( 542 ) 



pressure of some millimeters of mercury caused a definite deviation 

 of the needle-shaped prolongation, the glass spring was fused into 

 a wide vessel A, as will be clear from fig. 2 withont further 

 description. On the wide tube A two marks have been made in 

 such a way that in case of equal internal and external pressure 

 the end of the needle is exactly between these two marks ; we may 

 also make a mark on one side, and a scale on the other side 

 to be able easily to determine eventual deviations from the zero 

 position after the scale has been gauged. As for our purpose an 

 accuracy of about 1 mm. of mercury sufficed, we have not made 

 any attempts as yet to make the manometer more sensitive. In our 

 opinion it is, however, certainly possible (among others with improved 

 arrangement for reading — see Jackson) to raise the accuracy to tenths 

 of millimeters of mercury. 



4. The preparation of nitrogen tetroxide and the filling of the 

 apparatus. 



The substance was prepared according to the method described 

 by GuYE and Drouginine by bringing NO, generated from a solution 

 of natrium nitrite by means of moderately diluted sulphuric acid 

 and dried by sulphuric acid and pho8})horpentoxide, into contact with 

 oxygen dried in the same way; then ths N,0^ vapours are condensed, 

 after a repeated contact with phosphorpentoxide, in a tube placed 

 in carbonic acid and alcohol. Generally two such tubes were used 

 in the preparation ; an inner tube fused into the two reservoirs then 

 afforded an opportunity to treat with oxygen the obtained substance, 

 which is coloured green or blue for the case that NO should be 

 present in excess in the gas mixture for a moment. The green or 

 blue hue e\entually present is easily and quickly removed, when 

 the substance has melted. 



Then the reservoirs filled with nitrogen tetroxide were melted off 

 (foi" which purpose beforehand capillary constrictions had been made 

 in the apparatus) and thus se\ered from the remaining part of the 

 preparative apparatus, wiiich was quite made of glass. In this care 

 was taken, that the reservoir continued to be provided with a tube 

 with phosphor pentoxide. 



In fig. 2 such a reservoir C is represented, which was sealed at 

 D to the rest of the apparatus, after a capillary point had been cut 

 open; in this operation the tetroxide is frozen by carbonic acid and 

 alcohol. A represents the manometer already described in ^ 3, ^ 

 the vessel which serves for receiving the tetroxide. The whole ap- 

 paratus was then connected with a GAEDE-purap at E, and evacuated 



