But to proceed regularly, it was necessary first to see how much light the sun loses by reflection at different distances, and what are the matters which reflect it the strongest; I first found, that glasses when they are polished with care, reflect the light more powerfully than the best polished metals, and even better than the compounded metal with which telescope mirrors are made; and that although there are two reflectors in the glasses, they yet give a brighter and more clear light than metal. Secondly, by receiving the light of the sun in a dark place, and by comparing it with this light of the sun reflected by a glass, I found, that at small distances, as four or five feet, it only lost about half by reflection, which I judged by letting a second reflected light fall on the first; for the briskness of these two reflected lights appeared to be equal to that of direct light. Thirdly, having received at the distances of 100, 200, and 300 feet, this light reflected by great glasses, I perceived that it did not lose any of its strength by the thickness of the air it had to pass through. I afterwards tried the same experiments on the light of candles; and to assure myself more exactly of the quantity of weakness that reflection causes to this light, I made the following experiments: I seated myself opposite a glass mirror with a book in my hand, in a room where the darkness of the night would not permit me to distinguish a single object. In an adjoining room I had a lighted candle placed at about 40 feet distance; this I approached nearer and nearer, till I could read the book, when the distance was about 24 feet. Afterwards turning the book, I endeavoured to read by the reflected light, having by a parchment intercepted the part of the light which did not fall on the mirror, in order to have only the reflected light on my book. To do so I was obliged to approach the candle nearer, which I did by degrees, till I could read the same characters clearly by the same light, and then the distance from the candle, comprehending that of the book to the mirror, which was only half a foot, I found to be in all 15 feet. I repeated this several times, and had always nearly the same results; from whence I concluded, that the strength, or quantity, of direct light is to that of reflected light, as 576 to 225; therefore, the light of five candles reflected by a flat glass, is nearly equal to that of the direct light of two. The light of a candle, therefore, loses more by reflection than by the light of the sun; and this difference proceeds from the rays of the former falling more obliquely on the mirror than the rays of the sun, which come almost parallel. This experiment confirmed what I had at first found, and I hold it certain, that the light of the sun loses only half by its reflection on a glass mirror.