THE FISHES OF BERING SEA. 473 



ventrals very sliort, not exceeding length of snout before mouth, not received into 

 longitudinal groove. We consider it very doubtful whether such a groove exists in 

 any of the other species of this group. It has been described as existing in the types of 

 P. gilberti (Collett) and P. peristethus, Gill. In both cases the type specimens were in 

 a ijoor state of iireservation, and the groove was probably due to a softening of that 

 longitudinal strip of the abdominal wall which includes the anal opening, and extends 

 backward from the base of the ventral fins and is interposed between the firm outer 

 series of ventral plates. That such a softening had occurred iu the type of Podothecus 

 peristethus is evident from Gill's statement that the ventrals had dropped out. This view 

 is rendered more ijrobable from the fact that P. peristethus is apparently identical with 

 the common P. acipenserinus, which contains no such groove. We have also examined 

 two of the type specimens of P. gilberti without being able to satisfy ourselves of the 

 existence of any special groove. The dorsal fins are closely juxtaposed, the interspace 

 including 1 or IJ pairs of plates. The base of the last ray of second dorsal is midway 

 between base of caudal and origin of spinous dorsal. 



Color dark or brownish above, with irregular spots or dashes of darker, which do not 

 form definite crossbars; a black streak from eye to tip of snout, passing onto lower side 

 of rostral spines ; a dark blotch on expanded limb of preopercle; a black spot on base of 

 middle pectoral rays, the fin very obscurely marked with dusky; dorsal spines and rays 

 with linear dark markings, one or two black spots near tip of spinous dorsal anteriorly; 

 under parts, including- fins, unmarked. 



Two specimens from Albatross station 3653, off Shana Village, Iturup Island, in 18 

 fathoms. A young individual from Albatross station 3646, o& Eobben Island, 18 

 fathoms, seems to belong to the same species, but has the snout less produced and the 

 dorsal VIII, 6; anal 8. (We take great pleasure in naming this species iu honor of 

 Hon. Charles Sumner Hamlin, late Assistant Secretary of the Treasury, under whose 

 auspices the investigations described in this memoir were undertaken.) 

 124. Podothecus thompsoni, Jordan and Gilbert, new species. (Plate LXXII.) 



D., VIII or IX-6; A., 6; P., 16. Head rather broadly triangular, its greatest 

 width across preopercular ridges greater than distance from anterior end of preoper- 

 cular ridge to tip of snout. Lateral ridge on head continuous from tip of snout along 

 suborbital bones to base of preopercular crest, the lateral spines usual in this genus 

 being represented by triangular processes borne on the ridge; preopercular ridge 

 produced posteriorly beyond gill opening, but not spine like; snout terminating 

 anteriorly in two rounded processes, each bearing on its upper surface a vertical crest, 

 and finely serrate along its margins; no terminal pair of strong spines as in other 

 species; under side of snout with an acute median spine directed downward and 

 backward; a pair of strong spines on upper side of snout behind terminal nostril 

 ridge; a pair of coalesced spines behind the nostril groove; a semicircular series of 

 spinelets below the eye; two small tufts of filaments on under side of snout, one on 

 middle of maxillary and one at its tip. A narrow band of sharp teeth in each jaw ; 

 vomer and palatines toothless. Gill membranes united to isthmus, without evident 

 free fold. Orbital rim much elevated; interorbital space narrow, deeply concave, its 

 width nine-tenths diameter of orbit; occipital ridges strong, elevated posteriorly, 

 ending in a backwardly directed spine which is much larger than those of the body 

 plates; occii^ital area narrow, deeply concave, its central portion sunk somewhat 

 below level of interorbital space, from which it is separated by a shallow transverse 



