The value of such praise was heightened, its apparent extravagance done away with, by the knowledge that dissatisfaction would be expressed in language equally unmistakable, and that either by the contributor or the editor the modifications which seemed to him desirable would be made. It was partly because he reserved to himself this power and accepted all the responsibility, that he insisted so strenuously on the anonymous character of the articles. But more even than that was his abhorrence of anything like 'log-rolling,' which, in his opinion, was inseparable from signed reviews. To the very last he discouraged, and indeed openly expressed his disapproval and dislike of the presumably inspired announcements of authors' names in the 'Athenaeum' or other journals. Here is an extract from a letter dated October 6th, 1891, which illustrates this objection:-- 'The only objection I have to the republication of articles with the name of the writer is that it destroys their anonymous character, which ought especially to be retained when they contain criticism of contemporaries.' So careful was he lest anything might warp the perfect fairness of criticism, which should 'nothing extenuate, nor set down aught in malice.' I, who write these lines, can say positively, after having written for the 'Review' under Reeve for upwards of twenty years, that in all that time I never received a hint or suggestion that any book should be dealt with otherwise than on its merits; and whilst engaged on this present work I have learned, for the first time, that men whose books I have reviewed, not always favourably, were personal friends of the editor. The following letter, addressed to Mr. T. N. Longman, is merely a concrete illustration of this:-- _December 26th_, 1891.--I thought it best to tell Froude frankly that the review of his book [Footnote: The Divorce of Catherine of Aragon,' in the _Review_ of January 1892.] in the 'Edinburgh' would be an unfavourable one. At the same time I disclaimed in the strongest language any disposition to make a personal attack on himself. Unfortunately he seems to ascribe adverse criticism of his works to personal animosity, which, in his case, is entirely wanting. It is a painful necessity. Froude and his book are too important to be passed over in silence. But the judicial character and consistency, and I may say honour, of the 'Review' absolutely require that the truth should be told about the book. I should consider it a derogation to my duty to the 'Review' if, from personal motives or affection, I suppressed an adverse criticism of a work which imperatively demands an answer. The independence of the 'Review' requires an independent judgement; but I expressly stipulated with the writer of the article that he should abstain from _bitterness_, which was carried too far in Goldwin Smith's article on the same subject in 1858. The 'Review' is pledged to the views already expressed on that occasion.