The combined use of a detailed inventory of Augustine's citations of St John's Gospel, with a commentary (p. 187–364), and a list of the distinctive readings of Jerome's revision (to become the Vulgate), according to the definition given on p. 106, allows us to delineate an evolution. Comparison may prove difficult for a number of reasons: (1.) Jerome's work on the Gospels, a revision, was in the fifth century only an additional form to the many divergent witnesses of the Old Latin; (2.) Its text is sometimes mixed with Old Latin readings; (3.) Late witnesses of the Old Latin are influenced by the Vulgate; and (4.) Editions of Augustine's works are not always above suspicion. The author's method of distinguishing valuable readings is sound. Augustine first met Italian and other local African Old Latin texts. From 403 onwards he discovered, and increasingly used, especially in his Tractatus in Iohannis Evangelium, Jerome's new translation (p. 157). Houghton distinguishes three periods: pre-403, 403–19, post-419. In the last period, Augustine's ‘mental text’ is no longer the Old Latin but Jerome. The evidence is confined to John. These conclusions could be presumably extended to the Gospels as a whole, but certainly not to the rest of the Bible, although for some books Augustine knew and quoted Jerome's translations of the Septuagint (Job and Psalms) and Hebrews. For other parts of the New Testament, Augustine did not use the type of text which would become the Vulgate (p. 84). It is a pleasure to observe that Bonifatius Fischer has found a capable reader of his Varianten zu Johannes (1991). Houghton, a research Fellow in theology at the University of Birmingham, has made a firm and longstanding contribution to Augustinian scholarship and to the history of the Latin Bible. History needs philology. As the preface indicates (p. ix), there is never an end to such work.