died in A. H. 257. It is still extant and is the most respected of all the collections of traditions. The principle of arrangement in it is legal; that is, the traditions are classified in these chapters so as to afford bases for a complete system of jurisprudence. Al-Bukhari was a strong opponent of speculative law and his book was thus a protest against a tendency which, as we shall see later, was strong in his time. Another point in which al-Bukhari made his influence felt and with greater effect, was increased severity in the testing of traditions. He established very strict laws, though of a somewhat mechanical kind, and was most scrupulous in applying them. His book contains about seven thousand tradition, and he chose those, so at least runs the story, out of six hundred thousand which he found in circulation. The rest were rejected as failing to meet his tests. How far the forgery of traditions had gone may be seen from the example of Ibn Abi Awja, who was executed in A.H. 155, and who confessed that he had himself put into circulation four thousand that were false. Another and a similar Sahih is that of Muslim, who died in A.H. 261. He was not so markedly juristic as al-Bukhari. His object was rather to purify the mass of existing tradition from illegitimate accretions than to construct a basis for a complete law code. He has prefixed a valuable introduction on the science of tradition generally. In some slight details his principle of criticism differed from that of al-Bukhari.

These two collections, called the two Sahihs—as-Sahihan—are technically jami's, i.e. they contain all

the different classes of traditions, historical, ethical, dogmatic and legal. They have also come to be, by common agreement, the two most honored authorities in the Muslim world. A believer finds it hard, if not impossible, to reject a tradition that is found in both.

But there are four other collections which are called Sunan—Usages—and which stand only second to the two Sahihs. These are by Ibn Maja (d. 303), Abu Da'ud as-Sijistani (d. 275), at-Tirmidhi (d. 279) and an-Nasa'i (d. 303). They deal almost entirely with legal traditions, those that tell what is permitted and what is forbidden, and do not convey information on religious and theological subjects. They are also much more lenient in their criticisms of dubious traditions. To work exclusion with them, the rejection needed to be tolerably uuanimous. This was required by their stand-point and endeavor, which was to find a basis for all the minutest developments and details of jurisprudence, civil and religious.

These six books, the two Sahihs and the four Sunans, came to be regarded in time as the principal and all-important sources for traditional science. This had already come about by the end of the fifth century, although even after that voices of uncertainty continued to make themselves heard. Ibn Maja seems to have been the last to secure firm footing, but even he is included by al-Baghawi (d. 516) in his Masabih as-sunna, an attempted epitome into one book of what was valuable in all. Still, long after that, Ibn Khaldun, the great historian (d. 808), speaks of five fundamental works; and others speak of