TJIIRRS: Number 2 of an Ongoing Series

In Pursuit of Yohen Tenmoku
Late April, 2005

When you decide to replicate an ancient glaze, there are usually a few paths you can take. If there’s an analysis, for example, you can try making a glaze with that composition and working from there. If there are modern glazes with similar appearance, you can start with one or more of those and tweak.

In the case of Yohen Tenmoku, unfortunately, we don’t quite have either of those options. The ancient glazes are among the rarest of Song pieces; there is an iridescent oilspot bowl (just one) in the Miho Museum, and three bowls in other museums, all of which resemble each other in terms of glaze and effect but are not standard oilspot. (We’ll deal with the Miho Museum’s bowl later: I haven’t even started on that one yet.) There are no shards of either type that I’m aware of, and so there are no analyses. Nobody knows exactly where or when these bowls were made, though I think they are very clearly Jian ware. All of these are designated National Treasures and Important Cultural Objects, if my information is accurate.

The three bowls that resemble each other are not like anything else I’ve seen, though there are indications that the same effect or a similar one does appear from time to time, rather unpredictably. You can see a photo of the best of these at the “Taste of Japan 2003” site, but it has fairly obviously been tweaked to bring out the blue luster. (Note the area of blue luster on the exterior. I believe that this is a significant feature, and I’ll discuss it later.)

[NOTE, added 2011.0407, early am: There is now a nice photo of the interior of what I believe is the Seikado Bunko bowl, widely acknowledged to be the best of the Song dynasty originals, here.]

Here is a closeup of a modern piece, showing what I think may be a similar type of luster:

(I don’t have any more pixels than that, sorry.)

I’m trying, somewhat against the odds, to learn how to make these. The fortunate thing about this effort is that it is repaying my effort in other ways, and will continue to do so even if I don’t manage to replicate the Song dynasty original. (I should say, the original Song dynasty glaze effects — it is beyond me to be a Song dynasty potter.) The unfortunate thing is that it is going to be impressively difficult to achieve.

The base glaze on the bowl pictured above (which is generally agreed to be the best of the three) is visually similar to the glazes on some other Jianware teabowls, and there are various analyses of Jian glazes in the literature, so at least I have a starting point. In fact, I have already begun to create an approximation of the base glaze, starting from J. M. Plumer’s 1935 analysis. (I’ll get back to this.)

The spots on these three bowls are, as I say, significantly different from ordinary oilspots; I’m testing washes to see how close I can get. They are also grouped oddly, and I am only beginning to have some ideas about how the wash may have been applied. (It is clear that they were applied, and did not simply appear out of the composition of the glaze.) The area of luster on the outside of the Seikado Bunko bowl suggests to me that when someone picked it up, they did so with a hand that was slightly wet with wash. There’s even a misformed partial spot, if I’m reading the photo correctly. This, in turn, suggests the possibility that the spots were put on by flicking wash off fingers, rather than dripping it from a brush, and I’ll be testing that. (I can already report that it would be very difficult to create groups of the right type with any brush I own.)

There’s a rumor that a master in Gifu prefecture, in Japan, has already replicated this glaze; but I can’t even find out the guy’s name, much less any details about what he did or how he did it. There’s a photo on this page, a little over halfway down, with a bit of text. At one point they published a larger one. (Note, if you click this small image, you’ll get one that’s 1104x1028.)

[NOTE, added 2011.0407, early am: It is now quite clear that several people in Japan have replicated the effect. Needless to say, I do not have any of the technical details.]

I’m writing an article about my attempt to copy this, and about how I have to think about it in order to do so. Until I actually make some concrete progress, of course, I can’t finish such an article; but writing it helps me marshal my thoughts, and sometimes gives me ideas about things I can try. Likewise, I make various glaze and wash tests, some of which suggest further testing. (Others indicate paths that aren’t fruitful here, but may be interesting in and of themselves. I have, for example, tested some washes that turned out to be quite lovely.)


As I mention above, the base glaze is probably similar to the glazes on other Jian wares; but if you look at the photos you’ll see that it is quite dark, and doesn’t show any hare’s-fur or oilspots of the ordinary sort. It is important to remember that the body clay is dark, and rich in iron; I am thinking about reproducing it from published analyses, though that may not be entirely necessary: I may be able to create a glaze formulation that fires out well on porcelain.

One of the peculiar aspects of this is the fact that when someone analyzes a high-fired glaze, an uncertain amount of the body shows up in it. This is because there is an intense reaction between body and glaze at high-fire temperatures, very different from the situation at low fire, where the glaze sits on top of the body almost like a layer of ceramic paint. In reproducing these glazes from analyses one is obliged to compensate, and because the amount of body that gets included in the analysis is uncertain, that is quite, uhh, challenging.

Another problem is the firing protocol: it’s fairly clear that Song dynasty potters fired large numbers of pieces, in stacks of saggars, in large kilns. This means that the firing took some days to reach peak temperature and some days to cool back down. The rate of heating is probably not horrendously important, but any reduction that took place is certainly an issue, and the cooling is likely to be absolutely crucial. I have seen examples of a high-iron glaze, cooled quickly and slowly, and it is difficult to tell that they are even particularly related, much less that they’re identically the same composition.

Leaving that aside for the moment because "firing down" is going to be difficult in anything other than an extremely tiny electrically-heated test kiln (yes, I’m building one), one of the things I would like to be able to do is fire to about cone 11 in oxidation, and then do a reducing soak at the peak of the firing. My big burner will easily take the Paragon to cone 11 or even 12, but won’t reduce — the blower on it puts out so much air that even when I crank the gas up as far as I can, the flame is still oxidizing to neutral.

I have swapped out the pair of small blowers on the little burner, replacing them with a single larger blower. My first actual firing was a good one, and it’s clear that this fan is an improvement, but I need to start reduction a bit earlier next time.

Then there is the issue of the wash. The spots, as they were described by Dr. Yamasaki in a 1982 article (he got to hold all three bowls and examine them closely with a hand-lens), are yellow, matte or granular, and slightly raised. They are in groups, and as you can see from the photos, while most of them are about 6 or 7 mm across, there are also smaller ones. The lustrous blue regions surround the groups of spots, which of course suggests that something coming out of the spots is responsible for the formation of the luster. According to Yamasaki, the luster is not iridescent: its color doesn’t change with viewing angle. On the other hand, it clearly is an interference effect of some sort, as the sizes and shapes of the lustrous regions do change with viewing angle. This tends to suggest that the luster is on the surface, and not within it.

You’ll notice, if you look carefully at that last photo, that some of the luster appears to have been scraped off, which again points to it being on the surface rather than in the glaze. There’s also an exceedingly peculiar thing about the wash spots: the centers of most (but not all) are yellow, and the edge is gray; each is then surrounded by a yellow ring, and the blue luster seems to be an extension of that yellowness.

I am quietly trying to learn how to make spots that correspond to what Dr. Yamasaki observed and what’s in the photos. It is not easy. I have begun to get spots that are more or less granular, but they tend to be brown rather than yellow, I don’t seem to get any gray, they aren’t grouped right yet, and there certainly hasn’t been any luster around them so far.

It seems reasonably likely that some volatile species may be involved in the formation of the luster. It is possible, alternatively or in conjunction, that something in the liquid wash diffused out into the glaze when the wash was first applied, and effloresced on the surface as it dried. The most obvious candidate for the latter is the soluble salts in wood ash, as it is almost certain that the Jian glazes were clay-and-ash based.

Quite a few chlorides are fairly volatile, and after I had less than satisfactory results with CaCO3 it occurred to me that perhaps I should try Mg instead of or in conjunction with Ca, and that perhaps I should try the chloride rather than the carbonate. Nigari, a side-product of the production of sea salt, is mostly MgCl2. It is used for coagulating soymilk to make tofu. Nigari is cheap, maybe two or three dollars a pound, if you buy it for that purpose. It is also easy to find on the Web, but difficult to find in my local area, as I discovered today when I went looking.

I’ll probably mail-order some, if it appears to be The Right Stuff; but in the meanwhile I went to the Home Depot and got some Muriatic (= Hydrochloric) Acid, a small amount of which I poured into 10 grams of MgCO3 a little while ago. Once I get this balanced so it is just slightly acid, I can make some wash tests with it. I also intend to test Epsom salt, which is MgSO4 and will also release some gases, probably at a higher temperature than the chloride.

There’s also a chance that Mn is involved in the luster, and I’ll be checking on that; but it is just as likely to have been present in the base glaze as in the wash, as the ash is a likely source for it. (There’s no obvious reason why they would have used a different ash for the wash, though of course that’s certainly possible.)

If I manage to get some luster, I’ll photograph it and report on it.


Publishing News

My article on Fluorescent Glazes is in the May-June 2005 issue of Clay Times, which has been printed and will be out any day now. The colors in the photos are quite nice.


Horticultural News

Behnke Nursery, not far from here, has bigleaf tea bushes. We were quite surprised to find them: tea is available as an ornamental camellia, but not very widely. (Raintree Nursery, in Morton, WA has had it for a few years, but I’m not particularly aware of other vendors.) The label claims that the plants are hardy here, which is possible — tea grows nicely in Seattle — but there are now two of them out by the greenhouse, and I’ll believe it when they’re still alive in five years or so.

[NOTE, added 2011.0408, early am: One of the tea bushes is not only alive, but has been covering itself in flowers every autumn for the past few years. Moreover, it is in a raised bed, several feet above ground level. Clearly, tea is reasonably hardy here.]

Various other plant news, but nothing particularly earth-shattering. I am finding that a small amount of colloidal silver in the water may help prevent tomato cuttings from rotting before they can root, but I have a sample size of one (1) so far, and I’m taking it with a grain of salt until I can do more testing.

Back to the Index


This work is supported by
the Joss Research Institute
19 Main Street
Laurel  MD  20707-4303  USA

Contact Information:

Email:, where you can replace a with my first name (jon, only 3 letters, no “h”) and b with joss.

My phone number is +1 240 604 4495.

Last modified: Fri Sep 2 13:51:25 EDT 2011