In the treatment of cerebrospinal fever by specific antiserum, the contention of Flexner and others, that without intrameningeal administration little or no effect is to be observed, has been amply confirmed clinically. A partial experimental basis for this conclusion was furnished by observations which showed that the meninges are normally quite impervious to certain foreign substances, but such observations have until lately been limited entirely to crystalloids. Recently, however, Lemaire and Debré1 have investigated the permeability of the meninges to antitoxins, and they have obtained results that confirm in a very striking manner those of the clinical work.
Their method was to administer to dogs, subcutaneously, large doses of tetanus antitoxin obtained from a horse. After allowing three days for complete diffusion of the antitoxin, cerebrospinal fluid was withdrawn with all possible precautions against admixture with blood, and this was then tested both for its protective power against tetanus toxin