te origin, as it is addressed
to Dis pater, who only became a Roman deity in 249 B.C.
(Wissowa, _R.K._ p. 257). The interesting feature in
this _devotio_, used at the siege at Carthage, is that
it is not himself whom the commander devotes--the common
sense of the Romans had got beyond that--but the enemy
as substitutes for himself. "Eos vicarios pro me fide
magistratuque meo pro populo Romano exercitibus do
devoveo, ut me meamque fidem imperiumque legiones
exercitumque nostrum bene salvos siritis esse." Thus the
enemy is made the victim, and this is why the only gods
invoked are the Di Inferi, Dis pater, Veiovis, Manes,
while in the older formula it is the gods of Romans and
Latins. Pacuvius in a praetextata called _Decius_ wrote:
"Lue patrium hostili fusum sanguen sanguine" (Ribbeck,
p. 280). This is the language Ennius used before him of
the sacrifice of Iphigenia: "ut hostium eliciatur
sanguis sanguine," where, however, the word _eliciatur_
shows that it is magic. The curious thing in this last
passage is that the parallel passage in the Euripidean
_Iph. in Aul._ (1486) does not suggest magic. Is the
idea Italian? The curse (for such it really is) is to be
witnessed by Tellus and Iuppiter, and the celebrant
points down and up respectively in invoking them, as
also in the _devotio_ of Curtis in the Forum (Livy vii.
6), which was an abnormal _procuratio prodigii_.

[434] Cp. the language used by Livy of the second Decius
(x. 29): "prae se agere formidinem ac fugam ...
contacturum funebribus diris signa tela arma hostium."
For spells or curses of this kind see Westermarck i.
563: a curse is conveyable by speech, especially if
spoken by a magistrate or priest. "Among the Maoris the
anathema of the priest is regarded as a thunderbolt that
an enemy cannot escape." See also Robertson Smith,
_Semites_, p. 434, for the Jewish ban, by which impious
sinners, or enemies of the city and its God, were
devoted to destruction. He remarks that the Hebrew verb
to ban is sometimes rendered "consecrate": Micah iv. 13;
Deut. xiii. 16; and Joshua vi. 26 (Jericho), which
exactly answers to the consecratio of Carthage. For
curses conveyable by sacrifices, as in all the cases I
have mentioned, see Westermarck ii. 618 foll. 624, and
the same author's paper on conditional