REVIEW OF ASTRONOMY PUISEUX. 135 



their interpretation is complicated and the results change according 

 as we consider some special spectrum line or the diverse parts of the 

 same line. For Evershed the dominant fact is the general expanding 

 out of the metallic vapors as they leave the border of each spot. St. 

 John finds that the centripetal tendency again becomes predominant 

 above a certain elevation. The analogies which have been attempted 

 between sun spots and cyclones or the Avhirlpools in water currents 

 give little satisfaction. 



The ascending movements which the spectroscope records toward 

 the center of the disk of the sun are not as rapid as the horizontal 

 movements, but it is not a rare occurrence for them to be accelerated 

 as if the weight was opposed effectively by a repulsive force. These 

 vertical velocities, in every case, are sufficiently great to make us 

 consider very hazardous the attempt of Schulz to revive the former 

 theory of Kirchhoff concerning the general constitution of the sun. 

 According to that theory the sun is liquid up to the level of the spots 

 and the latter are floating scum. Every difficulty is removed by 

 that theory relative to the existence of a continuous spectrum but 

 not relative to temperature and velocities. Fowler prefers to admit 

 the existence in the sun of some unknown physical agent capable of 

 maintaining certain refractory elements in a pulverulent state at 

 temperatures above 6,000° C, the temperature above which pyrhelio- 

 metric measures show that the sun must be. We must resign our- 

 selves for a long while yet perhaps to see Nature use in the stars far 

 more powerfid sources than those at our disposal in the laboratory. 



Deslandres and d'Azambuja continue to devote themselves to the 

 isolation of the light of the central parts of the strongest lines of 

 the solar spectrum and its use in their solar photographs, and that 

 choice is justified by the striking originality of the photographs ob- 

 tained. The astronomers at Meudon, despite the doubts raised by 

 A. Buss, maintain an essential distinction between " filaments " and 

 " alignements." The latter, fainter but more prolonged, are char- 

 acteristic of the upper layers. They appear as far as the greatest 

 latitudes and are not dependent upon the Schwabe cycle. 



The existence of the Zeeman phenomenon at the border of the 

 spots, shown by Hale, as we Imow, has furnished him w^ith a means 

 of measuring the local magnetic fields. We see no other probable 

 origin for these magnetic fields except the motion of electrified par- 

 ticles, but one would suppose that electricity would be conducted 

 with great difficulty in as rare a medium as that whicli surrounds 

 the sun. This objection has been very much weakened, although 

 not nullified, by the recent experiments of Harker, who found that 

 a rarified gas becomes an effective conductor for electricity in the 

 neighborhood of a body at a very high temperature. 



