The marble inscription The top three lines of the stele (Wick, 1995:plate XXIII) are given below showing its 48 character stoichedon (grid) pattern and the worn character of the marble.
A possible reconstruction from Meiggs and Lewis is shown as well.
However, he decided, from the inscription itself, that a dotted phi was appropriate in front of "ων" was to use letter forms to "formulate criteria" and "provide the framework" for dating fifth century Attic inscriptions that could not be dated by other means.
Meiggs and Lewis (19) note that 454/3 BCE was the "generally accepted" date for the alliance until sometime around 1969 (1964), and Mattingly (198) rejected this completely. the rho was not readable, but Habrōn was left as the probable reading because of the probable beta.
By 1971 Meritt (according to Wick, 197,n.8) had moved to Raubitschek's position. One reason was that the Lilybaeum mentioned was not founded before the fourth century (Ehrenberg, 196,n.138) and so there could not have been a war with that city in the mid fifth century.
The evidence for this alliance comes from an inscription on two fragments of a marble stele, labelled variously Inscriptiones Graecae I2 19,20 and I3 11.
The alliance has been dated as 458/7 or 454/3 or 418/17 BCE (Fornara, 20). Evidence for the first two dates is discussed and then the claims of letter-form analysis.
Since then Mattingly has argued copiously , on the basis of historical congruence and textual context, for a number of inscriptions to be put back into the 420s BCE.