Some Thoughts on Constant and Variable Components of Systematic Error

Systematic error has been struggling for recognition for almost fifty years. One of the first papers on the topic is due to Student [1] who in 1927 proposed treating some measurement error as semiconstant error. His suggestion appears to have been poorly received and we have no knowledge of Student's later thinking along these lines. We do know Student was ahead of his time and that he wrote during a period when Vigneron [2] ?nd others were still questioning the applicability of the theory of probability and statistics to laboratory measurements. It was thought by some that applying probability theory tc laboratory results would be wrong and unjustified.