20 
PACIFIC SCIENCE, Vol. XX, January 1966 
though the larger mussels could obviously accom- 
modate them. Second, the very large mussels, i.e., 
over 85 mm in length, seldom contain a crab. It 
is easy to see, as did Christensen and McDermott 
(1958:161), why small mussels only rarely con- 
tain relatively larger crabs. The reverse, how- 
ever, is not so easy to explain, i.e., why are 
smaller crabs seldom found in relatively larger 
mussels? Also, why is it rare that few crabs, of 
any size, are found in the very large mussels 
above 85 mm in length, the latter in spite of the 
fact that such mussels are relatively abundant? 
The occurrence of the early postplanktonic 
stages has been noted only in the mantle cavity, 
and not throughout the entire water-conducting 
system as noted for P. ostreum by Christensen 
and McDermott (1958:173). Therefore, it is 
not felt that such crabs have been overlooked in 
the microscopic examination of the larger host 
bivalves. 
Because the very small prehards and invasive 
stages (1.0 mm or less in carapace width) are 
most frequently found in the smaller mussels 
(15 mm or less in shell length), it is thought 
that the newly moulted true first crab stage, 
when settling out of the plankton, usually selects 
a mussel of this size for its host. Figure 5 in- 
dicates that 80% of the true first stage crabs are 
found in mussels ranging from spat to 20 mm 
in valve length. The mechanism affecting this 
selection is not definitely known, but at least two 
possibilities might be suggested. 
It is possible that the smaller mussels use 
qualitatively or quantitatively different food ma- 
terials than do the larger mussels. Further, it is 
quite possible that the invasive and other pre- 
hard instars require a similar size or type food 
and are thus obliged to infest initially the 
smaller, immature mussels. Small crabs which 
fortuitously find their way into compatible small 
mussels would survive, while those crabs that 
infest the larger mussels would not. The fact 
that occasionally small, invasive stage crabs are 
found in larger mussels (Fig. 5) could be ex- 
plained by regarding these crabs as in a transient 
situation, in which the mussels have only re- 
cently become infested by the invasive crab, 
which would soon be eliminated. The waste 
which would accompany this elimination is an- 
other example of the normal larval or juvenile 
"wastage” noted by Thorson (1950:3) in so 
many marine forms. 
The second suggestion involves the hypo- 
thesis that the crabs are attracted, selectively, to 
the smaller mussels by a "host factor.” Recent 
studies by Davenport ( 1950, 1953*2, b) , Johnson 
(1952), Hickok and Davenport (1957), and 
Sastry and Menzel ( 1962 ) indicate that in cer- 
tain cases of symbiosis the commensal or para- 
site is indeed attracted to the host by a diffusible 
factor from the host. The same factor may be 
used to maintain the relationship once it has 
been established. Preliminary work recently car- 
ried out by Davenport (personal communica- 
tion) at the Friday Harbor Laboratories did not 
reveal evidence to support the existence of any 
such mechanism between F. subquadrata and its 
host, M. modiolus. It is important to note, how- 
ever, that these experiments were conducted only 
with the Stage I and older instars; and it is quite 
possible, in fact probable, that such an inter- 
action might be found only between the invasive 
stage and its host, and perhaps even at a par- 
ticular time during the instar’s existence. 
Johnson (1952) reported that a chemotaxis 
existed between the pinnotherid, Dissodactylus 
mellitae, and the echinoid, Mellita. His work 
with two other pinnotherid species, however, 
did not reveal the existence of any attractive 
mechanism between them and their hosts. He 
suggested that the chemotaxis between Disso- 
dactylus and Mellita acted to enable the pinno- 
therid to maintain a continuous relationship 
with the host in an environment (heavy surf) 
in which they might readily become separated. 
In the Pinnixa chaetopterana-Chaetopterus re- 
lationship, as well as the Pinnotheres-0 strea 
relationship, both of which he studied, it was 
suggested that the negative evidence for the ex- 
istence of a host factor might be the result of 
using the experimental devices with a stage of 
the crab’s life cycle which is not attracted to 
the host. It might well be that other stages do 
respond. 
The recent study by Sastry and Menzel 
( 1962 ) , while indicating that Pinnotheres macu- 
latus is attracted both to the bay scallop, Aequi- 
pecten irradians concentricus , and the penshell, 
Atrina rigida, makes no mention of the stage of 
the females used in the experiments. They do 
