Coster himself was not heard of as a printer till about a hundred years after he was supposed to have printed, when Junius wrote in his _Batavia_ the wonderful legend of the letters cut in beech bark. That a person called Lourens Janszoon lived at Haarlem from 1436 to 1483 seems to be an established fact; but, at the same time, all the entries and notices relating to him show that he was a chandler or innkeeper. Von der Linde very justly, therefore, considers he was not a printer; and this view is certainly reasonable, for we can hardly suppose that a man could have printed all the so-called Costeriana and at the same time have attended to his business so carefully, that all the entries which relate to him speak of him only as an innkeeper, and no mention of any kind is made of him as a printer, though he was, so believers in him assert, the only printer in Holland for thirty years. Coming to the books themselves, what do we find? The first printed date is 1473, in which year books were issued at both Utrecht and Alost. M. Holtrop mentions that the Hague copy of the _Tractatus Gulielmi de Saliceto de salute corporis et animæ_ and _Yliada_ was bought by a certain Abbat Conrad for the library of his house; and as the Abbat in question was Abbat only from 1471 to 1474, the book cannot have been printed later than 1471-74; and this and the rubricated 1472 in the Darmstadt copy of the _Saliceto_ are at present the only dates which we can use for our purposes. There are, however, a large number of fragments of books known, printed in a rude type and with the appearance of early printing, all of which are frequently asserted to have been printed before 1473. These books, consisting for the most part of editions of the _Donatus_ or the _Doctrinale_, are known by the name of Costeriana, as being the supposed productions of Coster. Among them also are the four editions of the _Speculum_, which we have examined at length in Chapter I. Fragments of at least fifty books or editions are known, which may be separated by their types into eight groups. Concerning the types Mr. Hessels says: ‘Type 2 is inseparably connected with type 1; and as the former is so much like type 3 that some consider these two types identical, nothing would be gained by separating them. Type 4 and 5 occur in one and the same book; and as certain letters of type 5 are identical with some of type 3, they may all be linked together. Type 6 is identical with type 5 except the P, which is larger and of a different form. Types 7 and 8 are linked on to the types 1-6, on account of the great family-likeness between them, they all having that peculiar perpendicular stroke to the cross-bar of the _t_, and a down stroke or curl attached to the _r_, which is found in no other types of the Netherlands.’ [Illustration: PAGE OF A “DOCTRINALE.” (_One of the so-called “Costeriana.”_)]