FISHES FROM BERING SEA AND KAMCHATKA. 75 



the anus close behind it. Origin of dorsal over the gill slit, its anterior portion wholly concealed in a 

 thick covering of gelatinous material beneath the lax skin ; the rays increase regularly in length from 

 the origin, there being apparently no differentiated spinous part; the last rays are slightly shortened, 

 giving a rounded contour to fin, which joins the basal half of the caudal. Anal similar to dorsal, joining 

 basal 0.6 or 0.66 of caudal. The lower pectoral lobe consists of 11 thick rays with exserted tips, the 

 longest reaching the vent. 



Skin extremely lax, with much soft gelatinous tissue intervening between it and the muscular 

 mass; small open pores along margin of snout and across cheek and on mandible. An obscure series of 

 slit-like pits, apparently not perforate, mark the anterior portion of tlie lateral line and exhibit the 

 usual short curve above base of pectorals; no papills are present. 



Color in life: Varying in different specimens from light rose red to light lemon yellow; the spots 

 accord with the ground color, and are either reddish brown or yellowish brown, usually darker at or 

 toward the margins, the darker portion often forming a distinct brownish red ring, most conspicuous in 

 specimens with a yellow ground color. Surrounding all is a light pearly gray ring, which often spreads 

 over the adjoining parts as a suffusion and seems to cover the ground color. The spots are large and 

 roundish, varj'ing in size, shape, and position; usually a conspicuous median series along anterior part 

 of dorsal profile, either a single one or a pair on occiput, one above and an elongate one below and in 

 front of each eye. 



This is tlie species with which the name Liparis cychstigma Gilbert has usually been associated, 

 and is the species to which it was intended to apply that name. Through an unfortunate clerical error, 

 the name cyclostigma became attached to a very different species in no way resembling tlie one here 

 described, and wholly without strikingly distinctive color marks. As the body of the original descrip- 

 tion of cyclosligma, and also the type as designated, concern the tmadorned species, it will be necessary 

 to use tlie name as a palpable nomen ineptum. As a further source of confusion, a color description 

 applying to the present species was added to tlie diagnosis of cyclostigma as the latter was passing through 

 the press. 



An examination of the type of cyclostigma (no. 48621, U. S. National Museum) shows it to be a true 

 Cyclogasler, with two well-defined nostrils, but without anterior or posterior tube. It is a rather elon- 

 gate form, with broad depressed head, about as wide as deep, and a very broad heavy snout. The 

 color was apparently grayish, vermiculatcd or reticulated with darker; the vertical fins are mostly 

 black, the dorsal with lighter base, all the rays with whitish tips, the black of the fins more or less 

 variegated with lighter mottlings; pectorals similarly marked, darker on inner margin, the course of 

 the rays externally lined with lighter. The lighter markings on the fins may have been yellowish or 

 greenish in life. Inside of mouth and gill cavity and tlie peritoneum white. 



We append a table of measurements of the type of C. cyclostigma given in hundredths of lengtli 

 without caudal: Head 27.8; exposed portion of eye ;i.i; orbit 5.9; snout 10.4; maxillary 14; cleft of 

 mouth 11; interocular width 11; depth at occiput 21.5; width of snout 22; greatest depth 25; length 

 of disk 12.9; distance from disk to tip of mandible 17; to vent 12.9; vent to first anal ray 7; snout to 

 dorsal 30.6; lengtli of gill opening 11; width of pectoral base ig; longest pectoral ray 25;longest caudal 

 ray 18; longest ray of lower pectoral lobe 18.5. 



In the type description of Crystallichthys mirabilis Jordan and Gilbert, it is to be noted that the 

 smaller of tlie two specimens, taken near St. Paul Island at station 3638, and figured on plate Lxxvi, 

 does not belong to C. mirabilis, but is the young of C. cyclospilus. The specimen of C. cyclospilus which 

 has been heretofore erroneously figured as Liparis (or Crystallichthys) cyclostigma (see synonomy) is now 

 in such poor condition that it has been thought advisable to select another specimen as the tj'pe of 

 cyctospihis. 



Crystallichthys mirabilis Jordan & Gilbert. 



A single specimen 87 mm. long, from station 4794, off the southern coast of Kamchatka; depth, 58 

 fathoms. This is the only locality from which tlie species is known, and is the second specimen to be 

 placed on record. The young specimen from station 3638, near St. Paul Island, Bering Sea, mentioned 

 in the original description, was incorrectly identified and belongs to C. cyclospilus. 



