THE CARBOHYDRATES. 293 



with the food. Within the past ten years it has been found that in 

 a not inconsiderable number of cases of diabetes degenerative 

 lesions can be demonstrated in the pancreas, and there can be no 

 doubt at the present time that a certain percentage of cases are 

 directly referable to such origin. That the islands of Langerhans 

 are here primarily concerned has been demonstrated by Opie. In 

 this connection Cohnheim's researches are very significant, in which 

 he could show that while neither pancreas nor muscle-tissue contain 

 ferments which are capable of splitting glucose by themselves, 

 extensive glucolysis is effected if an extract of pancreas and muscle- 

 plasma conjointly are allowed to act upon glucose. A condition 

 thus exists which seems quite analogous to the relation between 

 trypsin and enterokinase. 



In the milder forms of diabetes an insufficiency on the part of the 

 muscle-tissue manifestly does not exist, as it is possible to prevent 

 the occurrence of glucosuria, temporarily at least, if the demand for 

 sugar is increased by abundant muscular exercise. That a hepato- 

 genic diabetes may coexist with a myogenic form, however, cannot 

 be doubted. 



The amount of sugar which may be present in the urine under 

 pathological conditions is exceedingly variable. On the one hand, 

 traces only may be found, which may be normal ; while, on the other 

 hand, the daily excretion may exceed 1000 grammes. In diabetes 

 an elimination of from 3 to 6 per cent, in an amount of urine 

 varying between 3000 and 6000 c.c. may be regarded as moderate. 



Tests for Sugar. Simple tests by means of which glucose can be 

 demonstrated directly in the urine as such are, unfortunately, not 

 available. Other sugars, it is true, enter into consideration only 

 under exceptional conditions, but if it is desired to prove that the 

 substance which gives the common sugar reactions is actually glu- 

 cose, a more detailed examination is necessary. Some of these re- 

 actions, moreover, may be simulated by substances which are not 

 carbohydrates, and deductions as to the presence or absence of sugar 

 are hence only warrantable when these can be excluded. If albu- 

 mins are present, they must first be removed. 



NYLANDER'S TEST. This test is to be preferred to the more 

 common one of Trommer, as the reagent does not react with uric 

 acid, kreatinin, or homogentisinic acid. With many of the conju- 

 gate glucuronates, however, a reduction is observed, and it is hence 

 necessary to eliminate this source of error when a positive reac- 

 tion is obtained, or to apply additional tests in which this possi- 

 bility does not enter into consideration. 



The reagent is prepared as follows : 4 grammes of the tartrate 

 of potassium and sodium, together with 2 grammes of subnitrate of 

 bismuth and 10 grammes of sodium hydrate are placed in 90 c.c. 

 of water. The solution is heated to the boiling-point, filtered on 

 cooling, and is then ready for use. It is kept in a dark-colored 

 bottle. 



A few cubic centimeters of the urine are treated with the reagent, 



