208 
DR. H. T. BARNES ON THE CAPACITY FOR HEAT OF WATER 
fitted up in the large tank, similar to the one described by Gouy (‘ Journ. de 
Physique,’ vol. 6, p. 479, 1897). Fig. 11 contains a drawing of this regulator with 
its attachments. A bulb of glass A, containing about 300 cub. centims. of toluene 
resting on a mercury surface B, is connected to a fine heavy-walled tube about 
1 millim. inside diameter, through which the mercury at B is made to pass by the 
expansion or contraction of the toluene. A three-way glass tap allows the mercury 
to pass either into the reservoir C, or up into the tube E. A platinum wire point is 
attached to a copper wire and drawn up and down about 4 millims. by a pivot on 
the wheel F, worked by a worm-wheel from the pulley C 1 . 
In this method, which is the distinctive feature of the Gouy regulator, the 
platinum point never sticks to the mercury surface, and consequently gives a sharper 
and more definite electrical connection between the thread of mercury in E and the 
wire. Connections were taken from this to a telegraphic relay, which was so 
arranged as to throw in and out a heating lamp placed in the tank. The arrange¬ 
ment is shown in fig. 12. 
When the relay is inactive, the terminals of the lamp are short-circuited by the 
arm extending between the mercury cups a and b , and the full current is permitted 
to pass through the lamp B. When the relay is excited, the arm ah is raised and 
the circuit broken at a , so as to bring the lamps A and B in series. Lamp A was a 
10-candle-power lamp of 200 ohms resistance, while B was either a 32-candle-power 
lamp of 100 ohms resistance, or a 50-candle-power lamp of about 60 ohms resistance. 
Either of the two heating lamps in series with lamp A was reduced in heating power 
over one-fourth of its full amount. The relay was made active by the closing of the 
