468 
PACIFIC SCIENCE, Vol. IX, October, 1955 
it appears that: (1) close to a half of the Line 
Island inshore fish fauna does not occur in 
Johnston or Hawaii and that the principal 
reason for this is the great area of deep water 
between the northernmost Line Island reef 
and Johnston; (2) the majority of tropical 
Pacific fishes that have reached Johnston also 
occur in Hawaii; (3) the relatively few species 
of tropical Pacific fishes that are known from 
Johnston but not Hawaii may have reached 
but have not survived in the latter islands 
because of differences in ecological conditions 
between Johnston and the Hawaiian chain. 
What may be termed the southbound fishes 
are now up for discussion. In the first place 
it must be noted that there is a very much 
smaller proportion of the Hawaiian inshore 
fishes restricted to the Hawaiian Islands (F of 
Fig. 4) than of tropical Pacific fishes that do 
not get north from the Line Islands (A of Fig. 
4). Indeed, it is quite certain that there is a 
considerably smaller number of Hawaiian 
"endemics” with the potentialities for moving 
south than of tropical Pacific fishes that 
might move north. Of the former group a 
rather high percentage (though a low number 
of species) have reached Johnston. It is for 
this reason that Johnston is to be considered 
primarily as an outlier of the Hawaiian faunal 
region rather than as a part of the tropical 
Central Pacific fauna. The example drawn 
from the Palmyra and Hawaiian poison sta- 
tions will bring out this point. In the Hawaiian 
rotenone station of 69 species, only 17 are 
restricted to Hawaii, but another 4 are re- 
stricted to Hawaii and Johnston. In the 
Palmyra poison stations of 62 species, 29 are 
not known north of Palmyra but only 1 is 
known from Palmyra and Johnston but not 
Hawaii. On the basis of these figures (fishes 
found both in Hawaii and Palmyra being 
excluded) the Johnston inshore fauna is 4/l7 
Hawaiian and 1 / 29 tropical Pacific. To what 
extent other Hawaiian endemics will turn up 
at Johnston remains to be seen. 
Finally, it is necessary to say something of 
those fishes found today in Hawaii, Johnston, 
and the tropical Pacific (C of Fig. 4). First, 
it seems certain that as more attention is given 
to the fishes in this category more of them 
will prove to show differentiation between 
the Hawaiian and Line Islands. Meanwhile 
there is no sure way of telling whether this 
group has moved north or south via Johnston. 
However, certain points regarding the hypo- 
thesis of Johnston as the original port of 
entry for the Hawaiian fauna may be men- 
tioned. On the one hand, it is certain that the 
Hawaiian inshore fish fauna was ultimately 
derived from that of the tropical Pacific. Fur- 
ther, there is no island that could or does at 
the present time provide a better stepping 
stone between the Hawaiian chain and the 
tropical Pacific fauna. Finally it has been in- 
dicated above that Johnston does at the pres- 
ent time serve as a terminal point for at least 
some northward movement. On the other 
hand, the age of Johnston is unknown, and 
it may be that Johnston is younger than the 
Hawaiian fish fauna as we know it at present. 
If this were so, it would be far easier to ex- 
plain why the following Hawaiian representa- 
tives of tropical Pacific forms rather than the 
latter forms themselves are present at John- 
ston: Muraenichthys cookei , Gymnothorax euros- 
tus, Cirrhitus alternatus, Chromis leucurus, Chae- 
todon multicinctus , and Acanthurus sandvicensis. 
Because of these features it seems best not to 
make categorical statements as to whether (or 
how much of) the Hawaiian fish fauna did or 
did not originally enter via Johnston. It can, 
however, be stated that a more plausible port 
of entry has yet to be found. 
Analysis of Individual Species 
The individual species investigated here are 
Johnston fishes represented by different geo- 
graphic variants in the Central Pacific and in 
the Hawaiian chain. They do not include all 
fishes in this category but only those for 
which sufficient information is available to be 
worth discussing. The following species com- 
plexes will be dealt with (in each pair the 
Hawaiian form is mentioned first): Murae- 
