The comparative method will then apply successively to the different parts of an organism, to the different ages of the same organism, and to the different organisms in the animal and vegetable series. It will even apply to embryonic life, Comte clearly formulates von Baer’s law, while making indispensable reservations. The primitive state of the highest organism, he says, must represent, from the anatomical and physiological point of view, the essential characteristics of the complete state which belongs to the more inferior organism, and so on successively “without our being able to find again the exact analogy of each of the principal terms of the inferior organic series in the sole analysis of the various phases of development of each superior organism.” This comparison, so to speak, allows us to realise in the same individual the growing complication of organs and of functions which characterises the whole biological hierarchy. Thus it is particularly “luminous.”[164] Von Baer’s book had appeared in German in 1827. Had Comte known it, it is most probable that, according to his habit, he would have quoted it. IV. In order to consider organisms in the regular sequence which allows of comparison, we must first have established the order in which they should be arranged. But, conversely, to establish this order, a knowledge of anatomy and physiology is indispensable. So between these two sciences on the one hand and “biotaxy” on the other there is a strict solidarity. The problem of classification is thus an essential part of general biology. In the natural classification sought after by science, the position assigned to each organism would suffice to define at once the whole of its anatomical and physiological nature, in relation to the organisms which precede and to those which follow.[165] Any natural classification cannot, however, be anything but imperfect. Accustomed as we are to artificial classifications, which admit of absolute and immediate perfection, we are surprised that the same should not be the case in natural classification. But, if the latter is a real science, we must own that, here as elsewhere, we can only reach more or less distant approximations. The co-ordination of living species is a problem like the static or dynamic analysis of a determined organism. Like this analysis, it only allows of solutions which are approached rather than realised.[166]