(M244) Secondly, we may distinguish the witnesses examined on oath. It is not clear that these were called by the same name. In the Code we read of _šîbi mûdi_, “the witnesses that know,” who seem to resemble very closely the Greek _Histores_. These, of course, were usually not on the jury. They testified, and were chosen by the parties to the suit. But the judge might examine persons who, in his opinion, would know. He selected and sent for them, directing the parties whom to produce. He might even adjourn the case for the production of witnesses.(99) (M245) Thirdly, we may distinguish the witnesses to a document. Very often we can discern that these had an interest in the case. They might be relatives of the parties, neighbors of the estate in question, officials whose rights were concerned. In later times they received the special name of _mukinnu_, “the establishers.” They may be presumed to have known at least the general purport of the deed which they witnessed. When the deed was called in question, they would be cited to state what they knew. In the case of legal decisions, both judges and jury occur as witnesses in this sense. Hence, in a great many cases the distinctions drawn above do not hold. Whether the term _šîbu_ was ever applied to the third class is doubtful. Their names are usually preceded by the sign which means “before,” however it was read. (M246) 4. *Cases of dispute settled out of court.*—When parties disagreed, they might discuss their difference between themselves and arrive at an agreement. Then they procured a scribe, who embodied the agreement in a binding compact, _duppu lâ ragâmi_. This took the form of a contract, the parties mutually undertaking not to withdraw from the agreement, re-open the dispute, or bring legal action, one against the other. To give sanction to this agreement, they swore by the gods and the king. Witnesses were called upon to be cognizant of and attest the contract; and their names were added to the contract. To authenticate their names both parties and witnesses often impressed their seals or, in default of seals, made a nail-mark. The date was then added. Each party seems to have taken a copy of the agreement and the scribe held a third, or deposited it in the archives. Such cases may be said to have been settled “out of court.” At any rate they contain no reference to a judge, or court. But it is possible that the administration of the oath was a judicial, or perhaps a sacerdotal function. Further, the witnesses may have been drawn from a body of men held in readiness at court to perform that function. It is certain in some cases, that agreements arrived at independently were taken to a judge for confirmation,(100) and the Code expressly directs some cases to be taken to a judge. But it is probable that many cases were settled by mutual agreement.