270 PROCEEDINGS OF UNITED STATES NATIONAL MUSEUM. 



from the beginning of the dorsal, one third the length of the pectoral 

 (nearly 1.20: 3.70). Small teeth. This form agrees with your species 2 

 {Mugil gaimardianus), except in the distance of the pectoral from the 

 dorsal. 



"The small example, in its whole appearance quite unlike the preced- 

 ing, shows: Anal, III, 9. Scales, 38 or 39. Distance of the end of the 

 pectoral from the dorsal very slight (about one-twelfth of the length of 

 the pectoral, nearly .03 to .34) ; teeth large, absolutely larger than iu 

 the preceding larger fish. 



"This animal differs from your species 3 {Mugil curefna=hrasiliensis 

 Auct.) again in the distance of the pectoral from the dorsal. To con- 

 clude: Only the dried example agrees exactly with your species 1 {tri- 

 chodon), (except in the number of anal rays, which cannot be counted), 

 and this example is certainly the one originally figured by Spix (I have 

 also asked Professor Zittle to verify this), the discrepancies of the figure 

 being the fault of the artist." 



In any case, therefore, whatever the two smaller specimens collected 

 by Spix and referred to M. hrasiliensis by Agassiz, may prove to be, it 

 is evident that the original type of Mugil hrasiliensis does not belong to 

 the species called by the latter name by Giiuther, Steindachner, and 

 other recent writers. There can be little doubt, in fact, of its identity 

 with Mugil trichodon Poey, for which reason we here retain for the latter 

 species the name Mugil hrasiliensis. 



Mugil curema is doubtless the present species, as I am informed by 

 Dr. Sauvage that the type preserved in the museum at Paris has 40 

 scales in a longitudinal series. 



Mugil petrosus is to all appearance also the same species, some of the 

 specimens (New York) being certainly the same. 



Mugil platanuSj Giinther is identified by Steindachner with this spe- 

 cies, the presence of but 8 soft rays in the anal fin being regarded as 

 accidental. As, however, in M. platanus, the dorsal and anal fins are 

 said to be naked, it is probably most nearly related to M. cephalus, of 

 which it may be a variety. 



7. Mugil hrasiliensis. Fan-tail mullet. 



Mugil hrasiliensis, Agassiz, Spix, Pise. BrasiL, 1829, 234, tab. 72 (Brazil) 

 (typical example ; not the two smaller ones). 



Mugil trichodon, Poey, Ann. Lye. Nat. Hist. New York, 1875, xi, 66, tab. 8, 

 f. 4-8 (Cuba) ; Poey, Enumeratio, 187.'>, 99 ; Jordan, Proc. U. S. Nat Mus., 



1884 (Key West). 



Habitat. — Cuba, Florida Keys, Brazil. Head, 4^ (5^) ; depth, 3| 

 (4f ). D. IV, I, 8. A. Ill, 8 ; Scales, 11-33. Length, 11 inches. 



Body rather robust, its depth somewhat greater than in M. curema. 

 Snout rather narrow and pointed, the upi)er and lower profile about 

 equally oblique. luterorbital space flatfish or slightly convex, 2J in 

 head. Upper lip thick; thicker than in any other species here de- 

 scribed. Space at the chin between the mandibulary bones oblanceo- 



