CHEMICAL ENVIRONMENT OF THE CENTRAL NERVOUS SYSTEM 



1869 



cal with cerebrospinal fluid insofar as electrolyte com- 

 position is concerned, and if the extracellular volume 

 is approximately 16 per cent, then there must be an 

 appreciable quantity of both chloride and sodium 

 in the intracellular compartment. This conclusion 

 was also entertained by Manery et al. (110, 111) in 

 their extensive studies upon equilibration of radio- 

 active chloride and sodium in various tissues. 

 Amberson and co-workers (5) produced evidence that 

 a large fraction of brain chloride did not varv pro- 

 portionately to serum chloride; hence it was to be 

 considered nondiffusable. Furthermore, the direct 

 analysis of axoplasm from giant axon of squid by 

 Hodgkin (81) has clearly revealed that sodium and 

 chloride exist intraccllularly and in concentration 

 ratios to outside medium of 1 :o and 1:14, respectively. 

 Despite the careful controls used by Allen in deter- 

 mining the inulin-ferrocyanide space, the result oi 

 in vitro techniques must be approached with con- 

 siderable caution. For example, if a rapid redistribu- 

 tion of' water from the interstitial compartment into 

 the intracellular were to occur in these isolated tissue 

 slices before the diffusion measurements were com- 

 pleted, a spuriously small extracellular volume would 

 result. That such a shift may occur has been proposed 

 by van Ilaneveld & Ochs (151) who found that the 

 in vivo cerebral cortex of rabbits develops a rapid 

 increase in electrical resistance within 5 min. after 

 circulatory arrest. They attribute this increase to 

 asphyxia! changes in cellular membranes which allow 

 sodium and chloride to enter the cell accompanied 

 by interstitial water, thus reducing the volume of the 

 extracellular compartment. Further evidence for 

 marked and rapid changes in ionic, and presumably 

 water, distribution in slices of cerebral tissue has been 

 presented by Krebs et al. (97) with their demonstra- 

 tion that brain slices lose about 40 per cent of their 

 potassium content within a few minutes after being 

 suspended in an incubating medium. It is therefore 

 quite possible that significant irreversible changes in 

 the relative volumes of cerebral tissue fluid compart- 

 ments ma) occur during the 10 min. required to 

 remove the brain, prepare the slices and immerse 

 them in the medium. 



Perhaps the most serious dissension to the concept 

 of 30 per cent or even 1 5 per cent extracellular volume 

 in central nervous tissue arises from the electron- 

 micrographs of osmic-acid fixed sections. In a recent 

 paper, from which figure 2 is taken, Schultz et al. 

 (135) present this view most emphatically. 



"In electronmicrographs of well-preserved cortex 

 the close apposition of cells and their processes is of 



iic. 2 Electronmicrographs of adult rat cerebral cortex 

 pieserved with buffered osmium textroxide. .1 The Literal 

 junction between two adjacent endothelial cells at high magni- 

 fication. The lumen of the vessel \cap) is to the left; astrocytic 

 cytoplasm (jastr) is to the right. The endothelium rests on a 

 basement membrane (Am) The endothelial margins overlap 

 slightly, the lapping here amounting to 330 mji. Structures re- 

 sembling adhesion plates arc present at the beginning and end 

 of the overlap as indicated by the arrows. X 45,000. B: A longi- 

 tudinal section of a capillary. The astrocytic sheath is incom- 

 plete, but of the elements abutting on the capillary wall, 

 astrocytic perivascular processes (astr) are far in the majority. 

 The wall of this vessel consists of a one cell thick layer of endo- 

 thelium for much of its course, but portions of a flattened peri- 

 vascular cell (pvc) may be seen in the lower part of the figure. 

 These cells may be rudimentary smooth muscle. A large neuron 

 (n) is immediately below the inset. Note the almost complete 

 absence of interstitial space. X 2,250. [From Maynard el al. 

 (114).] 



particular interest. It is particularly evident at high 

 magnification . . . that there are absolutely no large- 

 scale extracellular gaps or spaces present, as Dempsev 

 & Wislocki (30) and Farquhar & Hartmann (39) 

 pointed out. Where one cytoplasmic mass abuts 

 against another there are fairly constant gaps of 



