THE RADIATION OF THE SUN—ABBOT. 139 
magnetic declination and in the earth’s magnetic force for the corre- 
sponding years. It will be seen how exactly the sun-spot curve is 
reproduced in these fluctuations of the earth’s magnetism, but the 
cause of the connection which is so apparent is not yet well under- 
stood. In the upper curve of the figure are represented the depart- 
ures of temperature for the average of 17 stations in the United 
States, and there will be seen, although not so plainly marked, an 
apparent influence of the sun spots on the temperature of the earth. 
Recent work, much of it at the Mount Wilson Solar Observatory, 
has given us a good insight into the nature of sun spots. They 
appear to be whirls of material coming outward from the inner layers 
of the sun toward the surface, spreading out there like a waterspout. 
The expansion attending decrease of pressure on the gases causes a 
fall of their temperature, so that the sun spots are cooler than the 
surrounding parts of the sun, and this is the reason why they seem 
dark. The whirling matter contains electrical charges, which, by 
virtue of their rotation, give rise to magnetic fields, as shown long 
ago by Rowland. The presence of magnetic fields in sun spots has 
recently been established by Hale. 
At certain times the moon interposes between the earth and the 
sun and cuts off the sunlight, so that we are able to see the objects 
which are surrounding the sun and usually lost by the intense glare 
of the sky. Such occasions are called ‘‘total solar eclipses.” As 
the moon is but little, if at all, greater in angular diameter than the 
sun, the cone of shadow cast by the moon only a little more than 
reaches the surface of the earth, and sometimes, indeed, fails to reach 
it at all. When the cone reaches the earth’s surface, and we have a 
total eclipse, there will be a belt, not more than 200 miles wide, but 
sometimes several thousand miles long, upon the earth’s surface, in 
which the total eclipse may be observed at some time of the day. 
Frequently the belt of totality passes over inaccessible regions of the 
earth, as, for instance, the North or South Pole, or falls upon parts 
of the ocean where it is impossible to use delicate instruments. The 
longest possible period of totality at any one station is seven minutes, 
and in general the total eclipses average about three minutes in length 
Thus only a very little time can be used in eclipse observations, and 
yet the information to be gained at such times is so valuable that 
observers often spend months in preparation and travel thousands 
of miles to observe them. 
Figure 2 of plates 1 and 2 show the total eclipse of the sun. The 
first is from a drawing of Calvert prepared from photographs by 
Yerkes Observatory observers at Wadesborough, N. C., in the year 
1900, and the second is from a drawing by Mrs. Abbot from plates 
of the eclipse as photographed by the United States Naval Observa- 
tory parties in Spain and Africa in the year 1905. In each photo- 
