tance <strong>the</strong> image appears silvery gray. In sharp contrast, <strong>the</strong> Fogg ink appears densely pigmented and thick with a more viscous binder so <strong>the</strong> lines appear blacker (fig. 12b). In some areas <strong>of</strong> shading, <strong>the</strong> lines are so close toge<strong>the</strong>r it looks as if lines were added or widened (fig. 13). Determining whe<strong>the</strong>r an area has been reworked requires careful scrutiny and counting lines. Sometimes, because <strong>of</strong> condition or printing flaws, comparison <strong>of</strong> <strong>the</strong> first state with multiple second-state impressions was necessary. During <strong>the</strong> course <strong>of</strong> research, it quickly became clear that <strong>the</strong> inks used in second-state impressions also varied and that <strong>the</strong> character <strong>of</strong> <strong>the</strong> ink was a key factor in <strong>the</strong> appearance <strong>of</strong> <strong>the</strong> print. 31
The light, grayish quality <strong>of</strong> <strong>the</strong> Cleveland impression has long been praised as "typical" and "representative" <strong>of</strong> good impressions <strong>of</strong> early Italian engravings, re- flecting a preference for s<strong>of</strong>ter tonal effects, more akin to drawings and unlike <strong>the</strong> strong black and white contrasts found in contemporary Nor<strong>the</strong>rn prints. 8 Thus <strong>the</strong> Cleveland print has been viewed as a closer representation <strong>of</strong> Pollaiuolo's original intention. 9 It would seem, however, given <strong>the</strong> range <strong>of</strong> inks used, that a greater vari- ety <strong>of</strong> impressions may have been acceptable than previously thought. "Relatively early" second-state impressions are characterized by little sign <strong>of</strong> plate wear overall, so that fine modeling lines still printed fairly well, and <strong>the</strong>re is evidence <strong>of</strong> good line embossment—that is, <strong>the</strong> lines in <strong>the</strong> plate were deep enough to create a notice- ably raised line when <strong>the</strong> dampened paper was pressed into <strong>the</strong> inked groove dur- ing printing. At least three such impressions, those in <strong>the</strong> Fogg, <strong>the</strong> Rijksmuseum, Amsterdam, and <strong>the</strong> Biblio<strong>the</strong>que Nationale, Paris, seem to have fairly dense, vis- cous black inks that result in high-contrast images where some interstices between background forms print nearly solid black. O<strong>the</strong>r relatively early impressions, such as those at <strong>the</strong> Yale University Art Gallery and <strong>the</strong> Museum <strong>of</strong> Fine Arts, Boston, have somewhat s<strong>of</strong>ter inks with a less viscous binder, which produces a less defined line that appears lighter, more sunk into <strong>the</strong> paper, and less densely black. The ink variation in <strong>the</strong> relatively early impressions is interesting because nearly all (Yale, Boston, Rijksmuseum, and Biblio<strong>the</strong>que Nationale among <strong>the</strong>m) share <strong>the</strong> same bird watermark—suggesting that <strong>the</strong>y could have been printed around <strong>the</strong> same time. So even within a group printed close in time, ink type, and hence appearance, could vary. The ink in some late impressions from <strong>the</strong> clearly worn and apparently corroded plate, such as those in <strong>the</strong> collection <strong>of</strong> <strong>the</strong> Philadelphia Museum <strong>of</strong> Art [4] and <strong>the</strong> Kupferstichkabinett, Berlin, also appears densely pigmented, but somewhat dry and thin though not "s<strong>of</strong>t." Such ink may have been used to maximize <strong>the</strong> ex- tent to which <strong>the</strong> remaining lines would print from <strong>the</strong> worn plate. Yet, o<strong>the</strong>r late examples such as that in <strong>the</strong> Kupferstichkabinett, Dresden, have a less dense ink, resulting in a somewhat grayer image, so again <strong>the</strong>re is variety in <strong>the</strong> inks used even later on. If <strong>the</strong> reworking <strong>of</strong> <strong>the</strong> plate in <strong>the</strong> second state was based on an aes<strong>the</strong>tic deci- sion to alter <strong>the</strong> image, and given that <strong>the</strong> ink makes a substantial contribution to an impression's appearance, <strong>the</strong> question arises: what was <strong>the</strong> motivation behind <strong>the</strong>se choices? First, one must address <strong>the</strong> time span between <strong>the</strong> states. Let us begin by considering <strong>the</strong> implications <strong>of</strong> <strong>the</strong> existence <strong>of</strong> a single example <strong>of</strong> <strong>the</strong> first state out <strong>of</strong> forty-eight known impressions, which might initially suggest that <strong>the</strong> first state is a pro<strong>of</strong>, a trial run before <strong>the</strong> plate was completed. However, <strong>the</strong>re are two signed woodcuts, copies created around 1490-1500 that are <strong>the</strong> same scale as Pollaiuolo's image and clearly reproduce <strong>the</strong> first state: one by Johanes de Francfordia [5] and <strong>the</strong> o<strong>the</strong>r by Lucantonio degli Uberti [6]. 10 The remarkably faithful version by Francfordia even follows <strong>the</strong> direction <strong>of</strong> shading on both foliage and figures, as well as o<strong>the</strong>r small details that are extremely difficult to emulate in woodcut. Francfordia copied <strong>the</strong> modeling <strong>of</strong> <strong>the</strong> ax man's thigh in <strong>the</strong> first state (fig. 14), and <strong>the</strong> scale <strong>of</strong> <strong>the</strong> figures is so close—within millimeters—that one could imagine an impression <strong>of</strong> <strong>the</strong> engraving was used to trace <strong>the</strong> image onto his woodblocks. Aside from <strong>the</strong> absence <strong>of</strong> crosshatching and zigzag lines, which would have been 32