in August, the Saturnalia in December.

But the reasons

von Domaszewski gives for the arrangement, and the further speculation that where it does not occur we may find traces of an older system, as yet unaffected by the so-called Pythagorean prejudice, do not seem to me satisfactory. We may be content with the general principle as I have stated it, and note that while religious duties must be performed on days of odd number, civil duties were not so restricted: the days belonging to the gods, which were, so to speak, taboo days, were more important than those belonging to men. There are, as I have said, but two days marked in the large letters as festivals, which are on days of even number, 24th February and 14th March, the Regifugium and the second Equirria; and about these we know so little that it is almost useless to speculate as to the reason for their exception from the rule. Two others, 24th March and 24th May, were partly the property of the gods and partly of men, and are marked QRCF (quando rex comitiavit fas); but the sense in which they partially belonged to the gods is not the same as in the case of sacrificial festivals.

This calendar thus shows obvious signs of both military and political development; in other words, its witness to the religious experience of the Romans proves that they had successfully adjusted the forms and seasons of their worship to the processes of government at home and of military service in the field. But the most conspicuous feature in it is the testimony it bears to the agricultural habits of the people-to the fact that agriculture and not trade, of which there is hardly a trace, was the economic basis of their life. At the time when it was drawn up, the Romans must have been able to subsist upon the ager Romanus, though, as we shall see later on, it was probably not long before they began commercial relations with other peoples; for their food, which was almost entirely vegetarian, and their clothing, which was entirely of wool and leather,15 they depended on their crops,

flocks, and herds; and the perils to which these were liable remain for the State, as for the farming household, the main subject of the propitiation of the gods, the main object of their endeavours to keep themselves in right relation with the Power manifest in the universe.

We can trace the series of agricultural operations in the calendar without much difficulty all through the year. The Roman year, we must remember, began with March, and March, as we have seen, had under the military necessities of the State become peculiarly appropriated to the religious preparation of the burgher host for warlike activity. But the festivals of April, when crops were growing, cattle bringing forth young or seeking summer pasture, all have direct reference to the work of agriculture.16 At the Fordicidia, on the 15th, pregnant cows were sacrificed to the Earth-goddess, and their unborn calves burnt, apparently with the object of procuring the fertility of the corn; and the Cerealia on the 19th, to judge by the name, must have had an object of the same kind, though the supersession of Ceres by the Greek Demeter had obscured this in historical times. The Parilia on the 19th, recently illuminated by Dr. Frazer, was a lustration of the cattle and sheep before they left their winter pasture to encounter the dangers of wilder hill or woodland, and may be compared with the lustratio of the host before a campaign. On the 23rd the Vinalia tells its own tale, and shows that the cultivation of the vine was already a part of the agricultural work. On the 25th the spirit of the red mildew, Robigus, was the object of propitiation, at the time when the ear was beginning to be formed in the corn, and was particularly liable to attack from this pest.

17

The religious precautions thus taken in April were not renewed in May; but at the end of that month of ripening the whole of the ager Romanus was lustrated by the Fratres Arvales. This important rite, for some reason which we cannot be sure of, was a movable feast, left to the discretion of the brethren, and therefore does not

appear in the calendar. In June the sacred character of the new crops, now approaching their harvest, becomes apparent; the penus Vestae, the symbolic receptacle of the grain-store of the State, after remaining open from the 7th to the 15th, was closed on that day for the rest of the year, after being carefully cleansed: the refuse was religiously deposited in a particular spot. Thus all was

made ready for the reception of the new grain, which, as is now well known, has a sacred character among primitive peoples, and must be stored and eaten with precaution.18 This was the chief religious work of June; in July, the month when the harvest was actually going on, the festivals are too obscure to delay us; they seem to have some reference to water, rain, storms, but it is not clear to me whether the object was to avert stormy weather during the cutting of the crops, or, on the other hand, to avert a drought in the hottest time of the year. The true harvest festivals begin in August; the Consualia on 21st and Opiconsiva on 25th both seem to suggest the operation of storing up (condere) the grain, and between them we find the Volcanalia, of which the object was perhaps to propitiate the fire-spirit at a time when the heat of the sun might be dangerous to the freshly-gathered crops.

After the crops were once harvested, ploughing and sowing chiefly occupied the farming community until December; and as these operations were not accompanied by the same perils which beset the agriculturist in spring and summer, they have left no trace in the calendar. Special religious action was not necessary on their behalf. It is not till the autumn sowing was over, and the workers could rest from their labours, that we find another set of festivals, of which the centre-point is the Saturnalia on the 17th, Saturnus being the deity, I think, both of the operation of sowing and of the sown seed, now reposing in the bosom of mother earth.19 A second Consualia on the 15th, and the Opalia on the 19th, like the corresponding August festivals, seem to be concerned with the housed

grain harvested in the previous August; I am disposed to think that in all three we should see not only the natural rejoicing after the labours of the autumn, but the opening of the granaries and, perhaps, the first eating of the grain. For on the Saturnalia there was a sacrifice at Saturnus' altar, followed by a feast, which was afterwards Graecised, but doubtless originally represented the primitive feasting of the farm, in which the whole familia took part. This brings us practically to the end of the agricultural year as represented in the calendar; for spring sowing was exceptional, the joyful feasts of pagus and compitum are not to be found in our document, and the month of February is specially occupied with the care and cult of the dead (Manes).

At this point I wish to notice one or two results of the adoption of a religious calendar such as I have been describing, which are more to the purpose of these lectures than some of the details I have had to point out. First, let us remember that agricultural operations necessarily vary in date according to the season, and that most of the rural festivals of ancient Italy were not fixed to a particular day, but were feriae conceptivae, settled perhaps according to the decision of some meeting of heads of families or officers of a pagus. That this was so we may conjecture from the fact that those which survived into historical times, e.g. Compitalia and Paganalia, and were celebrated in the city, though not as sacra pro populo,20 were of varying date. But all the festivals of the calendar were necessarily fixed, and the days on which they were held were made over to the gods. Now by being thus fixed they would soon begin to get out of relation to agricultural life; just as, if the harvest festivals of our churches were fixed to one day throughout the country, the meaning of the religious service would sooner or later begin to lose something of its force. And how much the more would this be so if the calendar itself, from ignorance or mismanagement, began to get out of relation with the true season, as in course of time was frequently the

case? When once under such circumstances the meaning of a religious rite is lost, where is its psychological efficacy? In the life of the old Latin farmer, as we saw, his religion was a reality, an organic growth, coincident at every point with the perils he encountered in his daily toil; here, in the City-state, it must from the beginning have had a tendency to become an unreality, and it ended by becoming one entirely. Some of the old rites may have attached new meanings to themselves; it is possible, for example, that beneath the military rites of March there was an original agricultural significance; the Saturnalia became a merry mid-winter festival for a town population. But a great number wholly lost meaning, and were so forgotten or neglected in course of time that even learned men like Varro do not seem to have been able to explain them. The only practical question about them for the later Romans was whether their days were dies fasti or nefasti or comitiales,—what work might or might not be done on them.

Another point, closely connected with the last, and tending in the same direction, is that such a calendar as this implies rigidity and routine in religious duties. A well-ordered city life under a strong government must, of course, be subject to routine; law, religious or civil, written or unwritten, forces the individual into certain stereotyped ways of life, subjects him to a certain amount of wholesome discipline. The value of such routine to an undisciplined people has been well pointed out by Bishop Stubbs, in writing of the effect of the rule of the Norman and Angevin kings on the English people," where it was also a religious as well as a legal discipline that was at work. In neither case was it the ignorant and superstitious routine of savage life, which of late years we have had to substitute for old fancies about the freedom of the savage; it is the willing obedience of civilised man for his own benefit. But if it means a routine of religious rites which are beginning to lose their meaning; if the relation between them and man's life and work is lost;