Mommsen, Hist. of Rome (E.T.), vol. ii. p. 433.
 Cumont, Les Religions orientales dans le paganisme
romain, p. 36. Cp. Dill, Roman Society in the Last
Century of the Western Empire, p. 63. Gwatkin, The
Knowledge of God, vol. ii. p. 133.
 See some valuable remarks in Lord Cromer's Modern
Egypt, vol. ii. p. 135.
 Since this lecture was written this scholar has
passed away, to the great grief of his many friends; and
I refrain from mentioning his name.
 Ira W. Howerth, in International Journal of
Ethics, 1903, p. 205. I owe the reference to R.
Karsten, The Origin of Worship, Wasa, 1905, p. 2,
note. Cp. E. Caird, Gifford Lectures ("Evolution of
Theology in the Greek Philosophers"), vol. i. p. 32.
"That which underlies all forms of religion, from the
highest to the lowest, is the idea of God as an absolute
power or principle." To this need only be added the
desire to be in right relation to it. Mr. Marett's word
"supernaturalism" seems to mean the same thing; "There
arises in the region of human thought a powerful impulse
to objectify, and even to personify, the mysterious or
supernatural something felt; and in the region of will a
corresponding impulse to render it innocuous, or, better
still, propitious, by force of constraint (i.e.
magic), communion, or conciliation." See his Threshold
of Religion, p. 11. Prof. Haddon, commenting on this
(Magic and Fetishism, p. 93), adds that "there are
thus produced the two fundamental factors of religion,
the belief in some mysterious power, and the desire to
enter into communication with the power by means of
worship." Our succinct definition seems thus to be
 The Golden Bough, ed. 2, vol. i. p. 62.
 Liberal Protestantism, p. 64.
 For religio as a feeling essentially, see Wissowa,
Religion und Kultus der Römer, p. 318 (henceforward to
be cited as R.K.). For further development of the
meaning of the word in Latin literature, see the
author's paper in Proceedings of the Congress for the
History of Religions (Oxford, 1908), vol. ii. p. 169
foll. A different view of the original meaning of the
word is put forward by W. Otto in Archiv für
Religionswissenschaft, vol. xii., 1909, p. 533
(henceforward to be cited as Archiv simply). See also
below, p. 459 foll.
 See, e.g., Frazer in Anthropological Essays
presented to E. B. Tylor, p. 101 foll.
 Staatsverwaltung, iii. p. 2. This will
henceforward be cited as Marquardt simply. It forms
part of the great Handbuch der römischen Alterthümer
of Mommsen and Marquardt, and is translated into French,
but unfortunately not into English. I may add here that
I have only recently become acquainted with what was, at
the time it was written, a remarkably good account of
the Roman religion, full of insight as well as learning,
viz. Döllinger's The Gentile and the Jew, Book VII.
(vol. ii. of the English translation, 1906).
 Two fragments of ancient carmina, i.e. formulae
which are partly spells and partly hymns, survive--those
of the Fratres Arvales and the Salii or dancing priests
of Mars. For surviving formulae of prayer see below, p.
185 foll. Our chief authority on the ritual of prayer
and sacrifice comes from Iguvium in Umbria, and is in
the Umbrian dialect; it will be referred to in
Bücheler's Umbrica (1883), where a Latin translation
will be found. The Umbrian text revised by Prof. Conway
forms an important part of that eminent scholar's work
on the Italian dialects.
 F. Leo, in Die griechische und lateinische
Literatur und Sprache, p. 328. Cp. Schanz, Geschichte
der röm. Literatur, vol. i. p. 54 foll.
 Among Roman poets Ovid is the worst offender,
Propertius and Tibullus mislead in a less degree; but
they all make up for it to some extent by preserving for
us features of the worship as it existed in their own
day. The confusion that has been caused in Roman
religious history by mixing up Greek and Roman evidence
is incalculable, and has recently been increased by Pais
(Storia di Roma, and Ancient Legends of Roman
History), and by Dr. Frazer in his lectures on the
early history of Kingship--writers to whom in some ways
we owe valuable hints for the elucidation of Roman
problems. See also Soltau, Die Anfänge der römischen
Geschichtsschreibung, 1909, p. 3.
 Most welcome to English readers has been Mr. T. E.
Peet's recently published volume on The Stone and
Bronze Ages in Italy, and still more valuable for our
purposes will be its sequel, when it appears, on the
 Roman Festivals, p. 142 foll.; henceforward to be
cited as R.F.
 See Virgil's Messianic Eclogue, by Mayor, Fowler,
and Conway, p. 75 foll.
 Wissowa, R.K. p. 227.
 An account of this in English, with photographs,
will be found in Pais's Ancient Legends of Roman
History, p. 21 foll., and notes.
 Mannhardt, Mythologische Forschungen, p. 72 foll.
 Ibid., p. 156 foll.
 Lectures on the Early History of Kingship,
 Not long after these last sentences were written, a
large work appeared by Dr. Binder, a German professor of
law, entitled Die Plebs, which deals freely with the
oldest Roman religion, and well illustrates the
difficulties under which we have to work while
archaeologists, ethnologists, and philologists are still
constantly in disagreement as to almost every important
question in the history of early Italian culture. Dr.
Binder's main thesis is that the earliest Rome was
composed of two distinct communities, each with its own
religion, i.e. deities, priests, and sacra; the one
settled on the Palatine, a pastoral folk of primitive
culture, and of pure Latin race; the other settled on
the Quirinal, Sabine in origin and language, and of more
advanced development in social and religious matters. So
far this sounds more or less familiar to us, but when
Dr. Binder goes on to identify the Latin folk with the
Plebs and the Sabine settlement with the Patricians, and
calls in religion to help him with the proof of this, it
is necessary to look very carefully into the religious
evidence he adduces. So far as I can see, the limitation
of the word patrician to the Quirinal settlement is
very far from being proved by this evidence (see The
Year's Work in Classical Studies, 1909, p. 69). Yet the
hypothesis is an extremely interesting one, and were it
generally accepted, would compel us to modify in some
important points our ideas of Roman religious history,
and also of Roman legal history, with which Dr. Binder
is mainly concerned.