jVA rURE 



[August 15, 1901 



face subsequently to the grinding. But experiment is not limited 

 liy the capabilities of the eye ; and it seems certain that a finely 

 ground surface would be smooth enough to reflect without 

 sensible diffusion the longest waves, such as those found by 

 Rubens to be nearly loo times longer than the waves of red light. 

 An experiment may be tried with radiation from a Leslie cube 

 containing hot water, or from a Welsbach mantle (without a 

 chimney). In the lecture the latter was employed, and it fell 

 first at an angle of about 45' upon a finely ground flat glass 

 silvered in front. By this preliminary reflection, the radiation 

 was purified from waves other than those of consideralile wave- 

 length. The second reflection (also at 45°) was alternately 

 froi'ii polished and finely ground silvered surfaces of the same 

 size, so mounted as to permit the accurate substitution of the one 

 for the other. The heating-power of the radiation thus twice 

 reflected was tested with a thermopile in the usual manner. 

 Repeated comparisons proved that the reflection from the groiind 

 surface was about 076 of that from the polished surface, showing 

 that the ground surface reflected the waves falling upon it with 

 comparatively little diffusion. A slight rotation of any of the 

 surfaces from their proper positions at once cut oft the eftect. 

 It is probable that the device of submitting radiation to prelim- 

 inary reflections from one or more merely ground surfaces 

 might be found useful in experiments upon the longest waves. 



the emery is hacked by a hard surface, e.g. of glass, while 

 during the polishing the powder (mostly rouge in these experi- 

 ments) is imbedded in a comparatively yielding substance, such 

 IS pitch. Under these conditions, which preclude more than 

 a moderate pressure, it seems probable that no pits are formed 

 by the breaking out of fragments, but that the material is worn 

 away (at first, of course, on the eminences) almost molecularly. 



The progress of the operation is easily watched with a micro- 

 scope, provided, say, with a |-inch object-glass. The first few 

 minutes suffice to effect a very visible change. Under the micro- 

 scope it is seen that little facets, parallel to the general plane of 

 the surface, have been formed on all the more prominent 

 eminences.' The facets, although at this stage but a very 

 small fraction of the whole area, are adequate to give a sensible 

 specular reflection, even at perpendicular incidence. On one 

 occasion five minutes' polishing of a rather finely ground glass 

 surface was enough to qualify it for the formation of interfer- 

 ence bands, when brought into juxtaposition with another 

 polished surface, the light being either white or from a soda 

 ffame ; so that in this way an optical test can be applied almost 

 before the polishing has begun. - 



As the polishing proceeds, the facets are seen under the 

 microscope to increase both in number and in size, until they 

 occupy much the larger part of the area. Somewhat later the 



Fig. 2. 



In view of these phenomena we recognise that it is something 

 of an accident that polishing processes, as distinct from grinding, 

 are needed at all ; and we may be tempted to infer that there is 

 no essential diflerence between the operations. This appears to 

 have been the opinion of Herschel,' whom we may regard as 

 one of the first authorities on such a subject. But although, 

 perhaps, no sure conclusion can be demonstrated, the balance 

 of evidence appears to point in the opposite direction. It is 

 true that the same povvders may be employed in both cases. In 

 one experiment a glass surface was polished with the same emery 

 as had been used efiectively a little earlier in the grinding. 

 The difference is in the character of the backing. In grinding, 



The intensity and regularity 



im is found to depend, not 



but very es<;eiuially on the degree of 



But it may reasonably be asked how 



in a surface polished by art, when we 



in f.ict, nothing more than grinding 



s by the use of hard gritty powders, 



of' mechanical comminution we may give them, 



mparison with the ultimate molecules of matter, 



ly be considered as an irregular tearing up by the 



"n the surface. So that, in fact, a 



ewhat of the same kind of relation 



1 "Enc. Met.," .\rt. Light 



of reflection at the external surface of 

 merely on the nature of the medi 

 smoothness and pnlish of its surfai 



I take pi; 

 ; of polishing 



any regular reflect! 



recollect that proct 



down large asperit 



which, whatever d 



are yet vast masses 



and their action ca 



roots of every projection that 



surface artificially polished must bear 



>the 



NO. 1659, VOL. 64] 



parts as yet untouched by the polisher appear as pits, or spots, 

 upon a surface otherwise invisible. Fig. I represents a photo- 

 graph of a surface at this stage taken with the microscope. The 

 completion of the process consists in rubbing away the whole 

 surface down to the level of the deepest pits. The last part of 

 the operation, while it occupies a great deal of time and entails 

 further risk of losing the " truth " of the surface, adds very little 

 to the effective area or to the intensity of the light regularly 

 reflected or refracted. 



Perhaps the most important fact taught by the microscope is 

 that the polish of individual parts of the surface does not im- 

 prove during the process. As soon as they can be observed 

 at all, the facets appear absolutely structureless. In its subse- 

 quent action the polishing tool, bearing only upon the parts 

 already polished, extends the boundary of these parts, but does 

 not enhance their quality. Of course, the mere fact that no 

 structure can be perceived does not of itself prove that pittings 

 may not be taking place of a character too fine to lie shown by 



1 The interpretation is facilitated by a thin coating of aniline dye which 

 attaches itself mainly to the hollows. 



- With oblique incidence, as in Talbot's experiments (see /Vi;7. Mag., 

 xxviii. p. igi, 1SS9), achromatic bands maybe observed from a surface 

 absolutely unpolished, but this disposition would not be favourable for 

 testing purposes, 



