The direct effects of arsenic in the improvement of the quality of the blood seem to me incontestable; and its use for this purpose in anæmic neuralgias is certainly something over and above its special neurotic action. No one, who has employed it much in the cases of anæmic children suffering from chorea after rheumatism, can have failed to observe its frequently striking influence upon blood-formation even long before the nervous ataxia is materially reduced. The misfortune is, however, that we possess no indications by which to judge beforehand whether we may reckon on its most favorable action in any given (non-malarious) case, with certain special exceptions. In angina pectoris it has a most direct effect, which is rarely altogether missed, and is sometimes surprising: the cases in which it succeeds best are those distinguished by anæmia, but we may well suppose, from its remarkable action upon other neuroses of the vagus, that it is something more than an action on the blood-making process which produces such powerful effects in allaying the tendency to recurrence of the paroxysms. My attention was called to its action in this disease chiefly by the remarkable case published by Philipp;[37] this was a purely neurotic angina, but one of the severest type, and the influence of arsenic was very striking. Since that time I have employed it in several cases, and, after trying various forms of administration, I conclude that nothing is better than Fowler's solution, in doses of three minims (gradually increased, if the remedy be well tolerated, up to eight or ten) three times a day. Unfortunately, there are some neurotic patients who cannot bear arsenic, the irritability of their alimentary canal is such that the drug always provokes vomiting, or diarrhoea, or both; this was the case with one of my patients, in whose case I had allowed myself to hope for the very best results from arsenical treatment. But where the patient tolerates it--and usually he tolerates it extremely well--the prolonged use of arsenic seems really to root out the anginoid tendency, or at least to confine it to the more trivial and manageable manifestations. I believe that in at least three patients, I have so completely broken down a succession of cardiac neuralgic attacks as to substitute for them a mere remnant of a tendency to "tightness at the chest" after any severe bodily exertion or mental emotion. It might be a question, in cases where the stomach does not tolerate the ordinary administration of the agent, whether it would not be worth while to try the effect of subcutaneous injection (two to four minims of Fowler), or inhalation of the smoke of arsenical cigarettes. But, in truth, it is not certain that even in this case we escape the characteristic effects of the drug upon those persons who are abnormally sensitive to it.