The annual amount of the secular variation in the remote parts of the Archipelago is not positively known; the low values of easterly declination found in March, 1904, in the northern part of Sumatra, make it obvious that hypothetical values are not to be relied upon. The best way to eliminate, as far as possible, this uncertainty is to obtain values of the magnetic elements in the remoter parts of the Archipelago in one and the same year. For this reason also the northeastern and southern boundaries of the region to be surveyed were likewise visited during 1904. Besides two trips, respectively, to Engano in June, and to Christmas Island in July, a great number of stations in Celebes, the Molluccos and northern New Guinea have been visited during the months August, September, and October of that year.