If we examine the various methods of blood-vengeance which have been adopted by different peoples throughout the ages, we shall find that they may be divided broadly into four groups or categories. Amongst rude and savage races there exists or has existed a system of vengeance which we may describe as a barbarous and unrestricted vendetta. In the absence of any social machinery for the determination of blood-guilt, or for the estimation of its varying degrees, a single deed of blood provokes an endless series of retaliations: a hideous orgy of revenge rages through the land, an orgy which no one may escape; for old men and women and children perish, whether one by one, or in a general massacre. The vengeance is at once collective and hereditary. It strikes at the neighbours and at the most distant relatives of the murderer: it strikes, too, at the children that are born when the murderer has been gathered to his fathers. It ends only when there is hardly anyone left to kill, or when a paltry sum of money is offered to placate a glutted thirst for blood. It is a strange fact that such a system should have survived up to comparatively recent times[1] in the Balkan States. It is generally but, as we hope to show, erroneously maintained that such a system prevailed amongst the earliest inhabitants of Greece about whom we have any certain knowledge. A second mode of vengeance we may describe as a personal restricted vendetta. It is distinguished from the mode which we have just mentioned by the absence of collective or hereditary punishment. It refuses to visit the sins of the father upon his children or upon his neighbours. The right to avenge remains with the relatives of the slain. They may lie in wait for the slayer or, if he flees, they may dog his footsteps over land and sea. But they dare not strike the innocent for the guilty. There is some power, whether of military autocracy, or of public opinion, which prescribes the bounds of their avenging. The system does not generally include a regular tribunal for the trial of homicide, whether because there is little difficulty, in certain social groups, in determining the identity of the murderer: or because some primitive method of evidence, such as the ordeal of medieval Europe, takes precedence of human witnesses: or because a recourse to arbitration, in the private domain of a king or of a squire, is too insignificant a procedure to have found its way into any historical records. It is such a system that seems to have prevailed in Serbia up to very recent times. It is such a system that, we hope to show, existed amongst the Achaean caste in Homeric Greece.