116 Proceedings of Royal Society of Edinburgh. [sess. 
The second term on the right-hand side of the equation is the part 
depending on gravity, and it is very small relatively to the other 
term. In the case of mercury, taking the greatest wave-length 
employed, its value is only 1/2040 of the first term, a quantity far 
within the limits of observational error. I have accordingly used 
in all the calculations the form T = X z p/2Trt 2 . If in this A. be 
measured in centimetres and t in seconds, T is given in dynes per 
centimetre. Previous attempts made by myself and others to 
determine surface-tensions by this method failed on account of the 
difficulty experienced in measuring the lengths of the ripples, and 
since T oc A 3 , it is evident that the method is useless unless X can 
be measured with great accuracy. This difficulty has been over- 
come by the use of photography, and it is now possible to test the 
practical value of the method by comparing the results that have 
been obtained with those got by other means. The application of 
photography is easy, since the waves that are dealt with are “stand- 
ing” waves, and when the apparatus is in proper adjustment the 
photograph is sharp even with exposures of as much as 10 seconds. 
The usual exposure is from 1 to 2 seconds. Thus there is an essen- 
tial difference between this application of photography and that 
made by Lord Kayleigh, who photographs a moving stream in a small 
fraction of a second. 
The general plan of procedure is as follows :^A tuning-fork in 
which the vibrations are electrically maintained is employed to 
produce ripples on the surface of the liquid under experiment, 
which is contained in a shallow dish. A scale is suitably placed on 
the edge of the dish so that its divisions are as nearly as possible 
parallel to the crests of the ripples, and a photograph of the dish 
is then taken. This photograph is measured with a micrometer, 
and the measurements are interpreted by reference to those of the 
scale photographed on the same plate. 
In the earlier experiments much difficulty was experienced in 
getting suitable illumination, but ultimately it was found that the 
best photographs were got simply by placing the dish near a window 
and using the light from the sky or a bright cloud. In most cases 
it was found best to allow a shadow of a bar of the window or of 
the tuning-fork itself to fall on the surface, for the edge of such a 
shadow becomes regularly serrated, and this serration is easily 
