298 



G. R. PUTNAM CONDITION OF THE EARTH'S CRUST 



Anomalies (in dynes), g observed, - — g 

 computed. 



i ^ * 



Average elevation New method 



Station. reduction (Put- (Havford, 



nam, 1895). 1012). 



Dynes. Dynes. 



For 17 other stations : 



Yakutat Bay, Alaska 4- .027 — .032 



Pyramid Harbor, Alaska 4- .075 4- .054 



Juneau, Alaska 4- . 046 4- . 037 



Sitka, Alaska 4- .007 + .008 



Wrangell, Alaska 4- . 022 + . 018 



Burroughs Bay, Alaska — .036 — .023 



Port Simpson, British Columbia 4- . 029 4- . 023 



Seattle, Washington (high school) — .071 — .093 



San Francisco, California — .039 — .023 



Mount Hamilton, California —.022 —.003 



Hoboken, New Jersey 4- .016 . 4- .024 



Baltimore, Maryland —.026 —.011 



Saint Paul, Alaska 4- .034 .000 



Honolulu, Hawaii —.002 + .054 



Mauna Kea, Hawaii 4- . 076 4- . 185 



Saint Georges, Bermuda —.029 4- . 020 



Jamestown, Saint Helena — .037 4- .060 



For 42 stations, range of anomalies .147 .278 



Mean, with regard to sign 4- .002 4- .007 



Mean, regardless of sign . 025 . 024 



For 25 United States stations of 1894 and 1895, 



range .086 .062 



Mean, with regard to sign . 000 . 000 



Mean, regardless of sign .018 .014 



The above comparison shows that the average elevation reduction of 

 1895 gave nearly as good an accord of anomalies as the Havford method 

 of 1912, and also that the individual anomalies, with few exceptions, fol- 

 low the same trend. In this comparison most weight should be given to 

 the first group of 25 United States stations, ami for these there is a small 

 but appreciable advantage in the 1912 results. 



As the superiority of the more recent method is admitted, it should be 

 explained that this comparison is now made for two reasons: First, for ■ 

 historical completeness, as the work of 1895, very likely because of the 

 way in which it was published, has been largely overlooked in the exten- 

 sive comparisons which have been made of recent with older reduction 

 methods, although the 1895 method was as definite as and more logical 

 than those preceding and gave much more accordant results. Second, the 

 fact is important in drawing deductions as to the validity of assumptions 



