— 218 — 
Only the knowledge of the effect connected with other 
eclipses can elucidate the question, but opportunities for 
observation being rare and the eircumstances generally in- 
favourable, it will be of importance to examine the magnet- 
. ograms and observations during recent eclipses. 
There appears to be no necessity for us to confine ourselves 
to total eclipses only, because a partial eclipse will, no doubt, 
give an effect as well, as an effect has been observed beyond 
the time and belt of totality. 
For this reason I have made use of the opportunity afforded 
during my stay in Atjeh in March 1904, whilst occupied with 
magnetic surveywork, to take observations at Sabang on the ` 
day of the annular eclipse of March 17%. 
This eclipse also offered the peculiar circumstances of being 
observed from a spot near the magnetic equator and occurring 
at noon; but it is to be expected that, later on. when the 
effect has become better known, these very circumstances 
will prove to be extremely useful. 
For the sake of comparison copies of the magnetograms 
of Manila and Zi-Ka-Wei have: been asked for; from the 
latter observatory also for the eclipses of 1882, 1887, 1894 
and 1896. 
The direetor of the Manila observatory has had the kindness 
to send a calque of the curves of March 17, 1904, together 
with tables of hourly values for a certain number of days 
in February and March in behalf of the magnetic survey. 
The Rev. Father J. pe Moiprrr of the Zi-Ka-Wei observatory 
has been so kind as to forward a copy of the magnetograms 
for March 16—18, 1904, together with the original sheets 
for the eclipse-days in 1882, 1887, 1894 and 1896. 
Besides hourly values for Zi-Ka-Wei are to be found in 
the Bulletins of the latter observatory. 
The best plan will be to treat the eclipses in chronological 
order, confining ourselves to the observations at stations in 
ast- Asia, 
D 
