Febeuaby 11, 1910] 



SCIENCE 



231 



records the strongest diurnal deflection. Some 

 kind of solar control seems necessary to ex- 

 plain these movements. The method of its 

 action has not been made out. 



Type 2. — Correlation of the movements of 

 the pendulum with the movements of areas of 

 low and high barometric pressure across the 

 United States and southern Canada, shows an 

 intimate relation between them. An easterly 

 deflection of the E.-W. component begins 

 when an area of low pressure appears in some 

 westerly or southwesterly direction from the 

 station. The cyclone may be even 1,500 or 

 more miles away when the deflection begins. 

 The time of beginning seems to depend partly 

 on the movements of a high-pressure center to 

 the east, though such a relation can not be 

 definitely worked out until more complete 

 knowledge of conditions over the Atlantic 

 Ocean is available. There appears to be at 

 least a general relation between the amount 

 of pressure at the center of the cyclone, the 

 area covered by it, the rapidity of its move- 

 ment, and the time of beginning and the 

 rapidity of easterly travel of the pendulum. 

 As the depression moves east or northeast, the 

 pendulum also moves toward the east until the 

 cyclone is nearly over the station, and as the 

 depression passes off the coast, the pendulum 

 begins to travel toward the west. Inspection 

 of the current weather map shows an area of 

 high pressure, or one of less intense low pres- 

 sure, ' than that which caused the deflection, 

 to be approaching easterly in the western 

 quadrant. As the anticyclone comes nearer, 

 the westerly travel usually increases in rapid- 

 ity. When the center is approximately over 

 the station, the direction of travel is reversed 

 and the cycle repeated. These deflections 

 occupy any length of time, dependent wholly 

 on the time taken for the passage of the cy- 

 clone or anticyclone. They often begin many 

 hours before the barometer indicates the ap- 

 proach of minima or maxima. They do not go 

 on uninterruptedly; there are countless minor 

 variations the causes of which it is as yet im- 

 possible to determine. The diurnal deflection 

 is. superimposed on these longer, non-periodic 

 deflections. 



The N.-S. Component: Type 1. — A diurnal 

 deflection. This is indicated by a more or less 

 strong tendency of the pendulum to move 

 south during the forenoon, and north later in 

 the day. It is much less clearly shown than 

 the diurnal of the E.-W. component, and is 

 apparently dependent on the same causes. 



Type 2.- — ^Deflections cyclonically or anti- 

 cyclonically controlled. These include all 

 movements due to the approach of high or 

 low pressure areas from some westerly direc- 

 tion. They are somewhat less frequent, and 

 usually much less marked, and their period of 

 maximum activity is nearly always much 

 shorter than is the case with the deflections 

 of the E.-W. component. This is apparently 

 due to the parallelism of the IST.-S. component 

 to the mean cyclonic and anticyclonic tracks. 

 The approach of a high from the northwest 

 and its passage north of the station, or the 

 approach of a low from the south or south- 

 west and its passage south, is accompanied by 

 a northward deflection of the pendulum. This 

 reaches its maximum when the pressure gra- 

 dient runs due north, and becomes a southerly 

 deflection when pressure conditions are re- 

 versed. Often interrupting these deflections 

 are temporary movements for a few hours in 

 a contrary direction, followed by the renewal 

 of the long-period travel. These variations 

 do not affect the general tendency, and their 

 causes have not been made out. The diurnal 

 deflection is superimposed on these longer, 

 non-periodic deflections. 



The summary presented in the table below 

 shows for each component the per cents of 

 cases (on the basis of numbers of days out of 

 the total) in which the particular deflection 

 occurred. A more desirable basis would be 

 units by cyclones and anticyclones; but the 

 variability of the time taken for the passage of 

 these areas by the station, their complex dis- 

 tributions, and the impossibility of evaluating 

 the share that each has in producing a given 

 deflection, makes it impracticable to deter- 

 mine the limits of any one unit. The per 

 cents, of movements from lows toward highs 

 are considerably smaller than they would be 

 if computed on the latter basis, for they take 



