710 RETENTION OF CHLORIDES IN THE BODY 



pass into the filtrate although the membrane is impermeable to 

 proteids. 



In connection with this subject, Sollmann has investigated in 

 dogs the effect of different diuretics upon the excretion of chlorides 

 in the urine ; he divided them into four groups, (a) Diuretics 

 which reduce greatly the percentage excretion of chloride, but 

 which owing to the diuresis cause an increase in the total excretion, 

 e.g. sodium sulphate, sodium phosphate, dextrose, urea, &c. 

 (b) Those which decrease both the percentage and total excretion, 

 e.g. water, and also chloride starvation, which is not accompanied 

 of course by diuresis, (c) Those which do not affect the per- 

 centage excretion materially, e.g. caffein and allied compounds, 

 phloridzin, oil of juniper, &c. These diuretics do not dilute the 

 blood, (d) Those which increase both the percentage and total 

 excretion in urines originally poor in this ion, e.g. sodium chloride, 

 sodium iodide, sodium nitrate, &c. Comparing the first and 

 second groups in order to find the common cause for the decreased 

 percentage excretion, Sollmann concludes that the cause is neither 

 diuresis, nor the presence of a foreign salt in the blood, nor dilution 

 of the plasma, but must be a lowered percentage of chloride in 

 the plasma. However, direct determinations of the chloride in 

 the plasma do not show this to be the case. He, therefore, has 

 recourse to Forster's theory, and considers that the essential factor 

 in the production of a low percentage of chlorides in the urine is 

 a lessened amount of unbound chloride in the plasma. He explains 

 the power of a nitrate to increase the percentage of chloride in 

 the urine by supposing that it can displace chloride from an un- 

 filterable compound, and so liberate chloride for excretion. All 

 observers who have worked with rabbits instead of dogs have 

 obtained wholly different results. In rabbits the effect of all 

 diuretics is to approximate the percentage of chloride in the urine 

 to that of the blood as determined by direct estimation. Sollmann 

 explains this difference between dogs and rabbits by the assumption 

 that diuretics in the rabbit break down the resistance of the kidney 

 cells to the excretion of combined chloride, but not in the case 

 of the dog. It has been shown that the human kidney reacts 

 like that of the dog. 



The theory of the combination in the blood of crystalloids 

 with colloid is interesting, and gives a possible physical explana- 

 tion of many phenomena which have been considered by some to 



