David J. Perry wrote: [...] >It occurs to me that it may be a question of habit. We are used >to seeing E/Epsilon, P/Rho, etc. with the same shape. We see >Lunate Sigma less often and so in my mind, at least, the C with >serifs is somehow not associated with a Greek letter--it is >advertising itself as a Latin form, not a Greek one, so to speak. I think you hit the nail on the head. Certainly from the-point-of- view of strict formal homogeneity the serif-less Lunate Sigma stands out, but, whichever way a Greek reader looks at it, it is a Latin C. To my eyes the lunate forms are inextricably linked with Byzantine uncial forms which, albeit modulated, did not have anything relating to what I think of as serifs. John Hudson wrote: [...] >On which note I have to share the attached graphic, one of my >favourite examples of multilingual 'making do'. The amazing thing with this is that they used a reversed U for Pi, which is nonsense: especially in the absence of a Lambda in the syllables preceding it this would be read as a strange Lambda, not a Pi. Only when you take in the 'proper' Lambda in the next syllable do you get it... Gerry