By what steps Berkeley was led, and has since led so many after him, into so grave an error, he has himself acquainted us. Thus it is that he argues: By sensible things can be meant only such as can be perceived immediately by sense: and sensible _qualities_ are of course sensible _things_. But the only perceptions of sense are sensations, and all perceptions are purely mental. Wherefore, sensible qualities being, as such, perceptible immediately by the senses, must be sensations, and being sensations must be perceptions, and being perceptions they are of course purely mental, and existent nowhere save in the mind. Carefully, however, as Berkeley fancied he was picking his way, he really had tripped, and that fatally, at the second step. He calls the qualities of objects _sensible_ things; but sensible they are not according to his definition, for they are not capable of being immediately perceived by the senses. It is not sense which perceives, but reason which infers them. The senses, as Berkeley elsewhere repeatedly and earnestly insists, receive nothing from objects but sensations, and these they communicate to the mind without accompanying them by the slightest hint as to whence they originally came. The senses suggest nothing as to any qualities resident in or appertaining to an object corresponding with the sensations derived from the object. The existence of such qualities is an inference of reason which, taking for granted that sensations, in common with all other occurrences, must have causes, and observing that certain of them commonly occur in the presence of certain objects, and never occur in the absence of those objects, infers that the causes of the sensations must exist in the objects. To the causes thus inferred the name of qualities is given, to distinguish them from the sensations whereof they are causes; and the Berkeleian transgression consists in overlooking the distinction between things so diametrically opposite.