These radiations are not specially bright in the comet's nucleus, but they extend from 5° to 15° into the tail, whilst the abovementioned bright lines seem to be confined to the region of the nucleus. Altogether 12 lines or groups of lines have been measured on this plate, and the wave-lengths determined graphically, using the 12 hydrogen lines from Hẞ to Hy inclusive in the spectrum of Procyon and of Sirius to obtain the dispersion curve. In the spectrum of September 3, with Procyon photographed on the same plate, the pair of bright lines in the ultra-violet are seen to be a little more refrangible than H¿ (λ 3889); and as their distance apart corresponds with about 11 units in wave-length, there can be no hesitation in ascribing them to the cyanogen lines at A 38715 and A 3883 5. In this plate, however, the images are too faint for

measurement.

In the table which follows are summarised the results of measures of the spectrum of August 28. As most of the spectral images on this plate are broad and ill-defined, I give in the second column the approximate limits of wave-length of each line or band; and in the third column the mean wave-length, or the positions of maximum brightness. In the fourth column is entered the length of tail in degrees which can be traced in each radiation. I can assign no origin for the radiations, which seem characteristic of the tail, except the band at A 455, which, according to Campbell, may probably be due to cyanogen. The continuous spectrum, which

*

is faint throughout, ends at about λ 490.

Note on the Ancient Solar Eclipses discussed by Mr. Cowell.
By A. C. D. Crommelin.

The results derived from these eclipses by Mr. Cowell are so important and far-reaching that it is desirable to remove all uncertainty as to possible errors, especially in view of the fact that Mr. Nevill published some results purporting to be deduced from the same formula, but inconsistent with those given by Mr. Cowell himself; I accordingly undertook an independent calculation of the six eclipses - 1062, -762, 602, 584, 430 and which are the most important and best authenticated ones. obtain results practically identical with Mr. Cowell, the discordance in no case exceeding 2", which is an absolutely negligible quantity.

197,

I

Several smaller terms in latitude were then introduced, which had some slight effect on the residuals, but not enough to affect the argument in any way; they possibly produce a small diminution in the relative acceleration of the Sun and the node, but only to the extent of" or less.

-

Two points suggest themselves as worthy of mention regarding the eclipse of 1062; viz.—(1) Mr. King's recently published book gives a translation of the inscription in which it occurs, which is a record of omens occurring in the city of Babylon (wild beasts entering the streets, dogs entering the temples, etc.); hence there

can be no reasonable doubt that the extinction of the Sun, recorded as an omen, was likewise witnessed from Babylon itself.

(2) The date of the phenomenon is given as Sivan 26; now, as the months began with the first observation of the Moon, the 26th of the month must be within a day or two of new moon; this decidedly increases the probability that the recorded solar darkening arose from an eclipse.

Both Professor Newcomb and Mr. Nevill have rejected the suggested explanation of the apparent acceleration of the Sun as arising from a diminution of the Earth's rate of rotation. The point, however, is such a simple one that it seems worth while to restate it in other words, in the hope that they may more definitely indicate their reasons for dissenting from it.

(1) A slowing of the Earth's rotation from tidal friction necessarily involves an increase in the Moon's distance, in order that the moment of momentum of the Earth-Moon system may be conserved.

(2) A rough calculation was given showing that the lengthening of the Moon's period arising from this may be nearly as great as the apparent shortening due to the increase in the length of the day.

(3) In the case of the Sun's period, the shortening due to the latter cause would act unimpeded, so that an apparent acceleration of the Sun comparable with that of the Moon would result.

Note on the Permanency of some Photo-visual Lenses.
By William J. S. Lockyer, M.A., Ph.D. (Plates 1, 2, 3.)

The Solar Physics Observatory at South Kensington is in possession of numerous photo-visual lenses which are employed in several instruments in frequent use. All these objectives have in time developed curious markings on one or more of their internal surfaces. The object of the present note is to draw attention to, and to give a brief description of these markings.

Each of the lenses in question consists of a combination of three elements, the front and back lenses being double convex and the middle double concave. The front and middle lenses are nearly in contact, but a small air-space separates the middle from the back lens.

The material with which the lenses are made is as follows:The front lens is made of Schott's baryta light flint glass (0 543), the negative or middle lens is of Schott's boro-silicate flint glass (a variety of their O 164), while the third is made of a light silicate crown (Schott's O 374).

The lenses to which reference will be made vary in aperture from 3 to 12 inches, and have been left, as regards their inner surfaces, untouched in their respective instruments for periods varying from 23 to 83 months.

The preceding table contains a summary in convenient form indicating the intervals between the times of mounting and dismounting the objectives, the conditions of their situation, and brief notes on the markings developed.

In order, in the first instance, to indicate the appearance of the markings in question, the following four selected photographs will serve as examples. It may, however, be first mentioned how these photographs were secured.

The whole combination was first removed from the instrument in which it is employed and placed in a V-shaped wooden stand in front of a window. The camera was placed about 2 feet away from the lens and about 3 feet from the window, when the markings were carefully focussed on the ground-glass. The system of illumination of the lens under examination was as follows. The space between the lens and the camera was first covered in with black velvet. About a foot away from the lens, but on the opposite side of it to the camera, a large dull black piece of cardboard was placed near the window in nearly a vertical position to serve as a black background. The lens itself was illuminated by daylight passing through the window and falling obliquely on the back portion of the combination. Unfortunately it was found very difficult to eliminate the reflections of the window from appearing on the lens, but they are not sufficient to detract from the main object for which the photographs were taken. The photographs show, therefore, the peculiar markings on more of the inner surfaces of the lenses, as white on a dark background.

or

Fig. 1 (Plate 1) represents the 3-in. objective which has been employed in an instrument for 83 months. The markings, it will be noticed, are small, and near the middle give the lens a distinct mottled appearance. Nearer the edges their structure becomes more pronounced, and there is a greater distance between them.

Fig. 2 (Plate 1) represents the markings on another 3-in. lens which forms the camera objective of the same instrument. This lens was mounted and dismounted at exactly the same time as that shown in fig. 1 (Plate 1), and during this interval of 83 months it was situated within a foot of the other lens.

The different structure and number of the markings is very striking, and can be judged best by closely examining the photographs.

In fig. 1 (Plate 2) we have another example of quite a different appearance. This lens has an aperture of 3.5 inches, and has been in use as an objective to a small telescope for 40 months. The markings in this case are of a much finer type than those previously mentioned, and are far more elaborate in their structure.

This lens exhibits, in addition, a second set of markings, which appears on another of the interior surfaces of the combination, but which could not be photographed under the same illumination and distance, and therefore cannot be seen in fig. 1 (Plate 2). By