light,
but to have expended their forces in modifying the positions which the
molecules must normally assume in darkness.

_Practical applications_.--There are many peculiarities of action
occasionally found, and the causes of such actions are not always
discernible. In practice, I have been accustomed to find the
peculiarities and weaknesses of each cell by trial, developing its
strongest properties and avoiding its weaknesses, until, when the cell is
finished, it has a definite and known character, and is fitted for
certain uses and a certain line of treatment, which should not be
departed from, as it will be at the risk of temporarily disabling it. In
consequence of the time and labor expended in making cells, in the small
way, testing, repairing damages done during experiments, etc., the cost
of the cells now is unavoidably rather high. But if made in a commercial
way, all this would be reduced to a system, and the cost would be small.
I may say here that I do not make cells for sale.

The applications or uses for these cells are almost innumerable,
embracing every branch of electrical science, especially telegraphy,
telephony, and electric lighting, but I refrain from naming them. I may
be permitted, however, to lay before you two applications, because they
are of such general scientific interest. The first is my

_Photometer_.--The light to be measured is caused to shine upon a
photo-electric current-generating cell, and the current thus produced
flows through a galvano-metric coil in circuit, whose index indicates
upon its scale the intensity of the light. The scale may be calibrated by
means of standard candles, and the deflections of the index will then
give absolute readings showing the candle power of the light being
tested. Or, the current produced by that light and that produced by the
standard candle may be compared, according to any of the known ways of
arranging and comparing different lights--the cell being lastly exposed
alternately to the two lights, to see if the index gives exactly the same
deflection with each light.

This arrangement leaves untouched the old difficulty in photometry, that
arising from the different _colors_ of different lights. I propose to
obviate that difficulty in the following manner. As is well known, gold
transmits the green rays, silver the blue rays, and so on; therefore, a
cell faced with gold will be acted upon by the green rays, one faced with
silver by the blue ray