108 Mr T. H. Easterfield, Some observations on  [Oct. 31, 
decomposition took place, oxide of manganese being probably pro- 
duced. This permanganate was then mixed with about half its 
weight of phosphoric pentoxide, and the mixture gradually heated. 
No change appeared to take place at the temperature of 100°, but 
when the mixture was still further heated over a small gas-flame, 
action suddenly began at one point, and rapidly spread throughout 
the whole mass, though the lamp was at once removed. 
So long as the action was going on, pink vapours were given 
off, which condensed to small solid particles in the neck of the 
retort in which they were generated, These particles, however, 
gradually liquefied to a solution of the characteristie perman- 
ganate colour. This liquefaction was probably due to the fact 
that the phosphorus pentoxide used contained a certain quantity 
of water, which appears to be driven off when the pentoxide is 
acted upon by the permanganate and to dissolve the solid perman- 
ganic acid, after condensing in the neck of the retort. As a fur- 
ther proof of this, drops of colourless water condensed on certain 
parts of the neck of the retort, The products of distillation were 
tested for potassium and phosphoric acid, but gave no trace of 
either. The pink liquid was faintly acid to litmus paper. The 
quantity of distillate obtained was very small compared with the 
quantities of phosphoric pentoxide and permanganate taken, 
The yield was slightly increased by mixing the ingredients with 
fine sand so as to prevent the too rapid spreading of the reaction 
through the mass. When the reaction had once begun, and had 
been allowed to run to a finish, the application of heat caused no 
further evolution of the pink gas, though a little more water was 
expelled. It was noticed that if the mixture of permanganate 
and pentoxide was allowed to absorb moisture, by exposure to the 
atmosphere of the room for a few hours, the yield of pink vapours 
was considerably increased. When large excess of pentoxide was 
used a considerable amount of it volatilised unchanged and only a 
trace of the solid permanganie acid was obtained. (No attempt was 
made to ascertain the composition of the fused mass which re- 
mained in the retort.) 
(ii) Dry potassium permanganate was ground up with boric 
anhydride and the mixture heated. The phenomena observed 
were almost identical with those noticed when phosphoric pent- 
oxide was used, so long as the boric anhydride contained a small 
quantity of moisture. The smaller the quantity of moisture pre- 
sent, the smaller the quantity of pink vapour and liquid produced. 
A little boric acid came off with the vapours, volatilising in the 
steam given off. In order to prevent any possibility of spirting in 
these experiments, the products of distillation were made to pass 
through a thick layer of glass wool, When a drop or two of 
