306 



Utility of the Symplesometer 



portion of this is placed a quantity of highly rectified oil, whose sur- 

 face is exposed to the free action of the pressure of the atmosphere. 

 Supposing, though it is not the case, that common air filled the upper 

 part of the tube above the oil, then a difference in the level of the latter 

 would be the joint effect of two distinct causes. Firstly, a variation of the 

 temperature would produce a change in the elasticity of the included air, 

 and consequently the oil would rise or fall accordingly. Secondly, 

 changes in the density of the atmosphere resting on the surface of the 

 fluid would likewise produce proportional alterations of its level in the 

 tube. It is only with the latter of these causes which we have to do % 

 some method must therefore be found by which the change due to the 

 former may be eliminated. This is done by the insertion of a scale so 

 graduated that the errors due to thermometric influence are thereby cor- 

 rected—hence the thermometer B is an essential part of the instrument, 

 and indeed regulates every observation. The first step in each of these 

 is to note the degrees which the thermometer shews, the scale D being 

 fixed and exhibiting these thermometric degrees, the scale C, which is 

 moveable by means of the knob E, is then set with its fleur-de-lis oppo- 

 site the observed point at which B stands, and the indications then read 

 off* from C give the tabulated numbers or heights of the column of oil 

 due solely to the pressure of the atmosphere. On the proper construction 

 of the scale C depends the eflftciency of the instrument, and it requires a 

 most elaborate series of experiments to determine its different points to 

 that degree of precision requisite in an instrument to be used for philo- 

 sophical purposes — for practical use, its indications not being then ne- 

 cessarily so minute, the same laborious construction is scarcely requisite. 

 Unless, however, considerable, nay great, care is taken in determining the 

 differences due to the effect of heat, the Sympiesometer is scarcely to be 

 depended upon. G is merely a register for marking the last observation. 

 The whole is contained in a box about 2 feet long, 4 in. broad and 1 lin. 

 deep, so that it is very portable. The preceding theory had suggested 

 itself to me while on board, and on arrival here and making inquiries as 

 to the real principles of the instruments, I found I had to substitute 

 nitrogen gas for the common air in the upper portion of the tube, that 

 haying been found to afford the most correct indications ; the principle 

 however was correct. The instrument I used was one by the original 

 patentee Adie, which, having been constructed shortly after his invention 

 vvas made public, was of peculiarly beautiful workmanship. The ther- 

 mometer was read off" to hundredths of a degree, and of course, the scale 

 C was corresponding, each single division being however considered as 

 double, so that in a space divided to fifths, the readings go as far afe 



i 



