S3(> On the Affinity existing between 



of metallic oxide : and from the experiments hereafter to 

 be detailed, it will not appear rash to infer, that this in- 

 verse ratio of carbonation proceeded exclusively from an 

 approximation, however remote, to the state of diamond ; 

 arising chiefly from a new and more dense arrangement of 

 the particles of the coal, in consequence of parting with a 

 portion of oxygen. 



These experiments were arranged and conducted in the fol- 

 lowing order : — About 50 pounds of oxide of iron were pre- 

 pared, and thoroughly mixed, that there might not exist any 

 variety of quality arising from different preparations. This 

 was kept during the whole series of experiments in the same 

 temperature, that none of the results might be affected by 

 the moisture of the atmosphere. A parcel of hand-made 

 crucibles, all nearly of the same size, prepared from Stour- 

 bridge clay, with ground covers, made so as to form a water- 

 tight joint, were set aside for the experiments. These, 

 from time to time, before the introduction of the mixture, 

 were brought to a red heat : when in this state, the mixture 

 was introduced, the cover was slipped on, and the whole was 

 put directly into the assay furnace. This mode had not only 

 the advantage of facility, but, which in these experiments 

 is absolutely indispensable, prevented the carbonaceous mat- 

 ter introduced from being dissipated by the moisture, which 

 is always contained in the clay when crucibles are put into 

 the furnace in a green state. , 



The proportions of oxide of iron and oxide of carbon, 

 used in the most of these experiments, were oxide of iron 

 200 grains, of carbon 15 grains. 



These were intimately mixed, and put into a square of 

 thin paper containing about five superficial inches of mea- 

 surement, and productive exactly of half a grain of char- 

 coal : this, and 14y grains of the lubstance to be tried, 

 formed the quantity of 15 grains. So that in all these ex- 

 periments there is nearly l-2Qth of carbon of paper united. 

 When the experiments were directed to comparative views 

 betwixt the raw and coked materials, a quantity of the 

 former was used, that would, by accurate experiment, have 

 formed 15 grains of coal, or prepared oxide of carbon. The 



paper 



