At entrance she weighed 56 pounds, at discharge 54, and lost 4 pounds during starvation, part of which she gained back again. On the diet which she was taking at discharge, she was just about holding her weight. She never excreted much acetone or diacetic acid, and when she was discharged there was merely the faintest traces of these in the urine. It is not well to raise the diet quite so rapidly as was done in this case, but for special reasons she had to leave the hospital as soon as possible, and so her diets were pushed up a little faster than would ordinarily be the case. Below is a graphic chart, such as we use in recording our cases. It has been split up into several pieces here on account of its size: [Illustration: Case 8. A chart tracking Urine and Calorie Intake for the month of April.] [Illustration: A chart tracking Carbohydrate and Protein Intake for the month of April.] [Illustration: A chart tracking per cent. of sugar for the month of April.] [Illustration: A chart tracking sugar output for the month of April.] [Illustration: A chart tracking ammonia for the month of April.] [Illustration: A chart tracking acetone and diacetic acid for the month of April.] [Illustration: A chart tracking weight in pounds for the month of April.] EXAMINATION OF THE URINE. _Directions for Collecting Twenty-four Hour Urine._ Pass the urine at 7 a.m. and throw it away. Save all the urine passed after this up to 7 a.m. the next day. Pass the urine exactly at 7 a.m., and add it to what has previously been passed. _Qualitative Sugar Tests._ (1) Fehling's Test:--Boil about 4 c.c. of Fehling's[1] solution in a test tube, and add to the hot Fehling's an equal amount of urine, a few drops at a time, boiling after each addition. A yellow or red precipitate indicates sugar. For practical purposes in the following of a diabetic's daily urine, this is a valuable test, and the one which we always use. (2) Benedict's Test:--To 5 c.c. of Benedict's[2] reagent add 8 drops of the urine to be examined. The fluid is boiled from 1 to 2 minutes and then allowed to cool of itself. If dextrose is present there results a red, yellow, or green precipitate, depending upon the amount of sugar present. If no sugar is present the solution may remain perfectly clear or be slightly turbid, due to precipitated urates. This is a more delicate test than Fehling's. [1] Fehling's solution is prepared as follows: (a) Copper sulphate solution: 84.65 gm. of copper sulphate dissolved in water and made up to 500 c.c. (b) Alkaline tartrate solution: 125 gm. of potassium hydroxide and 178 gm. of Rochelle salt dissolved in water and made up to 500 c.c. These solutions are kept in separate bottles and mixed in equal volumes when ready for use. [2] Benedict's solution has the following composition: