Of course, in a given case of optic-nerve atrophy without spinal symptoms the question may fairly be raised whether it is right to place such a case in the locomotor ataxia group. All that can be said in reply is (1) that from cases of optic atrophy pure and simple one passes by an unbroken series of steps through cases with more and more distinct evidence of locomotor ataxia to, at the end of the series, optic atrophy in association with characteristic ataxic symptoms, and (2) that, as already stated, a simple case of optic atrophy may remain unchanged for many years, and yet in the end display undoubted evidence of the development of a spinal lesion. But if optic-nerve atrophy may be the primary symptom in the disease, if the occurrence of spinal symptoms may follow it after an interval of many years, and if again it may remain without at any time any existing ataxia, it is not unreasonable to presume that both the Argyll-Robertson pupil and an ocular paralysis may each have exactly corresponding relations to the development of the spinal evidences of locomotor ataxia. The collection of evidence to support this suggestion is even more difficult than in the case of optic-nerve atrophy. The latter condition must ere long compel the patient to seek medical advice, and thus the opportunity for a complete investigation of the state of his nervous apparatus is afforded at a relatively early date. But an Argyll-Robertson pupil may exist, and presumably exist for years, without any inconvenience to the patient. Such a patient, therefore, will not consult his medical adviser until spinal or other symptoms display themselves, and thus the precedence of the pupillary condition cannot be determined. In the case of an ocular paralysis medical assistance is, no doubt, usually promptly invoked. But such an occurrence is open to a number of ætiological explanations, for example, rheumatism, cold, etc., which it is difficult to exclude with confidence. Hence it is much less precise in its significance than either a double optic atrophy or the Argyll-Robertson pupil. It must be by the collection of observations extending over a long term of years that actual demonstration of the relationship of the ocular disturbances now in question to the occurrence of spinal disease can be established. But while falling short of the merit of actual demonstration, the presentation of the facts displayed by a number of cases which could only be observed over relatively brief periods is not without value. If no one case affords a complete history of all the stages of the disease the picture presented may none the less be fairly complete, provided the cases are sufficiently numerous, and they are seen at different points of development. It is believed that in the present series these conditions are fulfilled. The conclusions they afford, as far as the present purpose is concerned, are: (1) That an optic-nerve atrophy, an