like a solid mass in resisting the entry of any more of its own kind, while it would be freely pervious to the vapor of water or that of the other liquids.
A practical example will further illustrate this. Some years ago I was engaged in the distillation of paraffin-oil, and had a few thousand gallons of the crude liquid in a still with a tall head and a rising condenser. In spite of severe firing, the distillation proceeded very slowly. Then I threw into the still, just above the surface of the oil, a jet of steam. The rate of distillation immediately increased with the same firing, although the steam was of much lower temperature than the boiling oil, and therefore wasted much heat. The rationale of this was that at first an atmosphere of oil-vapor stood over the oil, and this was impervious to more oil-vapor, but, on sweeping this out and replacing it by steam, the atmosphere above the liquid oil was permeable by oil-vapor. This principle is largely applied in similar distillations.
But I am exceeding my limits, and must, therefore, defer the direct application of these principles to my next, though doubtless most of my readers w r ill anticipate, or, in vulgar but expressive phrase, "see what I am driving at."
XI.
Always keeping in view that the primary problem in roasting is to raise the temperature throughout to the cooking heat with the smallest possible degree of desiccation of the natural juices of the meat, and applying to this problem the laws of vapor diffusion expounded in my last, it is easy enough to understand the theoretical advantages of roasting in a closed oven, the space within which speedily becomes saturated with those particular vapors that resist further vaporization of these juices.
I say "theoretical," because I despair of practically convincing any thorough-bred Englishman that baked meat is better than roasted meat by any reasoning whatever. If, however, he is sufficiently "un-English" to test the question experimentally, he may possibly convince himself. To do this fairly, a large joint of meat should be equally divided, one half roasted in front of the fire, the other in a non-ventilated oven over a little water by a cook who knows how to heat the latter. This condition is essential, as some intelligence is demanded in regulating the temperature of an oven, while any barbarian can carry out the modern modification of the ordinary device of the savage, who skewers a bit of meat, and holds this near enough to a fire to make it frizzle.
Having settled this question to my own satisfaction more than twenty years ago, I now amuse myself occasionally by experimenting upon others, and continually find that the most uncompromising theoretical haters of baked meat practically prefer it to orthodox roasted meat, provided always that they eat it in ignorance.
