On some Phenomena connected with the Boiling of Liquids. 85 



Wave-lengths of Lines of Magnesium Oxide. 



Automatic spectroscope Direct-vision spectroscope 



with micrometer-eyepiece. with new micrometer. 



5007 * 50060 



4997 4996-0 



4986 4985-5 



4975 4974-5 



4963 4963-5 



4948 4948-5 



4934 4934 



*4918 ? 4924 



4914 



Each of these results is the mean of five readings. A re- mea- 

 surement of the first two bands by displacement of the lines in 

 the opposite direction gave 5006*2 and 4996*1. 



To test the capabilities of the new micrometer on faint spectra, 

 two determinations were made of the wave-length of a line in 

 the oxide-of-carbon spectrum given by a vacuum-tube enclosing 

 coal-gas. The wave-length of this line I found before by mea- 

 surement with a three-prism instrument to be 5195. It was 

 obtained very faintly by a feeble discharge through the tube, 

 and was measured by using as reference-line the lines given by 

 a Bunsen flame. 



Two successive determinations (each the mean of four read- 

 ings) gave 5196 and 5196. 



The advantages of the new form of micrometer seem to be 

 (1) great precision in results, (2) convenience in use. My thanks 

 are due to Mr. Browning for the skill with which he carried out 

 my wishes in the construction of the instrument. 



XIII. On some Phenomena connected with the Boiling of Liquids. 

 By Charles Tomlinson, F.R.S. 



[Continued from vol. xlix. p. 448.] 



THE boiling-point of a liquid in an open vessel may be de- 

 fined as the lowest temperature at which its vapour can 

 have the elasticity of common air. But the temperature of the 

 liquid itself and that of its vapour under the terms of the defini- 

 tion are not necessarily identical. Many circumstances, irre- 

 spective of pressure, tend to raise or lower the temperature of 

 the liquid, such as (1) the nature of the vessel, since the liquid 

 may wet some surfaces more or less completely, and others not 

 at all ; (2) the state of such surfaces with regard to chemical 



* Too faint to be measured exactly. 



