152 Proceedings of Royal Society of Edinburgh. [sess. 
the end. In those curves which are symmetrical about the 
central transverse line, the uninode is obviously not displaced 
from the centre. Not only does a curve of varying depth 
have its nodes displaced towards the shallower parts, as specified 
above, but definite experiments explicitly showed that the 
tendency of a shallow in the neighbourhood of a node is to 
displace the node from its normal position in the direction of 
the shallow. 
In the case of the concave parabola, of length 127 cm. and 
maximum depth 9 '4 cm., the binodal line is situated 27 cm. 
from the end. A rectangular obstacle of dimensions 5 cm. x 3 
cm. x 2 cm. was submerged. The average distance of the top of 
the rectangular body from the surface of the water was 2 cm. ; 
the mean distance of the edges from the end of the water was 
12 cm. Under these circumstances the position of the node was 
observed as being 23’5 cm. from the end; i.e. the node was 
displaced 3*5 cms., and towards the shallow caused by the 
obstacle. 
This particular experiment is of interest in connection with 
observations of the west binode of Loch Earn. This binode 
was calculated to be in the neighbourhood of the position of 
greatest depth of the loch, and slightly to the west of it. The 
ualculated binodal position lies between a particularly deep part 
and a rise in the bottom of the loch. Experimental investigation 
led to the conclusion that the node was really considerably to the 
west of the calculated position, thus showing that the deep and 
the neighbouring shallow had exercised an influence on the 
position of the node. 
Experiments were made with a model of Loch Earn, of which 
the longitudinal section was that along the line of greatest depth 
of the loch and in which the depth was increased ten times 
relatively to the length. While the model, being of uniform 
breadth, took no account of the varying breadth of the loch, as 
would be the case in a “normal” curve, the observed periods are 
nevertheless interesting. These are 
Uninodal period . . . .2*8 secs. 
Binodal . . . . L52 „ 
Trinodal ,, . . . IT ,, 
* 
