1458 
all reduced to the lowest value of the resistance, and reckoned from 
zero-positions that were found by interpolation between a series of 
zero-readings, made in the course of the eclipse with the thermopile 
shaded. The shift of the zero was small and regular. 
Plate XI‘) Fig. 1, is a reduced copy of the original mapping of the 
Table I.The deflections observed between 0"28"10s and 0>41™30s, plotted 
on a ten times larger seale, are shown on Plate XI, Fig. 2. These latter 
observations give evidence of the exceptionally favourable condition 
of the sky especially during the middle part of the eclipse. When uniting 
the observational points by a curve, I was quite surprised to find it 
so perfectly smooth and symmetrical, for in our country a sky without 
even invisible haze is a rare occurrence. 
Tie central part of this curve corroborates our conclusion drawn 
from the photographic curve, viz. that the minimum value of the 
radiation was '/,,,, of the maximum. Indeed, -the real minimum 
value could not be reached by the slow apparatus; but if we prolong 
the lower parts of the falling and the rising branch of the curve 
downward as nearly straight lines (beginning at points corresponding 
to 10 seconds before and 10 seconds after centrality), they meet at 
one millimeter above zero; and according to Plate XI Fig. 1, the 
maximum was represented by about 5000 millimeters. 
The rest of the observations ran somewhat less reguiarly, both 
in the falling and in the rising phase of the radiation. From notes 
on sky-condition, made by other members of the party, we could 
afterwards state that the depressions in the series of points exactly 
corresponded to hazy cloudlets passing before the sun. Yet some 
arbitrariness was left in the process of tracing the radiation-curve 
so as to answer to an ideally constant degree of transparency of the 
sky. We simply made the curve pass through the highest points 
(because the observed values could only be too small), and for the 
rest took care that the curvature should vary as regularly as possible. 
Special attention may be drawn to the points 5 (Plate XI Fig. 1), 
marked by small circlets. They are deduced from the Burgos obser- 
vations of 1905*) in the following way. 
In the course of that eclipse the sun shone sometimes for a few 
minutes in a beautifully clear patch of sky between heavy clouds, 
and happened to do so during the phases in which the radiation passed 
through one-half of its maximum value. The exact epochs at which 
. 1794000 
the intensity was EN = 897000 occurred 33738: before second 
- 
= 1) Cf. Astrophysical Journal, 3%, p. 232, 1913. 
*) Astrophysical Journal Vol. 23 p. 312, 1906. 
