141 



think of to insure accuracy, and I watched every step of the 

 process, in order to avoid all sources of mistake or uncer- 

 tainty. 



" Two ores of lead, both of them galenas, but very diffe- 

 rent in their qualities, are found in the valley of Glenmalure, 

 in the county of Wicklow ; one a steel-grained, hard kind, 

 very refractory in the fire, taken out of an exceedingly hard 

 quartz; the other of a softer nature, more easily reducible to 

 the metallic state, and either plumose or cubical in its frac- 

 ture. I shall distinguish these varieties by the names of hard 

 and soft ores. 



" Aheap of hard ore, weighing one ton, was exceedingly 

 well mixed with the shovel. A heap of soft ore, also weigh- 

 ing a ton, was equally well mixed, and kept separate from the 

 former. These were intended to be separately smelted. The 

 blast-hearth being filled with its proper fuel, and now at a 

 sufficient heat, ore not taken from either of the heaps intended 

 for the experiment was occasionally thrown on the fire, and 

 worked in the usual manner, until the well of the furnace was 

 filled with melted lead, and began to run down the gutter of 

 the apron, or, in technical language, until we had a running 

 sump. Without this precaution, whatever lead would be pro- 

 cured in the subsequent smelting process might be attributed 

 to the greater or less overflow of the sump, owing to greater 

 or less pressure of fuel on the surface of the melted metal. 

 This method was further corrected by ascertaining the num- 

 ber of inches which the lead fell in the sump, in each case, 

 after the fire had been rempved from its surface at night. 

 Those who are conversant with smelting operations will 

 readily understand me. The height and quantity of the su- 

 perincumbent fuel pressing on the surface of the melted lead, 

 at the moment when the sump began to run, was accurately 

 observed ; so that by leaving ofi" with the same quantity of 

 fire, the weight pressing on the melted lead was the same as 

 at the commencement of the process, and thus no part of the 



