A man walking. Collotype after Eadweard Muybridge, 1887.

"Here, a Dane, aged fifty-two years, with a history of a long-standing and increasing ataxia, was first photographed (see upper series) while walking with the eyes open. An examination of the figures shows that although the gait differs from the normal in the two principal factors described, yet, owing to the guidance afforded by his eyes and to the evident effort he is making, as shown in the throwing back of the shoulders and the fixation of the trunk, his walk is tolerably good. In the lower series, however, he was photographed while his eyes were closed. A remarkable change is at once observed, and it needs no elaborate study of the trajectories to show what the chief elements of the change are. It is seen at once that the foot is raised much higher from the ground, and that the lateral sway is enormously increased; also, that the successive steps are more irregular. In this series, also, the abrupt descent of the foot upon the ground is again noticed, the entire sole impinging or flopping, as it were, upon the ground at once. … To show how difficult it is to observe a moving limb, even when the movement is slow, it need only be stated that medical writers almost without exception describe this gait erroneously. Almost all lay stress upon rigidity of the leg and insufficient action of the knee-joint. It needs but a hasty examination of the photographs to show how utterly wrong this view is. Every one of the plates reveals the action of the knee-joint, and in fact of all of the joints, to be far in excess of the normal; and, further, the rigidity is simulated and not real. Doubtless this erroneous view has its origin in the difficulty of separating mentally the active from the passive phase of the leg. Plate 550, which is one of the most instructive in the collection, admirably suits the purpose of an explanation. It is seen in studying the upper lateral series that the passive leg, i.e., the one suspended in the air, presents exaggerated knee-joint action and that it at no time gives even the semblance of rigidity. The moment, however, that it impinges on the ground, i.e., the moments it becomes active, all action at the knee ceases. More than this, as the weight of the body is brought upon it, the extension becomes absolute, and finally the knee is, so to speak, locked ; that is, recurved. (This is well seen in No. 6 of the series.) By means of this expedient the strain is thrown upon the bones and ligaments of the joint rather than trusted to the treacherous muscles. The rigidity is, therefore, not muscular, which, in fact, all of our knowledge of ataxia forbids. How much voluntary effort on the part of the muscles assists in steadying the limb, of course depends upon the degree to which the disease has progressed. Regarding the action of the foot and the manner of its impact on the ground various accounts have been given, some writers maintaining that the heel strikes the ground first and considerably in advance of the plantar surface, and others maintaining that the entire sole strikes the ground at once. An examination of Plate 550 shows that the manner of impact probably depends upon the degree of the ataxia. In the upper series, in which the patient was photographed while having the guidance afforded by his eyes, the heel impinges distinctly before the sole, while in the lower series, in which the patient walked with closed eyes, the entire plantar surface impinged at the same time. Occasionally the toes came down first and sometimes even the heels, so that in advanced ataxia any part of the foot may strike the ground first. Regarding the lateral sway of the trunk, it needs no detailed study to tell us that it is much exaggerated. In ataxics who are still able to walk comparatively well this exaggerated sway is always towards the side opposite the advancing or passive leg. However, if the ataxia be increased, as, for instance, in the lower series of Plate 550, the sway may be towards the same side, and then the patient is in danger of falling."--Dercum, loc. cit.