generally, so that this barometer may be regarded as recording the true pressure of the atmosphere. This was more exactly secured in making comparisons of the two barometers by selecting only those cases when winds at the Fort William Observatory were light. As stated by the Committee in previous reports, the observations of the force of the wind are estimations on a scale of 0 to 12, the equivalent of each figure of the scale in miles per hour having been carefully determined by Mr. Omond by means of Chrystal's anemometer. The barometric observations at the two observatories were reduced to sea-level hour by hour, and the differences plus or minus were entered in columns representing the different wind forces at the higher observatory. The following is the result of the comparison :—

Thus in calm weather the two reduced barometers are practically the same, but with every increase of wind which sweeps past the higher observatory, the depression of the barometer inside steadily augments. It is not till a velocity of more than 20 miles an hour is reached that the depression amounts to one-hundredth of an inch. At 57 miles it is 0-050 inch, at 77 miles 0.104 inch, and at 99 miles 0·150 inch. In forecasting weather it will be necessary to keep this effect of high winds on the barometer constantly in mind, with the view of arriving at a better approximation to the geographical distribution of pressure at the time the forecasts are being framed.

These results are for all winds grouped together irrespective of their direction. The next inquiry grouped the winds according to their direction to sixteen points of the compass. During the time under examination, all the very high winds were from E.S.E. or S.E., these being the directions in which the wind blows freely along the slopes of the mountain to the observatory. In 11 cases the wind from these directions attained a velocity of 100 miles an hour or more, and the reduced barometer of the high-level station read about one-sixth of an inch lower than the barometer of the low-level observatory. In no other of the 16 directions was there, during the ten months, a higher velocity than 62 miles an hour observed, and indeed in the directions E., E.Ñ.E., N.E., N., N.W., and W. the observed velocity was never greater than 29 miles an hour. With these northerly winds the observations at the top of the mountain indicate a much lower speed than that which, from the drift of the clouds, is seen to be reached at a comparatively small height above the top of the Ben. The cause of this comparatively calm state of the air immediately on the top is the impact of the air on the face of the tremendous cliff, close to the top of which the observatory is built, by which the stream 1891.

L

lines are suddenly deflected upwards. Now in such cases the depression of the barometer is about three times as great as that which occurs with an equally strong wind from other directions, and indicates clearly the formation of a restricted region of low pressure around and outside the observatory. Another curious and highly interesting result observed with other directions of the wind is that the reduced high-level barometer exceeds the reduced low-level barometer when the wind blows at the rate of about 5 miles an hour. This increased pressure accompanying wind rising up the slope of the hill may perhaps explain the small clear space immediately on the top of a hill, otherwise cloud-topped, and the very different force of wind on the two sides of a ridge lying about a right angle to the direction of the wind.

An examination has also been made of the relations of differences of temperature at the two observatories to differences of the sea-level pressures at the same hours. During the ten months examined the temperature differences have ranged from the high-level observatory showing a temperature 26° lower to a temperature 6° higher than the temperature at Fort William at the time. A comparison has been made by sorting the differences into two-degrees amounts, and instituting a comparison only on those cases when the strength of the wind at either of the observatories did not exceed 26 miles an hour.

The following show for each two-degrees difference of temperature the difference between the reduced barometer of the top and the barometer at Fort William, the plus sign indicating that the top barometer was the higher, and the minus sign that it was the lower of the two :

The broad result is this, and it is clear and explicit, when the higher observatory has the higher temperature, and when the differences of temperature are small, then the reduced pressure at the top of the mountain is the greater of the two; but when the differences of temperature are large then the reduced pressure at the top is the less of the two. The regular progression of these figures show that what is substantially a true average has been obtained. The result, which is altogether unexpected, raises questions of the greatest importance, affecting the theory of storms, the effect of vertical movements of great masses of air on the barometric pressure which accompanies cyclones and anticyclones, and the necessity there is for some accurate knowledge of the absolute amounts of aqueous vapour at different heights in the atmosphere under different weather conditions. Ben Nevis, with its two observatories, one at the top, the other at the foot of the mountain, would, with a third halfway up the hill, afford unique facilities for the prosecution of this all-important hygrometric inquiry, which would, however, require considerable additions, for the time it is carried on, to the observatories' present appliances and staff.

Third (Interim) Report of the Committee, consisting of Professor FITZGERALD, Dr. JOHN HOPKINSON, Mr. R. A. HADFIELD, Mr. TROUTON, Professor ROBERTS-AUSTEN, Mr. H. F. NEWALL, and Professor BARRETT (Secretary), on the various Phenomena connected with the Recalescent Points in Iron and other Metals.

THE Committee reported at some length last year, and wish to postpone a further report till next year. They desire, therefore, to be reappointed without a grant.

Second (Interim) Report of the Committee, consisting of Dr. JOHN KERR, Sir WILLIAM THOMSON, Professor RÜCKER, and Mr. R. T. GLAZEBROOK (Secretary), appointed to co-operate with Dr. KERR in his researches on Electro-optics.

THE Committee report that Dr. Kerr is continuing his experiments on Electro-optics, and hopes to be able to get some definite results for the meeting next year. They wish to be reappointed.

Report of the Committee, consisting of Professor LIVEING, Dr. C. PIAZZI SMYTH (Secretary), and Professors DEWAR and SCHUSTER, appointed to co-operate with Dr. C. PIAZZI SMYTH in his researches on the Ultra-violet Rays of the Solar Spectrum.

THE first proceeding of this committee after authorisation was to inquire into all that their Secretary was proposing to do in the way of observation and record in the ultra-violet of the solar spectrum and the sufficiency or otherwise of the apparatus he had already collected for the purpose. Much correspondence followed through the autumn and in the winter of 1890–91, and it soon became evident that only a small part of what was scientifically necessary could be procured with the amount voted.

In February, 1891, however, a most agreeable surprise occurred, in the shape of a resuscitation of a still earlier application on the same general lines, but on a wider basis, by Dr. C. Piazzi Smyth to the Royal Society's Government Grant Committee in July, 1890, and which he erroneously imagined, from their silence after receiving it, had not been approved by that body. But it had been simply kept in abeyance, and was finally pronounced favourably upon and granted in 1891. This measure happily relieved the British Association Committee from attempting to do altogether too much for its small means, though still requiring the utmost economy in their disposition, as well as their limitation to the exact line pointed out in the resolution passed by the General Committee at Leeds, viz., 'to co-operate with their Secretary in his researches on the Ultra-violet Rays of the Solar Spectrum.'

Now this part of the spectrum being absolutely invisible to the eye, though otherwise known to be in the field of the Secretary's Grating spectroscope at the time, while the focus of the inspecting or photograph

ing telescope thereof varied rapidly with the smallest angular change of its direction in spectrum place, there arose a necessity for a considerable improvement of the focussing arrangement over and above what is usually supplied for the visible parts of the spectrum, or had been furnished in the present instance for all parts. But this improvement has now been accomplished by Messrs. T. Cooke & Sons, of York, according to a design by the Secretary, enabling the focus to be set distinctly and solidly to the thousandth of an inch without reference to anything but numerical tables prepared beforehand and tested by photographic record.

Again, however, in some of the most interesting of those ultra-violet regions of solar spectrum light a further and more intricate difficulty of a physical nature was found when photographing in the second order of the Grating's spectra. For, though that operation was performed under double shields of the darkest blue glass procurable, yet the red region of the first order of spectrum would insist on breaking in through all obstacles, and showing itself even brilliantly by means of the anomalous ultrared ray transmitted by the supposed most pure and densely blue, or violet, glass known! One possible method of getting rid of this difficulty immediately seemed to be by photographing only in the first order of the Grating's spectrums, throughout whose violet fields there is no red band of any other order to come in-blue glass in place or not. But could sufficient spectrum separation of lines be thereby obtained, and without any other drawback?

To meet this essential problem Messrs. T. Cooke & Sons, of York, were again applied to, and they constructed within the grant made to the Committee an extra-large Barlow photo-achrom-concave lens, which magnified the previous image of the inspecting telescope's object-glass by 2.3 times, or rather more than the first order of the Grating's spectrums is magnified, in separation only, by the second order. And if by the Barlow concave the magnifying is both in separation and in height of lines (and therefore weakening to the intensity of the image), it was hoped that longer exposures could be freely given. So that then, with them, would come the final trial, which has still to be made whether the exquisite definition of the first order of spectrum cannot be lenticularly magnified to the required degree, with less loss of that still more valuable feature, definition, than what takes place when it is diffractionally magnified (at least in the Secretary's Grating spectroscope) by resorting to its second order of spectrum?

This is the main point, then, up to which the Secretary's research has just arrived by aid of the British Association's grant of 1890. For while the whole of that sum has now been expended on the above-mentioned major subjects and a number of minor improvements and working particulars bearing on the same ends, and nothing further in the way of grant is now being asked for, it leaves sufficient material in Dr. C. Piazzi Smyth's hands for much work in the months to come. In earnest whereof he begs to send some of his accomplished work during the last nine months, in the shape of two album cases, each containing twenty-six of his separately mounted and scaled but continuous solar spectrum magnified photographs of lines in the violet and ultra-violet, besides a third and thinner album case of previously taken eye-and-hand-made drawings at the same instrument, but of the easier half only of the same subjects, for inter-comparison of the two methods which are past, and in preparation for the third, which is to come.

Report of the Committee, consisting of Professor W. GRYLLS ADAMS (Chairman and Secretary), Sir WILLIAM THOMSON, Professor G. H. DARWIN, Professor G. CHRYSTAL, Professor A. SCHUSTER, Professor RÜCKER, Mr. C. H. CARPMAEL, Commander CREAK, the ASTRONOMER ROYAL, Mr. WILLIAM ELLIS, and Mr. G. M. WHIPPLE, appointed for the purpose of considering the best means of Comparing and Reducing Magnetic Observations.

In accordance with the arrangements made last year for determining the mean diurnal range from the observations taken on five days in each month, the following list of quiet days during the year 1890 has been selected by the Astronomer Royal as suitable for the determination of the magnetic diurnal variations:

During the past year the magnetic survey of the United Kingdom, now in progress under the superintendence of Professors Rücker and Thorpe, has advanced rapidly. Messrs. Gray, A.R.C.Sc., and Watson, B.Sc., A.R.C.Sc., are at present working in Ireland and Scotland respectively. A body of computers has been organised at South Kensington, so that the reductions are proceeding pari passu with the observations, and by the end of this summer complete observations will have been made at more than 600 stations in the British Isles.

On June 18 last, in a paper read before the Royal Society on the 'Comparison of Simultaneous Magnetic Disturbances at several Observatories, and Determination of the Value of the Gaussian Coefficients for those Observatories,' the Chairman pointed out the importance of adopting the same scale-values for similar instruments at different observatories, especially at new observatories which have been recently established, and discussed special magnetic disturbances, especially the disturbances of a great magnetic storm which occurred on June 24 and 25, 1885, for which photographic records have been obtained from seventeen different observatories: eleven in Europe, one in Canada, one in India, one in China, one in Java, one at Mauritius, and one at Melbourne.

In this paper the records are discussed and compared, tables are formed of the simultaneous disturbances, and the traces are reduced to Greenwich mean time and brought together on the same plates arranged on the same time-scale. Plates I. and II. show the remarkable agreement between the disturbances at the different observatories, and the tables show that the amount of disturbance, especially of horizontal magnetic force, is nearly the same at widely distant stations.