Arnobius (v. 155) fortunately mentions that this
story came from the second book of Valerius Antias,
whose bad reputation is well known. It was plainly meant
to account for the cult-title of Jupiter Elicius, and
the origin of the procuratio fulminis, and was
invented by Greeks or Graecising Romans at a time (2nd
century B.C.) when all reverence for the gods had
vanished as completely as in Greece. Yet Dr. Frazer
writes of Numa as "an adept at bringing down lightning
from heaven" (Early History of Kingship, p. 204).
 On this subject, the evolution of the knowledge of
God, I may refer to Professor Gwatkin's Gifford
Lectures of 1904-5, published by Messrs. T. & T. Clark,
 The meaning of deus is well put by Mr. C. Bailey
in his sketch of Roman Religion (Constable & Co.), p.
 Guesses can be made about these, but little or
nothing is to be learnt from them to help us in this
 I adhere to what was said in R.F. p.
312 foll. We do not know, and probably never shall know,
the original deity concerned in that festival. The
ritual is wholly unlike that of the rustica Faunalia
(R.F. p. 256 foll.). I believe that it dates from a
time anterior to the formation of real gods--possibly
from an aboriginal people who did not know any. (I am
glad to see this view taken in the latest summary of
German learning on this subject, Einleitung in die
Altertumswissenschaft, by Gaercke and Norden, vol. ii.
p. 262.) At the moment of printing an interesting
discussion of the Lupercalia, by Prof. Deubner, who
treats it as a historical growth, in which are embodied
ideas and rites of successive ages, has appeared in
Archiv (1910, p. 481 foll.). See Appendix B.
 Wissowa, R.K. pp. 170 and 250 foll.
 Strabo, p. 164. Cp. Usener, Götternamen, p. 277,
whose comment is, "Die Götter aller dieser Stämme waren
'namenlos,' weil sie nicht mit Eigennamen sondern durch
Eigenschaftsworte benannt wurden. Für einen
griechischen Reisenden vorchristlicher Zeit waren sie
nicht fassbar." Arnobius iii. 43, Gellius ii. 28. 2 are
good passages for the principle. The latter alludes to
the anxiety of veteres Romani on this point, "ne alium
pro alio nominando falsa religione populum alligarent."
Hence the formulae "si deus si dea," or "sive quo alio
nomine fas est nominare," Serv. Aen. ii. 351;
"quisquis es," Aen. iv. 576. See also Farnell,
Evolution of Religion, 184 foll.; Dieterich, Eine
Mithrasliturgie, p. 110 foll.
 Serv. Aen. ii. 351. I am inclined to think it is
only an inference from the want of substantival names in
so many Roman deities; surely, it would be argued, the
pontifices must have had some reason for this. It is
contradicted by the fact that in such ancient formulae
as that of the devotio (Livy viii. 9) the great gods
are called by their own names, though the army was in
the field and in presence of the enemy. There was,
however, an old idea that the name of the special
tutelary god of the city was never divulged, lest he
should become captivus, and that the true name of the
city itself was unknown; see Macrob. iii. 9. 2 foll. I
believe that these ideas were encouraged by the
pontifices, but were not founded on fact.
 For the Indigitamenta see below, p. 159; R.F. p.
341; R. Peter's able article in Myth. Lex., s.v.
Scholars do not seem to me to have reckoned sufficiently
with the tendency of a legal priesthood, devoted to the
strict maintenance of religious minutiae, to elaborate
and organise the material for god-making which was
within their reach. To judge by the elaboration of the
ritual at Iguvium, the same tendency must have existed
in other kindred Italian communities, both to develop
ritualistic priesthoods, and through them to elaborate
the ritual. This is, I think, the weak point of Usener's
reasoning in his Götternamen, and as applied to Roman
deities it is the weak point of an interesting article
by von Domaszewski, reprinted in his Abhandlungen zur
röm. Religion, p. 155 foll.
 The best account of Tellus is in Wissowa, R.K.
p. 159 foll.
 R.F. p. 71; Ovid, Fasti, iv. 631 foll. This
was a festival of the populus as a whole, and also of
each Curia, like the Fornicalia in February. Both were
clearly agricultural in origin, though the Curia as we
know it was probably an institution of the city. I must
own that I am quite uncertain as to what the thing was
which was originally meant by the word Curia; my friend
Dr. J. B. Carter may have something to say on the
subject in his book on the Roman religion in the Jastrow
 Dieterich, Mutter Erde, pp. 11 and 73 foll.
 Virg. Aen. iv. 166, "prima et Tellus et pronuba
Iuno Dant signum"; commenting on which Servius wrote,
"quidam sane etiam Tellurem praeesse nuptiis tradunt;
nam et in auspiciis nuptiarum invocatur: cui etiam
virgines, vel cum ire ad domum mariti coeperint, vel iam
ibi positae, diversis nominibus vel ritu sacrificant."
There is little doubt that Tellus is frequently
concealed under the names of Ceres, Dea Dia, etc. For
Ceres and Juno in marriage rites, see Marquardt,
Privatleben, p. 49.
 See below, p. 206 foll.; Macrob. iii. 9. 11;
Deubner in Archiv, 1905, p. 66 foll.
 See De Marchi, La Religione, etc., i. p. 188
and reff. (The reference to Gellius should be iv. 6. 7,
not iv. 67.) Like some other operations of the Roman
religion, this became a form, and was used as a kind of
insurance, whether or no there had been any omission;
Wissowa, R.K. p. 160.
 That Ceres represented the fructus is shown by
the fact that in the XII. Tables the man who raided a
field of standing corn at night was made sacer to her;
Pliny, N.H. xviii. 12.
 Cato, R.R. 134. De Marchi, op. cit. p. 135.
Janus, Jupiter, and Juno are concerned in this rite,
Ceres coming last. Varro has preserved the part of
Tellus for us: "quod humatus non sit, heredi porca
praecidanea suscipienda Telluri et Cereri, aliter
familia non pura est" (ap. Nonium, p. 163).
 The verses are quoted by Dieterich, Mutter Erde,
p. 75, among others from Buecheler's Anthology of Roman
Epitaphs, Nos. 1544 and 1476. The story is told in
Suetonius' Life of Tib. c. 75, and again of Gallienus
by Aurelius Victor (Caes. c. 33).
 Marquardt, p. 326, who notes that the Romans
themselves derived the word from filum, a fillet;
e.g. Varro, L.L. v. 84, "quod in Latio capite velato
erant semper, ac caput cinctum habebant filo." Modern
etymologists equate the word with Brahman.
 Thus the Flamen Quirinalis sacrificed at the
Robigalia, R.F. p. 89, and with the Pontifices and
Vestals took part in the Consualia, Marq. 335.
 We may note here that the most general Latin name
for a priest was sacerdos, which seems to have
excluded all magic, etc.; it means an office sanctioned
by the State. On the general question of the origin of
priesthood see Jevons, Introduction, etc., ch. xx.,
with whose explanations, however, I cannot entirely
agree. I should prefer to keep the word priest for an
official who sacrifices and prays to his god. In this
view I am at one with E. Meyer, Geschichte des
Altertums, i.^2 p. 121 foll. God and priest go together
as permanent, regular in function, and entrusted by a
community with certain duties.
 Marquardt, p. 180; Wissowa, R.K. p. 427. The
popa or victimarius is seen in many artistic
representations of sacrifice, e.g. Schreiber, Atlas
of Classical Antiquities, plate xvii. figs. 1 and 3.
 Jevons, ch. xx.; Frazer, G.B. i. 245 foll., and
Lectures on Early History of Kingship, Lectures ii.
 Virg. Aen. viii. 352.
 In a valuable paper in his Gesammelte
Abhandlungen (p. 284) Wissowa says that "personal
conception of deity is absolutely strange to the old
Roman religion of the di indigetes." I believe this to
be essentially true; but my point is that localisation
and ritual prepared the way for the reception of Greek
ideas of personality. The process had already begun in
the religion of the house; but it was not likely there
to come in contact with foreign germs. When Janus and
Vesta, who were in every house (Wissowa, p. 285), were
localised in certain points in a city, they would be far
more likely to acquire personality, if such an idea came
in their way, than in the worship of the family.
 Aug. Civ. Dei, vii. 28, "quem alii caelum, alii
dixerunt esse mundum." Dr. Frazer, citing this passage
(Kingship, p. 286) in support of his view that Janus
was a duplicate of Jupiter, has omitted to notice that
some theorisers fancied he was the universe, which by
itself is enough to betray the delusive nature of this
kind of theological speculation. Varro elsewhere gives
us a clue to the liability of Janus to be exalted in
this unnatural fashion, L.L. vii. 27, "divum deo" (in
the Salian hymn), if this be taken as referring to
Janus, as it may be, comparing Macrob. i. 9. 14. But
this is easily explained by the position of Janus in
prayers; cp. Cic. Nat. Deor. ii. 27. 67, "cum in
omnibus rebus vim haberent maximam prima et extrema,
principem in sacrificando Ianum esse voluerunt." The
phrase "Deorum" or "Divum deus" is indeed remarkable,
and unparalleled in Roman worship; but no one acquainted
with Roman or Italian ritual will for a moment suspect
it of meaning "God of gods" in either a Christian or
metaphysical sense. I shall have occasion to notice the
peculiar use of the genitive case and of genitival
adjectives in worship later on. See below, p. 153 foll.
 Fasti, i. 89 foll.; R.F. p. 281 foll.
 Frazer, l.c. (a page of which every line appears
to me to be written under a complete misapprehension of
the right methods of research into the nature of Roman
gods); A. B. Cook, Classical Review, vol. xviii. 367
foll.; Professor Ridgeway, Who were the Romans? p. 12,
where, among other remarkable statements, Janus is
confidently said to have been introduced at Rome by the
Sabine Numa, and therefore to have been a Sabine deity,
an assumption quite irreconcilable with those of Dr.
Frazer and Mr. Cook. In striking contrast with such
speculations is a sensible paper on Janus in M.
Toutain's Études de mythologie et d'histoire, p. 195
foll. (Paris, 1909).
 Dr. Frazer is aware of this; see his Kingship,
p. 285, note 1. See also Roscher in Myth. Lex., s.v.
"Janus," p. 45 foll.
 For the evidence for this and the following facts,
see Roscher's article just cited, or Wissowa, R.K. p.
91 foll.; cp. R.F. p. 280 foll. The cult epithets of
Janus are thus explained by von Domaszewski,
Abhandlungen, p. 223, note 1, "Bei Ianus tritt
regelmässig der Begriff des Wesens hinzu, dessen Wirkung
er von Anfang an bestimmt, so I. Consevius der Anfang
der in Consus wirkenden Kraft, und in derselbe Weise I.
Iunonius, Matutinus," etc. This is reasonable, but it
does not suit with I. Patulcius-Clusius, and I cannot
accept it with confidence at present.
 Roscher, op. cit. p. 34.
 Wissowa, Gesammelte Abhandlungen, p. 284 foll.
 Festus, p. 185.
 It is due to the good sense and learning of Dr.
Roscher; he had previously, when working on the old
methods, tried to prove that Janus was a "wind-god"
(Hermes der Windgott, Leipzig, 1878); but a more
searching inquiry into the Roman evidence, when the
prepossessions had left him which the comparative method
is so likely to produce, brought him to the view I have
explained in outline, which has been adopted in the main
by Wissowa, Aust, and J. B. Carter, as well as by myself
in R.F. The last word about so puzzling a deity can of
course never be said; but if we indulge in speculations
about him we must use the Roman evidence with adequate
knowledge of the criticism it needs.
 This difference between Zeus and Jupiter has been
pointed out by Wissowa, R.K. p. 100; Jupiter stands
for the heaven even in classical Latin literature, as we
 See his papers in the Classical Review, vol.
xvii. 270 and xviii. 365 foll., and in Folklore, vol.
xv. 301; xvi. 260 foll.
 Kingship, p. 196 foll.
 Macrobius i. 15. 14. In historical times a white
victim, ovis idulis, was taken to the Capitol by the
via sacra in procession (Ov. Fasti, i. 56. 588).
Festus says that some derived the term via sacra from
this procession (p. 290); and to this Horace may be
alluding in Ode iii. 30. 8, "dum Capitolium Scandet
cum tacita virgine pontifex."
 R.F. pp. 86, 204.
 R.F. p. 160.
 No doubt Jupiter was specially connected with the
oak, as Mr. Cook has shown with great learning in the
paper cited above, note 36; but at Rome he had an
ancient shrine among beeches, and was known as I.
Fagutalis: Varro, L.L. v. 152; Paulus 87. For I.
Viminalis, see R.F. p. 229.
 See Aust's article "Jupiter" in Myth. Lex. p.
 Aust gives a cut of a coin of the consul Claudius
Marcellus (223 B.C.) dedicating spolia opima in this
little temple, according to the ancient fashion,
supposed to be initiated by Romulus, Livy i. 10.
 Dionys. Hal. ii. 34.
 R.F. p. 230.
 See De Marchi's careful investigation, La
Religione, etc., i. p. 156 foll.; Gaius i. 112. The
cult-title should indicate that the god was believed to
be immanent in the cake of far, rather than that it
was offered to him (so I should also take I. Dapalis,
though in later times the idea had passed into that of
sacrifice, Cato, R.R. 132), and if so, the use of the
cake was sacramental; cp. the rite at the Latin
festival, R.F. p. 96.
 There are distinct traces of a practice of taking
oaths in the open air, i.e. under the sky; of Dius
Fidius, unquestionably a form of Jupiter, Varro says
(L.L. v. 66), "quidam negant sub tecto per hunc
deiurare oportere." Cp. Plutarch, Quaest. Rom. 28;
R.F. p. 138. For the conception of a single great
deity as primitive, see Lang, The Making of Religion,
ch. xii.; Flinders Petrie, Religion of Egypt (in
Constable's shilling series), ch. i.; Ross, The
Original Religion of China, p. 128 foll.; Warneck, Die
Lebenskräfte des Evangeliums, p. 20 (of the Indian
Archipelago). The last reference I owe to Professor
Paterson, of Edinburgh University.
 Serv. Aen. viii. 552, "more enim veteri sacrorum
neque Martialis flamen neque Quirinalis omnibus
caerimoniis tenebantur quibus flamen Dialis, neque
diurnis sacrificiis distinebatur." It is, however,
possible that under the word caerimonia Servius is not
here including taboos, but active duties only.
 See my paper, "The Strange History of a Flamen
Dialis," in Classical Review, vol. vii. p. 193.
 Henzen, Acta Fratr. Arv. p. 26.
 Cato, R.R. 141; Henzen, op. cit. p. 48.
 Frazer, G.B. iii. 123, note 3; R.F. p. 40, for
further examples. It may be worth while to point out
here that the coupling of all farm animals except goats
took place in spring or early summer; Varro, R.R. ii.
2 foll. Isidorus (Orig. v. 33), who embodies Varro and
Verrius to some extent, derived the name Mars from
mares, because in the month of March "cuncta animalia
ad mares aguntur."
 I prefer, with De Marchi, to take Silvanus here as
a cult-title, though we do not meet with it elsewhere;
see La Religione, etc., p. 130 note; but Wissowa,
who has a prejudice against the view that Mars was
connected with agriculture, insists on taking Marti
Silvano as a case of asyndeton, i.e. as two deities.
 See, e.g., Varro, L.L. v. 36, "quos agros non
colebant propter silvas aut id genus, ubi pecus possit
pasci, et possidebant, ab usu salvo saltus nominarunt."
 Cato, R.R. 141. Mars is there invoked as able to
keep off (averruncare) evil influences and to make the
crops grow, etc.; he has become in the second century
B.C. a powerful deity in the actual processes of
husbandry, just as he became in the city a powerful
deity of war. But as he was not localised either on the
farm or in the city, I prefer to think that he was
originally conceived as a Power outside the boundary in
each case, but for that very reason all the more to be
propitiated by the settlers within it.
 See below, p. 235.
 So Wissowa, R.K. p. 131. Cp. R.F. p. 39, note
4. Deubner in Archiv, 1905, p. 75.
 Servius, commenting on line 3 of Aen. viii.
(utque impulit arma) writes: "nam is qui belli
susceperat curam, sacrarium Martis ingressus, primo
ancilia commovebat, post hastam simulacri ipsius,
dicens, Mars vigila." The mention of a statue shows that
this account belongs to a late period. But Varro seems
to have stated that there was originally only a spear;
see a passage of Clement of Alexandria in the fragments
of the Ant. rer. div., Agahd, p. 210, to which Deubner
(l.c.) adds Arnobius vi. 11. Deubner calls this spear
a fetish, which is not the right word if the deity were
immanent in it in the sense suggested by "Mars vigila."
See above, p. 116. If Servius correctly reports the
practice, it must be compared with the clashing of
shields and spears by the Salii, which may thus have had
a positive as well as negative object.
 Livy v. 52.
 Mr. A. B. Cook (Classical Review, 1904, p. 368)
has tried to connect both names with the Greek word
[Greek: prinos], and Professor Conway, quoted by him, is
inclined to lend the weight of his great authority to
the conjecture. Thus Quirinus would be an oak-god, and
Quirites oak-spearmen. We must, however, remember that
Mr. Cook is, so to speak, on an oak scent, and his
keenness as a hunter leads him sometimes astray. One is
a little perplexed to understand why Jupiter, Janus,
Mars, and Quirinus should all be oak-gods (and all in
origin identical as such!). On the other hand, it is
fair to note that the original spear was probably of
wood, with the point hardened in the fire, like the
hasta praeusta of the Fetiales: Festus, p. 101. If
quiris has really anything to do with oaks, it would
be more natural to explain the two words as springing
from an old place-name, Quirium, as Niebuhr did long
ago, and to derive that again from the oaks among which
it may have stood. But I am content to take quiris as
simply a spear, as Buecheler did; see Deubner, op.
cit. p. 76. Since the above was written, the article
"Quirinus" by Wissowa in the Myth. Lex. has appeared.
Naturally it does not add anything to our knowledge; but
Wissowa holds to the opinion that the most probable
derivation of the name Quirinus is from Quirium,
possibly the name of the settlement on the Quirinal; and
compares Q. pater (e.g. Livy v. 52. 7) with the
Reatinus pater of C.I.L. ix. 4676.
 The Nonae Caprotinae (July 7), the day when women
sacrificed to Juno Caprotina under a wild fig-tree in
the Campus Martius, is not known to us except from
Varro. See R.F. p. 178, where (note 8) is a suggestion
that the festival had to do with the caprificatio, or
method of ripening the figs, which Dr. Frazer has
expanded in his Lectures on Kingship, p. 270,
believing the process to be that of fertilisation.
 Classical Review, vol. ix. p. 474 foll. The same
view has recently been taken independently by W. Otto in
Philologus, 1905, pp. 215 foll., 221. It is perfectly
clear that the monthly sacrifice to Juno was the duty of
the wife of the rex sacrorum; a pontifex minor is also
mentioned (Macrob. i. 15. 19).
 Wissowa, R.K. p. 116.
 Ib. p. 114.
 See Ihm's article "Iunones" in Myth. Lex. vol.
ii. 615; Pliny, N.H. ii. 16.
 Dr. J. B. Carter tells me that he has abandoned
this explanation of the evolution of Juno. On the other
hand, von Domaszewski seems in some measure to accept it
(Abhandlungen, p. 169 foll.), when he says that
"similar functions, when exercised by different
numina, can eventually produce a god. Auf diese Weise
ist Iuno geworden." He means that the creative power is
called Juno in a woman, or in a people (Iuno Populonia),
or in the curiae (Iuno Curitis), and that an independent
deity, Juno par excellence, emerges from all these.
But so far I cannot follow him.
 There is no real evidence from purely Roman
sources of this fancied conjugal or other relation, if
we exclude that of the alleged cult of Juno by the
Flaminica Dialis. This has been well seen and expressed
by W. Otto, l.c. p. 215 foll.; see also Classical
Review as quoted above. As we shall see in the next
lecture, Dr. Frazer is much concerned to show that
Jupiter and Juno are actually a married pair, and
consequently he will have nothing to do with my opinion
on this point: Early History of Kingship, p. 214
foll., and Adonis, Attis, Osiris, ed. 2, p. 410,
 Wissowa, R.K. p. 141.
 Festus, p. 106; Macrob. i. 12. 6.
 I have discussed the Vestalia and the nature of
Vesta and her cult in R.F. p. 145 foll. See also
Marquardt, p. 336 foll., and Wissowa, R.K. p. 141
 Ovid, Fasti, vi. 296, says that he had been
stupid enough to believe that there was a statue in the
aedes Vestae, but found out his mistake:--
esse diu stultus Vestae simulacra putavi;
mox didici curvo nulla subesse tholo.
The passage is interesting as showing how natural it was
for a Roman of the Graeco-Roman period to suppose that
his deities must be capable of taking iconic form. For
anthropomorphic representations of Vesta in other places
and at Pompeii, see Wissowa, Gesammelte Abhandlungen,
p. 67 foll.
 See Lanciani, Ruins and Excavations of Ancient
Rome, p. 223 foll. The statues of the virgines
vestales maximae, discovered in the Atrium Vestae, all
belong to the period of the Empire. They are now in the
museum of the Baths of Diocletian.