380 



KNOWLEDGE. 



October. 1913 



Ilford Company, and two by the Imperial Plate 

 Company, of Cricklewood, were placed parallel, one 

 behind the other and a thick lead screen. The 

 plates were normal to the rays, but their edges were 

 at various angles to the horizontal, none of them 

 being the " way " of the plate-process. The distance 

 between the plates was a few centimetres, and that 

 of the first plate to the platinised-nickel anti-cathode 

 of a water-cooled X-ray bulb was fifty centimetres. 

 A constant current of -7 milliampere was passed 

 through the bulb, and the exposure continued for 

 eleven hours. Dark parallel bands showing a crossed 

 system appeared on all the plates, but these are not, 

 as they should be if one continuous form or structure 

 of radiations had been transmitted through the 

 series, in some one definite relation to the images 

 of three small circular apertures in triangular form 

 which had been made in the lead screen. Similar 

 bands, however, in so many other experiments have 

 been found in a definite quadrilateral pattern that 

 chance variations in the plate process or in develop- 

 ment can scarcely be the cause, though some definite 

 variations causing an inequalityof absorbing substance 

 may be. It appears that such causes varying the 

 effect must be considered as one of the inevitable 

 disadvantages of the photographic method, which 

 even in the case of experiments with crystals, giving 

 reflections or spots of variable form, should be 

 checked by such devices as we have used, or even by 

 the ionisation method. Mr. Keene, in a recent 

 letter to Nature, seems to suggest the convertibility 

 of the spots and bands by the mechanical structure 

 of rolled metals. 



Inference is either direct or indirect. A thoroughly 

 instructive experiment does not allow one to escape 

 its lesson. But there is another kind of inference, 

 less trustworthy, but still necessary, to guide re- 

 search, which demands theoretical knowledge or 

 ingenious insight into analogy. No doubt one 

 must first adopt the method of frontal attack which 

 leads to the inevitable form of inference, yet after- 

 wards probabilities or speculative suggestions may 

 be tolerable or even useful. 



Now, if the X-rays are simply light, the reasons 

 which may compel them to appear different may be : 

 (1) that the velocity is higher ; (2) that the mass of 

 the constituent rays or particles may be greater, the 

 total momentum being higher on one or other of 

 these grounds. I prefer to suppose that it is the 

 velocity, considered as a function of the inertia or 

 primary property of resistance, which is greater than 

 that of ordinary light.* It does not appear that 

 experiments have yet rendered untenable this 



reduction of light and matter to one and the same 

 substantial theory of electrical resistance. If not, it 

 may form a simple aim or guiding hypothesis which, 

 being proven, would constitute an advance towards 

 the simplicity of generalisation. One or two salient 

 facts which tend in support of such a theory may be 

 mentioned here, not because they prove it, but rather 

 that they are steps on the way. Thus the writer 

 found, following Tribe and Leduc, that if a 

 small quantity of potassium ferri-cyanide be com- 

 pounded with gelatine, and the mass subjected 

 to electric currents of low voltages, coloured rings 

 appear in the gelatine round the electrodes, and these 

 rings are undoubtedly due to the transport of ions from 

 one pole to the other after the chemical dissociation 

 of molecules. In one case a clear-cut impression 

 like a seal or intaglio was obtained round the zinc 

 electrode as though matter had been actually 

 scooped out and transferred to the other pole. Also 

 Scheffer, of Berlin, has recently shown that probably 

 particles of silver bromide in the gelatine of such 

 plates as were used for these X-ray experiments 

 explode and emit daughter particles having filaments 

 connecting them to the parents. It seems reasonable 

 to expect that there should be for freed electrons 

 some lateral principle of limitation of the distribution 

 answering to the vertical geometrical progression 

 discovered by Perrin for the arrangement of small 

 particles in fluids. For instance, he found (vide 

 " The Brownian Movement and Molecular Reality," 

 translated by F. Soddy, F.R.S., page 42) that the 

 concentrations of granules were determined in five 

 equidistant planes, the numbers being 100, 116, 146 

 170, and 200, whereas the numbers 100, 119, 142, 

 169, and 201, which do not differ from the preceding 

 by more than the limits of an experimental error, 

 are in geometrical progression, the altitudes (repre- 

 sented in Gregory Vincent's construction as areas 

 or volumes) being in arithmetical progression. This 

 is also suggested to the ordinary observer by the 

 form of the smoke from volcanoes in still air, which 

 spreads out into a flat plate bounded on the under- 

 surface by what resembles a hyperbolic curve. Such 

 a principle, if based on numerical experiments such 

 as those of Perrin, with the granules of X-ray 

 plates, might tend in the direction of setting free the 

 logarithmic or hyperbolic curve from the special 

 mode of energy-action we know as gravitation, 

 giving us a counter-principle of levitation or atomic- 

 segregation through the cosmos acting, in general, 

 contrary to weight, which is conformable to much 

 recent research and speculation, including Professor 

 Bickerton's new astronomy. 



;: This view was first expressed by the writer before Kaufmann and Abraham had given experimental evidence of it for 

 Cathode particles. Vide " New Theories in Biology," Zoological Record, 1899, where it is called the principle 



of " Cumulative inertia." 



