retain more traces of primitive popular thought than other divisions of the religious literature with the exception of the omens. The remodeling to which they were subjected did not destroy their original character to the extent that might have been expected--a circumstance due in the first instance to the persistency of the beliefs that called these texts forth. Many of the texts containing incantations were found by the modern explorers in so mutilated a condition, that one can hardly hazard any generalizations as to the system followed in putting the incantations together. From the fact, however, that in so many instances the incantations form a series of longer or shorter extent, we may, for the present at least, conclude that the serial form was the method generally followed; and at all events, if not the general method, certainly a favorite one. Deviating from the ordinary custom of calling the series according to the opening line of the first tablet, the incantation texts were given a distinct title, which was either descriptive or chosen with reference to their general contents. So one series which covered at least sixteen tablets was known by the very natural name of the 'evil demon'; the incantations that it contained being intended as a protection against various classes of demons. Another is known as the series of 'head sickness,' and which deals, though not exclusively, with various forms of derangements having their seat in the brain. It covered no less than nine tablets. Two others bear names that are almost synonymous,--"Shurpu" and "Maklu," both signifying 'burning,' and so called from the chief topic dealt with in them, the burning of images of the sorcerers, and the incantations to be recited in connection with this symbolical act. The "Maklu" series embraced eight tablets and contained, according to Tallqvist's calculations,[341] originally about 1,550 lines, or upwards of 9,000 words. The "Shurpu" series, although embracing nine tablets, appears to have been somewhat shorter. In view of the extensive character of these series we are justified in speaking of incantation 'rituals.' The texts were evidently prepared with a practical purpose in view. The efficacy of certain formulas having been demonstrated, it was obviously of importance that their exact form should be preserved for future reference. But a given formula was effective only for a given case, or at most for certain correlated cases, and accordingly it became necessary to collect as many formulas as possible to cover all emergencies. The priests, acting as exorcisers, would be the ones interested in making such collections, and we may assume, as already suggested, that each temple would develop a collection of its own,--an incantation code that served as a guide for its priests. The natural tendency would be for these codes to increase from generation to generation, perhaps not rapidly, but steadily. New