— 241 — 
ruption at 11" 10", i.e. 18 minutes before the first contact. 
From 1^3" p the curve seems to be affected by a slackening 
of the normal decrease, reaching its maximum 36 minutes 
before the last contact. 
Accordingly we have again: 
a decrease before the maximum phase of + 
an increase after „, HEE 
The total range thus mans to the same value viz. 6 y 
as that during. the eclipse of 1901. 
Vertical Intensity. 
The Z-curve shows some minor waves, which have no 
marked relation to the eclipse; the considerable positive 
wave during the eclipse of 1901 has not repeated itself. 
This is remarkable indeed, because this time the D-wave was 
larger and the H-wave the same as in 1901. 
Zi-Ka-Wei. 
Minimum distance of central line south of Zi-Ka-Wei + 10°. 
Maximum phase at + 3» 0” p. 
Declination. 
In the pencil copy of the magnetograph-curves may be 
clearly seen that'the D-curve has a secondary wave during 
eclipse-time. Having no readings for other days in March 1904 
at my disposal to deduce a normal, I have compared the curve 
with the one for the next day, also drawn in the copy. 
I have connected the points of the curve of March 17* for 
2Mp and 5^p by a curve parallel to the one for the next 
day and measured the deviations of the true curve from tbis 
one (see dotted part in the plate). 
4 y. 
29. 
gh ("p 0.'0 gh 45m 0.5 
15 0:3 A u 0. 5 
30 SE 15 0.4 
45 0.2 30 0.2 
9 9 0.4 45 0.1 
vi max. phase ni IA 0.0 
30 0.4 
