2. From Apes back to lowest Lemurs in the lowest Eocene. The date of Eocene being fixed at 3,000,000, we have about 2,100,000 years for this stage; assuming as much as five years for puberty, this results in 420,000 generations.
3. From Lemures to Prototheria. The earliest known mammalian remains come from the Rhætic, or top formation of the Triassic epoch; allowing for the Rhætic only 100,000 years, we have to add the whole of the Jurassic and Cretaceous, in all about 5,500,000 years. Assuming three years for a generation, we get 1,800,000 generations.
4. From Prototheria to something like the Theromorpha at the bottom of the Triassic strata. A duration of 1,700,000 years divided by four gives 425,000 generations.
5. From Theromorpha to Proreptilia, represented by Eryops and Cricotus from the Lower Permian of Texas. Allowing 1,000,000 years, each generation at four years, we obtain 250,000 generations.
6. From Proreptilia to Eotetrapoda, the first terrestrial Vertebrata, represented by something like the Stegocephali, the earliest of which are known from the Coal-measures. Assuming them to have come into existence at the bottom of the Coal-measures, for the duration of which we may guess 2,000,000 years, we get, with four years' allowance for puberty, 500,000 generations.
7. From Eotetrapoda to a not yet separated or differentiated group of Crossopterygian and Dipnoan fishes, both of which are known from Devonian strata. The duration of the latter has been computed at 4,000,000 years, which, with 1,000,000 for the Mountain Limestone formation, gives us 5,000,000 for this stage. Assuming, for the sake of round numbers, as much as five years for a generation, we get 1,000,000 generations.
8. Earliest stage, down to the first fish-like creatures. Teeth and spines indicating the existence of fishes are known from the Upper Silurian. By carrying the earliest fishes down to the bottom of the Silurian, with 2,700,000 years' duration, and allowing three years for attaining puberty, the calculation results in 900,000 generations.
Further back we cannot go. We do not know of any Vertebrate remains from the Ordovician and Cambrian, which together represent 6,700,000 years, enough for at least half as many generations of Prochordate creatures. The pre-Cambrian or Laurentian epoch lies quite beyond the reach of calculation, nor have we any trustworthy fossil remains of living matter from these strata, to which, however, Haeckel and others refer the first beginnings of life.
All the above calculations are, of course, only approximate. What we do know is the existence of representatives of the stages, our proofs being the fossils; but when we refer the origin of the Eotetrapoda, for example, to the bottom and not somewhere to the middle of the Coal-measures, we are guessing merely. Alterations in the levels assumed for the various stage-representatives will, of course, alter the result of the number of generations; but the leading idea, as a whole, is not thereby upset. The fact remains that in the Upper Silurian we have fishes; from the Coal-measures onwards, fishes and Amphibia; since the Permian, fishes, Amphibia, and reptiles; since the end of the Trias these three classes and the Mammalia; and lastly, at least since the Plistocene, man himself. If Evolution is true at all, the transformation from early fish-like creatures to man has come about within these epochs. Being able to assign a time of duration to each of them, with an approximate total of 21,000,000 years, we are also able to put the whole ancestral series to a test by expressing each great stage in generations. The result is very satisfactory. The whole enormous stretch from the lowest fish-like creatures to man has been resolved into more than 5,000,000 successive generations, and each of these means a little step forwards in onward Evolution.