Taking our position on the fact that change of accent in the vocative in Greek is due to the continued influence of an older system of Aryan accentuation, we now see how the change of nom. Ζεύς into voc. Ζεῦ, and of nom. +Dyaús+, into voc. +Dyaû́s+, rests on the same principle. In Sanskrit the change, though at first sight irregular, admits of explanation. What we call the circumflex in Sanskrit, is the combination of a rising and falling of the voice, or, as we should say in Greek, of an acute and grave accent. As +Dyaús+ was originally +Diaús+, and is frequently used as two syllables in the Veda, the vocative would have been +Díaùs+, and this contracted would become +Dyaus+. Thus we have +paribhvế+ from +paribhûs+. In Greek the facts are the same, but the explanation is more difficult. The general rule in Greek is that vocatives in ου, οι, and ευ, from oxytone or perispome nominatives, are perispome; as πλακοῦ, βοῦ, Λητοῖ, Πηλεῦ, βασιλεῦ, from πλακοῦς, οῦντος, placenta, βοῦς, Λητώ, Πηλεύς, βασιλεύς. The rationale of that rule has never been explained, as far as Greek is concerned. Under this rule the vocative of Ζεύς becomes Ζεῦ; but no Greek grammarian has attempted to explain the process by which Ζεύς becomes Ζεῦ, and nothing remains for the present but to admit that we have in it an ancient Aryan relic preserved in Greek long after the causes which had produced it had ceased to act. It would fall into the same category as εἶμι and ἴμεν. Here, too, the efficient cause of the length and shortness of the radical vowel _i_, viz., the change of accent, Sk. +émi+, but +imás+, has disappeared in Greek, while its effect has been preserved. But whatever explanation may hereafter be adopted, the simple fact which I had pointed out remains, the motive power which changed the nom. +dyaús+ into the vocative +dyaû́s+, is the same which changed Ζεύς into Ζεῦ. Those who do not understand, or do not admit this, are bound to produce, from the resources of Greek itself, another motive power to account for the change of Ζεύς into Ζεῦ; but they must not imagine that a mere reference to a Greek elementary grammar suffices for explaining that process. The passage in the Rig-Veda (VI. 51, 5) to which I referred is unique, and I therefore give it here, though it has in the meantime been most ably discussed by Benfey in his “Essay on the Vocative” (1872). “Dyaû́ḥ pítaḥ pṛthivi mấtaḥ ádhruk Ζεῦ πάτερ πλατεῖα μῆτερ ἀτρεκ(ές) Ágne bhrấtaḥ vasavaḥ mṛláta naḥ[13] Ignis φράτερ ϝέΣηϝες μέλδετε nos.”