Lastly, the same kind of feeling which added Parentalia to Lemuria developed the vague general notion of the Di Manes, not the deified spirits of the dead, but peaceful and on the whole kindly divinities holding sway in the world of dead spirits, yet accessible to the prayers of the living. The dead, then, were not themselves worshipped, but they needed commemoration and kindly gifts, and they had in their lower world deities to whom prayer might be made and worship given. FOOTNOTES: [5] It is right to state that there is a totally different theory, according to which the Lares were the spirits of the dead ancestors and the Lar Familiaris an embodiment, as it were, of all the family dead. [6] It is significant that even when the dead were cremated, one bone was carefully preserved in order to be symbolically buried. [7] We may note that, though it is a state festival, our information is solely of rites in individual households. [8] Their mention in sepulchral inscriptions dates from the time of the Empire, when a new conception of their nature had sprung up. CHAPTER VI WORSHIP OF THE FIELDS The life of the early Roman in the fields, his activities, his hopes and fears, are reflected in the long list of agricultural festivals which constitute the greater part of the celebrations in the Calendar, and follow closely the seasons and occupations of the agricultural year. We are, of course, in the Calendar dealing, to speak strictly, with the worship of the state, and not with the semi-private festivals of groups of farmers, but in many instances, such as the Robigalia, the state seems only to have taken over the cult of the farmers, preserving carefully the site on which the celebration took place; in others, such as the Terminalia and the Parilia, it seems to have established, as it were, a state-counterpart of a rite performed independently at many rustic centres: in both cases we are justified in inferring the practice of the early Roman agriculturalist. We shall see that in most cases these festivals are associated--though often loosely enough--with the worship of a particular divinity. Sometimes, however,--as in the case of the Lupercalia--it is very difficult to discover who this divinity was; in other festivals, such as the Robigalia, it looks as if the eponymous deity was a comparatively late development. We may, therefore, suppose, on the analogy of what we have already seen to be the general lines of development in Roman religion, that the festivals in origin centred round a purpose rather than a personality, and were addressed 'to all spirits whom it might concern'; and that later, when the _deus_ notion was on the increase, they either attached themselves to some god whose personality was already distinct, as the Vinalia were attached to Iuppiter, or 'developed' a deity of their own. Among these deities, strictly functional as a rule and existing only in connection with their special festival, we shall