September 18, 1896.] 



SCIENCE. 



381 



in most cases in one devoted as closely as 

 may be to their particular branch of bot- 

 any, provided it have a fair general circu- 

 lation, and especially provided it reach the 

 principal botanical libraries. 



Especially in the earlier years of their 

 work writers are sometimes given to distrib- 

 uting their papers among a number of 

 journals. Except for the purpose of special- 

 ization just referred to, this is usually a 

 mistake. Knowledge that a certain stu- 

 dent has published on a given subject is 

 often first obtained through incidental ref- 

 erence, lacking every element of precision. 

 The probability that all of his writings are 

 to be found in one or a few journals or 

 series of proceedings greatly simplifies the 

 completion and use of such references, since 

 the Royal ^ociety^s Catalogue, though perhaps 

 more complete as to titles, is necessarily 

 even farther behind than the Jahresbericht. 

 Where the subject of an earlier paper is 

 again passed in review by the author, only 

 the gravest necessity should lead to the se- 

 lection of a new medium for the publication 

 of the later paper. 



Whether the medium of publication se- 

 lected or accepted be a j ournal or the proceed- 

 ings of a society, the possibility of having 

 separates struck off for the mere cost of 

 press work, paper and stitching, makes it 

 possible for almost any paper to appear as 

 an independent pamphlet, accredited, to be 

 sure, to the journal from which it is an ex- 

 cerpt, but, like a book, necessitating author's 

 citation in catalogues, and admitting of more 

 ready arrangement in its proper place 

 where the works of a library are disposed 

 on the shelves according to subject. The 

 time was when a pamphlet was considered 

 of little value and quite certain not to be 

 preserved, but one of the characteristics of 

 the modern librarian is a great and grow- 

 ing appreciation of the value of this class of 

 works, leading to their careful preservation. 



No small part of the volume of M. De 



Candolle, already referred to, is devoted to 

 very explicit and well considered directions 

 for preparing the record of one's observa- 

 tions for the press ; and the general conclu- 

 sion is reached, after a careful analysis of 

 the subject, that the maximum value of any 

 manuscript exists at the exact moment of 

 its completion, indicating this as the most 

 suitable time for its publication. Though 

 it is probable that the publishing of any im- 

 portant work should not be unnecessarily 

 delayed after it has been pushed to what the 

 author considers completion — at least so far 

 as he can carry it, — there may be reasons in 

 some cases for publishing a preliminary 

 statement considerably in advance of the 

 completion of the work. Neglecting the 

 publication of an early abstract of unfinished 

 work as a means of securing priority — too 

 often a purely personal matter — I may say 

 that such abstracts, coupled with a request 

 for material or data, not infrequently bring 

 to the advanced student the means of greatly 

 increasing the completeness and value of 

 his work. 



Time does not permit me to go into a de- 

 tailed analysis of the many ways in which 

 an investigator may use his time so as to 

 make it productive of important results for 

 himself and others. Having passed in some- 

 what comprehensive, though hasty, review, 

 the main factors in the question, I desire, 

 in closing, to repeat that for most of us the 

 opportunity of life does not lie in a great 

 and abrupt change of condition, but that it 

 is composed of countless minor chances 

 which are great only when viewed collec- 

 tively. To see and use them calls for alert 

 senses, a knowledge and use of the means 

 of ascertaining what has already been done, 

 and, by exclusion, something of what re- 

 mains to be done, facilities adequate to the 

 task in each case, and indomitable perse- 

 verance and ceaseless activity. Great as 

 the value of facilities is, they are merely 

 means to an end. They accomplish noth- 



