WEATHER BUREAU. 73 



Special work carried out by the library during the year includes the 

 preparation, at the request of Mr. Melvil Dewey, of a scheme of classi- 

 fication for meteorology to be used m connection with his decimal 

 classification; also the compilation of a brief bibliography on weather 

 forecasting for use in the forthcoming forecasting manual. 



The number of promotion-examination papers rated during the 

 fiscal year was 80, of which 64, or 80 per cent, attained passing grades. 



SEISMOLOGY. 



It is generally recognized that the systematic collection and 

 arrangement for convenient reference of earthquake data serve 

 several useful purposes. As the data are accumulated and earth- 

 quake regions more minutely mapped the engineer is correspondingly 

 able to distinguish with greater certainty between safe and dangerous 

 localities for bridges, dams, aqueducts, or any other important 

 structures. A knowledge of the magnitude, rapidity, and other 

 peculiarities of earth movements is of especial value in the designing 

 of structures of whatever kind to be erected in earthquake regions. 

 Finally, detailed data concerning earthquake motions serve a very- 

 valuable purpose in the science of geophysics. 



To further these important studies as much as possible, the 

 Weather Bureau began on December 1, 1914, to collect nonmstru- 

 mental reports of earthquakes from all its regular stations, nearly 200 

 in number, and also from nearly all its 4,500 cooperative observers. 

 These data are pubhshed month by month in the Weather Review. 



In addition, the Bureau operates two seismographs — a Marvin 

 vertical pendulum seismograph, giving both horizontal components, 

 at Washington, D. C, and a two-pendulum Bosch-Omori instrument 

 at Nortlifield, Vt. The Washmgton records were resumed in October, 

 1914, and those of Nortlifield, Vt., began in December, 1914. 



Begimiing with January 1, 1915, the Bureau has also collected and 

 published as far as possible instrumental records of earthquakes 

 obtained at — 



Sitka, Alaska, United States Coast and Geodetic Survey. 



Tucson, Ariz., United States Coast and Geodetic Survey. 



Honolulu, Hawaii, United States Coast and Geodetic Survey. 



Cheltenham, Md., United States Coast and Geodetic Survey. 



Porto Rico, United States Coast and Geodetic Survey. 



Point Loma, Cal., Raja Yoga Academy. 



Denver, Colo., Sacred Heait College. 



Georgetown, D. C, Georgetown University 



Lawrence, Kans., University of Kansas. 



Cambridge, Mass., Harvard University. 



St. Louis, Mo., St. Louis University. 



Buffalo, N. Y., Canisius College. 



Fordham, N. Y., Fordham University. 



Balboa Heights, Panama Canal Zone. 



Ottawa, Canada, Dominion Observatory. 



Toronto, Canada, Dominion Meteorological Service. 



Victoria, British Columbia, Dominion Meteorological Service. 



AEROLOGICAL INVESTIGATIONS. 



Observations of the diurnal convective system have heretofore 

 been made by means of kites and have necessarily been limited to 

 heights of 3 to 3.5 kilometers. It seemed desirable to have a 24-hour 



