An Early Criticism of the Ahhe Theory. 267 



The conditions of illumination were in fact such — the source of 

 li,i>'ht being focused on the stage of the instrument — that even the 

 real diffraction grating showed no spectrum. For diffraction only 

 occurs when the same wave- front passes through two or more 

 apertures of a grating. It only occurs, therefore, when the incident 

 wave-front has a finite magnitude, and, consequently, if the grating 

 is itself the source of light, or if it coincides with any plane in 

 which an image of the source of light is focused, no diffraction 

 spectra can arise from it. This is not only a theoretical result : 

 it can be very easily verified to the eye by removing the ocular 

 from the Microscope when a diffraction grating is on the stage of 

 the instrument, and inspecting the back focal plane in which these 

 spectra are formed. If now, by moving the condenser, the image of 

 the source of light be brought nearer and nearer to the plane of the 

 grating, it will be found that, as the radius of the incident wave- 

 front shortens, the spectra crowd closer and closer together, and 

 when at last the image of the light source coincides with the 

 grating, these spectra have all disappeared, being absorbed in the 

 central image with which they all now coincide. This phenomenon 

 was very fully explained in my paper on the Abbe theory (p. 366), 

 and was illustrated experimentally at the meeting at which the 

 paper was read. 



We are now informed that this experiment, or something more 

 or less resembling it, was put forward by Professor Altmann. How 

 close the resemblance was we cannot well judge, for we are not 

 told whether Professor Altmann made it a condition of his 

 experiment that the source of light should be focused in the plane 

 of the original grating. It would indeed be most natural to 



EXPLANATION OP PLATE XIII. 



Fig. 4. — Contains pliotographs of the various images formed by the grating (an 

 Abbe diffractionsplatte) of the source of light when the position of this 

 latter is varied, as stated in the diagram. It is to be observed that as the 

 distaiace between the source of light and the grating gets shorter, the spectrum 

 becomes more and more compressed ; that when it reaches a point in the tube 

 at which the pencils from various openings in the grating separate out from 

 one another, the spectrum (at a distance of 0*002 in.) loses the character of 

 a grating spectrum, and conforms to the type of spectrum formed by a single 

 opening. At this jjoint Abbe's rule that two spectra or more are required for 

 the formation of a resolved image, ceases to hold good, since the central beam 

 of the spectrum formed by a single opening is alone sufficient for the purpose. 

 When the source of light coincides with the grating, there is an image of 

 the one superimposed upon an image of the other, but no spectrum of any 

 sort 



[To prevent a mistake which appears to be possible, it should be added, that 

 in these experiments the Abbe slit cannot be replaced by a lamp-flame arranged 

 "end on" towards the ilicroscope. The difficulty in that case is that the flame 

 is so extended along the optical axis, that wherever it may be placed some part 

 of its image will be far enough from the grating to give rise to a spectrum of 

 the grating type.] 



T 2 



