NO. 1687. THE THORAX OF INSECTS— SXODGRASS. 535 



The Ephemerid mesotergiim and metatergum are sufficiently 

 shown in the drawings of Ilexagenia hilineata (4, 3). Each presents 

 simply a confusion of elevations and depressions showing little sim- 

 ilarity to any other order. A posterior median lobe, however, sug- 

 gests the similarly situated lobe in Odonata (17), Acridiida? (53), 

 and Perlida? (75), and also the median part of the scutellum of Lepi- 

 doptera (150), of Hymenoptera (161), and of Diptera (175). It is 

 also comparable with the median shield-shaped area of the meso- 

 thorax of Coleoptera (127, 128). A pseudonotum (3, PN) is present 

 in each segment, but is hidden from above by a posterior membranous 

 fold margined by the axillar}'^ cords (4, Ax C). 



This review of the notal structure in the different orders will show 

 that the diagrammatic conception of the notum illustrated in fig. 2 

 is really the fundamental notal structure that prevails in all the 

 principal orders except the Ephemerida and the Odonata. It is evi- 

 dent that the three ventral ridges — the anterior, the V-shaped, and the 

 posterior — are constant characters and that the regions they mark oflE 

 can be regarded as homologous in all the orders. But such subdivi- 

 sions are not the ones usually apparent on the surface, and the latter, 

 though generally the same within an order, vary so much in different 

 orders that they can not be regarded as homologous structures except 

 within limited series. 



Three factors contribute to the formation of the notal subdivisions 

 as follows: (1) Topography, as elevations and depressions of the sur- 

 face forming more or less distinct regions, some of which appear 

 variously modified throughout nearly the entire insect series; (2) ven- 

 tral ridges, three of which are constant characters and, hence, to be 

 regarded as homologous in the different orders; and (3) depressed 

 suture-like lines and actual sutures, variable and not the same in dif- 

 ferent orders and, hence, not necessarily defining homologous sclerites. 



Therefore, it may be concluded from a study of development and 

 comparative anatomy that any scheme of thoracic structure in insects 

 is untenable which postulates four primitive transverse plates in the 

 tergum. Much less is there any evidence that the definitive tergum 

 is composed of the united terga of two or four primitive metameres. 



This conclusion is opposed to that of Berlese (1906), who finds in 

 the thoracic terga of all orders four exactly corresponding parts, 

 the " acrotergite," the " protergite," the " mesotergite," and the 

 " metatergite." An examination of his colored diagrams (1906, Plate 

 4) will show, however, that in order to carry out his scheme Ber- 

 lese has in many cases drawn purely arbitrary lines across the 

 notum. Moreover, he has disposed of the pseudonotum by making 

 it the " acrotergite " of the segment following the one to which it is 

 attached. Thus he equates the mesopseudonotum and its phragma 

 in Sphinx with the metaprephragma of Acridium. The pseudonotum 



