475 
GENERAL DISCUSSION OF SRECTROSCOPIC RESULTS. 
in all tlie images obtained before mid-eclipse, and in those obtained after that phase 
only the stronger lines are impressed, the moon’s liml) having occulted the stratum 
very rapidly, notwithstanding that the motion was nearly parallel to it. This would 
indicate an extreme sliallowmess of the layer near the pole. 
In the mid-eclipse photograph, No. 11, the continuous spectrum being Imoken up 
into narrow bands, the flash spectrum arcs can be traced right across the polar region 
near the more refrangible end of the plate. In the portion of spectrum between F 
and K the bands coalesce from over-exposure and obscure the bright arcs entirely. 
Ihe highest latitudes in which really good images of the flash S])ectrum occur are 
— 1 0 to — 77° on the east side in No. 9, and — 76° on the west side in No. 13 ; and the 
lowest latitude is in — 36° to — 41° on the west side in No. 11. Intermediate 
between these there are the excellent images in latitude — 56° west and — 64° east 
in^ No. 10. From this material comparisons can be made between the spectra at 
faiily high latitudes and those at mid-latitudes, and as a check on the results the 
east and west limbs at about the same latitudes can be compared. 
All these images are indicated on Plate 2 by arrows at the ends of the spectra, and 
the position of the south pole is similarly showm for each spectrum. In Plate 3 
a limited portion of the spectrum is shown for the three images which were measured. 
These aie on a scale equal to 4'3 times that of the original negatives, and the curved 
arcs have been converted into linear spectra by means of a cylindrical lens during the 
]uocess of enlargement. Great care was taken to avoid the production of spurious 
lines due to defects in the negatives. 
Comparing the two high-latitude spectra shown in the upjier and lower figures of 
Plate 3 with the mid-latitude spectrum placed between them, it is not easy to detect 
differences which can fairly be ascribed to latitude. It may be noticed that the 
titanium line at about A. 3900 and the aluminium line at \ 3944 are both relatively 
weak in the upper spectrum (latitude - 74° East) compared with the middle spectrum 
(latitude - 41° West). But in the lower spectrum, from an equally high latitude on 
the opposite side of the pole, these lines are as strong as in No. 11 spectrum. 
There are many other minor differences in relative intensities betw^een the three 
spectra, as will be a])parent on comparing the three columns of intensities given in 
I able I., but these seem to bear no relation to difference of latitude. 
A special effort was made to discover any modification of intensity in the enhanced 
lines near the pole, and the average intensity of all the more prominent enhanced 
lines of iron and titanium in Nos. 9 and 13 spectra was compared with the average of 
these hues ill No. 11 spectrum, making due allowance for the greater intensity of 
No, 11 spectrum, as a whole, compared with the others. 
Ihe result is shown in the following table ;— 
3 P 2 
