240 
faster than did the nauplius, which can be ex- 
plained by the apparent increase in body volume 
and the decrease in relative number and length 
of setae. Copepodites were kept up to 72 hours 
in culture. After about 10 hr they began to prod 
the bottom and sides of the dish with the an- 
terior end of their body. Inasmuch as the co- 
pepodite is the attachment stage, the prodding 
may have some relation to the attachment of 
the animal to its host. 
As was mentioned earlier, attachment was 
observed in only one copepodite. Attachment 
was accomplished by hooking the strong, claw- 
like terminal processes of each second antenna 
into the substrate. The formation of the attach- 
ment filament was not seen in the single copep- 
odite that attached in the laboratory. To de- 
termine the means by which the filament was 
formed, several copepodites were collected that 
had attached to the host and formed a filament. 
These animals were separated from the host by 
severing the filament close to the body of the 
copepodite. A piece of loosely woven nylon 
cloth was then placed in a finger bowl and the 
detached copepodites placed in with it. Those 
specimens that reattached did so by holding 
themselves to the cloth substrate by means of 
their maxillipeds and second antennae. They 
then placed their body at an angle to the sub- 
strate, with the anterior end in close proximity 
to it, and, by jabbing the substrate with the 
middle of the frontal region, they appeared to 
secure either the remaining attachment filament 
or a new filament formed by a secretion that 
hardened almost immediately on the substrate 
as an irregular lobe, not the disk found in co- 
pepodite and chalimus specimens normally at- 
tached to the host. After the filament was at- 
tached, the terminal processes of the maxillae 
were used to stroke the filament between the 
frontal region and the substrate while the co- 
pepodite backed off from the point of attach- 
ment. The backing off of the animal appeared 
to cause the materal secreted by the frontal 
organ, either before or after the animal started 
backing off, to be drawn out into a filament and 
the stroking by the maxillae to make the fila- 
ment more even. 
After the attachment, the activity of the co- 
pepodite decreased. This observed lower activity 
rate was noted for the first four chalimus stages 
PACIFIC SCIENCE, Vol. XVII, April 1963 
although activity did increase to some extent 
during the actual shedding of the old cuticle in 
moulting. 
Heegaard (1947:90-94) indicates that the 
various attached stages of Caligus curtus secrete 
the attachment filament anew each time a moult 
occurs. During the period of moulting observed 
in both copepodite and chalimus stages in L. 
dissimulates , the original filament remained at- 
tached to the animal, the cuticle rupturing on 
the anterior dorsal surface and the animal, by 
vigorous movements of its body, freeing itself 
from the old cuticle and passing it posteriorly 
over the posterior end of the body. The shed 
cuticle did not remain attached to the host or 
to the frontal filament. Heegaard (1947: fig. 
32) figures a portion of the fin ray of a cod 
with several attachment filaments of Caligus cur- 
tus hanging from it and (1947:92) suggests 
that these are left by the various chalimus stages. 
The only attachment filaments that were ob- 
served in host tissue and could be definitely tied 
in with a chalimus were those from specimens 
of the fifth and sixth chalimus stages of L. dis- 
simulatus that had just broken free and assumed 
the free-moving adult type of existence. 
Gurney (1934:184, 186, 188) indicates that 
the attachment filament is increased in length 
slightly at each moult as indicated by the small 
annuli at the proximal end of the attachment 
filament of later chalimus larvae. This could not 
be verified or disproved for L. dissimulatus, as 
the attachment filament is irregular at the proxi- 
mal end at all stages of development, from the 
copepodite to specimens of the sixth chalimus 
that are still attached. The mean length of the 
filament during each stage is given below in 
Table 6 and suggests that a small amount of new 
material is added although the lack of specimens 
in some of the stages does not provide substan- 
tial proof of this. 
Shiino (1959:305) reports chalimus larvae 
and young stages of L. dissimulatus from the 
gills of Bodianus diplotaenia captured in the 
Revilla Gigedos Islands and includes several fig- 
ures of these stages. This report is of interest 
because of the place of attachment of the chal- 
imus larvae. With only two exceptions, all of 
the chalimus larvae of Hawaiian specimens of 
L. dissimulatus were found attached to the up- 
per surface of the buccal cavity of the host. The 
