A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 



135 



TABLE 2.— PRECIPITATIONS AT DIFFERENT TEMPERATURES AND PRESSURE 

 DIFFERENCES, y = {p-p')/{p-p'-6p); p = 76. 



Since 7;; = nnd^/b, if there are w fog particles per cubic centim. each of the 

 diameter d, and since sd = D where 5 is the aperttn-e of the coronas with an 

 arbitrary goniometer and D the corresponding constant, 



sf-^wi = .524L)f' w, 



which is constant for a given nucleation. Thus the relation between 5' and 5 

 at (J/j = 22 cm. and 17 cm., respectively, may be written 



5'/-^ = {m/m'Y'^ = .952. 



In figure 4 the line s' /s has been constiiicted and the observations grouped 

 with reference to it, showing that the curve reproduces the experiments fairly 

 well. The exceptional cases are all too low; or, in other words, at the higher 

 pressure difference, 6p = 22, which requires a longer period of waiting after 

 influx ceases, relatively fewer nuclei are entrapped. From this one concludes 

 not only that from the meditun, if saturated, all the nviclei are ])recipitated at 

 dp = 17, but that at the higher pressure difference the time needed for adjust- 

 ment is excessive and that the time loss of nuclei in the receiver frequently 

 becomes appreciable. 



On the other hand, if Sp exceeds 22 cm. for the given apparatus, the con- 

 ditions of spontaneous condensation of dust-free moist air are initiated and 

 continue thereafter with increasing intensity for higher pressure differences. 



8. Precipitation per cubic centimeter. — To determine m, I have heretofore 

 proceeded as follows: In a mixture of x grams of vapor, y grams of air, and 

 i — {x + y) grams of water, the absorption of heat due to a rise of temperature 

 ds at constant volume was taken as C(i-{x + y) ) + hx + r dx/dB + cy, per 

 degree, where C, c, and h are the specific heats of water, air at constant volume, 

 and saturated vapor, respectively, and r the latent heat. Since h—C= dr/ds 

 -r/s, h may be eliminated. Again the absorption of heat due to a volume 

 increase dv, at constant temperattu-e is if y is the heat ratio r (dx/dv)dv + y^c 

 {y- i)dv/v. If the expansion is adiabatic the total heat absorption is nil and 

 the equation thus obtained may be reduced eventtially to 



(irf 



{rx/B + ((7(1 -y) + cy)lgB)dB + yc{y-i)d\gB 



