THE STRUCTURE OV THE NUCLEUS. 



55 



being longer (30 cm.) and adjusted by an a[>[)roi)iiat«; tangent screw, f.. In such a 

 case, the instrument as a whole must be swivelled around a horizontal axis, C, so 

 that both tangeucies may be quickly caught, s is the screen, n, n!, the needles, 

 G the axis. In practice, the jaws en, cii', are turned around C until they are nearly 

 horizontal, and the screen s vertical. 



The j-emainderof the appai'atus — receiver, in which the coronas are produced, 

 exhausting train, etc., Figured — did not differ materially from the forms given 

 above and need not, therefore, be specially described. E is the exhaust-pipe with 

 its cock, c ; F, the cottou filter with two stop-cocks, a and /; (the latter being a 

 check-valve insuring slow influx) ; G, the vacuum gauge. Q is the water level, 

 pp a fog line showing companulate coronas at N. 



3. liesults. — The first question to be answered relates to the sizes of the cloud 

 particles suspended in benzol. These are found frcnn measurements of the aper- 

 tures of the respective coronas. The results are given in tal)le 1, on the general 

 plan already detailed in the pi'ecediug chapter. The first three columns have an 

 obvious meaning. It is necessary, as stated, to shake the receiver to obtain annular 

 coronas without color distoi-tiou, and the apei'tures will not therefore sink below a 

 small value corresponding to the special nucleation so introduced. 



The foui'th column of table 1 gives the chord, .s>, (m the goniometer, subtending 

 the aperture of the coronas, whence the fifth column containing the diameters of 

 the water particles, d, in cm., is computed foi- the coronas examined, as explained 

 in the preceding paragraph. These results are plotted in the curve, figure 3, cor- 

 responding to the two parts of the table. The curves do not coincide, since the 

 initial nucleation is necessarily different, but they are of the same nature. Curve 

 II, which is the more conqJete, shows a sudden bi'eakdown beyond a, after 10 or 

 12 exhaustions, and it is here that the raining subsidence of cloud particles is 

 marked, and the nucleation thereafter, shaking presupposed, is appreciably constant. 

 The table shows that it would l)e possible to make an ai>pi'oximate allowance for 

 these extra particles. 



Why this rainy subsidence and purification begins in the sudden fashion 

 evidenced is an interesting question for fui'ther investigation. Something similar 

 will be noticed in the case of water below. 6^'. § 5, 7, 11. 



TABLE 



I— CORONAS IN BENZOL VAPOR, KEPT REGULAR BY SHAKING. 

 SULPHUR NUCLEI. EXHAUSTION, 76-58 cm. d„ = .000,044 cm. ; 

 computed d = d^io^' ; P — ■ 176- 



