MOUNT WILSON OBSERVATORY. 193 



The underlying causes of the sun's general magnetic field remain 

 obscure. The evidence bearing on the local-whirl hypothesis is 

 unfavorable to its acceptance as an adequate explanation of the 

 existence of the field. 



The investigation for determining the position of the sun's magnetic 

 axis and its period of revolution about the solar axis of rotation is 

 now approaching an end. Mr. van Maanen has measured 325 addi- 

 tional spectra, which completes the series, about 2,500 in all, photo- 

 graphed on 74 days between June 8 and September 25, 1914. 



The provisional reductions have been finished by Miss Wolfe, who 

 has done all of the extensive numerical work connected with the 

 investigation. The least-squares solution for the inclination, period, 

 and epoch have been revised, but with no essential modification of 

 the results previously reported. The data will now be discussed by 

 zones of heliographic latitude in order to see if the sun's field deviates 

 appreciably from that of a uniformly magnetized sphere. 



SOLAR ROTATION. 



The series of solar-rotation observations is now in its fifth year. 

 In accordance with the original plan, the same instruments, observers, 

 and measurers have been employed, and it is hoped to continue them 

 throughout the sun-spot cycle. 



The variations during the period 1914 to 1918 show no evidence of 

 periodicity and apparently depend upon temporary local conditions 

 in the solar atmosphere and in particular upon the high differentiation 

 of the sun's surface furnished by a solar image 420 mm. in diameter. 



The slight indications of changes in the sun's rotation shown in the 

 past are of such an order that they may in great probabiHty be referred 

 to temporary conditions then prevaiUng in the sun or to other cir- 

 cumstances of the observations. 



The reflecting-prism system in use consists of three prisms each 2 

 mm. long, two of the prisms receiving light from one limb and one 

 from the other limb. By rotating the spectrograph, the east and 

 west limbs are interchanged with respect to the prism system. It was 

 hoped by such an interchange to obtain a check upon the observations, 

 but differences of several per cent often appear. These differences 

 are real, for upon exposures taken in immediate succession, wdth 

 interchange of prisms, they may amount to 6 or 7 per cent, while the 

 individual lines are mea,sured with a mean deviation of a third of 1 

 per cent. Moreover, they must be due to solar conditions. In using 

 the red region of the spectrum it was found that the atmospheric lines 

 are undisplaced upon plates showing extraordinary displacements of 

 the solar lines. Further, observations with a duplicate prism system 

 and an image 164 mm. in diameter, in which case the integrating 



