all fall at periods of three, or a multiple of three, days, weeks, months, or years, from the date of death. These, I think, have been selected in deference to the mysterious virtue of the number three[1], and not improbably multiplied by the importunities of a penurious priesthood, to whom some half-dozen hearty meals in the course of the year do not appear an inappropriate remuneration for their services at death-bed and burial. But the fortieth day was originally devoted to this purpose, it may reasonably be supposed, because it was the last opportunity of setting before the dead man's neighbours and acquaintances savoury meat such as their soul loved, that they might eat thereof and 'loose' the dead man from any curse wherewith in his lifetime they had bound him; if dissolution was not to be retarded, the fortieth day was in popular reckoning the last opportunity for absolution.
From this it should follow that any memorial feasts held later[2] than the fortieth day are of purely ecclesiastical contrivance; and the correctness of this inference is attested by a curious local usage which clearly distinguishes the popular and the ecclesiastical feasts. At Sinasos in Asia Minor two classes of commemorations are recognised. The one is called [Greek: kaniskia], 'little baskets,' from the method in which food is distributed to the poor; this is held on the fortieth day. The other has usurped the name [Greek: mnêmosyna], which commonly belongs to all memorial-feasts, and is held on the three anniversaries of the death (for, after the third, exhumation generally takes place, and no further memorial-feasts are needed) and consists in the presentation of an ornamental dish of boiled wheat ([Greek: kollyba]) at the church and the reading of a service[3]. In other words, the fortieth day is the popular festival, and the observances of later dates are ecclesiastical. Clearly the reason for this distinction must lie in the fact that the common-folk believe, as the song from Zacynthos shows, that dissolution is normally complete by the fortieth day, while the Church has prudently fixed the date, after which exhumation is permissible, at the end of the third year. Presumably then a period of forty days was the old pagan period, for which the Church has tried, with partial success, to substitute three years., p. 82.]
