24

          At every change of diet the same routine was observed, i.e.,
     the prisoners were kept on the diet two or more days before collec-
     tion of their excreta for analyses.

III. The feeding of the prisoners was carried out on the following plan:—

          The amount of each separate item of the diet was weighed dry
     before cooking, again weighed after cooking, and then distributed
     to each separate batch of prisoners under observation. Any cooked
     food that remained uneaten was weighed, and its amount in the
     dry form calculated and subtracted from the total dry material of
     the diet. The total value of the diet in proximate principles.
     was then determined by analyses of the food materials actually
     in use at the time of the experiment.

IV. The collection of the excreta:—The pooled urine and pooled fæces
     of the prisoners forming the batch were collected, measured—or
     weighed as the case may be—and samples were taken as required.
     This was done usually for five days, sometimes more, for each
     set of observations.

          In the investigations carried out in Calcutta, there was no
     difficulty as the excreta were sent to the laboratory every morning;
     but for jails hundreds of miles away from Calcutta we had to make
     special arrangements. The plan that we finally adopted and which
     gave very satisfactory results, was to have the urine and fæces.
     measured or weighed, as the case be, on the spot and samples put
     up at once: for nitrogen with the usual amount of pure sulphuric
     acid; for ammonia with calcium hydrate; for urea with
     the barium hydrate and chlorate mixture, then the ether and
     alcohol solution was added: the fæces were weighed, thoroughly
     mixed and samples weighed and put up with sulphuric acid for
     nitrogen; or dried, weighed, and put up. By following this method
     we eliminated all source of loss of nitrogen from decomposition,
     and were able to make a week's collection before sending the
     samples to the laboratory. This was a matter of some importance
     for some of the jails dealt with were over 300 miles from
     Calcutta—one being 411 miles away.

          The samples obtained in this way were collected in reagent
     bottles with ground-in glass stoppers; for safety, the stoppers were
     sealed and tied tightly in, and the bottles were packed in a parti-
     tioned box—a partition for each bottle—with saw-dust, and care-
     fully guarded the whole way to the laboratory to prevent breakages.
     We had practically no breakages with these samples.