[Illustration:  Fig. 52.—­Dr. Robertson’s results.]

[Illustration:  Fig. 53.—­Service Guncotton for Cordite made at a Private Factory.]

Readings are taken thirty minutes from the time the nitro-cellulose is heated, and are taken at intervals of fifteen minutes for about four hours; fresh caustic potash is added every thirty minutes or so.  It is convenient to plot the results in curves.  The curves given in Fig. 53 are from gun-cotton manufacturers in England at a private factory.  The rate of evolution of nitrogen is as follows:—­

In 1 hour.  In 2 hours.  In 3 hours.  In 4 hours. 
   N. N. N. N. in milligrammes.
  1.25 2.55 4.5 5.75
  1.5 3.25 5.25 6.75
These results are very satisfactory, the gun-cotton was of a very good quality.  Several hours are necessary to remove all the air from the apparatus.  Dr Will stated fifteen minutes in his original paper, but this has not been found sufficient.  It has not been satisfactorily proved that Will’s test can be applied to gelatinised nitro-cellulose powders.  It is convenient to plot the results in curves; the nitrogen is generally given in cubic centimetres or in milligrammes, and readings taken every fifteen minutes.  The steepness of the curve is a measure of the stability of the nitro-cellulose which is being examined.  The steeper the curve the more nitrogen is evolved per unit of time, and the less stable the nitro-cellulose.  In the case of unstable nitro-celluloses heated under the conditions described, the separation of nitrogen is much greater at first than at a later period.  If the nitro-cellulose be very unstable, explosions are produced.  If the separation of nitrogen is uniform during the prolonged heating, then the nitro-cellulose may be regarded as “normal.”  If it be desired to determine the absolute amount of nitrogen separated from a nitro-cellulose, the following conditions must be observed:—­(1.) Accurate weighing of the nitro-cellulose; (2.) Determination of the amount of air in the CO_{2}, and deduction of this from the volume of gas obtained; (3.) Reduction of the volume of the gas to the volume at 0 deg.  C. and 760 mm. pressure.[A]

[Footnote A:  See also Jour.  Soc.  Chem.  Ind., Dec. 1902, pages 1545-1555, on the “Stability of Nitro-cellulose” and “Examination of Nitro-cellulose,” Dr Will.]

Bergrnann and Junk[A] describe a test for nitro-cellulose that has been in use in the Prussian testing station for some years.  The apparatus consists of a closed copper bath provided with a condenser and 10 countersunk tubes of 20 cm. length.  By boiling amyl-alcohol in the bath, the tubes can be kept at a constant temperature of 132 deg.  C. The explosive to be tested is placed in a glass tube 35 cm. long and 2 cm. wide, having a ground neck into which an absorption bulb is fitted.  The whole apparatus is surrounded by a shield, in case of explosion.  In carrying