62 DR. W. G. RIDEWOOD 0\ THE CRANIAL [May 3, 



homologous with the pre-epiotic fossa. If the pre-epiotic fossa of 

 Clupea were to become larger and deeper, the exoccipital would 

 probably form part of its margin. The relation of the supra- 

 temporal bone to the lateral cranial foramen in the one case and 

 to the pre-epiotic fossa in the other, suggests a morphological 

 equivalence. 



The lateral tempoiul groove is faii-ly well roofed over in Alhula, 

 Chanos^ and Heterotis, and pai'tially so in Elops, Arapaima, 

 Osteoglosstom, Hi/odon, and Notopterus. 



In Osteoglossum, Alhula, Elop)s, and Megalops there is a 

 depression of considerable size in the side of the cranium, situated 

 beneath the posterior part of the articular surface for the head of 

 the hyomandibular. This is the " subtemporal fossa " ; it is 

 bounded by the pro-otic, squamosal, exoccipital, and opisthotic 

 bones. The fossa is recognisable in Arapaima, but it is only a 

 shallow depression. According to Hay (Zool. Bull. ii. 1, 1898, 

 p. 30) it is well developed in the Cretaceous genus Xiphactinus, 

 in which it was erroneously described by Cope as a foramen. 



As already shown by Sagemehl (Morph. Jahrb. xvii.) the sub- 

 temporal fossa attains to great dimensions in such Cyprinoid 

 fishes as Cyp>rimi,s, Leuciscus, Labeo, sund Ahramis — it is practically 

 wanting, however, in Cobitis and Botia. In the former Cypri- 

 noids it is situated more posteriorly with respect to the articular 

 facet for the hyomandibular than is the case in the genera 

 mentioned at the beginning of the preceding paragi'aph, and it 

 opens vertically downward instead of downward and outward. 

 It is so large and deep as to be separated from the dorsal surface 

 of the skull by but a single layer of bone (epiotic bone). In these 

 fishes the subtemporal fossse serve for the lodgment of the great 

 muscles which, by pulling up the inferior pharyngeal bones (fifth 

 ceratobranchials), biing the teeth upon those bones forcibly against 

 the callous pad that is carried by the under surface of the basi- 

 occipital bone. 



The subtemporal fossa is not present in the Salmonidse. It 

 appears in the form of a wide but shallow depression in Pro- 

 chilodus, Alestes, and Hydrocyon^ , but the fossa, is unrecognisable 

 in the majority of the Oharacinidae. 



The " auditory fenestra," bounded by the pro-otic, exoccipital, 

 and basioccipital, and leading into the perilymphatic cavity, is 

 present in Chatoessus, Chirocenirus, Clupea, Dussumieria, En- 

 graulis, and Hyodon. In the last-named genus it is traversed by 

 a vertical bar of the pro-otic. 



There are csecal diverticula of the swim-bladder contained in 

 the squamosal and pro-otic bones respectively in Chatoessus, 

 Ckirocentriis, Clupeaf, Dussumieria, Engraulis, and Coilia. Mega- 

 lops has a cavity in its opisthotic, which is probahly occupied by 

 a similar diverticulum, although I have not had an opportimity of 

 tracing the connection with the swim-bladder ; but in Elops there 



* See, for instance, Sagemehl, Morph. Jahrb. x. 1, 1884, pi. 2. figs. 13 & 14. 

 "t" In Clupea sprattus the pro-otic vesicle alone is present. 



