'A new sword on an old anvil': W.B.Yeats, Robert Graves and the Anglo-Irish Tradition

Throughout his writing life, Robert Graves was consistently and often publicly hostile to the work of W.B. Yeats, whilst still also owing a considerable debt to the older poet (who he never met). This essay explores Graves' complex responses to Yeats, arguing that his antagonism may be understood in the light of his own Anglo-Irish background, and is implicated in his relations with his father, Alfred Perceval Graves, as well as his experience of the First World War. Probing the suggestiveness of Graves's claim in 1959 that his poems 'remain true to the Anglo-Irish poetic tradition into which I was born', it traces the relation between Yeats and Graves through correspondence, critical writings, and through a comparative reading of Yeats's A Vision and Graves's The White Goddess, and reveals underlying similarities in their critical and mythological thinking in spite of Graves's public disavowal of the Yeatsian aesthetic.

Comparison of two polyclonal antibodies for the detection of 19-nortestosterone in bovine bile by ELISA

19-Nortestosterone (beta-NT) is banned for use as a growth promoter in food animals within the European Union. For regulatory control purposes, urine and bile samples are routinely screened by immunoassay. The aim of the present study was to compare the ability of two immunoassays, using two rabbit polyclonal antibodies raised against two different NT derivatives, to detect NT residues in bovine bile. One antiserum cross-reacted with both alpha-NT and beta-NT (alpha/beta-NT), whereas the other was specific for alpha-NT. Bile samples from 266 slaughtered cattle were deconjugated and analyzed using both antibodies, with all screening positives (>2 ng ml(-1)) confirmed by high resolution gas chromatography mass spectrometry. The alpha/beta-NT and alpha-NT antibody-based ELISAs screened 39 and 44 samples positive, respectively, with NT confirmed in 22 and 39, respectively. The alpha/beta-NT antibody-based ELISA produced a false-negative rate of 44% compared to 0% for the alpha-NT antibody-based ELISA. Supplementary investigations concluded that a matrix effect was a major cause of the marked differences in false-negative rates. This result underlines the necessity to validate immunoassays in the sample matrix.