"Experimental investigation, to borrow a phrase employed by Kepler respecting the testing of hypotheses, is "a very great thief of time." Sometimes it costs many days to determine a fact that can be stated in a line."