Naturally, Murdock had many difficulties to overcome in these early days, but he possessed skill and perseverance. His first retorts for distilling coal were similar to the common glass retort of the chemist. Next he tried cast-iron cylinders placed perpendicularly in a common furnace, and in each were put about fifteen pounds of coal. In 1804 he constructed them with doors at each end, for feeding coal and extracting coke respectively, but these were found inconvenient. In his first lighting installation in the factory of Phillips and Lee in 1805 he used a large retort of the form of a bucket with a cover on it. Inside he installed a loose cage of grating to hold the coal. When carbonization was complete the coke could be removed as a whole by extracting this cage. This retort had a capacity of fifteen hundred pounds of coal. He labored with mechanical details, varied the size and shape of the retorts, and experimented with different temperatures, with the result that he laid a solid foundation for coal-gas lighting. For his achievements he is entitled to an honorable place among the torch-bearers of civilization. The epochal feature of the development of gas-lighting is that here was a possibility for the first time of providing lighting as a public utility. In the early years of the nineteenth century the foundation was laid for the great public-utility organizations of the present time. Furthermore, gas-lighting was an improvement over candles and oil-lamps from the standpoints of convenience, safety, and cost. The latter points are emphasized by Murdock in his paper presented before the Royal Society in 1808, in which he describes the first industrial installation of gas-lighting. He used two types of burners, the Argand and the cockspur. The former resembled the Argand lamp in some respects and the latter was a three-flame burner suggesting a fleur-de-lis. In this installation there were 271 Argand burners and 636 cockspurs. Each of the former "gave a light equal to that of four candles; and each of the latter, a light equal to two and a quarter of the same candles; making therefore the total of the gas light a little more than 2500 candles." The candle to which he refers was a mold candle "of six in the pound" and its light was considered a standard of luminous intensity when it was consuming tallow at the rate of 0.4 oz. (175 grains) per hour. Thus the candle became very early a standard light-source and has persisted as such (with certain variations in the specifications) until the present time. However, during recent years other standard light-sources have been devised.