NO.1301. OSTEOLOGY OF ITEMIBRANCHIATE FISHES— STARKS. 621 



It is not apparent wh}- Dr. Giinther ^ should write, referring to the 

 Fistularidw, "ventral tins abdominal, separate from the puhic bones, 

 which remain attached to the humeral arch." The "interclavicles" 

 mig-ht easily be taken for pubic bones, but the ventral tins are attached 

 directly to w^ell-developed pubic bones. 



Dr. Giinther," referring to ^EoUscus {strigatus) as seen externafl}^, 

 makes the following statement: '*Tho horizontal portion of the 

 humerus,' which is visible externall}-, is of moderate length, rather 

 broad at its middle, and obliquely truncated behind; the coracoid, 

 which is situated immediately below, has its basal portion not styli- 

 form, but is broad like the humerus above.'' He has evidently taken 

 the process of the clavicle for the hypercoracoid and the supraclavicle, 

 including, possibly, the ankylosed plate above, for the clavicle. 



Drs. Jordan and Evermann* give the number of "pectoral ossicles" 

 (actinosts) in Flstularki as three. There are four, as usual, the upper 

 one closely attached against the hypercoracoid. 



The Hemibranchs certainly do not deserve coordinate rank with the 

 Acanthopterl^ but should be included as a suborder under them, coor- 

 dinate with the Percesoces. Probably the Synentognath fishes should 

 also be so included. 



A representative of each of the four Synentognath families has been 

 examined as follows: In Belonidiv, Ti//osurus /odiatc/r iind Tylosicrus 

 niarinus,' in Hemiramphidie, Ihiporlmmphus wnifasciatus; in Scom- 

 bresocidse, CololahU saira; in Exoc<jetidie, Cyj^selurus californicrts and 

 Cypselurus agoo. They show relationship to the Hemibranchs in pos- 

 sessing the following characters: Exoccipitals not united over the 

 basioccipital;^ parapophyses developed on all abdominal vertebrse; 

 postclavicle absent (absent only in some families of Hemibranchs); 

 supraclavicle very small "^ (in all but Cololabls^ in which it is absent); 

 post-temporal, when present, usuall}^ simple^ (absent in Cololabis^ forked 

 in Tylosu)'us)\ parietals absent* (absent in all the Hemibranchs except 



'Cat. Fish. Brit. Mus., Ill, p. 529. 



•^Idern., Ill, p. 528. 



^Dr. Giinther in his Catalogue of the Fislies of tlie British iMuseum, p. 532, calls 

 the clavicle of Parker the humerus, and in his Introduction to the Study of Fishes 

 he calls it the cla'vicula. In the catalogue he calls the hypercoracoid of Gill the 

 coracoid, and in the Study of Fishes the scapula. 



■'Fishes of North and Middle America, Bull. U. S. Xat. Mus., No. 47, p. 756. 



^I put but little weight on this character, but give it for what it may be worth. 

 It is not an uncommon condition, especially among the lower fishes, while the 

 opposite seems to be the rule among the higher. 



^Cope says: "Epiclavicle (= supraclavicle) not distinct." Systematic Relations 

 of Fishes, Proc. Amer. Asso. Adv. Sci., 1871. 



'Said by Cope to be "slender, furcate." 



*Said by Cope to be "very much reduced." Cypselurux and Hyporhamphus 

 appeared to have small parietals, but maceration of the specimens proved the 

 contrary. 



