== E — 
1 
will first increase, show a maximum when H == NX, decrease 
after that and become zero when NX = 0. ijs vertical 
component will reach a maximum when X == 
At the beginning, the behaviour of the ren cOMm- 
ponent at Karang Sago fully agrees with this hypothesis. 
The first maximum has reached at 5.45, ie. four minutes 
after the middle of totality. 
The distance at which the magnetie phenomenon Bloes 
the central shade, is not known. Considering, however, that 
Herm CrayroN |) has found + 500 miles = + 15 minutes of 
time for the metcorological disturbance, we may in this case 
agree upon a smaller distance, as a matter of course; some 
few minutes will certainly come nearer the truth. As a 
minimum we find 4 minutes here. The rate of velocity of 
the central shade across Sumatra being 32 K.M. per minute, 
we find for H == 226, 271 and 317 K.M. resp. with a lagging 
of 5, 6 and 7 minutes. | 
The vertical component, decreasing in the United Se 
where the north-pole is dipping, must have increased in 
Sumatra where the south-pole is dipping, which is confirmed 
by the observations at Batavia. It is to be regretted that 
the variations for the vertical force at Karang Sago are 
wanting, because they might probably have recorded the 
moment of the crossing of the magnetic phenomenon. 
As a counterpart to the first hypothesis, drawn up in order 
to explain the variations of the magnetic force at Karang 
Sago and Batavia, it is possible to make another, in which 
the influence of the eclipse is assumed to have worked long 
afterwards and to have produced the second large loop at 
Karang Sago as also the eastward hump at Batavia. 
A close study of the material gathered by the different 
observing stations over the whole earth, will perhaps render 
it possible to decide between the two hypotheses. 
1) H. Herm Crayron, The eclipse cyelone and the diurnal eyelones. 
Annals of’ the Astr. Obs. Harvard College. Vol. XVIII p. 1. 
