﻿400 PROCEEDINGS OF THE NATIONAL MUSEUM vol. lOO 



short dorsal filament, posterior position of the pelvics, short basal 

 length of the anal fin, and few anal rays chavesi agrees rather closely 

 with smithi (tables 2 and 5). The number of vertebrae, usually 46, 

 is somewhat intermediate between that recorded for anale and smithi 

 (table 4). On the basis of present knowledge, it may be hypothe- 

 sized that the similarities between chavesi and smithi suggest that the 

 two arose from a common ancestral stock. The fact that Lake 

 Nicaragua was at one time a Pacific tributary (Hayes, 1899; Durham, 

 1944; and Marden, 1944) might help to explain these resemblances. 

 An alternate hypothesis, however, is proposed under the account ot 

 Dorosoma smithi. 



Variation. — Counts of fin rays, scales, and ventral scutes were made 

 on 22 to 36 specimens (tables 2-4). In addition, 24 measurements 

 were recorded for 21 specimens (table 7), Although no correlation 

 was noted between the samples from the two lakes and their meristic 

 characters, unmistakable differences were observed in certain measure- 

 ments. Thus five specimens from Lake Managua differed prominently 

 from five of comparable size taken in the basin of Lake Nicaragua in 

 having (1) a more posteriorly placed anal fin, as expressed by the 

 shorter distance between anal origin and caudal base; (2) a longer 

 and deeper head; (3) a longer snout + eye (the eye is only slightly 

 larger but the snout is much longer); (4) a shorter anal fin base; (5) 

 a longer upper jaw; (6) a more posterior dorsal fin ; (7) more posteriorly 

 inserted pelvic fins; (8) a broader body; and (9) a narrower caudal 

 peduncle (table 7). Since only five specimens of comparable size were 

 available from the basins of the tv/o lakes, and since it has been shown 

 that body form in the gizzard shad is subject to environm.ental 

 modification (Hubbs and Whitlock, 1929), I hesitate to give nomen- 

 clatorial recognition to the Lake Managua form. If and when large 

 samples from these lakes become available and the dift'erences shown 

 are further tested and found to be valid, then I would regard chavesi 

 as comprising two subspecies. There are falls between Lakes Managua 

 and Nicaragua which seem to prevent the migration into Lake 

 Managua of certain species found in Lake Nicaragua (Meek, 1907, p. 

 99; Marden, 1944, pp. 178-179). According to Marden, the stream 

 connecting the two lakes is normally subterranean. 



The caudal rays of 27 specimens were 19, excejDt for two with 18 and 

 one with 17 rays. The pelvic rays of 29 specimens were 8-8, except 

 for one with 7-8. 



Material examined.^ — In the following list of specimens all but 

 U.S.N.M. Nos. 16882 and 22138 represent paratype material col- 

 lected by S. E. Meek in March, 1906. 



• U.S.N.M. No. 30965, an adult collected by Capt. J. M. Dow, reportedly in Panama, represents D. 

 chavesi and obviously was not taken in Panama but probably came from Nicaragua, where Dow 

 also collected. 



