^{page 220 note 1} Even now that the general presence of titanic acid in rocks is more fully known than formerly, it is very rarely determined by analysts. This arises to some extent from the fact that its non-determination does not affect the total addition of the analysis, as in the ordinary course of the estimations it is weighed partly with the silica and partly with some of the bases. But the non-determination is more largely due to the fact that the exact quantitative serparation of titanic acid is a very tedious and difficult operation.

^{page 220 note 2} In the “American Journal of Science” for 12, 1891, is an interesting paper by Mr. F. P. Dunnington, “On the Distribution of Titanic Oxide upon the Surface of the Earth,” in which he gives the results of 72 determinations on soils from various parts of the world, as well as on a few rocks. The universal diffusion of this oxide in appreciable quantity is fully demonstrated. These many determinations were not made by Mr. Dunnington in the ordinary tedious gravimetric mauner, but by a rapid calorimetric process which is described in the paper. It is to be supposed that this method has been fully cheeked and proved to be reliable. That being so, the determination of titanic acid in our rocks and minerals ought to become the rule instead of the exception, since an easy and rapid method ia available for it.Google Scholar

^{page 221 note 1} It is also to be remarked that the small crystals of tourmaline, so usual in clays, shales, and slates, are never seen in these ashes.

^{page 221 note 2} Since the above was in print, I have had the opportunity of examining some of these minute auatase crystals by means of a inch oil-immersion objective. This has enabled me to see them very much better than I ever did before, and to ascertain that 1 was mistaken in stating that the prism-band is present on some of them. The apparent bund disappears when the crystals are seen under the great magnification, together with fine definition, which the use of this objective gives us. 23).

^{page 222 note 1} It is usually considered to be difficult, and indeed often impossible, to distinguish microscopically between kaoline and muscovite as alteration-products in decomposing felspars. Kosenbusch, for instance, points this out very emphatically (Physiographie der Mineralien, pp. 516, 561). The flaky alteration-product in felspars is certainly sometimes so minute, and occurs in such an indistinct manner, that decision as to its exact character is very uncertain or impracticable. But as soon as the flakes are larger and more distinct (and this is often eminently the case in the rocks now under consideration), I venture to think that the identification of the mica can be safely made. The bi-refraction of kaolinite is high, but so far as my observations go it is very distinctly less than that of muscovite, observing transverse sections in each case. Then again, edge-sections of the mica extinguish parallel, while edge-sections of kaolinite give such very decided angles that no confusion of the two minerals appears possible. 1 have used for these comparisons the well-known kaolinite crystals from Anglesey, and also those which occur abundantly (though of smaller size) in the interstices of many coarse sandstones and grits of the Coal-measures.