having made the preceding statement, delivered the opinion of the court.
The patent which is now in question is for the basic or generic invention of Jeavons, upon which he devised several improvements which were the subjects of other patents enumerated in the foregoing statement, and which have fallen out of the contest. His application for it was filed December 20, 1888. He stated therein that his invention related to hydrocarbon burners, and consisted in the method of construction described by him in his specifications. He recognizes the previous constructions, and their manner of use, in the language following, *854which we copy, believing that it fairly states the existing conditions at the date of his application. He says:
“In the construction and operation of burners prior to my invention, different ways of obtaining a distribution of hydrocarbon oils or vapors or carbureted air have been known, among which may be mentioned: First, the distribution of oil by capillary attraction, as by a wick, in which the oil is drawn to the surface of a wick and consumed, as in an ordinary lamp. In this style of cases vaporization occurs directly at the point of combustion, and the oil itself is distributed. Secondly, by spraying the oil by means of a jet of air or steam under pressure. Thirdly, by generating the vapor in a suitable retort, in which the vapor is subjected to a head or pressure, and depending on the artificial pressure in the retort to distribute or feed the vapor. This style of burner is exemplified in the well-known vapor burner which feeds through a jet orifice. Fourthly, by evaporating or vaporizing gasoline or other light hydrocarbon on an exposed surface by passing a current of air over the same and then feeding the carbureted air to the burner, the old and well-known carbureting devices being of this class.”
He then proceeds to state that his own method, differing from those above described, involves, first, the conversion of the oil into vapor by exposing the oil to a heated surface, and then distributing or conveying the vapor by its gravity to the place or places where or about which the vapor is supplied to the burner and maintains combustion. His explanation of his method and the means devised by him to accomplish it are somewhat lengthy, and we epitomize so much of it as seems necessary to understand the claim. To do this we insert Fig. 1 of the drawings, which shows one, and probably the most generally used, form of his burners, in central vertical section.
C and D are two perforated concéntric metallic cylinders resting on a base shown below them. Around the inner side of the base is a cir*855cular trough, closed on the outside, but opening upward into the space between the cylinders. This trough is seen at the right and left hand in the base. The asbestos rope, L, lies in the bottom of it. The oil is admitted into the trough through the tube, H, and is controlled by the valve, the stem of which is K. In operation, when oil or gasoline is used, enough is let in to saturate the asbestos cord, L, which being lighted heats up the trough. The oil as it comes out of the tube, H, is immediately vaporized by the heat, and, the vapor being two or three times heavier than air, sinks and flows around in the trough and fills it. The oil expanding into vapor fills several hundred times its own space, and as the latter piles up, so to speak, it passes into the perforated combustion tubes, when, on receiving the air through the perforations and being lighted, it burns with a blue flame ascending through the top of the cylinders, where its combustion is ended. The trough is kept hot by the conductivity of the heated metal of the burner, and condensation of the vapor on the bottom is thereby prevented. If the asbestos is employed for the initial heating up, it is not longer used after the process is under way. The patentee lays stress upon the fact that as the vapor is formed it falls by gravity and flows around through the trough, whereby an even foundation or source of supply throughout the entire circuit or length of the trough is secured, and, by consequence, an even flame in all parts of the combustion chamber. This is the purpose of constructing the parts in such form as that the trough shall be on a level below the entrance of the oil where the vaporizing takes place. We are satisfied that this is substantially the manner in which the vapor is distributed and supplied to the combustion chamber. It is evidently so for a time, at least, after the beginning of the vaporizing, and is probably true in a modified degree after the trough is filled by the vapor.
The claim is here set forth:
“A hydrocarbon vapor burner, consisting of a vapor holder constructed for the free and uniform distribution of the vaj>or therein by gravity, and having a free opening for the escape of vapor, in combination with perforated combustion w'alls having a flame space between them, in communication with the said holder, substantially as described.”
The court below found difficulty in believing that the principle of gravity had anything to do with the operation of the burner, and was disposed to discard the theory of the patentee, on which his apparatus was constructed, as unfounded. But the fact is that, by constructing the burner in the manner prescribed by him, the vapor is produced and distributed to and in the combustion chamber in a very satisfactory and useful way. That it is a successful improvement on all former methods is shown by the general adoption of it by the public, no less than 122,-000 burners of this kind having been sold within 2J^ years. It may be that tlie patentee did not fully understand the rationale of the manner in which his construction effected the results. And it may be that the expert witnesses have not in all respects correctly apprehended it. But if the fact be that his construction does effect the results, and they are beneficial, he is none the less entitled to the benefit of his invention though he may not have correctly understood the principles of its operation. Andrews v. Cross, 19 Blatchf. 294, 8 Fed. 269, per Judge (after*856ward Justice) Blatchford, approved in the Driven Well Case, Eames v. Andrews, 122 U. S. 40-55, 7 Sup. Ct. 1073, 30 L. Ed. 1064; Walker on Patents (4th Ed.) § 175, and cases there cited.
Certain prior patents are cited as anticipations of Jeavons’ supposed invention. We will examine them in the order of their dates. First in this order are three patents to Morrill: No. 18,465, dated October 20, 1857; No. 44,548, dated October 4, 1864; and No. 60,224, dated December 4, 1866. These were burners having a wick rising out of an oil reservoir. The oil was carried up by the wick, and the combustion was at its upper end on the instant of vaporization at that point. The resemblance of the structures covered by these patents, and their mode of operation, to that of the Jeavons patent, is so slight as not to require discussion. They belong to the wick-burner class, and the differences are radical. The next is a patent to Brown, No. 60,680, dated January 1, 1867, which is also a wick burner belonging to the same class as the Morrill patents. Patent No. 127, 236, to Rogers, dated February 13, 1877, was for an improvement on a former vapor burner invented by him which is not shown. Whether it was designed for using any heavier kind of fluid than gas or gasoline, we do not know. From the indications, we should suppose not. However this may be, he did not make any provision for taking down the vapor after it was generated into the trough below for distribution by gravity, nor was any such thing contemplated by him. In his device the vapor was formed as the fluid ascended, and the combustion took place in the chamber where the fluid was vaporized. The vapor was generated simultaneously all around the chamber, and not at one point, as in the Jeavons patent, after the initial heating. There are other differences not so important. Patent No. 281,107, to Miller, dated July 10, 1883, shows a burner with an indestructible wick, as asbestos, on the surface of which combustion took place; in short, combustion took place simultaneously with the formation of the vapor. There was no independent vapor trough such as is itsed in the Jeavons construction, nor was it intended to effect the diffusion of the vapor in the same way. A patent to Stephens, No. 286,862, dated October 16, 1883, for a lamp stove, showed a reservoir of oil over which was spread an absorbing mat which was saturated with the oil. Combustion took place on the surface of this mat. The vapor was burned as soon as it was generated, the two processes occurring together. It was another form of wick burners, and bears no resemblance that we can perceive to the patent in suit. There was no provision whatever for the distribution of the vapor, other than such as would ensue from the ignition of the surface of an outspread, saturated wick. A patent to Carsley, No. 397,630, dated February 12, 1889, shows a form of burner in which gas or vapor was used. In use the reservoir was located in the fire pot of a stove or some heating apparatus, whereby the action of the vapor or gas was accelerated. From this and other parts of the specifications, we gather that it was designed to use gas or vapor already generated. There is no mention of any method or means for generating vapor, but only of means for conducting gas or vapor and distributing it to the place of combustion. It lacks, therefore, the essential characteristic of the patent to Jeavons. It has no *857use for such features. It is proper to observe, also, that the means employed for effecting the distribution and delivery of the vapor into the combustion chamber are quite different.
A patent to Bodwell, No. 320,533, and some others are referred to. I >ut it is unnecessary to discuss them further in detail, for they are still more remote from the patent in suit than those already considered. It must be admitted that the devices we have described were designed to effect similar results, in general, but they did not do it in the same way as that contemplated by Jeavons, and the burners having no wick were not successful. We are therefore of opinion that nothing is shown which would justify the conclusion that the patent in suit, in respect to this first claim, is invalid because anticipated. Moreover, the proofs show that this invention contributed quite decisively to the success of the class of burners which dispenses with a wick, and effectively generates, distributes, and consumes the vapor by a well-arranged combination of means which had not been previously disclosed.
But the defendant contends that the proceedings in the Patent Office show that the applicant, Jeavons, was constrained to surrender the substance of his first claim, and acquiesced in the requirement of the office. If this was so, of course it would be fatal to his claim to the extent of the matter surrendered, and the claim would be limited to the bounds of that which he was permitted to retain. It is necessary, therefore, to find out what the original application was in respect of the matter of this first claim, and follow its essential characteristic through the proceedings to the final allowance of the claim, for it must be conceded that if that feature was not disclosed by the application, but was brought in by amendment, the claim would not be valid. We are not required to pursue other matters which were considered in the office. The drawings which accompanied the application when filed remained unaltered throughout the proceedings, so far as this special feature was concerned. The functions which the apparatus disclosed by the drawings performed were in no wise affected by any changes made in the specifications. A burner made according to their construction would operate in the way to be expected from the claim. It is true that the patentee did not fully perceive nor correctly perceive, as we think, the principles on which the operation of his burner proceeded. For instance, we more than doubt whether he appreciated that the principle of gravity co-operated in the method of his invention. But he did see and know that the burner he had devised would successfully accomplish the results he anticipated and was laboring for. It is clear enough that while the application was pending he came to understand the operation of his device better than he had, and that his first claim was founded on that better understanding. These are the essential facts on which the issue as to whether such new matter was brought in and claimed as would defeat the validity of the claim must be tested. We think this issue must be determined in favor of the patent. In the case of Michigan Central R. R. Co. v. Consolidated Heating Co., 67 Fed. 121, 14 C. C. A. 232, we had occasion to consider the question of the validity of amendments made while the application for a patent was pending. We there held that the validity of such an amendment de*858pended upon the question whether it brought in original matter, or was of something that might be fairly deduced from the original application. In the first instance the amendment would not be justified; in the latter it would. The case was decided upon that distinction. And this is the rule which is elsewhere recognized. Hobbs v. Beach, 180 U. S. 396, 21 Sup. Ct. 409, 45 L. Ed. 586; Western Electric Co. v. Sperry Electric Co., 58 Fed. 186, 7 C. C. A. 164; Sugar Co. v. Yaryan Mfg. Co. (C. C.) 43 Fed. 140. This distinction harmonizes with the doctrine that the benefit secured by an invention extends to all the uses of which it is capable, whether the inventor had them all in contemplation or not, and the other rule, to which we have already referred, that it does not matter that he does not understand the principles on which his device operates. If he discovers new uses to which his invention may be put, or discerns the principles thereof more clearly, while his application is pending, in neither case is there any new invention, nor any enlargement of the old. And, that being so, there can be no legal objection to his so molding his claims as to secure all his invention discloses.
Another objection urged is that the invention covered by this claim had already been patented by former patents issued to the same patentee during the pendency of the application for the patent in suit. The facts upon which this objection is based are these: The application for the patent in suit was filed December 20, 1888, but the patent did not issue until May 24, 1892. In the meantime Jeavons had filed three or four applications for as many several improvements on his original invention. In them he described his original structure, as well as, in each case, the improvement he proposed. He then added claims which embodied his improvements as applied to the original structure. But he did not therein claim his original structure. Nor did he refer to his original application or disclaim the subject thereof. On each of these subsequent applications patents were issued prior to the issuance of the patent in suit. In support of the defense that the patent in suit is void because thereby the invention was doubly patented, the defendant cites: Miller v. Eagle Manufacturing Co., 151 U. S. 186, 14 Sup. Ct. 310, 38 L. Ed. 121; Thomson-Houston Electric Co. v. Ohio Brass Co., 80 Fed. 712, 26 C. C. A. 107; Palmer v. Lozier, 90 Fed. 734, 33 C. C. A. 255; Dayton Fan Co. v. Westinghouse, 118 Fed. 573, 55 C. C. A. 390. This subject has been so fully elaborated by us, especially in the last-cited case of Dayton Fan & Motor Co. v. Westinghouse Co , that we do not think it necessary to go over the groitnd again. The cases of Miller v. Eagle Mfg. Co. and Palmer Pneumatic Tire Co. v. Lozier are not applicable. As was explained in the Dayton Fan & Motor Co. Case, when the patent first granted is distinctly and only for an improvement on another invention which is already the subject of a prior application then pending, and on which a later patent is granted, the patent for the improvement in no wise interferes with the other application or the patent issued thereon, for the reason that the patents are for separate and distinct inventions. In just such a case as this we held that the later patent, being one for the generic invention, was not invalidated by *859reason of the issue of a previous patent in which improvements upon the other only had been patented. We had already so held in Thomson-Houston Electric Co. v. Ohio Brass Co., supra. It is true that in the applications for these patents for improvements there was no express disclaimer or renunciation of the matter of the former application. But that was unnecessary. That application was pending and being prosecuted in the Patent Office, and the fact that the applicant for the improvement patents did not intend to release his former invention to the public was as well understood as if he had in express terms said so. In order to explain the basis of the improvement patents, it was necessary to state what the improvement was upon, and how it fitted it. Having done this, he claimed what was new, and thereby distinguished what his patent was intended to include.
It is further contended that the claim is for a function of the apparatus, and, again, that it is for a process, two rather inconsistent propositions, neither of which is well founded. It is clear that the claim, when construed by reference to the specifications and drawings, as it should be, is for definite means devised to perform desirable functions, but it is not for the functions themselves. That it is not for a process is obvious.
We think none of the defenses resting upon the ground that the claim in question is invalid can be sustained. If the matter were in doubt, we should still incline to the same conclusion, because of the concurrence of so many makeweights, the granting of the patent after protracted examination and discussion, the obvious utility of the invention, the favor with which it has been received by the public, and the recognition of its merit, testified by the defendant’s adoption of it. It is not denied that, if the patent is held valid, the defendant infringes it.
The decree of the Circuit Court, in so far as if relates to the first claim of patent No. 475,401, is reversed, with costs. In all other respects the decree is affirmed. The cause will be remanded with directions to enter a decree for complainant for an injunction, and for profits and damages, to be ascertained. No costs will be recovered by either party in this court or in the court below; each party having succeeded only in part.