Skip to main content
The Collation

An Experiment in Following a Worm Through a Folded Letter

Folger staff have long been interested in folding early modern letters for mailing. It comes up periodically when someone finds a letter with unusual folds. Both Heather Wolfe and Erin Blake have written Collation posts about discovering how certain types of folds were done and why. However, a digital surrogate is not usually a source that leaps to mind as a tool when confronted with folding questions. Somewhat surprisingly, the Imaging department once assisted with a folding project: we digitized the front and back of an autograph letter by Lord Burghley (X.d.494), and those images were used to create a realistic, foldable simulacrum of that letter for teaching purposes.

Then, in the past few weeks, we got involved a second time in a new and different way when we helped explore a question about worms.

It began with an on-line discussion about the nomenclature of holes in manuscript letters (yes, holes in the paper) for use in metadata. This is generally as interesting as it sounds, but catalogers and other librarians are all about very accurate and consistent description, and for researchers it can be a life-saver. However, the discussion went on a tangent with an investigation of X.c.61 (22), a letter with several types of holes.

Figure 1: Francis Hay to David Rattray of Craighall, Dupplin, 1647 October 23 (Folger MS X.c.61 (22), part of the Papers of the Rattray family of Craighall)


In addition to a hole where the seal has been torn away and wear holes, it has worm holes, and the question was how the letter had been folded when the worm(s) made them. While we don’t know the exact number of worms involved, for simplicity’s sake we will assume that there was only one, very energetic, worm. From whence did the worm come? Where did it go? There are three sets of these holes in the letter: a large pair in the middle of the address leaf; another, smaller pair in the middle of the text leaf (the leaf containing the text of the letter); and one very small one at the top of the text leaf. How are these holes related? The Imaging staff decided we could look for answers using a digital image while working from home, without any access to the original item.

Step 1: Preliminary work

First, we made a rough model in the usual way, by folding a scratch piece of copy paper, using our knowledge of standard letter writing procedures and folding techniques. This allowed us to work out the folds with a pretty high degree of certainty, and confirmed how the letter had been sealed. It also allowed us to determine that at some point, the letter had been folded in half while it was completely folded for mailing. That extra fold goes vertically in between the two worm holes on the address leaf. It seems relatively unlikely that the worm damage occurred while the letter was folded in this way, since it would require the worm(s) to enter and exit from each side of the letter, because we know that it did not burrow through all the sheets in the middle. However, our piece of paper wasn’t an exact enough copy of the letter to help us any further with the problem.

Therefore, we duplicated the image (never mess around with the archived copies!) and began to play with it in Photoshop. First, we created a neon green layer below the letter. Then we cut out all the holes in the letter, so you could see through them onto the layer below, making any unobstructed hole glow neon green. Since we were focused on actual holes, we only realized slightly belatedly that the rectangular area to the left of the address square on the address leaf is actually a big hole along the edge, so we cut that out also. Apart from that rectangle, we didn’t cut out the white background around the outside edges of the letter.

Figure 2: A green background highlights the holes found in X.c.61 (22)

Step 2: Cutting up the digital letter

We knew the first fold in the sheet was made vertically, from top to bottom, down the middle. Most letters in this period were written on a sheet folded this way, with the text of the letter written on the right-hand leaf and the fold on the writer’s left side. Letters were then folded further for delivery and sealed in one of several ways that allowed the address to be written on the previously blank left-hand leaf, with the text of the letter inside, hidden from grime and gossips. To simulate this first fold, we cut each of the two leaves out of the image and put them on separate layers, so we could manipulate each any way we wished, and we could see through the holes where they lined up.

Figure 3: Cutting a letter in half is less scary when it’s digital! Here, the two halves of the letter are shown on separate layers for easier manipulation.

Step 3: ‘Folding’ the letter

Since the worm holes were largest on the address leaf, we ‘folded’ the text leaf behind address leaf. To do this, we flipped the text leaf horizontally (left and right edges reverse positions). This made the writing Leonardo-esque and on the wrong side of the leaf, but as a sheet of paper is essentially a 2-dimensional object, the edges and holes of the text leaf still looked like they would if the leaf simply had been folded around behind the address leaf.

Figure 4: If you thought early modern writing was hard to read normally, try it mirrored! Here, the right hand page of the letter has been flipped so the writing appears as a mirror image.


You can get the same effect if you write on a piece of paper with a dark magic marker that bleeds right through the paper, and then flip the sheet horizontally, especially if you hold it up to the light.

Then we made the address leaf 65% opaque so we could see the text leaf through it, and slid the text leaf behind the address leaf (for an image of a semi-opaque layer with an opaque layer behind it, cf. figure 7 below). It took a little massaging to find what looked like the right positions for the two leaves when folded. Then we could step back and see bright neon green where all the holes on the two sheets aligned. Then we made the address leaf fully opaque again.

Figure 5: With the two pages of the letter layered on top of one another, it is easy to see where the holes are.


Most of the holes are wear holes around the address square. This makes sense, as those folds would be the edges when the letter was sealed and sent, so those edges received the roughest treatment during transit. It was very nice, though, that those wear holes lined up! It gave us confidence.

Figure 6: Wear holes around the area where the seal would have been – all lined up neatly.

Step 4: The worm holes

The worm holes also looked good. For those who are cynical, yes, we looked at the worm holes when we were “massaging,” so it would be more accurate to say that we were surprised and gratified that we didn’t have to take any tortured, extraordinary measures to make the folds and the worm holes both line up well. In fact, we lined up the folds first, but when we shifted slightly for the holes, the position of the folds looked better than before. This makes more sense when you look closely at the folds. You can see in figure 4 that the initial cut along the first fold of the address leaf is not exactly straight (the white left edge of the text leaf next to it is straight). Over time, after repeated folding and use, the folds became less thin, sharp, and clear, introducing some play or ambiguity (wiggle room?) into the folds, and therefore less exactness in our reconstruction.

After one fold, the worm holes in the center of the two leaves lined up, like the wear holes. However, there was one more small hole in the top left corner of the flipped text leaf in figure 4. It is only found on the text leaf, which is behind the address leaf in figure 5. We hypothesized that the second major fold in the sheet was the top third down behind the middle third. If that was true, that hole should align with one of the large worm holes in the center third. Even though the top third of both sheets clearly would move together after the first fold, in our reconstruction we only “folded” the text leaf, so we could see the hole alignment more easily. We knew that the worm didn’t make it in as far as the top third of the address leaf, as there was no hole there.

Then we went to work again. We copied the top third of the text leaf, flipped it vertically (top and bottom edges reversing positions), made the full address leaf disappear and the neon green background disappear, made the text leaf 65% opaque, slid the copied top third behind the center third of the text leaf, and aligned the top edge of the center third along the proper fold in the text leaf.

Figure 7: How to fold the top third of a letter – digitally, that is.


After restoring the two full leaves and the neon green background to normal, the image had 3 letter leaves superimposed: the address leaf on top; then the flipped text leaf in the middle; and finally the copied, flipped top third of the text leaf on the bottom.

Figure 8: So many layers in the middle third of this image.


The orphaned almost-round hole on the top of the text leaf aligned beautifully with the largest worm hole in the center sections of both leaves.

Figure 9: It is clear that an industrious worm created a hole in the middle of this section, most (but not all!) of the way through the layers of paper.


Confirmation that the top third was the second fold, at least when the worm arrived, comes from the secretary’s filing note at the bottom of the right edge of the address leaf. If the bottom third was the third fold, then that writing would be on the outside on the top, ready for filing. If this was done, and a stack of letters was stored together, one would expect another letter, stored right behind this one, also to have worm damage: the worm ate along the path of least resistance between the two letters, dipping over into this letter and then back into the other letter several times, without ever going all the way through our letter. This would explain why most of the holes are long rather than round, and why two sets of holes are in pairs, for the worm meandered (wormed its way?) up and down and side to side as it ate.

It was interesting to see how much information about paper folding we could extract from a digital image using Photoshop, without risking any damage at all to the original item. If a suitable candidate for the adjacent letter is found, it might also be possible to use these methods to make an argument for the two being stored together. Such information is usually lost completely.


Fascinating post – but if I may be permitted one facetious question, is there a particular reason why you chose to mirror the page rather than simply using the image of the verso?

Elisabeth Chaghafi — August 5, 2021

Thanks very much! Great question; it has two facetious and two serious answers. To wit: (1) It didn’t occur to us (see the serious answers for why). (2) It was kind of cool. (3) The ‘simply’ goes the other way in Photoshop: it was easier to flip the recto image. In this case the verso image would be exactly the same size as the recto (very helpful, but not a deal-breaker, just a huge time saver). However, using it would require locating and downloading a duplicate of the verso image, cutting the page out, and pasting it into the recto image. On the other hand, to flip the text side of the recto image, we need only cut, copy, paste, and flip. It is quicker and easier to do, and in the end we arrive at the same outline of the page. (4) We weren’t thinking of doing a post at the time. We were just answering a question for Dr. Wolfe, so ease and speed to the answer could justifiably outweigh the elegance and simplicity of description helpful in a blog post. Then we were stuck.

William Davis — August 5, 2021

Most interesting- and funny. Thank you so much.

Angela Cockburn — August 5, 2021

Thanks for the explanation! Sorry, for some reason I read past the bit that explains you are a senior photography person (so this would probably be a good occasion to say ‘I love your work’).;) I naively assumed you had been using the images on LUNA like a normal end user. Fair enough, then. If you are ever in the mood for creating a 3D model of worm holes for a follow-up post, though, may I recommend W.a.311? That one has a worm gradually working its way through the entire book.

Elisabeth Chaghafi — August 6, 2021