DEPARTMENT OP TERRESTRIAL MAGNETISM. 331 



experienced. Numerous experiments were made during the search for more 

 suitable ink or paper to improve the clearness and shaipness of the record; 

 thus it was attempted to adapt the principle of the thread-recorder as made 

 by the Cambridge Scientific Instrument Company, and also to replace the 

 ink by singeing the paper at each contact by means of an electric current and 

 a small V-shaped loop of fine platinum wire attached to the end of the "pen." 

 The latter method promised excellent results, but the experiments were dis- 

 continued when the supplj'- of new paper was secured and found so satisfactory. 

 The treatment of the theory of the instrument was extended, and as a 

 result of the new discussion it was found possible to make the scale-value 

 more unifonn throughout the width of the record. In the original form of 

 the instrument, the magnet-lever was vertical and the pen-lever was hori- 

 zontal; thus while the length of the magnet-arm was constant, the length of 

 the pen-ann became longer as the deflection from the axial line of the instru- 

 ment increased, and the scale-value was greater at the edges of the paper 

 than in the center. This difference, which amounted to several per cent, 

 was materially reduced by inclining each of the two levers at an angle of 45° 

 with the horizon. In that case, as the changing declination turns the pen- 

 arm from its mean position, the effective lengths of both levers increase and 

 hence the ratio of the angular displacement of the magnet to the angular 

 displacement of the pen, {. e., the scale-value, is more nearly uniform. It is 

 even possible by carefully choosing the two angles of inclination and bj' 

 curving one or both of the levers to make the scale-value perfectly uniform 

 across the paper; that degree of refinement is, however, unnecessary and 

 would be inadvisable owing to the probable increase in friction. 



Some notes on the occurrence of thunder at sea. W. J. Peters. Terr. Mag., vol. 21, pp. 

 21-22 (Mar. 1916). Washington. 



Baron von Humboldt appears to be responsible for the statement that 

 thunder is never heard on the ocean at any great distance from land, though 

 violent thunderstorms are often observed at sea and vessels are frequently 

 struck by lightning. Since this statement has provoked discussion from time 

 to time, observations are being made aboard the Carnegie, under J. P. Ault's 

 command, in accordance with directions issued to the vessel by the Director 

 of the Department of Terrestrial Magnetism. The present paper summarizes 

 the results of the observations made on the voyage from Dutch Harbor, 

 Alaska, to Port Lyttelton, New Zealand, between August 6 and November 2, 

 1915. Final conclusions are deferred pending receipt of additional data. 



The normal electric field of the Eartli. W. F. G. Swann. (Presented before the Philo- 

 sophical Society of Washington, October 2, 1915.) 



The paper comprised: (1) A general review of the principal atmospheric- 

 electric phenomena; (2) a discussion of the extent to which the known causes 

 contributing to atmospheric ionization are sufficient to account completely for 

 such ionization ; (3) a discussion of the origin and maintenance of the Earth's 

 charge. 



The chief sources of atmospheric ionization in the lower atmosphere arc 

 the radioactive materials in the soil and atmosphere, and the penetrating 

 radiation. Over the sea, the radioactive material is insufficient to account 

 for an appreciable fraction of the ionization observed there. If, however, we 

 are correct in assuming that an appreciable fraction of the ionization pro- 

 duced in a closed vessel is produced as a result of actions other than that of 

 the vessel itself, the cause responsible for such ionization is amply sufficient 



