Robert Boyle on the importance of reporting and replicating experiments

Boyle’s writings, although full of archaic language, repetition and digressions, give us detailed accounts so that we can understand his approach to experimentation. Many aspects of modern scientific work are reflections of these accounts. Boyle is very aware of contingent influences in a system, and he tries to control variables in experimental design and address them in discussion. He makes careful measurements but notes that obsessional precision is not always required. Reports of his experiments are accompanied by possible explanations for the findings, but with competing explanations and influences discussed. Boyle’s research approach was to set out heads of enquiry after ‘a general survey of the subject’ (p. xvxiv). He refers to these as ‘primary titles’. Then, ‘by reading, conference, meditation, and the experiments suggested by the heads of enquiry of the first class’, he proceeds to form a set of second titles. Sometimes a third set will be needed. This process forms the beginning of a ‘natural history’ of the subject. Boyle is far-seeing about the progressive and provisional nature of science: