350 F R. Bigelow — Further Study of the Solar Corona. 



The first process was to substitute the resulting values of 



r .6 in the formula for the lines of force, N= — . . The 



3 r 



computed values are collected in table No. II. 



It is observed that the values of N derived from measures 

 on the same ray are not equal. If all the circumstances are 

 fully accounted for, they would be alike and the angle Q 

 would show the same polar distance for the beginning of the 

 line N. But it is seen that there is a marked progression in 

 these values, which indicates that some systematic error ad- 

 heres to the work. An inspection of the case suggests that the 

 projection of the ray from its position in space upon the appar- 

 ent plane of the disk must be taken fully into account, for it 

 is not probable that any large number of the visible rays start 

 from the body of the sun in the very plane of the disk. To 

 such rays the formula should have applied. The problem is 

 therefore to discover at what point of the sphere each ray 

 originates, and to assign angular coordinates to the same. 



The figure illustrates how a ray springing from the surface 

 of the sun is seen projected so as to lie across a series of true 

 3 N-lines represented as 



proceeding from the 

 disk. What was com- 

 puted in the last opera- 

 ation is in fact the 

 particular N-line that 

 passed through the 

 point as measured in 

 its projected situation. 

 We must now discover 

 a means of determin- 

 ing through what angle 

 a the plane of the ray 

 N must be rotated in 

 order to be seen on 

 the plane of the disk. 

 Also we must carefully 

 distinguish between the 

 pole of the corona and 

 the selected pole on the disk, for although these two poles lie 

 in the same plane whose position angle with the sun's axis at 

 the center of the disk has already been given, yet the angular 

 distance of the coronal pole from the plane of the disk, being 

 at this time unknown, has an immediate effect upon the angle 

 a that is being sought. All the planes containing the rays 

 intersect in the coronal axis, and if this was in the plane of the 

 disk the rotation angle a would be the same for all measured 

 points on the ray; otherwise a measured at the pole on the 



Ray projected across a series of N-lines. 



