portion to the magnitude of the cylinder, the momentary expansion of the steam 

 passing to the cylinder from the boiler at each stroke would reduce the pros- 

 sure of the steam in a great proportion, and unless the pressure in the boiler 

 were considerably greater than that which the steam is intended to have in 

 the cylinder, the pressure in the latter would be reduced below the proper 

 amount. The proportion of the steam-space in the boiler to the magnitude of 

 the cylinder has been very variously estimated, nor can it be said that any 

 practical rule of a general kind has been adopted. It is held by some that the 

 steam-space will be sufficient if it contain five times the quantity of steam con- 

 sumed at each stroke, while others maintain that it should contain at least ten 

 times that quantity, and opinions vary between these limits. 



The proportion of water-space in the boiler to its evaporating power should 

 also be regulated, so that the introduction of the feed at a comparatively low 

 temperature may not unduly chill the water in the boiler. Supposing the feed 

 to be introduced in a low-pressure boiler at the temperature of 100, and that 

 the necessary temperature within the boiler be 225, the quantity of water it 

 contains should be about five times the quantity evaporated, and therefore also 

 five times the quantity introduced through the feed per hour. For every cubic 

 foot of water per hour, therefore, intended to be evaporated by the boiler, water- 

 space for five cubic feet should be provided. It is, however, right to repeat 

 that this (like almost every other so-called rule) is the result, not of any exact 

 general calculation, but one deduced from the custom which has obtained 

 among the manufacturers of steam-engines. 



The surface of the water in the boiler should always be above the range of 

 the flues. When the heated air in the flues acts upon a part of the boiler 

 within which water is contained, the water within receiving an increased tem- 

 perature becomes, bulk for bulk, lighter than the strata of water above it, arid 

 ascends. It is replaced by the descending strata, which, in their turn receiv- 

 ing increased temperature, rise to the surface ; or if the action of the heat con- 

 vert the water into steam, the bubbles of steam rise to the surface, fresh por- 

 tions of water continually coming into contact with the boiler-plates on which 

 the heated air or flame acts. By this process the boiler-plates are continually 

 cooled, either by being successively washed by water at a lower temperature, 

 or by the heat taken from tiiem becoming latent in the steam-bubbles formed in 

 contact with them. But if the heat act upon a part of the boiler containing 

 steam within it, which steam being a slow recipient of heat, and no currents 

 being established, nor any phenomenon produced in which heat is rendered 

 latent, the heat of the fire communicated to the boiler-plates accumulates in 

 them, and raises their temperature to an injurious degree. The plates may by 

 this means be softened, so as to cause the boiler to burst, or the difference be- 

 tween the expansion of the highly -heated plates thus exposed to fire in contact 

 with steam and that of the plates which are cooled by contact with water, 

 may cause the joinings of the boiler-plates to open, and the boiler to leak. 

 By whatever means, therefore, the boiler be fed, care should be taken that the 

 evaporation should not be allowed to reduce the level of the water in it below 

 the highest flue. 



As the water by which the boiler is fed must always have a much lower 

 temperature than that at which the boiler is maintained, the supply of the feed 

 will have a constant tendency to lower the temperature of the water, and this 

 tendency will be determined by the proportion between the magnitude of the 

 feed and the quantity of water in the boiler. 



Since it is requisite that the level of the water in the boiler shall not suffer 

 any considerable change, it is evident that the magnitude of the feed must be 

 equal to the quantity of water evaporated. If it were less, the level of the 



