We shall see later that, in the Apolline code, death was probably the invariable penalty for kin-slaying,[118] and there was no ‘purgation’: but in other cases purgation was possible, and in the purgation ceremony an animal was slain. The conclusion which is suggested prima facie by these facts, namely, that at one time human sacrifice was the only purgation for homicide, is not necessarily correct. We believe it is incorrect. We agree with Glotz[119] in deriving the purgation rite from Chthonian sacrifice in its general aspect. In such sacrifice, originally, human beings were probably offered, prior to, contemporarily with, and even subsequent to, the adoption of animal sacrifice. We cannot legitimately assume that the latter supplanted the former. Glotz points out that religion, being conservative, tends to preserve in ritual elements which civilisation has abandoned. Hence arose the mock-rites of human sacrifice which took place in historical times.
The belief that homicide-purgation originated in the sacrificial slaying of the murderer was encouraged by the similarity which existed between the rites of homicide-purgation and the ordinary ritual of Chthonian expiation. We shall see later that, in the ceremonial of purification which was applied to persons guilty of homicide, from the seventh century B.C. onwards, the blood of a slain animal was poured over the hands of the slayer, and allowed to flow away into the sea or into a running stream. Thus, homicide-purgation (καθαρμός) easily came to be regarded as a kind of expiation (ἱλασμός); but it differs fundamentally in meaning from expiation, inasmuch as it is symbolical of the fact that a social or religious obligation has been discharged, rather than of the fact that it is being thereby discharged. The sacrifice of an ox or a sheep or a ram to a god or a ghost was in itself a payment or a retribution. But homicide-purgation (καθαρμός) was never permitted until the slayer had re-established his normal social equilibrium, had suffered the penalty prescribed by law, namely exile, temporary or perpetual, and was ready to resume religious communion with his fellow-men. Since, therefore, homicide-purgation was rather a symbol of reconciliation than a medium of expiation, it was more closely allied to the rites which accompanied the swearing of oaths,[120] the giving of pledges and the making of contracts. The animal on which an oath was sworn could not be eaten: so, too, the pig or the lamb by whose blood a murderer was ‘cleansed’ could not be eaten. Now it is unfortunate that such ceremonies, which were really symbolic of reconciliation, should have been so similar to the general ritual of religious expiation that they could easily be confused. There is a vast difference in meaning between reconciliation and the aversion of evil, yet all these ideas were confused in the general system of Chthonian ritual. As an illustration of this confusion we may cite a passage from Vergil, in which is described a rite which is really an ‘aversion of evil,’ a kind of purgation by anticipation. Urging the farmer to be religious in the interest of his crops, he says[121]:
cui[122] tu lacte favos et miti dilue Baccho,
terque novas circum felix eat hostia fruges.
The milk, honey and wine here mentioned are the characteristic offerings in the placation of ghosts.[123] The rite was easily transferred to Demeter or Ceres, the Chthonian goddess, because of the natural tendency of Chthonianism to identify the ghost with the god. The ceremony of carrying a victim round the crops was not a symbol of atonement for moral guilt so much as an aversion of quasi-physical evil spirit which caused sterility.
Athenaeus,[124] describing the ‘purgation’ of an Arcadian city which was necessitated by the visit of certain citizens from a town which was polluted by bloodshed, says: ‘They made purgation of the city, carrying “victims” round the city territory.’ The similarity of this ceremony to the ‘aversion’ rite described by Vergil is obvious. Yet this ceremony is somewhat different from the purgation of an actual homicide, which we shall describe more fully later.[125] In the former a number of victims are slain; in the latter, only one. Now, if homicide-purgation originated in human sacrifice, and if, as Müller maintains,[126] wergeld was suggested to men by the de facto acceptance, on the part of the gods, of an animal substitute, why was the number of animals sacrificed in homicide-purgation limited to one? Why did men not offer to the gods at least the saraad or insult-price,[127] which generally consisted of a number of animals? The sacrifice of only one animal in such a ceremony cannot be explained by Müller’s hypothesis. It can, however, be made intelligible if we assume a direct derivation of the rites of homicide-purgation from the ritual which accompanied solemn oaths and reconciliations. In such a ritual, only a single victim was slain: its death was a kind of inductive symbol of the fate of its slayer, if he ever proved false to his oath. But in ceremonies of general purgation, such as Athenaeus describes, there was an element of expiation, or aversion, and hence there was no limit to the number of victims, for there was no such limit in expiatory sacrifice of any kind.
We shall see later how, in historical times, purgation for homicide was inadmissible in cases of kin-slaying, unless the dying man forgave; even then the slayer had to be exiled for one year before he could be purged in his homeland: in cases of wilful murder, purgation of the slayer in his own country was impossible at any time, but was possible, if not compulsory, abroad: in cases of manslaughter, purgation could take place at home when the conditions of exile and of the ‘appeasement’ of the slain man’s relatives had been fulfilled. From such regulations we can obviously infer that purgation was a symbol of reconciliation, but not an expiation of guilt.
The Homeric and the Tragic Erinnys
We must now contrast what we may call the Homeric Erinnys with the Erinnys of post-Homeric times and with the ‘tragic’ Erinnys. In the course of our discussion we hope to suggest some reasons, more satisfactory, even if they be more complex, than that which Müller[128] gives, for the refusal of the Erinnyes in the Oresteian legends of Attic tragedy to recognise the purgation of Orestes until they assume the rôle of Semnai Theai or Eumenides. In our view there are just two reasons for this refusal: one is the fact that the purgation-rites for homicide were a symbol of reconciliation, not with ghosts, but with gods: the other is the fact that the Erinnyes of Attic tragedy are a complex product, reflecting the attitude of the relatives of the slain at different periods, and from different points of view, in the post-Homeric era. We shall see later that there must have been several different variants of the Oresteian legend. The act of Orestes would have been approved or condemned according as social custom, at any given epoch, recognised the right of Apollo to command or to justify in advance the slaying of Clytaemnestra, or the right of a State court to approve, or at least to condone, an act which tribal society would have probably condemned.