is considered with special reference to De Laval's type of impulse turbine, of which type the Curtis is, as we have seen, an improvement. For it must be conceded that, whatever principle of pressure staging Moorhouse disclosed, anything he disclosed was not applicable to the high-speed impulse turbine which De Laval produced by his nozzles where there is no pressure difference at the inlet and outlet ends of the moving vanes, for, prior to De Laval, as we have seen, no one (and of course, Moorhouse) dealt with the then unknown condition of a pressure drop created solely in the nozzles. And, indeed, Gentsch, who in his Dampfturburen (an authority quoted by one of the respondent's witnesses as "a well-known member of the German Patent Office and a very high authority on steam turbines"), while classifying Moorhouse's turbine as an impulse one, wholly disassociated him and other designers from the De Laval type, saying:

"The steam which expands outside the nozzles, and which in the free jet wheels is mostly made to perform work during the period of expansion, is able to convert only a small portion of its pressure energy into current energy, so that the working of the velocity turbines hitherto discussed has not given a satisfactory economical result. * * * A better state of things was produced for the first time by the invention of De Laval."

Finding, then, as we do, that the disclosures of the Moorhouse patent had no helpful bearing or practical effect on the impulse turbine art, and supported in that conclusion by the fact that its vagueness is such that fair-minded witnesses in this record greatly differ as to what its disclosures really are, we are not warranted in attributing to it any effect in the way of vitiating, or even minimizing, the work of Curtis. We pass on to the Mortier article.

In 1890, Rateau, a French savant, read before the Society of Mineral Industries of France two papers on the Parsons turbine, which had been lately exhibited at the Paris Exposition. In his first paper, Rateau discussed that turbine, stating its advantages and disadvantages. Several members expressed their views upon it, following whom M. Mortier stated "that this form of motor utilizes the complete expansion .of steam," whereupon the president inquired, "What advantage is gained by using the steam in the form of velocity instead of using it in the form of pressure?" Mortier's subsequent remarks were evidently prepared with reference to this question, and in order to gather their significance it is important to determine what the president's question raised, and how it was understood by those present, and how it was acted upon. That it meant a comparison of the worth of a reciprocating engine and some turbine is clear. But what turbine? Respondent contends it covered impulse turbines. We cannot accede to this view. The question was raised by the president, not by Mortier, and, as we have seen, was called forth by the assertion of Mortier, who was apparently completely satisfied with the Parsons turbine, "This motor utilizes the complete expansion of steam." Mortier was seeking or suggesting no other form or type of turbine, and the president, then, in substance, put the question as one between the Parsons turbine and a reciprocating engine. Certainly Rateau so understood the question, for he answers "that he intends to treat this question and to complete his communication (which was based wholly on the Parsons tur

bine) at a future meeting"; and the society so understood, for its minutes state:

"Order of the day for the meeting of April 12, 1890: The Parsons' Steam Turbine."

Moreover, Rateau's subsequent paper was based on the question between Parsons and the reciprocating engine, opening with the state

ment:

"I wish to-day to enter upon some considerations, theoretical for the most part, which will permit me to compare this new kind of motor with ordinary steam engines and to arrive at an estimate as to the probable future in store for it."

As if to emphasize and limit himself to this single issue, he announces his satisfaction with the Parsons machine, saying, "New types will undoubtedly succeed one another, and there is reason to expect within a short time the complete solution of the question already fitly answered by the Parsons system," and disposed of another type (Dow's) lately introduced, which he estimates as * * inferior, from various points of view, to that of M. Parsons," and of which * * in its present condition the system would not be of a nature to be widely introduced in practical industry." He then takes up the Parsons, as the turbine basis of comparison with a reciprocating engine and states his conclusions, which need not be quoted.

The minutes then state, "Continuing the preceding communication, M. Mortier gives the following information on the same subject." Without entering upon a discussion of Mortier's statements and calculations, it suffices to say that to us the inherent proofs of the proceedings show that they are directed to the Parsons type, which, as we have seen, was a reaction turbine. There was nothing in the subject before the society to suggest the introduction or discussion of impulse turbines. That meeting was discussing a particular reaction turbine; it was practical and efficient; and they had seen it operate. It was the contrast of this practical device with steam engine practice the society was discussing. There was no necessity for discussing impulse turbines, for no one had then produced one that was practical and efficient. And, as we have seen, no engineering basis of fact existed prior to De Laval for any speculation as to the future of the impulse turbine. If the striking effects of pressure staging and velocity-compounding impulse turbines, which afterwards gave them efficient working value, were then realized and disclosed by M. Mortier's paper, he did not claim them in his paper, his subsequent acts were in conflict with such a claim, and the engineering world ignorantly suffered years to pass and misguided efforts, in other directions, to be made in the face of such disclosures. Indeed, if Mortier's address be assumed to apply to impulse turbines and to disclose Curtis's mode of overcoming their failings, Mortier's subsequent acts are inconsistent with such assumption. When he subsequently took up the subject of minimizing the steam speed, it was not, as shown by his two French patents of 1894 and 1895, on the principal of operation now alleged to have been disclosed by him, to wit, the principle of eliminating such speed, but on the principle of controlling such speed by mixing live steam with

hot water or exhaust steam. This system, which as now known resulted in a loss of from one-half to three-quarters of available steam energy, shows that Mortier, instead of anticipating Curtis in his disclosure, followed in the lead of those inventors of whom Stodola said:

"The majority of the older patents showed lack of knowledge of the laws of steam flow. One idea especially led inventors on in spite of constant failure; to decrease the velocity of the steam by mixing it with fluids or gases."

[4] We next turn to the question of infringement. The disclosures of Curtis's patent, as we have seen, consisted, broadly stated, of pressure staging an impulse turbine, the velocity-compounding thereof, and the abstraction, at each passage of the steam, substantially all or the principal part of the vis viva developed at the preceding stage. Without discussing the proofs in detail, we may say we find these features in the respondent's turbines. The proofs show the proposals made by them to the government for equipping certain vessels with turbines and a guaranty that certain results will be obtained. We are warranted therefrom in assuming the respondents meant to comply with their representations and contract guaranties, and, in the absence of any proof by them tending to give the court light on exactly what form of turbine they are constructing, we are, under the authorities (Peifer v. Brown [C. C.] 85 Fed. 780; Celluloid Co. v. Arlington Co. [C. C.] 85 Fed. 449), justified in resting on the proofs of complainants before us. These show that the principle of operation of respondents' turbine is distinctively impulse, that it is multi-pressure staged having 32 pressure stages, 12 stages having two velocity rows each and 20 stages one row each. On the same shaft is mounted also a reversing multipressure staged turbine having three pressure stages with two velocity rows each and the rest with one. We agree with the deductions drawn by complainants, based on calculations on data as to bucket speed and steam speed furnished by complainants' witnesses, that the abstraction of vis viva by respondent's turbines is substantially and practically complete, the unused velocity amounting to 2.29 per cent. the energy, and this conclusion is confirmed by the standard of efficiency guaranteed to the government by the respondent under the designed full speed conditions. That when operated under other conditions such turbines might abstract lesser amounts of vis viva does not relieve the turbine of its infringing character. Being designedly made capable of infringement, its capacity to infringe warrants the conclusion that it does infringe. It is contended, however, that infringement of the Curtis patent is not established unless there is an absolute and total abstraction of vis viva. We find no warrant for this contention in the specification or claims of that patent, and we find no ground in reason or thermodynamic practice for such extreme contention. The economies of fuel, power, and indeed all motive mechanism, are necessarily only approximately perfect. Waste, loss of motion and power are incident to all mechanical, thermal, and motor operations, and the effort is to reach substantial, practical results rather than absolute theoretical ones. And such substantial abstraction was the measure Curtis disclosed in his specification.

*

"My object is to develop mechanical power from steam or other elastic fluid under pressure by utilizing a large proportion of its vis viva in a turbine, whose speed of rotation shall be low. I deliver the flowing jet to a movable element of the apparatus consisting of one or more circular ranges of vanes forming passages through which the jet passes and in which the direction of flow is changed, so as to extract its velocity wholly or largely whereby the vis viva developed in nozzle or passage is wholly or largely converted into mechanical rotation."

And the same thing is embodied in several claims in the words:

"Said vanes being adapted to abstract at each passage there through substantially all or the principal portion of the vis viva developed at the preceding stage."

In the same way we find no warrant in the patent for restricting the nozzles or passageways to the expansion pipe. We have already pointed out earlier in this opinion that the patentee stated parallel and diverging nozzles were alternative constructions. It is contended, however, that Curtis by his definition of expansion nozzles in another application to which this patent refers so restricted himself. But the fact is that this definition was not embodied in that application when the reference was made. Its subsequent introduction in such former application was for reasons involved in that particular application, and just principles of construction do not necessitate it being retroactively applied to a patent which expressly negatived both in specification and figures any such restricted meaning. The partial peripheral introduction of the steam has been emphasized in complainants' testimony as a feature of marked advantage in impulse turbines and which distinguished them from the reaction type. In his specification Curtis lays stress on this feature as one characteristic of all his passages and as distinguished from introduction in annular form, saying:

"It is the design of my present invention, as of the apparatus of my prior application referred to, to employ at the delivery end of the nozzle and in the working passages a 'jet' of steam or other elastic fluid, i. e., a practically solid stream of fluid having an oblong form in cross-section whose thickness bears a considerable proportion to its width, so that its cross-sectional area will be large compared with its perimeter as distinguished from an annular film of elastic fluid whose cross-sectional area is small compared with its perimeter. By this means the frictional retardation is greatly reduced and the efficiency is largely increased."

It is manifest, therefore, that a turbine which while it delivers "a fluid jet to a portion of the vanes within the first shell," but not to the succeeding ones, does not infringe a claim, one of the elements of which is "intermediate passages connecting the different shells together and delivering the fluid jet to a portion of the vanes of the different sets in succession." Gauged by these general conclusions, we find that, with the exception of the seventh and tenth, all of the claims charged. are infringed.

[5] We next turn to patent No. 595,435, the first, second, third, and fourth claims whereof are alleged to be infringed. The object of Curtis, as stated in his application, was "to produce an elastic turbine op-. erating under conditions of high efficiency in which variations in speed may be effected without great variations in the efficiency of operation." This he accomplished by constructing and arranging the fluid passages

of the turbine and their connections in such a way that the elastic fluid may be caused to traverse the movable vanes a greater or less number in succession." He states, "The general plan of the elastic fluid turbine being such as is described in patent No. 566,969, issued to me September 1, 1896." The proofs show that for efficient operation the vane velocity should be about one-half the velocity of the steam action upon the vanes where the velocity is abstracted by a single set of vanes, and in like proportion if the velocity is fractionally abstracted by two or more sets of vanes velocity-compounded, and consequently, generally speaking, the vane velocity should be higher the fewer the number of stages into which the pressure drop is divided. This principle is used by Curtis, whose device, shown in the accompanying figure 1, is so arranged that the number of stages into which the pressure drop is divided may be varied according to the rotary speed at which it is desired the motor should be driven; a less number of stages being used for higher speeds and a greater number for lower speeds.