"Progressive medicine recognizes that among the most helpful aids to the advancement of the work of our profession are the critical analysis of errors, either in diagnosis or treatment the frank discussion of mistakes, and the unremitting investigation of the reasons why procedures in current practice fail."1 The following confession and analysis of errors is made in this spirit, and further to add its iota of emphasis to the importance of the correlation of clinical and pathologic data. An important purpose of the necropsy is surely lost if it is not observed at least indirectly by the clinician that he may learn of his mistakes or strengthen his convictions.
In the 198 necropsies that have been performed at the Boston Consumptives' Hospital since its foundation, twenty-three cases, or 11.5 per cent., have proved to be non-tuberculous in so far, at least, that there was no active lesion. The report