﻿390 



PROCEEDINGS OF THE NATIONAL MUSEUM 



Table 1. — Comparison of Dorosoma and Signalosa 



VOL. 100 



1 Counts based on 30 specimens in the U. S. National Museum as follows: Florida (2), Arkansas (2)' 

 Alabama (1), Louisiana (4, one a cotype of S. atchafalayae, U.S.N.M. No. 48971), Tampico, Mexico (6). 

 near Veracruz, Mexico (5), and Lake Petto, Guatemala (10, topotypes of /S. petenensis). 



' Where the two genera coexist, Dorosoma has 17 to 20 and Signalosa 14 to 17 (rarely 17) prepelvic scutes. 

 The counts of anal rays and ventral scutes for Signalosa are based on 186 specimens from Florida to Lake 

 Petto, Guatemala, all deposited in the U. S. National Museum. 



recently emphasized. Part of tlie difficulty has been the lack of 

 adequate material, particularly from Guatemala and Mexico.^ Thus 

 m the major division of his key, based on the number of abdominal 

 scutes, Weed (1925, pp. 141-142) had to rely upon the published 

 descriptions of S. petenensis, with the result that he widely separated 

 petenensis and atchafalayae. An examination of 52 topotypes of 

 petenensis from Lake Peten, Guatemala (U.S.N.M. No. 132269), 

 demonstrates that, in this character at least, there is virtually no 

 difference between these species. The total number of abdominal 

 scutes in this sample varied from 26 to 29, and in 68 specimens in the 

 National Museum from Louisiana to Florida (the range given by 

 Weed for atchafalayae) the number varied from 25 to 29. Material 

 from Mexico, however, shows far fewer scutes. In 38 specimens 

 from El Hule =Papaloapdn (U.S.N.M. No. 55739), Lake Catemaco 

 (U.S.N.M. No. 48213), and Tampico (U.S.N.M. No. 62271) the scutes 

 varied from 23 to 26, which agrees closely with the variation of 20 to 

 27 given by Weed (1925, p. 143) for his Mexican material (including 

 a sample from Brownsville, Tex.). On the basis of this character, 

 S. mexicana (Giinther) appears to be very distinct from both atcha- 

 falayae and petenensis, but Hubbs, after study of much material, 

 recognized (in Hubbs and Allen, 1943, p. 116) only a single species, 

 petenensis. Fowler (1945, pp. 22, 266, 366, 372), without presenting 

 data, followed Hubbs and Allen. A thorough revision is obviously 

 needed. 



2 Fowler's record (1911, p. 211) of "Dorosoma petenensis" from Panama obviously represents an erroneous 

 locality. Signalosa is not known to occur south of northern Guatemala and British Honduras (Rio Belize; 

 uncataloged material at U.M.M.Z). 



