THE MANOMETERS. 35 



safer to employ the experimental correction rather than that calculated 

 from the known diameter of the tube and the supposed spherical form 

 of the meniscus. 



One great advantage of the practice of deriving the meniscus correc- 

 tion from the calibration data is the excellent means which it affords 

 of detecting faulty calibration. It is known that the best work in 

 calibration leads uniformly to an approximately fixed value for the men- 

 iscus, hence it is to be inferred, when another value is obtained, that the 

 calibration which gave it is erroneous. 



The inverse relation of the importance of the meniscus correction to 

 the volume of the gas which is measured makes it desirable to increase 

 the quantity of nitrogen in the manometers as far as may be done with- 

 out creating other difficulties of a serious nature. This has been accom- 

 plished by the form of manometer seen in Figures 19, 20, etc., in which 

 the volume of nitrogen is relatively very large. In the manometers of 

 this kind which are in actual use, a length of 1 millimeter in the wider 

 part is about equal in capacity to a length of 16 millimeters in the nar- 

 rower portion of the tube. The column of mercury which occupies the 

 closed end of the manometer, being in the narrow portion of the ma- 

 nometer, is not easily dislodged by tapping. In this respect, the instru- 

 ment seen in Figure 19, etc., is not inferior to the earlier form seen in 

 Figure 18. During a measurement of pressure, the whole of the nitrogen 

 is compressed into the upper and narrower portion of the tube, hence 

 the column of the gas is much longer under any given pressure in the 

 latter than in the former instrument, and the errors due to faulty deter- 

 minations of the value of the meniscus and of the amount of capillary 

 depression are correspondingly less serious in their effects upon the 

 accuracy of the measurement. 



Manometers like that shown in Figure 19 are designed more espe- 

 cially for the measurement of the pressure of concentrated solutions 

 where errors of meniscus tell heavily on the results, unless large vol- 

 umes of gas are used. In the case of dilute solutions, large gas volumes 

 are obviously less necessary as a means of minimizing such errors. 



The length of the wider portion of the second form of manometer is 

 varied according to the range of pressure which it is desired to measure 

 with the instrument; e. g., if the pressures in question lie between 4 and 

 6 atmospheres, the wide and narrow portions are so related that the 

 mercury meniscus will appear in the latter at some pressure slightly 

 below 4 atmospheres. In instruments designed for use with normal 

 solutions, on the other hand, the nitrogen is not all compressed into the 

 narrower portion of the tube until a pressure of more than 20 atmos- 

 pheres has been reached. 



No considerable dilution of the solution results from the larger vol- 

 ume of gas in such manometers, because, at the time of closing the cell, 

 a mechanical pressure — the so-called initial pressure — is brought to bear 



