NO. 2 SOLAR CHANGES ABBOT I3 



sity of reception of long-range radio-transmission. Figure 8 gives 

 three curves, as plotted by Austin. The first represents the monthly 

 mean departures from a three-year's mean of the radio reception at 

 Washington from several distant stations. The second curve shows 

 similar departures of the Wolfer sun-spot numbers, and the third, 

 the corresponding departures of the Smithsonian provisional solar 

 constant values, given above in table 2. Dr. Austin has informed me 

 that the probable error of his observations of individual days on 

 radio reception is from lO to 20 per cent, so that in the monthly 

 means it must be from 2 to 4 per cent. The general accord of the 

 three curves seems to indicate that the departures of monthly mean 

 radio reception from average values are almost wholly dependent 

 on the state of solar radiation. 



In what has been said, we have been concerned only with long- 

 interval changes of the solar radiation, and associated terrestrial phe- 

 nomena. My colleagues and I have long believed that these changes are 

 due to changes in the effective temperature of the sun's radiating 

 surface, which depend on the activity of convection in the sun's 

 substance. We have noted, also, solar fluctuations of such short inter- 

 vals as a few days. These we attribute to the rotation of the sun 

 which brings successively opposite to the earth regions of unequal 

 radiating power, or i:)erhai)s. rather, of unequal absorbing or scatter- 

 ing power, on the sun's surface. 



In harmony with this idea, the planets, which lie in different 

 directions, viewed from the sun, will succcssiz'cly feel the changing 

 influence of each inequality of the solar surface, as the rotation of 

 the sun brings such inequalities into- line with the planets successively. 

 j\s the sun's equator is inclined to the ecliptic, the interval of time 

 to be allowed differs a little from that which would be the case if the 

 inclination were zero. Furthermore, if one observes from the earth 

 some effect ujjon a distant planet, due to a variation of solar emission, 

 the time of observation will be influenced by the time required for 

 light to travel out to the distant planet and return to the earth. For 

 the causal irregularity of the solar surface is moving by solar rotation 

 while the light is on the wa}". 



In Volume IV of the annals of Astrophysical (Observatory, page 

 190. figure 13, a not unfavorable test of this hypothesis is given, 

 de])ending on a comi)arison of observations by Guthnick of the planet 

 Saturn as compared with Smithsonian observations of the solar con- 

 stant of radiation. 



While we are considering short interval solar variations, I give in 

 figure 9 a series of curves taken from table 6 of Clayton's paper 



