Fish Fauna of Johnston Is. — GOSLINE 
465 
Gymnothorax gracilicauda 
Gymnothorax meleagris. 
Gymnothorax moluccensis 
Gymnothorax pictus 
Gymnothorax undulatus 
Rahula fuscomaculata 
Uropterygius fuscoguttatus 
Uropterygius tigrinus 
Aulostomus chinensis 
Fistularia petimha 
Holocentrus tiere 
Holocentrus microstomas 
Holocentrus lacteoguttatus 
Holocentrus sammara 
Holocentrus spinifer 
Holotrachys lima 
Myripristis argyromus 
Apogon erythrinus 
Apogon snyder i 
Pseudamiops gracilicauda 
Pseudogramma polyacanthus 
Priacanthus cruentatus 
Kyphosus vaigiensis 
Mulloidichthys auriflamma 
M ulloidichthys samoensis 
Parupeneus hifasciatus 
Parupeneus chryserydros 
Parupeneus pleurostigma 
Paracirrhites bimacula 
Abudefduf imparipennis 
Abudefduf sordidus 
Plectroglyphidodon johnstonianus 
Bodianus bilunulatus 
Cheilinus diagrammus 
Cheilinus rhodochrous 
Gomphosus tricolor 
Gomphosus varius 
Novaculichthys taeniourus 
Pseudocheilinus sp. 
Pseudocheilinus octotaenia 
Stethojulis axillaris 
Thalassoma lutescens 
Thalassoma purpureum 
Chaetodon auriga 
Chaetodon citrinellus 
Chaetodon ephippium 
Chaetodon ornatissimus 
Chaetodon quadrimaculatus 
Chaetodon trifasciatus 
Chaetodon unimaculatus 
Zanclus cornutus 
Acanthurus achilles 
Acanthurus e Ion gat us 
Acanthurus olivaceus 
Ctenochaetus strigosus 
Naso lituratus 
Zebrasoma flavescens 
Bathygobius fuscus 
Gnatholepis anjerensis 
Exallias brevis 
Cirripectus variolosus 
S alar i as gibbifrons 
Brotula townsendi 
Neomyxus ch apt alii 
Polydactylus sexfilis 
Bothus mancus 
Rhinecanthus aculeatus 
Melichthys buniva 
Melichthys vidua 
Amanses sandwichiensis 
Ostracion lentiginosum 
Arothron meleagris 
Canthigaster jactator 
Diodon bystrix 
For purposes of assessing the importance 
of Johnston as a filter bridge for species com- 
ing up from the south the following group- 
ings have been made (see Fig. 4): (A) those 
fishes that never reached Johnston, (B) those 
that got to Johnston but no farther, (C) those 
that apparently passed through Johnston on 
the way from the Line Islands to Hawaii or 
vice versa, and (D) those that apparently by- 
passed Johnston. Similarly, the Hawaiian spe- 
cies may be divided into (F) those that never 
reached Johnston, (E) reached Johnston and 
stopped, (C) passed through Johnston, and 
(D) by-passed Johnston. 
The stringency of the Johnstonian filtering 
effect on northbound and on southbound 
fishes will be shown by the relative strengths 
of each of the above categories (except D). 
An attempt to quantify each of these relative 
to one another is therefore made in Figure 4 
by means of the widths of the columns. 
Widths of columns B, C, E, and G are based 
directly on the relative number of Johnston 
species in each of these categories in the list. 
Column D is given no width, because it is 
impossible to know how much of column D 
is represented but as yet uncollected at John- 
ston. Widths for A and F were estimated in a 
very simple and admittedly imperfect fashion, 
and indicate only rough magnitudes. Two 
shallow-water poison stations run at Palmyra, 
the nearest island to the south of Johnston, 
by Mr. J. E. King, et al . , in approximately the 
same way as those made at Johnston yielded 
(among the same fish groups used here) 62 
species of which 29 are not known from 
Johnston or Hawaii. Thus, the number of 
species in these two collections (and pre- 
