While the forty-seven books show a considerable number of varieties of type, the editions being identified by these type differences, all the type faces show a strong family resemblance. They are designed from a common model, but not at the same time, and consequently they show marked resemblances and marked differences. The question may be asked why the same printer should use eight or nine different fonts of type for only forty-seven books. The answer is found in the fact that type-making was as yet in an experimental stage and that durable material had not yet been found for that purpose. When we come to the discussion of evidence of another class we shall find confirmation of this. There is no evidence of the second or third class connecting early printing with any Dutch town except Haarlem. There is, however, important evidence of the other classes which does connect printing with Haarlem. There are not, however, forty-seven different works. Twenty of the forty-seven books are different editions of the Donatus, that is to say they are Donatuses showing such typographic differences as to show that no two of them could have been printed from the same type form. Four of them are editions of the Speculum and eight are different editions of the Doctrinale. The Doctrinale was a brief compend to Christian doctrine approved by the church and widely circulated among the faithful. Nearly all of the fragments of these forty-seven books have been found in Haarlem or in the neighborhood. It is evident that the publications of this press, whatever its date, were locally sold and that neither its fame nor its product went far from the place of production. Having thus shown the reasons for believing that these forty-seven pieces of early printing came from Haarlem, let us see what they have to say for themselves as to the time of their production. It has already been pointed out that a small group of the best of them dated themselves no later than 1474, as is shown by their contents. So far as the contents themselves are concerned we have nothing to date the others. There are certain things about the books themselves, however, which show that their production must have begun long before 1474.