ASTRONOMY. 313 



by the observer ; next, the observer records his own time by the relai/ 

 beat of a chronometer every 1 second ; next by the chronographic method, 

 and lastly by the beat of the chronometer itself (every 0.5 second). Be- 

 ginners nse these methods in succession in the order named. 



Personal equation is studied by means of Eastman's machine (see 

 Wash. Ast. Obs., 1875), and an interesting table of results by the second 

 method of marking time (the relay beat) is given. 



The time determinations are given for each day of observations, with 

 the probable errors. 



Time determinations by sextant observations are also given, and by 

 means of the (known) correction of the standard chronometer the error 

 of each observation and observer is determined. 



We quote below the errors of the sextant Delta t's so determined. 

 (Usually 10 altitudes of an east star and 10 of a west were employed.) 

 4«.0; KO; 1«.5; 13«.7;8«.2; 24M; 1«.7; 0«.4; 1«.8; 0«.7; 3«.3; 2«.8; 5«.6^ 

 5^.0. These are the observations of students. 



The latitude is determined, first by zenith telescope, and the first 

 process is the evaluation of the level division. This is done by each 

 student, and of the set for 1881 the largest probable error is +0".008. 



The observations for latitude are given in detail. From 326 observa- 

 tions of 84 pairs in 1880, the latitude was found to be + 40° 47' 21".59 

 ± 0".082. From 591 observations of 104 pairs in 1881, it was found to 

 be -f 40° 47' 2L".47, a difference of about 12 feet only. Sextant observa- 

 tions for latitude were also made. The errors of the several determina- 

 tions were l".l; 30".5; 14".8; 7".4; 0".9; 0".5, respectively. 



The longitude is determined — 



1st. By lunar culminations. The errors were found to be (of one night's 

 observations), 7^01; 6^99 5 3M4; 15«.97; 3«.74; 21«.06; 14«.29; 24«.84; 

 20«.58; 25^04; 4«.06; 2«.28; 1«.81. The errors of each limb of the moon 

 are separately shown. 



2d. By Jupiter's satellites. Errors: 21«.02; 29«.48; 25«.95. 



3d. By lunar distances (sextant). Errors: 4«.0; 25^0J 40^.65 13«.7^ 

 34«.7; 6«.4; 5^7. 



The value of a revolution of the micrometer screw of the small (five- 

 inch) equatorial was twice determined, as follows : 



R = 19".362 i 0".018 Polaris. 

 = 19 .287 rt .057 " 



The displays of the aurora have been regularly noted since 1870, Feb- 

 ruary 1. The number for each year is as below : 



1870, 99; '71, 104; '72, 94; '73, 92; '74, 35; '75, 27; '76, 17; '77, 13 j 

 '78,4; '79, 16; '80, -2; '81, 44. 



What has been given as an abstract of one year's work in one depart- 

 ment only of this school of application for young engineer officers i» 

 sufficient to show that we have at present no better school of practical 

 astronomy in America. 



