RECENT ADVANCES IN SCIENCE i8i 



If, now, a narrow slit be placed over the objective, the 

 diffraction phenomena are entirely modified and a series of 

 diffraction fringes parallel to the edges of the slit are formed. 

 By using two parallel slits, the diffraction pattern is modified 

 according to the distance apart of the slits, and it is this fact 

 which proves to be of value in astronomical observation. 

 The theory of the phenomena which then result has been dis- 

 cussed by Michelson and also by Hamy. We need consider 

 only the two most important cases : 



(i) If a circular object of uniform brightness with a finite, 

 but small, angular diameter is observed, a series of fringes is 

 obtained; but if the separation of the slits is gradually in- 

 creased, the contrast between the fringes becomes less until a 

 certain stage is reached at which their visibility becomes a 

 minimum. The angular diameter of the object (a) is then 

 connected with the distance apart, d, of the slits by the rela- 

 tionship a = 1-22 \/d. The distance apart of the slits corre- 

 sponding to minimum visibility can be determined with great 

 accuracy, the phenomenon being very sharply marked. A 

 method of determining the angular diameters of satellites, etc., is 

 thus obtained which does not suffer from the causes of error 

 incidental to the method of direct observation. It has been 

 utilised to measure the diameters of the larger satellites of 

 Jupiter by Hamy, using a 12-inch equatorial at the Paris 

 Observatory, and by Michelson, using an equatorial at the Lick 

 Observatory, in each case in the year 1891 and with very 

 concordant results. It is not necessary that the slits should 

 be placed over the object glass : they can be anywhere in the 

 cone of rays, the corresponding separation being proportional 

 to their distance from the focal plane. 



Michelson has shown how, by using an interferometer to 

 increase by a known amount the relative retardation between 

 the rays from the two sHts, the effective aperture of the tele- 

 scope can be considerably increased. It does not seem beyond 

 the bounds of probability that by this means, with the 100- 

 inch telescope at Mount Wilson, it will be possible to measure 

 the angular diameters of the nearer stars. It is not improb- 

 able that some of these stars have angular diameter of o^-oi, 

 which quantity should be measurable. 



The disadvantage of this method is that the amount of 

 light utilised is only a small fraction of that transmitted by 

 the object glass. Hamy has considered how this defect may 

 be avoided. The method is still vahd if, instead of narrow 

 slits, openings having two perpendicular axes of symmetry 

 are used, one of which coincides with the line joining their 

 centres, and provided that the width of the openings is small 

 compared with their distance apart. Hamy proved that the 



