Works of limited issue are sometimes, but not always, nor indeed often, of especial value. It has been the practice for some years among publishers to issue works on what is nothing more nor less than the old subscription plan; but, unlike the hungry poets of old, who trudged the streets taking the price of copies in advance, the publishers keep faith with their subscribers. The edition is limited to a given number of copies, after which the type is distributed, and the plates--if the work is illustrated--broken up. Many speculators in books have endeavoured from time to time to "corner" editions so limited in quantity, buying at the published price, and subsequently selling again at an increased amount. In this way considerable sums have been _lost_, for works published on this plan have a decided tendency to fall in the market, and when this is the case they seldom if ever recover their former position. Hogarth's works, published in 1822, by Baldwin and Cradock, is a very good example of this tendency. The work was originally issued at L50, and the impressions, taken from Hogarth's original plates, restored, however, by Heath, are consequently of full size. There is a secret pocket at the end containing three suppressed and highly indecent plates, which considerably add to the value. I myself have many a time seen this large and sumptuous book knocked down in the auction room at sums varying from L3 to L5, and once bought a good copy by private contract for L4 10s. Ottley's _Italian School of Design_ is another example. This work when on large paper, with proof impressions of the 84 tinted fac-similes of original drawings by Cimabue, Giotto, Guercino, and other famous painters, is worth about L3 by auction. The published price in 1823 was no less than L25 4s. The issue of each of these works was limited, but neither have succeeded in retaining its position in popular favour, and in all probability will decline still further in the market as time goes on. The lesson to be learned here is that such phrases as "only 100 copies printed," or "issue strictly limited to 50 copies," frequently to be observed in publishers' and auctioneers' catalogues, should be taken _cum grano salis_. The description may be accurate, but it does not follow that the limitation necessarily increases the value of the book. On the contrary, it may be well imagined that the publisher hesitated to launch the book entirely on its own merits, seeking rather the extraneous inducement of a "limited number". The earlier editions of Ruskin's works are an exception to the rule, for that author's reputation is deservedly great, and he is, moreover, master of his own books, which from choice he has, until the last year or two, preferred to render difficult of access.