48 NATURAL SCIENCE. July. 



find that each of the orders of mammals requires four decimal places 

 (in Dewey they require one or at most two), and reaches its climax 

 when we arrive at 565.9395 for " Man," and are told it *' will want 

 much subdivision." 



This is followed by two tables " showing how genera and species 

 might be treated." I have puzzled over these for some time, but 

 cannot make them out ; but, as I am given to understand this idea 

 was early dropped for an alphabetical arrangement, any criticism of 

 them is unnecessary. 



Time and space fail me to go into further details ; but to sum up, 

 the Dewey classification is not perfect, but has been long before the 

 public and has been well tried, it may be had in print with an 

 elaborate index, it is simple and consistent and is based upon practical 

 experience of the subjects on which books and memoirs actually have 

 been written, and of the proportional numbers required by each 

 topic. The Royal Society's classification, so far as I have examined 

 it, has none of these advantages : it is crude, incomplete, and incon- 

 sistent, largely because it is the result, not of practical experience, 

 but of a priori consideration. I conclude with two passages from 

 Dewey's own Introduction to the fourth edition of his tables : — 



" The inexperienced user is very likely to feel entirely competent, 

 after once reading the tables ... to institute a series of improve- 

 ments. Experience proves that nothing could be more disastrous. 

 . . . Frequently, proposed changes, carefully studied out and sub- 

 mitted as improvements, are shown by our old records to have been 

 adopted and used in the exact form proposed till considerations which 

 had not been foreseen forced us to change the form as printed. Even 

 after years of experience one is not safe in pronouncing on an apparent 

 improvement without consulting the voluminous records of previous 

 experiments." 



" To make out a new set of heads would involve labor and cost 

 vastly beyond the dreams of any person who has not tried exactly 

 this work. The time actually spent on the tables as here printed, 

 by various competent workers, would aggregate several years and 

 cost thousands of dollars. The uniform and urgent advice of 

 the experienced is to adopt a poorer scheme already made rather than 

 undertake so herculean a labor. When done, the maker may 

 possibly be better suited with it, but it is doubtful if many others will 

 be. It is vastly wiser for any man whose time is of value, to use it 

 in something more practically useful to himself and his library than in 

 trying to construct a 'satisfactory' scheme of classification. No one 

 ever yet wholly suited himself or any one else, and probably no one 

 ever will. By adopting this scheme already workt out he saves 

 much time and money, gains the immense advantage of using a system 

 in common with hundreds of others, so that he may utilise their 

 labors and investigations and share with them the economies of 

 co-operation." 



The Manchester Museum. Wm. E. Hoyle. 



We are glad that we have succeeded in obtaining the foregoing 

 article from Mr. Hoyle. Since the International Conference on 

 scientific bibliography, to be held at the rooms of the Royal Society 



