134 | ANNUAL REPORT SMITHSONIAN INSTITUTION, 1944 
These plausible arguments may be confuted, but it is doubtful if so 
complex a proposition could be made altogether clear to the lay reader. 
The simpler course is to show that these same marches of temperature, 
at these same cities, are associated with another common system of dates 
in another series of years, which system of dates has an undoubted solar 
connection. This I shali now show. 
THE SPECTROHELIOGRAPH AT THE OBSERVATORIO DEL EBRO 
The eminent astronomer, Dr. George E. Hale, in his youth invented 
the beautiful instrument which he named the spectroheliograph. This 
device photographs the clouds of vapors of individual chemical ele- 
ments, such as hydrogen, helium, iron, or calcium which float above the 
sun’s surface. Hale’s spectroheliograph found instant favor all over 
the world, and many observatories were equipped with it. Among 
them is the Observatorio del Ebro in northern Spain, which is main- 
tained by the Jesuits. Every available day from 1910 to 1987 the 
monks at Ebro photographed the calcium clouds on the solar surface 
with their spectroheliograph. And not only did they observe, but they 
measured the areas of these clouds as well as their mean distances from 
the center of the sun’s disk, and they published all the measures. 
CHARACTER FIGURES OF THE SOLAR-FLOCCULUS ACTIVITY 
With the help of my assistants, Mrs. Bond and Miss Simpson, I have 
used these Spanish measurements of every day of observation from 
1910 to 1987 to compute character figures. These represent the solar 
activity of a given day as measured by the summation, according to cer- 
tain weights, of the areas of the calcium clouds, or “flocculi,” photo- 
* graphed that day on the sun’s disk. These character figures having 
been assembled by months in 12 groups, it was seen at once that they 
showed sequences of rise and of fall, for intervals of a few days each, 
just as the solar-constant values do. 
Going over the tables with care, I selected dates in each of the 12 
months in the years from 1910 to 1937 when the best examples of se- 
quences of rise and sequences of fall occurred. The period of 28 
years is so long that there was no difficulty in finding enough excellent 
sequences without including doubtful cases. I thus tabulated the 
zeroth dates of the rising and the falling sequences of flocculus char- 
acter figures for each of the 12 months covering the years 1910 to 1937. 
Then the Washington temperature departures from 5 days before to 
14 days after each zeroth date were tabulated in the same way as for 
solar-constant correlation. 
