THE sTuucruiti; of the nuclkus. 



ll'.» 



dt/it + 278) =.29 dp/p, so that roughly one centira. of pressure decrement will 

 corres[)oml to a little more than one degree of temperature decrement in a dew- 

 point apparatus. More than 10 or 15 ceutims. ol" pressure difference will I'arely be 

 required. Perhaps a fixed exhaustion, together with a measurement of the apertui'e 

 of the corona produced, is the pi'eferable method in pi'actice. What is particulai'ly 

 remarkable about these air fogs is their enormous rate (say meter/second and more) 

 of instantaneous subsidence. This is, of coui'se, mei'e evaporation from the finer 

 cloud particles at the top downwards while the air is quickly recovering its original 

 tem})eratui'e. 



With regard to its bearing on the present experiments, it will be seen that the 

 neetl of the desiccators, D and //, figure 1, in the aljove experiments is clearly 

 [lointed out. If they are not used half of the experiments are liable to prove fail- 

 ures, while the specific effect of the presence of even unsatui-ated vapor is yet to 

 be ascertained. Slight heating of the receiver should remove the danger of col- 

 lapse of the kind here in question in a measui'e. Experiment bears this out, and 

 shows that the danger of convection encountered in this way is less sei'ious than 

 might be supposed. 



11. Toluol. — Two sets of experiments wei'e made for this body, one in the 

 spherical receiver (30 cm. in diameter), the other in the diffusion tower. The data 

 of tables 3 and 4 and the graph which follows (figure 4) show the I'esults. 



TABLE 3.— DIFFUSION. TOLUOL. GLOBE. PHOSPHORUS NUCLEI, NOT 

 REFILLED. EXHAUSTION, 76-59 cm. 



TABLE 4.— DIFFUSION. TOLUOL. TOWER. EXHAUSTION, 76-680111. PHOS- 

 PHORUS NUCLEI. CLEAN VESSEL AND FRESH NUCLEI FOR EACH OB- 

 SERVATION. 



