SCIENCE AND INDUSTRY. 
cylinder is then once more heated up and carefully annealed. When 
cold, it is again put in a lathe and the neck is turned, bored, and tapped. 
The cylinder is now ready for testing. 
The question of the final and proper annealing of the cylinder pre- 
vious to testing is of prime importance. The actual temperature is 
naturally dependent on the composition of the steel, and is so definite 
that there’ is no scope for more than about 10 degrees F. variation either 
way. It can only be established by experimenting. 
A plant with 3 presses and 3 steam-hammers would be capable of 
turning out from 300 to 500 cylinders per week, working “ days only,” « 
after thé preliminary drawing operations have been carried out. In 
other words, during the period, say one weck, occupied with the initial 
drawing and shaping of the cylinder, there would be no output. The 
dies and rams would then be changed, and the ensuing week would be 
productive. For continuous output, the plant would obviously have 
to be doubled. 
Though reliable estimates of cost are not available, it is probable 
that the cost of such an installation, with buildings, gas producer, 
furnaces, presses and dies, lathes, steam-hammers, testing plant, 
hydraulic accumulator, and all necessary accessories, would amount to 
upwards of £50,000. This at once establishes a considerable overhead 
charge for depreciation, which alone would amount, at 10 per cent., to 
£1 per cylinder on an output of 5,000 cylinders per year. 
As regards the number of cylinders required for a given gas plant, 
experience in England and America shows that a cylinder is filled about 
ten times a year on the average of the whole lot. Thus, a plant, to turn 
out 1,000,000 cubic feet of oxygen annually, would require about 1,000 
cylinders, each of 100 cubic feet capacity. 
Pia 
Legetiy' 
UpigNN 
LxS west 
x 7) ll i | 
Yyhi) Qty 
694 
