BY R. J. TILLYARD. 383 



mined), proves the existence of two larval types, somewhat 

 parallel in their lines of development with those of Macromina 

 and Si/tUhcmiiia. The larva X belongs to an unknown species, 

 whose venation is exceedinyly close to that of Syncordulia (until I 

 have actually succeeded in breeding it out, I cannot say definitely 

 whether the species will go into Syncordulia or not). In many 

 respects (though certainly not in appearance), it resembles the 

 larva of Synthemis. Thus, it is a short thick-set larva, with 

 very short legs, is very inert and fond of concealing itself; it can 

 withstand starvation and drought almost as well as the larva of 

 Synthemis, and, in its labial development, it shows irregular and 

 deep incisions. Contrast this with the well-known Hemicordulia- 

 type of larva (Needham* has described the larvae of Epicot'dulia, 

 Tetrayoneitria, Somatochlora, Ilelocordulia, Cordulia, and Doro- 

 cordulia, all of this type). 



There we have a spider-like larva, with large head, broadly 

 oval body, and very long legs; one that relies for protection on 

 its colour-pattern, and lives uncovered on the pond or river- 

 bottom; apparently with no power to endure either starvation, cr 

 even a short drought; and with a labial development oi the 

 LibeUnline type, characterised by exceedingly shallow crenations, 

 often armed with sets of tiny spines, and with the lateral lobes 

 often spotted with warts and dots. In general form and habits, 

 these larvae resemble the Macromian larvte, though the latter still 

 retain the deeply indented labium characteristic of the original 

 Corduline stock. 



1 propose, therefore, to subdivide the group Cordidina into two 

 groups, which I am at present inclined to regard as coordinate in 

 value with the Macromina and Synthemina. But, as these two 

 groups include a larger number of forms, and those more widely 

 distributed over the earth, we must be prepared to find inter- 

 mediate genera connecting the two main lines of development. 

 Suc-h genera would probably be difiieult to place in any linear 

 classification. Also, further knowledge of life-histories may prove 



* New York State Museum. Bulletin 47, September, 1901. " Aquatic 

 Insects in the Adirondacks."' 

 33 



