showed a decrease of temperature.  The difference of the two temperatures, which was at starting 1.3 deg.  Cent., consequently increased to 7.2 deg.  Cent, after 17 min., the boiler having then its lowest temperature of 148.8 deg.  Cent.  After that both temperatures rose together, the difference between them increasing slightly to 9.5 deg.  Cent., and then decreasing continually.  After 2 hours 13 min., when the engine had made 12,000 revolutions, the soda solution had reached a temperature of 170.3 deg.  Cent., which proved to be its boiling point.  The steam from the engine was now blown off into the open air during the next 24 min.  This lowered the temperature of both water and soda lye by 10 deg. and re-established its absorbing capacity.  The steam produced under these circumstances had of course a smaller pressure than before, in this way the engine could be driven at reduced steam pressures until the resistance became relatively too great.  The process described above is illustrated by the diagram Fig. 1, which is drawn according to the observations during the experiment.

[Footnote 3:  Zeitschrift d.  Vereins Deutscher Ingenieur, 1883, p. 730; 1884, p. 69.]

[Illustration:  FIG. 1.]

[Illustration:  FIG. 2.]

The constant rise of both temperatures during the first two hours, which is an undesirable feature of this experiment, was caused by the quantity of soda lye being too great in proportion to that of water, and other experiments have shown that it is also caused by an increased resistance of the engine, and consequent greater consumption of steam.  In the latter part of the experiment, where the engine worked with expansion, the rise of the temperature was much less, and by its judicious application, together with a proper proportion between the quantities of the two liquids in the engines, which are now in practical use, the rising of the temperatures has been avoided.  The smaller the difference is between the temperatures of the soda lye and the water the more favorable is the economical working of the process.  It can be attained by an increase of the heating surface as well as by a sparing consumption of steam, together with an ample quantity of soda lye, especially if the steam is made dry by superheating.  In the diagrams Figs. 3 and 4, taken from a passenger engine which does regular service on the railway between Wurselen and Stolberg, the difference of the two temperatures is generally less than. 10 deg.  Cent.  These diagrams contain the temperatures during the four journeys a b c d, which are performed with only one quantity of soda lye during about twelve hours, and show the effects of the changing resistances of the engine and of the duration of the process upon the steam pressure, which, considering the condition of the gradients, are generally not greater than in an ordinary locomotive engine.  It can especially be seen from these diagrams that an increase of the resistance is immediately and automatically followed