280 DU. VV. G. UTDKWOOU ON Til 10 



mouth-border, leaving tlie hiteral parts of the niaxillse to support 

 the t^ides. Each mandibular raiiins consists of a slender hori- 

 zontal part, with a large process upstanding from near the 

 syinphysial end, somewhat as in the case of (rO)iorhi/7icJms, but 

 more aberrant from the normal proportions. The upstanding 

 parts of the two rami together form a semicircle as viewed 

 trom the fi-ont, and support the lower part of the border of the 

 gape. At their symphysial ends the rami are loosely united by 

 fibrous tissue, and each bears near its sympliysial end. a stout? 

 short, conical tootli. These two are the only teeth wdiich the 

 animal possesses ; it is not clear what purpose they serve, for 

 they liave no hard structure opposed to them against which they 

 might bite. 



Tht' quadrate is thrust forward far in advance of its normal 

 position, and the union between the quadrate and mandible is 

 eftected in such a manner that tlie ring of bones surrounding 

 the mouth-iipening is freely movable in every direction. The 

 h\ pohyal is large, and single on each side ; there is no descending 

 process of the second hypobranehial ; the first basibranchial is 

 unossitied. 



Even if one dismisses from consideration such aberrant features 

 as those of the mouth-bones, the preopercular andinteropercular 

 bones, and the suborbital bones, one cannot recognise in the 

 skull of JBhractolcEinus any characters that might be described as 

 " Osteoglossid." On the other hand, one notes that the nasal 

 bones are not incoi'porated into the cranium, the parietal bones 

 are separated, the mesethmoid and subopercular bones show no 

 sign of reduction, tliere is no lateral peg of the parasphenoid for 

 articulation with the hyopalatine arcli, there is no lingual nor 

 entopterygoid dentition, the hypohyal is large, there is no 

 descending process of the second hypobranehial, and the brauchio- 

 stegal rays are as few as three in number. 



Comparison of the skull of JPhractol<smus with that of the 

 ClupeidfE is equally unproductive of positive results, for Phracto- 

 Icemus has no diverticula of the swim-bladder in its pro-otic and 

 squamosal bones, no temporal foramen, no pre-epiotic fossa, na 

 auditory fenestra, no fontanel I e betw^een the anterior ends of 

 the frontal bones, no posterior wings to the parasphenoid, no 

 posterior opening to the eye-muscle canal, and no surmaxilhe. 



It is to be noted, however, that in respect of all tlie characters 

 just mentioned, and in the fewness of the branchiostegal rays, in 



