90 ANNUAL 11KCORD OF SCIENCE AND INDUSTRY. 



has discussed a large number of observations of contacts, and de- 

 duced the corrections to the American Ephemeris and the British 

 Nautical Almanac, respectively. In the former, Leverrier's old the- 

 ory of Mercury is used; and in the latter, his later tables; and the 

 importance of the comparison lies in the circumstance that these lat- 

 ter include a term due to the supposed attraction of an infra-Mercu- 

 rial planet. The mean corrections to the predicted times of contact 

 resulting from the observations are as follows: 



s. s. 



1st contact, C ob. corr. to Am. Eph. 45.7 to N. A. 4 

 2d " 15 " " - 61.8 " -20 



3d " 11 " " -124.4 " -18 



4th " 10 " " -141.7 " -35 



Means - 1)3.4 -40.25 



Thus it appears that the later tables, with the term due to an intra- 

 Mercurial planet, give a satisfactory representation of the fact. The 

 apparent corrections to the Nautical Almanac range from 10 s to 

 I s for first contact; 32 s to 10 s for second contact; 26 s to 4 s 

 for third contact; and 41 s to 24 s for last contact. Taking 

 simply the discordances from the mean in each case in the tables 

 given by Mr. Boss, the mean error of an observation is 2 s for first 

 contact, 6 s for second, and 5 s each for third and fourth. Considering 

 the small number of observations of first contact, we should not be 

 far wrone in taking the mean error as about the same for all four 

 contacts, and equal to 5 s . 



Micrometrical measures of the diameter of Mercury during the 

 transit gave 11.30"0.14", uncorrected for irradiation or possible 

 expansion of the screw. This value would give 176. G s as the inter- 

 val between external and internal contact. The observed interval 

 was 172.4 s between first and second, and 171.2 s between third and 

 fourth. 



