RECENT ADVANCES IN SCIENCE 533 



and formed a substantial advance on the knowledge then 

 available as to the angular diameter of a star. 



The method recently employed with success at Mount Wilson 

 was indicated by Michelson in a paper in the Phil. Mag. in 1 890. 

 By the ordinary interference method, already explained in these 

 notes, in which two slits are placed in front of the telescope, 

 the limiting angular diameter which can be measured is exactly 

 equal to the theoretical resolving power of the telescope. In 

 the case of the 100-inch telescope this is about o"*o5, so that 

 the angular diameters of the stars are just beyond its reach, 

 Michelson 's suggestion was to use a refractometer in which the 

 two mirrors could be separated by any desired amount, which 

 is in effect equivalent to using a telescope whose aperture is 

 equal to the distance between the mirrors. Full details of the 

 methods employed at Mount Wilson have not yet been pub- 

 lished, but from the information available at the time of 

 writing the following would appear to be correct in principle : 

 a steel girder, 20 feet in length, is fixed across the upper end 

 of the tube of the loo-inch telescope in such a manner that 

 it can be rotated about the axis of the tube. On the girder 

 are fixed two mirrors, equidistant from its centre, and each 

 inclined at an angle of 45° both to the girder and to the axis 

 of the tube. These mirrors can be moved along the girder in 

 such a manner that they always remain equidistant from its 

 centre. The light from a star is reflected by them towards 

 the centre of the girder, where two other mirrors reflect the 

 two beams down the tube of the telescope on to the 100-inch 

 mirror, which reflects them and brings them together in the 

 focal plane of the eyepiece, where they interfere with one 

 another and produce fringes. The distance apart of the mirrors 

 is varied until the fringes vanish. The girder is then rotated 

 into a different position, and, if the fringes still vanish, it is 

 evidence that their vanishing is not due to the star under 

 observation being a double star, in which case the measures 

 would merely give the separation of the components. From 

 the distance apart of the mirrors when the fringes vanish {d), 

 it is possible to compute the angular diameter of the star from 

 the formula a = i'22\jd, X being the mean wave-length of the 

 light employed. There are practical difficulties incidental to 

 the use of so large an interferometer which make the observation 

 a delicate and difficult one : it is stated that, after the mirrors 

 are shifted, it requires about one hour to rediscover the fringes. 

 The observers at Mount Wilson are therefore to be congratulated 

 on having succeeded in measuring the angular diameter of 

 a Orionis (Betelgeuse), the value obtained being o"'042. The 

 least angular diameter measurable with the present arrange- 

 ment is about o"-o2, though it is probable that before long it 



