596 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.59. 



The small specimens are the smallest Progomphus naiads the writer 

 has seen, and are interesting in that they show by the way the mature 

 naiads differ from them, the direction in which the latter are special- 

 zed. Figure 1 represents the mature naiad. The specimens 7 mm. 

 ong differ from the mature naiad as follows : 



1. The wing pads are minute triangular flaps. 



2. The dorsal spines on segments 2 and 3 are lacking. (See fig. 7). 



3. The superior pair of anal appendages are mere tubercles. (See 

 fig. 7). 



4. The fourth segment of the antenna is only one-third as long 

 as the third. (In the mature naiad it is nearly one-half as long as 

 segment 3). (See figs. 2 and 4.) 



5. Labium with two triangular teeth on the middle lobe besides 

 the usual row of truncate teeth. (In the adult there are not notice- 

 able triangular teeth). (See fig. 6). 



6. Lateral labial lobe with several broad though shallow teeth. 

 (See fig. 5.) 



It can be legitimately inferred that these ontogenetic changes 

 point out the phylogeny of this most specialized of burrowing Odon- 

 ate naiads. They must have developed through a Gomphus-like 

 form that had the characters of these very young naiads. It is 

 interesting to note that these specializations have developed more 

 in the naiad of borealis 1 , and that on naiadal characters alone obscurus 

 is more generalized than borealis. This pair of species agrees with 

 Amphiagion and a number of other pairs of closely related species 

 of Odonata in which the western form is more specialized than the 

 eastern. It is interesting to those studying distribution to find that 

 these obscurus naiads from Texas, where both obscurus and borealis 

 are found, are in no wise intermediate in characters between the two 

 species but are true obscurus naiads. 



GOMPHUS (ARIGOMPHUS) SUBAPICALIS Williamson ? or SUBMEDIANUS Williamson.?' 



Plate 115, figs. 8-12. 



Four specimens, 14-19 mm. in length. Wing pads minute, tri- 

 angular, barely reaching beyond the apex of segment 1. Probably 

 1 year old or half grown. 



By Doctor Needham's key, 8 to the naiads of the subgenera of the 

 genus Gomphus these run out to the subgenus Arigompbus. Mutt- 

 kowski and Whedon 4 discuss the naiads of this subgenus to which 



1 See Kennedy, Odonata of Central California and Nevada. Proc. U. S. Nat. Mus., vol. 52, pp. 527-529, 

 1917. 



' In the Bulletin of the Bureau of Fisheries, vol. 36, p. 250, Prof. C. B. Wilson records this speciesfrom 

 Illinois opposite Fairport, Iowa. In a letter to the writer he states that exuviae were collected there, but 

 were lost when the Fairport Laboratory burned. 



3 Needham, American Gomphinae. Can. Ent., vol. 29, pp. 167-168, 1897. 



• Muttkowski and Whedon, On Gomphus cornutus Tough. Bull. Wis. N. H. Soc., vol. 13, pp. 88-101, 

 June, 1915. 



