1904.] OSTEOLOGY OF THE ELOPIDiE AND ALBULID^. Gl 



from the pro-otic floor of the cranium by the eye-muscle vacuity. 

 The character is one which is very difficult of application ; and it 

 is a matter of individual opinion whether such a form as Clupea 

 is to be regarded as having a simple or double basis cranii, for 

 here the parasphenoid is produced backward into a pair of large 

 lateral wings, the space between which is freely open below ; and 

 again, to attempt to disci-iminate, as Cope does, between Noto- 

 pterus and Osteoglossum by the former having a " double basis 

 cranii " and the latter a " simple basis cranii" is futile. 



Since the terms posterior and lateral temporal grooves (or fossae, 

 as the case may be) have not always been employed in the same 

 sense in connection with the cranium of Teleostean fishes, it may 

 be well to explain that in this paper the prefix " posterior tem- 

 poral " is applied to that groove or fossa which lies immediately 

 external to the epiotic bone, and the " lateral tempoi-al " to that 

 depression which lies posterior to, and sometimes also above, the 

 postfrontal bone, just above the anterior part, or the whole, of the 

 articulation between the hyomandibular and the cranium. This, 

 I believe, is the most genei-ally accepted usage of the expressions. 

 The former space is occupied b}^ the trapezms and trunk-muscles 

 (Vetter, Jena. Zeitschr. xii. 1878), the latter by the dilatator 

 operculi and other muscles. The posterioi- tempoi'al groove is 

 roofed over to form a posterior temporal fossa in Arapaima, 

 Osteoglossum, Albula, Elops, Megalops, and Ghanos. 



In Clupea, Dassumieria, Chato'essus, Chirocentrus, and Eugrau- 

 lis thei-e is an aperture — the " temporal foramen " — in the side of 

 the cranium, bounded by the parietal and frontal bones. This 

 in life is occupied by a fatty mass, and in the dried skull leads 

 directly from the posterior temporal groove to the cavum cranii. 

 A short distance behind this is a lateral depression — the " pre- 

 epiotic fossa " — situated immediately in front of the epiotic bone, 

 and bounded by the parietal, squamosal, and epiotic. In Coilia 

 there is a small aperture immediately above the most dorsal part 

 of the upper of the two swim-bladder vesicles, which may possibly 

 correspond with the temporal foramen of Clupea and its allies, but 

 the relations of the parts are rather aberrant. Even in Engraxdis 

 the pre- epiotic fossa is largely obliterated by the bulging of the 

 squamosal vesicle. The bottom of the depression is comjDOsed of 

 cartilage in Dussumieria and in Clupea harengus, as also in 

 Osmerus, where, too, a shallow pre-epiotic fossa is recognisable. 

 In Hyodon and in Coregonus pollan, on the other hand, there is 

 no fossa, biit a similar tract of cartilage is present, bounded by 

 the parietal, squamosal, and epiotic bones. 



The large apertui'e — the " latei-al cranial foramen " — in the side 

 of the cranium of Notoptents and the Mormyridse is bounded by 

 the squamosal, epiotic, and exoccipital *, and may possibly be 



* It is not bounded in Notopterus by the postfrontal and squamosal as stated 

 by Boulenger (Poissons du Bassin du Congo, 1901, p. 115, and Ann. Mag. Nat. Hist. 

 1904., vol. xiii. p. 164), nor in the Mormyridee by the opisthotic and parietal (Poiss. 

 Bass. Congo, p. 49). 



