•70 Oh some Covilhialions of Plaliiiai 



In a third experiment S*9 cubical inches of hydrogea 

 were absorbed, and 7*58 grains of platina obtained: the 

 barometer being al 30°, and thermometer 60^ Fahrenheit. 

 This result agrees very nearly with the preceding one. The 

 difference between these experinients is not very great. No 

 allowance has been made, in the calculations, for a small 

 quantity of hydrogen gas which would have been absorbed 

 by the water. There is some difficulty in making correct 

 experiments in this wav, as there is otten a slight film of 

 the revived metal deposited on the sides of the retort, 

 which adheres to it with great obstinacy, and must of 

 course nialerially affect the results. It seems probable that 

 the oxys^'u is rather overrated in the two last experiments ; 

 hut I am inclined to think, froin a comparison of all the 

 experiments, and making every allowance, that it can 

 scarcely be less than l6])crccnt. From calculations de- 

 rived from the experiments of Professor Bcrzelitrs, it ap- 

 pears that the peroxide of platina (which may be presumed 

 to exist in the above solutions) contains only 14-1 grains 

 per cent, of oxygen. 



During the first of the preceding experiments, I wit- 

 nessed some phaenomena which seein to illustrate in a 

 beautiful manner the attraction of cohesion. After the 

 retort had been filled with hydrogen gas, and suffered to 

 remain for some hours in water, a number of minute me- 

 tallic particles scarcely visible to the naked eye made their 

 appearance. By degrees they increased in size, and were 

 perceived to be in motion. They appeared to be flowing 

 from all parts of the circumference to the centre, where 

 they ultimately found their place of rest, and attached them- 

 selves to a small mass of platina. As they approached the 

 centre, thev moved with greater velocity. This phseno- 

 njenon continued for some time, until a considerable part 

 of the surface was covered with the metal. It appears to 

 succeed best when the aqueous solution of platina contains 

 but a snjall quantity of metal. The most jierfect instance 

 I have witnessed was when a cubical inch of the solution 

 contained only 7*8 grains of pialina. In other cases, it 

 may be presumed that the force of cohesion will be coun- 

 teracted by the gravity which the reduced particles of metal 

 w.ill soon acquue. 



I shall not attempt to set limits to the application of 

 hydrogen gas, as a means of determining the quantity of 

 oxygen in metallic oxides; or to say, whether it may not 

 in certain circumstances revive all this class of bodies. 

 This however seems probable, that it will offer a simple 



and 



