A NEW SOLAK THP:()RY 



173 



,sutr,s .surface. In this rt'8})ect tli(^ theory will l)e submitted to an 

 exhaustive test in ui.v pTiper in the Annals. In one important point it 

 involves a radical deviation from the views hitherto held. So far 

 investigators have almost unanimously adhered to the traditionaUiew 

 that an increase in the dynamical forces at the sun\s surface indicates 

 at the same time an augmentation of his light and lieat radiation into 

 the universe. A theory founded on this assumption would ha\'c to 

 account not only for the extra expenditure of force into s])ace. but 

 also for the simultaneous increased development of force in the sun. 

 But in the theory here proposed the exactl}' opposite conclusion is 

 arrived at. Here the forces which we see acting on the sun are called 

 into existence hy the accunuilation of such parts of his radiating energy 

 as have been prevented from being thrown oti' into the universe. Thus, 

 a surplus of energy working on the sun means a deficit of energy com- 

 municated to space. 



It will be important, then, to ascertain how far this conclusion can 

 be verified b}' observed facts. Modern researches seem, indeed, to 



/ly. 2. 



D enatu>ns of (Ae Mean. Arinu.ai Tt'/ryoerature from Che Norma /.. 



Voj-KUicns LT. (AeArjxucU Mt'Ufi Temperatures 



corrol)orate this theoretical result. If the theor\ ))e true, the tem- 

 })eratui'eof the solar layers inside the al)sor)>ing atmospliere should be 

 higher at the maxima than at the minima of solar activity, while the 

 temperature of a bod}" in space, which receives its heat from the sun, 

 should vaiy inversely. In proof of the first conclusion I may refer to 

 Sir Norman Lockyer's results with regard to the beha\ior of the lines 

 widened in the spectra of sun spots, from which he infers that the 

 luatter composing the spots unist be of higher temperature at the 

 times of maxima. The second conclusion, on the other hand, is cor- 

 roborated ])y all the more important researches which have recinitly 

 been made regarding a connection b(>tween the changes of terrestrial 

 temperature and solar activity. Of some of these I subjoin the main 

 results in tigure 2, which exhibits the ()bserv(>d changes in the mean 

 annual temperature at tropical and subtropical stations and the corre- 

 sponding variations of solar activity. It will ))e seen that for the whole 

 period from 1821 until 18*,)S the temyjerature curve follows most accu- 

 rately the fluctuations of the inverted spot curve, thus so far proving 



