United States Court of Appeals
                         FOR THE EIGHTH CIRCUIT
           ___________

           No. 01-3181
           ___________

Kimberly R. Smith,                 *
                                   *
             Plaintiff-Appellant,  *
                                   *
      v.                           *
                                   *
BMW North America, Inc.; BMW AG; *
Roadshow Enterprises, Inc., doing  *
business as Roadshow BMW Inc.,     *
                                   *
             Defendants-Appellees. *

           ___________                   Appeals from the United States
                                         District Court for the
           No. 02-1016                   Eastern District of Arkansas.
           ___________


Kimberly R. Smith,                   *
                                     *
           Plaintiff-Appellant,      *
                                     *
      v.                             *
                                     *
BMW North America, Inc.; BMW AG; *
Roadshow Enterprises, doing business *
as Roadshow BMW Inc.,                *
                                     *
           Defendants-Appellees.     *
                                   ___________

                             Submitted: June 28, 2002
                                Filed: October 24, 2002
                                 ___________

Before HANSEN, Chief Judge, BRIGHT, and MURPHY, Circuit Judges.
                              ___________

BRIGHT, Circuit Judge.

        On August 24, 1997, appellant Kimberly R. Smith suffered severe injuries in
an automobile accident rendering her a quadriplegic while driving her 1994 BMW
automobile. She brought a product liability action against appellees BMW North
America, Inc., BMW AG, and Roadshow Enterprises, Inc., d/b/a Roadshow BMW,
Inc. (collectively BMW). Smith alleged that the air bag in her car was faulty and, had
the air bag properly deployed, it would have reduced or prevented her injuries. BMW
moved to exclude the testimony of two of Smith’s experts and for summary judgment.
The district court held a Daubert hearing and excluded the proposed testimony of
Smith’s expert witnesses under Fed. R. Evid. 702 and Daubert v. Merrell Dow
Pharms., 

(1993). Having determined that without the testimony of her
experts Smith could not establish a prima facie case against BMW, the district court
also granted BMW’s motion for summary judgment. The district court denied Smith's
subsequent motion under Fed. R. Civ. P. 60(b) for relief from the judgment. Smith
appealed to this court the order granting summary judgment and separately appealed
the order denying her Rule 60(b) motion. We consolidated the appeals. We now
vacate the district court’s grant of summary judgment and remand for proceedings
consistent with this opinion.




                                         -2-
I.    BACKGROUND

       Appellant Kimberly Smith suffered severe injuries in a car accident while
driving her 1994 BMW 318i. She lost control of the vehicle which then left the road,
struck an embankment, and rolled several times. Her injuries resulted in quadriplegia.
She brought a product liability lawsuit against BMW, alleging that the air bag in her
car was faulty and that, had her air bag deployed, it would have reduced or prevented
her injuries.

       Smith intended to offer at trial the testimony of Dr. Larry Williams, a certified
accident reconstructionist, and Dr. Stephen Erickson, a forensic pathologist. Dr.
Williams was to testify regarding his calculations of the barrier equivalent velocity1
of Smith’s vehicle when it struck the embankment. Dr. Williams’ testimony would
have established that the barrier equivalent velocity was great enough that a properly
functioning air bag would have deployed upon impact.2 Dr. Erickson was to testify
regarding his opinion that Smith’s injuries would have been prevented or reduced had
the air bag properly deployed.

      BMW moved for the exclusion of Dr. Erickson’s testimony and for summary
judgment. The district court set a Daubert hearing for June 12, 2001. One week
before the hearing, BMW moved to exclude the proposed testimony of Dr. Williams

      1
        A vehicle’s barrier equivalent velocity in a crash is the velocity with which the
car would have to impact an infinitely rigid barrier at the same angle of impact in
order to sustain the same damage as actually occurred. The barrier equivalent
velocity “combine[s] into one speed-like measure the energies, forces, and motions
that a vehicle experiences during a crash.” Crespo v. Chrysler Corp., 

, 227 n.4 (S.D.N.Y. 1999).
      2
      The parties disagree as to the threshold barrier equivalent velocity for air bag
deployment. The record contains numbers ranging from 12 to 15.5 miles-per-hour.

                                          -3-
as well. BMW informed the district court that it would be calling Mr. John Bentley,
an accident reconstruction expert, to testify at the Daubert hearing in opposition of
Dr. Williams.

      At the Daubert hearing, Dr. Williams testified regarding his calculations of the
principal direction of force3 and measurements of vehicle deformation, or crush,
which he entered into the EDCrash computer program to calculate the barrier
equivalent velocity of Smith’s vehicle when it struck the embankment.4 Dr. Williams
concluded that the barrier equivalent velocity was 22 to 26 miles-per-hour.

       Regarding the principal direction of force, Dr. Williams testified that he visited
the scene of the accident with the state trooper who viewed the accident scene in the
immediate aftermath of the crash. He also visited the auto salvage yard with that
trooper. This permitted Dr. Williams “to incorporate [the trooper’s] firsthand
observations with my work.” He stated that the trooper’s statements “about the
evidence of the debris and path of travel of the vehicle” helped him significantly. The
trooper’s statements that the vehicle climbed the embankment led Dr. Williams to
conclude that “the vehicle ha[d] to be going into the mountain.” Dr. Williams also

      3
        A vehicle’s principal direction of force during impact is the direction of the
force that causes the deformation and structural displacement in the damaged vehicle.
The principal direction of force is measured in degrees from a head-on impact along
the car’s longitudinal axis. A 0° principal direction of force corresponds to a twelve
o’clock or head-on impact. A 30° principal direction of force corresponds to an
impact at one o’clock and a –30° principal direction of force corresponds to an impact
at eleven o’clock.
      4
       The EDCrash computer program takes as input the principal direction of force,
measurements of displacement in the car’s structure (deformation or crush), and
known data regarding the car’s structure and materials from which it is manufactured.
The program then outputs the barrier equivalent velocity for a given vehicle accident.
Both Dr. Williams and Mr. Bentley used the EDCrash computer program and its
relevance and reliability is unchallenged.

                                          -4-
consulted the tow-truck driver who removed the damaged vehicle, read deposition
testimony and the police report of the accident, and viewed various photographs of
the vehicle. Based upon the entirety of this evidence, Dr. Williams concluded that
the principal direction of force was “no more than 20 degrees into the front of the
vehicle.” He used –20° as the principal direction of force input into the EDCrash
computer program.5 When asked to explain how such an impact could account for
the existence of passenger-to-driver side direction of crash damage and displacement
of the front end of the vehicle, Dr. Williams opined that the car sustained that damage
when it rotated counterclockwise at impact.

       As to his deformation measurements, Dr. Williams testified that he began with
a scale drawing of the vehicle around which he superimposed a rectangle fitted to the
dimensions of the undamaged car. He used the rectangle to represent the vehicle in
its undamaged state. From six different points along the edge of the rectangle
corresponding to the front of the vehicle, he measured the perpendicular displacement
from the rectangle to the damaged vehicle. It is from these displacement
measurements, the principal direction of force, and data pertaining to the vehicle’s
structure, that the EDCrash computer program calculates the barrier equivalent
velocity.

       BMW then offered the expert testimony of Mr. Bentley. Bentley testified that
Dr. Williams incorrectly measured the displacement of the car’s front end caused by
the accident. Mr. Bentley also testified that Dr. Williams’ estimate of the principal
direction of force could not generate the counterclockwise rotation he said accounted
for the passenger-to-driver side displacement of the front end. According to Bentley,
these mistakes in the data entered into the EDCrash computer program led to an
incorrect calculation of the barrier equivalent velocity.


      5
      The negative sign reflects that the driver’s side of the front of the car struck
the embankment first. See note 

.

                                         -5-
        Mr. Bentley stated that Dr. Williams’ use of a rectangle to represent the vehicle
in its undamaged state led to incorrect displacement measurements. Mr. Bentley
testified that using a rectangular representation of the vehicle overmeasures
displacement because such measurements from the rectangle to the deformed vehicle
include air gaps, particularly at the corners of the vehicle. He stated that Dr.
Williams, rather than employ a rectangular model, should have used an exemplar
representation of the vehicle (i.e., one that uses the actual parameters of the vehicle
in an undamaged state). Mr. Bentley asserted that use of an exemplar representation
of the vehicle is necessary under a protocol for displacement measurement as set out
by the Tumbas papers.6

       Dr. Williams testified that he had been trained to use a rectangular
representation of the vehicle in order to account for the restitution, the amount the
vehicle rebounds from the maximum engagement of the crush. Restitution occurs
nearly instantaneously in an accident; it is immeasurable from an examination of the
post-accident vehicle. According to Dr. Williams, post-accident measurements will
underestimate the actual deformation the vehicle sustained at its maximum
engagement because they do not include restitution. Dr. Williams commented that
measurement of restitution is “one of the real controversies currently in the
literature.”

      Smith also attempted to introduce the testimony of Dr. Erickson, a forensic
pathologist, to address the causation element of her claim. Dr. Erickson based his
opinions on his observations of the accident scene, discussions with the investigating
trooper and treating neurosurgeon, his review of their depositions, his examination
of Smith’s vehicle, his interview with a witness to the crash scene, his review of



      6
       Nicholas S. Tumbas & Russel A. Smith, Measuring Protocol for Quantifying
Vehicle Damage from an Energy Basis Point of View, SAE Paper No. 880072 (1988).

                                          -6-
Smith’s medical records, the opinions of Dr. Williams, and his experience as a
forensic pathologist.

       Dr. Erickson assumed that Smith was injured either when the car struck the
embankment or during the subsequent rollover.7 He determined that, to a reasonable
degree of medical certainty, Smith’s neck was broken at the initial impact of her
vehicle with the embankment, rather than in the rollover. He based this opinion on
the following factors: (1) according to her neurosurgeon, Smith’s neck injury was a
wedge fracture, more likely produced by neck flexion as the car quickly decelerated
than by a rollover in which a compression or burst fracture would be more typical; (2)
there was no medical evidence of trauma to Smith’s head; (3) Smith was wearing her
seat belt at the time of the accident.

       Dr. Erickson further opined that, to a reasonable degree of medical certainty,
the air bag would have prevented Smith’s neck flexion and prevented or reduced her
neck injuries had it deployed. Dr. Erickson reached this conclusion based upon his
understanding of how air bags operate, which he gained from personal observations
and films, from attending a national conference for forensic pathologists on air bag
operation, and from reading articles on what air bags are designed to do.

      By order dated July 5, 2001, and pursuant to Fed. R. Evid. 702 and Daubert v.
Merrell Dow Pharms., 

(1993), the district court excluded the testimony
of both Dr. Williams and Dr. Erickson, having determined that their opinions were
unreliable and would not be helpful to the jury. The district court then granted
summary judgment in favor of BMW because, without the testimony of her experts,
Smith could not succeed on her products liability claim.



      7
       The vehicle’s air bag is not intended to deploy during a rollover. Injuries
sustained during a rollover cannot be attributed to the failure of the air bag to deploy.

                                          -7-
       Smith moved the district court to vacate this final judgment pursuant to Fed.
R. Civ. P. 60(b)(3) or 60(b)(6).8 Smith argued that in granting summary judgment to
BMW, the district court relied on Mr. Bentley’s false and misleading testimony. The
district court denied the 60(b) motions in an order issued on December 5, 2001.
Smith now appeals both the summary judgment order and the order denying her 60(b)
motion. We determine that the district court’s grant of summary judgment of
dismissal to BMW be vacated and we remand for further proceedings. We dismiss
Smith’s appeal on the 60(b) issue.

II.   DISCUSSION

      Smith argues that the district court erred in excluding the testimony of her
experts, Drs. Williams and Erickson.

      Federal Rule of Evidence 702 governs admissibility of expert testimony. Rule
702 explains:

             If scientific, technical, or other specialized knowledge will assist
      the trier of fact to understand the evidence or to determine a fact in
      issue, a witness qualified as an expert by knowledge, skill, experience,
      training, or education, may testify thereto in the form of an opinion or
      otherwise, if (1) the testimony is based upon sufficient facts or data, (2)
      the testimony is the product of reliable principles and methods, and (3)
      the witness has applied the principles and methods reliably to the facts
      of the case.

Fed. R. Evid. 702 (2001).


      8
         “On motion and upon such terms as are just, the court may relieve a party
. . . from a final judgment, order, or proceeding for the following reasons: . . .(3)
fraud, . . . misrepresentation, or other misconduct of an adverse party, . . . (6) any
other reason justifying relief from the operation of the judgment.” Fed. R. Civ. P.
60(b).

                                         -8-
       We review the exclusion of expert testimony for abuse of discretion. General
Elec. Co. v. Joiner, 

, 143 (1997). We review de novo the district court’s
grant of summary judgment. Netland v. Hess & Clark, Inc., 

, 898 (8th
Cir. 2002). Summary judgment is appropriate if, after viewing the facts and all
reasonable inferences in the light most favorable to the nonmoving party, the record
demonstrates that there is no genuine issue as to any material fact and that the moving
party is entitled to judgment as a matter of law. Id.; see also Fed. R. Civ. P. 56(c);
Celotex Corp. v. Catrett, 

, 322-23 (1986).

       Under Weisgram v. Marley Co., 

(2000), an appellate court may
direct entry of judgment as a matter of law when it determines evidence was
erroneously admitted at trial and that the remaining, properly admitted, evidence is
insufficient to constitute a submissible case. In a like manner, an appellate court may
vacate a summary judgment on the ground that evidence that should have been
admitted or is shown to be available creates a genuine issue of material fact requiring
further proceedings.

A.    Exclusion of Dr. Erickson’s Proposed Testimony

       Dr. Erickson testified that, to a reasonable degree of medical certainty, Smith’s
neck was broken at the initial impact of her vehicle with the embankment, rather than
in the rollover. He also testified that to a reasonable degree of medical certainty the
air bag would have prevented or reduced her neck injuries, had it deployed.

       The district court concluded that Dr. Erickson’s opinions about the cause of
Smith’s neck injuries were not scientifically or medically reliable and therefore could
not assist a jury. The district court offered seven reasons for concluding that Dr.
Erickson’s testimony was not scientifically or medically reliable: (1) his experience
as a forensic pathologist was inapplicable to determining whether Smith sustained a
broken neck due to pure acceleration-deceleration forces because, as a forensic


                                          -9-
pathologist, he has never seen or read about a car accident where a person’s neck was
broken in this manner; (2) he did not know and could not estimate how much force
occurred on the initial impact of the accident or how much force would have been
required to break Smith’s neck; (3) he relied on medical knowledge as opposed to
scientific knowledge; (4) he did not perform a biomechanical evaluation using
engineering or physics; (5) he could not use engineering principles to calculate
possible accelerations upward in Smith’s accident; (6) he based his opinion, in part,
on Dr. Williams’ opinion as to the barrier equivalent velocity of Smith’s car at
impact; and (7) his “logically flawed” conclusion that the acceleration-deceleration
forces (i.e., the ones that occurred at impact with the embankment) acting on Smith’s
neck were more likely to have caused Smith’s injuries than the forces incurred during
the rollover portion of the accident because Smith had no head wounds, there was no
indication in the car that she hit her head on the roof, and she was wearing her seat
belt.

       The first five reasons9 offered by the district court all stem from what appears
to be the court’s ultimate reason for excluding Dr. Erickson’s testimony: he is not an
expert in biomechanics, physics, or engineering. The district court excluded Dr.
Erickson’s testimony because he did not know how much force Smith’s neck could
withstand, nor could he quantify the forces on Smith’s body during the accident.
However, Dr. Erickson was not called to offer expert testimony on those issues. Dr.
Erickson offered his opinion that Smith was injured when her vehicle struck the
embankment rather than during the rollover. He based this opinion on information
that fell within his field: the injuries Smith sustained (e.g., wedge fracture in neck)
and the injuries Smith did not sustain (e.g., no trauma to her head). He applied his


      9
       The sixth reason relies on the district court’s exclusion of Dr. Williams’
testimony, discussed in the text, infra. The seventh reason attacks Dr. Erickson’s
conclusion, and not the methodology used in reaching the conclusion. Under
Daubert, a court should review only the methodology of the expert, not his or her

.

                                         -10-
medical knowledge and his experience to the physical evidence and came to a
conclusion as to the cause of Smith’s neck injury. The fact that experts in other fields
might also be able to form opinions regarding the cause of Smith’s neck injury and
would base those opinions on factors other than those used by Dr. Erickson does not
disqualify Dr. Erickson from offering testimony that would be helpful to the jury. See
Smith v. Ford Motor Co., 

, 720 (7th Cir. 2000) (stating that an expert’s
testimony need not relate directly to the ultimate issue that is to be resolved by the
trier of fact, it only need be relevant to evaluating a factual matter); Clark by &
Through Clark v. Heidrick, 

, 915 (8th Cir. 1998) (stating that experts
who can offer a global understanding of the possible causes of an injury are useful to
a jury). The district court abused its discretion in excluding Dr. Erickson’s testimony
about when and how Smith suffered her neck injury.

       The district court also abused its discretion in excluding Dr. Erickson’s
proposed testimony about whether a properly deployed air bag would have reduced
or prevented Smith’s injuries. Dr. Erickson based his opinion on his knowledge of
the basic operation of air bags and his knowledge of how injuries of the type
sustained by Smith occur and can be prevented. That is a sufficient basis for him to
render an opinion on whether a properly deployed air bag would have reduced or
prevented Smith’s injuries. That Dr. Erickson was unable to quantify how much of
Smith’s assumed forward neck flexion occurred before a properly functioning air bag
would have deployed, or by how much Smith’s forward neck flexion would have
been reduced by an air bag, is not grounds to exclude his opinion, based upon factors
within his expertise, that to a reasonable degree of medical certainty a properly
deploying air bag would have reduced Smith’s injuries.

B.    Exclusion of Dr. Williams’ Proposed Testimony

      The district court determined that “the method used by Dr. Williams to put
information into the EDCrash program was fundamentally flawed. His opinions


                                         -11-
based upon those EDCrash results are, therefore, inherently unreliable.” In particular,
the court found fault with Dr. Williams’ estimates of principal direction of force and
his displacement measurements. The court excluded Dr. Williams’ proposed
testimony in its entirety.

Principal Direction of Force

       The district court concluded that Dr. Williams’ estimate of the principal
direction of force (“no more than twenty degrees into the hill”) was “contradicted by
the visible damage on the car and the principles of physics.” The court determined
that a -20° principal direction of force would not cause the passenger-to-driver side
displacement observed in photographs of the car. The court was “dubious” of Dr.
Williams’ explanation that counterclockwise rotation of the car as it struck and
climbed the embankment caused the passenger-to-driver side displacement
observable in photographs of the car. Dr. Williams “attempted to demonstrate his
theory using a tablet and his hand, but this courtroom demonstration did not appear
to [the court] to support his theory.” The court credited instead Mr. Bentley’s
testimony that there could be no significant counterclockwise rotation with a principal
direction of force of -20°,10 and concluded that “it appears to me to be a matter of
simple physics (not beyond the ken of knowledge of the ordinary lay person).”

      Dr. Williams based his estimate of the principal direction of force on all
information he could obtain, including interviews with persons at the scene of the
accident, his own inspection of the scene, and his own physical inspection of the
vehicle. Dr. Williams testified that the principal direction of force was “no more
than" -20° from being a head-on collision. From this range, Williams used the


      10
        The district court overstated Mr. Bentley’s testimony by taking him to say that
a principal direction of force of -20° would result in a clockwise rotation. Mr.
Bentley’s testimony actually supports the possibility of a small counterclockwise
rotation.

                                         -12-
principal direction of force resulting in the lowest barrier equivalent velocity. All
other things remaining constant, the closer the principal direction of force is to 0°, the
stronger the counterclockwise rotation created by a driver’s-side frontal impact, the
greater the barrier equivalent velocity, and the better Smith’s case.

       Neither Mr. Bentley’s testimony nor the district court's perception of Dr.
Williams’ hastily composed demonstration provides a valid basis for concluding that
Dr. Williams’ testimony about the principal direction of force is unreliable.
Therefore, the district court cannot base its exclusion of Dr. Williams' expert opinion
regarding the barrier equivalent velocity of Smith’s vehicle on his determination of
the principal direction of force.

Measurement of Frontal Displacement

       The district court concluded that Dr. Williams’ method for measuring the
frontal displacement of Smith’s car was flawed in that he failed to account for the
inclusion of “free space” or “air gaps” in his frontal displacement measurements.
Citing the Tumbas paper, the district court determined that air gaps should not be
included in the displacement measurements entered into the EDCrash computer
program. The court stated that it understood the rectangular method to be acceptable,
so long as the resulting air gaps are taken into account and subtracted from the
displacement measurements. With this understanding, the court determined that Dr.
Williams had “attempted to use the rectangular method” but “[a]t no point did he
account for free space.”

      We agree with the district court that Dr. Williams’ statements that the air gaps
account for restitution are unsupported by any evidence in the record. Smith offered
no evidence that remotely explains how air gaps measured between the curve on an
undamaged vehicle and the front of the approximating rectangle might compensate
for the omission of a restitution measurement. In addition, the district court


                                          -13-
determined that Dr. Williams’ measurement of a twenty-five-inch displacement at the
passenger corner of the front of the car was “clearly flawed” in that Dr. Williams
measured from the baseline of his rectangular grid to a point on the side of the car.
Again, we agree.

       We conclude that Dr. Williams’ measurement of frontal displacement was
sufficiently flawed to serve as a basis for the district court’s exclusion of Dr.
Williams’ testimony regarding the magnitude of the barrier equivalent velocity.11




      11
         That is not to say we agree with each and every reason the district court
offered for excluding Dr. Williams’ testimony. For example, the district court found
it significant that Dr. Williams’ method of measuring frontal displacement from a
rectangular form rather than from the dimensions of an exemplar (dimensions of an
undeformed vehicle) leads to a barrier equivalent velocity of approximately ten miles-
per-hour for a car that has been in no accident and has no damage. Note, however,
that if we consider a car in an accident at a low speed of, say, five miles-per-hour, so
that upon impact its bumper flexes and then rebounds to its pre-collision form, the
exemplar method of measuring frontal displacement would suggest there had been no
collision at all.

       The court also noted that Smith did not recall Dr. Williams to rebut Mr.
Bentley’s testimony or to explain why his (Dr. Williams’) opinions were reliable
despite Mr. Bentley’s “damning” testimony. This assumes that Smith considered Mr.
Bentley’s testimony damning. Smith’s decision to not re-examine Dr. Williams
regarding his disagreements with Mr. Bentley’s testimony does not provide ground
for the district court to exclude Dr. Williams’ testimony. Experts frequently disagree
and their opinions, if reliable, are for a jury's consideration. See Bonner v. ISP Tech.,
Inc., 

, 929-30 (8th Cir. 2001) ("As a general rule, the factual basis of an
expert opinion goes to the credibility of the testimony, not the admissibility, and it is
up to the opposing party to examine the factual basis for the opinion in cross-
examination. Only if the expert's opinion is so fundamentally unsupported that it can
offer no assistance to the jury must such testimony be excluded.").

                                          -14-
C.    Summary Judgment

      Dr. Williams’ displacement measurements contain errors as explained.
Therefore, the barrier equivalent velocity he obtained from EDCrash is inadmissible.
However, the district court erred in granting summary judgment.

      Mr. Blaisdell, BMW’s accident reconstructionist, corrected Dr. Williams'
displacement measurements to account for air gaps.12 Mr. Blaisdell entered the
corrected measurements, along with Williams’ principal direction of force estimate,
into EDCrash. The resulting barrier equivalent velocity of 14.4 miles-per-hour lies
within the 12 to 15.5 miles-per-hour range of barrier equivalent velocity values at
which there is an issue of material fact as to whether the air bags should deploy.13
BMW presented Mr. Blaisdell’s results in its Reply to Plaintiff’s Opposition to
Summary Judgment; these results were before the district court when it granted
summary judgment.

       This evidence, along with Dr. Williams’ admissible testimony regarding the
principal direction of force and Dr. Erickson’s testimony regarding causation, makes
summary judgment for BMW inappropriate in this case. We therefore vacate the
district court’s grant of summary judgment to BMW. See City of Tuscaloosa v.
Harcros Chem., Inc., 

, 566-67, 572 (11th Cir. 1998) (reversing grant of
summary judgment based in part on determination that a portion of expert’s
testimony, all of which had been excluded by district court, was admissible and that
the admitted testimony created a genuine issue of material fact).


      12
        Mr. Blaisdell also corrected what he perceived to be another error: a slight
alignment error between the axes of the vehicle profile and the “undamaged vehicle”
rectangle.
      13
           See note 

.

                                       -15-
      We remand to the district court for such further proceedings as may be
consistent with the Federal Rules of Civil Procedure and the district court's pre-trial
orders with respect to the disclosure, supplementation, and amendment of expert
opinions.

III.   Conclusion

      We dismiss the appeal on the denial of relief under Fed. R. Civ. P. 60(b) and
reverse the grant of summary judgment. The case is remanded for further proceedings
consistent with this opinion.

       A true copy.

             Attest:

                      CLERK, U.S. COURT OF APPEALS, EIGHTH CIRCUIT.




                                         -16-