Developing into series and returning from series: A note on the foundations of eighteenth-century analysis