A. .4. Mioheleon — Measurement by Light-waves. L19 



These forms may also be used for differentia] measurements with the advant 

 that the displacement is double thai of the corresponding instruments of Group II. 



Group IV. — This group is useful only in oases where the velocity of light varies 



with the direction according as this is positive or negative. The only application 



thus far made is in the problem of the speed ol light in moving media For this 

 purpose it is -ntial that no accidental variations in the two paths should 



have any effect Accordingly each pencil is made to return by the other's path. 

 Of this group fig. 1 is the type; ii is the form used by Kizcau in his celebrated 

 experiment; and 2 that employed in a recent repetition of the same work.* 



Group V. — In this case the paths of the two pencils are unequal. Figfl 1, 2 and 

 3 represent various forms of grating Bpectroscope, of which the la-t is a concave 

 grating in which the spectral image coincides with the slit. The transition to the 

 corresponding refractometer forms. -I and ;">. is apparent. 



Linear 3[easurement$. 



It may be assumed from experiments made by Professors 

 Rogers and Morley that with the most accurate and powerful 

 microscope at their command, and with the very best condi- 

 tions as regards the object observed and its illumination, the 

 utmost attainable limit of accuracy of a setting of the cross- 

 hair of the microscope on a fine ruled line was about two mil- 

 lionths of an inch or about a tenth of a light-wave. Now it is 

 usually granted that the cross-hair of an eye piece can be set 

 upon the middle of an interference fringe, if this is sufficiently 

 broad and clear, with an error of about one thirtieth of a fringe. 

 In the refractometer this would mean an error in distance of 

 only one-sixtieth of a light-wave. 



The following is a record of the results of three direct meas- 

 urements of the length of a wave of green light emitted under 

 the proper conditions by incandescent mercury vapor. I have 

 been fortunate in obtaining the first series from observations 

 by Professor Morley, whose results with the microscope have 

 just been quoted. The second series w T as taken by myself ; 

 and the third, by an observer who had no previous practice in 

 this kind of measurement. In all cases the scale readings were 

 unknown to the observer. 



(1) 



A 



(2) 



A 



(3) 



A 



1190 



2-0 



118-5 



o-o 



1170 



-2-0 



117-5 



0-5 



119-0 



0-5 



118-5 



-0-5 



1170 



o-o 



1195 



1-0 



118 5 



-05 



116-0 



-10 



118-0 



-0-5 



122-5 



3-6 



115-5 



-1-5 



1195 



1 



120-5 



1-5 



117-5 



05 



1190 



n-6 



1205 



1-5 



117-0 



o-o 



118-5 



o-o 



1190 



o-o 



1170 



00 



1190 



05 



1180 



— l-o 



117-5 



0-5 



117-5 



-1-0 



117-5 



—1-6 



117-5 



0-5 



116-5 



-2-0 



118-5 



-TO 



Means: 11715 



0-65 



118-50 



0-70 



119-05 



1-30 



Error in waves: 



00056 





00059 





o-oiio 



* "Influence of Motion of the Medium on the Velocity of Light:" Michelaon 

 and Morley, this Journal, xxxi, May, 1886. 



Am. Jocr. Sci.— Third Series, Vol. XXXIX. No. 230— Feb., 1890. 

 9 



