Table IV.-Secular Changes of the Declination and Inclination at the Cape of Good Hope on the Magnetarium. The secular changes of the magnetic elements at St. Helena are interesting from the fact that the declination period agrees well with the geographical position of this island, which is west of the meridians of London and the Cape of Good Hope. The epochs of the zeros of the declination at these places are in the order of their longitudes respectively, As St. Helena is south of the equator, the outward westerly march of the declination needle has a long period which is not yet completed (prop. V). The length of this period, as shown on the magnetarium, is 256 years, and will arrive at its maximum in the year 1939. Further proofs of the truth of the theory that the secular changes in the magnetic elements are caused by the rotation of an electro Table V.-Secular Changes of the Declination and Inclination at St. Table VI.-Secular Changes of the Declination and Inclination at St. Helena on the Magnetarium. dynamic substance within the earth's crust are afforded (1) by the dipping needle making only one downward or upward motion (outside the space comprised within the north and south limits of the magnetic equator) for the outward and return march of the declination needle, as instanced in the continued diminution of the dip in the British Isles during the westerly outward and return march of the declination needle since the year 1723. (2) From the secular changes of the dip in opposite directions about the same meridian in the northern and southern hemispheres, as instanced in the dip diminishing in the British Isles and increasing at the Cape of Good Hope and St. Helena for the same epoch (prop. X). (3) The rapid increase of the dip about the nodes of the magnetic equator, as first indicated by Sabine in the Gulf of Guinea and St. Helena. Before the time of my experiments no attempt had been made to assign a cause for the large and rapid change in the dip (about 17 minutes annually) on this part of the terrestrial surface until I found that the results obtained on the magnetarium agreed very closely with the observations. Subjoined are tables of the declination and inclination at St. Helena and on the magnetarium. On working the inclination backwards on the magnetarium chronologically, it will be seen from the Table VI that about the year 1747 the dip changed the sign from south to north. As no observations were made on the dip at St. Helena previous to the year 1825, there is no record of this interesting fact, nor has it hitherto been deduced from theory. A further confirmation of the agreement of the results obtained on the magnetarium with actual observations has recently been brought under my notice in the bulletin issued by the United States Coast and Geodetic Survey, No. 23, March 16th, 1891. In this publication Table VII.-Secular Changes of the Declination and Inclination at Ascension Island and on the Magnetarium. * Annual differential motion of internal sphere 22'50-375° from zero at London in the year 1657. a table is given of the declination and dip at Ascension Island for the epoch 1834-1890, in which the change of the dip from north to south occurs between the years 1842-1839. The same change is not only shown on the magnetarium but also the amount of the dip and declination for the epoch 1834-1890. The correlation of the maximum rate of change of the inclination with the minimum rate of change of the declination about the nodes of the magnetic equator is well seen in the observations, and is also demonstrated in my paper (prop. XVIII) and on the magnetarium. III. "Terrestrial Refraction in the Western Himalayan Mountains." By General J. T. WALKER, C.B., R.E., F.R.S. Received January 13, 1894. In the operations of the Great Trigonometrical Survey of India it is customary to determine the coefficient of refraction by reciprocal vertical observations between contiguous stations on the sides of all the principal triangles, and also as many as possible of the secondary triangles. [The sum of the reciprocal vertical angles, plus the angle at the earth's centre, would be exactly equal to 180° if there were no refraction; the excess gives the sum of the refractions in both the angles, and half of it is taken as the amount for each angle.—March 2.] The values of the coefficient thus obtained for the operations in the Western Himalayas-between the meridians of 73° and 80° east of Greenwich-have been grouped together for comparison in successive ranges of 2000 ft. of altitude between the elevations of 5,000 and 21,000 ft. above the sea level. The operations happen naturally to have been divided into two sections; for the regions lying between the great snowy ranges on the southern face of the Himalayas and the plains of India were first completed, and some time subsequently the still higher regions to the north, extending up to the Karakoram and Kuenlun Ranges, which look down on the plains of Turkestan. The first portion appertains to what is called the N.W. Himalayan Series, the second to what is called the Kashmir Triangulation. Thus the values of the coefficients of refraction were obtained separately for each section, and the results are shown in the following table, where the heights have reference to the middle points of the sides of the triangles, of which the number is given for each group : |