DETERMINATION OF SUGAR. 767 



For the physician the method in this form is not serviceable. Even when 

 the specific gravity is determined by a delicate urinometer which can 

 give the density to the fourth decimal, exact results are not obtained, 

 because of the ordinary errors of the method (BUDDE) ; but the errors are 

 usually smaller than those which occur in titrations made by unskilled 

 hands. 



When the quantity of sugar is less than 5 p. m. these methods cannot 

 be used. Such small amounts cannot, as already mentioned, be deter- 

 mined by titration directly, because the reducing power of normal urine 

 corresponds to 4-5 p. m. of sugar. In such cases, according to WORM- 

 MfLLER, it is better to first determine the reducing power of the urine 

 by titration with KNAPP'S solution, then ferment the urine with the 

 addition of yeast and titrate again with KNAPP'S solution. The dif- 

 ference found between the two titrations calculated as sugar gives the 

 true quantity of the latter. 



The determination of the sugar by fermentation can be so performed 

 that the loss in weight due to the CO2 can be estimated or the volume 

 of the gas measured. For this last purpose LOHNSTEIN l has constructed 

 a special fermentation saccharometer, and his " precision saccharometer " 

 is to be recommended. Based upon LOHNSTEIN'S instrument, WAG- 

 NER 2 has constructed a " fermentation saccharo-manometer," which has 

 certain advantages over LOHNSTEIN'S apparatus. 



ESTIMATIONS OF SUGAR BY POLARIZATION. In this method the 

 urine must be clear, not too deeply colored, and, above all, must not 

 contain any other optically active substances besides dextrose. The 

 urine may contain several levo rotatory substances such as proteids, /?-oxy- 

 butyric acid, conjugated glucuronic acids, the so-called LEO'S sugar 

 and less often cystine, all of which are unfermentable. The proteid is 

 removed by coagulation, and the others are detected by the polariscope 

 after complete fermentation. The fermentable levulose is detected in a 

 special manner (see below), and the dextrorotatory milk-sugar differs 

 from dextrose in its not fermenting readily. By using a delicate instru- 

 ment and with sufficient practice very exact results can be obtained by 

 this method. The value of this procedure consists in the rapidity with 

 which the determination can be made. In using instruments specially 

 constructed for clinical purposes the accuracy is less than with the less 

 expensive fermentation test. Under such circumstances, and as the 

 estimation by means of polarization can be performed with exactitude 

 only by specially trained chemists, it is hardly worth while to give this 

 method in detail, and the reader is referred to handbooks for hints in the 

 use of the apparatus. 



Levulose. Levogyrate urines containing sugar have been noted by 

 several investigators, although the nature of the sugar was not well known 

 to the earlier observers. In recent years several positively authentic 

 cases of levulosuria have been described, and also cases of diabetes 



Pfluger's Arch., 33 and 37; Budde, ibid., 40, and Zeitschr. f. physiol. Chem., 13. See 

 also Huppert-Neubauer, 10. Aufl., and Lohnstein, Pfluger's Arch., 62. 



1 Berlin, klin. Wochenschr., 35, and Allg. med. Central-Ztg., 1899; Goldman, Chem. 

 Centralbl., 1907, 1, 1149. 



2 Miinch. med. Wochensch., 1905. 



