GASOMETERS. 137 



greater, and therefore they are not generally used, except in cases before 

 mentioned. They are in use at the Chartered, Phoenix, and other companies' 

 works in London, and are worked with the same confidence as the more 

 simple vessels. 



I insert the following extract from Mr. Clegg's Journal, because the obser- 

 vations contained therein are founded upon a practice of forty years, and are 

 therefore valuable : 



" All the gasometers at the works must be connected together, as if they 

 were one vessel, and each one must regulate its own pressure, by a wooden 

 curb, assisted by weights where found necessary, when the gasometers are put 

 into action. The flow of gas from one to another must not, on any account, 

 be guided by a workman, nor must he be allowed to close any valve connected 

 with them, unless expressly ordered to do so. Indeed it would not be amiss 

 to remove all the valves, and seal the outlet- and inlet-pipes by water, when 

 the gasometer they belong to requires repair. 



" I would warn all young engineers, upon the necessity of making pneumatic, 

 hydraulic, and other machinery, as much as possible self-acting. In the first 

 place it is the perfection of mechanism ; in the second place it prevents danger 

 from neglect ; and in the third place it is universally found cheaper, and tends 

 much towards the oeconomy of machinery in general. 



" When the steam-engine was first invented, a boy was employed to open 

 and shut the valves connected with the steam-ways, above and below the piston, 

 when the engine was at the top and bottom stroke. A man was also em- 

 ployed to regulate the quantity of steam according to the power required. 

 The engine now regulates itself. The damper which regulates the draft of the 

 furnace rises and falls, as the steam is high or low, by the same contrivance 

 as that by which the water in the boiler is kept at one uniform level. 



" The water-wheel, if it does not regulate its own supply of water, accord- 

 ing to the quantity of work it has to perform, is a very imperfect machine. 

 The same may be said of windmills, blowing-engines, and even of switches on 

 railways, which ought also to be self-acting ; and in gas-works where it can 

 be, and is not done, it is the cause of many serious accidents. 



" In all the parts that are self-acting, such as the hydraulic main, no acci- 

 dent ever arises from a neglected valve, which was constantly the case when 

 there was a valve to each retort requiring attendance. The engineer of the 

 Glasgow gas-works, when they were first erected (and before Mr. Liddle had 



