pared the rainfall at the Fort William Observatory with the monthly amounts collected with the old gauge at Mr. Livingston's during the year and a half ending with December 1892, with the result that the rainfall at the observatory is 4.6 per cent. less than at Mr. Livingston's. The largest rainfall of any month at the top was 22:32 inches in January, and at Fort William 14:42 inches in September, the smallest amounts being respectively 5.42 inches and 1.84 inch, both occurring in March. The heaviest fall on any single day at the top was 5.76 inches on January 25, and at Fort William 2.83 inches on November 28. The number of days on which the rainfall was nil, or less than onehundredth of an inch, was 116 at the top, and 152 at Fort William. At the top the largest number of days was 16 in July, and the smallest, 5, in January; and at Fort William the numbers were 23 in March and 5 in January. At Fort William a fall of an inch of rain a day or more occurred on 13 days during 1892, but at the top of Ben Nevis on 50 days, or nearly four times as often. From one to nine such wet days were recorded at the top each month, whereas at Fort William no such wet days occurred in March, April, June, July, and September, and on four of the other months only one day each. At Fort William the mean atmospheric pressure at 32° and sea-level was 29.852 inches, and at the top 25.284 inches, thus giving a difference of 4.568 inches. The lowest pressure at the top for the year was 24.147 inches at 2 P.M. of February 2, and the highest 25.960 inches at 7 P.M. of March 22, the difference being 1.813 inch. Mr. A. J. Herbertson is carrying on at the two observatories the research on the hygrometry of the atmosphere referred to in our report of last year. During last autumn and early winter he spent upwards of four months at the Ben Nevis Observatory, and there measured the aqueous vapour by direct weighing, obtaining a new and valuable series of experiments at very low temperatures and humidities. Since July last a similar set of experiments are being conducted by him and two skilled assistants, involving measurements of the aqueous vapour simultaneously at both high- and low-level observatories. Observations are made at the same times with the dry and wet bulb, both in a Stevenson screen and in an Assmann aspirator psychrometer; with Regnault's hygrometer; and of the numbers of dust particles present, the general weather conditions as to barometric pressure, sunshine, wind, &c. Mr. Herbertson has communicated a preliminary report on the results of last year's observations and experiments to the Royal Society of Edinburgh, in which among other points of interest the very unsatisfactory character of the hygrometric tables now in use is clearly shown for low temperatures and great dryness. The experiments at present in progress give the corresponding data for summer temperatures and humidities; and, seeing they are conducted simultaneously both on the top of the Ben and at Fort William, the effect of large differences of pressure will be seen on the relations of the dry and wet bulb readings to the amounts of aqueous vapour actually present in the air at the time. The Directors regard these experiments as of great importance, not only as furnishing data towards a better knowledge of the hygrometry of the atmosphere, but also as leading to much needed improvements in the methods of reducing the readings of the dry and wet bulb thermometers. Mr. Omond has written a paper on hourly readings of a black bulb thermometer in vacuo as compared with the readings of the dry bulb thermometer in the Stevenson screen, which was published in the last 'Journal' of the Scottish Meteorological Society. The results are interesting and suggestive. He has prepared a paper on the height of the lower edge of the cloud layer on Ben Nevis, based on observations taken at Inverlochy during some months of the present summer; and has also instituted a comparison of the temperature observations from August 1890 to December 1891, made simultaneously by himself at the new observatory and by Mr. Livingston at the old station with the instruments in use there since 1883. At last year's Meeting of the Association a grant of 50l. was made to aid in the payment of assistants to perform the strictly routine work of Dr. Buchan and Mr. Omond, so that their time would be set free for the discussion of the observations of the two observatories beginning with August 1890. This arrangement has been carried out, and the following is a detailed statement of the work which has been completed or is still in progress. From the first eight years' observations of the rainfall at the top of Ben Nevis the mean hourly variations for the twelve months of the year have been calculated, and the hourly values reduced to percentages above or below the monthly means. The results were then 'bloxamed' in the usual XII+I+II; for way, that is, the value for 1 A.M. of January equals I+II+III 3 1 A.M. of February &c., where the Roman numerals represent the values of the months December, January, February, &c., for these years. In this way Table IV. has been constructed, which shows the diurnal variation in the precipitation throughout the year. As was to have been expected, the curves for the warmer months of the year are best marked. These show a clearly defined double maximum and minimum. The larger maximum occurs from 11 A.M. to 8 P.M., or during the warmer hours of the day after the ascending current has set in. Then a minimum occurs from 8 P.M. to 1 A.M., or during the hours when temperature falls most rapidly, and the evening maximum of pressure prevails. For the next six hours precipitation is above the average, the greatest increase being from 5 to 6 A.M.; and finally from 7 to 11 A.M. the next minimum occurs, or during the hours atmospheric pressure and temperature increase, and terminates with the formation of the ascending current, which is so pronounced a feature in the meteorology of Ben Nevis. During the colder months the curves are less distinctly marked, except a decided maximum during the coldest hours of the day, and a minimum during the hours of the morning barometric maximum, when temperature is rising. The hourly variation of the rainfall is more clearly shown than at any other observatory at which hourly observations have been made from results extending over a comparatively short term of years. The discussion of the hourly barometric and thermometric means at the two observatories for the three years is nearly complete. An inquiry into the diurnal variation of the barometer and thermometer on Ben Nevis during days of clear weather on the one hand, and days of fog or mist throughout on the other, is completed for the three years ending August 1893. The inquiry had not proceeded far when it was apparent that the curves for clear weather and those for clouded weather while fog or mist was absent were in all essential respects the same; but the curves were of quite a different character when fog or mist prevailed. TABLE IV.-Showing the Hourly Variation of the Rainfall on Ben Nevis, expressed in Percentages above or below the daily means. For Hour ending Jan. Feb. Mar. April May June July Aug. Sept. Oct. Nov. Dec. Year Hence, only those days were entered as foggy or misty when fog or mist was recorded for each of the twenty-four hours of the day. The hourly temperature and pressure of such days were extracted from the daily sheets, and the averages for each month calculated. Those days were regarded as clear days when the sun shone at least several hours, and when fog or mist was virtually absent. Means were similarly calculated for these clear days. Thereafter the monthly means for the three years were ascertained, and the hourly results 'bloxamed' as explained above. The results show two sets of curves, essentially different the one from the other, the monthly curves for foggy and misty days revealing a diurnal variation of pressure quite distinct from that of the curves for clear days. Table V. shows the differences between the two sets of curves, the plus sign (+) indicating a higher pressure for foggy days, and the minus sign (-) a lower pressure for those days as compared with the pressure for clear days at the same hours. With clear skies the daily maximum pressure occurs at 11 A.M. in winter, but at 2.30 P.M. in summer; whereas with fog or mist it occurs at all seasons between 10 and 11 P.M. With clear skies the minimum occurs at 4.30 A.M., but with fog at 6 A.M. From Table V. it is seen that, with fog, pressure is higher than with clear skies from 7 P.M. to 4 A.M., attaining the absolute maximum at midnight, but lower from 5 A.M. to 6 P.M., the absolute minimum being about noon. The important bearing of these results on solar and terrestrial radiation and other physical inquiries is obvious. No small part of the large excess of pressure during the night hours in fog is probably occasioned by the latent heat set free in the condensation of the aqueous vapour into fog or mist. This necessarily, in the circumstances, increases the barometric readings where it occurs, viz., on the top of the mountain, and particularly at night when the surface temperature of the mountain. is low. Owing to the high winds which often prevail at the time, and the formation of the fog being chiefly confined to the restricted area of the mountain top, the increased pressure is not relieved by the formation of an ascending current, and hence pressure is increased at the top, being the restricted area where the condensation takes place. TABLE V.-Showing Difference of Pressure, in Thousandths of an Inch, at the Ben Nevis Observatory, 4,406 feet, between foggy days and clear days respectively. The plus sign shows pressure on foggy days the greater; the minus sign, less. Hour Jan. Feb. Mar. April May June July Aug. Sept. Oct. Nov. Dec. Year +11 +7 +9 +12 +14 +16 +11 +9 +11 11 +22 +16 +15 +15 9 10 + 8 +14 +9 +15 +18 +12 +12 +12 +17 +15 +18 +19 +15 +16 +21 +20 +22 +24 +28 +23 +21 +20 Midnight +21 +24 +16 +19 +19 +27 +24 +25 +25 +29 +26 +27 +24 The work of recopying on daily sheets the hourly observations of both observatories, which show at a glance the relations of the two sets of observations to each other, is in progress, and already about half of the three years is finished. These sheets show the relative times of occurrence at the top and bottom of the mountain respectively of changes of pressure, temperature, humidity, and the other subjects of observation, together with their relative amounts. As the work proceeds the entries on the sheets are compared with the bi-diurnal weather maps of the weekly weather report of the Meteorological Office, and copious notes are made of the relations of the observations of the two observatories to the cyclones and anticyclones of north-western Europe at the time. One example may be here referred to. It frequently happens that the temperature difference of the two observatories, which is generally about 16° 0, becomes less and less during the time of the anticyclone, and occasionally temperature is higher at the top than at the foot of the mountain. But as the anticyclone gives way, and the cyclone advances, temperatures assume their normal difference. Now the observations show two marked types of weather in these circumstances: one type when the difference is brought about by a falling temperature at the top while temperature at Fort William remains practically stationary; and the other type when temperature rises at Fort William while at the top of Ben Nevis it is stationary. These different types at the present stage of the inquiry seem to point to mportant well-marked characteristics of the approaching weather. Earth Tremors.-Report of the Committee, consisting of Mr. G. J. THE Committee were appointed to consider the advisability and desirability of establishing in other parts of the country observations upon the prevalence of earth-tremors, similar to those now being made in Durham in connection with coal-mine explosions. In the present report descriptions are given of several instruments which have been used in the study of earth-tilts and earth-pulsations.. The list is not a complete one, some having already been included in the reports of the Committee on the Lunar Disturbance of Gravity for 1881 and 1882, drawn up by Messrs. G. H. and H. Darwin.' The first of these valuable reports also contains an account of the pendulum with double-suspension mirror, used in the well-known experiments in the Cavendish Laboratory, Cambridge. The extraordinary sensitiveness of this pendulum led Mr. Horace Darwin to design another form of the instrument, smaller and simpler in construction, but capable of measuring smaller angles than are required in these experiments. The method of determining the angular value of the scale-divisions has also been altered. The new pendulum was made some months ago by the Cambridge Scientific Instrument Company, and Mr. Darwin, being unable at the time to give much attention to it, has lent it to the Committee for trial. The preliminary experiments have been made by the Secretary at Birmingham, chiefly under Mr. Darwin's guidance, and the results are described below. The Committee think it most desirable that the observations should be continued in Birmingham, and also made in other parts of the country. One of the most delicate instruments that have been used for the observation of earth-tilts is the horizontal pendulum of Dr. E. von Rebeur-Paschwitz, who has described the results obtained with it in several valuable memoirs. The Committee considered that an account of these observations in English would be of great use, and at their request Dr. von Rebeur-Paschwitz kindly consented to write the very interesting and valuable summary which forms an appendix to this report. Nadirane of M. C. Wolf. The nadirane erected in 1863 by M. d'Abbadie at Abbadia, near Hendaye, is described in the reports for 1881 and 1882,2 as well as some of the results which have been obtained with it. 1 Brit. Assoc. Reports, 1881, pp. 93-126; 1882, pp. 95-119. In referring to these reports in the following pages they will be quoted as the Reports for 1881 and 1882. 2 Brit. Assoc. Reports, 1881, pp. 116-118; 1882, pp. 102, 103. The following is a list of M. d'Abbadie's papers on this subject :— |