1454 
Observations made with the bolometer. 
During the greater part of the eclipse the galvanometer deflections 
were only visually observed, by noticing the motion of the reflected 
image of a slit on a transparent scale; but from 5 minutes before 
until 5 minutes after centrality the image was received on a photo- 
graphic recording drum. 
For a reproduction of the photogram we must refer to. the 
Astrophysical Journal 87, p. 229, Plate X, Fig. 1. On the same 
plate, Fig. 2 shows the central part of the curve on a larger scale’), 
and Fig. 3 gives on the same scale a control of the volt-sensitivity 
of the galvanometer, effected immediately after the eclipse was over. 
It shows well the qualities of the instrument. 
The vertical lines are time-signals, produced by a small electric 
lamp flashing up at intervals of ten seconds in front of the slit of 
the recording apparatus; the first line following the minimum of the 
curve corresponds to O'34"57s Leiden M. T. 
Two of the zero-readings, obtained by screening the bolometer, 
are visible on the curve (Fig. 1), one at 0'30™, another at 0637™, 
A straight line joining them may quite safely be taken to represent 
the zero during the interval. The ordinate of the minimum thus 
comes out to be a quarter of a millimeter. At 11°30™ (6 minutes 
after first contact) a deflection of 6.1 mm.*) was observed visually, 
the intensity of the main current at that time being '/,,, of its value 
at the time of recording. Reduced to the latter value of the main 
current, the deflection corresponding to full sunshine would have 
been more than 195 x 6.1 = 1190 mm., or nearly 5000 times the 
deflection at minimum. 
A few irregularities in the curve, especially at 0°31™20° and at 
0"36™40s, require explanation. They are not genuine, but simply 
due to an excusable negligence of the assistant who had to point 
the bolometer at the sun. The emotions of the event making him 
forget to keep the tube continuously in the right direction, he had 
twice suddenly to make up for the loss. Fortunately the minimum 
is unaffected. 
Discussion of the bolometer results. 
If the apparatus had followed the radiation instantaneously, the 
minimum would have been lower yet. We may therefore certainly 
1) The striped aspect of the curve is connected with the click of the recording 
apparatus. 
2) As a basis for calculation we purposely select this small deflection, because 
the great deflections of the provisory galvanometer were not strictly proportional 
to the current. 
