The Indus script has been known for a little over 40 years, and it has so far resisted all attempts at decipherment. This is not surprising in view of the fact that the inscriptions are all very short, and that no long connected texts are preserved. It is also not known what language lies behind the writing: bilinguals are nonexistent, and given the historical circumstances, not likely to be found. In the period since the discovery of the inscriptions a number of attempts at decipherment has been made, but none has carried conviction. Quite recently there has been a number of fresh attempts, of which the most sustained and ambitious, by a group of Finnish scholars, has been described in the previous issue of this journal. It is based on the assumption that the language spoken in the Indus cities was a form of Proto-Dravidian. The detailed methods used in this attempt at decipherment have been briefly described in the previous articles, and they need not be repeated here. Instead it is proposed to examine the handling of the Dravidian material, which, as will be seen, is open to considerable criticism