I began my discoveries with an easy one...the Surveyor House, which still stands as the oldest structure in De Smet. It had been moved decades ago from its original location. Now it is lovingly maintained in a state much like it was when the Ingalls spent their first winter in Kingsbury County by the shores of Silver Lake. There is a simple guided tour of the first floor, but the upstairs is off limits. No photos are allowed to be taken, so I had to satisfy myself with pacing off the main room. It was 12 by 13 of my feet.

Having discovered paper models in 2004-2005, I thought this would make my first original creatiion. I found photos to give me near elevation views of the front and north side (it faces east now). Photos of an outside model of the house in its original configurations helped me find proportions and details to establish dimensions. I also wanted to have an interior. Those dimensions were deduced from the interior photos I found. I colored the structure as bare unpainted wood that was starting to gray with weathering. You cannot see it, but I used photos from around De Smet as views in the windows on the interior. Finally, I made quick models of the Ingalls girls from a family portrait of them and added my own children for comparison. Guessing the Ingalls' heights was just that. I think Laura's adult height was only 5'1", and she was only 12 or 13 during the "Long Winter" - not fully grown up.
It was a successful project that did give me the comparison result I was looking for. By the way, the model was made in 1/4 inch = 1 foot scale (1/48).

This birdseye view of De Smet in 1883 was drawn by Henry Wellge of Wisconsin. For a hand drawn illustration without a rigorously constructed perspective, it rectifies pretty well. BTW, the block of trees is not the cemetary, but a park! The scan was downloaded from Nancy Cleaveland's Pioneer Girl website. She has a wealth of collected and original research into all things Laura. It will be worth your time to browse thru her blog and pages - if she posts them again.

The rectified map corrected to the vertical frame the and horizontal baselines of the text. Then I "rotated" the image to look down on the town as though it were a map. One quickly discovers that this was not a tightly drawn perspective. I focused on the central four blocks and made them to be rectangular. I needed to bring the rotated image into Photoshop in order to disproportionally scale it so that the blocks would be square. Hugin made them taller than wide.
And that's how I got to the final result.
And when I took the look-down image and rotated and skewed it so Calumet Ave. was horizontal, I got the storefronts on the west side of Calumet (vertical scaling was arbitrary.)

This was just a "let's see if..." exercise. Nothing was really discovered by it. Maybe you, the reader, can find some utility from it.

ROUND ONE

One of the biggest challenges to my imagining Laura's experience is Pa's store in De Smet. They lived thru a brutal winter in it their second year in the De Smet area, in 1881. I imagine Pa's store to be... well... store-size. But how big was it really?

It turns out there are two public domain photos of it taken in 1912 as seen in two panoramaic photographs. These can be found at American Memory here and here. You can find these there by going to the panoramic photos section and searching on the photographer's last name, "McKibben" or on "South Dakota." The first shows a 3/4 front view of the building and the second shows a nearly complete side elevation of it. It had been moved from its original location facing Calumet Ave. to the back of the lot when the First National Bank of De Smet building was constructed in 1885. That building today houses the Gass Law firm. There also exists a number of private photos of it and one that can be purchased from the South Dakota State Historical Society. One researcher graciously took interest in what I was doing and shared a few of those private ones.

My procedure will be to compare Pa's store in the photos, whose dimensions are unknown, to the bank building/Gass Law firm building, whose dimensions can be known. So a good stereographic reconstruction of the bank would also yield a reconstruction of Pa's store and provide dimensions by comparison. Because a rectified image will not necessarily be in correct proportion width to height when using Hugin (in my experience), I will need to compare the 1912 photos to my own modern ones. I will bring my photos of the bank/Gass Law building taken when we visited in 2003 into Hugin to get a present day comparison of the building. Then I will "Huginize" the bank building taken in 1912 to match the modern views. I will then be able to use the same Hugin file to make elevations of Pa's store. So goes the theory, but measurements trump theory.
I did this by hand and came to the rough conclusion that Pa's store was about as big as the Surveyor House! In other words, it was small. Pa may well have used the Surveyor House as his model when he built the store. . . maybe. I wanted to order up large files of the panoramas and Huginize them to get better detail thru increased resolution, so I contacted American Memory, but what is on line is all one can get digitally.

ROUND TWO

A 2nd round of Huginizing produced the bottom row. What really helped was having the basic aspect ratio of the floorplan - or in this case the roof plan. I found that on Mapquest, and established the front/side ratio to be 1:1.73-ish ("-ish" because of not very great resolution). With that in hand, I was finally forced to admit that my belief about rotating the image in Hugin until I had two rectilinear side views, which would be in the same proportion to each other, was FALSE! I don't know why. It should be doable to my way of understanding. But it simply never worked out that way. But now I had some confidence about how to disportionally scale the two sides. I scaled the Calument (front) side to my own photo I took face on. I then scaled the 2nd Street (north) side in height to the front, and in width to my aspect ratio in which the front equals the "1" of 1:1.73. You get the idea. Then I scaled the side of Pa's store from the front view to match the height of his store in the side view, and violá, I had his store to scale to the Gass Law building today. Its aspect ration was very close to 1:1 for the front, and 1:1.5 front width to depth. The second story was barely tall enough to stand up in. The two outside corners of the store front 2nd story windows HAD to be touching the rafters! The peak of the roof barely comes up to the top of the 1st story windows of the Gass Law building. If we assume the front door is 6-1/2 feet tall, then that height would be just over 14 feet tall!! Yup. It's small.

ROUND THREE, FOUR, FIVE, AND SIX - KO!

ROUND 3 - This is the reason I have not posted in over a year - I have been trying to arm-wrestle this to the ground. I shared my findings from an 1883 so-called earliest photo with LIW researcher, Nancy Cleaveland, and it resulted in an excellent critique of my conclusion. I had identified one building in the photos as being Pa's store, a conclusion reinforced by one of the cartouches in the "Little Town on the Prairie" book. So back I went to REALLY work over the so-called earliest photo that I found on the web. Better files were shared, but cannot be reproduced here cuz of copyright permissions. Sorry about that, but it a nuisance worth protecting... especially some day when it is one of your OWN images! Long story short, she is right! It was NOT Pa's store, but what was later the meat market three doors down. You can see it in the 1912 panorama. But wait... after trial upon trial, I got a good elevation of the entire two blocks of Calumet from the earliest photo, and lo and behold (drum roll)... PA'S STORE IS NOT THERE!!! Given all the buildings one CAN see, it is our opinion that this photo would be from some time in the late spring/early summer of 1885. Pa's store was moved to the back of the lot to make way for a new brick bank building. The bank was completed later in the fall of '86. So, what we have is a photo of Pa's store and the new bank, neither of which is visible. Some days ya feel like Thomas Edison waiting for the light bulb to turn on.

ROUND 4 - Back to the drawing board, or, to the McKibben 1912 panoramas. Unlike a regular photo, the panoramas seem to have a disproportional stretching. By that I mean that when I rectify them and look at the entire block, the width of each lot increases with distance from the camera. The more distant ones are wider than the nearer ones. It should not do that, and does not in a well-done rectified photo. I tried determining the camera position and ended up wandering all over the intersection. I tried looking down on the intersection, but even knowing its dimensions (which I learned much later - 80x100 feet building corner to building corner), I could not get it level and rectangular. If you study the panorama you will notice something is off-quilter. The left and right ends of the photo, which are the street corners just across from each other, have a big elevation difference! For a rotating camera, it will (should) return to the same position. In other words, the greatest elevation difference should NOT be at the opposite ends of the panorama. The right end should be approaching the same elevation as the left end. Note that this was not a complete 360 degree panorama, but it was close - I guess between 310-330 degrees. But if one crops the panorama to about a 100 degree field of view (this encompasses over one block), one can get a fairly well rectified image. So I settled for that.

ROUND 5 - I next started working on some loaned images (not shown - copyrighted). But again, lacking known dimensions stymied my ability to know if my width and depth proportions were correct. But I COULD get details about the placement of windows and such, so they proved useful. Then, I was provided with measurements of the Gass Law Office building, so now I could check proportions over an entire block. I also had access to a Sanborn Insurance map (now available on American Memory!), which showed the locations and dimensions and heights of all the buildings in De Smet. Big Help! And Google and Bing now had better resolution aerial views of De Smet today (remember that roof proportion?). Things were coming together and I could make an educated guess to the basic dimensions of the building - safely to within inches I hoped.

ROUND 6 - But the clincher was a new direction I explored to verify my dimensions. I don't know why I did not think of it before, but I realized the best way to verify anything was to rebuild it! And so I did. Digitally! In effect, I created a blueprint of the structure. You could draft it or use a good drawing program or a simple CAD program. But how did they build structures back then in the 1880s?

It turns out they used a technique called "balloon framing", which had been recently developed around 1860. It is closely related to our modern platform framing, which is really just an improvement of it. It used 2x4 lumber and 16 inch on-center stud spacing. The studs though were full height single boards. It was a dramatic breakthrough in construction, producing a sturdy but light structure. More importantly, all one needed were basic carpentry tools and a helper! It made possible the overnight building of prairie towns along the railroad lines.

AND THE KNOCK-OUT! - And so I began, using my derived dimensions as my rough guide, and, EUREKA! It worked beautifully! The structure I came up with matched well my derived dimensions and the photos. The roof angle worked (30 degrees) and even the lap board siding (5 inch boards with an exposure of 4-1/2 inches). The door sill height indicated a floor on a 2x8 joist. The top windows extend down nearly to the attic floor, permitting an 8-foot main floor ceiling height to the bottom of 2x8 attic joists. I came across one final confirmation while searching for more birdseye maps by Henry Wellge, who created the De Smet 1883 map. He did many, which may be found at American Memory, but the one for Redfield, Dakota (now South Dakota) had a print shop in a corner cartouche. It looked exactly like Pa's store! Apparently Charles Ingalls built to a standard model. And so we had it cornered and it was just a matter of finishing it off.

There are still some assumptions and unknowns, of course. Reading thru Laura Ingalls Wilder's writings yields clues. From them we learn that:
• the store was left unpainted, inside and out
• It was ceiled and sided inside and out
• It had a root cellar somewhere under it
• The attic trap door was where the back door was, because the ladder to it was by the door
• The front room was large and the back room was small - very small

But just how small is unknown and can only be surmised from how it was furnished. That, and the fact that eight people lived in it by day. Yes, eight! It turns out that the Ingalls boarded (hosted more realistically - their guests never paid more than a token amount) a young couple they knew from Walnut Grove, plus their newborn who was birthed there that winter.

Then, there is the lean-to shed on the back. Wellge's drawing shows one less than half the width of the back of the store. But the 1912 photos show it now on the side and about 2/3 the width of the back. But look again and one sees a difference in texture or surface shading dividing it evenly in the middle, indicating that maybe, just maybe, it was expanded when the store was moved! That affects the back room layout significantly. The back door would now be well off center, at the corner even, and it would move the trap door closer to the center of the room. A project I might pursue later is trying different arrangements.

It was a good project because it finally came to a conclusion. And I learned to use Hugin much better and quicker to pull elevation views out of photos. And, I think I have made a small but real contribution to the Little House knowledge base.

REFERENCES

"De Smet, South Dakota, corner of 2nd and Calumet Ave."
J174028 U.S. Copyright Office / Copyright deposit; C. S. McKibben; October 10, 1912; no known restrictions on publication.
http://www.loc.gov/pictures/collection/pan/item/2007662776/

"View of Desmet [sic], S.D. from Court House"
J174029 U.S. Copyright Office / Copyright deposit; C. S. McKibben; October 10, 1912; DLC/PP-1912:44504; no known restrictions on publication.
http://www.loc.gov/pictures/collection/pan/item/2007662798/

"Ten men building a wood frame house, Omaha Reservation, Nebraska, 1877"
Photo by William Henry Jackson; The National Anthropological Archives, Smithsonian Institution
http://www.nationalbuildingmusuem.net/Large%20Images/house--home/05_Balloon-Framing.jpg

"1883 Bird's Eye View of Redfield, Dak." Drawn by H. Wellge. Copyrighted by H. G. Rising
Digital Id: G4184R Pm008910 Http://Hdl.Loc.Gov/Loc.Gmd/G4184R.Pm008910
http://memory.loc.gov/ammem/pmhtml/panhome.html

This was cited as the earliest known photo of De Smet in Aubrey Sherwood's "Beginnings of De Smet" and likely was taken by Mr. H.W. Cooledge, one of the town's first photographers. But is it really? How can we tell? Oh! I know... I will look for Pa's Store. It should be easy to see and was there from April of 1880, the year the town was settled. But for the life of me, I could see other buildings I recognize, but not Pa's store! It simply was not there! How could that be?
I tried doing perspective correction on a photo of this photo three years ago. The photo was taken not only from an off-center angle, but of a curved photo print. But the left half where Pa's store should be was pretty flat and so I was confident of my correction of that part. And sure enough it, too, showed an empty space where Pa's store should be. But I could not be absolutely sure.
So in 2015 I attended the "Laurapalooza" in Brookings. I took a half day to drop in on the Laura Ingalls Society in De Smet and met with the director, who graciously allowed me to look at some original photos in their collection. The earliest photo was a top item of my interest. And she even more graciously allowed me to scan the photo! And I did.
Returning home, I eventually started working on my high-res scan doing perspective correction. And still I found an empty lot where Pa's store should be. I did a back-and-forth dialogue with my go-to researcher on the matter who was still not sure my technique was working. So I reverted to a mechanical perspective analysis. I found the vanishing point and the perspective rays along the street and storefronts. I then tediously established each of the 28 lots along the two blocks shown in the photo. I then tinted each one to stand out and then ran the photo thru the Hugin perspective correction technique. And sure enough, once and for all, demonstrated that there is NO STORE where Pa's store should be.
So how could that be? What does that mean? Only one thing I could think of: the store had been moved. That event happened just before the new bank building construction was begun. But there is also definitely NO BANK in the photo either! Ahah! There was only one brief period of time when neither building occupied that location - after Pa's store was moved and before construction was begun on the bank. And when was that? In the summer of 1885.
So now we know when this so-called earliest known photo was taken. And from the location of the camera, it was taken from Cooledge's shop.
BUT WAIT!... about a year before this, my go-to researcher had established that another photo I was working on was taken maybe even earlier! Yes - even earlier!! That photo was of a celebration of some sort and the people were in strange-faced costumes. It turned out that Pa's store was hidden just behind the crowd, but one other detail showed up on close examination. Are you ready for this....? It showed the upper story of Pa's very first store!!! And when was the photo taken? Well, the costumes are typical of those worn by the Odd Fellows, which held gatherings in De Smet in 1883 and 1884. And Pa's first store was moved in April 1886 to make way for the Couse Opera House completed in October of that year. So the photo was taken some time between 1883 and April 1886.
What is most interesting about it is that is looks exactly like his more famous store from "The Long Winter"! Well, that does make sense. They were built in the same year, so why not make them the same? There was no reason to make them different. He could order the same amount of materials. And it would be easier to make two from one plan rather than two from different plans. Surely there must be photos of the first store after Couse moved it before building his opera house!

BUT WAIT! I have heard told that there were photos taken on July 4, 1883! I would love to see those!

While visiting the Ingalls Homestead in De Smet in 2015, I asked if there was a building I could make a model of for them. "Ma's House!" was the enthusiastic answer. There are, of course, no photos of the building. All we have to go on are other claim shanties as well as the dimensions listed on Charles Ingalls' claim proof document. The Ingalls Homestead worked off of those plus Laura Ingalls Wilder's description of how the house looked and was added to.
So I headed down to their recreation of it and took some documentation photos of the building inside and out. Things are hopping there, so it took two years to complete it and get the final approval, but you can buy it there now.

Instead of relying on Hugin to get my orthographic views, I loaded my photos into a Photomodeler project and began marking points. The exterior plotted nicely, but I needed more interior photos to get the same accuracy. Still, I had enough to do what I needed to make the model. From Photomodeler, I export the different orthographic views as pdfs. Because the artwork in the pdf is in vector format, I can open the pdf in an illustration program like Illustrator or CorelDraw or Inkscape and begin creating my line drawings. Once I have all my walls detailed and sized together, I can then lay them out for cutting and folding onto 8-1/2" x 11" sheets. The final model closely matches the building constructed at the Ingalls Homestead . . . but you will have to visit them to confirm it for yourself!