538 
MR. J. T. BOTTOMLEY ON THE 
The mode of applying heat in this second apparatus is different from that described 
above. When an experiment is to be made the tank is filled with water (or other 
liquid) up to the level, D, and after the water has been carefully stirred from top to 
bottom and has come to rest after the stirring, a “floater 5 ’ is placed on the surface. 
The floater is a large thin board with its edges carefully bevelled all round. It covers 
nearly the whole surface of the liquid below. Hot water is allowed to flow, slowly at 
first, but afterwards as quickly as is convenient, on the top of the floater, and a thick 
layer of very hot water is laid on the top of the colder water beneath. The edges 
of the floater are bevelled so that the hot water falling on it is shot off horizontally at 
Fig. 1. 
the edges without any downward motion, and therefore with no tendency to mingle 
with the colder water, of greater density, below; and if the operation of laying on a 
layer of hot water is carefully carried out, there is scarcely any mixing of the hot 
water with the cold. This can be seen if the experiment is tried in a large glass 
vessel. A layer of hot water can easily be deposited on the top of a quantity of cold 
water, with so little mixing that the layer of separation intermediate between the hot 
and the cold, which are distinguishable by the difference in refractive indices of hot 
and cold water, is seen to be not thicker than a sheet of paper. I have also tried 
experiments as to mixing on a large scale in the sheet-iron tank, fig. 1. On running 
