333 .Experimental Self Registering Barometer. [Oct. 



quite cold, this tube is to be fitted into a piece of wood, in which a 

 groove is cut to receive it — the vial part being received into a sort of 

 cup formed in the expanded upper part of the wood. To the lower 

 end of this, the float (also consisting of wood turned on a lathe to the 

 requisite size) is to be attached; and through the middle of it the 

 glass tube must descend to a level with the lower end. The wood and 

 glass are attached by means of wax, and it is necessary to coat the 

 float with the same, to prevent the mercury entering the pores of the 

 wood. 



About a foot above the float a cross piece of wood was fixed horizon- 

 tally, and from either end of this a wooden rod descends about four 

 feet— these rods ought to admit of being easily and firmly attached to, 

 or removed from, the rest of the apparatus. They support the wooden 

 shelf on which to place the weights which act as ballast, and it ought 

 to be broad to admit of moving the weights into the position that will 

 cause the whole instrument to float perpendicularly. The vial ought 

 to be filled with mercury before the tube is cemented to it — however, 

 this is not essential, as the mercury can easily be introduced down the 

 tube — care must be taken, however, to incline the tube as the mercury 

 rises in this, otherwise the great hydrostatic pressure may burst the 

 vial. The cistern which holds the mercury may easily be made out 

 of a common beer bottle, by cutting off the top narrow portion with a 

 red hot iron. The upper part of the float is considerably smaller 

 than the low T er, with the view to admit of the indications being mag- 

 nified. 



Compared hourly with a common barometer, the two rose and fell 

 simultaneously, but not regularly. This of course arose from the un- 

 equal diameter of the vial, and the irregular contour of the float— it was 

 intended that the indications should be magnified five times— but at 

 some hours the indications were greater, at others less, than those di- 

 mensions — this irregularity of course arose from the cause just mention- 

 ed — but, as already stated, all I contemplated, was merely to ascertain 

 whether a barometer on this principle will work, which was proved 

 most satisfactorily. 



This instrument, compared with the balance variety, recommends 

 itself for its great comparative simplicity of construction. 



The plan of bringing the centre of gravity below the float, would be 

 advantageously introduced in the variety of barometer shown by fig. 2, 

 plate 15, in the 14th Number of this Journal. 



Hoonsoor, 5lk October, 1837. 



