OF ARTS AND SCIENCES : FEBRUARY 14, 1865. 413 



The experiments from which the formulas are deduced were made 

 on pillars of smaller dimensions than those commonly used, and they 

 were very carefully cast from Low Moor iron, No. 3. The strength 

 of iron in small castings is usually greater than in large castings. 

 Low Moor iron has a very high reputation ; its transverse strength, 

 however, by Hodgkinson's experiments, appears to be less than that of 

 many other kinds of iron. The different qualities of iron are usually 

 classed as Nos. 1, 2, and 3, differing from each other in the appear- 

 ance and properties of the material. No. 1 includes the softest and 

 richest irons, those having the largest crystals, and containing gener- 

 ally the most carbon ; No. 3, the hardest and densest irons, those having 

 the smallest crystals ; No. 2 irons are intermediate between Nos. 1 

 and 3. No. 3 iron, although the least valuable, is usually the strong- 

 est. Castings made from a mixture containing a considerable portion 

 of old metal have usually greater strength than castings from a simple 

 iron (see Table 11.) The cheapest irons thus appear to be the most 

 suitable for making pillars. 



The experimental pillars, from which the formulas are deduced, 

 were cast in dry sand moulds, and most of them vertically, in order to 

 obtain, as far as possible, uniformity in the texture of the iron. This is 

 generally understood to be the best, as it is the most expensive mode 

 of moulding; but casting vertically does not appear to produce the 

 strongest castings. In the experiments made in 1856 at the Royal 

 Gun Factories at Woolwich, England, on the strength of cast-iron,* 

 fifty-three different kinds of iron were tested ; bars were cast 22 inches 

 long, and 2 inches square ; in one half of them the moulds were laid 

 horizontally, in the other half vertically ; in nearly all cases the bars 

 cast horizontally, when subjected to a transverse strain, had a greater 

 breaking weight than those cast vertically; the average being about 

 19.5 per cent greater. Founders prefer to cast pillars in moulds laid 

 horizontally, or nearly so, on account of economy ; there is, however, 

 much greater liability to imperfections in the castings than when cast 

 vertically. 



In the experimental pillars great care appears to have been taken to 

 make them straight ; a small variation from a straight line in the axis 

 of a pillar will, in general, weaken it materially ; this is a defect, how- 

 ever, which can be readily detected. 



* Cast-iron experiments ordered, by the House of Commons, to be printed, July 

 30, 1858. 



VOL. VI. 45 



