284 PETERS: MEASUREMENT OF SMALI^ DILATATIONS 



By using the Fizeau apparatus in the manner described in 

 this paper the absolute change of length of one, two, or three 

 small pins can be determined during one experiment. The 

 exact change in each pin is determined and not the mean change 

 of the three pins. The unequal change of similar pins is not 

 important in most practical tests, but where precision is de- 

 sired it is quite annoying. The differential changes in length 

 of three pins can also be determined without knowing the abso- 

 lute change of any of them. 



As it is rather difficult to watch the displacement of the fringes 

 past the three reference marks simultaneously, it is better to 

 determine the absolute change in length of one pin by equation 

 (i) and compute the values for the other two pins with equa- 

 tions (4) and (5). The results in either case are exactly the same 

 because they depend upon the same measurements of the frac- 

 tional orders. The accuracy of these results is limited only by 

 the accuracy with which the fractional orders are determined. 



