THE NITROGENOUS EXCRETION. URINE. 353 



perfectly automatic. When the osmotic pressure of certain 

 constituents is increased beyond a pretty definite limit, the fil- 

 tering mechanism in the kidney for those constituents becomes 

 active and the excess is allowed to pass. The simple laws of 

 diffusion and osmotic pressure do not help us greatly in ex- 

 plaining the actions of the kidneys where the flow of excreted 

 substances is usually from a level of low concentration to one 

 of higher. Attempts have been made to compare the separat- 

 ing medium between the urine and the blood to a semiperme- 

 able membrane, but the comparison is very imperfect unless 

 the degree of impermeability be specially limited for each sub- 

 stance passing from the blood to the urine. The limitation 

 would have to account for a concentration of salt from about 

 0.6 per cent in the blood to over i.o per cent in the urine, 

 while for urea the concentration would change from about 

 0.05 per cent or lower to over 2.0 per cent, that is forty fold. 

 Limitations as w r ide as these render the comparison of little 

 practical service. 



Percentage Variations. It is not possible to speak of the 

 mean strength of normal urine since the variations are ex- 

 tremely irregular, depending in health on a great many 

 factors. The volume excreted daily, as stated in the books, is 

 usually given much too high for the conditions obtaining in 

 the United States. In place of the 1,500 cc. as found in most 

 of the foreign works we should take 1,150 to 1,200 cc. as 

 nearer the average excretion for 24 hours. In some hundreds 

 of examinations made by the writer in the last few years on 

 people of both sexes engaged in various occupations the aver- 

 age volume comes within these limits. 



A number of complete analyses of urine are found in the 

 literature, but in most of them the uric acid content is placed 

 too low because of the faulty methods of determination for- 

 merly employed. In the following table are given some results 

 obtained in the author's laboratory in which the recognized 

 sources of error have been avoided as far as possible. In each 

 case the mixed urine of the 24 hours was used, and the am- 



2J 



