242 CARNEGIE INSTITUTION OF WASHINGTON. 



with the center of the sun) is not primarily due to the Stark effect. 

 Any residual influence of an electric field must be extremely small, 

 demanding special methods of measurement for its detection. 



(5) Further studies have emphasized the significance of the opposite 

 magnetic polarity which characterizes the principal members of binary 

 spot-groups. 



(G) Before and after the recent sun-spot minimum, the preceding (or 

 following) spots of bipolar groups have been found to be of opposite 

 polarity in the northern and southern hemispheres. 



(7) In each hemisphere the high-latitude spots of the new cycle are 

 opposite in polarity to the low-latitude spots of the old cycle. 



(8) On ''flash" spectra taken without an eclipse 1,024 bright lines 

 have been measured. These surpass in number the '' flash " lines photo- 

 graphed in the same region at eclipses and represent a lower level in 

 the solar atmosphere. 



(9) Double reversal universall}" characterizes the stronger dark solar 

 lines in these spectra. 



(10) Elements of high atomic weight are found to lie below the level 

 of iron and other lighter elements, in harmony with previous results. 



(11) As there appears to be no clear evidence of systematic displace- 

 ment of the bright lines, the results do not support the hypothesis of 

 anomalous dispersion. 



(12) The dark-line spectrum on these plates differs remarkably from 

 that of the sun's center, notably in the great weakening or strengthen- 

 ing of certain lines and the appearance of a few new ones. 



(13) An apparent band spectrum, possibly due to magnesium 

 hydride, has been found between X5050 and X5150. 



(14) A preliminary stud}' of the displacements of solar lines at the 

 sun's center gives no evidence of the existence of the gravitational shift 

 toward the red predicted by Einstein. 



(15) The results are equail}^ opposed to Julius's anomalous dispersion 

 theory. 



(16) A discussion of the displacements indicates that in the higher 

 levels of the solar atmosphere they are due to downward velocity of 

 the vapors, which decreases tow^ard lower levels and finally vanishes. 



(17) Below the level of solar lines of intensity 4 or 5 the effect of 

 pressure appears, increasing rapidly from about one atmosphere to 

 considerably higher values at greater depths. 



(18) The great mass of the vapors comprising the solar atmosphere 

 is condensed within a very thin layer close to the photosphere. 



(19) A new discussion of Mitchell's eclipse spectra yields results in 

 complete harmony with those derived from radial displacements in 

 spots. 



(20) These "flash" spectra indicate levels for iron lines ranging from 

 275 kilometers for lines of intensity 00 to 806 kilometers for lines of 

 intensity 10-40. 



