80 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. 39. 



the hind wings; and (7) the posterior composite mass of the thorax, 

 consisting of the metapostnotum (PKi) and the first abdominal 

 tergum (IT) above, of the metapleura (PI3) on the sides, and of the 

 metasternum below, carrying the first abdominal spiracles (ISp), 

 the pedunculate part of the abdomen (77, 777) and the hind legs 

 (CX3), and supporting the hind wings on the metapleural wing 

 processes ( WP^). 



The wingless Hymenoptera have the thorax the most highly 

 specialized and, at the same time, the most simplified. In the 

 apterous forms of Mutillidae and the workers of Formicidse (62) the 

 propectus is detached in the usual fashion from the rest of the thorax 

 and the protergum is separated from the mesothorax, at least on the 

 sides, by a cleavage suture, but otherwise the thoracic walls are 

 sohd. The back sclerites (62, Sct^, Scl^, N^, PN^ and IT) form 

 one continuous plate from the protergum (N^) to the second abdom- 

 inal segment (77). The indistinct line (k) across the back appears 

 to be the scuto-scutellar suture. The lateral margins of the dorsal 

 plates, are indistinguishably fused with the pleurites and these latter 

 are continuous with the sterna. The mesopleurum is partially 

 divided by an impressed line (r) into an upper plate (PZj) continuous 

 with the mesoscutum (ScQ and into a lower plate (^^^2) carrying 

 the middle coxa (CiCj). This sort of subdivision of the mesopleurum 

 has been pointed out in other forms (59). The mesepisternum alone 

 is frequently so divided (Pepsis, 61, EpS2, eps^). The first spiracle of 

 the worker ant is situated as usual behind the angle of the protergum 

 (62, ISp), the second (2Sp) is inclosed in the posterior margin of the 

 upper mesopleural plate (PZj). The first abdominal spiracle (ISp) is 

 located on the side of the propodeal region (IT). 



Other authors have made different interpretations of the morph- 

 ology of the ant thorax. Janet (1898) calls the proepisternum (62, 

 EpSj^) the " presternum " and the lower parts of the pleura of the 

 other two segments the "mesosternum" and the "metasternum". 

 Nassanoff (1889), Emery (1900), and Berlese (1908) name these parts 

 in the same way. The writer has already stated the argument against 

 such a disposition of these plates (see pp. 77 and 78). The 

 pleurum of any segment lies normally between the base of the wing 

 and the base of the leg, and it is inconceivable why a line midway 

 between these points should be regarded as the sterno-pleural suture. 

 At least some strong reason should be given for imagining such a 

 distortion to have taken place that would put it there. In the lower 

 Hymenoptera, as already shown, the true sterno-pleural sutures lie 

 ventrad to the articulations of the coxse (10, 11, 14, 16, 27, g). In 

 most of the higher forms these sutures disappear though they recur 

 in many scattered cases (40, 43, 50, q). Therefore, the line on the 

 middle of the side can be nothing else than a secondary subdivision 

 of the pleurum itself. 



