These experiments, Mr. Walker informs us, were made in the following manner:-- Three candles, the dimensions of which are given in the table, against 1, 3, and mould. These were first weighed, and then lighted at the same instant. At the end of the time inserted in the third column of the above table, they were extinguished and weighed again, and the loss of weight of each candle is contained in the fourth column. The three first experiments were made under such favourable circumstance, that there was little doubt of their results being more accurate than what practical utility requires, but the fourth experiment cannot be depended on so much, in consequence of the variable light of No. 5. This candle was moved so often to keep the two shadows equal, that it was found necessary to set down its mean distance from the wall by estimation; but as this was done before the candles were weighed, the experimenter’s mind could not be under the influence of partiality for a system. The method which Mr. Walker employed in comparing one light with another in each experiment, was that which has been described page 24. 1. The experiments were made at different times, and the light of the mould candle was made the standard, with which the lights of the others were compared; but it must not be understood, that this candle gave the same strength of light in every experiment. 2. The sign + in the 5th column, signifies that the candle against which it is placed, gave a stronger light than the others. From the experiments contained in the table, it appears to be an established law, where combustion is complete, that the quantities of light produced by tallow candles, are in the complicate ratio of their times of burning and weights of matter consumed. For if their quantities of matter be equal, and times of burning the same, they will give equal quantities of light, _by the experiments_. And if the times of burning be equal, the quantities of light will be directly as their weights of matter expended. Therefore the light is universally in the compound ratio of the time of burning and weight of matter consumed. If the law which Mr. Walker has endeavoured to prove, both by reason and experiment, be admitted, we have a standard with which we may compare the strength of any other light. Let a small mould candle, when lighted, be so placed as neither to produce smoke nor require snuffing, and it will lose an ounce of its weight in three hours. Let this quantity of light produced under these circumstances, be represented by 1.00. Then should this candle at any other time, lose more or less of its weight in three hours than an ounce, the quantity of light will be still known, because the quantity of light in a given time is directly as the weight of the candle consumed.[6]