This is the Karagashli carpet, number ten in Meik and Ian McNaughton's Making Miniature Oriental Rugs and Carpets.
It is worked in Appleton's crewel wool on 22-count Hardanger canvas, something of an experiment as I haven't been able to easily get hold of 24-count canvas, which is what the McNaughtons use throughout their book, and I had only 22-count and 28. The coverage of the Appleton's is erratic at this gauge, even inadequate in some places, I'm afraid. I noticed that the lighter colors have better coverage than the darker ones, so perhaps this is due to the peculiar effect of different dyes on the same kind of wool -- I've noticed this in knitting wool, certainly. I wanted a good-sized carpet, so I chose the larger gauge, going on various needlepoint websites that assured me that Appleton's is for 18-count. I did get the wool to bloom a little in the blocking, though -- soaked it in warm water for a few minutes, roughed it up a bit, whacked it on the side of the sink -- which did actually work a little, but you will have to take my word for it as I'm afraid I forgot to take a "before" photo. There are still some noticeably thin spots, but, well, I will just call it "wear and tear" on an "antique" carpet, I guess! I purposely didn't block it very hard, either, to leave it looking even more aged.
I had just enough of the green wool left to make a small sample to compare "before and after". I suspect now that much of the inadequate coverage has to do with the particular stitch used, as you can see on my afterthought-sample that the canvas shows through on regular diagonal lines every other stitch, and I have also noticed in other projects that the stitches lie differently when the continental is worked to the right than when it goes to the left -- so, like knitting, it apparently can make a noticeable difference what direction the wool goes in at the back of the canvas as it comes in or out of the stitch on the right-side. It is possible that continental stitch would have provided better coverage, but of course with more attendant warping in the finished piece.
(On closer inspection, I see that even some of the other carpets in the McNaughtons' book show a little canvas in the darker background sections, so perhaps even 24-count canvas is a bit skimpy in some projects. Oh, well.)
This is the second carpet I've worked from this book, and while the first chart was flawless, this one had a small but fairly glaring mistake in the outermost border (repeated, because apparently the chart was mirrored for printing) but it was so obviously wrong that it was not at all difficult to spot it before I worked that area. I did notice -- eventually! -- that the three red rhombus-ish motifs in the center are actually slightly different from each other, but whether this is a mistake or intentional, I don't know. ("Only Allah is perfect.") Mine are worked slightly differently, because I assumed that they were all the same and wasn't working strictly to chart! But otherwise I'm delighted with this design, and the planning that went in to making the odd-sized motifs fit into a regular field -- notice that the white "stars" touch the border at the bottom end, while there is a single line of green background on the opposite end, and that the whole arrangement of central motifs at the bottom is not quite in the same places relative to each other as in the top part -- by only one row -- but since it is only one row, you hardly notice it, only that everything fits neatly. This ever-so-slight imbalance is in order to accommodate a perfectly regular arrangement of the border motifs, and I can only admire the meticulousness of designers whose corners have full, continuous motifs, not lopped-off overlaps. And I would also highly recommend this particular chart for a beginner to petit point, as the border is very logical and therefore easy to work, yet the central panel is rather fantastical and interesting.
I suspect that the McNaughtons' designs are not based on specific antique carpets, as Frank Cooper's are for instance, but they label this one as "mid-19th century" so perhaps it is more in the style of a mid-19th century Karagashli than a miniature version of a particular one. They write, "Karagashli is a village near Kuba in the eastern Caucasus. Many carpets woven there display a series of banner-like red rhomboids on a turquoise background, surrounded by white cruciform stars and strange, bird-like motifs. The pattern in the main border is known as 'leaf and chalice' pattern. It is not peculiar to Karagashli, but is widely used in Caucasian weaving."
Here are two full-sized Karagashlis, in fact, from the very helpful and interesting identification guide at Azerbaijan Rugs in San Francisco --
which clearly show the inspiration for the miniature version. It would be fun to work the McNaughtons's chart again with the dark-turquoise background, though I do like the green. There are also ones in the guide where the abrash is quite noticeable, and a handsome one where the two background colors are reversed, with a mid-blue on the leaf-and-chalice border and a golden-ivory in the center panel. (Interestingly, the guide calls the border "leaf and calyx" -- like the McNaughtons, I guess, I certainly immediately saw chalices!)
New skill for me on this project: basketweave stitch. Most needlepoint guides, online and in print, tell you to do larger swathes of single-color stitching (backgrounds, for example) in basketweave, which unlike continental stitch has a tendency to keep the canvas true. I'm guessing that this comes from the strong angle of the back of the continental stitch, which you can see here in pretty much everything except the green background and the red in-fill of the large medallion. Presumably the vertical/horizontal alignments of the back of basketweave stitch counteract the slant of its front. But I got my nerves all a-jangle by various sources warning that you must alternate the vertical and horizontal rows or you get a little ridge visible from the front -- fair enough, but the instructions on how to do this seem to involve a whole lot of trying to keep straight which is the warp and which is the woof of the canvas and whether you have used vertical on the warp or on the woof, say by going "under" on the "overs" (or is it "over" on the "unders"??) and this all just made my brain go aaaauuggghhhhhh. But since this particular piece had already turned out to be a "teaching moment" I thought, "right, just try it" and dove in. After carefully working a good-sized plain section in one corner, I realized that this is the back of a "down" row --
and this is an "up" row --
It seems quite a simple matter to me to tell which is which, so I worry that I'm missing something and it actually is really tricky. As long as you have a long-enough row for the stitches to be clear, it shouldn't be a problem -- but since this revelation I have been making it a point to stop on a long row like these where it is quite obvious how the stitches are lying, or even better to start a row and pause it in mid-stitch so that I can see right away in which direction I've been traveling.
This is the overcast edging, as Janet Granger uses -- I wasn't sure if there was enough blue to work one of the "braided" edgings.
Now that I've had a chance to look at it finished, I'm less unpleased with it than I was before, so perhaps if I put it in a darker corner of some roombox, the wonkiness will look quite natural!