Your reviewer commences with some general remarks, against which I have nothing to say. He then proceeds to consider my “requirements for a logical definition.” Here he seems to find a difficulty—which may be due to my not having expressed myself clearly. If so I hope he will read what I say on the same subject in my _Essay on Reasoning_, which I cannot believe he will find “indefinite” or not well “issuable.” But indeed I cannot see where his difficulty comes in with my old statement of the case. I state perfectly clearly that requirements (3) and (4) are not _logically_ necessary for a definition, but are only required if that definition is intended to give a _particular meaning_ to the word. He tries to reduce my argument _ad absurdum_ by giving a definition of “troft.” But so far from being absurd his definition is perfectly good. According to it “troft” would include in its denotation all our percepts and concepts. When however he goes on to say “... These significant names must be so used that the intellectual sensibility shall be excited to perceive that which is intended to be defined,” I differ from him entirely. This is only required for a _description_, not for a _definition_ (see _Essay on Reasoning_, p. 53). Your reviewer’s only solid objection to my “requirements” seems to be that the fourth includes all the rest. This is only true if the term proposed for definition has an import which has already been determined; but even in such a case it is better to consider the requirements separately, as I have given them. For the force of objections under the different headings varies enormously. An objection under heading (1), if established, would be fatal to any definition whatever. One under heading (2) so far from being fatal would only be a suggestion for the improvement of the definition. Objections under either of the headings (3) and (4) would only be to the effect that the term as defined meant something different from what it was desired that it should mean. It is however convenient to consider (3) and (4) separately as it would generally be possible to decide (3) at once, whereas if a doubt were raised under heading (4) it might lead to a prolonged discussion before it could be laid. I do not however pretend that the “requirements” are laid down in my _Foundations of Geometry_ in the best possible form. Indeed I have altered the form in my second essay. There is moreover one requirement for a logical definition which is not included in my heading (1) in the _Foundations of Geometry_, though it is included in (4). This defect is remedied in the _Essay on Reasoning_ (p. 55). It is curious that your reviewer should have missed this point, as it is the very one on which he attacks my definition of “direction.” It is that the assertions in a definition must not be _independent_ of the meaning of the term defined. If they were, the assertion would be equally true (or false) whatever meaning the term might have. The _import_ of the