JOHN WESLEY POWELL. 45 



tion was paid the exceptional compliment of immediate considera- 

 tion and confirmation. 



He directed the work of both bureaus until 1894. During this 

 period the appropriations for the work of the Geological Survey 

 were greatly increased, and its functions were from time to time 

 enlarged, especially by the addition of investigations and surveys 

 connected with the utilisation of the waters of the arid region for 

 irrigation. In 1888 the Survey was instructed to classify the lands 

 of the public domain, and especially to set apart as agricultural 

 those which might be redeemed by irrigation. The provisions of 

 the law were such that the Secretary of the Interior felt compelled 

 to withdraw all public lands from sale pending their classification 

 by the Geological Survey. This withdrawal aroused a storm of in- 

 dignation, leading to the repeal of the new law and the reduction 

 also of the appropriations for other work of the Survey. The dis- 

 aster indicated diminished confidence on the part of Congress in 

 the Director of the Survey, and led him to resign his office as soon 

 as he could be sure- of the appointment of a properly qualified suc- 

 cessor. He retired gladly, as impaired health had for several years 

 made heavy executive responsibilities an onerous burden, and he 

 afterward watched with great pleasure the successful administration 

 of his successor, Mr. Walcott. 



Immediately after his resignation he submitted to a third ope- 

 ration on his wounded arm, which had given him much trouble, 

 and thereafter sedulously husbanded his physical resources, devot- 

 ing the remainder of his life to the elaboration and publication of a 

 system of philosophy to which he had already given much thought. 

 He retained the directorship of the Bureau of Ethnology, but dele- 

 gated the chief labor of administration to another. This work was 

 carried on despite a complication of bodily ailments, and his health 

 steadily declined until his death, which occurred on the 23d of Sep- 

 tember, 1902. 



The study of nature falls logically into three categories : ob- 

 servation, classification, and explanation. One great part consists 

 in the observation and description of phenomena, another in their 

 classification and generalisation, the putting of like phenomena to- 

 gether and the substitution of summary statements for the enume- 

 ration of details. A third part furnishes the explanation of groups 

 of phenomena, or constructs theories. The three interlock and in- 

 teract. Most good observation is guided by antecedent classifica- 

 tion or theory; the observer either gathers facts within a specific 

 category, or he seeks crucial facts to test an hypothesis. Before 



