tives
everywhere by day and night, crying in doleful tones, "Oh, how cold! oh,
how cold!" Hence when the body of a dead kinsman was lost because he had
been drowned at sea or slain on a battlefield, some of his relatives
would go down to the seashore or away to the battlefield where their
friend had perished; and there spreading out a cloth on the ground they
would pray to some god of the family, saying, "Oh, be kind to us; let us
obtain without difficulty the spirit of the young man!" After that the
first thing that lighted on the cloth was supposed to be the spirit of
the dead. It might be a butterfly, a grasshopper, an ant, a spider, or a
lizard; whatever it might be, it was carefully wrapt up and taken to the
family, who buried the bundle with all due ceremony, as if it contained
the body of their departed friend. Thus the unquiet spirit was believed
to find rest. Now the insect, or whatever it happened to be, which thus
acted as proxy at the burial was supposed to be the _ata_ or shadow of
the deceased. The same word _ata_ served to express likeness; a
photographer, for example, is called _pue-ata_, "shadow-catcher." The
Samoans do not appear to have associated the soul with the breath.[150]

[150] G. Brown, _Melanesians and Polynesians_, pp. 170
_sq._, 218 sq.; G. Turner, _Samoa_, pp. 150 _sq._; S. Ella,
"Samoa," _Report of the Fourth Meeting of the Australasian
Association for the Advancement of Science, held at Hobart,
Tasmania, in January 1892_, pp. 641 _sq._; J. B. Stair, _Old
Samoa_, p. 184. According to Brown and Stair the ceremony
described in the text was observed when a man had died a violent
death, even when the relatives were in possession of the body,
and in that case the insect, or whatever it might be, was buried
with the corpse. I have followed Turner and Ella in supposing
that the ceremony was only observed when the corpse could not be
found. As to the fear of the spirits of the unburied dead, see
also W. T. Pritchard, _Polynesian Reminiscences_, pp. 58 _sq._,
151.

They attributed disease and death to the anger of a god, to the agency
of an evil spirit, or to the ghost of a dead relative who had entered
into the body of the sufferer. Epilepsy, delirium, and mania were always
thus explained by the entrance into the patient of a god or demon. The
Samoan remedy for all such ailments was not medicine but exorcism.
Sometimes a near relative of t