78 Scientific Intelligence. 



2. The Boiling joints of the Alkaline Metals. — On account 

 of experimental difficulties, the boiling points of sodium and 

 potassium have not been accurately determined, and no attempts 

 have been made to determine this constant for lithium, rubidium 

 and caesium. Ruff and Johannsen, having- found that the 

 vapors of these metals do not act upon metallic iron, have dis- 

 tilled the metal in wi*ought-iron retorts consisting of long tubes 

 enlarged at the bottom, and have measured the temperatures of 

 the vapors by means of a thermo-electric couple protected by a 

 steel tube and introduced into the vapor. Accurate results were 

 apparently obtained in all cases except that of lithium, which 

 boils at too high a temperature for the application of this method. 

 The boiling-points found were as follows : caesium, 670° ; rubid- 

 ium, 696° ; potassium, 757^° ; sodium, 877-J° ; lithium, higher 

 than about 1400°. When these boiling-points are plotted with 

 the atomic weights a curve is produced which slowly rises from cae- 

 sium, through rubidium to potassium, then becomes steeper to so- 

 dium, and finally rises very steeply to lithium. The curve is quite 

 similar to that of the melting-points, but it does not permit the 

 calculation of the boiling-points as a function of the atomic 

 weights by means of a simple mathematical formula. — Berichte, 

 xxxviii, 3601. h. l. w. 



3. The Rusting of Iron. — The slow oxidation of iron at 

 ordinary temperatures was at first regarded as a simple process, 

 but more recently iron has been generally supposed to rust 

 through the combined action of carbon dioxide, moisture and 

 oxygen. Dunstan, Jowett and Goulding have recently per- 

 formed numerous experiments upon this subject, and their results 

 show conclusively that iron, oxygen and liquid water are alone 

 necessary for the rusting of iron to take place. Under atmos- 

 pheric conditions carbon dioxide plays a quite subordinate part 

 in this process. It appears that the simplest representation of 

 the process involves the intermediate production of hydrogen 

 peroxide ; for example : 



Fe + H 2 0=FeO + H 2 ; 

 H„ + 2 =H„0 2 ; 

 2FeO + H 2 2 = Fe 2 2 (OH) 2 . 

 Although the authors showed that hydrogen peroxide was formed 

 in the similar oxidation of several other metals, they did not 

 prove the existence of this compound in the case of iron. From 

 the analysis of a number of samples of rust, it appears that its 

 composition corresponds closely to the formula Fe 2 0„(OH) 2 , or in 

 another form, Fe 2 3 .H 2 0. — lour. Chem. Soc., lxxxvii, 1548. 



h. l. w. 



4. The Cause of Color in the Iron- Cyanogen Compounds. — 

 After having made a study of the various blue compounds con- 

 taining iron and cyanogen, Hofmann and Resenscheck have 

 reached the conclusion that the cause of the color is the presence 

 of iron atoms in different states of oxidation in the same molecule. 

 They call attention to the fact that an even more intensely col- 



