SECT, in.] CONDENSATION OF STEAM. 131 



under the supply boiler, as Oliver Evans proposed. When a supply boiler cannot 

 be used, somewhat more than one-fourth of the effect of the fuel will be lost by 

 the smoke escaping at such an elevated temperature. 



239. The construction of boilers for steam boats must be such as will render 

 them secure against danger from the fire, and also with as little of either bulk or 

 weight of materials as possible. When they are low pressure boilers, and I 

 would strongly recommend that no other should be used at sea, the force of the 

 steam does not prevent the use of plane surfaces, to bound the flues and fire. The 

 object then is to arrange the fire-place and flues within the boiler, so as to afford 

 the proper quantities of fire, and flue surface, and of capacity, and admit of being 

 cleaned with facility. 



Various methods are adopted, but I have observed that the common tendency of 

 a few years' practice is to simplify both the construction and the means of ob- 

 taining effect. 



240. The boiler is sometimes made so as to admit a clear passage of about 

 eighteen inches between the timbers and the boiler ; but this excellent practice is 

 by no means so general as it ought to be ; for it not only gives a great degree of 

 security against accident by fire, but also renders the examination and repair of 

 the boiler easy and satisfactory. 



241. The grate should not be less than about 2 feet from the floor, and the 

 sum of the areas of the flues of the fires should be somewhat larger than the area of 

 the chimney, or simply larger than the chimney when there is no more than one 

 fire. It will be an advantage to have as many separate fire-places as is convenient, 

 for several reasons. First, The fire is easier to manage, and a less interruption to 

 the generation of steam is caused by feeding it. Secondly, The flue and fire 

 surface are obtained in less space, because two flues have more surface than one 

 capable of conveying the same quantity of smoke. It is, however, scarcely pos- 

 sible to point out the limits which should determine the choice in different cases, as 

 first expense is too often avoided, under the impression that it is more than 

 equivalent to an unknown loss, which will become as regular as it is certain. 



242. A flue in about the proportion of 12 inches wide, and 18 or 24 

 inches high, with one of its ends easily accessible, is a good proportion ; height 

 rendering the flue more effective than width, in consequence of the hottest part 

 of the smoke pressing against the upper part of the flue, while the bottom gets 

 speedily covered with a coat of sooty matter, which being a bad conductor of heat, 

 the bottom surface has very little effect. Hence, in estimating the quantity of 

 surface, the bottom of the flue should not be calculated. 



243. The fire-place is necessarily surrounded by water, but there is no ad- 

 vantage in this ; for water is so rapid a conductor of heat, that it absorbs it too fast 



