Parallaxes of ft, and e Cassiopeiae. 11 



corresponding to a motion of 3". 784 upon a great circle whose 

 position angle is 113° 59'. The following are therefore the several 

 values of the proper motion, to be compared with each other: 



As just obtained 3-784 



Auwers'-Bkad. (1810), as used 



in the present paper . . . 3729 



Auweks'-Bkad., reduced to 1S72 3.729 



When we compare the parallax of fx, Cassiopeise derived in the 

 present paper, with the work of other observatories, we find large 

 discordances. Thus the Oxford photographic result is only about 

 o" 036 =1= o".oi8, while the Rutiierfurd plates give ©".249 ± ©".045 

 from the same pair of comparison stars. On the other hand, Struve 

 has obtained o".25i =b o".o75 from distance measures, and from 

 position angles ©".425 d= o".o'j2. It is therefore plain that the 

 photographic method of determining parallaxes cannot be regarded 

 as free from systematic error. An examination of the equations of 

 table Y. shows that negative parallax coefBcients invariably occur 

 in the case of plates exposed at eastern hour angles. This circum- 

 stance, which arises from the inconvenience of observing after mid- 

 night, may possibly produce systematic error. But the evidence 

 of the scale value table (IV. A) is against this supposition, as is 

 also the approximate equality of the parallaxes obtained from pairs 

 of comparison stars having widely different distances from ft. 



In conclusion, the results here deduced may be summed up as 

 follows: — 



Parallax of jit Cassiopeise 0.375 ± 0.024 

 Parallax of d Cassiopeias 0.233 ± 0.067 



But the above probable errors must not be taken as reliable esti- 

 mates of uncertainty, since a comparison with the work of other 

 astronomers seems to indicate the possibility of systematic error. 

 But if we are willing to accept the above results, it is perhaps 

 allowable to speculate upon fx, and Cassiopeias as a system re- 

 motely resembling that of 61 Cygni. The indication of equality 

 of parallaxes furnished by Bessel's observations, and the slight 

 evidence of variation in the proper motion of fi Cassiopeias obtained 

 from the equations on p. 10, would almost seem to favor such an 

 idea. 



