But in spite of its inherent improbabilities, M. Halévy succeeded in making many converts to his theory, including Prof. Friedrich Delitzsch and a number of the younger school of German Assyriologists. More conservative scholars, such as Sir Henry Rawlinson, M. Oppert, and Prof. Schrader, stoutly opposed the theory, maintaining that Sumerian was a real language and had been spoken by an earlier race whom the Semitic Babylonians had conquered; and they explained the resemblance of some of the Sumerian values to Semitic roots by supposing that Sumerian had not been suddenly superseded by the language of the Semitic invaders of Babylonia, but that the two tongues had been spoken for long periods side by side and that each had been strongly influenced by the other. This very probable and sane explanation has been fully corroborated by subsequent excavations, particularly those that were carried out at Telloh in Southern Babylonia by the late M. de Sarzec. In these mounds, which mark the site of the ancient Sumerian city of Shirpurla, were found thousands of clay tablets inscribed in archaic characters and in the Sumerian language, proving that it had actually been the language of the early inhabitants of Babylonia; while the examples of their art and the representations of their form and features, which were also afforded by the diggings at Telloh, proved once for all that the Sumerians were a race of strongly marked characteristics and could not be ascribed to a Semitic stock. The system of writing invented by the ancient Sumerians was adopted by the Semitic Babylonians, who modified it to suit their own language. Moreover, the archaic forms of the characters, many of which under the Sumerians still retained resemblances to the pictures of objects from which they were descended, were considerably changed. The lines, of which they were originally composed, gave way to wedges, and the number of the wedges of which each sign consisted was gradually diminished, so that in the time of the Assyrians and the later Babylonians many of the characters bore small resemblance to the ancient Sumerian forms from which they had been derived. The reading of Sumerian and early Babylonian inscriptions by the late Assyrian scribes was therefore an accomplishment only to be acquired as the result of long study, and it is interesting to note that as an assistance to the reading of these early texts the scribes compiled lists of archaic signs. Sometimes opposite each archaic character they drew a picture of the object from which they imagined it was derived. This fact is significant as proving that the Assyrian scribes recognized the pictorial origin of cuneiform writing, but the pictures they drew opposite the signs are rather fanciful, and it cannot be said that their guesses were very successful. That we are able to criticize the theories of the Assyrians as to the origin and forms of the early characters is in the main due to M. de