SOME INTERESTING DRAGON-FLY NAIADS FROM TEXAS. 



By Clarence Hamilton Kennedy. 



Of Ohio State University, Columbus. 



On October 20, 1920, the writer received from the United States 

 National Museum a small collection of Gomphine naiads for identifi- 

 cation. These were collected by J. D. Mitchell and labeled, "From 

 Colleto Greek and sloughs in creek bottom near Victoria, Texas, 

 July 2, 1920." This is in southeastern Texas in the valley of the 

 Guadaloupe River. The collection, while small, is of interest, first, 

 because all Gomphine records from" this southwestern boundary of 

 the Austro-riparian faunal region are interesting, and, secondly, 

 because it contains at least one new form. It is unfortunate that 

 these were not reared to make their identity positive. All are quite 

 immature, having been collected in July; they would have emerged 

 in the spring and early summer of 1921. 



No data are readily available on the nature of the stream in which 

 they were caught, but the evidence is that it is very sandy. 

 This is suspected because the collection includes Progomphus obscurus, 

 which is found only in the sandiest streams, and because none of the 

 naiads show traces of mud in their hairy covering, as Gomphine 

 naiads from sandy streams are spotlessly clean, while those from 

 mud bottom are always filthy. 



Because our knowledge of Gomphine Odonata is so limited the 

 writer has been to pains to illustrate these fully. 



PROGOMGHUS OBSCURUS (Rambur). 



Plate 115, figs. 1-7. 



The material consists of 35 naiads, ranging in length from 7 mm. 

 to 28 mm. The largest two, 25 and 28 mm. in length, would un- 

 doubtedly have emerged during the season of 1920, as the Progomphus 

 season is a long one. Whether the numerous smaller examples would 

 have emerged the same season or one or two seasons later it is impossi- 

 ble to say, as so little is known concerning the life history of Progom- 

 phus. Burrowing insects in general are slow feeders and consequently 

 slow growers. It is impossible to give a table of measurements 

 because about 20 of the specimens are broken. 



Proceedings U. S. National Museum, Vol. 59-No. 2390. 



595 



