— 226 — 
The D-curve (see the plate) exhibits a slight wave at 11^ a 
and afterwards a larger one at 19^ 15"; but such waves 
also occur on other days and at other hours of the day. 
Comparison of the eclipse-day curve with the average one 
for April allows two interpretations. 
The one is that an easterly deflection of the needle took 
place during the hours of greater obscuration 11^ a— 1^ 30” p, 
the other, that before the middle of the eclipse a westerly, 
afterwards an easterly deflection occurred; but the latter inter- 
pretation seems rather constrained and I will stick to the first. 
The amount of easterly deflection may be esteemed not to 
exceed 0.'5. 
In the H-curve we notice a wave between noon and 2^ p 
showing an increase before the middle of the eclipse and a 
decrease after it. The value of H at the secundary maximum 
at 12:30" equals that at the secundary minimum at 1" 15" p, 
but the normal curve requires an increase of 5 y, accordingly 
the amplitude of the eclipse-effect may be esteemed to amount 
to 5 y. 
The eurve of the Z-magnetogram shows small and rapid oscil- 
lations, typical for.a disturbed equilibrium of temperature, and 
probably the consequence of air currents within the instrument. 
As far as can be judged the curve does not show any 
wave during eclipse-time. Notwithstanding the comparison 
of the curve, which is drawn with the aid of readings and 
the published hourly values, with the normal curve obviously 
points to an increase of Z of about 4 y during eclipse-time. 
Total Eclipse of the Sun of August 9, 1896. 
Course of central line: Norway, Siberia, China, Japan. 
Minimum distance from Zi-Ka-Wei + 22°. Maximum phase 
at Zi-Ka-Wei + 0^ 52" p. 
As may be seen in the D magnetogram the lamp or the 
recording-cylinder was displaced six times between 0% 30" p 
i 
E 
| 
| 
4 
a 
P 
