1887. ] Permangame Acid. 113 
The ring thus formed was insoluble in water, but readily soluble 
in strong hydrochloric acid, and was probably some oxide of 
manganese. 
The gas was at once decomposed by contact with mercury, 
causing a purple film upon the surface. 
An attempt was now made to obtain some clue as to the 
nature of this gas by passing through weighed bulbs containing 
water, and provided with a calcium chloride tube, and then 
titrating the pink liquid so obtained with oxalic acid and centi- 
normal permanganate. The results were extremely unsatisfactory, 
giving 1°6, 2'1 and 0°8°/, as the oxygen value in three experiments. 
Since ozone is perceptibly soluble in water, the ozone evolved 
from the retort may, by dissolving in the water in the bulbs, have 
caused the greater part of the observed increase in weight. 
Attempts were made to determine the density of this gas, but 
failed, owing probably to two very apparent causes: (i) The 
volume of the gas evolved was only a minute fraction of the air 
passed through the retort ; (ii) ozone was largely given off. 
Experiments made without passing air through the retort were 
quite as unsatisfactory; the yield of the required gas appeared 
smaller, whilst that of oxygen and ozone increased. 
In the vapour density experiments, the gases were first passed 
through a small quantity of pure redistilled sulphuric acid at the 
bottom of the second test tube in the apparatus already men- 
tioned, then through a conical flask, and finally into a specially 
made bulb. This bulb, which was capable of holding rather more 
than half a litre was of thin glass, and had two narrow tubes 
entering it at opposite ends, for the entrance and exit of the gases. 
The tubes were closed by small glass plugs, kept in their places by 
little pieces of caoutchoue tubing. The exhaust tube was con- 
nected with a Mariotte’s bottle, so that a regular current of air 
could be drawn through the entire apparatus when required. 
When the apparatus was set going, the manganese gas was at 
first all absorbed in the sulphuric acid through which it passed, 
and the acid became olive green, the colour deepening until the 
acid was saturated, after which the gases passed through appa- 
rently unchanged. In no case did the increase of weight in the 
bulb above that of the bulb filled with dry air amount to more 
than two centigrammes, and since ozone was no doubt present, 
satisfactory results could hardly be expected. 
A calcium chloride tube was placed between the aspirating 
bottle and the vapour density bulb, to prevent aqueous vapour 
diffusing back. The arrival of the gas was at once indicated by a 
change in colour of the calcium chloride from white to reddish 
brown. 
The yield of the gas under consideration appears to vary 
