tion of seismic energy. A part of the downwardly traveling energy is reflected upwardly at each subsurface interface. A reflecting interface is a region where there is a change in the velocity [of the seismic energy] as between adjoining layers of the earth, such as a layer of rock (high velocity) and a layer of sand (low velocity). In addition to change in velocity of the seismic energy in the earth due to the velocity characteristics of layers, the velocity through the earth increases with depth.

At the earth's surface, the upwardly reflected energy is detected by a plurality of seismic detectors or geophones. These extend linearly along a line of exploration. After each explosion of dynamite [along the line of exploration], each detector over a period of several seconds generates a plurality of electrical signals representative, inter alia, of reflected energy, multiples, and noise due to random earth movements unrelated to the effect of the reflected seismic energy. [Bracketed insertions ours.]

A seismogram is produced by recording, on a magnetic tape for example, the electrical signals generated by each detector. Ordinarily, a "family" of seismograms is produced for each dynamite blast-there being one seismogram for each detector. A plurality of dynamite blasts along a line of exploration will therefore yield a plurality of families of seismograms.

Appellant refers to two ways in which the detectors may be arranged with respect to the shotpoints along the line of exploration, one being referred to as a "split-spread" and the other as an "expanded-spread." In a split-spread, the shotpoint is located in the center of a spread of detectors. In an expanded-spread the shotpoint is located on the line of exploration but at some distance from the spread of detectors. It is unnecessary for an understanding of this opinion to be aware of further details of these arrangements. It will suffice to note that appellant uses both arrangements simultaneously to produce two families of seismograms for each dynamite blast.

To render meaningful the seismogram produced as described above, it is necessary to apply to it a so-called "weathered-layer correction" and a so-called "normal move-out correction." With respect to the former, appellant explains that at the earth's surface there is an unconsolidated, weathered layer (commonly called "soil") of variable depth and inclination. The velocity of seismic energy passing through this layer is much lower than in the consolidated layer just beneath it. Since the weathered layer is of variable thickness and of low velocity, it is necessary to subtract the travel time of the seismic energy in the weathered layer from the total travel time.

Because of the high velocity contrast which exists between the base of the weathered layer and the adjoining consolidated layer, some

of the seismic energy produced at the shotpoint will travel downward to the interface of the weathered and consolidated layers and be reflected upward to the detectors. The time-occurrence of the first reflection on the seismogram (time-zero being the instant the dynamite is detonated) provides the time-correction needed to eliminate the effect of the weathered layer on the time or depth measurements of interest.

A normal move-out correction is necessary to compensate for the geometrical spreading of the detectors. Since the measurements of interest are depths below the earth's surface, the identification of reflections in terms of vertical travel time is desired. Obviously, the travel path, and therefore travel time, from a shotpoint to a given reflecting interface or "horizon" and then to a given detector is greater for a detector located some distance from the shotpoint than for a detector directly adjacent the shotpoint. In correcting a family of seismograms for normal move-out, however, it must also be taken into account that the effect of geometrical spreading of the detectors decreases with increases in depth from which a given seismic wave is reflected. Therefore, normal move-out corrections must be "dynamic"; that is, the magnitude of each correction for each detector must be varied inversely with the depth from which a wave is reflected the greater the depth the less the correction. Stated differently, the longer the time-occurrence of a given wave is from timezero on a seismogram, the less it is corrected for normal move-out. Correction of a family of seismograms for the weathered-layer and normal move-out yields, in effect, a new family of seismograms on which the positions of the representations of seismic waves relative to one another more nearly correspond to the relative depths of the horizons from which those waves were reflected. Perfect corrections would cause all the reflection signals corresponding to a given horizon to be lined up across the set of seismograms. However, since the corrections are ordinarily somewhat imperfect, further adjustments are made by reproducing the seismograms as traces on an oscilloscope and manipulating knobs on the oscilloscope to bring the reflections into horizontal alignment.

Refinement of this "new" seismogram is accomplished by identification and elimination of "multiples." Multiples represent unwanted. signals which must be eliminated to avoid errors in measurements of the time-occurrence of reflections. These unwanted signals occur by

reason of multiple reflection of seismic waves, for example, as shown in Fig. 5A:

Reflections R1, R2, R3 arise because of seismic waves reflected to the earth's surface from Horizons RH1, RH2, and RH3, respectively. Multiple M1 arises because a wave is reflected from the earth's surface to horizon RH, and thence again to the surface. Its travel time is twice that for reflection R1. M11 and M12 illustrate other types of multiple reflections. There are still others which may obscure the time-appearance of the reflections which are the features of principal interest.

Appellant describes still other techniques used to refine seismograms, such as removal of noise signals due to random earth movements, but these are not critical to appellant's invention.

Appellant's Discovery

Appellant has discovered that a family of seismograms obtained by using an expanded-spread of detectors can be most precisely corrected for the effect of the weathered layer by deriving the necessary time-correction from the time-occurrence of the first reflection on a corresponding family of seismograms obtained using a splitspread of detectors.

Appellant has also discovered that the reflection-wave front of energy detected by an expanded-spread of detectors is hyperbolic in character. Based on this discovery, appellant has developed a new technique for identifying the multiples which involves applying functions of hyperbolic character to a family of seismograms. In this way, the magnitudes of errors in the normal move-out corrections can be determined and multiples can be separated from reflections, making it possible to remove the multiples from the seismograms.

Appellant's application emphasizes that to use his techniques, the seismograms must be "phonographically reproducible, whether on

magnetic, photographic or other reproducible medium." Apparently, this is necessary because the refinement of seismograms as described above involves repeated recording and playing back of the signals representative of seismic waves.

The Appealed Method Claims

We consider claims 2 and 60 to be representative. For ease of reference and understanding we reproduce these claims in numbered paragraph format, contributed in part by us:

2. In seismic exploration, the method of establishing weathering corrections in the form of individual static time-corrections for the signals from each of a plurality of seismic detecting stations spaced one from the other along a traverse which comprises

[1] generating at generating stations seismic signals adjacent selected ones of said detecting stations whereby the magnitude of said static corrections at said selected stations are known,

[2] applying said known static corrections respectively to signals generated at said selected stations,

[3] applying relative to said known corrections interpolated static corrections to the remaining signals generated at the remaining of said detecting stations, and thereafter

[4] generating at generating stations further seismic signals at spaced locations along said line,

[5] detecting at the location of a first group of said stations and thereafter at other locations of other groups of said stations seismic signals, said locations being selected in reference to the locations of said second-named generating stations for the production of an expanding-spread seismicsection having applied to the signals from each of said detecting stations said static corrections, and

[6] applying dynamic normal moveout corrections to the signals of each group of said detectors to correct them for geometrical spreading.

60. In seismic exploration where a family of seismograms are produced, each seismogram including multiple reflection signals and a plurality of single reflection signals representative of waves reflected from subsurface reflecting points after travel to said points over a plurality of paths, each of which for any one of said seismograms differs from the path for any other of said seismograms, the method which comprises:

[1] generating signals from each of said seismograms,

[2] applying to said generated signals a succession of dynamic time-adjustments, one for each said seismogram, and of magnitude to correct for normal movement delays present in said seismograms.

[3] time-shifting said generated signals, the magnitude of the time-shifts varying across said family of seismograms in accordance with a plurality of approximately hyperbolic functions of different eccentricities, and

[4] adding together said generated signals for the production of summation signals representing (a) multiple reflections which add together cumulatively for certain of said hyperbolic functions, and (b) single reflections which add together cumulatively for other of said hyperbolic functions.

The Rejection

The sole rejection is based on 35 USC 101. We will refer only to the board's opinion since all points raised in the examiner's Answer are discussed therein and will refer only to the board's general remarks applicable to all the claims and specific remarks directed to claims 2 and 60. The board stated:

The Examiner rejects each of the claims on appeal on the doctrine in In re Abrams, 38 CCPA 945 *** 188 F. 2d 165, 89 USPQ 266. This, of course is a rejection based on 35 U.S.C. 101 and is a finding that the subject matter sought to be patented is not embraced by the patent statutes The Examiner acknowledges that certain of the claims [including claim 2] on appeal * * * set forth physical steps that are clearly old in the Salvatori et al. and Jolly patents [2] * * * but asserts that patentability of the method is not dependent on these physical steps but on the other non-physical or "mental" steps set forth in these claims. * * * the Examiner asserts that [the other claims including claim 60] * include no physical steps but set forth merely a method of processing data which does not require any tangible device or apparatus to carry out the method and hence could be carried out mentally.

Appellant *** strongly urged that the Abrams case is not applicable law where there is a disclosure in the specification, as here, that the process can be carried out with apparatus there specified even though the method also could be carried out within the human mind without the apparatus. Appellant further contends that the Abrams Rule 1 and Rule 2 * * * are not applicable to any claim on appeal.

we are not impressed by either logic or the authorities cited by appellant that a claim which embraces within its scope and is patentable only because it embraces non-statutory subject matter should be allowed on the basis of a disclosure not referred to in the claim, of a possible physical alternative to the non-physical or "mental" steps embraced by the claim. This would seem to be no more logical than it would be to allow a broad apparatus claim that read on the prior art devices solely on the basis of a particular new apparatus disclosed although not claimed specifically. In each instance it would be a case of over-claiming by an applicant to embrace by the claims that which cannot be patented under the statutes. 35 U.S.C. 100(b) provides a sanction for the claiming, as a method, the use of a known machine, and obviously would be extended to include a new use of a new machine, but the use of the machine there contemplated must be claimed and not merely disclosed in the specification.

Nor do we find any logic or authority for departing from the Rule 2 of Abrams so that claims which include both statutory physical steps and nonstatutory non-physical or "mental" steps can be patentable on the sole basis of the non-statutory subject matter included therein. Were this Rule not the case, then methods of telling fortunes or predicting the activities of the stock market would be patentable providing one included the use of playing cards or a desk calculator in a claim that otherwise is for a non-statutory algorithm, such as the hypothesized principles underlying human behavior or the fluctuating values of the stock market.

2 United States patents to Salvatori et al., 2,087,120, July 13, 1937, and Jolly, 3 105,568, Oct. 1, 1963.