In the earliest tetrapods, the fenestra vestibuli was a large hole in the braincase wall bounded by bones of different embryological origins: the otic capsule and occipital arch components, and also, in all except the Devonian Acanthostega, the dermal parasphenoid. This means that the hole lay along the line of the embryonic metotic fissure. Early tetrapod braincases were poorly ossified internally, and no specialized opening for a perilymphatic duct is evident. It is arguable that the earliest tetrapods had neither a perilymphatic duct crossing the otic capsule nor a specialized auditory receptor in a separate lagenar pouch. The primitive tetrapod condition is found in the earliest amniotes, and the separate development of (1) a fenestra vestibuli confined to the limits of the otic capsule, (2) a specialized pressure relief window also derived from components on the line of the metotic fissure, (3) a nonstructural, vibratory stapes and (4) increased internal ossification of the internal walls of the otic capsule, can be traced separately in synapsids, lepidosauromorph diapsids, archosauromorph diapsids, probably turtles, and amphibians. This suggests separate development of true tympanic ears in each of these groups. Developments indicating the existence of a true tympanic ear in amniotes are first found in animals from the Triassic period, and a correlation with the evolution of insect sound production is suggested.