190 



CARNEGIE INSTITUTION OF WASHINGTON. 



PRESSURE IN THE SOLAR ATMOSPHERE. 



St. John has made an investigation of the pressure in the sun's reversing 

 layer from changes in the relative wave-lengths in the sun and arc of lines 

 which belong to groups affected differently by pressure. This makes a severe 

 demand upon the accuracy of the wave-lengths and of the pressure-coefficients 

 per atmosphere, since, to determine fractions of an atmosphere with precision, 

 the ratio of the wave-lengths of lines in the two pressure-groups must be known 

 to a few parts in 10,000,000, A slight change in the adopted values of the 

 pressure-coefficient is sufficient to change the sign of the result. The determi- 

 nation from a single pair of lines is liable to great error and results of value 

 require a large quantity of data. 



Other conditions being the same, the use of lines showing great differences 

 in pressure shift would seem to be preferable, but such lines belong, unfortu- 

 nately, to the very unstable groups C5 and d, for which the determination is 

 seriously complicated by pole effect in the arc, even under pressures as low as 

 one atmosphere. 



Though the difference in pressure coefficient for lines in the comparatively 

 stable groups a and b is small, the precision of its determination and the 

 accuracy of the wave-lengths are so high that the mean from a large number 

 of lines of these groups should give a good idea of the possibilities of the 

 method. The results from the Mount Wilson data are: 



The conclusion from these observations is that the method is too sensitive 

 to slight errors in the data to yield results accurate to fractions of an atmos- 

 phere, but that the pressure in the reversing layer is below one atmosphere and 

 probably in the neighborhood of zero. 



The method assumes that effects depending upon wave-length are elimi- 

 nated by using pairs of lines in the same spectral region. While this would be 

 true for gravitational displacement, it does not eliminate the effects of radial 

 motion unless the levels of effective absorption are the same. It is possible, 

 and, in fact, probable, that the results are influenced by radial motions vary- 

 ing from level to level, and that more consistent results would be obtained 

 by closer attention to level. 



An effect which bears upon this result is the increase of ionization over 

 faculse, shown by the increased intensity of enhanced lines. If the amount of 

 this effect is interpreted by means of an ionization-pressure-temperature 

 diagram based upon Saha's tables, a probable partial pressure somewhat less 

 than 10 ~^ atmospheres is indicated. 



RADIOMETRIC OBSERVATIONS OF THE SUN. 

 The sensitive types of thermo-couples used in measurements of stellar 

 radiation have been employed by Pettit in observations of the sun's radiation, 



