=15. RESPONDET= introduces the agricultural image of 18 'sed siccum sterili uomere litus aro', for the word here means 'yield'. _OLD_ _respondeo_ 8c cites for the literal sense Virgil _G_ II 63-64 'truncis oleae melius, propagine uites / respondent', Columella II 1 3 'humus ... magno faenore ... colono respondet', Col III 3 4; for a transferred use see Sen _Ep_ LXXXI 1 'non respondeant [_sc_ beneficia] potius quam non dentur'. =16. SICCVM ... LITVS ARO.= Proverbial for a useless activity. See Otto _harena_ 4 and compare _Tr_ V iv 47-48 'plena tot ac tantis referetur gratia factis, / nec sinet ille [Ovid] litus arare boues'. _Sterili_ is transferred by hypallage from _litus_; _siccum_ serves no purpose beyond providing a balancing epithet. =17. VENAS EXCAECAT=, the reading of most codices, is obviously correct as against the VENAS CVM CAECAT of _BCHL_. Ovid uses _excaecare_ again at _Met_ XV 270-72 'hic fontes natura nouos emisit, at illic / clausit ... flumina prosiliunt aut _excaecata_ [_uar_ exsiccata] residunt'. =17. IN VNDIS= is probably corrupt; if it is retained, from the context it must mean 'in the water of springs' (Professor A. Dalzell). Williams suggests 'in the case of water', marking the analogy with _pectora sic mea sunt limo uitiata malorum_ in 19. For _undis_ as a corrupt hexameter ending, compare _Met_ XV 276 'redditur Argolicis ingens Erasinus in aruis [_codd_: in undis _Sen_ NQ _III 26 4_]', _Met_ VIII 162 'liquidus Phrygiis Maeandros in aruis [_uar_ liquidis Phrygius ... in undis]', and _Met_ XIV 155 'sedibus Euboicam Stygiis emergit in urbem [_uar_ sedibus euboicis stigiis emersus ab undis]'. The line seems to have passed without comment until Merkel's second edition: '_in undis_ minus bene positum uidetur; temptabam _hiulcas,_ quod expressisset Statius Theb. VIIII 450 _hiulcis flumina uenis Suggerit_ ['he (the river Asopos) opens his springs wide and adds his streams']'. There seems no obvious reason, however, for Ovid to define the springs as 'gaping'. Madvig conjectured INVNDANS, the corruption of which would be easy; but _uenas_ seems more in need of a modifier than _limus_--Professor R. J. Tarrant suggests APERTAS or AQVARVM, Professor A. Dalzell IN ARVIS. Professor Tarrant also suggests to me that _in undis_ could well have originated as a gloss on _uenas_. =18. LAESAQVE.= There seems no reason to replace this with Merkel's LAPSAQVE ('flowing back'?), which even seems to contradict the sense of _resistit_. The same sense of _laesus_ at _Am_ III vii 32 'deficiunt laesi carmine ['spell'] fontis aquae'. =20. VENA PAVPERIORE.= The same image of Ovid's poetic talent at _Tr_ III xiv 33-34 'ingenium fregere meum mala, cuius et ante / fons infecundus _paruaque uena_ fuit' and _EP_ II v 21-22 'ingenioque meo, _uena_ quod _paupere_ manat, / plaudis, et e riuo flumina magna facis'.