This latter definition surely was not not taken seriously because there have in all probability, lived bipedal dynosaurs in the Mesozoic.
It is also very probable that there are (or were) animals in the strict sense, which, if we knew them better as we do now, were such that we would be entitled (or forced) to call rational, surely if we had the opportunity to compare them in a detailed way with the first humans (which, as far as we know, have lived somewhere in Africa). The way of life of such animals could be still such that we call them animals ('brutes') as contrasted with humans, despite the fact that they are intelligent (like Orca's, Chimpansees, and the like). In that case rational could not serve to represent the specific difference which distinguishes between brutes and man (because some brutes are intelligent). Capable of laughing would then be a better feature to express the specific difference (provided it does neither occur in these intelligent animals, nor in any other animal in the strict sense). And then 'capable of laughing' would be an essential predicate.