In carrying out a series of photometrical experiments lately, I found that it was a matter of considerable difficulty to keep the flames of the standard candles always at their proper distance from the light to be measured, because the wick was continually changing its position (of course carrying the flame with it), and thus practically lengthening or shortening the scale of the photometer, according as the flame was carried nearer to or farther from the light at the other end of the scale. In order, therefore, to obtain a correct idea of the extent to which this variation of the position of the wick might influence the readings of the photometer scale, I took a continuous number of photographs of the flame of a candle while it was burning in a room quite free from draught; no other person being in it during the experiment except a photographer, who placed sensitive dry plates in a firmly fixed camera, and changed them after an exposure of 30 seconds. In doing this he was careful to keep close to the camera, and disturb the air of the room as little as possible. In front of the candle a plumb-line was suspended, and remained immovable over its center during the whole operation.
The candle was allowed to get itself into a normal state of burning, and then the wick was aligned, as shown in the photographs Nos. 1 and 2, after which it was left to itself.
VARIATION IN PHOTOMETRICAL STANDARDS.
With these photographs (represented in the cuts) I beg to hand you full-sized drawings of the scales of a 100 inch Evans and a 60 inch Letheby photometer, in order to give your readers an opportunity of estimating for themselves the effect which such variations from the true distance between the standard light and that to be measured, as shown in this series of photographs, must exercise on photometrical observations made by the aid of either of the instruments named.