This inference is confirmed by what we know of Ostade’s work on canvas. His first period dates from 1630 to 1635; then follows a middle period in which, influenced by Rembrandt, he adopted a warmer scheme of colour; lastly, in a third period, he began to repeat himself and decline. Beyond such general deductions it does not seem worth while to go. In Rembrandt’s case the question of chronology is of extreme interest and significance, but in Ostade there is no development to speak of, and to labour after exhibiting it would be waste of time. Next, as to the various states of the etchings. The reverence for first states and rare states, common to collectors, has from their point of view its own justification; but they are apt perhaps sometimes to confuse the æsthetic value of a print with its market value. Artists, on the other hand, are sometimes prone to dismiss the whole question of states as tedious and absurd. It is, however, of great importance that the etcher should be judged on his own merits and not on the merits, or demerits, of other people. Ostade undoubtedly made alterations in his plates during printing and thus created “states”; but many more states were created after his death by other hands re-working the worn copper. It is reasonable to suppose that the last state touched by the artist is the one that he would wish to be taken as typical of his perfect work. But the question arises: Which is the last state touched by the artist? The work of later hands, added to a plate after the artist’s death, does not concern us; but the development of the etching up to that state when the artist leaves it as a finished thing, must interest us greatly. How are we to decide? In the case of Ostade, we are helped a little by external data. As we [Illustration: _Fig. 8.--Peasant paying his Reckoning. By Ostade. B. 42._] have seen, the plates were sold at his death in 1685. We know also that they were sold again by their new possessor, Dirk van der Stoel, Ostade’s son-in-law, in 1686; and eight years later again, in 1694. What state they were in then we can only conjecture: but we may infer something from what we know to have been their state in 1710 or a little later.