54 THE ROYAL SOCIETY OF CANADA 
A previous comparison of the two Angstrom instruments showed 
that No. 147 was 0-4% higher than No. 78. The latter instrument 
was used as the standard and the readings of the former were ex- 
pressed in terms of the standard by multiplying them by -996. The 
curves for Angstrom No. 78 and the Callendar are similar to those 
obtained in the first comparison but the curve for No. 147 when covered 
with glass agrees fairly well with the Callendar in the morning, but as 
the day advances the Callendar readings become greater and continue 
so for the rest of the day. This result was obtained on all days on 
which comparisons were made. Taking the. percentage difference 
from the standard it shows that during the greater part of the day 
the pyrheliometer covered with glass bears about the same ratio to 
the standard but that as the sun gets near the horizon the readings 
increase relatively to the standard or the glass absorbs less and less 
of the sun’s radiation. 
The comparison further showed that by covering the Angstrom 
instrument with glass its readings were then less than the Callendar. 
This result indicated the desirability of comparing the Angstrom 
with the Callendar when the glass bulb protecting the bright and black 
platinum resistances was removed, and an old sunshine receiver was 
obtained from the Cambridge Scientific Instrument Co. In this re- 
ceiver the platinum thermometers are mounted on a circular disk of 
mica, 2 inches in diameter, one half being black and the other bright, 
(the protecting bulb was not exhausted) while in the other type they 
are mounted in checker board fashion on a square disk of mica. This 
receiver was not standardized when received and it was necessary 
therefore, to standardize it before removing the glass bulb. The 
receiver was mounted as before in a tube about 18” long, the surface 
of the receiver being exposed perpendicular to the sun’s rays. Its 
readings on the Callendar recorder with the standard charts were 
taken at regular intervals at the same time as an observation was taken 
with the Angstrom pyrheliometer and the amount of heat received 
from the sun as determined by the Angstrom instrument was divided 
by the reading of the Callendar; in this way the value of a scale 
division on the Callendar was obtained in terms of the Angstrom and 
it was found to be -0444 gramme calories per sq.cm. per min. Having 
determined the scale value with the glass bulb surrounding the re- 
ceiving plate, the next operation was to determine it with the glass 
bulb removed but in attempting to cut away the glass bulb just below 
the mica disk, the bulb and one of the wires were broken and then the 
disk was mounted on an ebonite base, the disk being about 3” from the 
base. In this case the value of a scale division was found 
to be -0420 gramme calories per min. The receiver was next 
