of Light and its Theory. 107 



It is impossible to trace in any layer of fat, or varnish, parallel 

 lines more accurately at equal distances. 



It was only by means of a diamond that I obtained still 

 finer systems of lines, when I was enabled by a machine, con- 

 structed for this purpose, to trace parallel lines with a diamond 

 point, in the greatest perfection, immediately on the surface of 

 the plate of glass. If, by good fortune, a very good diamond 

 point is procured, by the help of this machine lines may be 

 traced so fine that they cannot be perceived by the most pow- 

 erful compound microscope. It is, however, not sufficient to 

 be able to trace a very considerable number of lines within a 

 given space that shall leave interstices between them, but it 

 is requisite that these lines should be at distances in so high a 

 degree equal to each other that the greatest number of them 

 shall not be nearer to each other, or farther, than the hun- 

 dredth part of this small distance. With the help of my ma- 

 chine I have obtained a system of lines in which 8 = 0,0001223 

 of an inch, the lines of which are at distances so very equal, 

 that the fixed lines of the first and second spectrum that are 

 seen through it may be very clearly distinguished. * Through 



• With this machine parallel lines, with intervals as wide as the lines 

 themselves, may be etched at such a small distance from each other that 

 S2000 will be contained in a Parisian inch ; but I have not hithcBto suc- 

 ceeded in giving them such equal distances from each other, which ought 

 to be 0,00003125 of an inch. Not many faults of a hundredth part, that 

 is, of 0,00000031 occur ; and it is perhaps not possible for human hands, 

 with any machinery that can be employed, to surpass this, since 100 or 200 

 parallel lines are of little use ; and in such fine systems of lines several 

 thousands are always required in order to obtain perfect intense spectra, 

 it requires much good fortune, even with s = 0,0001223 to find a dia- 

 mond point which shall trace several thousands of such fine lines without 

 becoming altered. I have hitherto only succeeded with one system of lines 

 so fine. If, during the process of tracing, the diamond point becomes al- 

 tered, all the preceding labour is lost. The point, without any apparent 

 cause, often makes stronger or weaker lines ; nor can it be discovered by 

 the strongest microscope whether the point is fit to trace the proper lines. 

 Hence it is only by trials that a useful point can be obtained. What ren- 

 ders the thing still more difficult is, that a small alteration in the inclina- 

 tion or position of the diamond, relatively to the plane of the plate of glass, 

 materially alters the force of the lines. Since every line requires to be 

 drawn singly, with great care, it may be easily conceived how much time 

 and patience are necessary to trace two thousand lines with the requisite 

 accuracy. 



