286 Transactions of the Society. 



1. A radiant point is never brought to focus in the image as a 

 point, but always as a diffraction figure — conveniently styled by 

 Mr. Gordon an " antipoint." 



2. Where a radiant point is viewed through a circular aperture 

 that antipoint consists of a central false disc, surrounded by a 

 system of alternately dark and bright rings. 



We may for our particular purposes leave out of consideration 

 all but the innermost of these rings. 



3. When a point is viewed through a slit opening the anti- 

 point obtained corresponds to an optical section of the antipoint 

 referred to in (2). It consists, in other words, of a dash — the 

 optical section of the false disc— flanked on either side by a faintly 

 luminous point, corresponding in each case to the optical section 

 of the first bright ring. 



4. Where a diffraction grating takes the place of the simple 

 slit aperture, the outlying elements of the antipoint are emphasized 

 at the expense of the central elements. We obtain, in other words, 

 as the antipoint of a point, a less conspicuous central dash flanked 

 on either side by a brighter point. 



5. When a line, or file of points, is viewed through a diffraction 

 grating we obtain a composite antipoint pattern, consisting of a 

 central or principal line furnished on either side by a flanking 

 line. 



6. The elongation — meaning thereby the distance between the 

 principal line and flanking line — is determined (a) by the periodical 

 interval of the diffraction grating, and (b) by the distance between 

 the aperture of the lens and the screen upon which the image is 

 brought to focus. 



Having recalled to mind these preliminary points, I am 'in a 

 position to make my suggestion intelligible. This suggestion is 

 that we should use the elongation of the flanking line as a mea- 

 suring staff for the determination of the distance at which the 

 principal images of the two lines lie apart in the retinal image. 

 We can use this measuring staff to advantage, in particular, in the 

 case where the elongation of the flanking line corresponds exactly 

 to half the distance between the principal lines. When this con- 

 dition is fulfilled, the flanking lines which fall into the interspace 

 between the principal lines merge and furnish a single well-marked 

 intercostal line. 



Having called your attention to the fiduciary phenomenon 

 which is furnished under these conditions, it will be manifest to 

 you upon consideration that we can exploit this phenomenon in 

 the following manner. 



We can place before us at a distance of 10 in. a series of 

 paired lines ruled at progressively increasing intervals, and view- 

 ing them through a diffraction grating held, let us say, with its 

 rulings parallel to the object lines, select that particular pair of 



