There have been a number of pitfalls to glaze testing beyond the normal, where the tests just don't work. I have had a number of tests pan out to the second stage where they didn't make it and some go on to be made up in larger batches, over 2500 gr. and then cease to work as they had in several test phases previous. But by far the most frustrating and even devious scenarios are where you make up a batch, glaze multiple pieces and the glaze responds perfectly. Now out of the new test glaze, you go and make up more, glaze up more pieces and when you unload the kiln, the new glaze has failed.
This happened recently, using materials that I "inherited". I made up a cool, bright glaze and glazed a number of pots, all came out well better than expected, however, the glaze ran low and I made up more with the same exact materials and the glaze failed miserably. My current suspicion is that the original glaze was made up with chemicals that have been around since the 70's and the new batch from materials I bought as far back as last spring. I think I can ultimately solve the problem, but I won't hold my breath. There is never any glaze certainty when testing is involved.
Illustrated is one of the blend glazes I came up with about a month or so ago. This one is called T'pir after the initials of the two glazes it sprung from and has made it through two levels of testing. Unfortunately, the necessity for several cycles of terra cotta has put the next phase of the testing on hold, but once I get back to stoneware, I see some teabowls, small jars and a vase or two destined for the T'pir glaze.
"If we begin with certainties, we shall end in doubts; but if we begin with doubts, and are patient in them, we shall end in certainties." Francis Bacon (1561-1626)