316 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 51 



curves of condition T x T 2 and T t ' T 2 ' and in fact the altitude will 

 be smaller for the first case than for the second. 



A clear idea of the operation of the method here used is obtained 

 from the following consideration. 



Assume that the so-called standard pressure of 76o mm prevails 

 at sea-level and that we ascend step by step in altitude so that 

 for each step the pressure diminishes 76 mm then each layer con- 

 tains one-tenth of the atmosphere that is present above the given 

 locality. For 760 mm of the mercurial barometer there is a 

 pressure of 10,333 kilograms per square meter. Each of these ten 

 sections into which we have in imagination divided the prism above 

 one square meter contains therefore 1033 kilograms of air or more 

 than a metric ton. If now the average specific humidity in such 

 a section is 4, that is if a kilogram of air contains 4 grams of water, 

 then there are 4.13 kilograms of water in that section. In a corre- 

 sponding manner we find that for each such section of the column of 

 air there are needed 1033 X °- 2 37S = 2 45-3 or in round numbers 

 245 large calories (kilogram-calories) in order to raise the tempera- 

 ture of the corresponding air by i° C. 



But these sections, each of which contains one-tenth of the total 

 column of superincumbent atmosphere, have remarkably different 

 altitudes. For instance, whereas the lowest sections, according to 

 the mean temperature deduced from balloon observations will 

 reach from an initial elevation of 20 meters up to 890 m. the second 

 section will extend from 890 to 1850 meters. 



In the above-described method of representation we imagine these 

 layers of differing altitudes all brought to the same thickness 

 exactly as if the air in each were compressed to the same density 

 and thus formed a so-called homogenous atmosphere. 



Conversely, the strata which actually have equal altitudes in the 

 natural atmosphere would in such an ideal case occupy very 

 unequal volumes since the higher strata would be crowded together 

 more and more. 



In order to keep the relations clearly in sight, lines of altitude have 

 been introduced in all the diagrams drawn on this system. These 

 lines of altitude are based on the average distribution of tempera- 

 lure deduced from the balloon voyages and therefore correspond 

 strictly only to the conditions presented by the curve t m in figures 

 43 and 47. 



Notwithstanding this last-mentioned limitation with reference 

 to the applicability of these lines of altitude, still in general they 

 furnish an excellent summarv view of the distribution of mass 



