In 27 young healthy guinea pigs the cerebrospinal fluid was stained with rhodamine by suboccipital punction. After removing the neck muscles a small piece of plastic tape was glued to the atlanto-occipital membrane with Histoacryl. When the plastic foil and the membrane are punctured by a fine needle, the canal closes by itself through the elasticity of the tape. Glas capillaries with an outer diameter of less then 0.1 mm were filled with 0.1 μl of a 2% rhodamine solution. After freeze drying, the tip contains 2 μg of the dye. The capillary is then pushed through the preformed canal into the cerebrospinal fluid by the aid of a micromanipulator. 0.1 μl of the CSF is allowed to enter the capillary to dissolve the dye and is then restored into the cerebellomedular cistern. The advantage of this technique is that CSF is used as the solvent. This procedure is nearly pressure free. Moreover the system remains closed because the elasticity of the plastic tape keeps a tight contact to the capillary and the CSF cannot escape on the sides of the capillary. At different time intervals after the application, the distribution of the dye is stopped by quickly deep freezing the whole head with liquid air. The head is then severed, mounted on the table of a cryotome and cut in 20 μm steps. At certain levels the mounted object is fotographed under normal and UV-light with aid of an operating microscope. The position of the rhodamine can bee seen by the yellow fluorescense of the rhodaminelipid complex which forms in the cell membranes. The distribution of the dye was studied in twelf living and six dead animals to differentiate between active flow and passive diffusion. In the living animals the rhodamine spreads in the subarachnoidal space in a frontocranial direction. It reached the frontal part of the brain after 20 min and did not enter the cochlear aqueduct (Figs. 2 and 3). In the dead animals the distribution of the dye was much slower and the dye entered the aqueduct after 30 min and reached the cochlea after 1 h. These results support the conclusion that in the guinea pig there is a flow of cerebrospinal fluid from the cerebellomedular cistern to the frontal parts of the brain. Even when it passes the funnel shaped opening of the perilymphatic duct it does not enter it. Earlier investigators reported the finding that the cerebrospinal fluid enters the cochlea. This depends on the relative big volumes which were used. The fluid amounts which have been applied to the CSF addionally or by exchange to withdrawn CSF are 500-2000 times bigger than the volume we used. Our own results with big volumes (0.05-0.1 ml) are the same as other investigators have shown. Because the method we used is nearly pressure free and keeps the fluid spaces closed throughout the whole experiment, we can support the statement, that perilymph is produced in the cochlea and does not derive from the cerebrospinal fluid. © 1978 Springer-Verlag.
