deity; then between 1600 and 1200 BC a developing ‘‘feudal’’ concept had gradually
transformed the horse-god Hippos into the god Hippios, ‘‘master of the horses,’’
under the influence of the role now played by the harness. The hypothesis essentially
rests on texts that express a fundamental link between Poseidon and the horse. I once
sought to support this theory with a Mycenaean tablet in which L.R. Palmer
explainedi-qoas the name of a god (Jost 1985:283–4). But it seems that this is rather
the name of a man (Lejeune 1958:289). In any case, there is nothing to license the
existence of a horse-god. I had also drawn support for this position, following others,
from the Mantinean legend on the birth of Poseidon: ‘‘When Rhea had given birth to
Poseidon, she laid him down amongst a flock so that he could live with the sheep; the
spring Arne drew its name from the fact that sheep [arnes] surrounded it. Rhea told
Kronos that she had given birth to a horse and gave him a foal to eat in place of her
child [anti tou paidos], as she next gave him a stone wrapped in swaddling clothes’’
(Pausanias 8.8.2).
Fougeres (1898:227) spoke of the ‘‘so-called horse borne by Rhea and the foal presented to Kronos.’’ Thus, in interpreting Rhea’s ruse, he understood that Posei- don had been born in the form of a horse. He concluded that Poseidon had ‘‘the body of a horse’’ and that we were dealing with ‘‘an animal cult of the horse.’’ Along similar lines Festugiere (1944:38) wrote on the births of Zeus and Poseidon: ‘‘These
two parallel cases invite the same explanation: if Zeus replaces a sacred stone, Posei-
don was in origin a horse.’’ If the two cases are parallel, the parallelism consists rather
in the fact that Rhea both times gave birth to a child,pais, that she hid sometimes in a
cave and sometimes among a flock. The ruse consisting of giving Kronos a stone or
a foal takes place later. It cannot be denied, in Poseidon’s case, that the ruse indicates
a very strong bond between Poseidon and the horse, but it does not imply the
existence of a horse-god. The ruse explains the epithet Hippios that the god bore
at Mantinea, without necessarily being linked to an ‘‘archaic’’ conception of the god.
The tale’s degree of antiquity, like its origin (Arcadian or Boeotian?) cannot be
determined. Whatever it was, Arcadian gods were not animals in origin, any more
than Zeus was a sacred stone: like the wolf-god of Mount Lykaion, or Artemis
the bear (Jost, forthcoming (b)), Poseidon the horse should be forgotten.
Let us return to the Thelpusa tale. The metamorphoses of Poseidon and Demeter
are ruses employed by these deities. This is not the token of a divine essence: the
metamorphosis is temporary (in Demeter’s case we think of the metamorphoses of
Thetis when she wanted to escape Peleus and, in Poseidon’s case, of the metamor-
phoses of Zeus when pursuing his beloveds). It illustrates a very strong bond, as we
have said, between Poseidon and the horse. Besides, metamorphosis into a horse is
particularly appropriate to the theme of the violent relationship between Poseidon
and Demeter, as A. Avagianou has stressed (1991:145–63). It is not a matter of a
‘‘sacred marriage’’ (hieros gamos) between two deities, but of a rape. The god
accomplishes his will through constraint. The horse offers an appropriate image for
the wild character of this union. A tempestuous and impulsive character was also
attributed to Pegasus, who made the spring of Hippocrene gush forth with a blow of
his hoof (Strabo C379). For Poseidon, the violence he displays reflects the wild forces
of the chthonic god. For Demeter, the myth reveals two contradictory but comple-
mentary aspects. Her wild nature is embodied by the Poseidonian mare and by her
anger, which inflicts sterility upon the land of Phigalia. Once she is appeased and
The Religious System in Arcadia 277