"by means of the transference of the polar ice from "one hemisphere to the other."

I wish you to notice particularly the last sentence of that note, for it is to that my attention has been called by your secretary. I was quite unaware that by this statement I had placed myself under any obligation, but it so turns out. A friend suggests that the quotation marks as supplied to that last sentence be here deleted, and the passage given as Mr. Croll's own. Mr. Croll may well have adopted it, for after quoting my note he at once proceeds to carry out the intention which I expressed in the concluding sentence. For myself I can only say that I am now trying to fulfil my obligation, though I feel I can throw but little additional light on the subject. Nor do I need to do so it is quite unnecessary for me to carry out the intention, since Mr. Croll has done it with all the means that occurred to him as bearing on former estimates, in regard to this very important and difficult subject.

With regard to the effect on sea-level Mr. Croll's principle, as set forth in pp. 368-369 of his book, is

་

thoroughly correct, and shows the remarkable power he possessed of grasping the subject, and dealing with it by a simple geometric construction which led to the same result as Laplace's mathematical analysis. For the stability of the ocean it is necessary that the specific gravity of water be less than the specific gravity of the solid, and it is less as we know. The mean density of the earth is about 5 or more exactly 5.6-times the specific gravity of water. This statement favours Laplace's theory as to the requisite for stability, but it is curious that Laplace did not notice the simple view that if the solid part of the earth had a specific gravity less than that of water it would tend to float and leave the water on either side of it. For example, take a globe of liquid of any size (Fig. 1) and let us suppose a small spherical portion at A to become solid. configuration of solid and liquid would remain stable provided the solidified portion remained of the same density as the liquid. If the solidified portion acquired a density greater or less than that of the liquid the configuration as shown in the

The

figure would become unstable, and the small sphere A would take the position A' or A", in either of

A

FIG. 1.

which two cases the configuration would be again stable.

Look now at Fig. 2 and suppose the outer circle to represent the section of a globe 8,000 miles in diameter, and let the inner circle represent the section of an enclosed globe 2 miles less in diameter. Suppose the outer envelope to be water, and the inner globe to be solid matter, of density equal to or greater than that of water, then the configuration of Fig. 2 would evidently be stable. On the other hand, if we suppose the

inner globe to be solid but of density less than that of water, say equal to that of cork or wood, then the water would no longer form an envelope enclosing the solid, but would run together, and

the configuration of solid and liquid would be represented by a section such as Fig. 3.

Mr. Croll goes on to inquire what would be the probable effect, upon the level of the ocean, of changing the centre of gravity of the earth, if we suppose a quantity of ice (a polar ice-cap) to be transferred to one pole from the other. I shall just say one word as to the attraction of the polar icecap.

Suppose that

We have here (Fig. 4) a globe. somehow or other a portion of ice was placed on the Antarctic continent, what would be the result? It would be that the polar ice-cap would attract the water so that the water which stood at a certain height before that transference was

made, would be drawn up to a higher level all round the Antarctic continent by the attraction of this mass of ice. The calculation for the result is merely a piece of mathematical book-keeping with which I need not trouble you.

There is just one other point which belongs to further explanation of Laplace's theory. That first