173 
It is clear by the predominating of the positive sign, that the 
systematic error looked for exists, but also, that the error is not 
constant in the different declinations, and shows a marked increase 
for increasing declination. 
The following table gives the mean values of y, for the zones 
according to which the programme stars are divided: 
Zone 0°—20° 12°14  +05,00108 + 61 
Pee te 2 oe 270 0 ,00230 + 42 
Ma 130 AOR OD 0,00301 + 33 
ret 40°2—zenith 425 0 ,00572 + 60 
As the velocity of the stars is proportional to sec. Jd, | have 
endeavoured to represent the resulting values of y by a formula of 
the form : 
d 
y = A — secd= Atandsecd. 
: dd 
The four means given above, give four equations of condition for 
solving the quantity A. Each equation has been given a weight 
equal to the number of nights which have contributed to the equation. 
I find A= + 0.00401. 
y is thus represented by the formula: 
y = + 08.0040] tan d sec d. 
The residuals between the observed and calculated values of 7 
for the four zones are as follows: 
— 0s,00020, — 0s,00001, + 05,00057. — 08,00079 
The correction for the time of transit of a star with decl. d;, when 
the mean decl. of the fund. stars is d, now becomes: 
di 
0s.00401 
= —— for d sec d dd = 08,230 (sec J; —sec d) 
EE 
bg 
11° 
If d; > the time of the transit must be corrected by a positive 
quantity. 
If d; < d the time of the transit must be corrected by a negative 
quantity. 
in other words: 
Stars with high deel. are registered too early. 
i . smaalhéted. a: r 5, late. 
