EDITORIAL. 405 



worth while recording permanently. It represents a finished product 

 as far as it goes. Hence it is well worth care and consideration. 



The selection of material to be included is a vital matter, and one 

 which many writers find most difficult. It is very rarely practicable, 

 and nearly as rarely desirable, to include all the data secured in an 

 investigation. There are always gradations in value; a great deal 

 of that pertaining to intermediate steps is unessential in detail to 

 the majority of readers and does not justify the expense of publica- 

 tion. Even in the more important aspects of the work all the fig- 

 ures and observations and determinations can rarely be published. 

 Summaries must naturally be depended upon, with sufficient expla- 

 nation to make clear the scope of the work done, the range of varia- 

 tion found, and the justification for the figures and deductions as 

 given. 



Not a few persons contend for the publication of practically all of 

 their data — an essential transcript of their notebooks. Considering 

 the present congestion of scientific journals such a contention is not 

 a reasonable one. Rarely, indeed, are the data so intrinsically im- 

 portant as to warrant publication in full, and insistence on it de- 

 prives someone else of the privilege of publication and his prospective 

 audience of the opportunity of reading. Many authors seem to have 

 difficulty in judging what is the most important in their own articles. 

 Undue emphasis is laid on trivial points or side issues. The results 

 of large numbers of observations on a particular point are recorded 

 merely to show the reader that they were made and that there is 

 practical uniformit3% whereas a statement of the extent of the obser- 

 vations, the nature of the range, and the general result would answer 

 every reasonable purpose. 



Again, detailed individual records are made of negative results, 

 which in themselves have no permanent value not expressed in a 

 brief text statement. Hesults are platted, although they confessedly 

 show no definite relationship and permit of nothing more than nega- 

 tive inferences. The value of a negative result is not disputed, but 

 when it merely indicates the author was not on the right track it 

 rarely justifies detailing. Other data are recorded for their own 

 sake, in the evident hope that someone wdll be able to make more out 

 of them than the author apparently has, for he ventures no comment 

 or deduction. 



The desire of writers to make their published articles a complete 

 record of their studies is often founded in the criticism of some 

 readers, who seem to assume that if everything is not given some- 

 thing important is being withheld, a captious attitude which is diffi- 

 cult to combat successfully. Manifestly the reading public must take 

 the author's word for it that he has made the number of observations 



