the technical language of Logic is often repeated in the scientific parlance of the day, but with more materialised conceptions and in more concrete cases. The dynamic theory of matter represents it as a unity of attraction and repulsion. A distinguished Darwinian remarks that 'all the various forms of organisms are the necessary products of the unconscious action and reaction between the two properties of adaptability and heredity, reducible as these are to the functions of nutrition and reproduction[3]. The terms 'action and reaction' are hardly sufficient, it may be, to express the sort of unity which is called for: but the statement at least shows the reduction of an actual fact to the interaction of two forces, the meeting of two currents. The one of these is the power of the kind, or universal, which tends to keep things always the same: the other the power of localised circumstances and particular conditions, which tends to render things more and more diversified. The one may be called a centripetal, the other a centrifugal force. If the one be synthetic, the other is analytic. But such names are of little value, save for temporary distinction, and must never be treated as permanent differences which explain themselves. The centre is relative, and so is the totality. Thus it is that the so-called Evolutionist explains the origin of natural kinds. They are what they severally are by reason of a process, a struggle, by alliances and divisions, by re-unions and selections. They are not independent of the inorganic world around them: it has entered into their blood and structure, and made them what they are. To understand them we must learn all we can of the simpler and earlier forms, which have left traces in their structure: traces which, without the existence of such more primitive forms, we might have misunderstood, or have passed by unperceived. And, again, we learn that our hard and fast distinctions are barely justified by Nature. There, kind in its extreme examples seems to run into kind, and we do not find the logically-exact type accurately embodied anywhere. Our classifications into genera and species turn out to be in the first instance prompted by a practical need to embrace the variety in a simple shape. But though perfectly valid, so far as we use them for such ends, they tend to lead us false, if we press them too far.