414 



SCIENCE 



[N. S. Vol. XXVIII. No. 717 



among the spiny rayed fishes, though it is well 

 in keeping with the pneumatic duct to the CESoph- 

 agus, which some of the Berycoid fishes are said 

 to have. The presence of orbitosphenoids is com- 

 mon among the lower forms from the bony Gan- 

 oids up to and including the Salmoids. So far 

 as the author can ascertain, they hitherto have 

 not been foimd in forms more specialized than 

 the last. They have been searched for in vain in 

 the following: Aulopus, Synodus, Esox, Fundulus, 

 Aphredoderus and nearly all of the families of 

 the Hemibranchs, Synentognaths and the Perces- 

 oces." Den gegenuber kann man einwenden, 

 (1) dass das Orbitosphenoid nicht alien niedrig- 

 eren Teleostei eigentliiimlich ist. So fehlt es 

 unter den Malacopterygii bei Osteoglossum, Gon- 

 orhynchus, Chanos,^ Gromeria,* unter den Cohiti- 

 dini bei Cobitis, Misgurnus, Acanthophthalmus,^ 

 und (2) dass ein Orbitosphenoid bei mehreren 

 Formen, die im System hoher als die Salmoniden 

 stehen, bekannt ist. So besitzt nach Vrolik" 

 Aulopus filamentosus (Scopelidae) ein " sehr aus- 

 gedehntes Orbitosph " (p. 270, Taf. XX., Fig. 30). 

 Femer ist dieser Knochen bei Oalaocias^ vor- 

 handen, ebenso unter den Aeanthopterygii : bei 

 Micropterus salmoides,' bei Pomacanthus paru," 

 bei Grammicolepis,^' bei Regaleous.^ Wir finden 

 folglich das Orbitosphenoid unter den Aeanthop- 

 terygii bei den versehiedensten Familien, von so 

 niedrig organisierten wie die Berycidae angefangen 

 bis zu so hohen wie die Trachypteridae. Weitere 

 Forschungen werden wahrseheinlich zeigen, dass 

 dieser Knochen noeh mehr verbrietet ist, als man 

 fruher angenommen hat. 



Eeferring to part 1 of Dr. Berg's conclu- 

 sions I wisli to point out that in my above 



= Eidewood, Proc. Zool. Soc. Lond., 1904, p. 59; 

 op. cit., 1905, p. 485, fig. 140-1, p. 489. 



* Swinnerton, Zool. Jahr. Aht. Anat., XVIII., 

 1903, p. 63, fig. F. 



■^Sagemehl, Morpli. Jahr., XVII., 1891, p. 579. 



" Neid. Arch, fur Zool, I., 1873. 



^ Saplomi: Swinnerton, Zool. Jahr. Ait. Anat., 

 XVIII., 1903, p. 63, fig. G. 



= Centrarehidae, Shufeldt, U. S. Fish Comm. 

 Kept. 1883, XI., 1885, p. 804. 



° Chsetodontidse, Shufeldt, Jour. Morph., II., 

 1889, p. 290, fig. 10. 



'"Zeidje, Shufeldt, Jour. Morph., II., 1889, p. 

 280. 



" Trachypteridae, T. Parker, Trans. Zool. Soc. 

 Land., XII., 1886, p. 12, T. IV., fig. 7, 11. 



quotation I said that an orbitosphenoid was 

 common, in fishes below the Salmonidse, not 

 that it was always present as Dr. Berg ap- 

 pears to have inferred. 



In reference to part 2 of his conclusions 

 I have carefully examined a number of speci- 

 mens of both Micropterus and Pomacanthus 

 and several of their close relatives, and 

 find no orbitosphenoid, nor is there any 

 mention of this element in any other report 

 on these forms. The common yellow perch is 

 a close relative of Micropterus, and since the 

 time Cuvier first took it as an anatomical 

 type of spiny-rayed fishes its skeleton has 

 been described a great many times, but with- 

 out mention of an orbitosphenoid. Dr. Bou- 

 lenger, in the first volume of the second edi- 

 tion of the " Catalogue of Fishes in the 

 British Museum " has worked out the oste- 

 ology of a great nimaber of the basses, perches 

 and sun-fishes, but without finding an orbito- 

 sphenoid. 



Either the internal descending wing from 

 the frontal, or the anterior part of the ali- 

 sphenoid, has been mistaken for this element 

 in Micropterus and Pomacanthus. There is 

 often a slight mark across the alisphenoid 

 which may have been interpreted as a suture 

 dividing the bone into two parts. This ap- 

 pears to have been the ease in Orammicolepis, 

 to judge from the picture, though I have had 

 no opportunity for examining the skeleton. 

 As to Aulopus, my specimen (A. japonicus) 

 certainly has no orbitosphenoid, though the 

 picture of A. filamentosus published by Vrolik 

 shows a well-developed one. It seems prob- 

 able that somewhere there has been a mis- 

 identification of material. Dr. Berg has mis- 

 interpreted Dr. Swinnerton's statement in 

 regard to Galaxias. Swinnerton referring to 

 this genus, says, " owing to the absence of an 

 orbitosphenoid [etc.]." On the preceding 

 page, however, he gives a figure of the cra- 

 nium of Galaxias in the anterior part of the 

 orbital cavity of which is a portion marked 

 " OS." To " OS " I find no reference in the 

 text, though it may as well refer to a carti- 

 laginous or membranous orbital septum as 

 to an ossified orbitosphenoid. 



