Without these the desire of knowledge cannot arise; and they are therefore known, from the very nature of the matter, to be necessary prerequisites. To sum up: The root of bondage is the unreal view of plurality which itself has its root in Nescience that conceals the true being of Brahman. Bondage itself thus is unreal, and is on that account cut short, together with its root, by mere knowledge. Such knowledge is originated by texts such as 'That art thou'; and work is of no help either towards its nature, or its origination, or its fruit (i.e. release). It is on the other hand helpful towards the desire of knowledge, which arises owing to an increase of the element of goodness (sattva) in the soul, due to the destruction of the elements of passion (rajas) and darkness (tamas) which are the root of all moral evil. This use is referred to in the text quoted above, 'Brâhmanas wish to know him,' &c. As, therefore, the knowledge of works is of no use towards the knowledge of Brahman, we must acknowledge as the prerequisite of the latter knowledge the four means mentioned above. THE SMALL SIDDHÂNTA. To this argumentation we make the following reply. We admit that release consists only in the cessation of Nescience, and that this cessation results entirely from the knowledge of Brahman. But a distinction has here to be made regarding the nature of this knowledge which the Vedânta-texts aim at enjoining for the purpose of putting an end to Nescience. Is it merely the knowledge of the sense of sentences which originates from the sentences? or is it knowledge in the form of meditation (upâsana) which has the knowledge just referred to as its antecedent? It cannot be knowledge of the former kind: for such knowledge springs from the mere apprehension of the sentence, apart from any special injunction, and moreover we do not observe that the cessation of Nescience is effected by such knowledge merely. Our adversary will perhaps attempt to explain things in the following way. The Vedânta-texts do not, he will say, produce that knowledge which makes an end of Nescience, so long as the imagination of plurality is not dispelled. And the fact that such knowledge, even when produced, does not at once and for every one put a stop to the view of plurality by no means subverts my opinion; for, to mention an analogous instance, the double appearance of the moon--presenting itself to a person affected with a certain weakness of vision--does not come to an end as soon as the oneness of the moon has been apprehended by reason. Moreover, even without having come to an end, the view of plurality is powerless to effect further bondage, as soon as the root, i.e. Nescience, has once been cut But this defence we are unable to admit. It is impossible that knowledge should not arise when its means, i.e. the texts conveying knowledge, are once present. And we observe that even when there exists an antagonistic imagination (interfering with the rise