114 THE STRUCTUKK OK THE NnOLEUS. 



5. Hydrochloric acid. Concentration. Supersatnration. — The two concentra- 

 tions of 3 fc and 6 fil used show results wliieh, as seen in figures 3 and 3', need 

 a detailed investigation to interpret them. The effect of pressure difference is 

 miK-h more marked and iriegular in case of the G fr than in case of the 3 '■; solutions. 

 Since the operation of shaking can be only roughly controlleti, the endeavor must 

 be made to smooth the data by tiieory. 



The pressure effect can easily be explained, as the jiarticles will be laiger and 

 the coronas smaller for the greatei' pressure differences, in view of greater super- 

 saturation. Since the lunnber, n, of nuclei is fixed for a given intensity of shaking, 

 if //( l)e the nniss of water precipitated per cub. cm., the equation is nd'^n/^ = m, 

 or (/ =: -^ t) nt/n n, where in vanislies with the [)ressure difference. 



6. Number of particlen.—'Yi) determine m, the equation for the constant 

 entropy, -S', of the mixture in the ratio of *'/(! — J') of vapoi' and liquid, 



S= C\\\d-\-xr/d, 

 where 6' is the specific heat of the licpiid, 6 its absolute temperatuie, /• its latent 

 heat of evaporation, may be used. Initially (before exhaustion), a? = 1 and since 

 6^= 1 nearly, the conditions before and after exhaustion may be determined. The 

 equation if solved for the ensuing x\ becomes 



^ and y' being the initial and final tenq)eratures, etc. If^and^)' be the corre- 

 sponding [)ressures, the adiabatic equation 



0'_ V (v-0/r_ p' >-. 



may 1)6 assumed as a first approximation. Furthei'inore, if (? = 273° + 20° and 

 p z= 76 be genei'ally taken as the point of departure, ;/ = 1.41 and (^y — \)/y = .29, 



6' = 293(^776)-^ 

 After allowing for the heat due to condensation as shown above (Chapter II, 

 §§ 16 et seq.), the following data approximately express the grams of water con- 

 densed per cub. cm. of saturated air at 20°, for the different sudden falls of jiressnre 

 fiom 76 cm. : 



Consecpiently as d, varies aa^^ni, the character of the vaiiation sought may be pre- 

 dicted. The conqmtation may be conqdeted by finding n from the observations 

 made at the pressure difference <Sj9 = 16 or Sp — 20 cm., as these are obtained under 

 the best conditions of measui'ement, the coronas l)eing sharpest and least diffuse. 

 W ith this determination the theoi'etical curve for </ results. The reductions show 



that 



n= 64 s'-'m X 10« 

 and the mean results for the two series in question are n = 56 and n = 45, for the 

 concentrations of 3 ^f and 6 ^. The individual data indicate that no stress can be 

 placed on the differences between mean values of n thus obtained. 



It is now possible to compute the curves for d fiom d = .324 -^in and 



