5. Both in Northern tradition and in the Homeric poems the practice of cremation was associated with the belief in a common home for the souls of the dead. This practice does not appear to have been common even in the earlier part of the Viking Age; and consequently the tradition probably comes down from the Heroic Age, at which time we know that cremation was widely prevalent. The cheerless home of Hades[628] differs considerably from Valhalla, though there is some ground for suspecting that the Greeks of the Heroic Age had once cherished a belief endowed with greater vitality. But both conceptions possess certain essential features in common, namely the removal of the soul to a distant place—a belief really incompatible with the local worship of heroes or manes—and the fact that this distant place of souls was a universal home and not reserved for the souls of one tribe. This belief again was doubtless an anti-tribal force of considerable importance.
Briefly then we may define as the predominant characteristic of heroic religion, both Greek and Teutonic, the subordination of chthonic and tribal cults, which as a rule go together, to the worship of a number of universally recognised and highly anthropomorphic deities—coupled with the belief in a common and distant land of souls. These characteristics, at all events among the Teutonic peoples, seem properly to have belonged only to the religion of the royal and military classes. Hence, when the royal families are converted to a new faith, as in England, or when kingless states grow up, as in Iceland, we find in all cases more or less of a reversion to the more primitive forms of religion. It is on the same principle that I would account for the differences in religion between heroic and historical Greece.
FOOTNOTES:
[586] These are the three gods mentioned in the solemn oath which, according to Landnámabók, IV 7 (Hauksbók), had to be sworn on the sacred bracelet at all legal proceedings: hialpi mér svá Freyr ok Niörðr ok hinn almáttki Áss, etc. In the later Melabók (a compilation of the seventeenth century) it is suggested that Áss here means Othin; but I do not think this explanation is generally accepted. It is scarcely credible that Thor should be ignored on such an occasion.
[587] The description in Grímnismál, str. 23, curiously recalls what is said of Egyptian Thebes in Il. IX 383 f. The nearest approach to Valhalla to be found among Northern kingdoms is Ibn Fadhlan's account of the Russian court; cf. Frähn, l.c. (p. 367, note).
[588] Frey's connection with Sweden appears in Saxo's History (frequently) as well as in sagas, but not in the Edda.
[589] The full form, Yngvifreyr or Ingunarfreyr, is clearly connected with Ingwina frea, a title borne by the king of the Danes in Beowulf; cf. p. [367] and note.
[590] The two conceptions are sometimes confused, e.g. in Völuspá, str. 34. But the eschatological conception involved by the story of Balder is that of the 'house of Hel'; and there can be no doubt that this conception itself is ancient, although the description of Hel in Gylf. 34 is probably quite late.
[591] The text does not say (as is stated in several works on Northern mythology) that Thor occupied the central position, but merely that he was robbed after Thórgerðr and before Irpa. This is the only mention, so far as I am aware, of a cult figure of Thor in his car—a feature which occurs in Hýmiskviða and Gylfaginning and may possibly have some ethnological significance. It is somewhat remarkable that in the tract Frá Fornióti (in Hrafn's Fornaldar Sögur, II p. 6 f.) the ancestry of Guðbrandr is traced to the giant Thrymr, Thor's antagonist.
[592] Cults peculiar to certain families appear to have been common among the Lithuanians and kindred peoples; cf. Lasicius, De diis Samogitarum (Respublica ... Poloniae, etc.; Leyden, 1642, p. 280): sunt etiam quaedam ueteres nobilium familiae, quae peculiares colunt deos, ut Mikutiana Simonaitem, Micheloviciana Sidzium, Schemietiana et Kiesgaliana Ventis Rekicziovum, aliae alios.