﻿394 



PROCEEDINGS OF THE NATIONAL MUSEUM 



South Carolina, Tennessee, and Louisiana, accorded exile subspecific 

 status. My examination of 25 specimens (56 to 196 mm. in standard 

 length) from Texas (none from Galveston, type locality of exile) 

 and of 26 (61 to 211 mm. in standard length) taken at random from 

 Maryland south to North Carolina gave the following ratios (of 

 greatest body depth to standard length): Texas — 2.45 to 2.9 (avg. 

 2.72); Maryland to North Carolina— 2.3 to 2.95 (avg. 2.65). 



These preliminary data indicate that body depth is very variable 

 in D. cepedianum and that exile cannot be subspecificaUy distinguished 

 on this basis. I did find that large adults from Atlantic-slope waters 

 definitely tend to be deeper-bodied than do those from Texas waters — 

 perhaps a racial characteristic. The proportionate length of the head 

 and of the caudal peduncle in these specimens gave similarly variable 

 results. Careful measurements of many individuals of various sizes 

 from the entire range of cepedianum is obviously required before 

 deciding if subspecies are to be recognized. 

 Table 4. — Number of ventral scutes and vertebrae in four species of Dorosoma 



[The figures in the first line for each count are the observed range and, In parentheses, the mean; the figure 

 in the second line represents the number of specimens] 



Species 



Number of scutes 



Prepelvic 



Postpelvic 



Total 



Number of 

 vertebrae 



cepedianum. 



anale 



chavesi 



smithi 



17-20 (17.99) 



196 

 17-20 (17.96) 



71 

 1&-18 (16.91) 



35 

 15-18 (16. 84) 

 185 



10-14 (11.76) 



197 

 9-12 (10.23) 



71 

 10-12 (10. 58) 



36 

 9-12 (10.87) 

 185 



27-32 (29. 74) 



196 

 25-31 (28. 18) 



71 

 25-29 (27. 51) 



35 

 26-30 (27. 71) 

 185 



48-51 (49.83) 



42 

 46-48 (46.91) 



22 

 44-47 (46. 13) 



15 

 43^7 (44. 71) 



34 



Material examined. — In determining the range of variation in me- 

 ristic characters for D. cepedianum, I used the data published by Hubbs 

 and Miller (1941, p. 234, table 1) and in addition examined many 

 collections of this species deposited in the U. S. National Museum. 

 These included 33 specimens from 6 localities on the Atlantic Slope 

 (Potomac Kiver to Orlando, Fla.); 33 from 10 locahties along the 

 Gulf of Mexico (Florida to Texas) ; and 35 from 13 locahties in the 

 basins of the Mississippi Valley and the Great Lakes. Critical mate- 

 rial from northeastern Mexico was examined both at Michigan and 

 at Washington. In recording the number of vertebrae, I included 

 the 26 specimens counted by Hubbs and Whitlock (1929, p. 463). 



Range. — From Nebraska and Minnesota to the St. Lawrence Kiver 

 and the Ohio Valley in western Pennsylvania; south to the Gulf of 

 Mexico and to the Rio Pd,nuco Basin of eastern Mexico; coastwise 

 northward to about latitude 40° N. in New Jersey and eastern Penn- 



