thesis of each other.* Here literary documents fail us, but comparative philology sheds a light that can hardly be called dim. By this light we can perceive that there are fewer gods than in the Vedic age, but more than had existed prior to the departure of the European branches. The elaboration and increased importance of the worship, the appearance of a professional priesthood, the rise of new gods like SomaHaoma, Mitra-Mithra, and other things indicative of growth in religious doctrines and rites, can be discovered from a comparison of the names and words existing at this period with those common to the Indo-European family as a whole, while the absence of gods afterwards well known, of ceremonies and castes raised at a later period to prime importance, can be ascertained from a comparison of the Iranic-Indian deities, religious terms and rites, with those of the Vedas. § The process of simplification thus continues; the younger the Polytheism the fewer its gods.

But behind the Homeric poems, and the Vedas,

Lassen, "Indis. Alterthums.," p. 617; Spiegel, "Erânische Alterthumsk.," i. 489.

Spiegel, "Erânische Alterthumsk.,” pp. 432 ff.

Spiegel, ut supra. Some excellent materials for such a comparison can be found in Fick's "Vergleich. Wörterb. der Indoger. Sprachen," ii. Wortschatz.

§ Muir's "Sanskrit Texts," i. pp. 289-295, where views of Dr. Martin Haug bearing on this point are stated.

and the separation of the Iranic-Indian branches, lies the period when Celt and Teuton, Anglo-Saxon and Indian, Greek and Roman, Scandinavian and Iranian, lived together, a simple single people. And at this point comparison can be again instituted. The germs of many subsequent developments in arts and institutions can here be discovered; but the one thing sought, meanwhile, is, What can be determined as to the religious faith then held? The points of radical and general agreement are few. Resemblances that may be classed as coincidences evolved in the course of subsequent development, must, of course, be excluded. Under this head many of the points comparative mythology seizes may be comprehended. The same faculties in men of the same race, working under different conditions indeed, but with kindred materials, could hardly fail to produce similar results. The most of those Myths of the Dawn which Max Müller has so ingeniously analyzed and explained;* gods of the stormful sky, like the German Wodin and the Indian Rudra; gods of the sea, like the Indian Varuna in his later phase, and the Greek Poseidon; gods of the sun, like the Indian Savitri and Surya and the Greek Heliosare, whatever their mythical resemblances, developmental coincidences, creations of the Aryan genius, nationalized yet retaining its family features. Excluding, then, the

"Science of Language," ii. lect. xí.

coincidences natural to related peoples developing the same germs, we find two points of radical and general agreement-the proper name of one God, and the term expressive of the idea of God in general. The name is the Sanscrit Dyaus, the Greek Zeus, the Latin Ju in Jupiter, the Gothic Tius, the Anglo-Saxon Tiw, the Scandinavian Tyr, the old German Ziu or Zio. On this point scholars are agreed. Sanskritists like Dr. Muir* and Professors Müller,† Aufrecht, and Lassen, § Greek scholars like Curtius || and Welcker,¶ German like Jacob Grimm,** and Celtic like M. Adolphe Pictet,++ unite in tracing the cognates back to a common root, and, therefore, to a primitive name. A name for God had thus been formed before the dispersion. It remained the name, too, of the Supreme Deity of the Greeks and Romans. A distinguished Sanskritist supposes Dyaus to have been before the rise of Indra the highest God of the Indian, as well as of the other Indo-Europeans, ‡‡ and

*

"Sanskrit Texts," vol. v. p. 33.

+ "Science of Language,” ii. pp. 425 ff.

‡ Bunsen's "Christianity and Mankind,” vol. iii. p. 78.

S "Indis. Alterthumsk.," i. 891.

|| "Grundzüge der Griech. Etymol.,” p. 222 (3rd ed.).

¶ "Griech. Götterlehre," vol. i. pp. 131 f.

**

"Deut. Mythol.," vol. i. p. 175.

++ "Les Origines Indo-Européennes,” vol. ii. pp. 663 ff.

++

++

Benfey, "Orient und Occident," vol. i. pp. 48, 49, note; Muir's "Sanskrit Texts," v. pp. 118, 119, where the greater part of Benfey's note is translated, and the similar views of M. Michel Bréal stated.

his supremacy may have extended into the period of the Indian and Iranian unity.* The German scholar most distinguished for research in the mythology of his own land, thought he had discovered traces of the original supremacy of Tius or Zio among the Teutonic tribes;+ and a brilliant philologist has generalized these facts and opinions, and argued that Jupiter was the supreme Indo-European God.‡

Perhaps it is too much to argue that the general eminence and prevalence of this name proves the supremacy of the God it designated. Two inferences, however, may be meanwhile allowed-(1) that the word in its primitive form was the name of a deity, (2) that the deity it denoted was acknowledged and worshipped by the Indo-European family as a whole. Let us turn, before attempting any more definite deduction, to the term. expressing the idea of God in general. This term is in Sanskrit deva, in Zend daeva, in Greek Oeós (?), § in

* Spiegel, "Erânische Alterthumsk.," p. 436.

+ Grimm, "Deut. Mythol.," vol. i. pp. 77 ff.

Müller, "Science of Language," ii. lect. x.

§ Skt., deva, Zend, daeva, Pers., dew, Lat., deus, Lith., déva-s, Old Prus., deiwa-s, Old Ir., dia, Gen., déi, Cym., dew, Armor., doué, Corn., deu, Old Nor., tiva-r, are certainly cognates, but there is by no means the same certainty as to θεός. The current of philological opinion, once strongly in favour of identifying its root with that of deva and deus, seems now to have set as strongly against it. Bopp ("Compar. Gram.," i. pp. 4 and 15), Lassen ("Indis. Alterthumsk.,” i. p. 755), Grimm ("Deut. Mythol.,” i. p. 176), Welcker ("Griech. Götterl.," i. p. 131), Pictet ("Les Origines Indo-Europ.," ii. p. 653),

Latin deus, in Lithuanian déra-s, old Prussian deiwa-s, old Irish dia. The very existence of such a term is remarkable.* It indicates that the united Indo-Europeans had advanced so far in religious thought as both to form and formulate a conception of God. Names may express perceptions of sense or presentations of imagination, but general terms imply more or less practised powers of comparison and judgment, abstraction and generalization. But why had the general term come into use? In the sphere of theological thought, if the theology be an

Max Müller ("Science of Lang.," ii. pp. 405, 454), make deva, deus, and eós cognates. But Curtius ("Grundzüge der Griech. Etymol.," Pp. 222, 466-473), G. Bühler ("Orient und Occident," i. pp. 508 ff.), Mr. Peile ("Introduct. to Greek Etymol.,"), Fick ("Vergleich. Wörterbuch," pp. 96, 368), hold @eós to have no connection with devadeus. Their objections appear to me to be valid. The Greek and the Latin d do not correspond. Curtius is uncertain as to the etymology of Oebs, but supposes it may be from a root @eo, whence beσ-σá-μevol, which he had connected in his first and second editions with the Latin festus, festum, festivus, but not in his third, doubts having been started by the objections of Corssen and Pott as to the correctness of his earlier view. Fick derives it from a word dhaya, from a root dhî, to shine, to look, to be devout ("Vergleich. Wörterb.,” pp. 368, 102). If the latter etymology be correct, the word coincides in meaning with deva-deus. Then there is a significant and appropriate progress in the meaning of the word. The primary sense of this root is to shine (scheinen); then to look at, contemplate (schauen) what shines; then finally, what results from the contemplation, to be devout (andächtig sein). The difference of root thus only leads back to identity of meaning, while it helps to show how the contemplator became the worshipper.

* Max Müller, "Hist. Ancient Sans. Lit.," p. 527. "Words like deva for God' mark more than a secondary stage in the grammar of the Aryan religion."