On raising the gaseous atom to E 4, the falling away of the wall sets free the six contained bodies, and both the balloon and the egg round themselves, apparently with the removal of their propinquity, as though they had exercised over each other some attractive influence.  The smaller bodies within the egg—­marked q on E 4—­are not on one plane, and those within n and o form respectively square-based and triangular-based pyramids.  On raising all these bodies to E 3 we find the walls fall away as usual, and the contents of each “cell” are set free:  p of E 4 contains six small bodies marked k, and these are shown in k of E 3, as containing each seven little bodies—­marked e—­each of which has within it two ultimate atoms; the long form of p E 4—­marked l—­appears as the long form l on E 3, and this has three pairs of smaller bodies within it, f’, g and h, containing respectively three, four and six ultimate atoms; q of E 4, with its seven contained particles, m, has three particles m on E 3, each showing three ultimate atoms within them; e from n of E 4 becomes i of E 3, with contained bodies, e, showing two ultimate atoms in each; while e’ from o of E 4 becomes j of E 3, each having three smaller bodies within it, e’, with two ultimate atoms in each.  On E 2, the arrangement of these ultimate atoms is shown, and the pairs, f’, g and h are seen with the lines of force indicated; the triads in f—­from m of E 3—­are similarly shown, and the duads in e and e’—­from i and j of E 3—­are given in the same way.  When all these bodies are raised to E 1, the ultimate physical atoms are set free, identical, of course, with that previously described.  Reckoning up the number of ultimate physical atoms in a chemical atom of nitrogen we find they amount to 261, thus divided:—­

62 + bodies with 2 ultimate atoms, 62 x 2 = 124
24 — " " 2 " " 24 x 2 = 48
21 — " " 3 " " 21 x 3 = 63
2 + " " 3 " " 2 x 3 = 6
2 + " " 4 " " 2 x 4 = 8
2 + " " 4 " " 2 x 6 = 12
——­
261
This again approaches closely the weight-number assigned to nitrogen:—­

261 / 18 =14.44 +

This is interesting as checking the observations, for weight-numbers are arrived at in so very different a fashion, and especially in the case of nitrogen the approximation is noteworthy, from the complexity of the bodies which yield the number on analysis.