OF ARTS AND SCIENCES. 237 



XVIII. 



ON A POSSIBLE EXPLANATION OF TItE METHOD EM- 

 PLOYED BY NOBERT IN RULING HIS TEST PLATES. 



By William A. Rogers. 

 Presented, June 9, 1875. 



I RECOGNIZE the fact that no explanation of a purely mechanical 

 process can be regarded as either satisfactory or final, which does not 

 answer the crucial test of reproduction. I oiFer to the Academy what 

 I believe may prove to be an explanation of the pi'ocess followed by 

 Nobert in ruling his test plates ; the highest band which has been 

 resolved under the microscope, reaching 112,600 lines to the inch. 

 You properly ask me if I can reproduce these rulings. I frankly 

 answer that I cannot. Indeed, I can hardly hope ever to succeed in 

 producing lines which combine the wonderful delicacy, uniformity, 

 and distinctness found in nearly all of Nobert's plates. But I have 

 reached what I hope may prove to be a useful approximation to 

 Kobert's results. Beginning with 2000 lines to the inch in 1871, I 

 have now little difficulty in reaching 60,000, the width of each line 

 being a little less than one half of the intervening space. Several 

 of my plates have been correctly counted as far as 80,000 to the inch ; 

 the observer having no knowledge of the actual number ruled. Two 

 plates in the possession of Frederick Habirshaw, Esq., of New York, ' 

 contain bands proceeding by 10,000 as far as 120,000 to the inch. 

 The bands of both these plates were correctly counted by Samuel 

 Wells, Esq., of Boston, as far as 80,000, but beyond that point the 

 number counted was less than the number ruled. While the lines of 

 the higher bands seem to be nearly as distinct as Nobert's, they are 

 by no means as smooth and uniform throughout their whole length. 



The theory which I offer to the Academy is wholly the outgrowth 

 ot my own experience. In the various experiments which I have 

 made, I have noted the constant recurrence of certain results under 

 certain conditions, and these results form the basis of my conclusions. 

 Whether they form a true explanation of Nobert's process is, of course, 



