Ovid has given a far more significant hint, when he tells us that the Lemuria was the more ancient festival of the two. So far we have had no indication of anything approaching divinity in connection with the dead or the underworld as distinct from the earth-goddesses, but the evidence for it, though vague and shadowy, is not wanting. Certain mysterious female deities, Tarpeia, Acca Larentia, Carna, and Laverna, of whom late aetiological myth had its own explanation, have, in all probability, been rightly interpreted by Mommsen as divinities of the lower world: the commemorative 'sacrifice at the tomb,' which we hear of in connection with the first two, was in reality, we may suppose, an offering to a chthonic deity at a _mundus_. A rather more tangible personality is Vediovis, who three times a year has his celebration (_Agonia_ not _feriae_) in the Calendar: he, as his name denotes, must be the 'opposite of Iove,' that is, probably, his chthonic counterpart, a notion sufficiently borne out by his subsequent identification with the Greek Pluto. Finally, of course, there is that vague body, the Di Manes, 'the good gods,' the principal deities of the world of the dead; to them invocations are addressed, and they have their place in the formulae of the _parentalia_ and the opening of the _mundi_.[8] In connection with them, acting as a link with the female deities, we have the strange goddess Genita Mana, the 'spirit of birth and death.' Controversy is acute as to the interpretation of these facts, especially in regard to the question whether or no the spirits of the dead were actually worshipped. I would hazard the following reconstruction of history as consistent with what we otherwise know of Roman religion, and with the evidence before us. From the earliest times the Roman looked upon his dead relations as in some sense living, lying beneath the earth, but capable alike of returning to the world above and of influencing in some vague way the fortunes of the living, especially in relation to the crops which sprung from the ground in which they lay. At first, when his religion was one of fear, he regarded the dead as normally hostile, and their presence as something to be averted; this is the stage which gave birth to the Lemuria. As civilisation increased, and the sense of the unity of household and community developed, fear, proving ungrounded, gave place to a kindlier feeling of the continued existence of the dead as members of household and state, and even in some sense as an additional bond between the living: this is the period which produced the _sacra privata_ and the Parentalia. When the _numen_-feeling began to pass into that of _deus_, in the first place a connection was felt between the spirits of the dead and the deities of the earth associated with the growth of the crops, in the second the notion that the underworld must have its gods as well as the world above, produced the shadowy female deities and Vediovis.