pairing its lubricating quality. This was easily overcome by modifying the vents and employing glands of different construction.

Some time was also required after erection to make necessary adjustments to relieve the turbine of longitudinal end thrust. This would have been corrected at the shop had the opportunity then been present for making complete test. It was found, too, that the shaft, which had been designed to afford the utmost ease of dismantling, was subjected to a considerable unevenness of temperature under superheated steam, and means were taken to make the temperature at all points more uniform. Having in due time overcome these local defects, which partook in no sense of functional fault, the turbine was then in serviceable condition, and its operation has since been most satisfactory. And the Hartford Company, notably alert to adopt the newer thing if there seemed advantage in it, have found when their water supply ran short that it paid to run the turbine and allow their Corliss engines to remain idle. This turbine is seen in Fig. 4.

Is the steam turbine efficient? And what, if it may be so termed, is the character of its efficiency? Is it, like the various types of piston engines, peculiarly fitted to certain conditions which permit of little change if economical performance be retained, or is there evidence that the turbine has a greater inherent efficiency that is less affected by attending circumstances?

The interest of engineers in the turbine has, perhaps, been drawn chiefly to the evident possibilities of its steam economy, and to the data already acquired, with the discussion it has provoked, much more of value will be added. We may in a general way, however, without referring to its thermodynamics, obtain from the evidence of actual results some knowledge of its efficiency and determine if the standards of present practice may not be improved.

It is well that the makers of the turbo-generator have been compelled to adopt the practice of basing the steam consumption on the unit of output, so that their guarantees are given on the electrical horse power or kilowatts delivered on the switch board, and not on the indicated horse power developed. This at once eliminates the factors of engine friction and generator loss, and thus more definitely establishes a measure of performance.

One is impressed with two distinguishing features of the turbine's steam efficiency, namely, that it seems to vary but little over wide ranges of load, and, further, that the size of the unit has comparatively little bearing. It follows, then, that if good results are possible at all, they are neither restricted to the larger plants nor to the requirement of steady load.

Fig. 5 illustrates this. Herein are given the results of tests on

a 400 K. W. turbine, made at the builders' works before shipment; the machine having since been in daily operation some eight months. These tests were conducted under brake load, so that the figures are based on the brake horse power developed. The rated load would be about 600 B. H. P. The steam consumption curve is seen to be very flat, graduating from 14.47 lbs. at full load, to 16 lbs. at half rating, and to less than 19 lbs. at one-quarter capacity. The relation of the consumption of steam in pounds per hour to the brake horse power developed is also shown, this line being almost straight. In the tabulation may be observed the interesting comparative effect of vacuum and superheat.

If it is thus shown that with a unit as small as 400 K. W. we inay obtain a result of 14.47 lbs. of steam per brake horse power per hour, corresponding to less than 1334 lbs. per I. H. P., it is evident that moderate-sized plants may with the turbine be sufficiently subdivided to give the maximum flexibility of service, with insurance of relay, and yet possess an efficiency heretofore identified only with very large units. Further than this, a fluctuating load is not incompatible with high economical performance.

As the units become larger the turbine is then brought into comparison with the best steam engine practice, where it still preserves its uniform efficiency, and where its practical advantages are no less evident. In a recent instance, a result of 11.7 lbs. of steam per electrical horse power per hour was guaranteed on a turbine of 750 K. W. capacity, corresponding to about 10.17 lbs. per I. H. P., which, though the size is moderate, is perhaps within the ability of but few engines, of any size or type, that have ever been built.

It may be pertinent to cite a few results obtained in regular service. The turbine at Hartford, under test conducted by Prof. Robb, at an average load of 1,800 K. W., with 155 lbs. steam pressure, 27 inches vacuum and 45 degrees superheat, gave a result of 19.1 lbs. of steam per kilowatt hour; or an equivalent of about 11.46 lbs. per I. H. P. hour. An interesting comparison has been made at this plant of the relative efficiency, under regular operating conditions, of the turbine and their Corliss engines. They have one 18 and 34x48, and one 24 and 44x60 cross-compound horizontal Corliss engine. These engines drive direct by belt one 400 K. W. and one 600 K. W. generator. The turbine is, of course, direct-connected to its generator. They have made comparisons of operation based in each case on rather extended runs. It has been found that the turbine requires in delivering 1,900 K. W. on the board about the same amount of coal that is used with the Corliss engines to deliver 925 K. W., the steam pressure and vacuum being identical in both cases; and this with the engines running at about their point of best efficiency, and