208 ASTRON-OMY. 



tiou of wliicli depends ou au equal number of results completely inde- 

 pendent of one another, and obtained by methods so totally different, 

 and subdivides the various values assigned for the sun's mean jjarallax 

 as follows : 

 Geometrical methods, 8".82: 



8".So by Mars (Cassini's method), Newcorab. 



8' .79 by Venus, 17G9 (Halley's method), Powalky., 



8".81 by Venus, 1874 (Halley's method), Tupman. 



8".S7 by Flora (Galle's method), Galle. 



8". 79 by Juno (Galle's method), Lindsay. 

 Mechanical methods, 8".83: 



8".81 by the lunar inequality (Laplace's method), Kewcomb. 



8". 85 by the monthly equation of the earth, Leverrier. 



8".83 by the perturbations of Venus and Mars, Leverrier. 

 Physical methods, 8".81 : 



8". 799 velocity of light (Fizeau's method), Cornu. 



8".813 velocity of light (Foucault's method), Michelson. 

 After explaining the value of the different results, M. Faye gives hi& 

 preference to the physical result, and arrives at these conclusions : 



1. That the method of the physicists is superior to all others, and 

 ought to be substituted. 



2. That the value of solar parallax, 8".813 (by physical methods), is 

 now determined to about y^ of a second. 



3. That the seven astronomical methods of procedure converge more 

 and more toward that value, and tend to confirm it without equalling 

 it in precision. 



In other words, M. Faye believes that the true distance of the sun is 

 that found recently by Mr. Michelson. 



Professor Eastman has published "A Value of the Solar Parallax from 

 Meridian Observations of Mars at the Opposition in 1877." In 1876 a 

 circular letter was x>repared at the ]S"aval Observatory and distributed 

 to astronomers, asking their cooperation in making meridian observa- 

 tions of Mars in 1877. A programme of observing was proposed by- 

 Professor Eastman, based on that suggested by Dr. Winuecke in 1863, 

 and a list of comparison stars was selected. The iirogramme was strictly 

 followed at Washington, at Sydney, and partly at Leyden. At Mel- 

 bourne, the Cape of Good Hope, and Cambridge, Mass., the programme 

 was not followed ; and from San Fernando and Kremsmunster observa- 

 tions were received either unreduced or unexplained. Professor East- 

 man has not reduced the San Fernando observations and makes no use 

 of those at Kremsmunster. The observations of Mars when less than 

 four comparison stars were used on any nighty are also rejected. Upon 

 the remainder, the discussion is made. 



The probable error of a single declination of 31ars was found to be — 



At Washington ± 0".452 



At Melbourne ± 0".552 



At Leyden ± 0".311 



