{il\ THK STRUCrUKE OK THE NUCLEUS. 



The ratio inuy be foiiml geiieially. Since tlie micleatiou m = 6 X 790 x 10 " ^/ ttJ'^ 

 deuutes tlie number per cubic centiiu., yjn is the number per linear^cm., and ^\/n, 

 their distance apart. Otherwise l/n is the volume of each and ■\jl/n the edge of a 

 cube occupied by a single particle. Hence, 



^^ = J'~ ^^~I = 87. 



(i \6 X 790 X 10"" 



The edo-e of a cube containing one cloud particle and the diameter of that particle 

 are proportional quantities in the ratio of 87:1. The density of the suspended 

 water is thus on the average .8 X 10 '<', throughout. 



13. >Si3€ of the particle producinfj axial color. — It is tinally necessary to 

 make a similar computation for the uuml)er of particles active in the drum when 

 a.xial colors are seen. The correlation will be made liy means of identical 

 coronas. Writin g as usual, d = 1.44/s X 10'' =d^-^\/jS\ the initial diameter is 

 d = 1.44/(.s- X \^1/-A^ X 10''). Using the second series of axial colors, Chapter II, 

 table 19, as these are more complete, two distant coronas are eminently available 

 for compari.stm. These are the early member with the olive-green center and the 

 subsequent mendjer with an apple-green center. 



'r.ible 19 shows, (olive-green) yV = .368 2 = 8.5 |" 1 .\'=i.40 



(apple-green) A' = .1 13 3=18.5 [Ji'i/yV = 2.07 



Table 4 shows similarly (olive-green) (/ = 4.6 X io"< cm. </„ = 3.3 X 10"* 



(apple-green) d—Z.\ X 10"' cm. r/„ = 3.9 X 10"* 



It is hardly reasonable to look foi' l)etter agreement in two observations so 

 remote, and hence the mean value, J„ = 3.6 X 10 "* will be taken foi' the reductions 

 in case of the axial colors obseived in the drum. Hen ce, 



(7=3.6 X 10-4 Xx'l/.V. 



The precipitation per cubic centimeter in the drum was found to be 

 301 X 10-9 grams of water per cubic centimeter of moist air, the [)ressiire ratio 

 being smaller. Hence, finally. 



The data of table 7 were derived in this way. 



This is an average exhibit of the dimensions and numbers of particles active in 

 producing axial color. The final column shows theii' average distance apai't. That 

 the a])sorption phenomenon occurs for particles as large as d is in many ways sur- 

 prisinij, for it does not suggest optical interference. The particles are, a.s befoi-e, 

 thinly disseminated. The ratio of their distance apart and their diameters being 



VU'" 



= 113. 



d 



The density of distribution is somewhat smaller than before, but tiie length of 

 column more than makes up for the difference. 



If, therefore, a cube .04 cm. along each edge be imagined, this cube at the 

 outset contains one particle of the diameter of 300/10^ cm. Thus 113 cubes on 

 end would make a row of particles normal to, and 113 X H-^ cubes on end 



