522 
before the objective of the telescope, has been placed at O (cf. 
the above figure). This ensures that the light passes through the 
apparatus at the right moment. The experimental error of the deter- 
mination of the velocity appears to amount to at most 0.75 °/,. 
We still wish to draw attention to a peculiarity of the described 
method of the velocity determination, which seems interesting from 
a theoretical, though not from a practical, point of view. 
The measurement actually takes place with a moving measuring 
rod (length at rest = S, S,), a peculiarity which we do not remember 
having seen used in practice with any other method. 
We are indebted to Messrs. W. pr Groor and G. C. Disserz Ju. 
for their assistance in the theoretical and experimental work, and 
to Mr. J. van DER Zwaat for his help in the difficult adjustment of 
the apparatus and the manufacture of the auxiliary appliances. 
