The difference between new and stale bread is familiar enough, but the nature of the difference is by no means so commonly understood. It is generally supposed to be a simple result of mere drying. That this is not a true explanation may be easily proved by repeating the experiments of Boussingault, who placed a very stale loaf (six days old) in an oven for an hour, during which time it was, of course, being further dried; but, nevertheless, it came out as a new loaf. He found that during the six days, while becoming stale, it only lost 1 per cent. of its weight by drying, and that during the one hour in the oven it lost 3½ per cent. in becoming new, and apparently more moist. By using an air-tight case instead of an ordinary oven, he repeated the experiment several times in succession on the same piece of bread, making it alternately stale and new each time. For this experiment the oven should be but moderately heated—260° to 300° Fahr. is sufficient. I am fond of hot rolls for breakfast, and frequently have them _à la Boussingault_, by treating stale bread-crusts in this manner. My wife tells me that when the crusts have been long neglected, and are thin, the Boussingault hot rolls are improved by dipping the crust in water before putting it into the oven. This is not necessary in experimenting with a whole loaf or a thick piece of stale bread. The crumb of bread, whether new or stale, contains about 45 per cent. of water. Miller says ‘the difference in properties between the two depends simply upon difference in molecular arrangement.’ This ‘molecular arrangement’ is the customary modern method of explaining a multitude of similar physical and chemical problems, or, as I would rather say, of evading explanation under the cover of a vague conventional phrase. I have made some simple experiments which supply a visible explanation of the facts without invoking the aid of any invisible atoms or molecules, or any imaginary arrangements or rearrangements of these imaginary entities. I find that, as bread becomes stale, its porosity _appears_ to increase, and that when renewed by reheating, it returns to its original _apparently_ smaller degree of porosity. That this change can be only apparent is evident from the facts that the total quantity of solid material in the loaf remains the same, and its total dimensions are retained more or less completely by the rigidity of the crust. I say ‘more or less,’ because this depends upon the thickness and hardness of the crust, and also upon the completeness of its surrounding. Lightly-baked loaves shrink a little in dimensions in becoming stale, and partly regain the loss on reheating, but this difference only exaggerates the apparent paradox of varying porosity, as the diminished bulk of a given quantity of material displays increased porosity, and the increase of total dimensions accompanies the diminished porosity.