THE HYDROSTATIC PROBLEM FOR THE ATMOSPHERE. 73 



Columns 20 to 24 give the determination of the pressures at the heights where the observations 

 were taken. The lines in the diagram required for this determination are not given in fig. 6. The 

 determination of these pressures is independent of the solution of the inverse problem, and dependent 

 only upon the knowledge of the heights of the standard isobaric surfaces (column 9). 



The discussion and the comparison of the examples 1 and 2 is important in con- 

 nection with the practical question as to the choice of method of observation, as 

 well as in connection with the theoretical question as to the choice of method for 

 representing the result of the observations. 



It is seen that from the point of view of the computer it is no advantage at 

 all to have the height instead of the pressure as one of the observed quantities. 

 Observed heights being always geometric heights, they have first to be changed 

 into dynamic heights, and then the average virtual temperature of the standard 

 isobaric sheets must be found by the method of conjectures instead of by the direct 

 way which can be followed when the pressure is one of the observed quantities. 

 When some practice is acquired, these conjectures can easily be made with suffi- 

 cient precision to make repetitions of the operation superfluous. But still the con- 

 venience of the direct method can not be attained. Thus, as far as the observa- 

 tions of pressure can be obtained with the same precision as those of height, the 

 observations of pressure should be preferred as those giving least trouble to the 

 computer. In no case the values of pressure should be left out in the publications 

 of the result of the meteorological ascents, as it is unfortunately sometimes done, 

 height being substituted for pressure as the result of the calculations. But espe- 

 cially we must warn, as we have already done, against the use of barographs with 

 height scale instead of pressure scale. For in addition to the increased trouble to 

 the computer, this method will give much less trustworthy results. 



On the other hand, it is seen that the calculation of the height of standard pres- 

 sures is in all cases easier than the calculation of pressure in standard heights. 

 This is equally true whether it is pressure or geometric height which is observed. 

 Even if it may be possible to further simplify the methods developed for calculating 

 pressure in standard dynamic heights, it is not probable that it should be possible 

 to attain the simplicity of the method given for calculating the dynamic height of 

 standard pressures. 



In the choice between the two theoretically equivalent methods of representing 

 the distribution of pressure, viz, that of registering the height of standard pressures 

 or that of registering pressure in standard heights, we have thus found an important 

 practical reason for preferring the first method, that of registering the height of 

 standard pressures. 



Whichever method of observation be used, and whichever method of repre- 

 senting the results be preferred, it is seen that the fundamental operation remains 

 that of drawing and interpreting the virtual-temperature diagram. What can be 

 done to facilitate this work will therefore be of the highest practical importance. 

 In this respect the hints given in the next section (55) will be useiul. 



