318 Prof. R. W. Wood on tJi 



le 



to only 13 ohms by the cuts. I feel confident that the 

 diamond removed all of the metal from the cuts, for they 

 appeared as brilliantly illuminated under the microscope as 

 the clear glass from which all of the silver had been 

 removed. 



Photographs of the rulings were made with a 1/12 inch 

 oil-immersion objective, and the width of the cut detei mined 

 by comparing it with the distance between the cuts, the 

 measurements being made with a dividing engine. In the 

 majority of cases the width turned out to be no greater than 

 the diffraction maximum yielded by the objective, and in 

 seme cases it was probably less, the cuts appearing slightly 

 greyish, as would be the case if the width was less than the 

 resolving power of the objective. This makes the order of 

 magnitude about equal to the wave-length of ultra-violet 

 light, say 0*0003 mm. Photographs of the rulings will be 

 published in connexion with the work on heat-waves. 



A number of strips of silver were prepared, each one cut 

 across by 20 strokes of the diamond, and the resistances 

 measured by applying soft pads of silver-leaf as electrodes 

 so as to include, or not include, the cross-ruled portion. 

 The increase in the resistance was found to be very variable, 

 fcr in some cases the original resistance of about 10 ohms 

 was raised to three or four hundred, and in others to 15 or 

 less. With cuts somewhat wider, say two wave-lengths, no 

 trace of conduction could be found, and it seems probable 

 that the great variation can be safely attributed to slight 

 differences in the widths of the cuts, the differences being too 

 small to be shown by the microscope. The voltage applied 

 was that of one dry cell, and by means of a potentiometer it 

 was found that conduction occurred with potentials as low 

 as O'OOl of a volt. Covering the ruled portion with a drop 

 of oil did not alter the resistance, though this is a point that 

 has not been very thoroughly investigated. 



It appeared to me possible that the conductivity at these 

 very low voltages might be ascribed to the extent of the 

 opposed surfaces. In all previous work the electrodes have 

 been either spherical or " point and plane," and when brought 

 close together give us what amounts to two opposed points, 

 on account of minute irregularities on the surfaces. If this 

 explanation is the true one conductivity at greater distances 

 should be obtained in the case of two opposed planes. 



It is obviously out of the question to use metal plates as 

 ordinarily prepared for electrical experiments, and, as I shall 

 show presently, even gold films deposited on optical flats by 

 cathode discharge are scarcely suitable on account of minute 



