The headword phrase -- dragmh\ xalazw=sa here but draxmh\ xalazw=sa in two earlier versions, Zenobius 3.27 and Hesychius delta2351: see further on this discrepancy below -- has been claimed as a fragment of Attic comedy (Adespota 314 Kock, now 950 K.-A.). Be that as it may, the context is classical Athens. For the theoric payments there see theta 218, theta 219, theta 220. In the present instance Hesychius and the Suda transmit the cardinal phrase as to\ qewrhtiko/n, but Latte printed to\ qewriko/n in Hesychius on the strength of Zenobius (zeta 73), and that is the preferable reading here too. Uncomprehending scribal substitution of qewrhtiko/n (v.s.) for qewriko/n (v.s.) is seen elsewhere. In the Suda alone Adler noted two instances -- one common to all mss, one in ms F only -- in her apparatus to theta 218 [= Photius, Lexicon theta150], above, and the same phenomenon under pi 388 and pi 1181 (both ms V). At tau 515, linked with pi 1181, she printed the paradosis qewrhtika/ even though the context demands qewrika/.
The other textual divergences here should also probably be resolved, likewise, in favour of the Zenobian version, from which the Suda compiler or his source departs in notable (though self-consistent) respects. In Zenobius the hail "fell", e)/pese; here in the Suda it "stopped", e)/pesxe. (This sentence is independently lacking in Hesychius; Latte supplies it from the paroemiographer.) And the final phrase of all, xalazw=san au)th\n e)pe/skwpton ("the joke was that it [the drachma] was falling as hail") in Zenobius, becomes dragmh\n au)th\n xalazw=n e)pe/skwpton ("the joke was that it was a handful of hailstones") here. (For hail see generally chi 5.) This recasting of the joke seems to be a corollary of presenting the noun in the headword phrase itself not as draxmh/, i.e. unambiguously the coin of that denomination (delta 1516), but as dragmh/, a post-classical variant on it which, depending upon context, can signify either the coin or a "handful" of something (= the original sense of draxmh/ itself, as many oboloi as the hand could grasp: Plutarch, Lysander 17.3); see LSJ s.v., and cf. delta 1490 for neuter dra/gma. The Zenobian version of the joke, accepted here, suggests sheer pleasure in theoric money dropping from the sky like hail. The Suda version would be more rueful (or cynical): money slipping through the fingers like hailstones melting after a storm.
For a drachma as the theorikon cf. under theta 218 and theta 219 -- but contrast theta 220 (two obols), and see the detailed discussion in Pickard-Cambridge/Gould/Lewis, The Dramatic Festivals of Athens (revised edn. 1988) 265-268.
[1] This phrase has been understood by most scholars as an archon-date, i.e. 395/4. Latte, for instance, adds this in parenthesis in Hesychius; and such a (putative) datum has duly played its part in reconstructions of the history of the institution (chiefly J.J. Buchanan, Theorika (1962) 29-34, 48-60). However, since good evidence for theoric payments proper in the 390s -- as opposed to payments for e.g. assembly attendance (?Aristotle, Athenaion Politeia 41.3, etc.) -- is hard to come by, some have identified this Diophantus as the well-attested Sphettian politician of a later generation, who, while never eponymous archon, is linked with theorika by a scholiast to Aeschines 3.24. (So G.L. Cawkwell, 'Eubulus', JHS 83 (1963) 47-67, at 55 n.53; P.J. Rhodes, A Commentary on the Aristotelian Athenaion Politeia (1981) 514.) Such a connection must remain conjectural, however, and scholars still disagree on when theoric payments began. (A recent contribution, its argument conveyed by its title, is D.K. Roselli, 'Theorika in fifth-century Athens', GRBS 49 (2009) 5-30.) As regards the present item of evidence, epi Diophantou probably should, on balance, be construed as an archon-date -- whether or not one accepts the historical implications of doing so. It would have been helpful if Zenobius had made this certain by adding the participle 'as archon' (for which see sigma 77, sigma 1386, phi 761, and cf. e.g. omicron 764 for a non-Athenian instance); nevertheless, an important source for what lexicographers et al. learned about theorika was phi 441 Philochorus (cf. under theta 219), and his convention seems to have been to leave the participle implicit (see the note at mu 801).