DEFLECTION OF LIGHT — DYSON AND OTHERS. 141 
of the mirror of the 16-inch coelostat. By inserting an 8-inch stop 
this was reduced to a hirge extent, and this stop was henceforth 
used throughout; but the defect was of such a character that it was 
clear that it would be necessary to stay at Sobral and obtain com- 
parison plates of the eclipse field in July when the sun had moved 
away. 
The focus of the 4-inch was determined in a similar manner. 
The images, though superior to those of the astrographic, were not 
quite perfect, and here again comparison plates in July were neces- 
sary. Once the focus had been decided on, the breech end was 
securely screwed up to avoid any chance of subsequent movement. 
A few check plates of the field near Arcturus were taken, but 
have not been used. 
11. The following is a summary of the meteorological conditions 
during our stay. The barometer record was interesting in that it 
showed very little change from day to day, in spite of changes in 
tlie type of weather; there was, however, a very well marked semi- 
diurnal variation, with range of about 0.15 inch. The temperature 
range was fairly uniform, from a maximum of about 97° F. toward 
3 p. m. to a minimum of about 75° at 5 a. m. The relative humidity 
(as shown by a hygrograph belonging to the Brazilian Commission) 
followed the temperature closely, varying from 30 per cent in the 
afternoon to 90 per cent in the early morning. 
May is normally the last month of* the rainy season at Sobral, 
but this year the rainfall was very scanty; there were a few after- 
noon showers, each ushered in by a violent gust of wind; and on 
May 25 there was very heavy rain, which was welcome for its 
moistening effect on the ground, the dust hitherto having been 
troublesome to the clockwork although every care had been taken 
to protect it. There was a fair amount of cloud in the mornings, 
but the afternoons and nights were clear in the majority of cases. 
Mount Meruoca, 2,700 feet high, about 6 miles to the northwest, 
was a collector of cloud, its summit being frequently veiled in mist. 
In spite of its cooler climate, the summit would thus not have been 
a suitable eclipse station, and, in fact, nothing of the total phase 
of the eclipse was seen from it. 
12. Although water was generally scarce, we were very fortunately 
situated as we enjoyed an unlimited supply of good w^ater laid on 
at the house. This was of great benefit in the photographic opera- 
tions. Ice was unobtainable, but by the use of earthenware water- 
coolers it was possible to reduce the temperature to about 75°, and 
by working only at night or before dawn development of the plates 
was fairly easy. Formalin was used in every case to harden the 
films, and thereby minimize the chance of distortion due to the 
softening of the films by the warm solutions. 
