385 



CHAPTER XLI. 



ADVANTAGES OF NUMBER OF EXPERIMENTS. 



IN research it is often a good plan to settle off-hand the 

 questions which arise at the moment, and thus clear one's 

 mind of doubts as they arise ; but if this practice is 

 carried too far, as it very easily may be, it is certain to 

 lead the investigator away from his subject into a multi- 

 plicity of researches which he will never be able to com- 

 plete. In searching for new truths, instead of waiting 

 until we can predict with certainty, which in most cases 

 would be to wait a very long time, it is frequently better 

 to make experiments at once, because the making of a 

 number of different experiments, if varied and important 

 ones are included, is one of the great and prime con- 

 ditions of discovery. Some of the discoveries made by Dr. 

 Priestley are examples of this. The experiments must, 

 however, be new and well-conducted ; if they are not new, 

 the results will not be novel ; and if not well-conducted 

 in all essential respects, the results will be erroneous, and 

 more or less misleading. A single well-chosen and pro- 

 perly carried out experiment is far more valuable than a 

 multitude of imperfect ones ; if an experiment is well made, 

 a repetition of it is almost unnecessary, because it would 

 add but little to the certainty of the results. It is not 

 the number of experiments alone which lead to great dis- 

 coveries ; Potts of Halberstadt, Professor of Chemistry in 

 Berlin, and a favourite of the King of Prussia, is said to 

 have performed 30,000 experiments in six years, for the 

 purpose of ascertaining the ingredients and process em- 

 ployed in making Dresden porcelain, but discovered no 



c c 



