REVIEWS 403
remarkable publications are always indicated by detailed and favorable notice ; as for the publications detestables, they have been purely and simply passed by in silence."
The policy announced in the introduction seems to have been in general carried out through the book. No one would ever suspect, from consulting the lists, that such persons as Power, Sabine, Petzholdt, Valine, Ottino, and Fumagali had lived and had written bibliographies of bibliographies. This policy must be responsible for the omission from the lists of bio-bibliographical works of Michaud's Biographic universel and the inclusion of Hoefer's Nouvelk biographie generale. Can it also be responsible for the omission of Sabine's Bibliotheca Americana, of Laior's Cyclopadia of Political Science, of McCulloch's Literature of Political Economy, and many other useful works that might be mentioned ?
This policy of throwing out books because later books have appeared in the same general field I maintain is all wrong. In the libraries of Paris, with its Bibliotheque Nationale of two million volumes, and its secondary libraries aggregating two million more, it may be proper to help the reader by putting only the latest bibliographies within his reach; but Paris is not the world, and probably not France, in this respect. Most of us are troubled with too few bibliographies, rather than too many. Not all libraries can afford to buy all the bibliographical works they would like. If they cannot have the latest, is this any reason why their readers ought not to be referred to one that is nearly as late and nearly as good ? Besides, it requires the most careful comparison of two works to be able to decide with justice that one work does entirely replace another. Take the case of Hoefer and Michaud, mentioned above. Granted that Hoefer is a little later than the second edition of Michaud, and that the concise bibliographical references appended to the articles in Hoefer do contain substantially all the references mentioned in Michaud, can M. Stein deny that the running comments which accompany the references in Michaud are in many cases very suggestive? He surely would be one of the last to admit that bibli- ography stops with a mere listing of the title, else he ought in strict logic to suspend further publication of the "Partie litt^raire." of Polybiblion.
It also appears from the passage quoted, and the lists confirm it, that the manual is not Petzholdt brought to date. To me it seems that the best part of Petzholdt is left out. M. Stein has doubtless lived in an atmosphere saturated with bibliographical information for
