Accmthums triostegus sandvicensis— Randall 
contained 10,814 large eggs. The total number 
of large eggs in this ovary was therefore esti- 
mated at 40,000. 
DEVELOPMENT 
Fertilization of the eggs of the manini was 
effected artificially, using live fish obtained from 
trap fishermen. As previously discussed, ripe 
female fish are detected by the release of per- 
fectly transparent eggs following gentle pres- 
sure on the abdomen. The milt of running ripe 
male manini oozes even more readily from the 
fish than the eggs of ripe females; however, it 
was found that eggs could be fertilized with 
sperm from males in which definite squeezing 
was necessary to bring forth the milt. Eggs 
were fertilized both by stripping ripe males and 
females concurrently into sea water with genital 
apertures in close proximity and by stripping 
them separately into different containers and 
subsequently mixing the contents. 
The fertilized eggs are spherical, 0.66 to 0.70 
mm. in diameter, and contain a single oil globule 
which is 0.165 mm. in diameter. If suspended 
in sea water, the eggs slowly rise to the sur- 
face. They exhibit no stickiness and do not ad- 
here to objects or one another, although they 
often lie in single-layered groups at the surface. 
Unfertilized (but fertile) eggs are also less 
dense than sea water. No difference in diameter 
of unfertilized and fertilized eggs could be 
detected. 
Developing eggs were placed in large finger 
bowls, an aerated battery jar, and an aerated 20 
gal. aquarium. In all of these containers heavy 
mortality of the larvae ensued because of set- 
tling to the bottom (probably through contact 
with bacteria there and not to any physical in- 
jury). Efforts were made to keep the larvae in 
suspension by placing a fine-mesh false bottom 
in a battery jar and locating a bubbling aerator 
stone beneath a hole at one side and by installing 
a paddle which was slowly turned by an electric 
motor. Although these methods reduced the 
amount of settling, mortality was still high and 
only a few larvae survived to an age of 5 days 
or more, when oriented swimming began. 
In view of the difficulty of rearing larvae in 
rigid containers, a new approach was tried for 
the manini. A fine-mesh bag, 1.2 m. on a side, 
241 
with cork along the top was floated in the Ala 
Wai Yacht Basin. Fertilized eggs were placed 
within and allowed to develop. Approximately 
the same mortality resulted, however. 
Once the manini larvae were capable of ori- 
ented swimming, they were maintained as easily 
in unaerated finger bowls as any other way. At 
this time the critical factor became the supply 
of food. 
The principal stages in early development 
from 0 hr. to 5 days 22 hr. are shown in Figures 
9-12. The drawings of eggs before hatching 
were made shortly after placing the eggs in 10 
per cent acetic acid, a treatment which makes 
the blastodisc opaque white and the cleavage 
pattern more visible. The ages given are based 
on an average development temperature of 
24° C. During development the temperature 
was 24° C. zb about 1°. 
Although temperature control was not suf- 
ficiently precise to permit an assay of the change 
of rate of development with changing tempera- 
ture, it was obvious from the rearing of one 
batch of eggs at 26° C. dz 1°, that temperature 
has a marked effect on the rate. The stage at- 
tained at 28 hr. 40 min. at 24° C. was reached 
in about 22 hr. 30 min. at 26° C. The 42 -hr. 
stage at 24° C. was reached in about 32 hr. at 
26° C. 
No attempt was made to determine the tem- 
perature limits within which normal develop- 
ment of the manini is possible. Due to proximity 
of a light bulb, the temperature of one finger 
bowl in which 37-hr. larvae were developing 
was inadvertently raised in a short time to 29.3° 
C. None of the larvae died at this time. 
The blastodisc of the fertilized egg is 0.4 mm. 
long. One hour after fertilization the first cleav- 
age becomes apparent. At the end of the second 
hour 8 cells are present, and after 4 hr. about 
55 cells can be counted in the blastodisc. 
At the age of 6 hr. gastrulation is under way. 
Epiboly is evident from the thin layer with a 
wavy edge which is extending downward over 
the yolk. The cells are very small and are not 
shown in the drawing of this and subsequent 
stages. 
At 14 hr. the primitive streak is apparent and 
the cephalic region of the embryo well devel- 
oped. By about 18 hr. epiboly is complete, and 
