Unicode Greek Fonts for Numismatists
                                                          by Chris Hopkins
                                                     Version 5, 28 Feb 2023

[Instructions: Open this file in Windows Notepad, then use the Format... Font menu to select different fonts to see how the Greek characters on Lines 1-4 are displayed. A web version is available at http://parthia.com/fonts/unicode_greek.htm]

With the continued advancement of font technology, I have revised my methodology: I used BabelMap to test each font as a single font (no substitutions permitted) to detect the supported Greek letterforms. I can now recommend that you select a font that includes all the required Greek characters.

The following Greek characters are the minimum needed to represent Greek legends on ancient coins (the modern koppa is included only for illustration of a problem).

Line 1. Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω
Line 2. Ϛ Ϝ Ϡ
Line 3. Ϙ (archaic "lollipop" koppa), Ϟ (modern "lightening bolt" koppa)
Line 4. Ϸ Ϲ Ϻ

The characters on Lines 1-4 will be displayed in Windows applications (including Internet browsers) if the viewer is using a Class I font. Lines 1-3 are almost always displayed if viewed using a Class I, II or III font. The Class IV and V fonts should not be used for Greek characters other than those on than Line 1.

Class I Fonts:
Lines 1-4 are fully supported.
Examples:
   Alphabetum (for sale at http://guindo.pntic.mec.es/~jmag0042/alphaeng.html) 
   Arial Regular (Microsoft) 
   Bahnschrift  (Microsoft) 
   Calibri (Microsoft)
   Cambria (Microsoft)
   Cardo (free at http://scholarsfonts.net/cardofnt.html)
   Code 2000 (free at http://www.code2000.net/code2000_page.htm)
   Consolas (Microsoft)
   Courier New (Microsoft)
   Microsoft Sans Serif 
   Segoe UI (Microsoft)
   Tahoma (Microsoft)
   Times New Roman (Microsoft)

Class II Fonts:
Lines 1-3 are fully supported.
Line 4 is not supported. 

Class III Fonts:
Lines 1-2 are fully supported.
Line 3 partially supported. These fonts are missing archaic koppa which is usually substituted from another font by Windows. The modern koppa is drawn as an archaic koppa, causing confusion (except in Lucida Sans Unicode).
Line 4 is not supported.
The problem is that these fonts incorrectly contain an archaic koppa character at the modern koppa position (Unicode codepoint U+03DE). Because there is no correctly coded archaic koppa in the font, Windows will substitute an archaic koppa from another font, if available. The user cannot readily determine if an archaic or modern koppa is used. Where the modern koppa codepoint is used, it will appear as an archaic koppa in these fonts, but appear as a modern koppa when the text is copied to another application that uses a correctly constructed font.
Examples:
   Aisa Unicode
   Georgia Greek (also missing the F-style Digamma, usually substituted by Windows)

Class IV Fonts:
Line 1 is fully supported.
Line 2 unreliably supported. Font does not contain any of the Line 2 characters, but Windows attempts to supply them through substitution, so these fonts usually display all the Line 1 and 2 Greek characters.
Line 3 unreliably supported. Font does not contain either archaic or modern koppa, but Windows attempts to supply them through substitution.
Line 4 is not supported.
Examples: Gentium, Warnok Pro 

Class V Fonts:
Line 1 is fully supported.
Lines 2-4 unsupported, although some characters may be substituted by Windows.
Unuseable. These fonts contain unreliable support, or glyphs unsuitable for representing numismatic inscriptions.
Examples:
   Arial Unicode MS (the most widely used Unicode font)
   Candara (Microsoft)
   Comic Sans MS
   Corbel (Microsoft)
   Ebrima (Microsoft)
   Franklin Gothic Medium (Microsoft)
   Gabriola Microsoft)
   Gadugi (Microsoft)
   Garamond
   Georgia (Microsoft)
   Lucida Sans Unicode (Microsoft)
   Minion Pro 
   Palatino Linotype (Microsoft)
   Trebuchet MS - the U+039E Greek capital letter XI is the rare "crossed I-beam" form
   Verdana (Microsoft)

In the text above, I mentioned that the Windows operating system will try to substitute a character for one that is missing in a font. To see how inconsistent Windows is in choosing the characters to substitute, do this:
   (a) Open this document in Notepad
   (b) Set the font to Arial Unicode MS
   (c) Turn off word wrap
   (d) Select All, then copy the Lines 1 to 4 text to the clipboard
   (e) Open a new document in MS Word. If the default font is not Times New Roman, set the font to TNR
   (f)  Paste text from the clipboard. Examine the two koppas
   (g) Select All, then change the font to Arial Unicode MS. Examine the koppas. Select All, then change the font to Times New Roman
   (h) In Micorsoft Word 2003 and later, you can examine the Unicode codepoint of a character by placing the cursor immediately after the character, then pressing Alt-X

You will likely see some interesting substitutions for the koppa, different from the substitutions used in Notepad. In my trials, the archaic koppa missing in the font is substituted with an archaic koppa (supposedly from the MS Sans Serif font which we know does not have any Greek characters in the font!) and the modern koppa is replaced with a poorly designed modern koppa from Palatino Linotype; it is recognizable neither as an archaic nor a modern koppa. Another strange aspect of substitutions is Windows' unpredictable use of different fonts for substitutions depending on point size of the font.

Conclusion: Use the characters on Lines 1-2 in any web page or document as they will be reliable reproduced in Windows applications (including Internet browsers) by Class I or Class II fonts. Use one of the fonts that has all the Greek characters in the font for Lines 1, 2 and 3, else don't try to use the characters on the third line. Important: remember that the user must have the font on his computer to see a web page designed with a specific font. You therefore must use a commonly distributed font or advise the user to be sure he has the font you specify.

=============================================================
Unicode characters suitable for describing Greek coin inscriptions:

Line 1:
Α U+0391 : GREEK CAPITAL LETTER ALPHA
Β U+0392 : GREEK CAPITAL LETTER BETA
Γ U+0393 : GREEK CAPITAL LETTER GAMMA
Δ U+0394 : GREEK CAPITAL LETTER DELTA
Ε U+0395 : GREEK CAPITAL LETTER EPSILON
Ζ U+0396 : GREEK CAPITAL LETTER ZETA
Η U+0397 : GREEK CAPITAL LETTER ETA
Θ U+0398 : GREEK CAPITAL LETTER THETA
Ι  U+0399 : GREEK CAPITAL LETTER IOTA
Κ U+039A : GREEK CAPITAL LETTER KAPPA
Λ U+039B : GREEK CAPITAL LETTER LAMDA
Μ U+039C : GREEK CAPITAL LETTER MU
Ν U+039D : GREEK CAPITAL LETTER NU
Ξ U+039E : GREEK CAPITAL LETTER XI
Ο U+039F : GREEK CAPITAL LETTER OMICRON
Π U+03A0 : GREEK CAPITAL LETTER PI
Ρ U+03A1 : GREEK CAPITAL LETTER RHO
Σ U+03A3 : GREEK CAPITAL LETTER SIGMA
Τ U+03A4 : GREEK CAPITAL LETTER TAU
Υ U+03A5 : GREEK CAPITAL LETTER UPSILON
Φ U+03A6 : GREEK CAPITAL LETTER PHI
Χ U+03A7 : GREEK CAPITAL LETTER CHI
Ψ U+03A8 : GREEK CAPITAL LETTER PSI
Ω U+03A9 : GREEK CAPITAL LETTER OMEGA

Line 2:
Ϛ U+03DA : GREEK LETTER STIGMA
Ϝ U+03DC : GREEK LETTER DIGAMMA
Ϡ U+03E0 : GREEK LETTER SAMPI

Line 3:
Ϙ U+03D8 : GREEK LETTER ARCHAIC KOPPA (a correct font has "lollipop" koppa)
Ϟ U+03DE : GREEK LETTER KOPPA (a correct font has "lightening bolt" koppa, but it is a modern character, never used as koppa on ancient coins)

Line 4:
Included in the Unicode specification, but not yet supported by most fonts:
Ϸ U+03F7 : GREEK CAPITAL LETTER SHO [Bactrian letter]
Ϲ U+03F9 : GREEK CAPITAL LUNATE SIGMA SYMBOL
Ϻ U+03FA : GREEK CAPITAL LETTER SAN