It is necessary here to inquire, more particularly than we have yet done, into the nature of the "painful points" first signalized by Valleix as a distinctive symptom of neuralgia. Very great differences of opinion have prevailed among subsequent writers, both as to the frequency and the significance of these points. It may be said, however, to be now quite settled that the presence of definite points, painful on pressure, and also corresponding to the foci of severest spontaneous pain, is far from universal in neuralgia. Upon this point there is probably no reason to doubt the correctness of Eulenburg's observations made in the surgical clinic of Greifswald and the polyclinic of the University of Berlin; he says that he discovered the existence of tender points in "Valleix's sense," in rather more than half the cases of superficial neuralgia, but in the rest he could not by any means discover them. In many other cases, however, he found more indefinite points of tenderness, not accurately corresponding to nerve-branches, but affecting individual portions of skin, bone, or joints; the relation of these to the neuralgic symptoms was difficult of explanation. Eulenburg lays down the principle that "hyperæsthesia" may depend on three sorts of causes--(1) On local disease of the peripheral ends of nerves; (2) on alterations of the psychical centres; and (3) on morbidly exaggerated conduction in the nerve-trunks themselves; and it is to this third source that he attributes many of the phenomena of the neuralgic painful points, and especially their multiplicity, in many cases. The _locus in quo_ of the mischief which sets up this exaggerated conduction of sensory impression is, upon this theory, between the psychical centre and the main point of branching of the nerves; hence a large number of peripheral nerve-termini might be practically sensitive to touch, because the mischief, though localized in a comparatively small spot, might easily affect many bundles of fibres, which diverge widely from each other in their course. It will be seen presently with what limits and for what reasons we believe this to be a true theory. But to return to the question of painful points in Valleix's sense, we must state one or two facts which seem certain from our own experience, but have not been adequately recognized, we believe, by others. The first is, that localized tender spots, accurate pressure on which will set up or aggravate the neuralgic pain, are not early phenomena, save in neuralgias of exceptional severity of onset; but that a certain persistence and severity of neuralgia are always followed by the formation of one or more true points douloureux. The second fact relates to the clinical history of migraine. Roughly speaking, it is true, as Eulenburg states, that, in pure migraine, painful points in Valleix's sense are not to be found; in place of them we observe, after the paroxysms have passed away, a more generalized soreness of considerable