April 28, 1916] 



SCIENCE 



597 



tion of the difierenee of longitude between the 

 observatories of Paris and Washington, as 

 deduced under the direction of M. Renau, 

 who makes the following statement: 



It is now three fourths of a century since 

 the first attempts were made to connect Eu- 

 rope and America in longitude. Gilliss in 

 1838, by meridian observations of the moon, 

 and later Walker, Peirce, and others, by means 

 of eclipses and occultations, obtained results 

 which were not accordant and showed a range 

 of 2.5 seconds. 



About 1849 new determinations were made 

 with the aid of chronometers, but these gave 

 results of little greater precision. Since 

 1866, several determinations were made by 

 the exchange of telegraphic signals. Gould 

 in 1866, Dean in 1870 and Hilgard in 1872 

 determined the difference between Cambridge 

 and Greenwich. Hilgard in 1872 determined 

 the difference between Cambridge and Paris, 

 and in 1892 a determination was made be- 

 tween Montreal and Greenwich. 



In 1912 Captain Jayne, superintendent of 

 the Naval Observatory, with the approval of 

 the Acting Secretary of the Navy, proposed 

 that a determination be made of the difference 

 of longitude between the observatories of 

 Paris and Washington. 



Early in 1913 the Bureau of Longitudes be- 

 gan to study the conditions under which this 

 important work could be undertaken. For 

 some years Messrs. Claude, Driencourt and 

 Ferrie, had been developing the idea of ap- 

 plying radio signals to the determination of 

 the differences of longitude, and due to their 

 remarkable initiative the observatory had been 

 able to successfully measure the differences 

 between Paris-Bizerta, and Paris-Uccle, in 

 1911 and 1912. 



These previous operations seemed to fix the 

 most suitable methods, and to assure the suc- 

 cess of the undertaking, though it was neces- 

 sary to take account of the difficulty of hear- 

 ing radio signals at a distance of 6,175 kilom- 

 eters. 



Operations began in October, 1913, and con- 

 tinued until early in March, 1914, with an 

 interchange of observers near the middle. 



The astronomical observations were most sat- 

 isfactory. It was not until after the middle 

 of November that satisfactory exchanges of 

 radio signals were effected. 



Owing to the perfect installations of the 

 clocks at Paris and Washington, which en- 

 abled their rates for many days to be deter- 

 mined with a precision at least equal to that 

 of the observations, it was found possible to 

 utilize the evenings on which radio signals 

 were exchanged, when astronomical observa- 

 tions were made at one station only (incom- 

 plete), as well as those when such observations 

 were made at both stations (complete). 



In the first part of the operations, 7 com- 

 plete and 14 incomplete evenings were secured 

 and in the second part 10 complete and 20 in- 

 complete evenings were secured. The results 

 are as follows: 



Complete Evenings 



h m B 



First part •. . 5 17 36.53 



Second part 5 17 36.75 



Weighted mean 5 17 36.65 



For all Evenings 



First part 5 17 36.53 



Second part 5 17 36.75 



Weiglited mean 5 17 36.67 



The value 5" 17" 35^67 is adopted as the 

 definitive result of our work. 



The difference 0^22 between the results of 

 the first and second parts should not be re- 

 garded as excessive in view of the peculiar 

 conditions of the enterprise and of the diffi- 

 culty of the exchange of radio signals. It 

 does not seem capable of explanation without 

 further labor. 



In a preliminary publication of the results 

 of the work of the American astronomers, the 

 definitive result is given as 5" 17" 36^62^ which 

 is within 0^01 of our result, and there is a pre- 

 cisely similar difference between the two 

 parts, 36^56 and 36'.76, corresponding to 

 36'.53 and 36^75 as given above. 



2 • Astrono7mcal Journal, Mareli 15, 1915. 



