406 
MR. J. EVERSIIEI) ON THE SOLAR ECLIPSE OF 1900, MAY 28. 
the 32 lines exceeding Howland's intensity 8 are present in the flash and none 
under his intensity 3. 
This general law of correspondence of intensity between bright lines and dark 
lines is, however, far from being exact in detail even with the selected lines used in 
these comparisons, and the average intensities of the bright lines are in some 
instances made up of rather widely diverging units. 
This is more particularly tlie case witli the weaker dark lines 'of each element, 
which are often of abnormal intensity in the tlasb. In the case of chromium most of 
the flasli lines corres]:)onding with solar lines of intensity 2 and 3 may be considered 
almormally strong, for the average intensities for these lines are greater than the 
average of the lines corresponding with the solar lines of intensity 4. 
The percentage columns show also that many dark lines of medium intensity may 
be absent in the flash, whilst other weaker lines are present. 
It must be remembered that estimates of intensity in the flash spectrum, however 
carefully made, are liable to considerable errors for 'several reasons. The great 
weakening of the spectrum near the ends of the materially affects the 
percentages of the weaker lines as given aljove, and the low dispersion of the plates 
com[)ared with tlmse cm which Howland’s estimates were based introduces other 
sources of discrepancy. Moreover, Howland’s table itself is admittedly a 
“ j)reliminary ” table, in which some of the assignments of origin may be erroneous 
or inconpdete, lines having a single origin assigned being really made up of two or 
more elements. 
Iron Lines in Sun and Flash. 
Including all isolated lines in Howland’s talde assigned to Fe only l)etween X 3500 
and X 5000, exce|)ting those wHich are obscured in the flash by strong hydrogen 
• and calcium lines. 
liowtAXn’s intensity 
in O. 
Number of line.? 
in O. 
Nnmlier of lines 
in flash. 
Percentage of lines 
in flash. 
Average intensity 
ill hash. 
Under 3 
Very large number 
0 
0 
3 and 4 
94 
12 
13 
5 
5 „ 6 
G6 
26 
40 
o 
O 
7 „ 8 
27 
23 
85 
G 
9 to 14 
13 
13 
100 
8 
15 „ 20 
14 
14 
100 
17 
25 and over 
5 
5 
100 
24 
* Including tliree linos ascribed to Fe only by LoCKYEll at X,\.4179, 4233, and 4515. 
