146 



THE STRUCTURE OF THE NUCLEUS. 



In constructing the preceding graphs, figure 4 ei aeq., it will be assumed that 

 the rise in all cases is from the same lower level corresponding to the instant from 

 which time is measured. If the gi'aphs are straight lines the initial level sliould in a 

 measure he indicated by the graphs tliemselves. These curves give amjde evidence 

 of the difficulties of observation, but nevertheless map out a locus of a sufficiently 

 definite character to prove tliat moie than an incidental result is in ipiestion. 



AVhat is particularly remarkable is the number of times the nuclei rise after 

 apparently complete precipitation. Thus, in the first part of the table there ai'e 

 nearly 20 exhaustions, but the nuclei letain their rate of motion throughout. The 

 same is true of other parts of tiie tal)le. In case of the half-hour periods, the plane 

 of demarcation gradually loses definiteness, shading off to notliing at the top, while 

 the motion of the nuclei is eventually faster, their rise higher than at first. These 

 observations are marked a in figure 4. The nuclei which finally appear are thus 

 somewhat smaller, and there is a correspondingly slight gradation of size, but the 

 time needed to bring out this gradation shows that it must be almost negligible. 

 Thus the nuclei are all very nearly the same size. 



TABLK I.— DIFFUSION IN BENZOL. EXHAUSTION, 76 cm. TO 62.6 cm. PHOS- 

 PHORUS NUCLEI. DIFFUSION TUBE CLEANED BY 5 EXHAUSTIONS. 



I ','uarter-hoiir periods. 



