456 E. Botlie on Oxygen. 



oxygen is absorbed with formation of binoxide of barium ; on 

 heating this to a higher temperature pure oxygen is given off. 



The author has effected in this manner, without cessation, as 

 many as 122 alterations of oxidation and deoxidation, and thus 

 separated by an easy and practical process the oxygen from the 

 nitrogen of the air. 



The following method of Bothe for the same purpose is a mo- 

 dification of one devised by Tessie du Motay and Marechal*. It 

 is based on the following reactions : — Peroxide of manganese 

 mixed with soda, and submitted to a high temperature in con- 

 tact with air, is changed into manganate, 



MnO 2 + Na 2 6 + O = Mn Na 2 O 4 . 



This manganate gives up part of its oxygen at 450° under the 

 influence of aqueous vapour, 



MnNa 2 4 + H 2 0=Mn0 2 + Na 2 + H 2 + 0. 



The apparatus used by Bothe in his experiments is a cast-iron 

 cylinder a metre in length by a third of a metre in diameter. 

 This is divided into two parts by a plate perforated with holes. 

 In the front part of the crucible is a spiral iron tube in which 

 the steam is superheated ; after traversing this tube the steam 

 passes through the apertures to the middle of the retort and 

 acts on the manganese ; then, mixed with oxygen, it traverses a 

 worm where the steam is condensed, while the oxygen passes 

 into a gas-holder whence it can be used. When the reduction 

 is complete, a current of air is passed over the mixture of per- 

 oxide of manganese., which at a red heat becomes changed into 

 manganate, which can again furnish oxygen, and so on. 



M. Fizeauf, in continuing his researches on tbe expansion of 

 bodies by heat, has observed that iodide of silver forms an im- 

 portant exception to the ordinary law of expansion. The sub- 

 stances he examined were the chlorides of potassium, sodium, 

 ammonium, and silver, bromide of potassium, iodides of potas- 

 sium, mercury, lead, and cadmium. All of these experience a 

 considerable increase in volume by heat, an increase which is 

 even higher than that of the most expansible metals, such as 

 zinc and lead. Yet iodide of silver, which both in chemical compo- 

 sition and general relations belongs to this group, constitutes a 

 very important exception ; it contracts by heat and expands by 

 cold : the phenomenon is perfectly regular and continuous be- 



* Bulletin de Chimie, December 1867. 



t Comptes Rendus, vol. lxiv. p. 314. Bulletin de la Socie'te Chimique, 

 September 1867. 



