34 



THE VOYAGE OF H.M.S. CHALLENGER. 



" No. 5. The cylinder in its present condition may be worked with safety up to 18 or 20 tons per 

 square inch. Of course the breaking strength, calculated in the resistance of the several parts of the 

 cylinder, is very much greater. 



" No. 6. The weights of the several parts are as follows : — 



" Cylinder, 

 Key, . 

 Plug, . 



Tons, cvvts. qrs. lbs. 



3 10 



3 14 



1 27 



Total, 



3 



1 13 



17/2/79. 



(Signed) C. W. Younghusband, M. Gen 1 -' 



The instrument was erected in a basement room on the north side of the College, on a stone slab 

 IS inches thick, which was supported by a large mass of concrete imbedded in the ground below the 

 floor, and in no way connected with the building. 



A wooden platform, 3 feet 9 in. high, was erected round the instrument to facilitate access to the 

 chamber from the top. When very high pressures were to be applied, a wooden screen, lined with 

 sheet-iron, was erected under the platform as a defence against nuts or other pieces of metal, in case 

 part of the lower fittings (by far the weakest part of the whole) should give way. This precaution 

 was taken in consequence of an accident which had happened with the old pressure apparatus. 

 Although no great pressure had been reached, a screw was stripped, and the nut projected with 

 considerable velocity. 



The nature of these lower fittings will be seen at once from the woodcut annexed. A block of 

 iron, with lateral attachments for the pump and for the external gauge, was fixed by three powerful 

 vices to the external flange of the steel core of the pressure- chamber. It was pierced by a hole 

 of -j^th inch diameter, exactly in line with the hole in the steel cylinder above. This hole was closed 

 below by a screw-tap, so that by withdrawing the tap a steel wire could be easily passed into the 

 pressure-chamber in case of obstructions in the narrow tube. This hole was intersected at right 

 angles by another bore, communicating at its ends with the lateral attachments. To prevent leakage 

 where the block was pressed against the lower flange of the steel cylinder, we first employed a leather 

 washer. But for this we afterwards substituted a washer of block tin. This was found to work 



