32 THE STRUCTURK OK THE NUCLEUS. 



tiiiu', as sliuwii ill table 5. Hence, anything in tbe nature of spontaneous nuclea- 

 tion tloes not occui". 



Nuclei produced l>y sliaking, though removable with diHiculty on exhaustion, 

 come out after the lapse of one or two hours. Probably the accession of licpiid due 

 to eacii e.xhaustion is too small to materially att'ect the rate of subsidence, a circum- 

 stance suggested b}' the large coronas. The early decay of nuclei produced by 

 shaking is an indication of the absence of foreign nniterial in solution. 



Bethzol. — In case of this re-agent it will i)e expedient to describe the phenomena 

 more at length than was done for the i)receding liquids. 



17. Coroiuis. — When suljjhur or other nuclei are i)Ut into the globe containing 

 benzol vapor, the result is peculiar. Instead i)f ilistiibuting themselves homogene- 

 ously throughout the leceiver, they usually collect in a heavy band near the 

 l)ottoni. This is invisible until revealetl by the first exhaustion, when a heavy 

 sluggish fog-bank is seen, only a few centimeters high. After violently shaking 

 the benzol in the receiver, strong c.oi-ouas gradually show themselves in the succes- 

 sive exhaustions. They are always of the normal type, however, being white cen- 

 tered and giadually narrowing to a vanishing diameter, as the nuclei aie more and 

 more precipitated. Without shaking, the coronas are soon distoi'ted again, evidencing 

 as in the case of gasolene, etc., a marked tendency to stratification. Out of this 

 confusion, normal coi'onas may be regained at any time while the nucleatiou lasts, 

 and the droplets produced by shaking can have but a secondai-y influence on the 

 main phenomenon. Careful inspection of the experiment during exhaustion shows 

 that tiie earlier coronas are flashed through ; but the normal type alone persists, the 

 whole indicating drops of large size. A vanishing corona after exhaustion expands 

 and clears. 



IS. Jidtihc'd foij'i. — The most curious feature in connection with ben/ol, as 

 well as the preceding litpiids, is the subsidence of the invisible nucleated air imme- 

 diately after influx and without exhaustion. It is not necessary to load the nucleus 

 by special exhaustion. This cxpeiiiiuMit recalls in some I'espects the very slow 

 subsidence of clay j)articles in [)ure water, and its almost tempestuous descent in 

 ether and similar licpiids; but it is moi'e probable that the nucleus loads itself with 

 a definite amount of liipiid invisibly and without supei'saturation, and that the .suc- 

 ceeding arrangement of strata is gravitatitnial, the nucleated air being heaviest. 

 The subject will be fully treated in the next chapter. 



19. ]>ijfii>ii<>n. of iiHchi. — Anothei- feature easily observed with lienzol, is the 

 apparent issue of nuclei from the licpiid, already I'efei'i'ed to. Supposing the air in 

 the receiver to have been practically freed from nuclei, good coronas will appear in 

 5 minutes within the lower third of the globe, while the upper two-thirds is still (piite 

 clear. A single exhaustion will usually throw them out again. After 20 minutes 

 waiting the ves.sel is ipiite full of nuclei, and the coronas found are perfect through- 

 out. Several exhaustions are needed to clear the vessel. At the surface of 

 <lemarcation the coronas are campanidate, showing that the particles are graded, 

 with the smallest uppermost, either from evaporation or from belated precipitation. 

 Eventually the coronas return to the annular form, compatible with the uniform dis- 



