580 
DE. J. P. JOULE PEOEESSOE W. THOMSON ON 
part of the time were traced to various causes, one of the principal being the length 
of time which, on account of the large capacity for heat and the small conductivity of 
the box-wood nozle, elapsed before the first large thermal effects consequent on the 
getting up of the pressure were dissipated. No doubt the results we arrived at were 
very accm’ate with the elastic fluids employed, viz. atmospheric air and carbonic acid ; 
but we possessed an unlimited supply of the former and a supply of the latter equal to 
120 cubic feet, which was sufficient to last for more than half an hour without being 
exhausted. In extending the inquiry to gases not so readily procured in large quantities, 
it was therefore desirable to use a porous plug of smaller dimensions enclosed in a nozle 
of less capacity for heat, so as to arrive rapidly at the normal efiect. 
Various alterations of the apparatus were made in order to meet the new require- 
ments of our experiments. A small high-pressure engine of about one horse-power was 
placed in gear with a double-acting compressing air-pump, which had a cylinder 4^ inches 
in diameter, with a length of stroke of 9 inches. The engine was able to work the piston 
of the pump sixty complete strokes in the minute. The quantity of air which it ought 
to have discharged at low pressure was therefore upwards of 16,000 cubic inches per 
minute. But much loss, of course, occurred from leakage past the metallic piston, and 
in consequence of the necessary clearance at the top and bottom of the cylinder when 
the pressure increased by a few atmospheres ; so that in practice we never pumped more 
than 8000 cubic inches per minute. 
The nozle we employed will be understood by inspecting Plate XNVI. fig. 1, where 
a a is the upright end of the coil of copper pipes. On a shoulder within the pipe a perfo- 
rated metallic disk (b) rests. Over this is a short piece of india-rubber tube (c c) enclosing 
a silk plug (d), which is kept in a compressed state by the upper perforated metallic 
plate (e). This upper plate is pressed down with any required force by the operation of 
the screw/ on the metallic tube g g. A tube of cork {hh) is placed within the metallic 
tube, in order to protect the bulb of the thermometer from the efiects of a too rapid 
conduction of heat from the bath. Cotton-wool is loosely packed round the bulb, so as 
to distribute the flowing air as evenly as possible. The glass tube {i i) is attached to the 
nozle by means of a piece of strong india-rubber tubiag, and through it the iadications 
of the thermometer are read. The top of the glass tube is attached to the metallic 
tube 1 for the purpose of conveying the gas to the meter. 
The thermometer (m) for registering the temperature of the bath is placed with its 
bulb near the nozle. The level of the water is shown by and oo represents the 
wooden cover of the bath. 
When a high temperature was employed, it was maintained by introducing steam into 
the bath by means of a pipe led from the boiler. The water of the bath was in every 
case constantly and thoroughly stirred, especially when high temperatures were used. 
The general disposition of the apparatus will be understood from fig. 2, in which A 
represents the boiler, B the steam-engine geared to the condensing air-pump C. From 
this pump the compressed air passes through a tram of pipes 60 feet long and 2 inches 
