MOUNT WILSON SOLAR OBSERVATORY. 211 



for close pairs. The investigation promises a means of fixing the 

 limits beyond which the measurements of an individual observer are 

 reliable and of determining approximately his systematic error within 

 such limits. 



The multiphcity of causes which may be concerned in the displace- 

 ment of Unes in terrestrial and solar spectra rendered much preliminary 

 investigation necessary before a hopeful attack could be made upon 

 the problems involved. Investigations of the preceding year showed 

 that, within the limits of error, systematic effects of anomalous refrac- 

 tion are absent from the general disk of the sun, while those of the 

 present year have shown the absence of displacements required by the 

 equivalence principle of relativity, and have developed a source of 

 iron Unes free from the disturbing influences of pole effect. By the 

 ehmination of anomalous refraction and relativity, and the develop- 

 ment of a source by comparison with which the relative displacements 

 of the iron lines in solar and arc spectra may be referred to conditions 

 in the solar atmosphere, it is possible to attack the related solar prob- 

 lems with greater prospect of success. 



It becomes increasingly evident that data based upon several ele- 

 ments and covering an extended range of wave-length will be necessary 

 for definitive determinations, and that further investigation should 

 be made of conditions obtaining near the solar limb. Investigations 

 now in progress bear upon these questions and have for an inunediate 

 end the establishment of a system of solar standards in terms of 

 international units. These standards will eventually be derived from 

 three sets of observations, for which an interferometer and two spectro- 

 graphs are available. 



SOLAR ROTATION. 



The series of plates at solar latitudes 0° and 45° is being continued 

 with the 150-foot tower telescope, with the object of obtaining obser- 

 vations over an extended period under as nearly as possible the same 

 instrumental conditions. In view of the differences found by obser- 

 vers, it is desirable to have a long homogeneous series for determining 

 whether the indicated variations in solar rotation period are real or 

 depend upon instrumental conditions and personal equation. Obser- 

 vations with the 150-foot telescope will be compared with simultaneous 

 observations made with a duplicate prism system used, sometimes with 

 the 60-foot tower telescope and sometimes with the 60-foot Snow 

 instrument, the advantage of the latter being the complete achrom- 

 atism of the solar image. 



The total variation found in the results for the years 1914, 1915, 

 1916, and 1917 is about 3 per cent. This difference, however, is not 

 between the values for the extreme dates and can not be ascribed to a 

 secular or long-period variation in the rotation rate. It is probably 

 due, in part at least, to local drifts in the sun's reversing layer. 



