OK. Here’s the post folks have asked for. Warning. It’s long.
I don’t claim this to be totally inclusive (I’m always stumbling across new-to-me things as I browse museum on-line photo collections), but it’s a start. Feel free to comment with additional examples.
There’s been lively discussion on what stitches and techniques were used for the backgrounds of voided works. I’m going to try to present as many examples as I can.
To start – voided pieces are a family of works that feature a more or less uniform background treatment that leaves the main design of the piece plain (or minimally worked. It results in a visual “reverse silhouette” look. There are many manifestations of this aesthetic over time. One widely known subset is Assisi work – a simplified but charming 19th century revival inspired by earlier Renaissance era embroideries. The revival used cross stitch (aka “plain old cross stitch” or POCS) ornamented by back or double running stitches. Earlier styles were more varied.
One of the most common treatments was a tightly pulled four-sided stitch, worked to completely cover the threads of the woven ground. None of the ground threads were cut – they were just bundled together, making an extremely durable net-like texture. How do I know it’s durable? I’ve stitched some, made a mistake, and found it absolutely impossible to rip back or deconstruct (perhaps that’s why so many fragments of it exist, even after the towels, pillowcases and other linen they adorned have frayed to death).
The border above is in the Art Institute of Chicago (accession 1896.112, and is attributed to Italy, in the early 1600s. I believe the outlines were established first, in either double running or back stitch, and then the background was filled in, working right up to and in some cases, encroaching on those outlines. Close examination of the photo where the outlines are broken shows no cut ground threads, just distortion. The “wing shapes” in the connecting meandering branches are very amusing to me. I know from experience that working in closed areas is challenging. it looks like the stitcher saved some time and effort by drawing a diagonal between the bud and the side sprig on the branch, and just not filling in between them.
Here’s another example, Italian, but undated, resident in the Harvard Art Museum collection (accession 1916.388). Also outlined with the meshy stitch worked up to the outlines. Note that companion edging though. I can’t tell for sure, but the branches that little leaves grow on at least may be cross stitches. Not sure about the leaves themselves. On this one it’s very clear that the ground cloth threads are bundled, not cut.
Here is a variant – a similar tightly stitched mesh, over a somewhat coarser linen ground, BUT in this case the stitcher did NOT establish an outline and then fill in the background. The piece is most definitely done on the count (not on a freehand outline), but the only stitching that established the motifs is the background mesh. This bit is also from the Cooper Hewitt (accession 1946-42-9a), dated 17th century, but has no posted place of origin. One other thing to note is a bit of directionality in the mesh. Mesh can be worked either on the diagonal or back and forth across succeeding rows. In this case the stitcher did the latter. But it’s NOT long armed cross stitch. It’s still the tightly overworked mesh.
This variant of meshy was done by someone who didn’t encroach on the established outlines. Instead this stitcher left a “halo” of unworked ground around the foreground motifs. There is no companion line on the outer edge of the halo area – the mesh stitch simply starts. I’ve mentioned this piece before in my series on long-lost siblings, and it’s in the Harvard Art Museum (accession 1916.377), but bears no date or location notes.
Here’s a piece that the holding institution claims was done by withdrawing threads, but the detail photo at left (a section where the red stitching has been lost) clearly shows the distortion of groups of 3×3 threads, with no snips or darns. I maintain that this is the pulled meshy stitch, too. Another Cooper-Hewitt sample (accession 1971-50-90), Italian from the 1500s. Love that needle lace edging detail, too!
Cut and Withdrawn/Overstitched Mesh
What about withdrawn thread work, where threads are snipped or turned back and the edges secured, with the remaining scaffolding overstitched to make a meshy background? I’m pretty sure it exists, but I need to find a well documented and clearly photographed sample that explicitly shows the snipped rather than distorted threads of the ground fabric’s weave. Have a reference? Feel free to share it in the comments. If a good one shows up I’ll edit this and include a cut thread heading and photo here.
Long Armed Cross Stitch (LACS)
Another popular ground treatment was long-armed cross stitch. This produces a distinctive almost braided texture when worked back and forth across the piece. The piece below is in the Cooper-Hewitt (accession 1971-50-100 ), with a provenance of Spain, of the 16th-17th century. Again the main design is outlined with back or double running stitch, and the background is filled in later. Note that the stitcher kludged this a bit where the rows of LACS meet up with angles, and that POCS is used for edge ornamentation.
But again, working with an linear outline is not mandatory. Here’s a jaunty falconer on his mount. He is also worked in LACS, but without the double running or back stitch outline, in spite of the complexity of the design. And yes, there ARE some plain old cross stitch bits in there. Much of the surface detail in the otherwise unworked foreground areas are done in POCS. I’d even entertain an argument that outlining was also done in POCS, but is mostly disguised by encroachment of the background LACS. However, the bulk of the background is clearly LACS. You can find this piece in the Cooper-Hewitt (accession 1904-17-4), dated to the 17th century, no provenance. I do wonder about the dating though. The design seems a bit “modern-revival” to me, unless there was a nostalgia movement in the 1600s that presented folk in “antique dress.” Also that cross stitch for outlining thing is very, very rare. (I’ll wait for the experts on dating to chime in on this one.)
More long-armed cross stitch – but more tightly pulled. It’s not true meshy – the plaited like texture and 1×2 crossings are still evident. This time with outlines. In green. This Italian piece is from The Art Institute of Chicago (accession 1937.779), and is dated from 1500s/1600s or so.
Another one just for fun. Clearly LACS-like, and you can make out that 1×2 cross on the very uniform top legs. From the uniformity of those legs I think that this piece was not worked in stitch-by-stitch mode (the standard way of working LACS, but as an entire row, with the stitcher first laying down the “short legs” and then covering them by a second pass working just the “long legs” in the opposite direction. This supposition is borne out by the way the successive rows cross. Note that there has been absolutely no effort to keep the successive rows of LACS either alternating left to right as is done when it’s worked in the usual manner, or all aligning in the same direction. Instead the rows “bounce” when they encounter an obstruction, and do so in a way that’s congruent with the in-two-passes approach. Obviously this one has outlining done in a different color, and the ground done in a very atypical yellow. Sprightly, even with the massive loss of the now blue/green thread. It’s from the Cooper-Hewitt collection (accession 1971-50-77), and dated to the 1500s (no provenance.)
There are a few pieces that use an effective but simple fill. The final appearance is that of boxes. The samples I have seen have all been double-sided, and from the pattern produced by unevenly dyed or faded threads, I suspect most of them were worked in double running on the diagonal. No proof though without picking one out, and that would be heresy. This piece is from the Philadelphia Museum (accession 1894-30-116). It’s Italian, of the late 1500s. In addition to the boxed fill the foreground is ornamented with cutwork, which makes it a double-curiosity. On some of these the outlines of the motifs are also done in double running. In others, in back stitch (or possibly very neatly done outline/stem stitch), so that the reverse presents a heavier line defining them. Whether or not those who first used these considered the heavier outlined side the public side is something we may never know.
Here’s the most well known sample of the boxed substyle – the Metropolitan Museum of Art’s apron (accession 38.19.8) – Italian, 16th-17th century. This one doesn’t use outlines to define the motifs. The edges of the box ground units themselves define the edges of the foreground motif.
Here’s another example of the squared filling style (with outlines). This piece is from Musées Royaux d’Art et d’Histoire, in Brussels (accession T.1578), and is dated to the 1500s. It combines variants of two of my favorite designs, the “lettuce” pattern on the left, and another that shows up again and again on the right. Both of these designs turn up in other voided and un-voided presentations, with meshy or LACS as the ground treatment. Or none at all. Variants of these two will be in my ever forthcoming book.
Plain Old Cross Stitch (POCS)
Yup. You had to peek to see what I would say here. Sadly, although I’ve examined hundreds of samples of voided pieces, I have found none with a ground worked in plain cross stitch until the mid/late 19th century revival of that style. That doesn’t mean there aren’t any – just that I haven’t stumbled across them yet. Got one? Feel free to send the reference to me. I’d love to find one and add it to the greater family.
But here’s a prime example of the most complex end of the revived style. These two designs have clear Renaissance era precursors (well, close at least – maybe not exact pedigrees), but are rendered using POCS, with and without linear outlines. This is from The Antique Pattern Library’s copy of Album des Broideries au Point de Croix, compiled by Therese de Dillmont, probably an edition of the 1880s,
Other Modern Treatments
I can cite no historical precedent for these treatments – I admit, I was just riffing on the squared box theme. But they do work and are interesting. These are my own: diagonals, diamonds, and steps. I like the mirroring on the diagonals in the top sample, the second one has all of the diagonals going in the same direction for the entire strip. All of these are worked on designs for which I have citations, and that have or will appear in my books.
Voided work is a catch-all term for a family of embroideries where the background is covered by stitching, and voids in that solid stitching make up the motifs (the foreground). Sometimes the foreground is further ornamented by additional stitching, sometimes not. There are many different styles of this work, lots of posited points of origin/provenance, and just as many design or pattern groupings that have come and gone in and out of style over the centuries that voided work has been done. While modern Assisi (simplified motifs done with cross-stitch backgrounds) is the form of voided work most widely known today, it’s not the only type, and there is a lot to explore in the allied family of voided styles.
Here’s one subgroup – Story Panels. This is a family of works that I’ve run across as I’ve researched counted voided styles, that hangs together as a subset based on a number of commonalities.
First, the examples:
l. From the Cooper-Hewitt collection, Band. Italy, 16th–18th century; silk on linen; H x W: 24.1 x 172.1 cm (9 1/2 x 67 3/4 in.); Gift of Richard C. Greenleaf; 1954-167-5. These four panels show elements of the Adam and Eve story, and the workaday life after Eden . It’s done in red silk on linen, with a densely overworked meshy background. I don’t necessarily agree that it’s long armed cross stitch – that has a different look of directionality. This has more of a meshy appearance. Foregrounds are outlined (back stitch according to the listing), and ornamented by knot stitches.
2. From the Art Institute of Chicago, Fragment (from a border), Italy, 1575-1625, silk on linen. 22.8 x 41.4 cm (9 x 16 3/8 in.); Art Institute of Chicago Purchase Fund; 1907.827 Part of the story of Noah. Outlined foreground elements with spot decoration, ground in long armed cross stitch aka LACS (that back and forth almost plaited looking directonality is evident.)
3. Another from the Art Institute of Chicago. Fragment (from a border) Italy, 1575-1625, silk on linen, 19 x 40.6 cm (7 1/2 x 16 in.); Art Institute of Chicago Purchase Fund; 1907.826. Joseph and his brothers. This may or may not be part of the same original (or series of originals) as #2, above. Similar color, and LACS technique, but the heights are different, and the motifs are simpler in this one – less ornamented, less detailed.
4. And also from The Art Institute of Chicago, Fragment (from a border), Italy, 1575-1625, Linen with silk. 276 x 44.2 cm (10 7/8 x 17 3/8 in.), Art Institute of Chicago Purchase Fund; 1907.825. To my eye based on these photos, it looks like this panel (Joseph and Potiphar’s wife?) is done the same way as #4, above.
5. From the Cleveland Art Museum, Embroidered Border: The Baking of Unleavened Bread, Italy 16th-17th century. Silk on linen. 18.1×45.4cm (7 1/8 x 17 7/8 in.) Gift of the Textile Arts Club; 1939.354. From Probably LACS (no prominent holes like the meshy style). Foreground lightly outlined with what looks to be a thinner thread, foreground details in back or double running. No knot stitches. I’ve discussed the group of four panels from which this comes once before.
6. From the Metropolitan Museum of Art, Strip, Italy, 16th century, silk on linen, 9 1/8 x 25 in (23.2 x 63.5cm), Gift of Mrs. Harry Ge Friedman; 48.57. I’m guessing from the inscription that this is part of the Joseph in Egypt narrative, where he has dealings with his half-brother Simon. Again, probably long armed cross stitch, with either double running or back stitch outlines of the voids.
7. Also from the Metropolitan Museum of Art, Panel with Noah’s Ark. Italy, late 16th/early 17th century. Silk on linen. 14 1/4 x 39 1/4 in. (36.2 x 99.7 cm) with lace. Robert Lehman Collection, 1975.11784. Another Noah’s Ark. Note that the base drawing of the ark section is very, very close to the green one, although the follow on panel is different.
8. Still more. From the Cooper-Hewitt, Band, Italy, Late 16th, early 17th century, linen, silk; H x W: 150 x 19 cm (59 1/16 x 7 1/2 in.); 1950-29-8. The center panel is probably David avoiding Saul’s spear, but the rest of the iconography is hazy and there’s no top line inscription to help. Very clearly long armed cross stitch, possibly double running on the outlines (there are also a few later repairs done using another color, to reunite the stitched ground with the open foreground but that doesn’t count).
9. From the Cooper Hewitt, Band, Italy. Late 16th, early 17th century. silk embroidery on linen foundation; H x W: 23.5 x 60.6 cm (9 1/4 x 23 7/8 in.); Bequest of Richard Cranch Greenleaf in memory of his mother, Adeline Emma Greenleaf; 1962-52-1. This one doesn’t have lettering at the top, but it’s clearly the story of Isaac. LACS, outlined foreground, some ornamentation of foreground with straight stitches (possibly double running or back stitch).
10. And finally, from the Yale University Art Gallery, Unknown Fragment, Italy, 17th century. Linen ground with red silk, 19.1 x 43.2 cm (7.5 x 17 inches), Gift of Howard L. Goodhart; 1928.151. Very had to tell from the photo but it’s probably LACS, with double running or back stitch for the lines. This bit is probably Jonah and the whale, and is clearly part of a multi-panel piece (or once was).
Now, I am sure there are lots more of these out there, that I haven’t included here. And there are narrative panels done in other stitched styles, but these do seem to hang together, more or less.
First, unlike most (but not all) other voided work examples, they display no symmetry. There are no reflection or bounce points; the designs are not aligned in balance around center urns, trees, or other elements. Each one of these panels stands alone, without a clear repeat inside its sequence.
Second, most (but not all of them) rely on similar framing techniques – a narrative with a very similar looking style of letter representation on top, and the curious mix of birds, dogs, and leaf/branch/flower elements below (which does repeat).
Third, none of these were done on the count. By that I mean that the foreground elements were not carefully copied from a graphed source. They incorporate strange angles and curves, and the ground stitching behind them – which was done on the count – looks to have been “mashed in” around the designs where they present those odd curves and angles.
I posit that these were hand drawn onto the cloth, overstitched using double running or back stitch (or possibly even SINGLE running in some cases); the foreground ornament was done, and then the backgrounds were stitched, in neat lines going back and forth across the cloth. BUT it’s pretty clear that some sort of common cartoon (in the tapestry sense) was used for the two Noah’s Ark panels – #2 and #7. Same ship, same placement of birds, bit players and leaves around it, but with just enough difference of detail and odd angles to look like tracings from the same original, not copies of the same chart.
Fourth, for #1-8 above, there are clear divisions into panels, with strangely familiar fat-fruited, full-leaved vegetation or ruled dividers separating the scenes.
Fifth, all appear to be Old Testament scenes. Given the time and place, it’s kind of strange that no large scale New Testament scenes are included. Now those may exist elsewhere, I don’t claim to have gathered a definitive collection of these fragments, but one would think that there would be a Last Supper, Passion or stray Saint among the lot. The closest we get is the Agnus Dei (lamb with cross standard) in the Jonah panel – #10, and even that is background – not the “featured scene.” It’s also worth noting that even with the popularity of Greek/Roman myth images at the time (just look at emblem books and early pattern books) – we’ve got no Aphrodites, Sieges of Troy, or other mythic representations.
Now, what conclusions can we draw from all this? Sadly very few without further research.
Who made these and why? I am tempted to say there was a small number of professional ateliers producing these in late 16th century Italy, due to the strong similarities of style, and the fact that these examples are relatively few among the large number of other voided work fragments we have today. Given the elaborate nature of the non-repeats and the scale of these sequential multi-panel narratives, I somehow doubt that these were loving-hands-at-home works created for household use.
Many of those other bits are probably domestic works – with designs that are symmetrical, with clear easy to replicate repeats. While it’s certainly possible that these panels were bed or other secular hangings, but I think it is more likely they were made for liturgical/didactic use.
And #9 and #10 – the odd outliers? I think they were clearly influenced by the group as a whole, but given the difference in their visual styles and details, I would not be surprised to find out they were done a bit later – or possibly even by competing contemporary workshops – in emulation of the established style.
Have you found other examples of these stitched comic books (biblical or not)? Share!
Readers have most graciously pointed out additional examples! Thank you – keep them coming
11. Holly found this in The Jewish Museum in New York, Embroidered Panel: The Story of David and Bathsheba. Greece, 19th century. Silk on linen. 10.5 x 29 inches (26.7 x 73.7cm). From the H. Ephriam and Mordecai Benguiat Family Collection, Accession S 202. The date and provenance are different from the rest, but it does appear to have some stylistic commonality with #8, above.
12. Melinda Sherbring alerts us to a holding in the Los Angeles Museum of Art, Embroidered Textile Panel Depicting Scenes from Genesis. Iberian Peninsula (Spain or Portugal), late 16th century. Linen plain weave with silk embroidery. (a): 9 7/8 × 64 1/4 in. (25.08 × 163.2 cm); (b): 35 3/4 × 9 5/8 in. (90.81 × 24.45 cm) Costume Council Fund (M.87.230a-b) . Sadly, there is no shared image available there, but from her detailed descriptions, it’s another version of the Adam and Eve panel (second panel in #1, above), and the Ark panel (#2 and 7 above), done in long armed cross stitch, in red silk. The foreground ornament of both is a bit simplified compared to the other versions posted here.
Melinda and her co-conspirator in textile history high-jinks, Robin Berry, had the opportunity to examine the piece up close. They have given me permission to share their notes on technique:
- Fabric thread count approximately 96 tpi.
- Embroidery floss is filament silk, finer than a single strand of Eterna; possibly Kreinik size 0.
- Motif colors: background color card 19-12 and 19-11 for Genesis, approximately DMC 3687.
- Technique: long armed cross stitch background with backstitch for details and outlines. Looks like the same thread was used for background and for details. Stitches over 3 threads, approximately 18 stitches per inch.
- There are holes along the edges clearly where fabric was nailed or tacked to a support.
Robin additionally points out that voided works with Iberian origins are properly termed “Reserve.”
Melinda agrees with me that the base layout of these pieces were probably traced or drawn rather than established by count. Having three examples of such a work is quite special.
Sorry folks. This has nothing to do with anyone’s search for companionship. Be warned, it’s all about embroidery, and this is a post that only a stitching geek will love.
As I fill out the last few pages of The Second Carolingian Modelbook, I’ve decided to take a stab at a design that seems to be everywhere. Except modelbooks, that is. I call it “Pelican with Harpies and an Urn.” It is one of a set of patterns that crops up again and again in museum holdings worldwide, most often as a fragment. It’s clear that unlike many other snippets, these all came from different works, often executed in different styles or stitching media. I’ve posted about this before, but my collection of examples continues to grow, and with it, the general confusion level.
The dilemma comes in because (to my knowledge) there is no existing printed pattern to establish a point of temporal or geographic origin. But there are lots of examples and they all express the details of the design slightly differently. Now if there was an authoritative point source that became unavailable, one would expect later iterations to be less detailed, or details to become blurred, through succeeding generations of copyist errors. We can see that with the oft-studied “boxers” sampler motif in Colonial American samplers – which probably started out as a cherub bearing a flower, but over time became less specific and more stylized, until what remained was a barely discernable chubby humanoid with a club fist. But I can’t arrange the Pelican/Harpy/Urn designs in an ironclad continuum of graduated detail.
Here’s the parade. The thumbnails are not clickable, please visit the links to see the museums’ higher resolution images.
CH-1. First is this example from the Cooper-Hewitt’s collection (Accession 1931-66-144). They date it as being a 17th century work, but do not offer a provenance. It’s done in silk on linen, with a characteristic tightly drawn background that produces the appearance of mesh, but does not involve withdrawn threads. Details are rendered in straight stitches, and may include double running or back stitch (it’s hard to tell without seeing the reverse). The museum acquired it in 1931, as a gift from Sarah Cooper Hewitt.
CH-2. The Cooper-Hewitt has another example (Accession 1931-66-142). This one is specifically called out as being Italian, and is also dated to the 17th century. It’s a particularly prime piece because it is a full span cut across the end of the towel, cover or cloth it came from. We see the orientation, the top and bottom borders, and how the slightly different side borders framed the work. The museum acquired it in 1931, also as a gift from Sarah Cooper Hewitt.
HERM-1. The Hermitage Museum has two examples. This one is entitled “Valence Embroidered with a Grotesque Motif (fragment),” but the on line page has no accession number. The full description calls out the linear stitching as being double running (Holbein), and the background as being drawn thread. They attribute it to Italy, and the 16th century. The museum got this piece in 1923, via the Stieglitz School, and ultimately from D. Flandin, an antiquarian dealer in Paris.
MET-1. Yet another example in the same style. This one is from the Metropolitan Museum of Art (Accession 14.134.16a). The MET cites it as being Italian, and 17th century. Although this one is at a different museum, and is clearly not a separate piece of either artifact, there’s a connection with the two above. It was acquired in 1914, via the Frederick C. Hewitt Fund.
HERM-2. On to another stitching style. “Valence Embroidered with a Grotesque Motif” from the Hermitage also has no listed accession number. This piece is lacis (darned filet net). It’s dated 16th century and placed in Italy. Although filet work doesn’t allow for the linear details of the red examples above, it’s amazing how much fidelity to the design can be included. Like the other Hermitage piece, it entered their collection via the Stieglitz School Museum in 1923, but came from the collection of J. Kraut, in Frankfurt-am-Main.
MET-2. More stitched net, and not another piece of the one above. This one is also from the MET (Accession 06.582). It’s cited as being Spanish, from the 17th century. This piece was acquired in 1906, via the Rogers Fund.
First of all, I’d agree that the source for these was probably Italian, regardless of where the final objects were collected from. I’d also agree that very late 16th century, but more probably the early 17th century is reasonable for the whole pattern family based on the style, usages, media, and iconography, plus parallels to other contemporary designs.
CH-1 presents the most detailed urn and pelican of the set. Both are encrusted with small linear features, although the placement of those features is not always symmetrical, nor is it identical from repeat to repeat. Feathers on the harpy’s body are shown in neat rows, but her wing feathers are very stylized, using right angles rather than diagonals. I’m unsure what she’s holding – a cup or panpipes (perhaps a fancy on my part, to think of that flower as the music of the pipes). From the patterning, it’s clear that the thing across her middle is her tail, wrapped up from between her legs.
It’s hard to see clearly, but there are lots of differences between the urns and pelicans in CH-1 and CH-2 (blurry pix above). It’s clear that both have less detail. But one of Mother Pelican’s chicks has moved up near her bent head, and another now floats over her back. The nest detail seen in CH-1 is now symmetrical right and left at the top of the urn, instead of looking like leaves on one side, and scrolls on the other. The sprouts on either side of the urn’s bulbous body have changed attachment points, and now hang down, instead of growing up as pomegranates.
CH-2’s harpy has retained her hairdo, but her wings are a bit more gracefully rendered, employing 45-degree angles to round off some of the shapes. Her feathers are more evenly spaced, but her tail is less pronounced, and whatever small markings covered her haunches have been lost in favor of more, smaller feathers. The thing she’s holding has lost its hatching, and now looks more like a cup than panpipes. She has also inherited another wayward pelican chick.
MET-1’s urn is in between the other two in terms of detail. The nest/scroll unit at the top underneath the big pelican has transformed into a chick. The stitcher chose not to fill in the background in the loop defined by the pelican’s neck. There is something unidentifiable between the pelican’s legs, and her fathers are somewhat simplified compared to CH-1. The lower ornament is again descending from the bowl of the urn as leaves, rather than rising from the base.
The harpy too has changed a bit. In this case, I’d say the sipped/sounded thing has parted company from the hand, and now looks more like panpipes, vaguely supported rather than held. She’s gotten a bit more balloon-like, and her breast feathers now march row by row. Her wings however have gotten a bit stunted, and return to a stepwise rendering similar to CH-1, but slightly more clumsy. The tail is suggested, and the haunches have been returned to stippling rather than feathers.
OK. It’s clear that detail is going to be lost when you move from ornamented surface stitching to the negative/positive lacy mesh look of the all-white technique. But even so, a tremendous amount has been preserved. We see the plumage of Mother Pelican, and even some details on her brood (she’s managed to gather three of them together on top of the urn). Her nest is symmetrical. The urn preserves the shapes and proportions of the red stitched pieces, and has grown back the two small pomegranates that grow from the base.
The harpy too preserves a lot of detail, down to the proportions and shapes of its flight feathers, and a bit of the detail inside of the wing. She’s lost some weight, although her hairdo is less detailed. Breast feathers are present, as is a pretty clearly defined tail. Stippling on the haunches looks different from the breast plumage, and her feet are now nicely shaped lion paws.
The final example, MET-2, the Spanish piece, is a bit simplified. The harpy is less prominent, and the largest space is given over to the urn and pelican, and to the foliate ornament between the repeats. Mother Pelican’s brood is more suggested than rendered, although her feathers are nicely done. The urn has the two upward growing pomegranates emerging from the base.
The harpy’s cup/instrument has become less detailed. It’s unclear what it might be. Her feathers have given way to geometric ornament, and her tail is suggested in shadow rather than being clearly defined. Her wings are somewhat like the Hermitage example’s, though. It’s worth noting that her proportions and body shape are more like CH-1 than the other examples.
One other thing that’s of interest is the presence of the little filled boxes that bead the motif’s edges. You can see them along the curve of the pelican’s neck, along the harpy’s breast, and lower legs. They give a lacier appearance to the composition. I also find little protrusions like this to be extremely valuable as I stitch my motifs because they help me confirm counts and stay true to the design. Note that they are absent in the other renditions.
Now, having our fill of urns and harpies, what can we say about them?
It’s obvious that there is an as-yet unidentified but unifying source for this design. I posit that there might originally have been a broadside or model sheet that showed the composition. I guess that it may have been on the count, and that its broad outlines were used to establish the placement of the main design elements. But I don’t believe that it was followed exactly. Instead I think each stitcher used it to establish the first iteration of the design, filling in the details and roughly eyeballing their placement, taking inspiration rather than ironclad direction from the model. Once the first repeat was worked, subsequent repeats and mirrorings were copied from that, with no more call to look at the original. That’s why the baby birds wander around, while the relatively easy to place urn decoration remains more stable.
Because of the different media and slightly different interpretations of the pattern (especially the pomegranates on the urn, and some differences I didn’t detail in the filler between the main motifs), my guess is that the same design branched into two slightly different but recognizable pattern “traditions,” which in turn spawned child works of their own. One of those traditions (marked by the upward pomegranates) made the leap from surface work to darned net.
Now. Which came first? I can’t say. On intuition alone I’d go with the fat, balloon-bodied harpy (MET-1) being later than CH-1, and the two white filet pieces belonging to the same “tradition” as CH-1. That leaves MET-1 and CH-2 as child works of the other branch.
Which came first? What chronological order can be used for these pieces? Aside from these idle thoughts, your guess is as good as mine. If you’ve managed to make it this far, please feel free to differ. Without detailed analysis or forensic investigations into fiber and dye, we’re all just speculating, anyway.