parallel—another case in which a lamp is thought to have done duty for a real fire. There was in old time a custom, to which several ancient writers refer[1], of keeping a lamp burning both day and night in the Prytaneum or in the chief temple of a Greek city; and both Athens and Tarentum are said to have had these lamps so constructed that they could hold a supply of oil sufficient to last a whole year. Such lamps, it has been suggested[2], represented the fire on the city's hearth which was not allowed to go out. The purpose of the lamp was clearly not to give light—for then it need not have been kept burning by day as well as by night—but it was a labour-saving appliance for keeping the sacred fire ever burning. The small flame was in fact a rudimentary fire. Thus all that I am supposing is that a lamp could represent a real fire just as well at the tomb as in the Prytaneum.
If then my explanation of the modern custom is right, the fact that the common-folk, though they have for many centuries employed inhumation as the ordinary Christian rite, have clung at the same time to a ceremonial form of cremation which they still connect in some way with the dissolution of the buried corpse, is additional proof of the favour with which the quicker and surer rite was formerly, and perhaps here and there still is, regarded.
Thus then the study of ordinary funeral-usage has confirmed the conclusions drawn in preceding chapters from the study of a certain abnormal state of after-death existence. As incorruptibility was the greatest bane to the dead, so dissolution was the greatest boon that the living could give them. This dissolution was to be effected by one of two methods, cremation and inhumation, which in theory were alternative but in practice were frequently combined. The combination of them was due in the first instance to the amalgamation of two races to which they respectively appertained; but in later times the racial difference between the two rites was obliterated, and they were judged on their own merits, with the result that a preference for cremation manifested itself in funeral-usage. This preference was due to a recognition that cremation was a quicker and surer method of dissolution, and is itself strong testimony to the desire to effect dissolution. The end to which both rites were directed was the same, but since one led
