Jordan and Evermann. Fishes of North America. 2031 



inches in length, collected in a rock pool on Channel Rocks, near Point. 

 Orchard, Puget Sound, l>y Miss Adella M. Parker, of Seattle. North 

 Pacific, from Sitka to Monterey; scarce, but not infrequent in Puget 

 Sound, at about 2 to 10 fathoms ; a most singular fish. (Named for John 

 Richardson, naturalist and explorer.) 



Ithamphocottus richardsoni, GUNTHER, Ann. Mag. Nat. Hist., xiv, 1874, 370, Fort Rupert, 

 Vancouver Island, British America; BEAN, Proc. TJ.'S. Nat. Mus. 1881, 252; JORDAN 

 & GILBERT, Synopsis, 722, 1883 ; JORDAN & STARRS, Proc. Cal. Ac. Sci. 1895, 813, pi. 87. 



Family CLXXXI. AGONID^E.* 



(THE SEA POACHERS.) 



Body angular, commonly 8-angled, the caudal peduncle 6-angled, cov- 

 ered with 8 to 12 longitudinal rows of imbricated, radially striated plates, 

 the anterior edge of each plate overlying the posterior edge of the plate 

 next in front of it; plates'spinous or not. Teeth small, even, in villiform 

 bands on jaws, and in most species on vomer and palatines, sometimes 

 wholly obsolete; gills 34, no slit behind the last; pseudobvanchiae large, 

 extending down the inner side of opercle; gill rakers small; gill mem- 

 branes united, free, or joined to isthmus; ventral fins thoracic, narrow, 

 their rays I, 2; vent usually close behind ventrals; spinous dorsal large, 

 small, or absent; anal without spines; caudal rounded, about 3 times as 

 long as wide at base, with 10 to 12 long rays; base of pectorals usually 

 broad, the lower rays sometimes produced; all rays of all fins simple; 

 branchiostegal rays 6; myodoiue (tube of recti muscles) with membra- 

 uaceous roof ; basisphenoid absent; post-temporal not bifurcate, continu- 

 ously articulated with epiotic and ptcrotic; pyloric caeca few, about 4 to 7; 

 vertebra' numerous, 35 to 50. Fishes of the cold seas, living among rocks 

 or kelp, most of them of small size and fantastic form, not valuable as 

 food. Genera 20; species about 40. The species are extremely varied, and 

 must be placed in very many genera, or else reduced to a single one in each 

 sub-family. The plates vary somewhat in number in all parts of the body 

 in most if not all of the species, although not to the same extent in all the 

 genera. Even the pattern on the breast, which is definite for the species 

 of all the genera except Hippocephalns, varies in the different individuals 

 of the same species, so that it is probable that even where no variations 

 are indicated by the numbers given in the following descriptions they 

 would be found by comparing large numbers of individuals. The plates 

 in the dorsal series vary from 1 to 3 in number in most species; they cor- 

 respond closely with the number of vertebra 1 , there being usually 1 or 2 

 fewer of the former than of the latter. It seems probable that the vertebra? 

 vary a little in number within the species, and the rings of plates cor- 

 respond in number with them at least throughout most of the length of 

 the body. There is no definite proof, though the balance of evidence 

 seems to indicate, that the superior lateral series of plates in Aspidopho- 

 roidcs, bearing the lateral line, corresponds with the median lateral series 



*The account of the Agonidce is contributed by Mr. Frank Cramer, a graduate student 

 in Lelaud Stanford Jr. University, some additions having been made by Jordan &. 

 Evermann. 



