Circumstantial Evidence of a Head Shot From The Grassy Knoll
                 (c) - Copyright 1993 by W. Anthony Marsh
         Presented at The Third Decade conference June 18-20, 1993

     As much as we would like to have direct evidence of a head shot from the
grassy knoll, such evidence may be missing, inconclusive, or suspect.
However, there may be a body of circumstantial evidence which would indicate
that the fatal shot which struck President Kennedy's head at Z-313 came from
the grassy knoll. This paper will not present conclusive proof of a head shot
from the grassy knoll, but it will cite examples of circumstantial evidence
which strongly suggest that the head shot came from the grassy knoll. Some of
the examples are well known, but need to be reexamined.
     The Zapruder film is the most well-known evidence of the head shot.
Various studies and interpretations of it have been made. Some studies, such
as the one done by Itek, have analyzed the movement of President Kennedy's
head around the time of the head shot. They note that President Kennedy's
head moves forward significantly from Z-312 to Z-313 and cite that as proof
of a shot hitting the head from behind. What they and everyone else has
failed to do is analyze the movements of all the occupants of the rear
compartment of the limousine, including the Connallys. That is what I have
     My analysis of the movements of the Kennedys and the Connallys is not,
unfortunately, based on the same reproductions of the Zapruder film as used
in other studies, due to cost considerations. I made measurements in 1/60th
of an inch increments on a photocopy set of prints from Zapruder frames 312
to 321, as reproduced by Robert Cutler in his dividend to The Grassy Knoll
Gazette of X-79. Bob's reference line is drawn through the center of the
window knob. I made all measurements starting at the front edge of his
reference line. However, I noticed that the distance from the reference line
to the rollbar is not constant. This means that we can not use unadjusted
measurements from these prints to calculate precise positions, but can
estimate relative movements. This may be due to a variety of factors, such as
variations in printing and copying each frame, changes in perspective,
mismeasurements, or blurring. Some Zapruder frames are too blurred to allow
accurate measurements. Each measurement of Nellie Connally's position is to
the front edge of her hair. Each measurement of John Connally's position is
to the front edge of his forehead. Each measurement of Jackie Kennedy's
position is to the front edge of her pillbox hat. Each measurement of JFK's
position is to the edge of his hair at the rear of his head. All measurements
were lined up against the chrome strip in the background for better contrast.
Be sure to remember that increasing measurements for the Connallys represent
forward motion, while increasing measurements for the Kennedys represent
rearward motion. Notice the direction and amount of movement of each person
listed in Figure 1. Between Z-312 and Z-313, all the occupants of the rear
compartment of the limo moved forward by about the same amount. Unless all
four were hit by bullets (a practical impossibility), their forward movement
must be caused by something else. The most likely cause is inertia due to the
limousine having suddenly slowed down. Dr. Luis Alvarez noted in his study [1]
that the average velocity of the limousine going down Elm Street sharply
decreased just before the head shot. Some researchers have theorized that
Secret Service agent Bill Greer jammed on the brakes or took his foot off the
accelerator. Whatever he did, the limousine very quickly changed from an
average velocity of about 12 MPH to about 8 MPH just before the head shot.
Obeying the law of inertia, passengers in the limo were thrust forward in
relation to their previous positions in the limousine. Further evidence of
this effect is the fact that the Connallys continued to move forward while
President Kennedy was being thrust backwards. I have not done a similar
analysis of previous Zapruder frames to pinpoint the start of the occupants'
forward movement, so I would urge others to do so themselves, in order to
verify my results and observations. Figure 1. 
            Z-frame    rollbar    Nellie     Connally    Jackie      JFK
 x/60"      Z312        136        106         77          44        159
difference                9      forward 5  forward 10 forward 6  forward 7
            Z313        145        111         87          38        152
difference                2      forward 2  forward 1  rearward 1 forward 1
            Z314        147        113         88          39        151
difference                4          0      forward 4  forward 6  rearward 6
            Z315        151        113         92          33        157
difference                3      forward 2  forward 1  rearward 1 rearward 9
            Z316        148        115         93          34        166
difference                2      forward 2  forward 7  forward 4  rearward 11
            Z317        150        117        100          30        177
difference                3      forward 5             forward 2  rearward 5
            Z319        153        122        NA           28        182
difference                0      forward 4             forward 3  rearward 14
            Z320        153        126        130          25        196
difference                4                            rearward 1 forward 1
            Z321        157        NA         NA           26        195

     Perhaps the most controversial evidence produced by the House Select
Committee on Assassinations would be the acoustical studies. In my opinion,
the conclusion that there was a conspiracy should not rest entirely on the
acoustical studies. But the acoustical studies are useful for establishing
the time between shots. All times are measured in seconds after the
microphone became stuck open for several minutes. BBN found 4 shots on the
tape, 3 of which came from the TSBD at 137.70, 139.27, and 145.61
respectively. The grassy knoll shot was found by Weiss and Aschkenasy to be
recorded at 144.90. There was a fifth set of impulses which was rejected by
HSCA as being a false alarm at 140.32. I have looked at the waveforms more
closely to try to determine when the muzzle blast of each shot was recorded,
to a greater degree of accuracy. My best estimate for each shot is 137.702,
139.268, 140.339, 144.895, and 145.608. You can get a general idea of the
spacing between shots by subtracting one time from another. But there is an
additional variable which must be taken into account. BBN found that the
recorder used that day was running about 5% slow, so all times must be
multiplied by about 1.05 in order to restore the original spacing. A more
accurate correction factor might be borrowed from the work which W&A did on
the grassy knoll shot. They found that a correction factor of 1.043 produced
the best fit for echo delays compared to their predicted model. Another
possible corroboration for the 1.043 correction factor is the 'bell' sound
found by BBN at 152.5. Although Todd Vaughan believes that it is only
electrical interference, if we can determine its true frequency, we can
derive the most accurate correction factor. That holds true for many other
sounds on the tape, such as car horns, tire squeals, police sirens, etc. BBN
found that the 'bell' sound had a nominal pitch of 420 Hz. This is close to
the note A, which is usually 440 Hz. If the sound is really a bell, it might
have been tuned to A=440. We do not know for sure what type of bell it was.
Most people have assumed that it is a carillon bell, but it could be a train
bell, a ship's bell, or a victory bell on a college campus. There are a
couple of other possible tunings which would produce a correction factor
close to 1.043. If the bell had been tuned using a mean-tone temperament
scale, it might have a real pitch of 438.075 Hz. Dividing that by 420 would
give us a correction factor of 1.0430357. If the bell had been tuned to an
old English standard of A=438.9, dividing that by 420 would give us a
correction factor of 1.045. Applying the correction factor to the spacing
between shots as found by BBN will give us the true spacing between muzzle
blasts picked up by McLain's cycle. If we want to then translate those into
Zapruder frames, we must multiply each interval by 18.3 frames. Figure 2 is
a rough approximation of how many frames there were between all 5 muzzle
     Matching these times to the Zapruder film is more complicated and
depends on making several real-world assumptions such as the speed of the
bullets. We can be fairly confident in ruling out the first three shots as
matching the head shot at Z-313, as such a match would place the first shot
after Z-255, when we can clearly see in Altgens 1-6 that President Kennedy
and Governor Connally have already been hit. The HSCA matched the last shot
with Z-313, because their medical evidence indicated that the head shot came
from behind. My alternative matchup tests the idea that the head shot came
from the grassy knoll.

Figure 2.
origin   tape time  spacing   *1.043  *18.3    Z-frame    Z-frame
TSBD     137.702                                162        176
                >   1.566     1.633   29.89
TSBD     139.268                                192        206
                >   1.071     1.117   20.44
TSBD     140.339                                212        226
                >   4.556     4.752   86.96
Knoll    144.895                                299        313
                >   0.713     0.744   13.61
TSBD     145.608                                313        327

     The first problem we notice with the HSCA version is that the first shot
is much too early. No other evidence supports a shot that early and clearly
President Kennedy was not hit by a bullet close to that frame. The first shot
was probably a miss. The HSCA places the hit to JFK's back at around Z190-
192. The problem with that is that we can see President Kennedy in the
Zapruder film during the range Z-190 to Z-210. He does not yet appear to be
reacting to being hit by a bullet. There is absolutely no indication that
Governor Connally was struck by a bullet at about that time, nor at about Z-
210 to Z-212, if we accept the fifth shot which HSCA rejected.
     My matchup would indicate a hit to JFK's back somewhere in the range of
Z-206 to Z-210, and a hit to Connally's back somewhere in the range of Z-226
to Z-230. I believe this is more consistent with previous studies of the
Zapruder film and eyewitness testimony. If there is some way to prove exactly
when either President Kennedy or Governor Connally received their back
wounds, that would force us to choose between the HSCA version and mine,
regardless of other evidence.
     Just as Altgens 1-6 helps us eliminate the first three shots as matches
with Z-313, it may also help us eliminate the last shot from the TSBD as
matching Z-313. Everyone is familiar with the fact that CE350 shows a crack
on the windshield and that it is not seen in Altgens 1-6, but is seen in
Altgens 1-7. There has been some doubt about which shot from which direction
caused that crack. I believe that I am the first person to notice something
in CE350 which would resolve the doubt. If you look carefully at CE350, you
will notice that the back of the rearview mirror is dented, but you can see
that it was not dented in Altgens 1-6. This damage was caused by a bullet
fragment which struck the windshield from the inside and ricochetted into the
rearview mirror. Many people believe this fragment came from the head shot,
which would been fired from the TSBD. I tend to feel that all the damage to
the limousine, consisting of the crack in the windshield, dented rearview
mirror and dented chrome topping, was done by the same shot. If we can find
evidence which pinpoints when that damage was done, we may be able to show
that it came several frames after Z-313. Photographic enhancement of the
Zapruder film, Muchmore film or the Bronson film might reveal that the
windshield was not cracked by Z-314. If that turns out to be true, then the
last shot from the TSBD must have missed JFK's head and hit the windshield.
In turn, that would mean that the fatal head shot came from the grassy knoll.
I seriously doubt that there is any photograph which would show exactly when
the rearview mirror was dented, but perhaps some as-yet-undiscovered
photograph would show when the chrome topping was dented. That dent could
only have been caused by a shot from behind the limo. If it was undented at
Z-314, that would prove that the head shot came from the grassy knoll.
     Another factor which might influence our choosing the HSCA version or
mine would be the jiggle analysis of the Zapruder film. Even the HSCA
admitted that the jiggle analysis matched better when the grassy knoll shot
was lined up with Z-313. [2]  Figure 3 compares the timing of the impulses
to the Zapruder film. The HSCA rejected the shot which is indicated in
brackets as being too fast for Oswald to have fired. The jiggle analysis
measured the amount of panning error by Zapruder. To simply and clarify,
I have put the groups into ascending order. The group with the largest
amount of blur is marked 'A', the second largest 'B', etc. I have chosen
the Hartmann figures to be representative, as his are usually midway
between Alvarez or Scott's figures. Zapruder's camera ran at 18.3 frames
per second on average.

     Figure 3. Two comparisons of jiggle analysis to acoustical data
Frames    Group 
158-159   D 
191-197   B                Note that the start of a jiggle group may not 
227       C                coincide with the firing or impact of a bullet. 
290-291   E                In most cases, it takes several frames before 
313-318   A1               Zapruder reacts to a stimulus. 
331-332   A2

HSCA   Z-#   hit?    origin  jiggle   Marsh    Z-#   hit?     origin  jiggle
137.70 161   miss    TSBD    D        137.702  179   miss     TSBD VI#1  B
139.27 191   JFK/JBC TSBD    B        139.268  209   JFK      TSBD VI#1
[140.32]                              140.339  230   Connally TSBD VI#10 C
144.90 297   miss    knoll   E        144.895  312.6 JFK      knoll      A1
145.61 312   JFK     TSBD    A1       145.608  328   Connally TSBD VI#1  A2
     The jiggle analysis can not be used as absolute proof of when a shot
occurred, but it matches up better for the head shot from the grassy knoll.
     Could eyewitness testimony help resolve the question of which shot hit
what? Secret Service agent Clint Hill testified (2H144) that the last shot he
heard sounded as though it had hit some metal place. If he in fact had heard
the last shot from the TSBD hit the chrome topping, that would not, in and of
itself, prove that the TSBD shot missed JFK's head, as the dent could have
been caused by a fragment from the head shot. But it would narrow the range
during which the chrome topping was dented to between Z-313 to Z-331 and make
it more likely that the chrome topping was dented at the same time that the
windshield was cracked, rather than much earlier as some have speculated.
     On pages 126-129 of Six Seconds in Dallas, Josiah Thompson cites the
statements of several witnesses who thought that a shot came from the grassy
knoll. William Newman felt that he and his family were in the direct path of
gunfire. Given their position, it seems more likely that the head shot came
from behind the fence on the grassy knoll than from the TSBD. Emmett Hudson,
who was standing on the steps leading up to the pergola, said that the shots
sounded as if they came from behind him, above his head and to his left. That
would place the origin near the fence. Zapruder felt that the head shot had
come from behind him and whistled past his right ear. Between these two
witnesses and behind them is the corner of the fence. W&A found a probable
shockwave at 24 ms. before the muzzle blast of the grassy knoll shot.
Assuming the weapon was aimed at the limousine, we can make a rough
calculation of the velocity of the bullet and the resultant angle of the
shockwave. Although the calculation for the decay of the shockwave is too
difficult for me, a rough calculation yielded an exit velocity of the bullet
at about 1564.5 fps. This would create a shockwave of at least 45 degrees on
either side of the flightpath of the bullet. All three witnesses were within
the cone of that shockwave and would have felt it very strongly.
     How can we know whether the weapon was aimed at the limousine? If it had
been aimed in some opposite direction, the open microphone would have been
outside the cone of the shockwave and thus the shockwave would not have been
recorded. Another indication of where the weapon was aimed can be found in
the statement that Sam Holland made to Josiah Thompson when he was shown a
very clear copy of the Moorman 2 Polaroid. He felt that the viewpoint was
looking right down the barrel of the gun. Given Mary Moorman's position, the
gun would have been aimed at the limousine. In Moorman 2 we can see the head
of a man peering over the fence, about 9 feet from the corner. Interestingly,
this is the same spot where W&A located the origin of the grassy knoll shot,
unaware of the existence of Moorman 2. In Moorman 2 we do not see a side view
of any weapon as we might expect if it was not aimed at the limo. Whoever
this man was, he moved from that position very quickly after the head shot.
There is no one there in later Zapruder frames or in the Stoughton
photograph, taken shortly after the head shot. The Stoughton photograph has
never been analyzed. The Warren Commission and the House Committee were
unaware of its existence. I believe that no one had ever studied it before I
found it at the JFK Library. Unfortunately, I do not have the resources to
properly analyze it and obviously the government will not, as it might reveal
the presence of a gunman on the grassy knoll. One independent researcher,
Dale Meyers, has done some computer analysis which suggests that there is a
person behind the fence, several feet to the right of the corner.
     Many witnesses can be seen reacting to the shots in various films. Some
fell to the ground very quickly while others did not seem to realize the
danger. We would expect those witnesses closest to the gun on the grassy
knoll to react very quickly and dramatically to the shockwave whizzing past
them from the grassy knoll. One of the best films to observe their reactions
was taken by Marie Muchmore. In Muchmore frame 42 we can see Hudson and his
two companions on the steps leading up to the pergola. They do not seem to be
reacting to any gunshots. In frame 55 we can see that two of the men are
reacting. It seems inconceivable that these men would not have reacted to the
sound of a shockwave coming from so close behind them. If the grassy knoll
shot was the miss before the head shot, we would expect to see these men
react before the head shot. The fact that they did not react until after the
head shot would seem to indicate that the head shot came from the grassy
     Based on the circumstantial evidence we have now, I believe it is more
likely that the head shot came from the grassy knoll. Existing evidence can
and should be examined further. More information can be gleaned from existing
data by novel analyses. The release of withheld data could provide new clues
and allow us to verify certain methods, such as the acoustical studies. I
urge other researchers to look for new evidence and perform new analyses.

1. W. Peter Trower, ed., "Discovering Alvarez", (Chicago: The University
   Press, 1987), pp. 210-224. Also HSCA Vol. I, pp. 428-442.
2. House Select Committee on Assassinations, Report, p.80, footnote 16