B. TERRESTRIAL PHYSICS AND METEOROLOGY. 67 



ABNORilAL DEFLECTIONS OF THE PLUMB-LINE. 



Lieutenant F. V. Greene, Engineer Corps, U. S. A., has re- 

 cently discussed the station-errors of the 41 astronomical 

 stations established on the 49th parallel of latitude by the 

 joint commission of English and American engineer officers, 

 for the purpose of marking the boundary-line between the 

 United States and the British possessions. The 41 stations 

 were all observed with the zenith telescope, and of them 19 

 were observed by the English alone, and the average prob- 

 able error of the final latitude of a station was 0.088" ; 17 

 were observed by the Americans alone, and the probable 

 error of a station established by them was 0.059" (76 ob- 

 servations) ; and 5 were jointly observed by the two parties, 

 their independent results differing by 0.28", 0.27", 0.07", 0.29", 

 and 0.31" respectively. The various stations were connect- 

 ed geodetically by the method of tangents and offsets ; and 

 the difference between the astronomical and creodetic de- 

 terminations was assumed to be due to abnormal variations 

 in the local directions of gravity, and was called " station- 

 error." 



Sufficiently accurate topographical work was done to ena- 

 ble 200-feet contours to be laid down on the maps. From 

 these contours Lieutenant Greene has calculated A (see 

 table), which is the deflection of the plumb-line produced by 

 all masses above the ground within a radius of ten miles, 

 and B, which is the deflection which would be produced by 

 masses between ten and sixty miles distant. This has been 

 done with reference to each of the 41 stations a Avork of no 

 little difficulty and the results are given in the accompany- 

 ing table. D is the observed deflection, and D (A+B) is 

 the outstanding residual and unexplained deflection. These 

 quantities for 37 stations in the Ordnance Survey of Great 

 Britain were D=:2.05", and D-(A4-B) = 1.35". From this 

 it appears that for 29 stations the abnormal phenomena are 

 partially accounted for and in the right direction; for 12 

 stations the computed deflection is in the wrong direction. 

 The quantity D 2.15" which is to be accounted for is about 

 the same in magnitude as the mean error in latitude (2.64") 

 found by Bessel from a discussion of all the great arcs pre- 

 viously measured, and goes to show that the mean hetero- 



