Epilepsy has long been classified as a neurosis without known anatomical basis and although its precise origin has not yet been proven, recent investigation has contributed much information as to the factors which predispose to it. Some thirty years ago Meynert attributed epilepsy to the accumulation of a toxic substance of a proteid nature and the work of passing years tends to confirm this view. The convulsive element is very prominent in the pathogenesis of nearly all the auto-toxins described in Part II and the clinical pictures described often very closely resemble that of epilepsy. The hyper-sensitiveness of hypoxanthin poisoning, in which slight stimuli caused muscular and nervous phenomena out of all proportion to the exciting cause, the fibrillary twitchings, rapid respirations, dilitation of the pupils, dyspnea, labored breathing and clonic convulsions with unconsciousness, of methyl guanidin; the dyspnea, salivation, mydriasis, irregular and labored respirations, unconsciousness and clonic convulsions of neurin: - all furnish a closer analogue to the convulsive phenomena of epilepsy than can be found in any other class of poisons. Intoxication of this kind must then be reckoned as at least one of the causes of epileptic and epileptoid convulsions.

This might account for an occasional epileptiform convulsion, the result of an auto-toxin developed from time to time in the digestive tract, but in order that the epilepsy continue in regularly repeated attacks it is necessary that there be a predisposing cause, set in operation by a toxic exciting cause, or else the intoxication cause a permanent structural change. Medical literature abounds in case histories which show that instability of the nervous system may be dependent upon reflex influences such as ovarian or uterine enlargements, adherent prepuce, and upon toxic materials circulating in the blood. It may be contended that the auto-toxins furnish both of these elements, the toxic exciting cause and, by causing cellular changes in the cerebral cortex, the predisposing condition.

It has been found that the toxicity of the urine varies in attacks of epilepsy even more than it does in migrim. Rachford (6) has found that paraxanthin is excreted in excess during attacks of a certain form of epilepsy and he terms this form, paraxanthin epilepsy. This substance does not occur in the urine of other disorders and does not normally exist in the urine in any appreciable amounts.

Pellagrini (39) has found that the cerebro-spinal fluid taken from an epileptic directly after an attack, possesses markedly toxic properties, and when injected into guinea pigs causes epileptoid convulsions and even the status epilepticus. The nearer to the time of the attack, the more toxic was the fluid. That these symptoms were not caused by any bacterial contamination was shown by the fact that the cultures from the fluid remained sterile.

Clark and Prout (40) show that the predisposition necessary for the production of epilepsy is an organic anomaly in the cerebral cortex. Toxins of autogenous or other origin cause actual lesions with swollen cell nuclei, destruction, of the nuclear membrane and intranuclear reticulum, easy abstraction of the nucleolus by the knife, diffuse chromatolysis and protoplasmic changes in the cells, tending to their disappearance as living units and their replacement by a gliosis.

Ceni (41) has given to science the most rational explanation of the philosophy of this disease and the most practical application of such philosophy. His theoretical conclusion is that epileptic blood contains two active principles of different natures and origins. One of these circulates in' a free state and becomes toxic only when injected into the blood of another epileptic: the other active principle circulates in a latent state and is endowed with properties which have a stimulating effect on those cells which are concerned in the elaboration of the epileptogenous toxins. These active principles, when injected in serum form, exert an influence upon nutrition and epileptic manifestations which may be restoring and therapeutic, or weakening and poisonous. In the former case the patient improves in physical condition and nervous disturbances become less frequent and milder in grade. In the latter, there is diminution in weight and strength, and a retrograde change in organic life with its consequent aggravation of nervous manifestations. As a practical application of these investigations Ceni has elaborated a serum containing the epileptogenous bodies and by injection into patients has found that the cellular sensibility of the epileptogenous nerve centres is lessened by repeated and progressive doses of this serum. This treatment was markedly successful in eight out of ten cases experimented upon.

These studies show that the exciting irritant of epilepsy is undoubtedly of autogenous origin, and the fact that no immunity occurs from repeated attacks and that the serum has no inoculating power shows that the process is not bacterial in its nature.

As might be supposed from the pathogenesy of cholin, it has an important relation to the production of epileptic seizures which will be discussed in the researches of Donath, (42) shortly to be published.