ASTRONOMY: A. VAN MA AN EN 
189 
ing the time when the dome is open, the mirror occasionally shows as- 
tigmatism. Under such conditions the images are more or less asym- 
metrical and the measures may not be as accurate as usual. This hap- 
pened with three of the exposures on the last star given in the table 
below. 
The result of the investigation has been rather promising; 5 stars 
have been finished, the results of which are given in the table. 
Number of 
Name of Star tt P.E. Exposures 
BossP. G. C. 96 +0^026 0\007 14 
672 - 0 . 009 0 . 0045 14 
1549 + 0 .001 0 .001^ 16 
2921 +0 .078 0 .006 10 
3233 + 0 .003 0 .010 12 
For the mean of 13 exposures the mean probable error of a parallax 
is not quite 0".006. If we compare this result with the best known 
photographic determinations of parallaxes listed below, it will be seen 
that we have gained considerably. 
Number of 
Instrument Observer Mean P.E. Exposures 
{Dormer 
Kapteyn 0^.023 12 
de Sitter 
(Kiistner 
Kapteyn 0^.020 12 
de Sitter 
Cambridge, England (12 . 5-inch) ^ Russell O'' .028 27 
Yerkes (40-inch)4 Schlesinger 0^.013 37 
— (40-incH. " 
— (40-.C.). - 
Swarthmore (24-inch)7 Miller 0^.011 Not published 
iGroningen, Pub. Astr Lab., 20, 28; 1908. ^ Astro phys. J. 38, 25; 1913. 
^Ibid., 23, 56; 1909. ^ Pub. Astr. Astrophys Soc Amer., 16th Meeting, 
3 Pub. Carnegie Inst Wash., 147, 65; 1911. 19; 1914. 
* Astrophys. J., 34, 27; 1911. ^ Ibid., 17th Meeting, 1914. 
The material is of course insufficient to yield any information as to 
possible systematic errors, but we have one evidence that these can not 
be very large; the mean parallax is 4-0'^020, while according to Kap- 
teyn's table for stars of this type, magnitude, and proper motion it is 
+0".014. 
