DIABETES MELLITUS. 837 



ranges from 1020 to 1030, and in severe ones from 1030 to 1050, de- 

 pends almost exclusively upon the quantity of sugar in it ; for although 

 the common belief has not proved true, that the absolute amount of 

 urea and salts is reduced in diabetic urine, yet, owing to the enormous 

 quantity of water in which they are dissolved, their relative quantity 

 is very small indeed. Liebermeister and Reich found that the produc- 

 tion of urea of their patients not only was greater than that of healthy 

 persons moderately supplied with mixed food, but that it also exceeded 

 that of a healthy man, who ate as much food as the diabetic patients 

 did. The amount of urea fluctuated in the patients from thirty-two to 

 fifty-five grammes, and in a healthy subject from twenty-nine to thirty- 

 two grammes. 



Although the profusion of the urine, its high specific gravity, and 

 its sweetish taste, are decidedly indicative of the presence in it of 

 sugar, yet its positive existence can be easily proved by means of one 

 of the many " sugar-tests." Although it is important to select the most 

 delicate and surest of these, when we wish to determine the presence 

 of very minute quantities of sugar, and although the supposed discov- 

 eries of traces of sugar in healthy urine, and in many other pathologi- 

 cal and physiological conditions, are in a measure due to our mistaking 

 other substances, having similar reaction, for sugar, yet, when it appears 

 in large quantities, as it does in diabetic urine, Trommels test is quite 

 sufficient to settle the diagnosis. An excess of liquor potassae is to be 

 mingled with a portion of the urine to be examined ; a dilute solution 

 of sulphate of copper is then to be added, drop by drop, as long as the 

 precipitate which first forms will redissolve when stirred. The liquid 

 (which, if sugar be present, will show a blue color) is then to be 

 filtered and warmed. The precipitation of red oxide of copper in the 

 liquid, while still below the boiling-point, furnishes absolute proof that 

 the solution contains sugar. If albumen also be present, it must pre- 

 viously be separated by boiling and filtration. Another more simple 

 test consists in warming the urine after adding a solution of potassa. 

 The presence of sugar is indicated by the liquid's assuming a yellow 

 color, which gradually deepens into a dark brown. If this color does 

 not appear, the non-existence of sugar is certain ; but, even when visi- 

 ble, it is always advisable to try Trommels test also. For the details 

 of the somewhat elaborate fermentation-test, and the other tests for 

 sugar, we refer to the text-books upon organic chemistry. Fehling's 

 " liquid " furnishes a very sure means of determining the percentage 

 of urea, and also of the sugar, if the quantity of urine passed in twenty- 

 four hours be measured. Liebermeister and Heich^ who, in their many 

 experiments, examined each portion of the urine separately, only 

 noting as reliable those observations which agreed, hardly ever had 



