Author: Frank A. J. L. James

Title: How Comte was proved wrong: The development of spectro-chemistry and its early application to the study of the stars

Abstract: In the second volume of his Cours de philosophie positive published in 1835, Auguste Comte asserted, as an example of something that could never be known, that it would be impossible ever to determine the chemical composition or mineralogical structure of the stars. This did not, however, prevent work on the subject being undertaken. These related especially to the problem of how the sun produced its heat and light which, following the establishment of the first two laws of thermodynamics, became a pressing problem in the 1850s. In turn these theories of solar heat production made predictions about the chemical composition of the sun. However, it was not until Robert Bunsen and Gustav Kirchhoff working in Heidelberg in 1859-60 showed that a chemical element emitted a uniquely characteristic spectrum that firm experimental knowledge could be gained about the chemical composition of the sun and stars. With knowledge of the already existing solar theories and their predictions, especially that of Hermann Helmholtz, who was also at Heidelberg at the time, Kirchhoff, in particular, was able to use the new knowledge of spectra to provide evidence for the chemical composition and structure of the sun, and thus of other stars. This story illustrates the crucial role of apparatus in the development of new knowledge across traditional disciplinary boundaries; it also illustrates how the German university system facilitated the development of such a new discipline.