334 



NATURE 



[December 27, 1917 



Galongs. This is written by Capt. Sir George Duff- 

 Sutherland-Dunbar and is beautifully illustrated. 



To mention by name even all the more interesting 

 papers given in these Memoirs and Journals woula 

 occupy many pages; as already suggested, they give 

 abundant evidence of a new lite in our Eastern Em- 

 pire. The suggestion might be offered, however, that 

 the division of these publications into at least three 

 sections, each with its own separate volume, would be 

 both an economy and a convenience. 



NEW FRENCH MAGNETIC CHARTS.^ 

 T N France terrestrial magnetism is included in 

 •*■ meteorology, and the actual survey upon which 

 the present work is largely dependent was made by 

 M. Moureaux, director of Pare St. Maur Observatory, 

 then the central magnetic station for France. Prof. 

 Angot, who is director of the French meteorological 

 service, was responsible for the last magnetic charts 

 relating to the epoch January i, 1901. Whether fresh 

 charts will continue to be published every ten years 

 appears as yet to be undecided. Two methods were 

 considered of obtaining the secular change data, neces- 

 sary to derive results for January i, 191 1, from those 

 for 190 1. The first consisted in taJiing fresh field 

 observations in a sufficient number of places, and some 

 observations having this end in view were taken by 

 M. Ebl^ in 1912 and 1913. These have served to 

 some extent as a control, but the second method was 

 that actually depended on. It consists in utilising the 

 secular change data published by observatories in 

 France and adjacent countries, including Potsdam, 

 De Bilt, Valencia, Greenwich, Kew, Falmouth, Val 

 Joyeux, Munich, Pola, Naples, Coimbra, and San 

 Fernando. The ten-year secular changes at these 

 stations were plotted in a map, and curves of equal 

 secular change drawn^ from which were deduced the 

 secular changes appropriate to each station. The 

 method is obviously more suitable for France than for 

 the British Isles. _ But even in the case of France, in the 

 absence of positive knowledge that. secular change is 

 unaffected by local disturbance, it is doubtful whether 

 it will be universally admitted that the method is alto- 

 gether satisfactory for the deduction of charts showing 

 the local anomalies. It is obviously simpler, however, 

 than the carrying out of observations at a large num- 

 ber of repeat stations. 



The values deduced for the epoch January i, 191 1, 

 for declination, inclination, horizontal and vertical 

 force, north and west components, and total force are 

 given for from 500 to 600 stations, arranged alphabetic- 

 ally_ under the several de:partments. The declination, 

 inclination, horizontal force, and vertical force data 

 are also embodied in four charts. Omitting a few 

 incomplete or obviously disturbed stations, the remain- 

 ing 538 were arranged according to geographical posi- 

 tion in twenty groups or areas. Taking any one 

 group, the mean of the observed values of, say, declina- 

 tion was assigned to an imaginary station, the geo- 

 graphical co-ordinates of which were the mean of those 

 of the actual stations. In this way values were found, 

 practically free from accidental irregularities, for 

 twenty different points. It was then assumed that 

 these twenty values could be represented by an expres- 

 sion, a + b<p + c\ + d<p^ + e(pX+fX^, where ^ + 47° and 

 A -1-2° represent the latitude and easterly longitude of 

 any station. The constants were determined both by 

 least squares and by Cauchy's method, with very satis- 

 factory results, showing that a simple quadra'tic ex- 

 pression suffices to give normal magnetic values with 

 high accuracy for the whole of France. 



./ J'.'^^*^^" magn^tique de la France et de I'Afrique du Nord (Tunisie, 

 Algeria, Maroc) au ler Janvier 191 1." By Prof. Alfred Aneot. Ann. du 

 bureau central m^t^orologique de igri, tome i., pp. 59-95-1-4 charts. I 



NO. 2513, VOL. IO0I 



Tunis, Algeria, and Morocco are treated by them- 

 selves (pp. 86-95). The available data consisted of ob- 

 servations taken by Moureaux at thirty-three stations 

 in 1887, and of recent results obtained by the observers 

 of the Carnegie Institution of Washington. The latter 

 had observed at thirteen of Moureaux 's stations, thus 

 obtaining data for secular change which were supple- 

 mented by results from the observatories of San Fer- 

 nando, Coimbra, Tortosa, Naples, and Helwan. A 

 six-constant formula of the type already described 

 seems to fit the observations reasonably well. Prof. 

 Angot would like, however, to have fresh observations 

 throughout North Africa, at a considerably larger num- 

 ber of stations. Declination, inclination, and hori- 

 zontal force charts, representing normal values for 

 North Africa as given by the interpolation formulae, 

 appear in the text, but on a reduced scale as compared 

 with that adopted for the French charts, which show 

 the actual anomalies. C. Chree. 



RAINFALL IN NORWAY DURING 1916.^ 

 'T^HE director of the Norwegian Meteorological In- 



-^ stitute has, with commendable promptitude, pub- 

 lished the twenty-first annual volume of rainfall data, 

 viz. that dealing with last year's returns. The daily 

 rainfall is given in extenso for about 200 stations, 

 additional information regarding the nature of the 

 precipitation, whether in the form of rain, snow, or 

 sleet, being afforded by the international symbol 

 affixed to the reading when the downfall was other 

 than rain. A monthly summary shows, for each of 

 476 stations, the actual precipitation, the maximum 

 daily fall, and date of occurrence, along with the 

 number of days with more than 01 mm. and more 

 than 10 mm. of rain respectively; the mean depth of 

 snow is also given and the greatest depth recorded. 

 The monthly and annual rainfall expressed as a per- 

 centage of the average is shown for sixty-four 

 stations. 



No general summary of the results appears, but 

 there is an excellent large-scale map in two sections 

 showing the distribution of the annual rainfall for 

 1916 by isohyetal lines drawn for each 200 mm. The 

 maximum rainfall, shown by the isohyet of 3000 mm. 

 (118 in.), appears in three small patches close to the 

 coast, between lat. 60° and 61° N., the highest 

 rainfall, 3127 mm. (123 in.), being at Indre Matre 

 (height 15 m.), in lat. 60° N., long. 6° E. The smallest 

 rainfall, about 200 mm. (8 in.), occurs in several areas 

 of no great extent north of the Arctic Circle, the most 

 extensive being an oval patch about eighty miles long 

 and fifteen miles broad, situated due south of Ham- 

 merfest. The isohyets in some districts near the coast 

 are very crowded, especially in areas contiguous to 

 the wettest spots, where the rainfall in rather less 

 than forty miles falls off from about 120 in. to 32 in. 



As compared with the average, the rainfall of 19 16 

 on the mean of sixty-four stations was 5 per cent, in 

 excess, but individual stations varied from 51 per cent, 

 above to 41 per cent, below the average. Rainfall 

 was much above the average at most stations to the 

 south of lat. 63°, but north of Trondhjem (lat. 63-4" 

 N.) there was a pronounced deficit, ranging in general 

 from 15 to 40 per cent. The only marked exceptions 

 were at Gjesvair and Vardo, stations to the north 

 of 70° and far to the east. In no month did the rain- 

 fall show a general excess or defect over the whole 

 country, although March and August were dry, and 

 January wet nearlv everywhere. In February, June, 

 and October to December there was a pronounced 

 tendency to rainy _ conditions in the south, while a 

 drought was experienced in the north of the country. 



1 Nedb<|.riaattagelser i Norge, utgit av Det Norske Meteorologiske 

 Institut. Aargang xxi., 1916. 



