207 
Experiments at San Antonio. 
On the Friday (the 16th) preceding the eclipse the instru- 
ments were fitted up at San Antonio, three miles or so from 
Puerto de Santa Maria; in this operation Lord Lindsay kindly 
afforded most valuable assistance. On that and every sub- 
sequent day till that of the eclipse, Thursday the 22nd, I was 
employed practising the manipulations of the telescope and 
observing the light from the sky, the clouds, the moon and 
various terrestrial objects for polarization. 
At first the light from almost every object seemed to be 
polarized : this was accounted for by the want of perfect black- 
ness in the tube of the telescope, and supposed to arise from 
reflexion in the interior. To correct this a projecting nozzle 
lined with black paper was first tried but found insufficient. 
Black velvet was next had recourse to and apparently with 
success. 
On the night of the 17th or rather the morning of the 18th 
I was observing the moon and noticed that the light appeared 
to be polarized. Not expecting to have been able to detect this 
polarization of the light reflected from the moon I called out to 
Mr Moulton who was observing at the same time, and he had 
also seen bands across the moon with his instrument. 
On repeating the observation I detected no polarization and 
Mr Moulton saw no bands. We were led to account for this by 
the presence of thin clouds: when the clouds were thick and 
when the sky was quite clear I could detect no difference of 
intensity on turning the analyser. When there were thin 
clouds, and even when no clouds were visible to the eye, but 
when their existence was rendered probable by the neighbour- 
hood of visible clouds, the polarization became manifest. 
I subsequently repeated these experiments near the sun at 
noon when the sun was so clear that I had to use a dark glass; 
there were however light clouds about. I observed the sky in 
the immediate neighbourhood of the sun, approximately within 
16—2 
