4019

The compiler must have noticed these difficulties, for he made an attempt to remove them. In his preface he adopted the Abbot’s version, but he modified it in one particular. The painter, he says, signed a bond with the Devil in 1669 in ink, but afterwards (‘deinde vero’) in blood. He thus overrode the express statement of both reports that one bond was signed in 1668, and he ignored the Abbot’s remark in his deposition to the effect that there was a difference in the year-number between the two bonds. This he did in order to keep in harmony with the dating of the two documents that were given back by the Devil.

In the Abbot’s deposition a passage appears in brackets after the words ‘sequenti vero anno 1669’. It runs: ‘sumitur hic alter annus pro nondum completo, uti saepe in loquendo fieri solet, nam eundem annum indicant syngraphae, quarum atramento scripta ante praesentem attestationem nondum habita fuit.’ This passage is clearly an interpolation by the compiler; for the Abbot, who had only seen one bond, could not have stated that both bore the same date. The placing of the passage in brackets, moreover, must have been intended to show that it was an addition to the text of the deposition. It represents another attempt on the compiler’s part to reconcile the incompatible evidence. He agrees that the first bond was signed in 1668; but he thinks that, since the year was already far advanced (it was September), the painter had post-dated it by a year so that both bonds were able to show the same year. His invoking the fact that people often do the same sort of thing in conversation seems to me to stamp his whole attempt at an explanation as no more than a feeble evasion.