GENERAL OBSERVATIONS.* ON CLEARNESS IN PRESENTATION 645 



and more or less irksome process, many new ideas occur to me, 

 and better ways of stating ideas already expressed. It helps 

 also, I find, to put aside the completed paper and come back to 

 it months later, as to a new subject, or to one by another author. 

 Occasionally there is a person who can write a thing as it should 

 be the first time trying, but I have known only one or two 

 such persons. Generally, easy writing is hard reading. Dar- 

 win sometimes recast his paragraphs a dozen times, and most of 

 us may expect to reach a good style, if at all, only by dint of 

 much labor and repeated re-writing. Yet who can doubt that 

 it is an end worth all it may cost ? 



You publish to convince your readers and advance your 

 own branch of science, and incidentally to enhance your own 

 reputation. Look to it, then, that your writings are not only 

 permeated with a love of the truth, but are forceful and limpid as 

 a mountain stream. To this end, avoid technical terms when 

 common words will serve, even if you must do so at the expense 

 of some conciseness. Nothing is more discouraging to the 

 general reader than a book or paper bristling with a newly in- 

 vented terminologv, or full of mathematical formulae. 



ON COMPLETENESS OF PRESENTATION 



If you wait for absolute completeness, you will never publish 

 anything but be always following up some one of the many side 

 paths ramifying entrancingly in every direction from the great 

 central subject under consideration. 



Nature is boundless and our own working lives are very 

 short. There must, then, be some compromises. The investi- 

 gation must be broken off somewhere. The question is, where? 

 This is solved, partly, but not altogether satisfactorily, by not 

 undertaking very complex problems. All I can say is — Do each 

 piece of work as thoroughly as time permits, but publish, 

 otherwise, especially on the assumption that you have something 

 really worth publishing, your generation is more or less 

 defrauded. 



Granted that you intend full publication, how complete the 

 first paper should be, whether it should include all, or only 



