The headword is a word of non-Greek origin which has many variant spellings (including kandu/lh ka/ndaulos, etc.; see LSJ s.v. ka/ndaulos). The notion (in Athenaeus: below) that it is Lydian may arise from the mere similarity to the name of the famous Lydian king Kandaules. For alternative recipes see Athenaeus, Deipnosophists 4.172A-C [4.72 Kaibel] = Menander fr. 451 Sandbach; Pollux 6.69; Hesychius kappa648. At least one ancient authority would read ka/ndulon in place of ko/ndulon ('knuckle') at Aristophanes, Peace 123 (Demetrius (?)of Phaleron; cf. scholia ad loc.), and some of the commentary may arise from that source.
cf. generally theta 489, omicron 251.
[1] alpha 103.
[2] Very similar entries in Etymologicum Gudianum 297.22 and Etymologicum Magnum 488.54 (s.v. kandu/lh). The second part is unintelligible in all three versions, but in different ways: Etymologicum Gudianum has "and sour-sauce qroiko\n food", and Etymologicum Magnum has "as also *)Amoita/khn e)/roikon food". In place of the nonsensical words in these two sources, and the Suda's incongruous 'restraining' (e)ru/kwn), we should probably read karukw=n ('of blood-sauces'); cf. generally kappa 435, kappa 436. Plutarch, Quaestiones Conviviales 664A (cf. 644B) mentions kandu/lai, a)burta/kai ('sour-sauces') and karu/kai ('blood-sauces') together as exotic and luxurious dishes. Compare also Menander fr. 397 Körte-Thierfelder, now 351 Kassel-Austin, where both blood sauces and ka/ndaulos are classified as u(pobinhtiw=nta brw/mata ('sex-stimulating foods') the latter worthy of a *)iwniko\s plou=tac ('rich Ionian twit'). The present entry, therefore, should have meant something like 'kandylos is a dish etc.; as also sour-sauce [made] of blood-sauces [is] a dish'.
[3] alpha 1001.