Here, however, without entering on moot questions, we may draw the _à priori_ conclusion that these earliest living things were double-natured, in so far that they must have had the ability to assimilate from the inorganic world all the materials of which protoplasm consists--must therefore, along with the power of appropriating carbon from its gaseous compound, also have had the power of appropriating nitrogen, either from one of its combined oxides or directly from the air with which water is more or less charged. For before organic substances existed there could have been none but inorganic sources from which nitrogen could be obtained. This conclusion concerns us only because it implies homogeneity of nature in these primordial forms of life. There could not at first have existed among these minutest of _Protozoa_ even such vague distinctions as are now presented in a shadowy way by their modern representatives. And the implication is that during the period throughout which these smallest, lowest, and simplest living things alone existed, there could have been, in the absence of kinds, no mutual dependence. Since, among various of the lowest types now known to us, the same individual exhibits a life which is now predominantly vegetal and now predominantly animal, we cannot err in assuming that there eventually took place differentiations of this original plant-animal type into types permanently unlike: some in which the traits were more markedly vegetal and others in which they were more markedly animal. As fast as this differentiation arose, there came the beginnings of co-operation between the predominantly vegetal types which by the aid of light formed organic matter from the inorganic world, and the predominantly animal types which, in chief measure, utilized the matter so formed. Evidently with the rise of such a differentiation came an incipient mutual dependence. If to the implied algoid type and the animal type there be added the fungoid type, somewhat intermediate in character, which in a large proportion of cases lives on the decaying remnants of the other two, we are furnished with a rude conception of the primary differentiations and the accompanying vague mutual dependences.