He also finds, from an analysis of the same cases, the fact noted by others, that the prognosis is favorable in inverse proportion to the duration of the disease.

Attacks which chiefly occur in the daytime are much more amenable to treatment than the nocturnal seizures, and especially is this the case in the tongue-biting form. Sudden blows upon the head or falls have been known in isolated cases to effect an amelioration in the patient's disease, but these examples are rare.

The existence of an aura is much better than if none existed.

Death from the attack itself is rare, yet in the large pauper institution with which I was connected for many years I have known of several cases. More often the death results from asphyxia resulting from a bolus of food which chokes the patient or from a fall in some dangerous place—into the fire or elsewhere. Accidental death from drowning is more common than any other form.

The status epileptica into which patients sometimes pass who have had many convulsions is occasionally a fatal termination of the malady, and is always a serious feature.

The influence of different epochs in life is worthy of consideration: of menstruation, of marriage, of pregnancy, and of the menopause there is much to be said. I have sufficiently spoken of the establishment of the menstrual flow, and I would only add another word of caution against giving a too favorable prognosis except in those cases of very recent origin. Marriage appears to have very little to do with changing the attacks, unless they be of an hysteroid character. I have never known epilepsy to influence the course of a pregnancy in any unfavorable way, and I think this has been the observation of others. Gowers refers to cases in which the attacks ceased during the time the mother was carrying the child.

The occasional bad influence of the pregnant state has been illustrated by a case reported by Terrillon.43 This example was a woman who had been the victim of epilepsy of hereditary origin since her seventh year. At the commencement of menstruation her attacks became periodic, and recurred every two months, and she had several two or three days before the flow. Two pregnancies followed several years afterward. During the periods they were increased in number and severity, and occurred several times daily. She had more attacks at this period than in all the time after delivery.

43 Annales de Gynécologie, June, 1881, p. 401.

I have found that the relief of uterine flexion or the establishment of menstruation has exercised a decidedly modifying influence for the better in several cases.

Sometimes the disease is interrupted by the menopause, but very often in my experience it has changed in type and been followed by mental degeneration.