Osteology of Kaupichthys diodontus — GOSLXNE 
313 
tb'ys are fused; the orbitosphenoid is short 
and rounded; the otic bulla is little devel- 
oped; the parietals and pterotics are far 
smaller; the suspensorium is somewhat for- 
wardly inclined; and the branchiostegal rays 
are long, fine, and numerous. Still other dif- 
ferences between the two genera are to be 
found in the pectoral girdle and vertebral 
column. 
Fig. 7. a, First four vertebrae, lateral view of 
right side; b, terminal vertebra (cartilage stippled); 
c, seventy-ninth vertebra; d, last four vertebrae plus 
endoskeletal and bases of exoskeietal elements of 
fin rays, ce, Centrum; ee, endoskeletal elements 
of fin rays; fo, foramen; ha, haemal arch; hs, 
haemal spine; na, neural arch; ns, neural spine; 
tp, transverse process. 
Since both Muraenichthys and Kaupich- 
thys have always been included in the Echeli- 
dae, a question of the allocation of the family 
name arises. It is obvious from Regan’s diag- 
nosis (1912: 380 and 386) of his family 
Echelidae that he investigated the Muraenich- 
thys type of eel. On the other hand, from 
the descriptions of Echelus myrus , the geno- 
type of Echelus , given by Richardson ( 1844: 
108), Gunther (1870: 50), and Jordan and 
Davis (1892: 642), it appears that the family 
name Echelidae belongs with Kaupichthys . 
The Muraenichthys type can probably best 
be placed within the Ophichtfayidae, follow- 
ing the suggestion of Myers and Storey 
(1939: 157), as a subfamily, Myrophinae. 
However, the final applications of the names 
Echelidae and Myrophinae (or Myrophidae) 
must await an osteological examination of 
Echelus myrus , a Mediterranean species un- 
available to me. 
The family Echelidae in the sense used 
just above, i.e., limited to the Kaupichthys 
type of eel, shows distinct relationship to the 
Moringuidae and Heterenchelidae in the 
paired frontals, and particularly in the en- 
larged otic bulla. It differs from either of 
the latter families in possessing relatively 
high vertical fins and labial posterior nostrils. 
It resembles the Heterenchelidae but not the 
Moringuidae in the long, narrow orbitosphe- 
noid, in the laminar palatopterygoid, in hav- 
ing the trunk shorter than the tail, and, ap- 
parently, in the movable articulation between 
the first vertebra and the skull. It resembles 
the Moringuidae but not the Heterenchelidae 
in lacking a suture between the ethmoid and 
the vomer. Thus the relationships of Kau- 
pichthys — and provisionally of the Echeli- 
dae— are closest with the Heterenchelidae. 
The fact that Kaupichthys and Muraenich- 
thys belong to different families also raises 
the problem of the family allocation of the 
remaining genera assigned to the Echelidae 
auctorum. Some, such as Garmanichthys, ap- 
pear to belong with Kaupichthys. Others, 
