MEASURING MACHINES. 109 



read, it does not by any means follow that an object 

 can be actually measured to this degree of accuracy ; 

 in fact, it is well-nigh impossible, except by the 

 most practised observers. It is not an easy thing to 

 realise what a millionth of an inch means. Let the reader 

 look at an inch divided into hundredths and try to imagine 

 a tenth of one of these : this will be a thousandth. Then 

 let him try to still further imagine one-thousandth of one 

 of these thousandths : this will be a millionth of an inch. It 

 is almost inconceivable. The smallness of one of these 

 millionths will be appreciated when it is remembered that 

 an inch bears the same ratio to a millionth part of itself as 

 one mile bears to a sixteenth of an inch, or nearly so. It 

 is easy to measure to a thousandth or even a ten-thousandth 

 of an inch ; it requires a much greater delicacy of touch 

 and skill to measure to one hundred-thousandth ; in fact, 

 there are few people who are capable of it. But when 

 it comes to measuring to so great a degree of precision as a 

 millionth, the observer has a task before him which is well- 

 nigh impossible of achievement. The only way a millionth 

 can be detected is by a slight difference of touch. The 

 effect of the warmth of the observer's hand is quite suffi- 

 cient to cause this difference. So that it will be seen that 



FIG. 42. 



for testing, as well as for almost all other purposes, measure- 

 ments to so great a degree of accuracy as involve 

 millionths of an inch are not n6cessary, and at the same time 

 are most difficult to carry out. A much more workable and 

 serviceable machine for ordinary purposes than that just 

 described is one, also made by Messrs. Whitworth, 

 which is capable of making measurements up to ten- 

 thousandths of an inch. The fine pitch screw of this 

 machine has 20 threads to the inch, and the hand-wheel 

 used to rotate this screw is of a large diameter, and has its 

 circumference divided into 500 equal parts. One of these 

 divisions therefore corresponds to T ^ x ? V or Tu^th of an 

 inch. Such a machine as this will be found very useful 

 for many purposes in a testing laboratory, chief among 

 these being the calibration and checking of other measur- 

 ing instruments. 



