Explanatory Notes, &c. By Trof. W. A. Rogers. 21 



cleaned by rubbing in the direction of the lines only. It will be 

 seen that the particles of glass are more completely removed in this 

 plate tban in No. 4. It should be noted here that the best lines 

 are always obtained by rubbing in the direction of the lines and 

 never by rubbing at right angles to the lines. That the width of 

 the lines remains 2 • 2 /i seems to he ^positive froof that the portion 

 removed was a real windrow of minute particles. 



No. 6. This plate, after ruling, presented the same appearance 

 as No. 2. It was then sent to Prof. Hamilton L. Smith to be 

 mounted in his new medium, with the expectation that the bril- 

 liancy of the spectrum would be sensibly increased. It will be 

 seen that this expectation has not been realized. The only expla- 

 nation which I can offer for this unexpected result, is that the 

 apparent lines being elevations, appear as projections. By focus- 

 ing upon the bottom of the real grooves, a very fine line is seen 

 which was not noticed before. In a previous plate of heavy lines, 

 kindly mounted for me by Prof. Smith, the sharpness of definition 

 was increased to a very marked degree. 



No. 7. This plate is half of a slide ruled upon my old machine 

 in 1881. One set of lines was ruled on the slide, and another set 

 upon the cover. Mr. Tolles aided me in a thorough examination 

 of these bands. We were never able to see the fine lines which 

 form the continuation of the 1/24000 band. This plate was sent 

 to Prof. Smith for experiment with his new medium. He re- 

 moved the cover from the rulings upon the slide, remounted the 

 bands in his medium, and after cutting the slide into two parts 

 sent one half to me. The lines are in every way improved, and 

 the fine lines of the 1/21000 band are easily seen. 



Fifth Series. Buled in 1885. 



No. 1. A convenient form of stage micrometer. Each succes- 

 sive band ruled with a lighter pressure of the diamond. As there 

 may have been a slight disturbance of the diamond produced by 

 removing the weights, the bands may not be exactly equidistant. 

 The measurements, therefore, should be from lines composing the 

 the bands. 



No. 2. Metric stage micrometer similar to No. 1. 



Nos. 3, 4, 5, 6, 7, and 8. Eye-piece micrometers of various 

 patterns. 



