Here it is that I regret most of all that I cannot print the working-drawings with this paper. If you will cut open the seed-vessel of Spergularia Rubra, or any other carpel that has a free central placenta, and observe how the circular seeds cling around the circular centre, you will have some idea of the arrangement of a transverse horizontal section of the completed MOON. Lay three croquet-balls on the piazza, and call one or two of the children to help you poise seven in one plane above the three; then let another child place three more above the seven, and you have the CORE of the MOON completely. If you want a more poetical illustration, it was what Mr. Wordsworth calls a mass "Of conglobated bubbles undissolved." Any section through any diameter looked like an immense rose-window, of six circles grouped round a seventh. In truth, each of these sections would reveal the existence of seven chambers in the moon,--each a sphere itself,--whose arches gave solidity to the whole; while yet, of the whole moon, the greater part was air. In all there were thirteen of these moonlets, if I am so to call them; though no one section, of course, would reveal so many. Sustained on each side by their groined arches, the surface of the whole moon was built over them and under them,--simply two domes connected at the bases. The chambers themselves were made lighter by leaving large, round windows or open circles in the parts of their vaults farthest from their points of contact, so that each of them looked not unlike the outer sphere of a Japanese ivory nest of concentric balls. You see the object was to make a moon, which, when left to its own gravity, should be fitly supported or braced within. Dear George was sure that, by this constant repetition of arches, we should with the least weight unite the greatest strength. I believe it still, and experience has proved that there is strength enough. When I went up to No. 9, on my return from South America, I found the lower centring up, and half full of the working-bees,--who were really Keltic laborers,--all busy in bringing up the lower half-dome of the shell. This lower centring was of wood, in form exactly like a Roman amphitheatre if the seats of it be circular; on this the lower or inverted brick dome was laid. The whole fabric was on one of the terraces which were heaved up in some old geological cataclysm, when some lake gave way, and the Carrotook River was born. The level was higher than that of the top of the fly-wheels, which, with an awful velocity now, were circling in their wild career in the ravine below. Three of the lowest moonlets, as I have called them,--separate croquet-balls, if you take my other illustration,--had been completed; their centrings had been taken to pieces and drawn out through the holes, and were now set up again with other new centrings for the second story of cells.