Among the followers of non-self-revealing theory the supporters of the Bhâtta Mimâmsâ claim that a cognition is inferable. Now we are going to discuss their view in a general way. According to this view a cognition, such as the cognition of a pot, reveals the object pot, but not the cognition of a pot. The cognition of a pot is revealed by an inference. The probans (hetu) of a cognition, according to this view, is the property of being cognised (jMtata). When an object such as a pot is revealed, we come to know the pot. Hence the property of being cognised is produced in the object pot, and this is the property of a known object. This property of being cognised is the probans by means of which we infer a cognition (jnana). According to this view the property of being cognised (jnatata) is not the same as the property of being the object (visayata) of the Nyâya and some other systems who have explained it in terms of the relation between a cognition and its object. For, this type of property of being the object which is cognised cannot be used as a probans to infer a cognition. The cognition of the probans is necessary prior to the inferential cognition. If the cognition of the probans is dependent upon the cognition of the probandum, then we cannot infer the probandum, because the probans cannot be cognised unless the probandum is cognised and the probandum cannot be cognised unless the probans is cognised. If the property of being the object (visayata), which is a relation between a cognition and its object, is considered as the probans, then the cognition of the probans is dependent upon the cognition of the probandum which is a cognition in this case and one of the relata of this relational property.1 In other words, the cognition of the probans is dependent upon the cognition of the probandum, because the cognition of a relation presupposes the cognition of its two relata.2 For example, we cannot cognise the relation of conjunction between the pot and the floor unless we have