Since midcentury the use of laboratory experimental methods in economics has developed into a major field of inquiry within microeconomics. The slow but steady development of the 1950s and 1960s was superseded by accelerated development in the 1970s and 1980s. Development, not growth, is the right word because the methodological purpose and function of experiments in economics has undergone in-depth inner change as well as quantitative growth. The idea that an experiment might be described as a “simulation” – a word used in my first (1962) paper before that word had become clearly associated with a different meaning – has yielded to the realization that in an experiment we create a certain type of controlled market or nonmarket allocation process that is real in the sense of rewards, people, and institutional rules of exchange in which all trades are binding. The issue of parallelism, or the transferability of results from laboratory to other environments, which is of ever-present interest to experimentalists, is most constructively viewed as an empirical question applying to any particular data set whether in the laboratory or in the field. Thus data from one field environment may or may not have relevance to another field environment. All data are specific to particular conditions and there is no means by which one can bootstrap finite data sets into a theory or generalization of any kind without falling prey to the fallacy of induction.