A slight digression from the original post, so please excuse us for a moment:
But I don't put metaborate in the regular D23. Should I start? What does it do that metol and s. sulfite don't?
The recipe is in the Ansel Adams book
The Negative, which sayeth:
"The most effective two-solution formula that I have found consists of D-23 for solution A (developer), and, for solution B (alkali bath) a 1 percent solution of Kodak Balanced Alkali (Kodalk), or 10 grams per liter. I have not conducted exhaustive tests with current films, but our preliminary data indicate that two-solution development can be effective for supporting shadow values while reducing high-value densities. I include the method here for those who wish to experiment with it. It is important to expose the low values about one zone higher than normal if this procedure is to be used.
The negative is immersed in solution A for 3 to 7 minutes, followed by a 3-minute bath in Solution B, without agitation. In the past, if the film tended to fog, I have added small amounts of a 10 percent solution of potassium bromide solution to the second solution. The strength of the high densities will be controlled primarily by the time the negative is developed in the Solution A; during the immersion in the alkali solution, the developer is soon exhausted in the high densities while its activity continues in the lower values. the Kodalk solution can be made stronger---up to about a 10 percent solution, or 100 grams per liter---and its activity will increase, but with the possibility of increased grain. I advise careful testing before using this process."
Sulfite is a weak alkali, and also a reducing agent; it activates the Elon and protects it from oxidation. Kodalk (metaborate) is somewhat more alkaline, but with a high buffering capacity, so that the effectiveness doesn't change much as the reactions go forward. This two-bath system is reasonably inexpensive, and very flexible---I used it when I was exploring the Zone System. It is an excellent compensating developer, and I would probably still be using it if I hadn't discovered PMK and Pyrocat.
So, technically, we're not
adding Kodalk to D-23 (that would have deserved different nomenclature) but rather using D-23 as the first half of a divided developer. There are a number of variants of divided D-23, as I'm sure you know; I think a few of them have both sulfite and metaborate (or some other alkali) in the second bath, and at least one omits most of the sulfite from the first bath.
We now return to your regularly scheduled topic