^{page 166 note 1} Of course, it does not follow that nothing but alkalies would be transferred, though alkaline compounds, mainly silicates, would be far away the principal material. We may well suppose that other matters would also be present, and that among them would be small proportions of fluorine-compounds, which have long been known to exercise such a very powerful influence in synthetic experiments on the formation of minerals, and which may, as has, indeed, been frequently supposed and stated about them, exercise a similar power in starting and maintaining the metamorphic processes in the rocks.

^{page 166 note 2} If this be so, and a rock cannot be metamorphosed simply by heat and the action of its own confined water, we cannot look for any success in experiments on samples of rock heated in platinum, or other sealed metallic containers, to retain the water. This may be the reason why so little is recorded in this direction, and even a portion of what is recorded is not satisfactory evidence. There may have been many failures not recorded at all, though it is almost as valuable, in many cases, to record failures as successes. I will set the example of recording failures by stating that recently I have heated a fire-clay, rich in alkali, in a moist condition, to a low red heat for one month in a wrought-iron bomb. All the water was retained perfectly, but the clay was not affected in any way, except that its combined water was I per cent. less after the experiment, showing that some of its hydrous minerals had been dehydrated by the red heat. This clay would have been very susceptible to metamorphism if it had been invaded by a granite or a diabase. Subsequent experiments at a strong red heat failed entirely, as my bombs gave way under the pressure of the contained water-vapour, the iron getting too soft at the higher temperature.

^{page 169 note 1} Mr. Barrow has pointed out (op. cit.) how greatly intense folding of beds may increase their depth of cover, and so their temperature. He also alludes to some of the evidence that greater depth, and its consequent higher initial temperature, can be shown to have influenced the degree of contact-metamorphism.