190 T. A. Hirst on Ripples, 
24. The moment the instrument rotates with sufficient velo- 
city the ripples make their appearance, and the general resem- 
blance between their forms and that of the one drawn according 
to theory in fig. 7, Plate I., is very striking. The external branch 
A X, and the portion of the internal one between the jet A and 
the cusp C are quite distinct and well defined ; the return branch 
CBD is far less distinct, the height of this part of the ripple 
being small, and its breadth great; its existence, however, may 
be established by a closer inspection, or by a glance at the dis- 
torted image at the bottom of the bath caused by the irregular 
refraction of the light incident upon the rippled surface. Besides 
the principal ripple there are anumber of secondary ones, similar 
in form but less in extent, which precede the former, and of 
course tend to diminish its prominence. 
25. In order to compare the actual with the theoretical ripple 
more precisely, the following simple expedient was adopted :— 
A circular plate of glass K La foot in diameter, and having a cir- 
cular aperture at its centre to admit the screw O, was attached 
by means of this screw to the top of the lead weight. Its lower 
surface was silvered, and the silvering afterwards coated with 
shell-lac to protect it from the action of the water. The image 
of the ripple in this plane mirror was not only more distinct than 
the ripple itself, but the arrangement had the additional advan- 
tage of obviating errors due to parallax, since the image of the 
observer’s eye could always be made to coincide with that of any 
part of the ripple under examination. Lastly, a second and 
transparent glass plate A A’, 10 inches in diameter, was attached 
by means of a screw E E! to the cap at the bottom of the mill. This 
plate turned of course with the mill, and could be graduated in 
any required manner; it had a small indentation at a poimt in 
its circumference, in order that the axis of one, A, of the vertical 
jets (or cylinders) might be made to describe, precisely, the 
circumference of a circle 5 inches in radius. 
26. By means of the equations (32) and (40), one or two rip- 
ples AaC and Aa,C, (fig. 2), and the locus A C C, of their cusps, 
were carefully plotted on a sheet of paper; they corresponded to a 
value of a=5 inches, and in the case of the ripples, to arbitrarily 
assumed values of the ratio a between the velocities X and w of 
the wave and jet. These curves were next transferred, in white 
paint, to the glass plate A A’, and one of the jets A made to 
coincide with their origin. The instrument being now filled 
with water, the jet at A’ was made to issue horizontally in order 
to turn the mill, whilst the jet at A descended vertically into 
the bath and produced the ripple to be examined. As the velocity 
of rotation diminished, the cusp of the actual ripple described very 
accurately the curve C,C A traced on the glass plate AA’; and 
