144 



THE STRtrcnTRE OF THE NUCLEUS. 



B 



Cotton 



CaCl, 



Fig. 2. — Drying Tube with Nucleator, P. 



Fig. 3. — Tubulated Partition for Diffusion of Water Nuclei. 



partially so, and c :xik1/' open. Then let the jet pump be started and act until the 

 vacciim in 6' has a predetermined value (pressure decrement usually 50 cm.), after 



which c is closed. Suddeulj' 

 opening e exhausts the receiver, 

 AB, as shown by the gauge (r 

 (pressure decrement usually 

 al)out 16 cm.), upon which the 

 observations for diffusion are 

 made at once. The cock e is 

 now closed and c opened for 

 further exhaustion of O, while 

 either filtered air or nucleated 

 air is introduced through F'; 

 or nucleated air may be intro- 

 duced through F' to the bottom 

 and filtered air subsequently 

 through F to the top of the 

 receiver AB ; or vice versa. All these adjustments are at times needed. To clean 

 the vessel of nuclei by aspiration, which is by far the most effective method, the 

 cocks a' b' f, e, c, are all left open while the suction proceeds for 10 or 15 minutes. 

 AVitli all reasonable precautions it is often very difficult to entirely exclude 

 water vapor, though its presence is not always a menace. Nuclei of water vapor 

 in the above sense seem to occur simultaneously with the nuclei of hydrocarbon 

 vapor, and they tend in a measure to remain independent of each other. But the 

 former diffuse instantly and fill the vessel, draining the fog bank of nuclei, and 

 soon wiping out the plane of demarcation. It is neces.sary in all such cases to clean 

 the vessel by aspiiation and begin over again. Very slow influx of air is a safeguard 

 both against imperfect filtration and against the danger of errors from convection. 



RESULTS. 



9. Benzol. — The following data show the interval of time elapsed, in minutes, 

 and the height to which the nuclei have risen in this interval. In the case of 

 benzol, in particular, jiains were taken to l)egiu witli a layer of nuclei as near the 

 bottom of the ve.ssel as possible and then to find the rise in periods of 15, 30, and 

 60 minutes. Accordingly, in the case of the ipiarter-hour periods, the rise was first 

 studied as dated from the time when the nucleation was introduced. The height 

 reached by the fog bank is here 25 to 30 cm. The endeavor was then made to 

 adjust the lower level as near zero as possible, by one or two auxiliary exhaustions, 

 made after the nuclei had been introduced. The rise of the fog bank in this case 

 is naturally smaller. It reaches a limit, if time is taken for the fog to subside with 

 eacii exhaustion, while the luiclei which have escaped precipitation by evaporating 

 their loads are available for further diffusion. Tiiere are many of these when the 

 subsidence is slow, and this practice was therefore followed in most of the succeed- 

 ing experiments. Thus, in tlie last series of values for quarter-hour periods, the 

 fog on exhausting 2 min. after nuc^leation (at niin.) was 5 cm. high. Filtered air 



