I9IS.] SCHLESINGER— VARIATIONS OF LATITUDE. 355 



certain definite purposes that they served as well as could be ex- 

 pected for their time ; but they were not intended and are not well 

 suited for precise determinations of the latitude. Close acquaint- 

 ance with the subject has taught us that exceedingly delicate ob- 

 servations are necessary to define the variations with adequate ac- 

 curacy. If I held in my hands two plumb lines half a meter apart, 

 they would not be quite parallel to each other, though both are 

 exactly vertical ; if they were prolonged, they would meet some- 

 where near the center of the earth, 4,000 miles below. The angle 

 between them is a little less than o".02 and represents approximately 

 the accuracy that is demanded and that has recently been attained 

 in latitude observations. This success is due chiefly to the Inter- 

 national Geodetic Association which has organized an " international 

 latitude service " of high efficiency, and to whose efforts and ex- 

 perience are due the improvements in instruments and methods that 

 have made possible this extraordinary degree of precision. Since 

 1899, the Association has maintained six observing stations for this 

 sole purpose, two of these being in our own country. One of the 

 minor effects of the war that is now raging in Europe will be the 

 discontinuance of some of these stations. One of the American 

 stations has already been abandoned and the same fate will over- 

 take the other in June, 1916, unless some independent means of 

 maintaining it, at least temporarily, presents itself soon. An in- 

 terruption of these observations would be a great pity, for this is 

 one of the cases where a continuous record is highly desirable. 



To return to Chandler and his work on these variations, per- 

 haps the most important of his achievements was to show that the 

 principal term in the variations, instead of having a period of ten 

 months in accordance with Euler's theory, has in reality a period 

 of fourteen months. This difference explains the failure of Bessel 

 and all the others who preceded Kiistner to find a latitude variation 

 in their observations ; for, relying upon Euler's results, they had 

 all tested their observations for the ten-month variation and had 

 sought for no other variation. For the same reason, Chandler's 

 announcement of the longer period was received with incredulity 

 in some quarters, and this feeling did not vanish until Newcomb 



