This is an appeal from the decision of the Board of Appeals of the United States Patent Office affirming the holding of the Primary Examiner rejecting as unpatentable claims 29 through 33, the only remaining claims in appellant's application for a patent relating to titrating apparatus.

The embodiment of the invention sought to be patented, as set forth in the claims, relates to an automatic apparatus for continuously titrating a moving stream of fluid. The apparatus basically consists of a first pair of electrodes immersed in the fluid for electrolytically generating a reagent which reacts chemically with the fluid which is to be titrated. The amount of reagent which is generated depends on the amount of current which is supplied to the first pair of electrodes, this current being controlled by the potential existing at a second pair of electrodes. The second pair of electrodes (or other sensing means) detects when an end point has been reached, that is, when the amount of reagent being generated by the first electrodes is sufficient to completely react with the fluid of unknown concentration which is being titrated. If there is any departure from the end point, the second electrodes, by supplying a signal to an electronic amplifier, cause said amplifier to change the amount of the current supplied to the first electrodes, this change in current in turn changing the amount of reagent being generated, so that an end point is again reached. The amount of current supplied to the first electrodes to maintain the reaction at an end point is proportional to the amount of reagent which is generated, and a measure of this current is therefore a measure of the concentration of the fluid being titrated. The "electronic power amplifier," which couples the first and second pairs of electrodes for the purpose of causing the system to operate in the aforementioned manner, possesses, as a component thereof, a reactive circuit which, in the terms of the claims, "more strongly attenuates high frequency components" of the amplifier input than low frequency components. The function of the reactive circuit, as we glean it from the specification, is to minimize amplification of rapid fluctuations of the amplifier input which are caused when the reaction approaches an end point, said fluctuations not being representative of true end point values. This results in more accurate titration.

Claim 29 is the broadest claim on appeal, and is representative of those appealed insofar as is pertinent for the purposes of this opinion. This claim reads as follows:

"29. In an automatic titrating mechanism for the quantitative determination of a first reactant in a fluid, the combination which comprises a cell into which the fluid is introduced, electrolytic means for generating a second reactant for reaction to an end point in the cell with the first reactant, end point sensing means in operative association with the cell for producing a potential difference which changes as the excess of one of the reactants in the cell changes from the end point, an electronic power amplifier of high transconductance having its input connected to the end point sensing means and its output connected to the electrolytic means to supply power thereto for generating the second reactant and so arranged that its output current varies in response to the change in the potential difference, and means including a reactive circuit connected between the end point sensing means and the electrolytic means for more strongly attenuating high frequency components of the amplifier input than low frequency components of the amplifier input."

The references relied on by the Patent Office are: Eberhardt et al., 2,297,543, Sept. 29, 1942; Eckfeldt, 2,621,671, Dec. 16, 1952 (Filed Nov. 21, 1944).

The Eckfeldt patent relates to an electrolytic titrimeter having the basic elements of appellant's device. It differs from appellant's titrimeter in that it employs an electro-mechanical system for changing the amount of current generated by a pair of reagent generating electrodes in response to a change in potential at a pair of sensing electrodes. The electro-mechanical system consists of a reversible electric motor, driven in response to the change in potential at the sensing electrodes, which adjusts the setting of a variable resistor in the circuit of the reagent generating electrodes to change the current supplied to said electrodes, and thereby change the amount of reagent being generated in order to achieve an end point.

The Eberhardt et al. patent relates to a direct voltage amplifier. Insofar as pertinent here, a part of the amplifier consists of a reactive circuit which acts as a bypass for undesired alternating voltage components of the direct voltage which is to be amplified.

The examiner and the Board of Appeals were in substantial agreement in holding that the appealed claims were unpatentable over Eckfeldt in view of Eberhardt et al. on the ground that it would not involve invention to substitute the electronic amplifier of Eberhardt et al. for the electro-mechanical amplifier of Eckfeldt primarily because they are equivalents and also because such a substitution would be apparent to those skilled in the art and would therefore not involve invention. The examiner came to this conclusion notwithstanding his recognition that appellant's "* * * end point voltage is a slowly fluctuating D. C. signal, because the total amount of reagents present require a finite time to complete their reaction * * *," and his further recognition that "Applicant provides his amplifier with circuit elements to minimize amplification of rapid fluctuations of signal voltage, which would not be representative of end point values. * * *"

The first question we must consider is whether the electronic amplifier of Eberhardt et al. is the equivalent of the electro-mechanical amplifier of Eckfeldt under the present circumstances. A test of equivalency is whether the substituted element operates in substantially the same way to produce substantially the same result as the element replaced. In re Lindberg, 194 F.2d 732, 39 C.C. P.A., Patents, 866; In re Husted, 39 F.2d 713, 17 C.C.P.A., Patents, 1002. We think that this criterion should be used in determining whether equivalency exists in the present case. The electronic amplifier of Eberhardt et al., as constructed, is capable of preventing undesirable fluctuations of output current from the amplifier; the electro-mechanical amplifier of Eckfeldt cannot do this. The use of the former therefore produces different results in the titrating system since it minimizes the output of fluctuating current from the amplifier. It is our opinion, therefore, that since the electronic amplifier does not operate in the same way as the electro-mechanical amplifier, and since it produces a different result, it cannot be considered the equivalent of the latter.

Having determined that there is no equivalency between the electro-mechanical amplifier and the electronic amplifier in the present case, we must now determine whether claim 29 defines an invention. However, there is a closely related ancillary question which must be considered, namely, whether the references were properly combined. If they were, there is no question but that claim 29 was properly rejected since the combination of references meets all the limitations of the claim.

It is too well settled for citation that references may be combined for the purpose of showing that a claim is unpatentable. However, they may not be combined indiscriminately, and to determine whether the combination of references is proper, the following criterion is often used: namely, whether the prior art suggests doing what an applicant has done. In re Fridolph, 134 F.2d 414, 30 C.C.P.A., Patents, 939; In re Dalzell, 152 F.2d 1013, 33 C.C.P.A., Patents, 808. Furthermore, when references are combined to negate patentability, it should also be considered whether one skilled in the art with the references before him could have made the combination of elements claimed without the exercise of invention. In re Goepfrich, 136 F.2d 918, 30 C.C.P.A., Patents, 1181. The foregoing cases, in our opinion, stand for the proposition that it is not enough for a valid rejection to view the prior art in retrospect once an applicant's disclosure is known. The art applied should be viewed by itself to see if it fairly disclosed doing what an applicant has done. If the art did not do so, the references may have been improperly combined.

At this point we will analyze claim 29 in the light of both the prior art and the foregoing law to determine whether the references were properly combined. It is to be noted that claim 29 recites, as an element of the titrating apparatus, an electronic amplifier including a reactive circuit for "more strongly attenuating high frequency components of the amplifier input than low frequency components." The purpose of this structure, in the words of the examiner, is to "* * * minimize amplification of rapid fluctuations of signal voltage, which would not be representative of end point values. * * *" The Eckfeldt patent teaches the use of an electro-mechanical amplifier for use with titrating apparatus of the general nature disclosed by appellant. There is no doubt that the Eberhardt et al. electronic amplifier possesses the structure recited in the claim, and is capable of performing the function of appellant's amplifier. However, there is no teaching in the Eckfeldt patent that a problem exists in that rapid fluctuations of the signal voltage are not representative of end point values. While the Eberhardt et al. amplifier is inherently capable of minimizing the effect of undesirable fluctuating signals, there is no teaching in the patent that the amplifier can be used with titrating apparatus for the purpose of minimizing the effect of undesired fluctuations of signal voltage which occur in a titrating operation of the nature involved here. In short, there is no teaching in either of the patents which would suggest that they could be combined for the purpose of producing a more accurate titration. In fact, a person having the references before him who was not cognizant of appellant's disclosure would not be informed that the problem solved by appellant ever existed. Therefore, can it be said that these references which never recognized appellant's problem would have suggested its solution? We think not, and therefore feel that the references were improperly combined since there is no suggestion in either of the references that they can be combined to produce appellant's result.

We deem it necessary at this point to review the criterion of patentability which is applicable to this case where a known combination, such as shown in Eckfeldt, is modified by the use of a known element, such as shown in Eberhardt et al. In cases involving analogous factual situations, it has been stated that a new combination of old elements may be patentable; but not all new combinations are patentable; and that if a new combination of old elements is to be patentable, the elements must cooperate in such a manner as to produce a new, unobvious, and unexpected result. In re Kaufmann, 193 F.2d 331, 39 C.C. P.A., Patents, 769. In re Lindberg, supra. Furthermore, as a general matter, in determining patentability, the conception of a new and useful improvement must be considered along with the actual means of achieving the improvement. In re DeLancey, 159 F.2d 737, 34 C.C.P.A., Patents, 849; In re Bisley, 197 F.2d 355, 39 C.C.P.A., Patents, 982. And it should be remembered that in many cases the discovery of a problem is often an essential element in an invention correcting such a problem; and though the problem, once realized, may be solved by the use of old and known elements, this does not necessarily negative invention. In re Hamilton, 64 F.2d 141, 20 C.C.P.A., Patents, 987; In re Bisley, supra.

In the present case, appellant has recognized a problem which is inherent in the titration of fluids with the general type of apparatus involved here. This problem, as stated above is that rapid fluctuations of signal voltage are not representative of true end point values. To overcome this problem, which was not recognized in the prior art, appellant has combined elements, all old in the art. However, since this combination of elements evidences the discovery and solution of a heretofore unknown problem to produce a new, unobvious, and unexpected result, namely improved titration, we feel that appellant has met the tests for patentability which are set forth by the above-cited cases.

Claim 29 recites all of the elements necessary to achieve appellant's result, and we are of the opinion that it is patentable. Claims 30 to 33 contain all the limitations of claim 29. We deem it unnecessary to discuss these claims in detail. It suffices to say that in our opinion these claims are also allowable.

Since the foregoing law as applied to the facts is determinative of the question of patentability, we deem it unnecessary to consider the other arguments which have been presented.

For the foregoing reasons, the decision appealed from should be reversed.

Reversed.