240 
which has long filamentous projections from the 
caudal fin; in this species it was impossible to 
determine how much, if any, of these filamen- 
tous rays were broken; therefore the total length 
was taken from the anterior tip of the upper 
lip to the posterior end of the shortest branched 
•caudal fin ray ) . 
Standard length: the distance from the an- 
terior tip of the upper lip to the posterior end 
of the hypural plate. When necessary the end of 
the hypural plate was determined by dissection 
or probing with a needle. 
Greatest depth: the greatest vertical distance 
of the body at its deepest part. 
Depth of caudal peduncle: least vertical dis- 
tance at the caudal peduncle. 
Snout length: the distance from the anterior 
tip of the premaxillary ( upper lip ) to the front 
edge of the orbit. 
Eye diameter: the greatest horizontal distance 
between the margins of the bony orbit. 
Postorbital length: the distance from the rear 
edge of the orbit to the rearmost tip of the 
fleshy operculum. 
Interorbital width: the least distance between 
the bony edges of the orbits on the dorsal surface 
of the head. 
Head length: the distance from the anterior 
tip of the upper lip to the posterior tip of the 
fleshy operculum. 
Gill rakers: the anterior rightmost arch was 
exposed by dissection. The count includes all 
gill rakers and rudiments on the first arch. 
Fin rays: spines are represented by Roman 
numerals, simple soft rays by lowercase Roman 
numerals, and branched soft rays by Arabic 
numbers. Each ray with a separate base was 
counted. 
Abdominal vertebrae: those vertebrae from 
the head to, but not including, the first vertebra 
with transverse processes that meet midventrally. 
Caudal vertebrae: those vertebrae from the 
most anterior one with transverse processes that 
meet midventrally to, but not including, the 
urostylar vertebra. Vertebral counts were made 
directly from radiographs and were confirmed 
in some instances by dissection. 
Scales in a longitudinal row: the number of 
oblique rows of scales that cross the first length- 
wise row just above the lateral line, counting 
PACIFIC SCIENCE, Vol. XVIII, July 1964 
from the upper angle of the gill opening along 
the body to the base of the caudal rays. 
Scales from the soft dorsal to the lateral line: 
the number of scales in an oblique row running 
down and back from the first soft dorsal ray to 
the lateral line. The lateral line scale is not 
included in this count. Small scales near the 
dorsal fin are counted as whole scales. 
Scales from the anal opening to the lateral 
line: the number of scales in an oblique row 
running up and back from the anterior margin 
of the anus to the lateral line. The lateral line 
scale is not included in this count. 
Scales around the caudal peduncle: the num- 
ber of scales in a zigzag row around the least 
vertical depth of the caudal peduncle. 
DIAGNOSTIC CHARACTERS 
Neither the measurements nor meristic data 
may be used alone to identify species. Although 
most of the characters vary within themselves 
and overlap with those of other species, in 
certain instances there is no variation of a 
character within a species nor overlapping be- 
tween species. 
The color pattern is of great value in identi- 
fying members of this genus when specimens 
are recently and/or well preserved. Very old 
and poorly preserved specimens, even recent 
poorly preserved specimens, tend to fade, and 
in time lose all or most of their color pattern. 
No sexual dimorphism in color pattern was 
observed. 
The mean (y), variance (s 2 ) , and standard 
deviation (s) were obtained for all characters 
involving measurements or count data. In those 
instances where a comparison was made between 
species involving these data, the standard error 
of the difference was calculated in order to 
determine the statistical significance of the 
difference of the sample means. 
GILL RAKERS: The total number of rakers 
varies between species from 9 to 25, but within 
each species the variation is usually 3 or 4 
(Table 6). The upper arch has from 1 to 11 
rakers and the lower from 7 to 15. 
fin RAYS: There is little variation, if any, 
in the fin ray counts. Without exception the 
pelvic fin ray count was I, 5. The caudal fin ray 
formula for branched rays was 8 above and 7 
