The intercomparisons of magnetic instruments made by the various investigators during the past fifteen to twenty years have served to call attention to the discrepancies existing between the determination of the magnetic elements by various observatory and magnetic survey instruments. These discrepancies frequently exceed the errors of observation. At the recent meeting of the International Meteorological Conference held at Innsbruck in I9o5, renewed consideration was given this important

more »

... by its Committee on Terrestrial Magnetism. The extensive magnetic operations of the Carnegie Institution of Washington have necessitated giving the adoption of an "International Standard" correct to within at least the allowable error of field observations early consideration. In the course of the work thus far executed, comparisons have been obtained between the following magnetic observatories' Cheltenham t3IaD'land), Toronto -Agincourt (Canada), Baldwin (Kansas), Mexico (SIexico), Havana (Cuba), Sitka (Alaska), Honolulu (Hawaii), Apia (Samoa), Tokio (Japan" Zi-ka-wei ½China), Hongkong (China), Christchurch ½New Zealand, Sydney and Melbourne (Australia). Further comparisons will be obtained in •9o8-9 in Europe, Africa and Asia in the course of magnetic work to be executed during this period. Combining all the comparisons made up to date by various invest/gators and organizations, for which the data are at present on hand, it would appear possible to come to some preliminary decision as to the approximate corrections of some of the observatory standards upon an "International Standard." For example, tabulating the results of the horizontal intensity comparisons made since •897, it is found that the Potsdam standard invariably is /'o:i' on Kew, Parc St. 5Iaur, Pawlowsk, Rome, and Pola, by amounts varying from 5 On the average, of the observatories named, the values of horizontal intensity, Jar, observed with the Potsdam Standard would require to be tit is not correct to assume that the intensity correction of a magnetometer, expressed in absolute units, will remain the same with change of magnetic field, the amount of the correction depending, in fact, upon the absolute value of the intensity at the place of observation. From whatever source the correction arises, it can be expressed, with close approximation, by a simple ratio change. z. e., a factor multiplied into the first power of the value of the intensity' only in certain extreme cases will a second term involving the second power of the intensity enter in appreciably, 3 •6•