62 REPORT — 1869. 



Si. Thomas's Supply. — The supply for St. Thomas's is taken chiefly, if not en- 

 tirely, from a well iu the surface-gravel, and is pumped up into a reservoir, from 

 vphich it is distributed to the houses. 



This water was cloudy when drawn, but after the suspended matter had been 

 deposited, the water was poured off, and was then clear, transparent, and colour- 

 less, and contained, in an imperial gallon : — 



grains. 



Inorganic matter 25-6C 



Organic and volatile matter 1'82 



Total solid matter 27-48 



Chloride of sodium 5"o5 



Ammonia < 0-003 



Nitric acid 2-00 



o 



Hardness before boiling 20-2G 



„ after „ 7-73 



The inorganic constituents consisted of carbonate of calcium, sulphate of calcium, 

 carbonate of magnesium, chloride of sodium, nitrates, and a little oxide of iron. 



It is evident, from the large quantity of chloride of sodium, nitrates, and 

 organic matter in this water, that it is obtained from a well which receives sur- 

 face-drainage, and on this account, as well as from its hardness, is not a fit supply 

 for domestic use. 



A considerable portion of the district is still supplied by the Exeter "Water 

 Company. 



All these samples were taken in the first week in August 18G9, after a continu- 

 ance of drv weather. 



On the Decomposition of Carhonic O.vide hij Spongy Iron. 

 Bij I. LowTui\N Eeli., F.C.S. 



In a communication on the chemistry of the blast-furnace, made to the Chemi- 

 cal Society of London last June, the author mentioned a circumstance in con- 

 nexion with the action of iron on carbonic oxide, which, so far as he knows, had 

 not previously been observed. 



On exposing fragments of ironstone, either raw or calcined, containing in the 

 one case FeO CO^, and in the other Fe^ 0.,, to the escaping gases of blast-fumaces, 

 there was, when the current of heated gas had a temperature sufficiently elevated, 

 an impregnation of black matter, which was ascertained to be carbon. 



The heat required to produce a slight appearance of this change was apparently 

 a little above that of melting lead, but at that of melting zinc the deposition of 

 carbon was very marked. 



The explanation ventured iipon at the time was, that the oxide of carbon was 

 resolved into carbon and carbonic acid, even at the low temperature of 337° to 

 3G1° C, as may be expressed by the formula 2CO = C4-C02. 



It has since been suggested that the action in question might have been caused 

 by traces of hydrocarbons still existing in the coke employed in the sraelting- 

 process. 



To satisfy himself that this explanation was not the correct one, the author has, 

 upon several occasions, repeated in the laboratorj^ the decomposition performed by 

 the blast-fm-nace, employing carbonic oxide, prepared both from oxalic acid and 

 from ferrocyanide of potassium, iu the usual way. 



The author was induced to submit the results of these experiments to the 

 Association, not only from the interest thc^^ may have in a chemical point of view, 

 but iilso from the circumstance that it is, from recent observation, possible that 

 the decomposition of carbonic oxide at this low temperature may have a practical 

 value in the smelting of iron. 



To prevent the reactions which accompany the deoxidation of an iron-ore compli- 

 cating the inquiry, a quantity of calcined Cleveland ironstone was reduced to a 



