20 PROGRESS OF SCIENCE IN THE CENTURY. 



(&) Classification of Facts. In many cases after 

 the accumulation of data, much time must be spent 

 in their arrangement. A careful worker at the prob- 

 lem of migration in birds, like Mr. Eagle Clarke, 

 may require for the classification of his data a longer 

 time than was spent in their collection. If the facts 

 are to form part of the body of science, they must 

 be made readily available, and this process of diges- 

 tion is often slower than that of ingestion. If 

 the aim be to detect similarities of sequence the facts 

 must be grouped in ordered series. Here, in many 

 cases, the use of graphs, curves, and mathematical 

 methods has proved itself invaluable, notably, for in- 

 stance, in Galton's work on inheritance, or in the re- 

 cent statistical studies on variation. 



It has been a common experience in the arrange- 

 ment of data that some minute discrepancy has re- 

 vealed itself, and that the following of this at first 

 perhaps puzzling occurrence has led to the elucida- 

 tion of the whole problem. Thus it has become a 

 maxim in science that no apparent departure from 

 the rule or general sequence should be treated as 

 trivial, and no minute discrepancy disregarded. 

 That nitrogen obtained from chemical combinations 

 should be about one-half per cent, lighter than that 

 obtained from the atmosphere, may seem a very 

 minute fact, but it led Lord Rayleigh and Professor 

 W. Ramsay to the discovery of Argon. 



(c) Analysis. With scientific problems of a cer- 

 tain order, there is often need for a preliminary 

 process of analysis before the desired data can be 

 obtained. Whenever we get below the surface phe- 

 nomena of life patent to the observer we have to 

 dissect, to cut sections, to take advantage of chemical 

 analysis and so on. The end desired is a re-state- 



