36 
PACIFIC SCIENCE, Vol. XIV, January I960 
FIG. 7. Pseudo gramma, a, Lower portion of left pectoral and pelvic girdles, from outside; b, upper and 
inner portions of right pectoral girdle, from inside and drawn to about half the scale of a. ac, Actinost; cl, 
cleithrum; co, coracoid; pc, postcleithra; pd, pad on cleithrum for articulation with pelvic girdle; pe, pelvic 
girdle; pm, posttemporal; sc, scapula; si, supracleithrum. 
nearly parallel, longitudinally oriented lamellae 
(instead of the usual rosette). Mouth large; 
supramaxillary present. Teeth mostly villiform, 
many of them depressible. No enlarged scales on 
operculum. Upper portion of preopercular bor- 
der usually with one to four projecting spines. 
Two (or one) pyloric caeca. Dorsal spines II to 
VIII. Anal with no or with III spines, in the 
latter instance with the last spine shorter than 
the penultimate. Pelvics I, 5, the inner ray at- 
tached to the abdomen by a membrane. 
The group of genera defined above seems 
to be sufficiently distinctive to warrant family 
recognition. That they are all interrelated is 
indicated by the rather large number of presum- 
ably independent minor characters held in com- 
mon. That they deserve family rather than 
subfamily status is a more open question, and 
the author admits to considerable vacillation on 
this matter. 
The major points in favor of family recogni- 
tion would seem to be the following: (1) The 
genera here attributed to the Grammistidae ap- 
pear to be easily and clearly distinguishable from 
all other percoid fishes. If intermediates exist 
between them and other families, they have 
escaped the author’s notice. Apparently the 
closest relatives are certain members of the 
Serranidae, e.g., Diploprion , with somewhat 
elongate nasal organ and membranes between 
the pelvics and the abdomen. However, in 
other characters Diploprion falls clearly with 
Epinephelus rather than with the grammistids. 
( 2 ) Although the characters differentiating the 
