Diagram showing the positions of Jupiter's Satellites VI, VII, and VIII, from photographs taken at the Royal Observatory, Greenwich, during the opposition of 1907-8. (Plate 15.)

(Communicated by the Astronomer Royal.)

The diagram is a graphical representation of the observations of Jupiter's Satellites VI, VII, and VIII, from photographs taken at the Royal Observatory, Greenwich, with the 30-inch reflector during the opposition of 1907-8, printed in the Monthly Notices, vol. lxviii. p. 582. In all, 38 photographs of Satellite VI, 21 of Satellite VII, and 13 of Satellite VIII have been obtained.

The positions plotted in the diagram are the means when two or more photographs were taken on any night, and the curves are drawn through these points.

For comparison, the orbits of the four inner large satellites are shown, the major axes being plotted to scale, but the minor axes intentionally exaggerated.

The apparent convexity of the path of the eighth Satellite with regard to Jupiter is due to the Earth's motion during the period over which the observations extend (about three months), and the consequent change in the point of view.

Royal Observatory, Greenwich: 1908 September 23.

Diagram showing the positions of Saturn's Ninth Satellite-Phoebe, from photographs taken at the Royal Observatory, Greenwich, during the opposition of 1907. (Plate 16.)

(Communicated by the Astronomer Royal.

This diagram is a graphical representation of the observations of Saturn's ninth Satellite-Phoebe-from 16 photographs taken at the Royal Observatory, Greenwich, with the 30-inch reflector during the opposition of 1907, printed in the Monthly Notices, vol. lxviii. p. 211.

The observed position of Phoebe with respect to Saturn on each day of observation is plotted in the diagram, and the curve drawn through these points.

For comparison, the orbits of the inner satellites are shown plotted to scale.

It may here be noted that during the present opposition (1908) 23 photographs on 15 nights have already been obtained up to the present time, the first being on July 31, some few days before Phoebe reached eastern elongation.

Royal Observatory, Greenwich:

1908 September 23.

The Relation between Star Colours and Spectra.
By W. S. Franks.

In a former paper on this subject (Monthly Notices, June 1907) I gave the result of an investigation into the relation between the colours and spectra of 1360 bright stars, in both hemispheres ; the colours depending on the observations made by members of the British Astronomical Association in England and Australia. The stars observed in this country, from the N. pole to 25° south declination, amounted to 928 of the total number. The recent publication of the Revised Harvard Photometry, which contains over 9000 stars down to 6 magnitude, has enabled me to carry this inquiry a stage further. Making use only of my own colour observations, contained in a MS. synoptic catalogue of 4175 stars (which includes all my previous work in this direction), I have carefully compared this with the Harvard revised spectra. The number

of stars between the N. pole and -25°, common to both lists, amounts to 3497; which is nearly four times greater than the B.A.A. total, and it includes stars down to 6 magnitude. As the colour estimates in my earlier catalogues were not expressed by symbols, like the B.A.A. and my later observations, I have used the old notation in this paper-dividing the colours into seven groups, as here shown :

1. White

2. Yellowish White

4. Yellow

Stars-including bluish-white and greenish-white

3. Pale Yellow

-(normal tint) including orange yellow and full yellow

-(normal tint) including also full orange

-including all ruddy tints

The foregoing colours are distributed as follows:

Analysis of Colours and Spectra.

Od Oe Oe5B B

Classification of Spectra in Tables II. and III.

(Intermediate) Procyon

Solar lines Capella

"K" line intense Arcturus,,

(Intermediate) Aldebaran K5M

Banded spectra Betelgeuse,,

Ma

= II-III.

= III.

do. (reversed) 19 Piscium,, N (ruddy stars) = IV.

The last class (N) would be considerably increased but for the difficulty of photographing objects near the red end of spectrum— hence they fall short of the number that can be observed visually. A number of red and orange stars of 6 magnitude, contained in my catalogue, are, from this cause, entirely omitted from the Harvard spectra results.