100. LlPARIDID.i: AMITRA. 739 



<f. Vcntrjil disk entirely '.vaiitinj;. (Amitriuw.) 



b. Lower .jaw iiuliided ; skin tliin. lax Amitka, 392. 



lut. Ventral disk present. {Liparimv.) 



c. Ventral disk very small, nuder the head, the vent elose behind it; lower jaw- 

 projecting ; skill lax, thin Careproctus, 3'.>:i. 



cc. Ventral disk well developed ; vent well behind head Lipakis, 31M. 



392.— AI?HTRA Goode. 



(Goodc, Proc. U. S. Nat. Mus. ISSO, 478: type Amiira Uparina Goode.) 



Body elongate, atteimaro backwarrl. covered witli lax, smootli, slimy 

 skill, which is separated from the body and fins byatilmy miicons inter- 

 tissne. Head small, thick, convex between eyes; snout convex, pro- 

 truding; mouth small, liorizontal, the lower jaw included; teeth small; 

 eyes lateral; gill-openings restricted to small slits above the base of the 

 pectorals, covered by the very small opercle; no barbels; dorsal and 

 anal fins many-rayed; dorsal continuous, almost concealed by the skin; 

 the spines scarcely different from the soft rays; anal similar to dorsal; 

 both connected with the caudal, which is pointed; no trace of ventral 

 fins or sucking disk; pectoral fin very broad, jirocurrent, its lower rays 

 inserted nearly under the eye, the fin emarginate; vent well behind 

 head. One species known ; its position intermediate between the Cot- 

 t(d(e {^Cottuncuius, Psyclirolutes) and LiparUlUla^ two groups much more 

 intimately related than most writers have conceded.* («, without; 

 luzfia^ stomacher.) 



1129. A. lipariiia Goode. 



Yellowish white, dusky toward the tail and front of head; perito- 

 neum black. Body posteriorly compressed, the tail very thin; eyes 

 half width of interorbital space, 5 in head; upper lip with i)ores. Dor- 



*I should put the family of Gobiesocidm far away, at least a suborder off, from the 

 (yi/clopteridw and Liparididce, which are far more closely united to the true CoUUUv, 

 represented by Cotttis and HemilripleruH^ than to either the GobiesoddcB proper or to the 

 Gobies and Blennies. In fact, Liparis has as close affinities, as shown by its skeleton, 

 with Cottiin and Hemitripterun as with Cycloplcrus ; and wo have in the three groups. 

 rei)r('sented by Cottufi, Liparin, and Ciicloptrriis, well-marked fa.iiiilies of the same sul-- 

 order. The only character by which the (UicJopteridiv and Liparidkhr are closely united 

 cou^isfs in the peculiar formation of tin; ventral disk by the union of the ventral fins: 

 but as Ibis structure is simply l>ronght about by the modilieation of th(> rays in a 

 maniKir conunon to the several genera, and not by any marked anatomical dillerence in 

 the structure of the same (ins in Cottua, I can only look upon it as a generic (•haia(t( r 

 common to the known representatives of both families of Cjicloptiridcp nud Lipoiidlda : 

 ;ind the discovery of a representative of either family with ventral hnsof the ordinary 

 form would not necessitate the establishment of a family for its recejition, as in th.it 

 cat.0 we should simply consider the structure as of generic value. (Putnam, Proc. 

 Am. Ass. Adv. Sei. 1873, 337.) 



