their death. Still the superior check and authority of a record must in practice have gradually superseded reliance on unassisted memory. Collections of the earlier traditionists fell sometimes into the hands of later authors, and we find Wâckidi and others making use of these treasures in a manner inconsistent with the canons of the Sunna. 

Sprenger states the following as the successive stages of record: (1) Notes or memoranda; (2) School or college collections; (3) Regular books. Our previous remarks refer exclusively to the first, that is, notes professing to be used simply for the refreshment of the memory. Towards the end of the first century, the second class, or School collections, began to be in vogue. Orwa and Zohri, for example, used such records in their prelections. The pupils were at liberty either to trust solely to their memory, or to make copies of their Master's collection ; but so rigidly was the oral canon still followed, that the copies thus taken had no authority until they were first rehearsed by the scholar in the hearing of his Master; and the date of each rehearsal (árz) was usually noted upon his manuscript by the copyist.¹ 

The third class of documents, answering to our published Books, was of much later rise. A Mahometan authority tells us that Ibn Jureij and Ibn Abi Rabia, who both died about the middle of the second century, were the first who wrote books. Mussulman writers themselves understand this passage as meaning that these persons were the first to make use of manuscript tradition in any shape. But this appears a mistake: the simple purport being that these were the first to put forth "Books," or collections of tradition, which carried their own authority with them, the condition of oral repetition being no longer required. It had become a question of accuracy of manuscript and edition; no longer pure accuracy of recollection. 

The use of books gradually displaced the old and cumbrous