I9IS.] MARSHALL-BANKS— NEW NEPHELOMETER. 133 



serum albumin added. Quantities of urine and of standard were 

 so taken that R would be in the neighborhood of one half. Urines 

 containing large amounts of albumin (i per cent, or over) were, 

 after suitable dilution, compared directly with standard serum solu- 

 tion. In the case of such urines the high dilution necessary to 

 obtain suitable nephelometric clouds eliminated the differences of 

 color mentioned above. The results were compared with gravi- 

 metric determinations made according to Scherer's method. The 

 clear filtrates from the coagulated protein were tested with sulpho- 

 salicylic acid to make sure that none of the protein remained in 

 solution. Duplicate gravimetric determinations gave good agree- 

 ment. It was immediately evident that the nephelometric de- 

 terminations were considerably higher than the gravimetric. More- 

 over, in the case of determinations on daily specimens of urine from 

 one patient, the nephelometric results were consistently about 25 

 per cent, higher than the gravimetric, while in a similar series from 

 another individual the ratio between nephelometric and gravimetric 

 determinations was very variable, ranging from i to about 1.5. 

 This at once suggested that the different proteins of the serum, while 

 closely related chemically and equally precipitable by sulpho-salicylic 

 acid, might give, in the nephelometer, clouds of different intensities. 

 It is a significant fact that in the case of patient R where the ratio 

 of nephelometric to gravimetric was variable, half saturation of the 

 urine with ammonium sulphate gave a considerable precipitate of 

 globulin. 



In order to determine what differences might exist between the 

 opalescences produced by equal amounts of the various serum pro- 

 teins on precipitation with sulpho-salicylic acid under identical con- 

 ditions, albumin, euglobulin, and pseudoglobulin were prepared from 

 horse serum. Solutions of these when compared in the nephelom- 

 eter gave surprisingly different results. The albumin gave about 

 two and one half times as great an opalesence as the euglobulin and 

 about three times as great as the pseudoglobulin. Compared with 

 casein'' suspensions, the following ratios, expressing the light reflect- 



6 As standard solutions of casein are easily prepared and also give very- 

 satisfactory clouds on precipitation with sulpho-salicylic acid, this substance 

 forms a very convenient standard of reference in nephelometric work with 

 various proteins. 



