DISCUSSION OF SPECIES AND THEIR DISTRIBUTION. 167 



were based on different specimens, a theory accepted without criticism by later writers, l)ut 

 which we can not believe a true one. 



Kisso was a careful and experienced worker, and it would be unjust to the memory of 

 one of the best Italian ichthyologists to admit that he could be guilty of such an error. 

 Then, too, he states positively that he had only a single specimen. It is much more prob- 

 able that the German typesetter, in the oftice of Wicgmunn^s ArcMv, mistook a ''7" for » 

 "9" in Kisso's manuscript. 



Eisso's figure is a good one of a young iV\ mediferraneKS, and his descriptir)n agrees 

 with it perfectly with the exception of this one figure in the text. 



The specimen described and figured by Giinther under the name N. mediterraneus is 

 not a Mediterranean form, but one from the southern Pacific, and has been referred by us 

 to a new genus and species. Moreau is in error in referring the figures of Bloch and of 

 Cuvier and Valenciennes to this species (see discussion under Kotacaiithns nnsii.s). 



N. Boiuqjartii was described under thename N. mediterraneus, by DeFilip])i and ^'erany 

 in 1857 from a specimen obtained at Nice, and preserved in the Zoological Museum at Turin. 

 Two others from the same locality, referred by Moreau to this species, are in the Museum 

 in Paris. The TravaiUeur and Talisman obtained four additional individuals: one from 

 the coast of Soudan, at a depth of 1,232 meters, and another from the same region at 932 

 meters; two from the Banc D'Arguin at 1,495 meters. These last have been made the sub- 

 ject of an elaborate description by Vaillant, who also publishes a good figure. 



This species is distinguished from N. sexspinis (Fig. 192 A, B), described by Eichard- 

 son from Australian seas (Voyage Erebus and Terror, Fishes, 54, pl.xxxii. Figs. 4-11), and 

 subse(piently described by Giinther, wlio also gives an excellent figure (Challenger Eejiort 

 XXII, 243, pi. LXi, Fig. «), by various characters, most striking of which is the diflerence 

 in the relationships of the position of the dorsal spines and the soft anal fin. In V. sex- 

 82)inis the dorsal and soft anal do not pass the same vertical, whereas in J\r. mediterraneus 

 the last three dorsal siiines are placed over the anterior part of the soft anal. 



The type of N'. mediterraneus from Nice was examined by Giglioli at the Tiuiu Museum 

 iu 1882. Its total length is 203 millimeters, and its radial formula 1). (J/1; A. 12/132?; Y. 

 3-4/8; C. 5?. 



Prof. Giglioli informs us that in his "Central Collection of Italian Vertebrata" at 

 Florence he has four specimens of N. Bonapartii, as follows : 



a. Nice, August 11, 1S82. Total length 153 millimeters. D. 8/1; A. 6-7/120; V. 3/6-7; p. 9-10; C. 3-4 f. A 



large, curved spine in upper corner of mouth on either side. 

 6. Nice, March 7, 1891. Total length 205 millimeters. D. 7/1; A. 14-120; V. 3 7; P. 12; C. 4». Buccal 

 spiuea hidden in skin. 



c. Nice, June 15, 1892. Total length 203 millimeters. D. 7/1; A. 8?/140; V. 3/5-7; P. 10-12. Buccal spines 



large and prominent. 



d. Syracuse, 185.5-60?. D. 7/1; A. 11/25; P. 9-10; V. 3/5. 



Another specimen, collected by Bellotti at Messina, December 12, 1882, and now in the 

 Museo Civico at Milan, was examined by Giglioli, who states that it was 104 millimeters 

 long, and had D. 7/1; A. 7/?; V. 3/6; P. 10-12; C. 5?. 



NOTACANTHUS PHASGANORUS. Goode. (Figure 186.) 



Notacanthus phasganorus, Goode, Proc. U. S. Nat. Mus., ni., sig. 34, .535, Apr. 18, 1881. — GOnther, Chal- 

 lenger Report, XXII. 249. — Jordan and Gilbkkt, Hull, xvi, U. S. Nat. Mus., 900. — Vaillant, Exp. Sci. 

 TravaiUeur et Talisman, 318, et siy. 



A N'otacanthus, with its body a little higher over the ventrals than over pectorals, and 

 comparatively elongated; witli its lateral line slightly arched above the pectorals, sinking 

 to median line of body in advance of first dorsal spines, and its last dorsal spine over the 

 fifth from the last anal spine. 



Its body is much compressed, it greatest width slightly more than one-third height of 

 the body at vent. Scales round, thin, flexible, very small upon the head (not wider than 

 the diameter of one of the dorsal spines) but upon the antericu- half of tin; body about three 

 times as large, decreasing in size upon posterior half, until upon tail they arc smaller than 



