UNPUBLISHED

UNITED STATES COURT OF APPEALS

FOR THE FOURTH CIRCUIT

MARSH-MCBIRNEY, INCORPORATED,
Plaintiff-Appellant,

v.                                                             No. 95-1381

ISCO, INC.,
Defendant-Appellee.

MARSH-MCBIRNEY, INCORPORATED,
Plaintiff-Appellee,

v.                                                             No. 95-1726

ISCO, INC.,
Defendant-Appellant.

Appeals from the United States District Court
for the District of Maryland, at Baltimore.
J. Frederick Motz, Chief District Judge.
(CA-94-1392-JFM)

Argued: June 5, 1996

Decided: July 30, 1996

Before WILKINSON, Chief Judge, MICHAEL, Circuit Judge, and
MACKENZIE, Senior United States District Judge for the
Eastern District of Virginia, sitting by designation.

_________________________________________________________________

Affirmed by unpublished per curiam opinion.

_________________________________________________________________
COUNSEL

ARGUED: John Alan Galbraith, WILLIAMS & CONNOLLY,
Washington, D.C., for Appellant. James Arthur Johnson, SEMMES,
BOWEN & SEMMES, Baltimore, Maryland, for Appellee. ON
BRIEF: J. Roger Williams, Jr., WILLIAMS & CONNOLLY, Wash-
ington, D.C., for Appellant. Richard A. Froehlinger, III, SEMMES,
BOWEN & SEMMES, Baltimore, Maryland; Charles E. Iliff, Jr.,
ILIFF & MEREDITH, Baltimore, Maryland, for Appellee.

_________________________________________________________________

Unpublished opinions are not binding precedent in this circuit. See
Local Rule 36(c).

_________________________________________________________________

OPINION

PER CURIAM:

This case involves two companies' competing flow meters, devices
that measure the rate of liquid flow in an open channel. Appellant
Marsh-McBirney, Inc. ("MMI") brought a claim under § 43(a) of the
Lanham Act, 15 U.S.C. § 1125(a), alleging that appellee Isco, Inc.
falsely advertises the accuracy of its flow meter in product specifica-
tions. The district court, following a ten day bench trial, ruled that
Isco had not mischaracterized its meter's capabilities so as to violate
the Lanham Act. We affirm.

I.

Flow meters measure the rate at which liquid flows through an
open channel, such as a sewer or storm drainage pipe. For some time,
appellant MMI's flow meter, the Flo-Tote, was the industry leader.
Recently, however, appellee Isco developed a competing device, the
4150, that now captures a significant share of the market. MMI claims
that Isco's sales result from advertisements that mischaracterize its
meter's accuracy.

                    2
Evaluating MMI's allegations against Isco requires some under-
standing of the technological context. Flow meters calculate the flow
rate at a specific point in an open channel by multiplying two vari-
ables, the cross-sectional area of the flow stream at the particular
point and the average velocity at which the stream is passing through
that point. Evidently, cross-sectional area can be calculated quite reli-
ably. Average velocity is much more difficult to estimate, largely
because velocities vary greatly across any individual cross-section of
the flow stream (the velocity near the pipe wall is often far less than
the velocity just below the surface).

MMI's and Isco's competing flow meters use different technolo-
gies for estimating average velocity. MMI's Flo-Tote employs an
electromagnetic sensor which measures the velocity of the flow
immediately surrounding it. From this measurement of"local" veloc-
ity, the Flo-Tote extrapolates an estimate of average velocity across
the entire channel through a series of stored internal adjustments and
site-specific calibrations. Isco's 4150 flow meter, meanwhile, uses
ultrasonic sensors. Unlike the Flo-Tote, which extrapolates average
velocity from local velocity, the 4150 directly measures the average
velocity of a subsection of the flow stream and uses this as an approx-
imation of the average velocity of the full stream.

The central question in this case is whether Isco fairly represents
the accuracy with which its flow meter senses velocity. When devel-
oping the 4150, Isco used both a "tow tank" test and a "flow bench"
test to check the accuracy of its velocity measure. In a tow tank test,
the meter is pulled through still water, and the velocity at which the
meter is being moved is compared to the meter's velocity reading. A
flow bench test more closely approximates genuine open channel con-
ditions, because water is moved through a pipe at varying velocities.
Flow bench testing, though, involves some inherent degree of error,
the extent of which is disputed by the parties.

In their published specifications for velocity, both MMI and Isco
use results obtained from tow tank testing -- both meters sense veloc-
ity to +/- 2% in a tow tank test, and both cite this figure. According
to MMI, however, Isco's use of this figure is misleading. Because a
tow tank test entails moving a meter through still water, that test,
MMI claims, primarily gauges a meter's ability to sense local veloc-

                     3
ity. In MMI's view, this measure is relevant to its Flo-Tote, because
the Flo-Tote uses local velocity as a basis for estimating average
velocity. But since the 4150 purports to measure average velocity
directly, MMI maintains, that meter's accuracy in a tow tank has little
relationship to its performance in an open channel. MMI concludes
that Isco should thus base the 4150's velocity specification on flow
bench testing.

MMI points to several flow bench tests performed on the 4150 in
charging that it is accurate only to +/- 20% in measuring average
velocity (and flow rate), not +/- 2% as its specifications allegedly sug-
gest. The Flo-Tote performs much better in flow bench testing, MMI
asserts, to within +/- 5% in measuring flow rate. Isco disputes these
allegations, arguing that its meter is much more reliable than +/- 20%
when one factors out the degree of error inherent in flow bench test-
ing. Moreover, Isco observes, MMI likewise does not use flow bench
results in its velocity specifications, nor does it publish a specification
for average velocity (or flow rate) -- instead, both Isco and MMI
publish a figure simply for "velocity" and both use the results of tow
tank testing. Finally, Isco notes, it has added the term "uniform veloc-
ity profile" to its velocity specification to convey that the +/- 2% fig-
ure was obtained in an environment without varying velocities.

After first attempting to challenge the 4150's published specifica-
tions in trade journal advertisements, MMI filed suit against Isco on
May 24, 1994. MMI alleged that Isco's velocity specification violated
§ 43(a) of the Lanham Act, 15 U.S.C. § 1125(a), by misleadingly sug-
gesting that the 4150 could sense average velocity (and flow rate) to
within 2%, when that figure was from tow tank (not flow bench) test-
ing and the 4150 could allegedly sense average velocity only to
roughly +/- 20%. The district court held a bench trial beginning on
January 23, 1995. Trial continued for ten days, after which the court
ruled in favor of Isco. MMI now appeals.

II.

MMI contends that the district court erred in rejecting its claim
against Isco. According to MMI, Isco's specifications for the 4150
plainly violate the Lanham Act, and we should thus overturn the dis-
trict court's judgment. We disagree. The district court's ruling was

                     4
based on a host of factual findings developed in the course of a
lengthy trial and with the aid of voluminous deposition testimony and
hundreds of exhibits. In a highly technical, fact-intensive case such as
this one, a district court's conclusions are entitled to special regard.

"In order to constitute a violation of § 43(a) of the Lanham Act, the
contested statement or representation must be either false on its face
or, although literally true, likely to mislead and to confuse consumers
given the merchandising context." Mylan Laboratories, Inc. v.
Matkari, 

, 1138 (4th Cir. 1993), cert. denied, 

(1994). The +/- 2% velocity specification is not "false on its
face"; MMI does not contest that the 4150 senses velocity to +/- 2%
in tow tank testing. Nor is the specification rendered false by Isco's
failure to specify whether it represents average velocity or local
velocity, or which sort of test it comes from. MMI also uses a result
obtained from tow tank testing in relating the Flo-Tote's accuracy,
and, like Isco, it simply characterizes the figure as a "velocity" speci-
fication. We thus will not disturb the conclusion of the district judge
who, in assessing whether Isco was "justified in publishing the veloc-
ity spec of plus or minus two percent from the tow tank test," deter-
mined "from the evidence that the answer is absolutely yes."

Given our conclusion that the +/- 2% specification is not false on
its face, we turn to whether it is "likely to mislead and to confuse con-
sumers." 

. Again, while a closer question, we
agree with the district court that MMI failed to carry its burden on this
score. Particularly significant in this regard is the district court's fac-
tual finding that "the people to whom [the specification] is directed
are not mere casual readers. They are people with engineering and
technical backgrounds. They have the sophistication to analyze the
information provided to them from a professional perspective." Isco's
inclusion of the term "uniform velocity profile" further diminishes the
possibility that such a sophisticated audience would misinterpret the
4150's velocity specification as reflecting average velocity readings
in flow bench testing.

According to MMI, Isco's specifications mislead consumers by
causing them to believe that they are purchasing a meter that mea-
sures average velocity (and flow rate) to +/- 2% when in fact the 4150
is allegedly accurate only to +/- 20%. If consumers were not misled

                     5
in this manner, MMI asserts, they would purchase the Flo-Tote, since
it is accurate to +/- 5%. The district court found, however, that Isco
"has not sustained its burden" with respect to demonstrating "the
extreme of a two percent or four percent or five percent meter versus
a 20 percent meter." Instead, the district judge was "inclined to
believe that . . . both of these meters may be plus or minus ten percent
instruments with a four percent range." The various tests relied upon
by MMI to show a gross disparity between the two meters amount to
"an inconclusive morass." For these reasons, we agree with the dis-
trict court that the 4150's specifications did not mislead consumers
into purchasing a far less accurate meter than they envisioned.*

III.

Because the +/- 2% velocity specification is not false on its face
and does not impermissibly mislead and confuse consumers, it does
not amount to a Lanham Act violation. The judgment of the district
court is

AFFIRMED.
_________________________________________________________________
*We also reject Isco's cross-appeal, concerning the flow bench tests
performed on the 4150 at Alden Research Laboratory. We agree with the
district court that Isco has failed to demonstrate that the Alden tests were
not independent.

                    6