EXPERIMENTAL DIABETES 373 



the possibility of a part of the acids being derived from proteid. 

 The acid cannot be derived from carbohydrate alone, for it is 

 produced in abundance on a carbohydrate-free diet, and at a stage 

 in diabetes when practically all the tissue carbohydrate must have 

 been used up. 



Two other possibilities remain to explain the source of 

 yS-oxy-butyric acid and its derivatives; the decomposition of fat 

 and a synthetic process. For this synthesis, bodies with two and 

 three carbon atoms would be available, and these might be derived 

 from proteids or fats or carbohydrates ; so that, indirectly at least, 

 these bodies may come from any of the proximate principles of 

 food. 



Their derivation from fats would, for certain fats at least, 

 be a very simple chemical process. Thus from butyric acid, 

 CH 3 CH 2 CH 2 COOH, the replacement by hydroxyl (OH) of one of the 

 hydrogen atoms of the second methyl group of the chain is all that 

 would have to occur CH 3 CHOH CH 2 COOH. In the case of higher 

 fats, with 16-18 carbon atoms (e.g. palmitic, C 15 H 31 COQH, and 

 stearic, C 17 H 35 COOH), it is possible that a gradual breaking away, 

 by an oxidative process, of the ends of the chain occurs, till at last 

 an acid with a chain of four carbon atoms, viz. buy trie acid, is left, 

 and that, in the diabetic state oxidation can only proceed a short 

 distance farther and convert this into /3-oxy-butyric acid, aceto- 

 acetic acid, and acetone, but not, as normally, into C0 2 and H 2 0. 

 By such a process only one molecule of /3-oxy-butyric acid would be 

 derived from one of fatty acid. It is conceivable, however, that 

 the larger fatty acid molecule breaks down into several smaller 

 fractions, each of which then changes into butyric acid. 1 By 

 such a process, several butyric-acid molecules might be derived 

 from each fatty acid molecule. Let us see now how far actual 

 observation bears out these suppositions. 



Schwarz ( 30 ) noticed a distinct increase in the amount of acetone 

 excreted in the urine of a diabetic when large quantities of butter 

 were added to a proteid, or proteid and carbohydrate diet. An 

 increase likewise followed the administration of the various fat 



1 Such a disruption of a large molecule into several smaller molecules un- 

 doubtedly occurs in the case of carbohydrates (e.g. lactic acid fermentation). 

 These contain, however, within each molecule sufficient oxygen for the disruption, 

 and it is a question whether fatty acid with its lack of oxygen could undergo the 

 same cleavage. 



