ased on
Usener's theory of Sondergoetter. It is ingenious and
imaginative, but in my view does not square with the
facts as far as we know them. His stages are: (1)
momentary function of _numina_, _e.g._ lightning; (2)
elevation of this into a permanent power or function;
(3) consequent limitation of the numen to a special
well-marked function; (4) elevation of the numen to a
_deus_, conceived in the likeness of man, and male or
female, because man cannot think of power otherwise than
on the analogy of male or female creative energy.
Lastly, when the _deus_ is complete, the functions of
the former numen become attributes or qualities, traces
of which we find in the pairs of deities in Gellius,
xiii. 23, which are discussed later on in this lecture.
Some of these, of course, eventually became separate
deities--Salacia, Maia, Lua. As I cannot accept the view
that the earliest Roman idea of the supernatural is to
be found in _comprecationes_ of a comparatively late
period, _i.e._ in the so-called Indigitamenta, this
charmingly symmetrical account has no charm for me
beyond its symmetry.

[299] Henzen, _Acta Fratr. Arv._ pp. 144, 146; Cato,
_R.R._ 139; _C.I.L._ vi. 110 and 111. Other references
are given by Wissowa, _R.K._ p. 33, note 2.

[300] For Pales, _R.F._ p. 80 note; for Pomona, Wissowa,
_R.K._ p. 165.

[301] The passage runs thus (Aug. _C.D._ iv. 32):
"Dicit enim (Varro) de generationibus deorum magis ad
poetas quam ad physicos fuisse populos inclinatos, et
ideo et sexum et generationes deorum maiores suos (id
est veteres credidisse Romanos) et eorum constituisse
coniugia." There is an amusing passage in Lactantius, i.
17 (_de Falsa Religione_), which Dr. Frazer might read
with advantage. It begins, "Si duo sunt sexus deorum,
sequitur concubitus." Then he goes on mockingly to argue
that the gods must have houses, cities, lands which
they plough and sow, which proves them mortal. Finally
he takes the whole series of inferences backwards,
finishing with "si domibus carent, ergo et concubitu. Si
concubitus ab his abest, et sexus igitur foemineus,"
etc. All this, he means, can be inferred from the fact
that gods are of both sexes; but that they have
_concubitus_ can no more be inferred from his argument
than that they plough and sow.