992 



SCIENCE 



[X. S. Vol. XXXI. Xo. 



delayed somewhat the carrying out of the 

 program, but fortunately the velocity re- 

 mained nearly constant for several weeks, 

 until satisfactory observations were secured. 



With the spectrograph adjusted for the 

 orange region, which is rich in water vapor 

 absorption, spectrograms of Mars were se- 

 cured on January 26 and 27 under poor at- 

 mospheric conditions, and on February 2 

 under excellent conditions, our atmosphere on 

 this night being exceedingly dry. Measures 

 of the available water vapor lines on these 

 spectrograms, 8 to 22 in number, establish 

 that they were displaced with reference to the 

 lines of solar origin in the observed Martian 

 spectrum by amounts on the three dates cor- 

 responding to velocities in the line of sight of 

 19.7, 20.2 and 18.3 km. per second; weighted 

 mean value, 19.2 km. The relative velocities 

 of Mars, computed from our knowledge of the 

 orbits of the earth and Mars, amounted to 

 19.1 km. per second. The dispersion and slit- 

 width employed were such that the water 

 vapor lines originating in our atmosphere and 

 any originating in Mars's atmosphere should 

 have appeared side by side, though not clearly 

 separated. If the absorptions by the two 

 planets were equal, the two sets of lines of 

 equal intensities should, in effect, have ap- 

 peared as broad lines of double width, and 

 the measured velocities should have been but 

 half the computed velocities. The facts are 

 that the terrestrial lines were not bordered 

 nor increased in width by companion lines. 

 WTien the micrometer wire was set succes- 

 sively in the positions which Martian ab- 

 sorption lines would occupy, no traces of ab- 

 sorption were found in these positions. In 

 effect, Martian absorption did not exist to 

 such an extent as to be visible in the spec- 

 trum, or to influence the measurements re- 

 ferred to. 



With the spectrograph adjusted to the so- 

 called " alpha " region at ^ 6280, which in- 

 cludes a large number of oxygen absorption 

 lines, two spectrograms were obtained on 

 February 3. The observable oxygen lines, 

 seven and six in number, were displaced with 

 reference to the lines of solar origin by 



amounts corresponding to velocities in the 

 line of sight of 18.8 and 17.4 km. per second. 

 The velocity computed from the elements of 

 the orbits amounted to 19.1 km. The dis- 

 crepancy of 1.0 km. is within the unavoidable 

 error of measurement. Here again the ter- 

 restrial oxygen lines were not bordered nor 

 doubled in width by Martian lines. 



The conclusions to be drawn from this in- 

 vestigation are : The quantity of any water 

 vapor existing above unit area in the equa- 

 torial atmosphere of Mars was certainly less 

 than one fifth that existing above Mt. Hamil- 

 ton under the excellent conditions prevailing 

 on February 2. The air temperature was 0° 

 Centigrade, the relative humidity 33 per 

 cent., the absolute humidity 1.9 grams per 

 cubic meter, and the zenith distance 55°. 



Likewise, the quantity of oxygen above 

 unit area of Mars must be small in compari- 

 son with that in the earth's atmosphere. 



It should be repeated that the rays of light 

 utilized had passed in effect twice through 

 the Martian atmosphere. 



W. W. Campbell, 

 Sebastian Albrechx 



Lick Obsekvatoey, 

 Univeksitt or Califobnia, 

 April, 1910 



SOCIETIES AND ACADEMIES 



THE AMEEIOAN PHILOSOPHICAL SOCIETY 



At the meeting of the American Philosophical 

 Society, on May 20, the following paper was read: 



On the Principle of Relativity and its Signifi- 

 cance: Dr. RoBEET H. Hough, University of 

 Pennsylvania. 



The question was considered only in its philo- 

 sophical aspect. The idea was developed from the 

 fundamental concepts of dynamics as formulated 

 by Newton and Hertz, and extended to the field 

 of electro-dynamics and optics. The validity of 

 the principle as a mathematical concept was 

 maintained. The equations of transformation 

 were derived by purely mathematical steps from 

 two initial equations representing experimental 

 laws to the present probable error of observation : 

 and the consequent relations of the distances and 

 times involved and their respective units con- 

 sidered. '^ 



^ 



