little over a meter in length, closed at the top by a board, in which is a circular aperture 15’8 cm. in diameter.  Over this board may be placed a cover, in the center of which is a hole 0.05 cm. in diameter, which therefore lets through 0.00001 as much light as the full aperture.  The silver chloride is placed a distance of just one meter from the larger aperture, and over it is placed the photographic scale, which might be made of tinted gelatines, or, as in the present case, constructed of long strips of tissue paper, of varying widths, and arranged like a flight of steps; so that the light passing through one side of the scale traverses nine strips of paper, while that through the other side traverses only one strip.  Each strip cuts off about one-sixth of the light passing through it, so that, taking the middle strip as unity, the strips on either side taken in order will transmit approximately—­

1 2 3 4 5 6 7 8 9 2.0 1.65 1.4 1.2 1.0 0.85 0.7 0.6 0.5

The instrument is now pointed toward the zenith for about eight minutes, on a day when there is a bright blue sky.  On taking the apparatus into the dark room and viewing the impression by gaslight, it will be found that the markings, which are quite clear at one end, have entirely faded out by the time the middle division is reached.  The last division clearly marked is noted.  Five strips cut from sensitized glass plates, ten centimeters long and two and a half in width, are now placed side by side under the scale, in the place of the chloride.  By this means we can test, if we wish, five different kinds of plates at once.  The cover of the sensitometer containing the 0.05cm. hole is put on, and the plates exposed to sky light for a time varying anywhere between twenty seconds and three minutes, depending on the sensitiveness of the plates.  The instrument is then removed to the dark room, and the plates developed by immersing them all at once in a solution consisting of four parts potassium oxalate and one part ferrous sulphate.  After ten minutes they are removed, fixed, and dried.  Their readings are then noted, and compared with those obtained with the silver chloride.  The chloride experiment is again performed as soon as the plates have been removed, and the first result confirmed.  With some plates it is necessary to make two or three trials before the right exposure can be found; but if the image disappears anywhere between the second and eighth divisions, a satisfactory result may be obtained.

The plates were also tested using gaslight instead of daylight.  In this case an Argand burner was employed burning five cubic feet of gas per hour.  A diaphragm 1 cm. in diameter was placed close to the glass chimney, and the chloride was placed at 10 cm. distance, and exposed to the light coming from the brightest part of the flame, for ten hours.  This produced an impression as far as the third division of the scale.  The plates were exposed in the sensitometer as usual, except that it was found convenient in several cases to use a larger stop, measuring 0.316 cm. in diameter.