434 
inhabitants of all islands, with the exception of 
Abaiang, agree that the liver of large sharks can 
be very poisonous. The Abaiang people main- 
tain that their magic prevents sharks from hav- 
ing toxic livers in the same way as it prevents 
their fish from producing ciguatera; they are not 
really believed by other islanders, who ridicule 
them, but not within their hearing. The Mara- 
kei people say that sharks caught at night are 
more likely to have poisonous livers than those 
taken during the day. Arunukans, Abemamans, 
and Butaritarians say that a toxic liver is always 
longer than a nontoxic one and has one lobe 
doubled back underneath itself. Others say that 
a toxic liver is darker than a nontoxic one or 
has dark spots on it. There are the same super- 
stitions as there are for toxic fish; i.e., flies will 
not settle on a toxic liver, a silver coin is black- 
ened, or grated coconut turns green when cooked 
with a toxic liver. 
Several people died on Beru in 1957 after 
eating some tiger shark liver; again in I960 and 
1961 people died on Tabiteuea, poisoned by 
liver from the same species. 
Shark flesh has never been reported to be toxic 
in the Gilbert Islands. 
Poisoning by Tetraodont and other 
Plectognath Pishes 
Puffer fishes, species of the family Tetraodon- 
tidae, are rather uncommon in the Gilberts, al- 
though on certain islands in other parts of the 
colony, particularly Fanning Island, they are very 
plentiful. Puffer fish poisoning is well known to 
the Gilbertese who, on the whole, do not espe- 
cially like to eat these fish. Nevertheless, there 
are many people who do eat them; and they say 
that, provided the fish is skinned and gutted and 
the ovaries are removed immediately while the 
fish is still alive, the flesh will not be poisonous. 
Sometimes, despite these precautions, a puffer 
may still be deadly. On Maiana a large blue 
puffer fish, probably Lagocephalus lagocephalus 
(Linnaeus), which was taken at sea on a bait for 
flying fish, severely poisoned all who ate it, kill- 
ing two, in spite of being "correctly” cleaned 
when it was caught. 
Porcupine fish, Diodontidae, are considered 
far too toxic to risk eating. The toxin is thought 
by the Gilbertese to be concentrated in the 
PACIFIC SCIENCE, Voh XVIII, October 1964 
ovaries, and they say that "even if only one egg 
is broken” the fish will be deadly. Sharp back 
puffers, Canthigaster spp., are also known to be 
deadly poisonous, but these little fish are too 
small to be used as food; Canthigaster solandri 
( Richardson ) is the only species of any interest 
to the Gilbertese, as this species is used for 
fighting in the same way as are the Siamese 
fighting fish. 
Trigger fish, Balistidae, are eaten whenever 
caught, but the majority are not particularly 
liked. Large specimens of Pseudobalistes flavi- 
marginatus (Ruppell) or Rhine canthus acule- 
atus ( Linnaeus ) are considered to be more tasty 
than other kinds of Balistidae. These fish do not 
appear to be toxic all the time, and the Gil- 
bertese say that they have always eaten them 
with impunity. Trigger fish may be very poison- 
ous, however, in a ciguatera-producing area, but 
only when the outbreak of poisoning is at its 
height. 
File fish, Monocanthidae, are not at all liked 
as food fish, but are occasionally eaten and do 
not appear to cause poisoning. The flesh of 
Aleutera scripta (Osbeck) is described as being 
rather bitter, but this fish was eaten in times of 
famine and does not have the reputation of 
being poisonous. 
Hallucinatory Mullet Poisoning 
The heads of certain species of mullets, Mu- 
gilidae, and of certain species of surmullets, 
Mullidae, have the reputation of causing a mild 
form of poisoning, described by some Gilbertese 
as a "madness,” by others as a "forgetfulness,” or 
"sleepiness.” It was very difficult to get Gil- 
bertese to admit to any knowledge of this form 
of poisoning. Fuller inquiries showed that the 
heads of certain of these fish were eaten with 
the full expectation and possible enjoyment of 
the hallucinations or dreams which followed. 
Helfrich and Banner ( I960) reported that in 
Hawaii this form of poisoning is restricted to 
certain localities and times of the year. Sufficient 
information was not obtained from the Gil- 
bertese to be able to decide if this was true in 
the Gilberts or not. 
It may be of interest to note that the Gil- 
bertese consider two more species of fish capable 
of causing a "madness” or "forgetfulness” form 
of poisoning, Epinephelus corallicola (Cuvier 
