938 



SCIENCE 



[N. S. Vol. XXXII. No. 835 



ing worm and the graduate student. Ele- 

 mentary facts about raw material are not 

 the advancement of knowledge. They are 

 killing to those who have a capacity for 

 something better. The listing of "Ter- 

 ence's terminations in T. " is a type of 

 work which at the best bears the same re- 

 lation to research that forge-work bears to 

 engineering. It is worth while to the engi- 

 neer to know what it is like and to be able 

 to handle a hammer if need be. Moreover, 

 a hammered-out horse-shoe is an actual 

 reality. But to make a horse-shoe, even 

 one of a form never seen before, is not the 

 final thesis for which the engineer enters 

 the university. 



Much of the graduate work in non-math- 

 ematical subjects receives an appearance 

 of accuracj^ from the use of statistics, or 

 other forms of mathematics. This seems to 

 make the results " scientific." Mathe- 

 matics is a science only when its subject- 

 matter is science — when it deals with re- 

 sults of human experience. At other times, 

 it is simply a method — a form of logic. A 

 mathematical enumeration, or even a 

 formula, does not give exactness where it 

 did not exist before. 



The statistical enumeration of the "pre- 

 fixes in P.," or the pebbles in the bank, is 

 held to give the method of research. It 

 teaches patience and accuracy, two funda- 

 mental virtues in the progress of science. 

 Patience, perhaps, if the student persists to 

 the bitter end. Accuracy certainly not. 

 Sooner or later the student will discover 

 that to multiply by ten one of his columns 

 of figures or to divide another by five will 

 have no effect on his final conclusion, for 

 there isn't going to be any conclusion. He 

 will then learn to supplement his tables by 

 the quicker and more satisfactory method 

 of guess work. He turns from the methods 

 of pedantry to the method of journalism. 

 At the best, he will find that the less labori- 



ous methods known as qualitative have the 

 advantage over quantitative methods, where 

 matters of quantity have no real signifi- 

 cance. 



No one should begrudge any amount of 

 time or strength or patience spent on a 

 real problem. In that regard, Darwin's 

 attitude towards the greatest of biological 

 problems is a model for all time. But we 

 should believe that there is a problem, and 

 that our facts point towards the truth in 

 regard to it. A fact alone is not a truth, 

 and ten thousand facts may be of no more 

 importance. A horse-shoe is not an 

 achievement. Still less are ten thousand 

 horse-shoes. "Facts are stupid things," 

 Agassiz used to say, "unless brought into 

 connection by some general law." In 

 other words, facts signify nothing, except 

 as the raw material of truth. 



A graduate student of an honored phil- 

 ologist in a great university lately ex- 

 plained her graduate work to me. A chap- 

 ter in Luther's bible was assigned to her, 

 another to each of her fellows. This was 

 copied in longhand, and after it, all the 

 variant German versions of the same 

 chapter. Her work was to indicate all the 

 differences involved. There may have been 

 something behind it all. The professor 

 may have had in mind a great law of vari- 

 ance, a Lautverschiebung or Entwickelung 

 of pious phraseology. But no glimpse of 

 this law ever came to the student. More 

 likely, the professor was at his wits' end to 

 find some task in German which had never 

 been accomplished before, and which had 

 never before occurred to any German task- 

 master. No wonder the doctor's degree is 

 no guarantee of skill as a teacher ! Among 

 the first essentials of a teacher are clearness 

 of vision and enthusiasm for the work. 

 This is not cultivated by these methods. It 

 is not even "made in Germany." The 

 "law of time relations of iron and sul- 



