234 PROCEEDINGS OF UNITED STATES NATIONAL MUSEUM. 



It has been sug^ges'^ied that the type of Citharichthys microstomus Gill 

 (Proc. Ac. Nat. Sci., Phila. 1864, 223) is Etropus crossotus rather than 

 Githarichthys spilopteruSj to which species it has been referred by Jor- 

 dan & Gilbert (Syn. Fish N. A., p. 817). The fin rays and scales agree 

 fairly with either, but the statements that the height enters 2'j5 times 

 in the extreme length, and that the mouth is "rather small" (for a 

 Citharichthys), show that Dr. Gill's fish could not have hiMnan Utrojms. 



37. Aphoristia plagiusa (Liunaius) Jordan & Gilbert. 

 One specimen taken. 



38. Paralichthys ommatus Jordan & Gilbert. 

 Rather cominon. 



39. Etropus crossotus Jordan & Gilbert. 



Four specimens of this species, each with about 42 scales in the 

 lateral line, and 70 developed rays in the dorsal fiji. Tlie type of this 

 species from Mazatlan had 48 scales in the lateral line, and 80 rays in 

 the dorsal. The specimens from Cedar Keys have the body rather 

 deeper than those from Mazatlan, If in total length without caudal. 

 We are not, however, j)repared to consider the Atlantic fish as a dis- 

 tinct species. 



40. Malthe vespertilio (Linnseus) Cuvier. 



Very abundant on the sandy bottoms in the harbor. Among the 

 eighteen specimens of this species brought from Cedar Keys the forms 

 known as Malthe cubifrons and Malthe nasuta [notata ; truncata), both 

 occur. The characters, however, ui)on which these species liave been 

 separated from M. vespertilio are so variable that we can consider them 

 as of individual value only, and we refer both cubifrons and iiasuta to 

 the synonymy of M. vespertilio. The form of the rostral process varies 

 in these specimens from that of a button-like tubercle, not projecting 

 beyond the snout, to a long conical process, one- tenth the length of the 

 fish to base of caudal. All intermediate forms and lengths are found 

 among these specimens. The rostral i)rocess a])pears to become shorter 

 with age, but there are exceptions to this rule. The width of the head 

 between anterior angles of orbits is usually greater in the sj)ecimens 

 with button -like rostral process. The height of the rostral cavity is 

 greater than the width in all our specimens from Cedar Keys, but a 

 fisli from Egmont Key, which is evidently not specifically ditferent, has 

 this cavity broader than high. The round bla<;k spots on the back are 

 conspicuous in life, but they grow fainter, and sometimes disappeai-, in 

 spirits. The belly in life is of a coppery red. 



41. Tetrodon nephelus Goodo & Bean. 



A single specimen was obtained. It has no prickles anywhere on 

 the body, but otherwise is not evidently different from T. nephelus. 

 Many similar specimens, as well as others prickly in various degrees,, 

 have been since obtained by Professor Jordan at Key West. 



Indiana University, January 25, 1884. 



