Battlefield Guide Services

Thursday, January 12, 2012

Gettysburg - Harvest of Death - Location Found, January 11, 2012, Confirmed June 14, 2012

REVISION: I have floated the revised analysis I am posting here for several years, but have not, till now, published it on this blog. I stated it publicly in April 2016, on location, in the presence of NPS Historian John Heiser, and numerous other attendees. I am very convinced that the dead seen in these images are members of the 151st Pennsylvania Infantry Regiment. Initially I would have thought them to be the dead of the 24th Michigan, but what I understand now shows me that the Iron Brigade skirted to the north of this location, falling back, past the 151st Pennsylvania who were going to take the brunt of Virginia and North Carolina units pressing through the woods. The 151st Pennsylvania made their stand on the crest, in front of the edge of the woods where Reynolds fell, until falling back across the field bordered by fences to the north and south. This field is the reason why the "Field Where Reynolds Fell" captioning is derived by Gardner. I am officially publishing this revision on 10/12/2018.

Be sure to read the post attached to our February 25, 2013 link. Click here. It further demonstrates this site is located on the first day's field. Clear visual evidence. 

This post has been revised to reflect the corrected findings of my October 6, 2012 return visit.
Be sure to read the October 12, 2012 post on this subject by clicking here. It contains more current information on the location.
Additional information can be found on the June 14, 2012 post by clicking here.

     One of the most elusive group of Civil War battlefield photographs has been the image known as “The Harvest of Death” and its companion images, showing the same bodies from a near opposite angle. The group was made in the days following the battle, most probably July 6, 1863. There have been three prevailing theories as to the precise location of this ghastly scene. Since 1975, historian William A. Frassanito has felt comfortable in his belief that the group of photos was taken somewhere near the Rose Farm and the Emmitsburg Road. The greatest challenge has been in trying to locate a piece of land where all terrain features cooperate, in both directions, to make the theory work. Granted, doing this has been made even that more difficult by over a century’s growth and/or removal  of wood lines and other obstructions, as well as potential modifications to the landscape surface. The mission has continued for another thirty-seven years as numerous investigators have scoured the battlefield landscape to find the right combination of elements.

Gardner caption, "View in field on right wing."
Library of Congress collection.

     In May 2011, Gettysburg National Park Service Historian Scott Hartwig published his own theory, located in an area below the Chambersburg Pike, near McPherson’s Woods that seemed to work, and was much more in keeping with the original captioning of the published images by photographer Alexander Gardner. It runs along the Park Service road, Reynolds Avenue South.

     Most recently, historian Jerry Coates has spent considerable time investigating an area south of Gettysburg, and west of the Emmitsburg Road. This theory stayed more within the confines of the original Frassanito analysis, but was supplemented by a more in depth study of the uniforms the dead are wearing and attempting to use this criteria to narrow down the numerical potential as to who they might have been and thus the field of battle these men would have been engaged on in July 1863. He has come up with very similar terrain and some good arguments.

     However, my own investigative work has concluded that of all these prevailing theories, the one that seems to come the closest is Scott Hartwig’s on the first day’s battlefield. But, although very well founded and assembled with sound judgment, Hartwig’s theorized location has its faults and can be termed with the old expression, “Right church, wrong pew.”

     What all the prevailing theories lacked was a concrete feature or landmark that would anchor the image to an indisputable location. My examination of the series of images has discovered what appears to be that necessary clue. The clue has been available in plain sight to all that have looked at the glass negatives for the stereo pairs, and quite possibly the 8 X 10 glass negative, although that full image is not printed or available online presently by the Library of Congress, but is referred to by William Frassanito and shown in cropped form on page 227 of his Gettysburg: A Journey In Time. In many prints of the image entitled “Field where General Reynolds Fell”, the full image has been cropped and has thus removed the landmark feature. The landmark looks like none other than the residence of Mrs. Mary Thompson, known more famously as “Lee’s Headquarters”. In the upper right hand corner, again invariably cropped out in published formats, the house is easily discernible along the horizon line. It is understandable that it could be easily overlooked as just another shadow in the line of trees, but when magnified, and compared to wartime images of the structure, it becomes more than obvious. In utilizing the Thompson house as the anchor for this image, it becomes far easier to establish a camera position, and thus, possibly prove that Gardner’s original captioning was spot on as to the vicinity of the battlefield.

Is this the home of Mrs. Mary Thompson, "Lee's Headquarters,
as seen in an enlargement of one half of a stereo glass negative?
Annotated copy of the previous enlargement showing key elements.
Find these same elements in the image below for comparison.
All images are clickable for enlarged viewing.
Photograph of Mrs. Thompson's house, attributable to Mathew Brady.
Note the white fence, chimney, and trees on either side of the building compared with
 those features in the enlarged views preceding. Is this the rock solid clue to these images?
Library of Congress collection.

Overlay with Thompson house image by Brady at 75 percent
transparency over top shadows seen in Gardner image, right horizon line.
Brady's street level image was of course taken close up, but even at
464 yards, the shape of the house and surrounding trees is discernible.

     Two field reconnaissance trips to the Gettysburg field had given me the evidence to support this theory. On November 18, 2011 and January 11, 2012, my investigations concluded that the dead in this series of exposures were likely killed south of the Chambersburg Pike, between McPherson’s Ridge and Seminary Ridge. This section of the field was just below the northernmost of two fence lines that ran parallel to the Pike and enclosed a field that included the Herbst Woods in its western extreme. If so, these men fell approximately 100 yards east of the spot marked where General Reynolds fell, and roughly 466 yards southwest of Mrs. Thompson’s house. The camera location is roughly 190 yards northeast of Scott Hartwig’s theorized conclusion, which was south of the lower enclosing fence, and too far south of the Thompson house to have allowed for the structure to appear as it does in scale along the Chambersburg Pike Ridge. Jerry Coates theoretical location is approximately one and three quarter miles south of the Seminary Ridge killing field.

Gardner's camera position adjusted October 6, 2012 with further corrections October 12.
 Click on this and all images to enlarge.
Green triangles indicate approximate field of view of lens for both images.
Light tan lines indicate approximate war-era fences separating field.

The Gardner image, note the faint Thompson house at upper right, along horizon.

Approximate same view as of October 6, 2012. Thompson house to right of motel complex.
Gardner's "Harvest of Death"
See my September 28, 2012 posting regarding the horizon line and
 apparent doctoring of the image by Alexander Gardner.
Approximate same view as of October 6, 2012.
For questions regarding the horizon line, visit this additional post on the subject.
Initial investigation, and attempt to set up the shot towards the Thompson house.
Seminary Ridge is in the background. November 18, 2011.
 Photo by James Anderson.
Looking for the correct view toward the Thompson house on November 18, 2011.
Photo by James Anderson

Be sure to visit the June 27, 2012 post
for the complete story
Monument to the 24th Michigan overlooking
Willoughby's Run. January 11, 2012.
These men, and others of the Iron Brigade, exited this area, and fled past the
151st Pennsylvania who would cover their retreat, taking heavy casualties.
Mrs. Thompson's house as it appeared January 11, 2012 with the addition of
dormer windows expanding the second floor. At the time of this photograph
the home served as the Lee's Headquarters Museum.

View of the field from Reynold's Avenue South, June 14, 2012.


Richard Goedkoop said...

Make that the 24th Michigan of Meredith's Brigade.

Garry Adelman said...

Looking forward to checking this out, John, and I am glad to have what I think is the 31st theory advanced! Suffice it to say that I do not as of now consider this site confirmed "beyond the shadow of a doubt" but it is in many ways infinitely superior to Hartwig's. Superior is not what we need, however. We need it to not contain any issues that make your proposed site impossible and in fact, make it the reverse--absolute. I have some real concerns right off the bat. More later...

Garry Adelman
Vice President, Center for Civil War Photography

John Cummings said...

Commanded by Meredith, yes, but still called the the Iron Brigade.

John Cummings said...

Thanks for checking out my theory and I will be happy to chat with you and/or walk the fields any time you want, schedules permitting. I welcome your thoughts as I am quite comfortable with my findings but encourage dialogue. I can't imagine finding a way to dismiss the Thompson house, and some of the trees across the Pike in the Orchard. Everything comes down to lenses, angles and distances. Move any direction five to ten feet maybe a fine tune adjustment, but there she sits.

John Cummings

Garry Adelman said...

Well... there are numerous ways, John, but I won't have time to get into them as quickly as I'd like. Start with exactly how you are 100% certain that what indeed seems to resemble a white-sided (the chimney would is dark but somehow the similarly shaded house is invisible) version of the of Lee's HQ is in fact that building and why the trees you say are next to the house are absolutely much closer to the camera, especially in 3D (that view is not an 8x10plate as you say, but half of a cropped 4x10 stereo negative) and why the fence posts are not visible anywhere along the road in any of the three views). Then, I would address why the ridge in the distance does not appear as tall (again, see 3D) and then I would take the July 1863 Gutekunst shot looking westward on the Chambersburg Pike and confirm exactly which trees in that shot are the same trees behind General Lee's HQ in yours. Then, I would address the rise on the right in the southward shot. That'll be a start! Explain these and you'll go a long way to at least consider why in the world Gardner would not use the seminary as a backdrop for any of the five views--or even record a picture anywhere else on the first day! More later...

John Cummings said...


The view I used is indeed one half of a stereo negative, but there is, according to Mr. Frassanito, an O'Sullivan plate, an 8 X 10, that depicts members of the burial crew standing by, and which is also cropped in his use of it in "A Journey In Time", page 227. There is very little variation in the camera angle.
As for the way the house, at a distance and out of focus of course, id devoid of the sharp details you ask of. The fence posts are simply not discernible. As for the house looking whiter than it is has everything to do with the reflection of light on the stone wall surface.
Now, as to why Gardner didn't photograph anything else it seems in that vicinity is anyone's guess, but we can not be confident that everything he may have taken still exists or even turned out well enough to keep. But one thing is for certain, he did refer to one of the views as being "Field where Reynolds fell", which either was a manipulation to have something Brady photographed represented in his catalog, as Frassanito suggests, or he did indeed have this image taken in the very field that I suggest in close enough relation to the scene of Reynold's death.
Regarding the "rise to the right" in the southward image, one has to see it or not, and again it comes down to two degrees here or five feet there. It can all be tweaked, as these were not done with the assistance of a computer, simply by eyeball. However, there are terrain features that do make themselves apparent, too much so to be coincidentally. These images had to be taken someplace, and eventually the time arises that they do manifest themselves. The real problem with the "Harvest of Death" image is that it looks different in any one of the exposures/prints made of it. The background is far too unfocused to expect sharp, contrasting lines.
I have worked with several images from the Spotsylvania area that have structures in the distance, all soft and washed out, yet distinctly shaped to leave no doubt as to there identity when compared with known images of the same building.
We really should not try to hash this out through the comments. Lets spend eye to eye time with some hopeful field work and I am confident I can win you over.
I do appreciate your prior yet apparently partial thumbs up.
Talk to you soon.

John Cummings

Richard Goedkoop said...

Your text still reads "29th" Michigan. See my Gettysburg Daily Series at

John Cummings said...

Richard, yes, a slip of the digit. Always meant to say 24th. Did so in depicting their monument. Corrected. Thanks.

Jim Lamason said...

Hi John,
The only problem I have with this conclusion is that I dont see the Thompson house in the old photo. There is something there, but even after staring at it, trying to trace it, I still dont see the house. And I have been on that part of the field a lot. Let me be clear. I am not saying your position is wrong. I just dont see the house in it. JIM ;-)

patt said...

John the over lay helps tremendousley.....i would have bet hard money before that the two pics were takin at different locations based on the apperant differance in topography.again I am blown away by your rock

John Cummings said...

I have added yet another overlay, concentrating on the shadows vs. the Brady image of the house. I believe you will find that it demonstrates the lines of the roof, chimney, and surrounding trees. Let me know if this makes it better.
Thanks for looking in.

John Cummings

Jim Thomas said...

John, I like the topography. I too have issue with "seeing" a house, but there is something there. I'm not sure the white picket fence would be visible. Allow me to suggest that the numerous white blotches around "the house" could be tents. And, if that's the case, perhaps it is a few white tents in the foreground that are making the house appear white...with only the roofline and chimney visible over them.

John Cummings said...

Jim Thomas,
Thank you for the comment. This morning I added an annotated version of the enlargement, with red lettering. If you have not seen it yet, please take a look and see if it helps. There are too many shapes in the right areas to be just a coincidence, especially when aligned with the topography. As you mentioned yesterday, the darker contrasting bricks of the chimney are relflecting light differently than the stone walls. And, as you suggest, anything else that might be reflective in the area could add to the washout effect.

John Cummings

Anonymous said...

John, I'm in awe, as always, of your sharp eye, hard work, and great graphics. As a veteran of blog postings with extensive comments, I'd actually vote for erring on the side of too much "hashing-out" in the comments, rather than too little, so that your readers can benefit from a full discussion and the perspectives of folks who've spent a lot of time on the ground. Especially in this case, when, as you say, the topic is a set of images that has long fascinated the photohistorical community at large. Again, great work! Noel H.

John Cummings said...

Thank you for your kind words. I always welcome your input and value your opinions.
Hopefully we can get together soon and talk shop. It has been a while.

John Cummings

Todd Berkoff said...

Hi John --

I would posit that the dead could be from any of the regiments of the Iron Brigade, not only the 24th Michigan--the entire brigade retreated across that field late in the day. Also, the 149th and 150th Pennsylvania advanced and retreated across these fields, finally falling back to the Seminary Ridge line where the two regiments took position on the right flank of the Iron Brigade. The dead could be from these Pennsylvania regiments too. Any signs of Federal Dress Hats/Hardee hats among the debris that soldiers in the Iron Brigade would have worn? If only forage caps, then most likely the dead are from Stone's Brigade. Also, I believe the 95th New York, 6th Wisconsin, and 14th Brooklyn charged across these fields toward the railroad cut and took casualties, although most of the dead in this attack would be north of the Chambersburg Pike I would think. Great work, as usual.

John Cummings said...

Thank you for your comment. You are very correct that many regiments passed over the fields west of Seminary Ridge, but within the confines of the two bordering fences of this field, the 24th Michigan would have been essentially confined by the north fence with the 19th Indiana to its left. This is the field where they fell back with the 26th North Carolina in hot pursuit. Although in retreat, regiments attempted to maintain ranks and stay together, not scattering hither and yon. Veteran regiments like those of the Iron Brigade would be especially disciplined to stay together. The action funneled through the corridor created by the north and south fences, fences that during this fight, would not have been previously dismantled as they were not an immediate impediment to the opening Federal advance toward West McPherson Ridge and Willoughby's Run.
Stones Brigade would have been on the opposite side of this north fence, within their own corral so to speak, just below Chambersburg Pike.

As for the lack of Hardee hats on the ground around these dead, it has to be remembered that these bodies lay behind Confederate lines until after July 4, and were clearly stripped of seviceable clothing and equipment. Hats, in particular anything that could be reshaped as a slouch hat, would be prime targets of opportunity.

I hope to meet up with you again this year as 2011 passed us by somehow.

John Cummings

Clark Buckner said...

Great work! There is a Military History online blog about these photos. Someone has blown up these pictures and posted sections of them. The detail is insane! You can make out a kepi[maybe not all Iron Brigade after all], and the contents of a spilled haversack.

John Cummings said...


Thank you for your comment and the information on the MHO discussions. These images have been the center of great debate for a long time. Lots of theories. I think the real problem most have had has been the belief Gardner was not on the July 1st battlefield. His discovered manipulation of the "sharpshooter" body at Devil's Den has made many hesitant to believe his narratives all around. In this case, he was being truthful about it being the near where Reynolds fell.

John Cummings

Todd Berkoff said...

Jerry Coates claims that the Union dead were from the skirmish line of the 5th New Jersey located at the Henry Spangler farm. This seemed fishy to me. There is way too much debris, and more importantly, too many dead bodies for a skirmish line--and the dead are in too close proximity as well. Soldiers in a skirmish line would be spaced farther apart and one would not expect so much debris if the 5th New Jersey occupied this position for a relatively short period of time. I think these facts corroborate your analysis that the pics show the dead from a battle line, not a skirmish line. John, I hope we can meet up for some treking in 2012.

Garry Adelman said...


The location of the famed Harvest of Death photo series, first detailed and outlined by William A. Frassanito, is what I and many consider the greatest Civil War photo mystery. I have been actively searching for this site since 1988. More importantly, however, I have been examining the theories advanced to Frassanito for almost twenty years now and I have seen that most of these theories, which now number around 30, distinct sites, have plenty in common. Namely, each theory was full of problems and most theorists were reluctant to deal with these problems. In many cases, the theorist simply chose to ignore the evidence that makes their theory impossible while focusing only on evidence that supports it. Many have proclaimed their sites to be absolutely correct—not theories at all—even in the presence of contrary evidence and before other historians weighed in on the matter. No historian, or detective for that matter, can succeed in this manner.

I will advance a discussion of the five photos in the series and provide very specific examples of problems with the recently-advanced theories—John Cummings’ (January 2012), Scott Hartwig’s May 2011 (although first advanced to Frassanito in 1999 by Mr. John Stewart) and Jerry Coates (first advanced in the late 1980s) in a series of dedicated posts on the Gettysburg Daily, probably in February 2012. I also hope to examine several other sites advanced by Messrs. Harman, Teague, Gibson, Martin, myself, and others.

For photo research in particular we need to apply a different standard than to work that involves textual accounts. With the latter, we regularly assemble evidence and weigh which are more likely to be correct and why. If 400 witnesses say one thing and two say another, and they are otherwise equal in bias and apparent accuracy, we usually side with the 400. With photo research you can have 400 things that support a theory but if one single photo taken that same week, or month shows features that should, but fail to, appear within the field of a proposed location’s photo, the 400 pieces of potentially supporting evidence become irrelevant, unless the difference in features can be explained. In short, photos can provide what I regularly call “deal breakers” that unless truly explained make a theory impossible. All three of my own, proposed Harvest of Death sites ultimately fell into this category and I never advanced them publically (although I still have some small hope for one of them!). With photo research we must be our own, harshest critics and actively work to prove our own theories wrong. There is no theory more correct than another—a site is either absolutely correct or completely wrong.

So, for now, I hope you will examine the images and locations of the various theorists for yourself. Ask why none of them leverage the 3D format in which 3 of the 5 photos were taken? Why do the features on early battlefield maps alone make almost all Harvest of Death theories impossible, alone? Why are the theorists tending to focus on one of the angles to the exclusion of the other? Why none of them leverage other historic photos that overlap their proposed fields of view? Form your own conclusions and I hope you’ll check out my take on the matters soon.


Garry Adelman
Vice President
The Center for Civil War Photography

John Cummings said...

Garry and all following:
I have more that I will be posting as a follow up piece to my 1-12-12 blog post. Be patient for that as well. In time.

John Cummings

Richard G. Williams, Jr. said...

This is quite fascinating John. My wife and I have stayed at Lee's headquarters before when in Gettysburg. I'll link to this post soon. Thanks for sharing the info.

Phil Spaugy said...

The problem I have with this view, is the lack of confederate dead, or the signs of freshly dug graves.

It is correct to say this was a killing field on the afternoon of July 1st, but it was that for both sides, particularly during Penders final advance against the First Corp position just west of the Seminary. I also seem to remember that Battery B of the 4th US Artillery, the left half section of which was in position just to the west of the Thompson house, wheeled to the left to enfilade the left flank of the confederate advance with canister with great effect.

Just my two cents. That said, I sure appreciate the the time, field work and scholarship that goes into trying to solve these historic mysteries.

John Cummings said...

Thank you for your comment.
Yes, Stewart did swing his guns around and fire into the left of the advancing North Carolinians after having held off those advancing on his front, north of Chambersburg Road. The effective range of the canister would have been contained within the area of the orchard on the south side of the road. It is noted they took shelter behind the trees. As for burials, the Confederates would have taken care of burying their dead over the next two days while in possession of the Seminary Ridge. The Elliott map indicates all the burials in that area, both Union and Confederate, are contained on the north side of the fence that dividing the field, all of which is in the soft focus area of the image looking toward Thompson. It is difficult to tell any details in that part of the field, although there is one slightly darker line that runs to the left of the two big trees and that may either be a burial ditch or a breastwork that is also indicated on Elliott's map. A very good question, and hard to answer with certainty.

John Cummings

HankC said...

so far it seems that no 'very specific examples of problems' with the current theory are forthcoming on the gettysburg daily site...

John Cummings said...

Revisions have been posted that should provide a clearer understanding of my findings. More recent postings have addressed many of the questions presented by Garry Adelman and Tim Smith on the Gettysburg Daily site.