( '^5 ) 



Lightship Tersciielliii.ucih.uik, N. Lai. 53°27', L. G. 4°52' 



watch-observ. ((i limes daily), magnetic, (1884 — 1908). 



Winter v — 3.9 cos {nl — 344°) + 11.8 cos (2 nt — 319°) 

 u — 3.1 cos { nt — 29 3°) +11.1 cos (2 n^ — 56°) 

 A^"5T° A ^^263° 



Spring V = 20.3 co.v (y^/ — 249°) -\- 12.6 cuv (2 7it — 344°j 

 'w = 1 8.9 co.v (?2i! — 350°) + 10.1 cos (2 w^ — 75°) 

 A~^l59° ^^^^269° 



Slimmer v = 21.6 cos {nt — 240°) -f 9.3 cav (2 nt — 348°) 

 M = 30.9 cos {nt — 354 °) + 11 .5 tw (2 nt — 94°) 

 A = 246° ~^==r254° 



Autumn v =: Ö.0 cos {nt — 248°) + 11.9 cos (2 ?i^ — 326°) 



u — 13.1 cos {nt — 358° ) + 17.6 cos (2 ?i^ — 52°) 



A = 250° A^274°^ 



Year v = 12.3 cm' (?z^ — 249°) + 11.1 cos (2 nt — 334°) 



?f =: 1 6.1 cos {nt — 351° ) + 11.9 cos (2 n^ — 66°) 



A = 258° ""A'^^^S^ 



If, for the present, we lea\'e the monodiurnal motion out of con- 

 sideration, it appears that, whereas the anguhir values of the semi- 

 diurnal variation show a close agreement, the amplitudes do not 

 agree in so far that sometimes the north- and sometime-^ the east- 

 component is the greatest ; on the average the east component is 

 somewhat greater, but the difference is so small and variable, that 

 a serious objection arises against calculating the friction coefficient 

 by means of formulae (JO) and (11), when «', according to theory 

 is equalized to zero because 



X^—k, = 270°. 



If A:=i B the windellipse approaches to a circle because cos A 

 is also a small quantity, according to theory as well as to observa- 

 tion, and the angle of deviation becomes undetermined. 



Different other methods however can be chosen for calculating 

 the value of k by means of form. (5), (6) and (9), as it is sufficient 

 if only one quantity or relation be assumed to be equal to its 

 theoretical value. 



We might e.g. assume that the theoretical value 



H, = 7, //, cos rp 



were accurately true; then, putting 



