In Narcolepsy we have the very opposite condition to insomnia, in which the patient is seized with a sudden and uncontrollable desire to sleep. Fisher claims that if hysteria, diabetes and nephritis can be excluded that the disease is almost pathognomonic of cerebral syphilis and that there is syphilitic endarteritis. Important in this connection are the researches of Barker (59) in which he shows the correlation between mental diseases and cerebral endarteritis, and the latter to be due to a variety of poisons. Stern (51) reports a case in which these may be excluded and which seems to be the result of intrinsic metabolic causes. His patient was mentally active, showed no evidences of hysteria, and the seizures occurred independently of the ingestion of food. The following arguments show the intrinsic origin of this case: -

1. The stomach was dilated and hyperchlorhydria was present. 2. There was high urinary density and acidity. 3. Urinary chlorids in great excess. 4. Diminished alkalinity of the blood. 5. The salts of oxalic, sulphuric and uric acids, carbamid, creatin and creatinin were in great excess. 6. There was a low degree of urinary toxicity before and after the attacks of somnolence. His explanation of the disease is that deficiency in the chlorids of the blood causes low osmotic pressure and as a result of this there is a temporary diminution in the nutrition and electrical conductivity of the nerves.

The records of experimental poisonings with autotoxins suggest an explanation of that class of cases which simulate typhoid fever, but which differ from it in that they do not give the serum reaction, do not show symptoms of perforation or result fatally, are more prone to constipation than diarrhea, and do not have as high or as long a temperature. Bruce (52) reports cases of this kind in which there were nervous depression, languor, drowsiness, vertigo, cephalalgia, pyrexia, vomiting, diarrhea and abdominal tenderness. The writer has observed a number of these cases and finds them usually accompanied by a cutaneous eruption similar to that described under infantile convulsions. All of the writer's cases were in adults. It may be 'that research will soon show these cases to be of the para-typhoid type or discover a specific bacterium as the cause; their similarity with known intoxications however, is so marked that the possible auto-toxic origin should not be overlooked.