LYMPH FORMATION AND CIRCULATION 121 



to the amount of salt in the diet. A very considerable retention of 

 water usually occurs before there is any evidence of edema; indeed, as 

 a result of giving salt, the body weight may increase from five to seven 

 kilograms (10 to 15 pounds) within a day or two without the appear- 

 ance of puffiness. 



The cause of the edema during salt retention is no doubt closely re- 

 lated to the action of lymphagogues. In a normal person excessive in- 

 gestion of salt is immediately followed by excretion of the excess through 

 the kidney. Where the kidneys are diseased, this excess of salt is re- 

 tained in the blood, raising its osmotic pressure and attracting water 

 from the tissue fluids. This leads to excessive thirst, the imbibed water 

 being used to replace that lost from the tissues. But all the crystalline 

 lymphagogues do not, when present in excess in the blood of nephritic 

 patients, necessarily cause edema; urea, for example, may accumulate 

 considerably without any such effect. The different action is usually 

 attributed to inequality in the diffusibility of the two crystalloids through 

 animal membranes, sodium chloride diffusing much less readily than 

 urea. 



It is most important to note that the fluid in edema is loose in the 

 tissues and can be drained away by the insertion of tubes. There is 

 absolutely no evidence in support of the claim of Martin Fischer that 

 edema is due to imbibition of water by the colloids of the tissues. This 

 question has been fully discussed elsewhere (page 63). 



THE CEREBROSPINAL FLUID 



Considerable attention is now paid to the cerebrospinal fluid which is 

 present in the subarachnoid spaces of the spinal cord and brain. This is 

 because the fluid is readily collected by the method of lumbar puncture, 

 which is performed for the purpose either of relieving increased inter- 

 cranial pressure as in hydrocephalus, delirium tremens, eclampsia, enceph- 

 alitis, etc., or of collecting some of the fluid for diagnostic purposes. 

 The rationale of the removal for the relief of pressure in the brain case 

 is somewhat difficult to understand, as will be evident from a perusal of 

 the chapter on the circulation of blood in the brain (page 254) . There is no 

 doubt, however, that the procedure gives relief. It is more particularly 

 with the biochemical characteristics of the fluid that we are concerned here 

 (of Levinson 19 ). The normal fluid (i.e., obtained from patients not suf- 

 fering from inflammatory processes of the cerebrospinal membranes) is 

 colorless, it contains only from 4-6 cells per cubic mm., its specific gravity 

 varies between 1,000 and 1,008 and it does not clot. The H-ion concentra- 

 tion of perfectly fresh fluid, removed during life, can be accurately meas- 

 ured by the colorimetric method without dialysis (page 32) and has been 



