What is the bearing of this account upon the "empirical datum"? Just this: The correlation of correlative series of changes which defines the object of sense perception is in no sense an original historic or psychologic datum. It signifies the result of an analysis of the usual crude empirical data, and an analysis which is made possible only by a very complex knowledge of the world. It marks not a primitive psychologic datum but an outcome, a limit, of analysis of a vast amount of empirical objects. The definition of an object as a correlation of various subcorrelations of changes represents a great advance--so it seems to me--over the definition of an object as a number of adjectives stuck into a substantive; but it represents an improved definition made possible by the advance of scientific knowledge about the common-sense world. It is a definition not only wholly independent of the context in which Mr. Russell arrives at it, but is one which (once more and finally) assumes extensive and accurate knowledge of just the world professedly called into question. II I have come to the point of transition to the other part of my paper. A formal analysis is necessarily dialectical in character. As an empiricist I share in the dissatisfaction which even the most correct dialectical discussion is likely to arouse when brought to bear on matters of fact. I do not doubt that readers will feel that some _fact_ of an important character in Mr. Russell's statement has been left untouched by the previous analysis--even upon the supposition that the criticisms are just. Particularly will it be felt, I think, that psychology affords to his statement of the problem a support of fact not affected by any logical treatment. For this reason I append a summary statement as to the facts which are misconstrued by any statement which makes the existence of the world problematic.