as a dredge collects marine specimens.  Nothing came amiss to them.  Wood, ivory, silver, bronze, marble, plaster—­they repudiated no material or period.  Stuffs, glass, pictures, porcelains, potteries—­it was all one to them so the object were old and rare.  Inevitably, then, they had come to primitive pots, and simultaneously, for they not only watched each other closely, but almost read each other’s minds.  And when they came to primitive pots it was certain that they would beg, borrow, or steal a well, since in old wells, and cisterns, besides less mentionable places, primitive pots abide.  Many pots were there, as we shall see, from the first, and the maids and children of the centuries, by way of concealing breakages, have usually made notable secondary contributions.  So when amiable Stanton Mayhew freely conceded a most ancient well to Cleghorn and Webb, it was like receiving Pandora’s box, with the difference that the well might safely be opened.

Here had ensued a most delicate negotiation concerning the division of the spoil.  A mathematical partition of the fragmentary material that an old Italian well contains is extremely difficult if at all possible.  After much debate it was agreed that after they struck pay dirt, each should dig in turn, each to have the bucketful that came under his trowel or fingers.  Scattered fragments of the same pot and other complications were to be adjudicated by Mayhew, whose ignorance and disinterestedness were safe to assume.  But the well gave up quantities of noncontentious matter before Mayhew’s services were required.  The first five feet had revealed nothing but fragments of kitchen pottery of our time and a fairly perfect hoopskirt of Garibaldian date.  A little lower had emerged the skeleton of a cat.  Similar tragedies were in evidence, on an average, at every quarter century of depth.  Between the second and third cat, lay Ginori imitations of Sevres and Wedgewood, scraps too of gilded glass—­the earnest of better things below.  Five cats down, some eighteenth-century apothecary pots, damaged but amenable to repair, had inaugurated the alternation of buckets under the agreement.  It were tedious to follow the ascending scale of excellence as the digging went deeper.  Enough to say that below the mixed ingredients and the nethermost cat they found a homogeneous layer of beautiful fourteenth-century shards, affording many buckets full, and promising delicate adjudication to the referee.

Before the lustred pots themselves shed a baleful gleam over this narrative, something should obviously be said about Italian wells and why they contain pots.  Beyond those casually acquired from careless or secretive servants, there is, if the well be old and of good make, a certain number of intact pieces put in to serve as a filter.  Often a group of pitchers or similar crocks is imprisoned between the two bottom-stones.  Sometimes there are two such layers.  After this filter had been made there was frequently scattered a bushel or more of small shards above.  From these by careful sorting complete or nearly complete pieces may be recovered.  Through all this mass of whole or broken pottery the water had to find its way up, for the cement sides of an Italian well are watertight.  Thus, barring the indiscretions of housemaids and cats, the early Italians drank pure water.