On the meam of safety in Steam Boats. 7 



by a rod from its centre. I suppose this end to be conveniently four 

 inches long, and the other twelve inches ; of course the depression 

 of the short end one inch, will raise the long end three inches ; I 

 then firmly fix a bell, triangle of steel, or other sonorous body, as 

 large perhaps as the space ordinarily vacant above water, or occu- 

 pied by steam will admit, and fix its tongue or hammer externally, 

 near to it, projecting downwards from an axle, from which an arm 

 extends horizontally towards the float, perhaps one third the length 

 of the tongue, the under side smooth and flat. I then make the 

 said lever, (which extends from the float-rod,) to underlap this arm 

 one, two, or three inches, according to size ; but at about one half 

 its length from its fulcrum or support, I bend it downward, and at a 

 point that clears the arm of the tongue make an elbow joint, from 

 which a short piece rises at about an angle of 45°, to contact with 

 the said arm, at about two inches from its end ; and in the end of 

 this short part of it, I place a small roller to prevent friction while 

 moving on each other ; and I put in tlie angle of the joint a small 

 spring. When therefore the float pulls down the short arm of the 

 lever one inch, the long arm rising three inches, or in due propor- 

 tion, raises the ai^m of the tongue three inches, while it at the same 

 time recedes from the bell six or nine inches, as the proportions may 

 be ; and as the ends of the lever and arm describe opposite curves, 

 they separate at their intersection, and the arm being liberated the 

 tongue falls (and with the more velocity if the reaction be increased by 

 a spring above the arm) and the bell is struck, and the alarm sounded 

 precisely at that moment when the surface of the water will have settled 

 to a level beyond which it would become dangerous to allow it to fall. 

 The danger being perceived and the boiler consequently better 

 supplied, the float rises and the lever returns to its position under the 

 arm of the tongue by means of the joint, which closes a little by the 

 pressure of the end of the arm till past- it, when the spring throws it 

 out to its proper position for raising the tongue again. 



The lever does not, of course, remain in contact with the arm ; but 

 when the water rises still more than one inch, it recedes downwards 

 from it. If on the contrary, the water falls and raises the tongue short 

 of the point that would produce the blow, yet returns in consequence 

 of the seasonable restoration of the water, the roller in the end of the 

 lever is to permit this motion easily as the tongue rises and lowers. 



Thus independently of constant human agency and care, the water 

 in the boiler will itself, by its motion, give notice of its own undue 

 and dangerous diminution ; which is of the more importance as steam 



