436 DOOLITTLE — VARIATION OF TERRESTRIAL LATITUDE. [Oct. 11, 



from various series of observations made at different places and by 

 different methods, began to awaken a widespread interest in the 

 subject. That the changes were real admitted of but little doubt. 

 It was equally obvious that they could not be fitted into the ten- 

 month period of Euler. 



Finally, Mr. Seth C. Chandler, by a most laborious investigation, 

 embracing an analysis of many thousands of observations, extend- 

 ing over a hundred and fifty years, succeeded in solving the mys- 

 tery, at least in so far as existing evidence can solve it. The result 

 shows that the earth's axis of rotation moves about the axis of figure 

 not in one simple period of ten months, but in two periods of 

 twelve and fourteen months respectively, with perhaps a third of 

 several years. The combination of these motions results in a some- 

 what complicated curve. If we suppose circles drawn about the 

 north and south poles of the earth having diameters of about fifty 

 feet, the extremities of the axis of rotation will always be found 

 within these circles. 



For the purpose of perfecting our knowledge of these motions, for 

 obtaining data for a correct explanation of the same from a theo- 

 retical standpoint, very accurate and carefully executed series of 

 latitude observations at different points are necessary. It is such a 

 series we have undertaken at the Flower Observatory.- This series 

 is a continuation as far as may be of that which was kept up for 

 several years at South Bethlehem. 



In this work the instrument employed is the Zenith Telescope. 

 The stars observed are arranged in groups so selected that the 

 errors in the positions of the stars are almost completely eliminated 

 in so far as they effect the apparent change of latitude. Regular 

 observation at our Observatory for this purpose was begun October 

 I, 1896. Since this date observations have been made on nearly 

 every favorable night. Two series constitute a complete night's 

 work, the first in the early evening soon after dark, the second in . 

 the morning before sunrise. Each series embraces ten pairs of 

 stars, requiring about two hours of actual observation. As will be 

 seen, the work is laborious. This with the necessary numerical com- 

 putation might very well be regarded as sufficient employment for 

 one person. 



Some 1700 observations, extending from 1896, October t, to 

 1897, August 26, have been reduced so that we can form some 

 judgment of the results obtained and of the character of the work. 



