I have obtained a reconciliation of this paradox by careful examination of the structure of the crumb. This shows that the larger or decidedly visible pores are cells having walls of somewhat silky appearance. The silky lustre and structure is, I have no doubt, due to a varnish of dextrin, the gummy nature of which I have already described. On looking a little more closely at this inner surface of the big blow-holes with the aid of a hand-lens of moderate power, I find that it is not a continuous varnish of gum, but a net-work or agglomeration of gummy fibres and particles, barely touching each other. My theory of the change that takes place as the bread becomes stale is, that these fibres and particles gradually approach each other either by shrinkage or adhesive attraction, and thus consolidate and harden the walls of each of the millions of easily visible pores, these walls forming the solid material of which the loaf is made up. In doing so they naturally increase the dimensions of the visible pores, while the microscopic interstices or spaces between the minute fibres of the cell walls are diminished by the approximation or adhesion of the fibres to each other. This adhesion is probably aided by an oozing out or efflorescence of the vapour held by the fibres, and its condensation on their surfaces. This point, be it understood, is merely hypothetical, as the efflorescence is not visible. All the other phenomena I have just described are visible either with the naked eye or by the aid of a lens. When the stale bread is again heated, a general expansion occurs by the conversion of liquid water into aqueous vapour, every grain of water thus converted expanding to 1,700 times its former bulk. As this happens throughout, _i.e._ upon the surface of every one of the countless fibres or particles, there must be a general elbowing in the crowd, breaking up the recent adhesion between these fibres and thrusting them all apart in the directions of least resistance; _i.e._ towards the open spaces of the larger and visible pores, producing that _apparent_ diminution of porosity that I have observed as the easily visible characteristic of the change.