If formation and absorption of the cerebrospinal fluid (c.s.f.) occur at different sites an intervening circulation of the fluid is obligatory, and such a circulation may appreciably modify the spread of drugs introduced into the c.s.f. Nevertheless, few controlled observations have been made on this subject. The problems involved, with particular reference to spinal anaesthesia in man, were described by Maxson (1938), who also studied the movements under various conditions of red ink in an artificial spinal fluid contained in a glass model. Sachs, Wilkins & Sams (1930) compared by direct observation the movements of a marker dye introduced at various sites into the spinal subarachnoid space of anaesthetized dogs with those occurring in a glass model.The general aim of the present investigations is to study the effects of bodily activity upon the c.s.f. circulation. In this paper the movements of a marker dye in the spinal subarachnoid space of the prone horizontal anaesthetized cat have been compared with those occurring in a glass model. The effects of tilting or shaking the animal have been studied and also the effect of picrotoxin convulsions.
METHODSCats were anaesthetized by intraperitoneal injection of sodium pentobarbitone (Abbott Laboratories, 30-40 mg/kg). In each experiment a cannula of known length and volume was inserted into the dorsal spinal subarachnoid space at one of three sites. When the cannula was to be inserted at the sacral site the laminae of the three sacral vertebrae were resected and the underlying extradural fat removed. A nylon cannula (No. 1 'Portex', Portland Plastics) with two oppositely-placed holes close to its sealed tip, filled with Ringer-Locke solution and stoppered, was introduced and tied in place with its tip pointing cranially at the first sacral vertebral level. In experiments in which the cannula was to be inserted at the cervical site, after laminectomy the right eighth cervical dorsal nerve root was exposed for about 5 mm from the cord. A cannula, as before, was introduced via the nerve-root sheath and tied so that it lay transversely across the subarachnoid space. For insertion at the thoracic site, after laminectomy the cannula (No. 00 'Portex') was introduced via the right twelfth thoracic nerve-root sheath and fixed with its tip cranially directed at the eleventh thoracic vertebral level.