Method of Investigation. 13 



Food, feces, and urine. Although the object of the experiments here reported 

 was primarily the study of metabolism during fasting, there was opportunity 

 to study the effect of fasting on metabolism when food was given after fasting ; 

 hence the chemical examination of food was included in many of the studies 

 here reported. 



The chemical determinations made in these experiments included water, ash, 

 nitrogen, carbon, organic hydrogen, ether extract, sulphur, and phosphorus of 

 food and feces. In addition to the above the examination of the urine included 

 determinations of chlorine, phosphoric acid, sulphuric acid, creatine, and 

 creatinine. 



Sampling and weighing for analyses. In so far as possible, the determi- 

 nations were made on fresh material. This was always the case with foods. 

 With feces a preliminary drying was necessary. This drying was done in a 

 vacuum desiccator. Urine was delivered from carefully calibrated pipettes. 

 Water. While formerly the water was determined by noting the loss in 

 weight of the substances when heated at about 100 C, experience in this 

 laboratory has shown that this method is open to criticism. Accordingly, water 

 determinations in this series of experiments were made by the vacuum method. 37 

 Ash. Ash was determined by charring a sample of the dried material, 

 extracting the charred mass with water to remove the more volatile salts, 

 igniting the residue, and evaporating the extract with the residue at very 

 low red heat. The determination of ash has by no means the scientific accuracy 

 of the other determinations made in connection with these experiments, since 

 the composition of the material weighed as ash may vary noticeably with the 

 method of treatment. Arrangements could not be perfected in time to make the 

 determinations of the calcium, magnesium, and potassium, but the crude ash 

 determination is not without value. 



Nitrogen. Nitrogen was determined by the Kjeldahl method, mercury 

 being used to facilitate oxidation and potassium sulphide to destroy the mercur- 

 ammonium compounds in the distillation. The distillation was carried out in 

 a special form of still. 38 



The accuracy of the nitrogen determinations was very frequently checked 

 by testing one or more materials of known nitrogen content, such as ammonium 

 sulphate, ammonium ferrous sulphate, urea, and. uric acid. In such tests the 

 digestion was carried out with the addition of 1 gram of pure sugar, so that 

 the conditions might be identical with those of the regular analyses. 



Carbon and hydrogen. The carbon was determined by the Liebig method, 

 with such modifications as have been found to be practicable from extended 



"Benedict & Manning, Am. Jour. Physiol. (1905), 13, p. 309. 

 38 Jour. Amer. Chem. Soc. (1900), 22, p. 259. 



