16 THE VOYAGE OF H.M.S. CHALLENGES,. 



water. The plug was forced down by the weight of an assistant standing on it, while a 

 stop-cock at the bottom of the cylinder was kept open for the escape of water, until a 

 massive steel key could be put in through a slot in the side of the cylinder to lock the 

 plug in its definite position. 



To the lower end of the steel cylinder were adapted a series of fittings by means of 

 which it could be connected with a powerful force-pump, and simultaneously with a gauge 

 whose construction will be afterwards described. The gauge enabled the experimenters 

 to know at every stage of the operation what amount of pressure had been reached in the 

 interior of the cylinder. The pump was worked at first by hand. Of late a more 

 powerful pump has been procured, and it can be fitted when necessary to the gas-engine 

 of my laboratory. 



Only one real difficulty was met with in working this apparatus ; viz., the difficulty 

 of making the plug fit perfectly tight. At first, when it came from Woolwich, the plug 

 was finished by a piece of leather in the form of a cup ; but this was found to leak seri- 

 ously even at very moderate pressures, so that even the comparatively small pressure of 

 a ton weight per square inch was unattainable. 



But by taking off the leather from the plug and furnishing it with a ring of steel 

 turned into cup form with an exceedingly thin and sharp edge, on the same principle as 

 that on which the piston of the pump was constructed, this difficulty was completely got 

 over. The flexible steel edge was pressed against the interior of the tube more forcibly 

 the greater the applied pressure, and it was found that the apparatus was then, except 

 under the most unfavourable circumstances, perfectly tight, at least so far as the plug was 

 concerned. Very great care was, however, requisite in cleaning the plug and the upper 

 part of the bore of the cylinder before each experiment. The smallest fragment of cotton- 

 waste, getting behind the edge of the cup, almost invariably produced serious leakage 

 when high pressure was applied. The cup form was objectionable for one reason, that it 

 always took down a considerable quantity of air, of which it was impossible to get rid. 

 This difficulty was overcome by putting into the cup a quantity of tallow which com- 

 pletely filled it up and projected considerably below it, so that the apparatus, when 

 pressure commenced, contained at the most a few small air bubbles only. 



Later, when I found it was impossible to obtain certain necessary data, on account of 

 the slowness with which pressure was got up in so large an apparatus, I procured a very much 

 smaller apparatus of similar character, in which the cylinder was only an inch in bore, 

 and rather less than a foot in effective interior length. With this apparatus two or three 

 strokes, only, of the pump were required to get up the desired pressure, and there was 

 the great additional advantage that temperatures could be independently measured by 

 means of thermo-electric junctions. [This could not be done in the large cylinder 

 without seriously affecting its strength, and rendering it at the same time almost un- 

 manageable.] 



