ROSSBY DIAGRAM PLOTTING ROUTINE 15 
the expression for potential tempera- 
ture. But e is generally very small 
compared with p, and for this reason 
6a and @ do not differ appreciably. 
It is doubtful if, in practical meteor- 
ological work, it is worth while to 
obtain $a rather than §. However, 
if numerical accuracy is desired, the 
partial potential temperature may be 
obtained by entering as the ordinate 
of the point on the adiabatic chart, 
not the actual pressure, but the total 
pressure minus the pressure due to 
the vapor. This graphical process is 
facilitated by locating the actual point 
of pressure on the adiabatic chart, 
then displacing this point vertically 
upward by the number of millibars 
of pressure exerted by the vapor. 
On one side of the adiabatic chart is 
generally found a set of curves by 
means of which the specific humidity 
may be determined. A more conveni- 
ent method of computing this quan- 
tity will be given below. The reader, 
however, will find it helpful to dis- 
cover how to use the specific humidity 
curves on the adiabatic chart. 
It has been pointed out that the 
specific humidity may be computed 
from the formula: 
where q is the specific humidity ex- 
pressed in grams per kilogram of air; 
€, vapor pressure, and p, the pressure 
of the air. The units for e and p 
may be chosen arbitrarily since they 
constitute a ratio. They are gener- 
ally expressed in millibars, however, 
since upper air observations are 
transmitted in these units. The value 
of e, the vapor pressure, is readily 
obtained by multiplying the relative 
humidity by the saturation vapor 
pressure at the temperature of the 
particle. The saturation vapor pres- 
sure for different temperatures may 
be found in most texts on meteorol- 
ogy or physics. Thus if one wishes 
to find the vapor pressure‘at the tem- 
perature of 10°C and relative humid- 
ity of 50%, he finds in the saturation 
vapor pressure tables the value of 
12.28 mb corresponding to 10°C. 
Since the air is only 50% saturated, 
the vapor pressure is one-half of 12.28 
mb or 6.14 mb. Note that vapor 
pressure is entirely independent of 
pressure, being a function of tem- 
perature and relative humidity alone. 
After e has been determined, the 
specific humidity, gq, is calculated by 
means of the above formula. A slide 
rule is most convenient for this com- 
putation. 
It should be also noted that, strictly 
speaking, the specific humidity is not 
used as the abscissa of the Rossby 
diagram, but rather a very nearly 
equivalent quantity called the mixing 
ratio. This term is defined as the 
mass of water vapor per unit mass 
of perfectly dry (absence of water 
vapor) air. In algebraic form: 
mixing ratio (w) = 622 uae 
p—e 
From the above formula it is clear 
that since e is very small compared 
with p there will be no appreciable 
error introduced by the use of q in- 
stead of w. In fact, with the present 
inaccurate method of measuring the 
relative humidity in the upper atmos- 
phere with the hair hygrograph, it is 
ridiculous to try to obtain such an 
accuracy as the difference between 
the formulae for gq and w indicate. 
The use of the partial potential tem- 
perature and mixing ratio rather 
than potential temperature and speci- 
fic humidity as codrdinates of the 
Rossby diagram, was made necessary 
by the construction of the lines of 
