GERMANIUM (PLATE XVI., 2, 4).

The central globe, with its two cigar '-bearing tetrahedra, need not delay us; the tetrahedra are set free and follow the occultum disintegration, and the central four atoms is the sodium cross that we had in titanium. The ovoids (XVI., 4) are liberated on the proto level, and the cigar,' as usual, bursts its way through and goes along its accustomed path. The others remain linked on the meta level, and break up into two triangles and a quintet on the hyper.

TIN (Plate XVI., 3, 4).

Here we have only the spike to consider, as the funnels are the same as in germanium, and the central globe is that of titanium, omitting the eight-atomed centre. The cone of the spike we have had in silver (see p. 729, May), and it is set free on the proto level. The spike, as in zinc, becomes a large sphere, with the single septet in the centre, the remaining six bodies circling round it on differing planes. They break up as shown. (Tin is Sn.)

IRON (PLATE IV., 1., and XVII., 3).

We have already dealt with the affinities of this peculiar group, and we shall see, in the disintegration, even more clearly, the close relationships which exist according to the classification which we here follow.

The fourteen bars of iron break asunder on the proto level, and each sets free its contents-a cone and three ovoids, which as usual, become spheres. The twenty-eight-atomed cone becomes a four-sided figure, and the ovoids show crystalline contents. They break up, on the meta level as shown in the diagram, and are all reduced to triplets and duads on the hyper level.

COBALT (PLATE XVII., 4).

The ovoids in cobalt are identical with those of iron; the higher ovoids, which replace the cone of iron, show persistently the crystalline forms so noticeable throughout this group.

NICKEL (PLATE XVII., 5).

The two additional atoms in a bar, which alone separate nickel from cobalt, are seen in the upper sphere of the central ovoid,

RUTHENIUM (PLATE XVIII., 1).

The lower ovoids in ruthenium are identical in composition with those of iron, cobalt and nickel and may be studied under iron. The upper ones only differ by the addition of a triplet.

Rhodium has a septet, which is to be seen in the c of titaninum (see k in the titanium diagram above) and differs only in this from its group.

PALLADIUM (PLATE XVIII., 3).

In palladium this septet appears as the upper sphere in every ovoid of the upper ring.

OSMIUM (PLATE XVIII., 4).

We have here no new constituents; the ovoids are set free on the proto level and the contained globes on the meta, all being of familiar forms. The cigars, as usual, break free on the proto level and leave their ovoid with only four contained spheres, which unite into two nine-atomed bodies as in silicon, (see above).

IRIDIUM (PLATE XVIII., 5).

The twenty-one-atomed cone of silver here re-appears, and its proceedings may be followed under that metal (see diagram, p. 729, May). The remaining bodies call for no remark.

PLATINUM (PLATE XVIII., 6).

Again the silver cone is with us. The remaining bodies are set free on the proto level, and their contained spheres on the meta.

LITHIUM (PLATE IV., 2, and XIX., 1).

Here we have some new combinations, which recur persistently in its allies. The bodies a, in Plate XIX., 1, are at the top and bottom of the ellipse; they come to right and left of it in the proto state, and each makes a twelve-atomed body on the meta level.

The five bodies within the ellipse, three monads and two sextets, show two which we have had before: d, which behaves like the quintet and quartet in silicon, after their junction, and b, which we have had in iron. The two bodies c are a variant of the squarebased pyramid, one atom at the apex, and two at each of the other angles. The globe, e, is a new form, the four tetrahedra of the proto level making a single twelve-atomed one on the meta. The body a splits up into triplets on the hyper; b and d follow their iron and silicon models; d yields four duads and a unit; e breaks into four quartets.

POTASSIUM (PLATE XIX., 2).

Potassium repeats the lithium spike; the central globe shows the 'nitrogen balloon,' which we already know, and which is surrounded

on the proto level with six tetrahedra, which are set free on the meta level and behave as in cobalt. Hence we have nothing new.

RUBIDIUM (PLATE XIX., 3).

Again the lithium spike, modified slightly by the introduction of an ovoid, in place of the top sphere; the forms here are somewhat unusual, and the triangles of the sextet revolve round each other on the meta level; all the triads break up on the hyper level into duads and units.

FLUORINE (PLATE IV., 3, AND PLATE XVII.,

1).

The reversed funnels of fluorine split asunder on the proto level, and are set free, the balloons' also floating off independently, The funnels, as usual, become spheres, and on the meta level set free their contained bodies, three quartets and a triplet from each of the eight. The balloons disintegrate in the usual way.

MANGANESE (PLATE XVII., 2).

Manganese offers us nothing new, being composed of 'lithium spikes and nitrogen balloons.'

(To be continued.)

ANNIE BESANT.

All life is exile from an unguess'd home,
Worlds half-remembered, where perhaps we trod
(For memories haunt us of a calm, grand peace)
When we were flowing in the veins of God.
There is a rhythm in the windy trees,

A gleam in th' apple-bloom,

A measured music in the hidden brook,

That seems the counterpart of something gone :
Whole lives forgotten live in a bell's tone,

A sudden landscape, or a sister's look.

ROBERT W. Cruttwell.

(From the Newdigate Prize Poem, 1907.)

SOME OCCULT INDICATIONS IN ANCIENT ASTRONOMY.

(Continued from p. 1136.)

HEN so great a discordance as 5.6 seconds is found between

WH

two admittedly great authorities in recent times, and that upon the one element of all others which is supposed to be determined with the greatest exactitude, what possible estimate are we to make as to the value assignable to the periods of, say, Jupiter and Saturn, where the matter is complicated not only by the error of the solar year used in reducing the observations, but also by the "great inequality" and other equations of their mean motions, about the exact values and periods of which astronomers are divided? We may gain some ideas on this point from an examination of the elements of the planet Uranus, as respectively published by Professors Newcomb and Leverrier about thirty years ago, within a year of each other; in which we note that the common epochal value of the mean longitude in the ecliptic differs by five minutes of arc, and in the periodic time they disagree to the extent of some fourteen hours.* So that neither of these two astronomers (assigning equal weight to their respective determinations) could be certain as to the exact position of the planet for any given date, within ten days. And this as the result of 95 years of consecutive observations many hundreds in number; but in the case of a still more difficult planet (Neptune) under observation for a much shorter time, we are told that its elements are known "with a high degree of precision "†.

And yet, in face of such grave discrepancies as these, we are gravely told, and in the name of official science are doubtless expected to believe that:

"The astronomical tables have been carried to such an astonishing degree of accuracy, that it has been said, by the highest authority, that an astronomer could now predict, for a thousand years to come, the precise moment of the passage of any one of the stars over the meridian wire of the telescope of his transit-instrument with such a degree of accuracy that the error would not be so

* Ball's Elements of Astronomy, 413.

† Mechanism of the Heavens, 181, ed., 1850, by Denison Olmsted.