In several of the best known manuals of pathology and clinical methods, it is stated that the fermentation test by means of yeast is the only absolutely certain and specific test for glucose.
Since 1904, one of us1 has called attention to the fallacy of this statement on various occasions, and some years ago we2 published a paper, in which we described a new mycologic method for the detection of glucose and other carbon compounds, first theoretically devised by one of us (A. C.) in Ceylon. Our results were confirmed by various authorities. We3 have since carried out some experiments on bakers' yeast (so-called German yeast), which is the yeast generally used in pathologic laboratories for the detection of glucose in the urine, and we can confirm what we stated in that paper. We have examined eight samples of bakers' yeast obtained in various quarters of London,