= J 
- 
J. M. Schaeberle—Lateral Astronomical Refraction. 471 
To obtain the correction to be applied to the level constant 
bd 
b, of the instrument, we have but to multiply the values a@ D' 
Sh 
and @ > taken from table I, with the arguments D’ and D, by 
the sines of the angles. 180°+ ¥ and ¥ respectively, and take 
the algebraic sum of the products. Similarly the quantities in 
table IIT are to be multiplied by the cosines of the angles 
180°+ ¥ and ¥ respectively. For stars north of the zenith dz 
must have the opposite sign from that given by the equation. 
An inspection of the weather maps of the Signal Service, as 
published by the War Department, will show that the distance 
etween two adjacent isobars or isotherms is frequently less 
than thirty miles, and at times not more than one-half of this 
distance. “As the times of observations will not in general 
correspond with the times for which the maps are constructed, 
the values of D,¥ and D’,¥%’, can when necessary, be interpo- 
lated for the middie tirne of the observations and assumed to 
remain constant for the series. A glance at the map will at 
once indicate whether the corrections will be sensible or not. 
At an isolated station, if the hourly thermometric and barometric 
changes are noted, the observer can still deduce the corrections, 
provided he has any means for finding the veloctiy and direc- 
tion of motion of the thermometric and barometric waves. 
The most reliable data would of course be obtained from — 
simultaneous observations made (while the astronomical work 
1S going on), at three or more nearly equidistant stations on the 
same level and lying within a radius of fifty miles. One of 
these stations would be at the observatory itself. 
Where the aim of an observatory is to determine absolute 
positions, an arrangement of this kind, if not indispensable, is 
at least most desirable. 
Ann Arbor, Mich., March 18, 1884. 
_ Posrscrrer.—The formule for finding the magnitude and diree- 
tion of inclination of the strata from observatio 
Le anal eee he—hy 
ee | eae and D, 
=to, 
(A, —h,=4h, etc. being found according to the method already 
given). If 55 =# denotes the inclination of the strata to the 
e horizon, we have, with all desirable accuracy, the equations 
¢ cos (A,— V)=7, t cos (A,— Y)=%,. 
