[texhax] Spurious kappa1 glyph in cmmi fonts?

[Douglas McKenna]

Using the fonttable package, one can get a one-page picture of all 128 glyphs for the TFM character codes in any of the cmmi (Computer Modern Math Italic) fonts, using, i.e., 

%%%%%%%%%%%
\documentclass[11pt]{article}
\usepackage{fonttable}
\begin{document}
\fonttable{cmmi12}
\end{document}
%%%%%%%%%%%

But if one converts the "cmmi" file's corresponding .pfb file to an OpenType font file, and then browses the glyphs in that resultant ".otf" file, there appear to be two extra glyphs that aren't represented in the original TFM font.  The PostScript glyph names associated with these two extra glyphs are "hardspace" and "kappa1".

There are a small number of variant lower case greek glyphs in cmmi whose PostScript names end in "1" (e.g., "psi1", "rho1" , etc.).  The "kappa1" variant glyph, however, is not to be found in the fonttable output.  I can imagine someone went a little too far, realized they'd created something that wasn't in the original font, but left it in anyway so as not to lose the work, because doing so wouldn't cause any problems.

Are these spurious glyphs in the .pfb file left over from some hand conversion back in the day, or is there more method to the madness than I can surmise?  How is it possible to draw this "kappa1" glyph using just a TFM cmmi font?

Doug McKenna
Mathemaesthetics