724 



SCIENCE. 



[N. S. Vol. III. No. 73. 



of planetoids seems to be untenable. (3) The 

 hypothesis of an extended mass of diffused 

 matter like that which reflects the zodiacal 

 light has insurmountable difficulties. (4) 

 The hypothesis of a ring of planetoids be- 

 tween the orbits of Mercury and Venus is 

 very unsatisfactory. 



N'ewcomb finally regards Prof. Hall's 

 hypothesis as not inadmissible. This hypo- 

 thesis is a startling one, no less than that 

 gravitation toward the sun is not exactly 

 as the inverse square of the distance. Prof. 

 Paul Harzer has recently published a 

 memoir dealing with this subject, which 

 obtained the prize of the Jablonowski 

 Society. Harzer is disposed to attribute 

 the greater motion of Mercury's perihelion 

 to an irregular distribution of the sun's 

 mass within its surface, admitted to be 

 spherical ; being denser in the parts near 

 the solar equator. He appears to think 

 the solar corona may have something to do 

 with it. Harzer's theory seems to have the 

 advantage over ISTewcomb's modification of 

 Newton's law in that it leaves the latter 

 intact. 



Newcomb considers that, with the ex- 

 ception of the motions of the moon and of 

 Mercury, " all the motions in the solar 

 system, as far as known, agree perfectly 

 with the results of the theory of gravitation. 

 The little imperfections which still exist in 

 the astronomical tables seem to proceed 

 mainly from errors in the data from which 

 the mathematicians must start in com- 

 puting the motion of any planet. The time 

 of revolution of a planet, the eccentricity of 

 its orbit, the position of its perihelion, and 

 its place in the orbit at a given time, can 

 none of them be computed from the theory 

 of gravitation, but must be derived from ob- 

 servations alone. If the observations were 

 absolutely perfect, results of any degree of 

 accuracy could be obtained from them ; but 

 the imperfections of all instruments and even 

 of the human sight itself prevents obser- 



vations from attaining the degree of pre- 

 cision sought after by the theoretical as- 

 tronomer and make the consideration of 

 ' errors of observation ' as well as ' errors 

 of the tables ' constantly necessarj\" 



One of the most important and interest- 

 ing investigations going on now deals with 

 the subject of variation of latitude. Certain 

 theoretical considerations led the astron- 

 omers fifty years ago to look for changes of 

 latitude which showed a period of 305 

 days. Maxwell and Bessel examined the 

 matter and Bessel found that his lati- 

 tude diminished 0".3 in two years (1842). 

 Other observations at various places 

 showed aparent changes of small amounts. 

 The results for several reasons that 

 then appeared sound were not regarded 

 as satisfactory, so that it was doubted 

 by many that any measurable vari- 

 ation of latitude would be found. The 

 problem assumed a new aspect, however, 

 when Dr. Kiistner, of Berlin, published 

 the results of his observations made in 

 1884-85. These results showed unmistaka- 

 bly that a small but quite a I'apid change 

 had occurred in the latitude of Berlin, 

 amounting to from 0".2 to 0".3. The ex- 

 amination of other observations showed 

 similar results. 



A crucial test was made by sending an 

 expedition to the Sandwich Islands, which 

 is 180 degrees (nearly) in longitude from 

 Berlin. If, it was known, the latitude of 

 Bei"lin increased, then a point in the north- 

 ern hemisphere, 180 degrees away from 

 Berlin, should simultaneously show a de- 

 crease in latitude, for if the pole moves 

 toward Berlin it must move from the point 

 on the other side of the earth. Our own 

 government joined in the effort. Marcuse, 

 of Berlin, and Preston, of Washington, 

 spent more than a year on the Sandwich 

 Islands observing for latitude, while at the 

 same time observations were continued at 

 Berlin, Prague and Strasburg, in Europe, 



