PHYSIOLOGY OF THE LIVER AND SPLEEN. 733 



sented in substance by the formula C 6 H 12 6 H 2 = C 6 H 10 O 5 . 

 There is no doubt that both dextrose and levulose increase markedly 

 the amount of glycogen in the liver; and, since cane-sugar is inverted 

 in the intestine before absorption, it also must be a true glycogen- 

 former, a fact that has been abundantly demonstrated by direct 

 experiment. Lusk* has shown, however, that, if cane-sugar is in- 

 jected under the skin, it has a very feeble effect in the way of increas- 

 ing the amount of glycogen in the liver, since under these conditions 

 it is probably absorbed into the blood without undergoing inversion. 

 Experiments with subcutaneous injection of lactose gave similar 

 results, and it is generally believed that the liver cells can not convert 

 the double sugars to glycogen, at least not readily; hence the value 

 of the hydrolysis of these sugars in the alimentary canal before 

 absorption. We may assume, therefore, that dextrose, levulose, and 

 galactose are the true glycogen-formers that occur normally in the 

 blood, and that the disaccharids (cane-sugar, milk-sugar, etc.) and 

 the polysaccharids (starches) are true glycogen-formers to the ex- 

 tent that they are converted into dextrose, levulose, or galactose. 



Effect of Proteid on Glycogen Formation. In his first studies 

 upon glycogen Bernard asserted that it may be formed from proteid 

 material. Since that time there have been much discussion and 

 experimentation upon this point. The usual view is that proteid 

 must be counted among the true glycogen-formers in the sense that 

 some of the material of the proteid molecule is directly converted to 

 glycogen. The proteid in digestion undergoes, it will be remembered, 

 a splitting process the limits of which are not definitely settled. It 

 is assumed, however, that the nitrogenous split products are acted 

 upon in the liver, the nitrogen with some of the carbon being changed 

 to urea, while the non-nitrogenous residue is converted to sugar. 

 Among the split products of proteid that have been especially 

 investigated in this relation the results with leucin and glucosamin 

 have been chiefly negative.! Experimentally observers find for 

 the warm-blooded animals at least that feeding with proteids, 

 even in the case of those proteids, such as casein, that con- 

 tain no carbohydrate grouping, causes an increased production of 

 glycogen. % The conclusion to be drawn from these experiments 

 is strengthened by clinical experience upon human beings suf- 

 fering from diabetes. In severe forms of this disease all the car- 

 bohydrate material of the food appears in the urine. If under 

 these conditions the individual is given an exclusively proteid diet 

 sugar still continues to appear in the urine, and it would seem that 

 this sugar can only arise from the proteid food. In the similar 



* Voit, " Zeitschrift f. Biologie, " 28, 285, 1891. 



t Halsey, " American Journal of Physiology, " 10, 229, 1904. 



% See Stookey, " American Journal of Physiology, " 9, 138, 1903. 



