DISCUSSION OK SPECIES AND THEIR DISTRIBUTION. 1!>3 



2. Corselet developed. 



«. Palatine teeth strong; vomer toothless; vertebrae normal Sakha 



h. Palatine teeth villiform. 



Vomer with teeth : vertebral normal, the lower foramina small. 



Lateral line single Orycni a 



Lateral line double Grammatorcynus 



Vomer toothless; abdominal vertebras with the lower foramina enlarged and a portion 

 bet wren the vertebral proper and the haornapophj aes developed in the form of a nel 



work or trellis GYMNOSAKDA 



II. Body very elongate, compressed. 



A. Spinous dorsal long; pectorals comparatively low: caudal not keeled; pyloric casca rather 

 few Thyrsi us i 



1. Ventrals well developed. 



(i. Finlets present; teeth on palatines; hod} in greater part naked. 



' Lateral line descending below posterior part of the spinous dorsal : body rather elongate; 

 dorsal and anal Unlets 6 (Type. Thyrsites tihin. South Africa and Van Diemen's 

 Laud) Thyrsites 



* * Lateral line nearly straight; dorsal finlets 5, anal 1; body rather elongate. .Thyrsitops 



* * * Lateral line absent; tin lets few; skin spinigerons ; abdomen keeled; bodj somewhat 



elongate Ruvettus 



6. Finlets absent ; palatines toothless ; scales minute. 



* Dorsal iu two portions. Dagger-shaped spine behind vent NesiaRCHDS 



* * Dorsal continuous. Two lateral lines Ei'Innci.a 



2. Ventrals reduced to a single spine. 



((. Finlets present; teeth on palatines, none ou vomer. 



* Dagger-shaped spine behind vent. Body much eompressed, incompletely covered with 



delicate scales NkaLOTUS 



* * No dagger-shaped spine. Body slender, fusiform; lateral line descending in strongly 



oblique line below front of spinous dorsal Promethichthys 



b. Finlets absent; teeth on palatines and vomer. 



* Ventral spine long, crenulate; preoperculum spinigerous Dicrotcs 



B. Body very elongate (height 18 to 20 times in length); spinous dorsal very long, with numerous 

 spines, continuous with the second; caudal not keeled; pyloric caeca not numerous Gkmpyi.in i: 



1. Body scaleless; no palatine teeth Gempyltjs 



Subfamily SCOMBRIN^E. 



Scombrids with fusiform, moderately elongate bodies, and short (less than lii spines) first 

 dorsal fin, separated by a wide space from the second dorsal. Pectorals nigh up. Caudal 

 nut carinated. 



Several of the forms in the subfamily occur at times far out at sea, but this is no reason 

 for considering them properly pelagic. Like the common mackerel (Scomber scombrus), 

 which also is found in schools remote from land, they probably all breed near the coasts, 

 and depend for their food upon the shoals of small fishes and crustaceans, which are abund- 

 ant near the hind. Scomber spawns at. considerable depths, but probably never beyond 

 the KMi fathom line. It is probable that many of the ScombHdcB sink at times far below 

 the sin face strata, especially in the winter. These matters are but little understood, tow- 

 ever, and it is indeed probable that our knowledge of the bathic distribution of the mem- 

 bers of the mackerel family is no more imperfect than our supposed knowledge of their 

 geographic range and migrations. The remarks of Valenciennes in the "Histoire Xaturelle 

 des lies ( Janaries" (Poissons, p. 49), though written fifty years ago, are worthy of consider- 

 ation to-day. 



Subfamily ORYCNIN^G. 



Scombrids having a fusiform, moderately elongate body, ami spinous dorsal long, com- 

 posed of more than 12 spines, and contiguous to the second. Pectorals low. Caudal carinate. 

 Pyloric casca very numerous. 



The tunnies, bonitoes, and horse mackerels composing this family are even more charac- 

 teristically pelagic than those oi I he preceding subfamily, but never have as ye1 been found 

 below 100 fathoms. There is every reason to suppose that at times they sink below the 

 upper strata. 



1980S— No. 2 13 



