It was written by him in 1737, the year of Chauvelin's disgrace. It is upon this work or upon notices of it that d'Argenson's reputation as a political thinker is principally based. It is necessary to examine it carefully, and in the light of the circumstances under which it was composed. The author would turn in his grave could he hear the verdict which has generally been passed upon it; for if there was one imputation which d'Argenson abhorred and strove to avoid, it was that of being a visionary possessed by chimeras. The consideration of this book is not so easy as it at first appears. It is less known than it certainly deserves to be, and we are dependent for our knowledge of it upon a couple of old editions, published in 1764[415] and 1784. The work consists of a number of dissertations on the past history and present condition of the French Government, which serve as the setting for its central feature, a Plan or project of reform. In the preface to the edition of 1784, which was issued by the d'Argenson family, the previous publication is noticed, and we are told that it is grossly imperfect, and that unwarrantable liberties have been taken with the author's text. Upon consulting the earlier edition, the discrepancies are at once apparent, and nowhere so noticeably as in the very Plan which is the most important section of the work. To begin with, the Plan which, in 1784, is given in the form of a royal proclamation in thirty-four articles, appears in the issue of 1764 as an informal draft consisting of fifty-two. And the difference is not of form alone; for on comparing the two Plans we find that they are as far apart as are the poles asunder, and that the second might represent, at the ordinary rate at which history proceeds, a century's advance in breadth of thought. In some bewilderment we turn to the four manuscripts preserved in the Library of the Arsenal;[416] and confusion is further confounded by the discovery that it is with the very Plan of this maligned edition of 1764 that that of the manuscripts coincides. The manuscripts themselves are clearly authentic: for they bear annotations in d'Argenson's handwriting. A question at once arises as to the authenticity of the scheme of 1784. A comparison of the texts establishes the fact that in spite of the impeachment of one editor and the apologies of another, both editions are equally genuine, and represent different stages of the work. Moreover it is evident that that of 1784 is mainly a revised and enlarged version of the original, and was probably executed about the year 1750; and further, that in it the editor, without a word of explanation, has incorporated a Plan, a Chapter and a Conclusion, of much later date and with a very different purpose.