Lettie S. Multhauf, "The Light of Lamp-Lanterns: Street Lighting In 17th-Century Amsterdam" Technology and Culture 26 no. 2 (1985): 236-252.
"Van der Heyden's first oil lamps had apparently been quite unsatisfactory. Great amounts of oil had spilled, drenching the lanterns and posts. The lanterns had been so greasy that they had been difficult to handle and the lamplighters had raged against the manufactureres of the lamps, who, not realizing the cause of the spilling, had vainly searched for leaks. But van der Heyden had come to realize that the oil was veing pushed out by air, which was enclosed in a small space and expanded when warmed by the flame or by sunlight: "The air, as can be seen in weather glasses, is such that it shrinks by the cold and expandes by the heat." To solve this problem, he experimented with a glass lamp in which he could observe the displacement of oil by the expanding air. Figure 2 shows both the unsatisfactory initial model and the improved final design, whereas figure 3 shows mode of constructing the entire lantern.
"In his treatise, van der Heyden devoted considerable attention to the placement of the streetlights. At the begining of the installation, some large lanterns, each holding four lamps, had apparently been placed 300 feet apart. But he showed mathematically that less oil was consumed and the light was brighter when smaller lanterns with only one lamp were placed closer together. After a detailed calculation taking into account the price of oil, wages, and the cost of lanterns, repair, and interest, he concluded that the highest efficency was obtained with single lanterns spaced at a distance of 125 to 150 feet. He suggested that in the poorer districts the lanterns could be somewhat further apart, since "poor people usually have less light inside, so that their eyes are used to to it."