A close look at the use of γελαν, αλλος and ετερος shows that none of the interpretation of Pi. I. 1,67-68 proposed by Farnell, Norwood, Bundy, Privitera and Most can be upheld. The alternative is either to adopt an interpretation close to Thummer's which is linguistically sound but makes little sense in the context, or to alter the text. A conjecture which is palaeographically close and makes good sense is to read γέλω in place of γελα.