ASTRONOMY: HALE AND ELLERMAN 
105 
correct impression of the hydrogen fiocculi from positive reproductions 
of spectroheliograms. For this reason the stereoscopic picture is repro- 
duced as a negative, and in this the bright structure gives a fairly good 
idea of the appearance of the absorbing hydrogen (corresponding in 
reahty to dark floccuH) on the original negative. Figure 4, in order 
to be directly comparable with figure 2 and figure 3, is reproduced as a 
positive, and the floccuH shown by it are therefore the dark regions, 
which occupy a somewhat smaller total area than the brighter spaces 
between them. Langley estimated that the bright 'rice-grains' cover 
less than one-fifth of the total area of the photosphere. The corre- 
sponding ratios for the calcium (Ho) fioccuh range from 10%, when 
only the smallest and brightest floccuH are included, to 30% and more 
when the fainter floccuH are also measured. In the case of hydrogen 
{Ha) the ratios vary from 40 to 60%, depending upon the Hmit of dark- 
ness adopted for the faintest objects included. These results are derived, 
moreover, from only a small number of measures of minute floccuH in 
thinly occupied areas, and are subject to much uncertainty because of 
the very wide range of intensity of the floccuH, which renders it difficult 
to distinguish the less conspicuous ones from the background. The 
ratios may therefore be considerably modified as the result of a special 
study of this question, which is now in progress.-' 
The smallest of the dark hydrogen floccuH shown in figure 4 are about 
2" in diameter, or approximately twice the diameter of the smallest 
calcium floccuH shown in figure 3 at the H2 level. 
The foregoing results, in harmony with those of Langley, Secchi, and 
Evershed, support the view that the photosphere and the gaseous atmos- 
phere above it are formed of columns of hot gases, rising by convection 
from the interior of the sun. At the photospheric level precipitation 
may occur of any materials refractory enough to withstand the high 
temperature, or if these substances do not exist, the conditions may be 
such as to cause the gases to emit a continuous spectrum. Higher up, 
where calcium and hydrogen persist, the radiation of the H and K lines 
in the gaseous columns still exceeds that of the intervening regions, 
and bright calcium floccuH are consequently recorded. At the still 
higher level represented by the Ha line, the reduced temperature causes 
the hydrogen to show its presence mainly by absorption, so that the 
hydrogen floccuH are darker than the background of generally diffused 
gases.'' The stereoscopic picture (fig. 5) permits this hypothesis to 
be further tested. 
Helmholtz estimated the minimum angle between two objects just 
separated by the unaided eyes to be one minute of arc, but Pulfrich has 
