No. 422.] NOTES AND LITERATURE. 161 
The family was evidently a provisional one, and several writers, 
notably Dr. Gill and Dr. Bleeker, have noted the incongruity of its 
members and suggested the removal of certain of its subdivisions to 
other groups. Nevertheless, in default of any study of the skeletons, 
these families have been kept in some sort of association by subse- 
quent writers. The study of the skeletons shows plainly the neces- 
_sity of a complete revision of the assemblage. This has been done 
by Dr. Boulenger. : 
Dr. Boulenger shows first that in Trachinus the scapular fenestra, 
as in the codfishes, lies between the scapula (hypercoracoid) and the 
coracoid (hypocoracoid) instead of piercing the former, as in ordi- 
nary fishes. This character is shared by Notothenia, Chznichthys, 
Parapercis, Eleginops, Harpagifer, Trichonotus, Callionymus, and 
their relatives. All these lack the supraocular lamina, which is 
present in Trachinus. Trichonotus and Callionymus show other 
osteological characters, which separate them as families, although 
not invalidating their general trachinoid relationship. The genus 
Percophis agrees closely in osteology with Trachinus, but the scapular 
fenestra is entirely within the scapula as in the percoid fishes. 
Bembrops, Chimarrhichthys, and Leptoscopus agree with N otothenia 
in the absence of a subocular lamina, but differ in having the fenestra 
within the scapula. These forms show affinities with the Batrachoi- 
dide. Boulenger refers them to Leptoscopide. 
Dactyloscopus, which has reduced ventrals and the pectoral arch 
of Clinus, is regarded as a true blenny, notwithstanding its apparent 
likeness to Uranoscopus. Gillellus, Dactylagnus, and Myxodagnus 
will doubtless go with it. 
Uranoscopus and its allies (Anema, Ichthyscopus, Kathetostoma, 
Ariscopus) have also the scapular fenestra in the scapula, but show 
a number of other osteological peculiarities. They are, however, 
unquestionably trachinoid in general relationship. E 
The group Trachinoidea in Boulenger's view, therefore, includes 
the following families : Trachinidz, Nototheniide, Percophide, Lep- 
toscopide, Uranoscopide, Trichonotidz, and Callionymidz. All 
these agree with the Blenniidz, Batrachoidide, Ophidiidz,, and Gadida 
in which the fin rays are frequently not 
in having jugular ventrals, 
Te uborder of 
of the normal number which is I, 5. For the division or suborder 
fishes thus characterized, containing these families e thei allies, 
Dr. Boulenger proposes to revive the old name “ Jugulares. due 
The remaining genera referred to Trachinidz or to Trachinoidea 
show no real affinity with Trachinus, Callionymus, and Uranoscopus. 
