selten, 

 hier einzeln. Ihr ganzes Betragen ist das einer Poneride." 



Owing to the close relationship of Cerapachys with both Ponerinae and 

 Dorylinae, the recent discovery by Emery of a form (Aneuretus ) intermediate 

 between the Ponerinae and the Dolichoderinae, and the patent relationship 

 between the Ponerinae on the one hand and the Myrmicinae and Camponotinae on 

 the other, it is evident that the Formicidae constitute a very compact group of 

 Hymenoptera. This unitary character of the group is still further emphasized by 

 its comparatively recent geological origin. Hence it should be designated as a 

 family, and its five divisions as subfamilies, in accordance with the views of Euro- 

 pean myrmecologists like Emery, Forel, Mayr, Wasmann, etc.; and the recent 

 tendency of some Cisatlantic hymenopterists to regard the ants as a superfamily 

 (" Formicoidea "), consisting of the five families Myrmicidae, Poneridae, etc., is not 

 to be commended merely for the sake of making the nomenclature of this group 

 look like that of some of the other divisions of the Hymenoptera. 



American Museum of Natural History, 

 Oct. loth, 1903. 



1 1 have not seen Emery's description of the larva of the Ponerine Ectatomma, but I have been able to examine the 

 larvae of E. {Holcoponera ) strigatum Norton from Guadalajara, Mexico. These larvae are nontuberculate like those of 

 Stigniatoimna pallipes (Biol, bull., 1900, vol. 2, p. 61, fig. S) but covered with long, multifurcate, somewhat flexuous hairs. 

 An exhaustive monographic study of ant-larvae would certainly repay the investigator, as they present a bewildering 

 array of interesting characters in the various tubercles, " poils d'accrochage, " etc., with which they are provided. 



