138 ASTRONOMY: E. DOOLITTLE 
weighted in proportion to the number of nights, except that at least double 
weight was always given to the measures of Aitken and Hussey with the 36- 
inch and to those of Burnham with the 40-inch. 
The approximate rectilinear path having been found, the differential cor- 
rections Av, AT and Ad to the mean motion, the time of closest approach and 
the direction, respectively, were next found by a least square adjustment, 
measures of angle, only, being employed; the distance at closest approach was 
next found from the weighted means of the measured angles and distances. 
It was assumed that the line as thus finally adjusted was the best determina- 
tion of the path obtainable upon the hypothesis that the path is rectilinear. 
The average velocity, v, along the path was finally compared with the observed 
velocity at different epochs in order to ascertain any systematic variation that 
might exist. 
It is found convenient to group the pairs examined into five classes, deter- 
mined by the results of the investigation as follows: 
Class A . — The velocity shows a decided and certain increase. Rapid 
motion in angle is to be expected, and careful measures, particularly of angle, 
are especially needed at this time. 
OS 20, (Burnham's General Catalogue, No. 479). Though for 50 years 
after discovery the velocity remained nearly constant, (about 0". 01 6), it has, 
since 1897, very decidedly increased, (to 0". 036), and orbital motion is certain. 
The period will, however, much exceed the value, 136.2 years, assigned by 
Glasenapp. 
2 208, (1074). It was found quite impossible to represent both the earlier 
and the later observations by a single rectilinear path. The velocity is now 
nearly three times the average velocity and the stars are near their minimum 
distance. 
2 963, (3625). The motion is here very slow, but there is a systematic 
increase in v, from 0".0070 in 1839 to 0".0131 in 1908. 
2 1306, (4923). The observations from 1845-65 are discordant but there 
is a remarkable increase in v during recent years. Undoubtedly binary, and 
now so near the least distance that an entry into the first quadrant may be 
expected during the next few years. The brighter component (5.0 magnitude) 
is one of Newcomb's Fundamental Stars and is contained in the American 
Ephemeris. As we, of course, know nothing of the location of the center of 
gravity between this and its 8.5 magnitude companion, it would seem that this 
is unsuitable for use as a fundamental star. 
2 1834, (6832). The distance in this pair is now but little greater than 
0'M and the measures since 1892 are hopelessly inconsistent. Analysis of the 
motion from discovery until the date mentioned shows a steady, and finally 
rapid, increase in velocity. The angle of the companion has probably increased 
180° during the past few years. 
2 113, (707) and 2 2576, (9602), also show a definite and certain increase of 
velocity in recent years. 
