THE BAROMETEK. 



Fig. 2. 



whatever flows from the tube must return to the cistern. If the 

 magnitude of the surface in the cistern be very considerable 

 compared with the bore of the tube, and if extreme accuracy be 

 not necessary, the effects arising from this cause 

 will be too minute to need any correction; but if 

 that extreme accuracy is desired, which is necessary 

 in barometers used for philosophical experiments, 

 then means must be provided for keeping the mercury 

 in the cistern at a fixed level, or for measuring the 

 change of level. 



In fig. 2, the cistern A B is represented as having an 

 index at p, showing the point at which the level of 

 the mercury in the cistern should stand. A screw is 

 represented at v, by turning which the bottom can 

 be elevated or depressed, so that when the level in the 

 cistern falls it may be raised, or when it rises it may 

 be lowered, and thus the level may always be 

 adjusted so as to correspond with the point of the 

 index. The scale represented at D E is divided 

 with reference to the level determined by the point 

 of the index p. 



8. In most of the uses to which the barometer 

 is applied the variations in the altitude of the mercu- 

 rial column are very minute, so minute indeed that 

 the most refined and delicate expedients are neces- 

 sary in the observation and measurement of them. It 

 is a fact familiar to every one, that all bodies, and more especially 

 fluids, expand and contract by changes of temperature. Mercury, 

 like others, expands when its temperature is raised, and^ contracts 

 when it is lowered. Now when it expands it becomes bulk for 

 bulk lighter, and when it contracts it becomes bulk for bulk 

 heavier. It follows, therefore, that the barometric column, even 

 when its altitude is the same at different times, will have different 

 weights, the weight being less when the temperature is higher, 

 and greater when the temperature is lower. 



The dilatation and contraction of mercury has been ascertained 

 to amount to the 9990th part of its entire volume for each degree 

 of temperature by which it is raised or lowered, and a column of 

 30 inches would therefore suffer a change of the 333rd part of an 

 inch for each degree of temperature. 



The extreme variations of temperature in this climate between 

 summer and winter being about 50, the barometric columns which 

 indicate equal pressures at the two extreme temperatures, would 

 differ in height by about the sixth or seventh part of an inch. 

 When barometric observations made in different places are 

 180 



