23

          The Carbohydrate of the food-stuffs analysed was estimated
     by completely changing all the carbohydrate material into sugar
     by the action of hydrochloric acid, and then estimating the per-
     centage of sugar by Fehling's method, checking the results by
     the Pavy-Fehling method or the Picro-Saccharometer.

          The Fat was estimated in the usual way by means of
     a slightly modified Soxhlet apparatus manufactured to our
     drawings by Messrs. Baird and Tatlock, London.

         The moisture and ash require no comment.

II. The method of procedure when we came to investigations on prisoners
     was almost invariably to use as our unit a batch of five picked
     prisoners—in a few cases we had four in the batch, in some
     six and in some ten. The excreta from these five men—or from
     the batch—were pooled, and in every case two samples of the urine
     and two of the fæces—when examined—were put up fór nitrogen,
     and a sample of each for chlorides; one of the urine for ammonia
     and sometimes the urea, phosphates, purin bodies, etc., of the
     urine were also estimated. The specific gravity and reaction of
     the urine was always taken. The weight of the prisoners making
     up the batch was taken every day; it was noticed that while
     individual prisoners varied in weight from day to day, the average
     weight of the batch after the first three or four days remained
     practically constant.

          The prisoners to form the subjects of the investigations having
     been chosen, were isolated from the other prisoners. In the larger
     jails we were able to make use of a block of cells with a common
     yard where the work given to the prisoners could be carried out,
     and where they were able to get air and exercise while at the
     same time the facilities for feeding and collection of excreta were
     much greater than was the case in the smaller jails. In the
     smaller jails a large block of cells was not available for this
     purpose; we were always able however to obtain some godown
     or store for the isolation of the batches.

          The next step was to put the batches on the diet that we wished
     to investigate. This was usually done for some days before any
     collection of the urine and fæces for analyses was made, in order to
     give time for the proper amount of protein to be absorbed and
     nitrogenous equilibrium to be established, and to allow of the usual
     increase in body-weight which we found to occur when the
     prisoners were first put on a diet weighed and distributed by us.