46 Shufeldt — Robert Collett on Pterycombus brama Fries. 



interspinals form together an almost solid perpendicular wall of bone, 

 which is entirely different from anything known as pertaining to the other 

 genera of the Scombroids. In this respect it so far departs from what we 

 find in its apparently nearest relative Brama, that it should probably be 

 placed in a family by itself. 



The cranium departs but slightly from what w r e find in Brama * is pro- 

 portionately higher and shorter than in Brama, the height here being 

 greater than the length. The lofty fronto-occipital crest so characteristic 

 of Brama, Lampris, Coryphzcna and the majority of the Scombroidea, 

 and which in the first-named attains a height which about equals the 

 diameter of the orbit, is in Pterycombus low and short. It arises here 

 I >i interior to the anterior margin of the orbit in the frontal region (conse- 

 quently completely posterior to the ethmoid) about midway between the 

 first interspinal and the premaxilla. The greatest height of the crest, 

 which is about opposite the middle of the orbit, is only one-fourth the 

 diameter of the latter. This crest is formed almost entirely by the frontals, 

 the supraoeeipital making up but a very small portion of it; that part 

 between the first interspinal and the posterior margin of the supraoeeipital 

 is triangular in outline, membraneous in structure and is found above the 

 occipital bone. The level area, which is seen on the superior aspect of 

 the cranium, which the supraoeeipital and frontals together form, and 

 from which the crest arises, is in Pterycombus relatively short, its entire 

 length being about equal to the orbital diameter. It exhibits but a few 

 rugose lines which radiate from a point directly beneath the center of the 

 supraoeeipital crest, or the frontal crest, and pass directly outward to 

 the margins. In Brama this surface is considerably longer, almost twice 

 as long as the orbital diameter, and exhibits a couple of raised longitudi- 

 nal lines which pass in an anterior direction.! 



Post-temporal, Parker, ( Supra-Scapula, Owen), is bifurcated as in 

 Brama; either fork is broad and fiat, without any elevations or lateral 

 processes at the base of the superior limb. 



*In the specimen of Pterycombus before me, the entire hyoidean apparatus is missing'. 



tThe more I study Professor Collett's skull of Pterycombus (herewith reproduced) 

 and his description of that part of the skeleton in Brama, and compare both with my 

 figures and descriptions of the corresponding structure] in Grammicolepis, the more I 

 am inclined to believe that Pterycombus and Grammicolepis are related genera, witli 

 Brama not very far removed. Further, there seems ground to believe, that in the case 

 of all three of these genera of fishes, each is a representative of a separate family. Two 

 of them have already been established, as the Bramidx and the Grammicolepidx (Poey), 

 and, unless some other naturalist has already done so, it would seem that Pterycombus 

 brama represents a group having similar rank, — that is, the family Pterycombidx. Pro- 

 fessor Poey was of the opinion that the Grammicolepidx came nearer to the Carangidx 

 than any other family known to him at the time he established the former, and I was 

 disposed to concur in his opinion ; but since reading Collett's paper, I am inclined to 

 think otherwise, and adopt what would appear to be a most natural arrangement, or 

 the one above suggested. All three would appear to be related by a variety of charac- 

 ters more or less closely with the Berycidoz; the Balistidx; Acanthuridx, and the Scom- 

 bridx, especially the last-named. 



The skeletal and other characters given above by Professor Collett found in Ptery- 

 combus brama are ample, in my opinion, to justify the establishment of the family 

 Pterycombidx. R. W. S. 



