To this question, the fossil remains of Vertebrata do not give us a very definite, but still they do give an approximate, answer. For with the exception of two lizard-like animals found in the Permian system (the Proterosaurus and Rhopalodon), all the fossil remains of Amniota, as yet known, belong to the _secondary_, _tertiary_, and _quaternary epochs_. With regard to the two Vertebrata just named, it is still doubtful whether they are genuine reptiles, or perhaps Amphibia of the salamander kind. Their skeleton alone is known to us, and even this not perfectly. Now as we know nothing of the characteristic features of their soft parts, it is quite possible that the Proterosaurus and Rhopalodon were non-amnionate animals more closely allied to Amphibia than to Reptiles; possibly they belonged to the transition form between the two classes. But, on the other hand, as undoubted fossil remains of Amniota have been found as early as the Trias, it is probable that the _main class of Amniota_ first developed in the Trias, that is, in the beginning of the Mesolithic epoch. As we have already seen, this very period is evidently one of the most important turning points in the organic history of the earth. The palæolithic fern forests were then replaced by the pine forests of the Trias period; important transformations then took place in many of the classes of Invertebrata. Articulated marine lilies (Colocrina) developed out of the plated ones (Phatnocrina.) The Autechinidæ, or sea-urchins with only twenty rows of plates, took the place of the palæolithic Palechinidæ, the sea-urchins with more than twenty rows of plates. The Cystideæ, Blastoideæ, Trilobita, and other characteristic groups of Invertebrata of the primary period became extinct. It is no wonder that transforming conditions of adaptation powerfully influenced the Vertebrate tribes also in the beginning of the Trias period, and caused the origin of Amniotic animals. If, however, the two Lizard and Salamander-like animals of the Permian system, the Proterosaurus and Rhopalodon, are considered genuine Reptiles, and consequently the most ancient Amniota, then the origin of this main class must necessarily have taken place in the preceding period, towards the end of the primary, namely, in the Permian period. However, all other remains of Reptiles, which were formerly believed to have been found in the Permian and the Coal system, or even in the Devonian system, have been proved to be either not remains of Reptiles at all, or to belong to a more recent date (for the most part to the Trias). (Compare Plate XIV.)