10 PROCEEDINGS OF THE NATIONAL MUSEUM vol. 82 



Gordon Gunter, who is a member of the staff of investigators of the 

 Bureau of Fisheries working on the shrimp problem, in appreciation of 

 the aid he rendered in my studies of the Gulf coast fishes during 1931 

 and for other courtesies shown. 



Family ELEOTRIDAE 

 ELEOTRICA, new genus 



Genotype. — Eleotrica cahleae, new species. 



Definition. — Ventrals closely approximated, but altogether sepa- 

 rate. Entirely scaleless. First dorsal with seven flexible spines. 

 The two dorsal fins separate. Second dorsal with more rays than the 

 anal, 12 and 11 rays, respectively. Teeth pointed, in bands, outer 

 row of both jaws more or less enlarged, but no definite canines; no 

 teeth or vomer or palatines. Body moderately elongate. Caudal 

 short, rounded. Maxillary moderate, not extending past posterior 

 margin of eye. No spine on preopercle. Shoulder girdle without 

 flaps of skin. Tongue rather emarginate, the cleft not deep. Cuta- 

 neous papillae on cheek in transverse and longitudinal rows; a series 

 of short transverse rows along middle of body; some of papillae on 

 top of head and over opercle greatly developed to form short tubules. 



Remarks. — Of the known gobioid genera, as far as their external 

 characters are concerned, Eleotrica seems to be most nearly related 

 to Chriolepis Gilbert. ^^ A reexamination of the type of Chriolepis 

 minutillus brings to light some inaccuracies and omissions in the orig- 

 inal description, which are here corrected in order to show the 

 divergence between the two genera. C. minutillus has some well- 

 developed scales posteriorly. There is a single row of four strongly 

 ctenoid scales on the caudal, at its base. The spinules at the hind 

 margin of these scales are conspicuously well developed, especially 

 on the upper and lower scales, which are considerably larger than the 

 other two, and the spinules laterally are very long, becoming grad- 

 ually shorter toward the center. There are also two large scales on 

 the caudal peduncle, one behind the other, near the base of the caudal 

 fin. Whether other scales were originally present is not evident on 

 the type specimen. The scales as described above are present only 

 on the right side ; the left side having but a single scale left, the others 

 apparently having fallen off. No tubular pores are present. The 

 teeth are in more than one row in both jaws, as in related genera, not 

 in a single series in the mandible as described originally. Cutaneous 

 papillae on the cheek are not evident in the type specimen, but this 

 may be due to its state of preservation. Eleotrica, therefore, differs 

 from Chriolepis chiefly in the total absence of scales and in some of 

 the pores being markedly tubular. In its physiognomy Eleotrica is 

 quite similar to Gobiosoma (see p. 13). 



» Proc. U. S. Nat. Mus., vol. 14, p. 557, 1891. 



