376 MORPHOLOGY OF THE ABDOMEN IN THE ODONATA 



of solution. It is not yet discovered in what past period the 

 elongated abdomen was acquired, nor is it certainty known 

 whether it first existed as the Zygopterous or the Anisopterous 

 type. Most writers would probably lean toward a decision in' 

 favor of the former, but it is interesting to note that Tillyard in 

 his recent work (1917) figures additions to Brongniart's restora- 

 tion of the Protodonate Meganeura monyi which give an inter- 

 mediate character to the thorax and abdomen (basal portion 

 only). On the basis of wing structure he distinctly places this 

 Carboniferous form with the Anisoptera. Most fossil species 

 are known from wings and fragments only. 



An examination of the second problem is, however, quite within 

 the present limits of researcii. Each of the great lines of Odonate 

 evolution arc today represented by numerous highly perfected 

 genera and species, and to determine the structural differences 

 or similarities between these various groups, and between the 

 larva and the adult in each group is not only possible but of prime 

 necessity; while in the next place an attempt can be made to 

 correlate these structural conditions with the life activities of the 

 insects. The outcome should be a decision as to the adaptive or 

 nonadaptive nature of the variations in the abdomen. This 

 paper concerns itself with these questions. 



MAT-ERIALS STUDIED AND THEIR PREPARATION 



The materials for this stud}- were gathered from several sources 

 and at times covering a period of more than fifteen years. From 

 1900 to 1911 the writer collected specimens in Iowa, and from 1911 

 to 1915 similar work was carried forward in the counties of south- 

 ern Minnesota. Man}' of the larvae used came from these regions. 

 More recently larvae and imagoes were collected in the vicinity 

 of Philadelphia, and were especially prepared according to a 

 variety of methods for dissection. Dr. Calvert, also, kindly 

 furnished material from this region. The specimen of Megalo- 

 prepus coeridatus 9 imago dissected was collected by him at 

 Juan Villas, Costa Rica, in 1909. 



The earlier material from Iowa was for the most part simply 

 dropped into alcohol or placed in papers. It was thus of little 

 use for dissections. This was to some extent also true of the 

 Minnesota specimens, but most of the larvae from there were 



