Intrinsic Evidence
Subscribe to Intrinsic Evidence's Posts

Resorting to Extrinsic Evidence Is Necessary When Intrinsic Evidence Is Muddied

The US Court of Appeals for the Federal Circuit determined that the limitation “a pH of 13 or higher” could not be construed using the asserted patents’ intrinsic evidence and therefore remanded to the district court with instructions to consider the extrinsic evidence and its impact on claim construction. Actelion Pharms. Ltd v. Mylan Pharms. Inc., Case No. 22-1889 (Fed. Cir. Nov. 6, 2023) (Reyna, Stoll, Stark, JJ.)

Mylan Pharmaceuticals sought market entry for its generic epoprostenol—a small molecule hypertension drug—via an abbreviated new drug application (ANDA). Mylan certified under Paragraph IV with respect to two Orange Book-listed patents that Actelion Pharmaceuticals owned. Actelion timely asserted both patents and achieved a favorable claim construction for the term “a pH of 13 or higher.” This limitation is central to the pharmaceutical breakthrough that the asserted patents disclose, which is stably solubilizing the otherwise unstable epoprostenol. The asserted patents teach that epoprostenol bulk solutions’ pH should preferably be adjusted to about 12.5 to 13.5. Mylan contended that this language meant the claim should not encompass anything below pH 13 (i.e., leaving only a fraction of the disclosure’s preferred range within the scope of the issued claims). The district court disagreed, adopting Actelion’s position that the limitation encompassed values that “rounded” to pH 13 (i.e., pH 12.5 and above). This prompted a stipulated infringement judgment. Mylan appealed.

The Federal Circuit reviewed the district court’s construction de novo, explaining that the district court’s construction was a ruling as a matter of law because no extrinsic evidence was considered. The Court then examined the claim language, specifications and prosecution histories in turn.

Regarding the claim language, the Federal Circuit rejected the parties’ invitations to acknowledge certain prior decisions as articulating bright-line rules. Instead, the Court appeared to endorse analyzing the intrinsic evidence on a case-by-case basis when construing similar range limitations. For example, the Court explained that “there is no blanket rule that ranges, or specifically open-ended ranges, must foreclose rounding.” The Court also rejected the following rules concerning range precision:

  • Avoiding rounding requires terms of precision, such as “precisely” or “exactly.”
  • The absence of approximation language dictates a precise value.

In contrast, the Federal Circuit signaled that district courts should properly account for a range’s technical implications (or at least ranges concerning the pH scale) even if that means looking to extrinsic evidence.

Regarding the specifications and prosecution histories, the Federal Circuit disagreed with the district court’s finding that both were inconsistent with Mylan’s position that the disputed limitation reflected a higher degree of precision. The Court concluded that there was little to glean from either, finding that the specifications were as clear as “muddied water” and the prosecution histories provided no insight into the relevant pH range—between pH 12 and pH 13.

The Federal Circuit ultimately concluded that this was an instance in which properly construing a claim limitation required the aid of extrinsic evidence, invoking the Supreme Court of the United States’ guidance from Teva v. Sandoz that extrinsic evidence [...]

Continue Reading




read more

Dictionaries Don’t Know Best: The Intrinsic Record Prevails (Again)

The US Court of Appeals for the Federal Circuit addressed the tension between the intrinsic and extrinsic record in claim construction, holding that the intrinsic record should be relied on first. The Court therefore reversed a district court finding of indefiniteness based on dictionary definitions and expert testimony. Grace Instrument Industries, LLC v. Chandler Instruments Company, LLC, Case Nos. 21-2370; -2370 (Fed. Cir. Jan. 12, 2023) (Chen, Cunningham, Stark, JJ.)

Grace owns a patent for a liquid pressurized viscometer used commonly to test the viscosity of drilling fluid that is used to drill oil wells. Grace’s viscometer uses an “enlarged chamber” between a lower chamber and a pressurization fluid inlet. According to the patent specification, this enlarged chamber was designed to reduce the measurement error in prior viscometer models that was caused by the mixing of sample fluid and pressurization fluid or by friction emanating from a seal in the viscometer. Within the lower chamber of the patented viscometer, there is a rotor having a “means for driving said rotor to rotate.”

Grace sued Chandler, claiming that Chandler infringed Grace’s viscometer patent. During claim construction, the district court found that the term “enlarged chamber” was indefinite, and that because it was a “term of degree,” it must be compared against something objective. The district court entered its final judgment in favor of Chandler. Grace appealed.

The Federal Circuit reversed, explaining that the “enlarged chamber[’s]” size did not need to be compared against any baseline object, but rather needed to be “large enough to accomplish a particular function.” The Court relied on the specification, which explained that the viscometer described in the patent reduced the mixing of sample fluid and pressurization fluid that was common in older viscometer models by using an “enlarged chamber.” The Court also cited the prosecution history, where the applicant explained that the invention solved a “long lasting problem” by reducing the measurement error caused by the friction of the seal or mixing of the fluids in older viscometer designs. Thus, the Court reasoned that a person of ordinary skill in the art would understand that the “enlarged chamber” must be large enough to prevent the mixing of the pressurization fluid and sample fluid to avoid the measurement errors associated with prior art viscometers.

Chandler argued that the term “enlarged chamber” was not a term of art. The Federal Circuit agreed but explained that “the intrinsic record sufficiently guides a skilled artisan as to the meaning of the term” as used in the patent. The Federal Circuit found that the district court erred in relying on a dictionary definition (extrinsic evidence) to contradict the meaning of the term found within the intrinsic record. The specification’s instructions for the meaning of a claim term should prevail over extrinsic evidence. Rebuffing Chandler’s argument, the Court explained that the specification does not need to explicitly define the claim term to govern the interpretation of that term.

The Federal Circuit remanded the case to the district court to reconsider indefiniteness in [...]

Continue Reading




read more

Bad Connection: Claim Construction Argument without Explanation Given No Weight

The US Court of Appeals for the Federal Circuit affirmed a Patent Trial & Appeal Board (Board) obviousness decision after finding that the patent owner failed to explain how its cited extrinsic evidence supported its proposed claim construction. Uniloc 2017 LLC v. Netflix, Inc., Case No. 21-2085 (Fed. Cir. Dec. 15, 2022) (Dyk, Taranto, Hughes, JJ.) (nonprecedential).

Uniloc owns a patent directed to a “more efficient” method for encoding videos by only coding at the pixel level when necessary and, where possible, reusing code for macroblocks in the background area. Netflix petitioned for inter partes review of several claims in Uniloc’s patent. The Board ultimately found the claims unpatentable as obvious. Uniloc appealed.

Uniloc argued that the Board erred in its claim construction of a limitation that required “dividing the stationary background region and the object region from an inputted video in a macroblock-by-macroblock basis by using a difference between the previous frame and the current frame.” The dispute before the Federal Circuit was whether the “macroblock-by-macroblock basis” required that the act of dividing be done macroblock-by-macroblock, as urged by Uniloc, or whether the dividing resulted in separate macroblock-based regions, as urged by Netflix and found by the Board. The Court concluded that Netflix and the Board were correct.

The Federal Circuit first addressed the intrinsic record, finding that it was ambiguous as to the meaning of “macroblock-by-macroblock basis.” Although the claim language supported either interpretation, Uniloc’s argument would require the Court to read language into the claim that was “simply not there.” The specification did not clarify the claim construction issue either. Although Uniloc pointed to one example where the dividing step occurred one macroblock at a time, the Court stated that it did not limit claim language to examples used in the specification. Lastly, the Court explained that the portions of the specification describing the purpose of the invention and avoiding problems in the prior art also failed to clarify the meaning of the macroblock-by-macroblock basis. Because the purpose was merely to make coding more efficient and avoid dividing at the pixel level, either interpretation could apply.

Relatedly, the Federal Circuit rejected Netflix’s argument that Uniloc forfeited the arguments on appeal related to the portions of the specification explaining the purpose of the claimed invention and how it purportedly solved problems in the prior art. The Court explained that the forfeiture doctrine does not “preclude a party from proffering additional or new supporting arguments, based on evidence record, for its claim construction.” Here, Uniloc “merely cited additional support in the specification to support the same argument it had always made” related to the macroblock-by-macroblock basis.

The Federal Circuit next addressed the extrinsic evidence and agreed with the Board’s decision that a “macroblock-by-macroblock basis” required dividing results in separate macroblock-based regions. Uniloc’s only extrinsic evidence was a dictionary definition of “basis,” and Uniloc offered no additional expert evidence to explain technological facts or usage in the field that would support its interpretation. In contrast, Netflix provided expert testimony and [...]

Continue Reading




read more

Terms of Degree Not Always Indefinite

The US Court of Appeals for the Federal Circuit overturned a district court determination that the claim terms “resilient” and “pliable” were indefinite. The Federal Circuit found that the claims, while broad, were sufficiently definite in view of both intrinsic and extrinsic evidence. The Federal Circuit also upheld the district court’s findings of no induced infringement, finding zero evidence of predicate direct infringement of the properly construed method claims. Niazi Licensing Corp. v. St. Jude Medical S.C., Inc., Case No. 21-1864 (Fed. Cir. Apr. 11, 2022) (Taranto, Bryson, Stoll, JJ.) The Federal Circuit also affirmed entry of sanctions excluding portions of the plaintiff’s technical and damages expert reports for failing to disclose predicate facts during discovery and also affirmed exclusion of portions of plaintiff’s damages expert report as unreliable for being conclusory and legally insufficient.

In reaching its decision on indefiniteness, the Federal Circuit focused on the terms “resilient” and “pliable” as used in a claim directed to a double catheter structure. Citing the 2014 Supreme Court decision in Nautilus v. Biosig Instruments, the Federal Circuit explained that language has “inherent limitations,” and stated that a “delicate balance” must be struck to provide “clear notice of what is claimed” and avoid the “zone of uncertainty” relating to infringement. The Court noted that under Nautilus, claims must provide “objective boundaries,” but the Court distinguished the present case from those in which “subjective boundaries” created uncertainty and rendered the claim indefinite. The Court pointed to its 2005 decision in Datamize v. Plumtree Software as a “classic example” of subjectivity where the term “aesthetically pleasing” was deemed indefinite because the patent provided no way to provide “some standard for measuring the scope of the phrase.” The Court also noted that a patent’s claims, written description and prosecution history—along with any relevant extrinsic evidence—can provide or help identify the necessary objective boundaries for claim scope

The Federal Circuit concluded that there was sufficient support in the intrinsic evidence, both in the claims themselves and the written description, to allow a skilled artisan to determine the scope of the claims with reasonable certainty. The Court explained that the claim at issue recited “an outer, resilient catheter having shape memory” that “itself provides guidance on what this term means—the outer catheter must have ‘shape memory,’ and ‘sufficient stiffness.’” The Court also cited to “[n]umerous dependent claims [that] further inform the meaning of this term by providing exemplary resilient materials of which the outer catheter could be made. . . . The written description provides similar guidance . . . . Thus, a person of ordinary skill reading the claims and written description would know of exemplary materials that can be used to make a resilient outer catheter, i.e., one that has shape memory and stiffness such that it can return to its original shape.”

The Federal Circuit distinguished this case from Datamize, where the claim scope depended on the eye of each observer, finding it more akin to its 2017 decision in Sonix Technologies. In that [...]

Continue Reading




read more

Game Reset: Extrinsic Evidence Can’t Limit Claim Scope Beyond Scope Based on Unambiguous Intrinsic Evidence

The US Court of Appeals for the Federal Circuit reversed a district court’s grant of summary judgment of noninfringement after concluding that the district court erred by relying on expert testimony to construe a claim term in a manner not contemplated by the intrinsic evidence. Genuine Enabling Tech. LLC v. Nintendo Co., Ltd., et al., Case No. 20-2167 (Fed. Cir. Apr. 1, 2022) (Newman, Reyna, Stoll, JJ.)

Genuine owns a patent directed to a user interface device (UID) that, in the process of synchronizing and merging data streams into a combined data stream, directly receives microphone speech input and transmits speech output via a speaker. During prosecution, the inventor distinguished “slow varying” physiological response signals discussed in a prior art reference from the “signals containing audio or higher frequencies” in his invention, arguing that the latter posed a signal “collision” problem that his invention solved. In distinguishing the prior art, the inventor explained that his invention “describes, in its representative embodiments, how to combine the data from a UID (mouse) and from a high-frequency signal, via a framer, which is unique and novel.”

Genuine subsequently filed a lawsuit against Nintendo, accusing Nintendo’s Wii Remote, Wii Remote Plus, Nunchuk, WiiU GamePad, Switch Joy-Con Controller and Switch Pro Controller of infringing the patent. During claim construction, the parties asked the district court to construe the term “input signal.” Genuine proposed the construction of the disputed claim term to be “a signal having an audio or higher frequency,” whereas Nintendo proposed the narrower construction of “[a] signal containing audio or higher frequencies.” Relying on expert testimony, Nintendo also argued that the inventor “disclaimed signals that are 500 [Hz] or less . . . generated from positional change information, user selection information, physiological response information, and other slow-varying information.”

The district court found that the inventor’s arguments amounted to a disclaimer. Crediting Nintendo’s expert testimony, the district court construed “input signal” as “signals above 500 Hz and excluding signals generated from positional change information, user selection information, physiological response information, and other slow-varying information.” The district court subsequently granted summary judgment of noninfringement, finding that the accused controllers produced the types of slow-varying signals that the inventor had disclaimed during prosecution. Genuine appealed.

Genuine argued that the district court erred in construing “input signal” by improperly relying on extrinsic evidence and improperly finding that the inventor disclaimed certain claim scope during prosecution. The Federal Circuit agreed, reiterating that although extrinsic evidence can be helpful in claim construction, “the intrinsic record ‘must be considered and where clear must be followed,’” such that “where the patent documents are unambiguous, expert testimony regarding the meaning of a claim is entitled to no weight.” In this case, although the parties agreed that the inventor had disavowed claim scope during prosecution, there was a dispute as to the scope of the disclaimer beyond signals below the audio frequency spectrum.

The Federal Circuit explained that for a statement made during prosecution to qualify as a disavowal of claim scope, it [...]

Continue Reading




read more

Improper Claim Construction Requires Partial Remand of Obviousness Determination

The US Court of Appeals for the Federal Circuit issued decisions in two separate inter partes reviews (IPRs), one involving a patent related to radio frequency communication systems and the other involving a patent related to multi-processor systems. Intel Corporation v. Qualcomm Incorporated, Case No. 20-1664 (Fed. Cir. Dec. 28, 2021) (Prost, Taranto, Hughes, JJ.); Intel Corporation v. Qualcomm Incorporated, Case Nos. 20-1828, -1867 (Fed. Cir. Dec. 28, 2021) (Prost, Taranto, Hughes, JJ). Based on issues of claim construction and obviousness, the Court affirmed in part and vacated in part the Patent Trial and Appeal Board’s (Board) decision in the radio frequency communication systems patent IPR and vacated the Board’s decision in the multi-processor systems patent IPR.

Radio Frequency Communication System Patent IPR (1664)

In the IPR related to the radio frequency communication systems patent, Intel proposed that the claim term “radio frequency input signal” should take its ordinary meaning of an input signal having a radio frequency. Qualcomm argued that a person of skill in the art reading the patent would understand the phrase to reference the radio frequency signal that is received before down-conversion, and thus proposed that the term should mean “a signal centered at a carrier frequency at which the signal was transmitted/received.” The Board agreed with Qualcomm based on the intrinsic evidence.

Intel argued before the Board that certain claims of the radio frequency communication systems patent would have been obvious in light of the Der reference and the Valla reference. Qualcomm argued that a skilled artisan would not have been motivated to combine Der and Valla, because Der’s transistor would defeat the intended purpose of Valla’s amplifier. The Board agreed with Qualcomm. Qualcomm also submitted substitute claims. The Board accepted the substitute claims after finding that a skilled artisan would have lacked reason to combine Der and the Burgener reference to achieve the substitute claims. Intel appealed.

The Federal Circuit first addressed the threshold question of whether it had jurisdiction since no lawsuit had been filed against Intel. Despite the absence of any lawsuit against Intel itself, the Court found that Intel had standing because it had engaged in acts that previously resulted in assertion of the patent against one of Intel’s customers. Because Intel continues to sell the relevant products to that customer and others, it must address the risk of an infringement suit by Qualcomm. Qualcomm also refused to offer a covenant not to sue or stipulate that it would not reassert its prior infringement allegations involving the Intel products. The Court found that this refusal made Intel’s risk more than “mere conjecture or hypothesis.” Therefore, the Court found that Intel had standing to pursue the appeal.

Turning to the merits, the Federal Circuit affirmed the Board’s construction of “radio frequency input signal.” The Court explained that while both parties’ proposed constructions had appeal when considered in a vacuum, the proper inquiry required analysis of the surrounding claim language and specification. The Court found that linguistic clues in the claims suggested that [...]

Continue Reading




read more

Rounding Error: Intrinsic Evidence Informs Plain and Ordinary Meaning

Vacating a stipulated infringement judgment based on an incorrect claim construction, the US Court of Appeals for the Federal Circuit explained that it is improper to isolate claim language from the intrinsic evidence when determining the plain and ordinary meaning of a disputed term. AstraZeneca AB v. Mylan Pharms. Inc., Case No. 21-1729 (Fed. Cir. Dec. 8, 2021) (Stoll, J.) (Taranto, J., dissenting).

AstraZeneca sued Mylan Pharmaceuticals for infringement of three patents listed in the US Food and Drug Administration’s (FDA) orange book covering the Symbicort® pressurized metered-dose inhaler for the treatment of asthma and COPD. 3M submitted an abbreviated new drug application (ANDA) to the FDA to manufacture and sell a generic version of the Symbicort® inhaler and certain interests to the ANDA were later transferred to Mylan. After receiving a Paragraph IV letter from Mylan, AstraZeneca filed an infringement suit.

Shortly before trial, the district court held a claim construction hearing to determine the meaning of “0.001%,” the claimed concentration of PVP (one of the active ingredients). The district court construed the term based on its “plain and ordinary meaning, that is, expressed with one significant digit.” Based on this definition, Mylan stipulated to infringement and the district court entered judgment. The district court held a bench trial on invalidity, ultimately determining that Mylan did not prove that the claims were invalid as obvious. Mylan appealed the stipulated judgment stemming from the claim construction determination and the judgment of no invalidity.

First, Mylan challenged the district court’s claim construction of “0.001%.” AstraZeneca argued that the district court improperly construed the term to encompass a range from 0.0005% to 0.0014%. Mylan contended that, in view of the specification and the prosecution history, the term was to be defined precisely at 0.001% with only “minor variations” allowed. The Federal Circuit agreed, finding that Mylan’s proposed construction was more properly aligned with the patent’s description as further informed by the prosecution history.

The Federal Circuit stated that the proper construction of 0.001% only allowed minor variations from 0.00095% to 0.00104%. There was no dispute that the term 0.001% would ordinarily encompass the range of 0.0005% to 0.0014%. AstraZeneca argued that this “ordinary meaning” would control absent lexicography or disclaimer. The Court disagreed, finding that it would improperly isolate the term from the claim language, specification and patent prosecution history. The Court explained that the “ordinary meaning” is not the ordinary meaning in the abstract but is instead the “meaning to the ordinary artisan after reading the entire patent,” and therefore the claims must be read in view of both the written description and the prosecution history. The Court’s rationale for narrower construction was based on the intrinsic record reflecting that the written description and prosecution history showed that very minor differences in PVP concentration would impact stability.

The Federal Circuit found that the written description explained that stability was one of the most important factors and that even very minor differences in PVP concentration could impact stability. The written description also [...]

Continue Reading




read more

Stick to the Fax: Conflicting Statements Made During Prosecution Lead to Indefiniteness

In deciding whether use of the term “passive link” to define a connection between a computer terminal and a fax machine rendered a patent claim indefinite, the US Court of Appeals for the Federal Circuit affirmed the district court’s finding of invalidity based on conflicting statements made by the patent owner during prosecution. Infinity Computer Products, Inc. v. Oki Data Americas, Inc., Case No. 20-1189 (Fed. Cir. Feb. 10, 2021) (Prost, C.J.)

Infinity owns a patent directed to providing a circuit for interfacing a personal computer with a facsimile machine to enable the facsimile to be used as a scanner or a printer for a personal computer. The patent seeks to accomplish all of the objectives of a scanner or a printer in a simple, straightforward manner through the use of a circuit of highly simplified design and low cost. The patent claims recite that this functionality is accomplished “through a bi-directional direct connection via a passive link between the facsimile machine and the computer.” Infinity asserted the patent against Oki in district court.

The term “passive link” does not appear in the patent specification. Infinity introduced this term during prosecution to overcome rejections based on a prior art patent to Perkins. During prosecution, Infinity unsuccessfully argued that unlike Perkins, the claimed invention permits uninterrupted transfer of signals between the facsimile and the computer without the use of intervening circuitry. Infinity engaged in multiple rounds of amendment and response with the examiner before finally overcoming the rejections based on Perkins by arguing that the invention “creates a passive link between the facsimile machine and the computer [and] therefore does not require any intervening apparatus as does Perkins.” Perkins used a modem, characterized by Infinity as the “intervening apparatus,” internal to the computer. Infinity argued that the modem “should be regarded as a peripheral device to the computer which processes data before it is transmitted to the I/O bus of the computer,” effectively drawing the boundary of the “passive link” at the I/O bus of the computer.

After allowance, the patent was the subject of three ex parte re-examination proceedings. The patent was a continuation-in-part of a parent application, and in order to overcome a prior art reference asserted in the re-examination proceeding, Infinity argued that the claimed “passive link” element was entitled to the priority date of an earlier parent application. Infinity specifically noted that the patent’s description of “the RJ11 telephone cable and use thereof in communicating data between the fax machine 30 and the PC computer 40 meets the definition of ‘passive link.'” In doing so, Infinity pointed to certain figures in the parent application specification that disclosed fax modem circuitry internal to the computer, effectively drawing the boundary of the “passive link” at the computer’s external port—before the I/O bus.

The district court found that there was a discrepancy on the boundary of the “passive link” because during prosecution it was defined as at the I/O bus of the computer, but during the ex parte re-examination it [...]

Continue Reading




read more

“Gradual” and “Continuous” Includes Step-Wise

The US Court of Appeals for the Federal Circuit affirmed a priority decision in favor of the senior party, upholding a claim construction that was based upon a verbatim definition set forth in the patent specification of the application from which the count in interference was copied. Chevron U.S.A. Inc. v. University of Wyoming Research Corp., Case No. 19-1530 (Fed. Cir. Nov. 4, 2020) (Schall, J.) (Newman, J., dissenting).

Wyoming Research provoked a patent interference proceeding by copying into its pending application a claim from Chevron’s pending patent application. Under the now-discontinued interference statute, the patent for an invention claimed by more than one party was awarded to the first-to-invent party. If the Patent Trial and Appeal Board (Board)determined there was an interference in fact—that is, two patent applications claimed the same subject matter—then the Board could proceed to determine priority of inventorship. A finding of interference in fact required the copying party’s patent specification to contain adequate written description and enablement to support the copied claim.

The copied claim was directed to a method of fractioning asphaltenes from crude oil. The technology used a mobile phase into which an alkane solvent was introduced and its concentration “gradually and continuously” changed over time, and the resulting eluted fractions were analyzed. The disputed claim limitation was: “gradually and continuously changing the alkane mobile phase solvent to a final mobile phase solvent.”

Chevron argued that “gradually and continuously changing” referred to the act of feeding alkane mobile phase solvent into the inlet of the column. Relying on intrinsic evidence, the Board instead adopted Wyoming’s construction, concluding that the limitation’s “gradually and continuously changing” referred to the change of solvents in the column and not to changes at the inlet to the column. The distinction was important because at the inlet, the Wyoming invention introduced solvent in a step-wise manner. The parties agreed that Wyoming’s specification supported only the construction adopted by Board, and Wyoming was declared to be the senior party for the priority contest.

Because Chevron had filed a priority statement that indicated that its earliest corroborated conception coupled with diligence date was later than Wyoming’s priority date, the Board determined that Chevron was unable to prevail on priority and entered judgment in favor of Wyoming. Chevron appealed.

On appeal, Chevron argued that the Board’s construction was inconsistent with Chevron’s patent specification. Chevron contended that its application disclosed that the solvent was “gradually and continuously” changed at the column’s inlet and that the Board’s construction rendered the limitation meaningless because it encompassed even “sudden, abrupt immediate solvent switches.”

The Court affirmed the Board’s construction, holding that the broadest reasonable construction of “gradually and continuously changing” did not require a change of solvents at the column inlet. The Court reasoned that the Board’s construction was consistent, and indeed tracked verbatim, with the Chevron application’s express definition of “gradually.” While the Court acknowledged that certain examples in the Chevron application illustrated that one way to implement a “gradual and continuous change” of the [...]

Continue Reading




read more

BLOG EDITORS

STAY CONNECTED

TOPICS

ARCHIVES