5^4 



NATURE 



[May 6, 1909 



appeared, but the responsibility for the worli seems entirely 

 his. The volume represents a complete magnetic survey 

 of the United States for the epoch January i, 1905, based 

 on observations at 4149 stations, including 3311 in the 

 United States itself, 575 in Canada, 201 in Mexico, and 

 62 in the West Indies. The great majority of the stations 

 were occupied by Coast and Geodetic observers, but 

 acknowledgments are made to Sefior Felipe Valle for 

 results from some seventy of the Mexican stations, and 

 to Dr. King and Prof. Stupart for a good many results 

 from Canada. In the United States the density of the 

 stations varied from one per ninety square miles in Mary- 

 land to one per 2924 square miles in Idaho, the average 

 being one per 973 square miles. Details as to instru- 

 ments, methods, and sites of stations are not given, having 

 been dealt with in previous volumes of the Survey. 



Table I., pp. 18-S7, summarises all the observations. 

 It gives the name, latitude, and longitude of the station, 

 the date of observation to o-i of a year, the observed 

 values of the three elements declination (D), inclination 

 (I), and horizontal intensity (H), the values of the 

 elements reduced to the epoch 19050, and, finally, the 

 authority. D and I are given to o'-i, and H to 

 o-oooi C.G.S. (or 10 7). The results are grouped under 

 the States of the Union. Table II., pp. 91-5, summarises 

 results obtained at sea in the Atlantic and Pacific Oceans, 

 and- in the Gulf of Mexico, at 241 stations, between 

 January i, 1903, and midsummer, 1907. In this case 

 results are given for total intensity as well as for D, I, 

 and H. Table III., pp. 101-7, shows the secular change 

 in D at eighty stations representing specified portions of 

 different States of the Union. Values of D are given at 

 ten-year intervals from 1750, when available, down to 

 1900. The values for 1905 and the estimated annual 

 changes at that date are added. Table IV., pp. 114-9, 

 gives secular-change data for D, I, and H for five-year 

 intervals, from 1S40 downwards, for forty-seven geo- 

 graphical positions ; e.g. twelve have latitude 45°, their 

 longitudes being respectively 65°, 70° . . . 110°, 115°, and 

 I22°.s W. Table V., pp. 123-150, contains values of D, 

 I, H (along with its north and east components), as well 

 as values of the vertical intensity (V) and total intensity 

 (T), at the intersection of all degrees of latitude and longi- 

 tude on the North .American continent between 17° N. 

 and 49° N. Values of D and I are given to o°-i, values 

 of H to 0001 C.G.S. These data for D, I, and H are 

 obtaiiied by scaling from the charts, the process not 

 claiming an accuracy exceeding o°.o5 in D and I, or 00005 

 in H. The other force elements were apparently computed 

 from these, but they are given to four significant figures. 



The first five charts, dealing respcctivelv with D, I, H, 

 V, and T, are each about 28 inches by 22 inches. As to 

 the method of construction of the charts for the three 

 first elements, we are told (p. 1531 that " the reduced 

 values for 19050 . . . were plotted on a base map of 

 about four times the size of the charts. Next the lines 

 (isogonals, isoclinals, isomagnetics) were drawn to con- 

 form as strictly as possible with the plotted results." The 

 y and T charts are based on values calculated bv com- 

 bining values of H and of I scaled from the charts for 

 ?hese two elements. The aim is at least to indicate all 

 local irregularities of any importance, and, as Dr. Bauer 

 truly remarks, " one cannot fail to be impressed by the 

 manifold irregularities shown by the lines." This remark 

 is especially true of the isogonals, but the V and T 

 isomag-'etics are also exceedingly irregular in the regions 

 bordering on the great Likes. In the charts, successive 

 D^ and I lines differ by 1°, successive H, V, and T lines 

 by ctoi C.G.S. These lines are drawn in red, geographical 

 details being in black. The D, I, and H charts also 

 con-tain blue lines, drawn to pass through the places where 

 the secular change of the element is the same. In 1905 

 the agonic line — along which the needle points to the true 

 north-^ran from a little to the west of Charleston, in 

 South Carolina, in a north-westerlv direction to the north- 

 east corner of Lake Michigan. The line of no secular 

 change ran, roughly, parallel to the agonic line, but about 

 250 miles to the west of it. To the east of the line of 

 no secular change the needle Is moving to the west, and 

 to the west of this line it is moving to the east. The 

 extreme annual change — met with on the Pacific coast — 

 NO. 2062, VOL. 80] 



is only about 4'. The phenomena, in a general way, are 

 such as would ensue from a southerly movement of the 

 north magnetic pole, and this is in general harmony with 

 the secular changes in the other elements. H is falling 

 and I increasing over nearly the whole United States, 

 except in the extreme north-east, west, and north-west. 

 There have, however, been remarkable changes in the set 

 of the secular change of late years, showing that the real 

 phenomenon is of a very complicated character, which 

 renders any forecast for the future very uncertain. 



The two last charts are of a different character from the 

 others. No. 6 shows " magnetic meridians," defined as 

 horizontal lines which have for their tangent at every 

 point the direction of the compass needle. No. 7 gives- 

 secular-change curves of two types, one showing the 

 change of absolute direction in space of the freely dipping 

 needle, the other showing changes in the horizontal 

 intensity. 



The work is one which merits, and will doubtless receive, 

 close attention from all who are engaged, or are likelj' 

 to be soon engaged, in magnetic surveys. It is interesting 

 to learn (p. 13) that corrections for diurnal inequality were 

 applied only in the case of the declination. In the case 

 of the inclination and horizontal intensity. Dr. Bauer's 

 opinion seems to be that corrections for diurnal inequality 

 " are, in general, of the order of the error of observation, 

 and certainly much less than the ' station error ' due to 

 the irregular distribution of the earth's magnetism." He 

 was presumably influenced, in part, by the consideration 

 that secular change in H varies over the United States 

 from +20 7 to —70 7 per annum, so that a knowledge of 

 the annual change, even to the nearest 10 7, must be 

 difficult to acquire in the regions more remote from mag- 

 netic observatories. How disturbances are dealt with does 

 not seem to be stated. If Dr. Bauer's views are correct, 

 and they are based probably on a wider experience than 

 that of any other man living, one cannot help thinking 

 that extreme refinement in field instruments or observa- 

 tions may be largely thrown away in the case of a general 

 survey of a large area. If we may borrow and extend a 

 military metaphor, supreme importance attaches, not so 

 much to the gun, or even to the man immediately behind 

 it, as to the general. 



In the second volume we have a survey for the epoch 

 19055 of the Dutch East Indies, made under the direc- 

 tion of Dr. W. van Bemmelen, of the Batavia Observa- 

 tory. The area included extends from Sumatra in the 

 west to the borders of Dutch and German New Guinea 

 in the east (95° to 141° E.), and from Timor in the 

 south to Mandanao (Philippines) in the north (11° S. to 

 8° N.). The area is thus very big, including much sea 

 and many small islands, in addition to Sumatra, Java, 

 Celebes, and parts of Borneo and New Guinea, "rhere 

 were 158 stations, a considerable number being on the 

 smaller islands, but none at sea ; they were occupied during 

 1903 to 1907. Owing, no doubt, in part to the rela- 

 tively small number of his stations, Dr. van Bemmelen's 

 practices are in many respects the antithesis of Dr. 

 Bauer's. The former attaches comparatively little import- 

 ance to the exact site of his stations. On the other hand, 

 though having continuous records from only one observa- 

 tory, he applies corrections to all the elements to eliminate 

 the diurnal inequality and irregular variations, going to 

 o'.i in declination and dip, and to i 7 in horizontal force. 

 The Dutch charts, again, unlike the American, take no 

 account of local peculiarities, but resemble the world 

 charts of the British Admiralty in the bold sweep of their 

 lines. They are drawn on thick paper in the body of the 

 volume, and, again unlike the American, present quite an 

 artistic appearance. 



Dr. van Bemmelen gives the values of the magnetic 

 elements at intersections of degrees of latitude and longi- 

 tude, going to o°-oi in declination, o"-! in dip, and to 

 o-oooi C.G.S. in the total force and in its vertical, hori- 

 zontal, southerly, and easterly components. He repro- 

 duces, with some corrections, the results of two previous 

 surveys of the East Indies, one for the epoch 1848-0. by 

 Captain C. M. Elliot, the other for 1876-5, by Dr. _ E. 

 van Ryckevorsel, and makes use of these in discussing 

 the secular change. Though representing a much smaller 

 amount of field work than the American, the Dutch survey 



