140 A. E. Bostwick—The Influence of Light 
When plates A and B were illuminated by sunlight as men- ~ 
tioned above, they were connected as in this method. The 
change of resistance corresponding to a given deflection was 
therefore calculated as above, except that as the light was not 
shifted from one plate to the other, the effect must be considered 
as produced by a change in the resistance of one plate instead 
of being the sum of equal changes in two plates, x was thus 
found in this case to equal ‘013 per cent. The per cent of 
change corresponding to any deflection D was of course —, 
S being the sensibility. 
The light was shifted from one plate to the other by moving 
the lens every fifteen minutes. During the last five minutes 
of each exposure, observations of the first elongation of the 
index were made once a minute and the means taken. Care 
‘was taken to read the zero point before taking each elongation. 
The circuit was closed for a few seconds only of the fifteen 
minutes of exposure. Plates A B, A’ and B’ were thus treated. 
With plates C, D and E the method of observation was the 
same, but the bridge was arranged as at first, only one plate ‘ 
being connected at a time, and this plate was alternately lighted 
and darkened at intervals of fifteen minutes. The change in 
the resistance of the plate corresponding to a deflection of one : | 
scale division was calculated as in method I. 
accidental causes was considerable. In this method there was 
observed a steady increase of the plate resistance caused prob 
ably by the heating effect of the current, or it may be by the 
passage of the current itself as claimed by Bornstein. 
he ca NS Sc oi es ay 
Bee Seis tana Tea a oh ad ae i ace 
| She Sh a Pe 
ages) IP 
