Maboh 19, 1909] 



SCIENCE 



475 



coil from the effects of atmospheric moisture, re- 

 duces the danger of oxidation due to imperfect 

 covering of shellac, and protects the coil from 

 mechanical injury. The resistances when prop- 

 erly prepared and mounted and protected in this 

 manner, remain remarkably constant in value, 

 whereas open coils in oil almost invariably have 

 a higher resistance in summer than in winter, and 

 fluctuate more or less in value from time to time 

 with the weather. As a result of the discovery 

 of the effect of atmospheric humidity on the re- 

 sistance of standards made at the Bureau of 

 Standards, the National Physical Laboratory of 

 England has sealed its standards, and the German 

 Eeichsanstalt is keeping its standards in a 

 chamber at a constant humidity. The use of the 

 new sealed resistances at the Bureau of Stand- 

 ards has increased the accuracy of resistance 

 work appreciably. 



A Proposed Modification of the KircJilioff Method 

 for the Absolute Measurement of Resistance: 

 Peank Wbnneb, Bureau of Standards, Wash- 

 ington. 



An Instrument Designed for More Precise De- 

 termination of Magnetic Declination at Sea: 

 William J. Pbtees, Department Terrestrial 

 Magnetism, Washington. 

 . The cruises of the Galilee in the Pacific Ocean, 

 among other results, made very apparent the ne- 

 cessity of more accurate determinations of mag- 

 netic declination than could be made with the 

 ordinary instruments of navigation. A coUi- 

 mating instrument has been constructed by Mr. 

 A. Widmer, mechanician of the Department of 

 Terrestrial Magnetism of the Carnegie Institu- 

 tion of Washington, which will be used in experi- 

 mental work on the vessels now being built for a 

 magnetic survey of the ocean areas. Experi- 

 ments were made on the last cruise of the Galilee 

 which assured the practicability of using such an 

 instrument and indicated the possibility of a high 

 order of accuracy. 



Many parts of the Ritchie ten-inch liquid com- 

 pass were used. The card was altered to a four 

 direction collimator by the addition of four eon- 

 cave mirrors with a scale of seven divisions in 

 the focus of each. This alteration increased the 

 original mass by one twentieth part, but decreased 

 the radius of gyration. The period of the colli- 

 mator arrangement in liquid is now about eleven 

 seconds at Washington. The angle between a 

 collimator axis and a celestial body is measured 

 by a pocket sextant, the scale being viewed 

 through windows in the bowl. The instrument is 



not intended for a navigation instrument, but as a 

 step in the attainment of the highest precision in 

 determinations of magnetic elements at sea. 



( The paper will be printed in full in the March 

 nimiber of Terrestrial Magnetism and Atmos- 

 pheric Electricity. ) 



The Electrical Conductivity of the Atmosphere 

 Over the Pacific Ocean: Paul H. Dike, Car- 

 negie Institution, Washington. 

 The work described was done on board the Mag- 

 netic Survey Yacht Galilee during the cruise of 

 1907-08. The purpose was to obtain data as to 

 the earth-air current at sea, to compare with 

 similar results obtained on land. The method 

 involves the measurement of two quantities^ the 

 specific conductivity of the air and the vertical 

 potential gradient. The latter measurement was 

 found to be impracticable on board ship, and only 

 a few values were obtained, during a calm. 

 These were of the same order of magnitude as are 

 ordinarily observed on land. Tlie specific con- 

 ductivity of the air was measured by means of 

 the Gerdien apparatus, consisting of a cylindrical 

 condenser, the inner cylinder of which is con- 

 nected with an electroscope. The conductivity of 

 the air is computed from the rate of dispersion of 

 a charge put upon the inner cylinder when a uni- 

 form current of air is drawn through the appa- 

 ratus. The reading of the electroscope offered 

 the principal difficulty. 



The mean values of the conductivity from all 

 the observations of the voyage were as follows: 



Xj, = 1.603 X 10"* electrostatic units. 

 \n = 1.433 X 10-« " 



XpA„ = 1.12 



Assuming a potential gradient of 100 volts- 

 meter these values of conductivity give a vertical 

 earth-air current 3 X 10"" amperes per square 

 centimeter of the earth's surface, slightly larger 

 than the usual value on land. No variation with 

 latitude was discernible, though the observations 

 extended from 65° 41' north to 45° 07' south. 



Ultra-violet Aisorption and Fluorescence and the 

 Complete Balmer Series of Sodium Vapor: R. 

 W. Wood, Johns Hopkins University. 

 The absorption spectrum of dense sodium vapor, 

 contained in a steel tube one meter in length, pro- 

 vided with quartz windows and heated red hot in 

 a combustion furnace, shows the lines of the 

 principal series (Balmer formula) reversed. But 

 seven lines of this series have been previously 

 observed, the observations having been confined to 

 the emission spectrum. 



