The methodological or procedural principles of Puhvel 1981 are two: 1. He limits his study of horse myth to "such Indo-European mythic and religious structures as are congruent with a tripartite concept of social and religious organization" (184). 2. The only viable starting-point in already existing scholarship on the subject is a triptych consisting of the Ancient Indic as’vamedha ritual; the Roman October horse; certain Celtic analogues.
It turns out that in the as’vamedha the mahisî or chief queen cohabited with the stallion that was the centerpiece of the ritual (161); the sexual element is lacking in the Roman ritual; in the Celtic ritual, the king mates with a mare (164; cf. Ford 1977).
Puhvel takes the Celtic material as the key to the whole problem and concludes that the Indic ritual has "floundered" (193): while "the king’s role as the patron of a great horse festival persisted, the detail was elaborated and absorbed Pre-Aryan matter, ... the typical Near Eastern ritual union of queen or goddess and beast. The usual bull of these rituals has, however, not supplanted the horse, or there would be no as’vamedha." (193)
Puhvel’s basic point: "The Indo-European pattern of theriomorphic hierogamy was clearly King and Mare, the near Eastern and Aegean one Queen and Bull (e.g., Europa, Pasiphaë in Cretan saga, wife of Archon Basileus in Greek religion, and so on)" (193-94). He concludes: "The Indic as’vamedha is thus a halfway house of transformation." (194)
Puhvel is dismissive of myths concerning Poseidon and Demeter (195-96).
Cardini 1981: 3-110, starting from the earliest times, surveys the long history of the use of the horse for military purposes. He also discusses beliefs and myths associated with horses.