From the introduction of spinal anesthesia (in 1899) to the present time speculation has been almost continuously rife concerning the behavior of the anesthetic after injection. Although the suggestion was made as early as 1901 by J. S. Miller1 that it would be important to study the actual concentration of the anesthetic in the cerebrospinal fluid, most attempts to solve the problem have been based on studies of the behavior of fluids injected into inanimate models, into cadavers and occasionally into experimental animals. On the basis of such studies and on a priori reasoning of doubtful validity these authors have concluded that anesthetic solutions of greater density than the cerebrospinal fluid flow downward toward the head if the patient is immediately placed in the Trendelenburg position.
In a recent paper2 this point of view has been adopted again, and various measures have been indicated to prevent untoward results