SOME REFINEMENTS OF MECHANICAL SCIENCE. 1-49 



The threads of the silkworm, although of great value as a com- 

 mercial product, are so coarse and rough compared ^Yith the silk of 

 the spider that they can not be used in such instruments. 



Platinum wires are made sufficiently fine, and make most excellent 

 cross wares for instruments where low" magnifying powers are used, 

 yet as the powder increases they become rough and imperfect. 



Spider lines, although of but a fraction of a thousandth of an 

 inch in diameter, are made up of several thousands of microscopic 

 streams of fluid, which unite and form a single line, and it is because 

 of this that they remain true and round under the highest magnify- 

 ing poAver. 



An instance of the durability of the spider lines is found at the 

 Allegheny Observatory, Avhere the same set of lines in the micrometer 

 of the transit instrument has been in use since 1850. 



The placing of the spider lines in the micrometer is a work of great 

 delicacy, and in some micrometers there are as many as thirty, which 

 form a reticule, with lines two one-thousandths of an inch apart 

 and parallel with each other under the highest magnifying power. 



Step by step, from the methods of the Arabian astronomers to 

 the time of Tycho Brahe and on down to the present day, improve- 

 ments in the instruments and methods for the measurement of angles 

 have been going on, until astronomers can measure double stars with 

 a separation of one second of arc, and within less than one second 

 they can define their positions in the heavens. 



In the realm of the measurements of minute linear distances and 

 the perfection of curved and flat surfaces the refinements are even 

 greater than those pertaining to the measurement of time and of 

 angles. 



Most important in the linear dividing engine is the screw, and 

 although much had been accomplished in bringing such engines to a 

 high degree of excellence, it was for Professor Rowland to make an 

 engine wdiicli has a practically perfect screw; and without doubt it 

 is in all respects the nearest perfect of all the mechanisms that have 

 been employed for ruling lines exactly parallel and equally spaced. 



The Rowland engine was made especially for ruling diffraction 

 gratings which are made of speculum metal, and with it a metal sur- 

 face has been ruled with IGO.OOO lines, there being about -20,000 to the 

 inch, and as many as 43,000 lines to the inch have been ruled. 



The gratings mostly used have from 14,000 to '20,000 lines to the 

 inch, and with such exactness is the cutting tool moved by the screw 

 that the greatest error in the ruling does not exceed one-millionth of 

 an inch. 



The production of these gratings, which has enabled the physicist 

 in his study of the spectrum to enter fields of research before un- 

 known, has not only called for the highest degree of perfection ever 

 attained in the spacing of linear distances, but it has also called for 



