322 



THE SAYRE OBSERVATORY. 



these corrections to the micrometer readings and using an approximate value of the 

 screw. The temperature coefficient deduced from the 1889, '90 series lias been used. 

 This was derived before the complication due to wearing of the screw had become 

 noticeable and seemed reliable. The value assumed is as follows : 



It — 50". 5194 — .000307 [T — 45°] 



T being the thermometer reading. 



Each latitude! determination now gives an equation of the form 



]/p\A'l> — (M—M>) Af 2 Ii = »] 



Where Aty and AE are the corrections to the assumed latitude and micrometer 

 values respectively. An example illustrating this process may be found in the published 

 results for 1894, '95, before referred to. 



Sixteen series of equations wen; formed for determining the values of Ac/> and AlE, 

 the time embraced by each series therefore averages something less than a, month. 



The results for Aiis! follow. No use was made of A$ : 



1 1892 October 10-Oclol>cr 24 



2 October 30-November 26 



3 December 1-December 23 



4 December. 26-.T;muary 20 



5 1893 January Si-February 16 



— .0111 

 — .0090 

 | .0029 

 + .0017 

 — .0001 



Adjusted 

 A%R 



— .0088 



— .0056 

 — .0024 



[- .0009 

 + .0041 



C-0 



I 23 

 -1-34 



— 53 



— 08 

 4- 42 



6 



February 



20-March 



18 



+ .0100 



+ .0073 



— 27 



7 



March 



18-Apiil 



4 



+ .0131 



+ .0106 



— 25 



8 



April 



16-May 



7 



1 .0124 



+ .0138 



| 14 



9 



May 



8- June 



8 



+ .0209 



4- .0170 



— 39 



10 



Juno 



8-.Tuno 



30 



+ .0216 



+ .0203 



— 13 



11 



July 



2-.Tuly 



23 



-(- .0205 



|- .0242 



-|-37 



12 



July 



27-August 



10 



4- .0244 



•f .0288 



-1- 44 



13 



September 



2-September 



23 



■f .0336 



[ .0337 



I 01 



14 



September 27-October 



24 



+ .0366 



-[- .0381 



1- 15 



15 



October 



25-November 



30 



-|- .0461 



|- .0425 



— 36 



16 



December 4-Decomber 27 



|- .0484 



.0470 



■14 



The adjustment was graphic and, as will be seen, the increase in AIR differs but little 

 from the linear form. 



The micrometer frame carried five threads numbered consecutively. The respective 

 distances were made as nearly as possible equal to ten revolutions of the screw. In 

 observing a latitude pair the bisection was made with thread III, unless the difference of 



