=D 
The eurve for Batavia, where the middle ot the eclipse 
occurred at 6.0, has a regular course till 5.30, after which, 
however, a decidedly western deflection with increased 
depression of H shows itself, changing a short time after 6.0 
into an eastward movement with increase of H and stopping 
at 7,15. It will be noticed that the curve at 5.45, i.e at 
the very moment when the striking loop movement alluded 
to above took place, is quite smooth, whereas it-shows a 
slight irregularity at 6.15. (Vide table). 
The first hypothesis made in order to explain these irreg- 
ularities is, that the westward movement and first loop at 
Karang Sago as well as the movement towards. the west at 
Batavia were caused by the eclipse; that afterwards, however, 
a common disturbance set in bringing about the second loop 
at Karang Sago and the eastward deflection at- Batavia. 
Looking at the variation-curve for the vertical intensity at 
Batavia on May 18** together with the normal one (May 1887), 
we see that the course is «a. normal one till 5.10, after which 
time a decided defleetion occurs, following which in the 
direction indicated by the broken line, we first get an 
incrcase until 6.20, then a decrease until 7.35, and further 
an inerease again. 
According to our first inns the diein increase may 
be aseribed to the eclipse, the strong decrease to the second 
disturbance. 
From the American observations of 1900 Baver concludes that 
some magnetic phenomenon, attracting the north-end of the need- 
le and following the central shadow, traversed the atmosphere. 
Let us imagine this phenomenon to be connected with a 
height H above the surface of the earth and to be above 
some point at a-distance X from the station of observation. 
The horizontal component of the attracting force will then be 
Gras 
a + 9% 
and, when approaching from-a distance oo, this component 
