220 MEMOIR OF EATON HODGKINSON. 



And I believe that if the subject were thoroughly examined it would be found 

 that the different appearances shown by iron when broken arise from the com- 

 binations of the causes producing fracture as often as from any change in the 

 texture of the material itself. This opinion was strengthened !)y various speci- 

 mens of irons broken, some with a fibrous fracture by means of a slow heavy 

 blow, and some with a crystalline fracture by means of a sharp, short blow. Mr. 

 May refers to the case of a cast-iron beam of a steam engine, vibrating hundreds 

 of thousands of times per annum, being as good at the end of 20 or 30 years as 

 when first put up. In this case, though the strain has been in opposite directions, 

 and constantly varying, still the vibrations have not weakened the beam. On 

 the other hand he says, " I have seen a cast-iron gun absolutely broken across 

 by many years' dropping pig-iron upon it." 



In order to facilitate the calculation of the strength of shoi-t pillars, Mr. 

 Hodgkinson has given the crushing strength of a great variety of timbers used 

 in practice. The above is but a hasty and imperfect glance at this important 

 paper, which appeared at the time when the railway system was developing 

 itself by means of the application of cast and wrought-iron pillars to the con- 

 struction of bridges, &;c. No engineer who has in future to deal with this sub- 

 ject must omit the reading of this paper. 



In the Philosophical Transactions for 1857 there is another paper by Mr. 

 Hodgkinson on the strength of pillars. The object here is to confirm the con- 

 clusions of the first paper by means of larger experiments, made by an appara- 

 tus three times as great as the apparatus used on the former occasion. Having 

 been unsuccessful in finding the weight producing incipient flexure, Mr. Hodg- 

 kinson devoted his attention to finding the breaking-weight, the deflection, and 

 decrement of length produced by the weight laid on the pillars. The pillars 

 with both ends rounded broke in one place, in the middle ; but the j^iHars with 

 both ends flat broke in three jDlaces — the middle and at each end. When one 

 end was flat and the other rounded it broke at one-third the distance from the 

 rounded end. 



The formulae in the fonner paper are here slightly corrected, as being more in 

 accordance with the results of larger experiments. 



Thus, in pillars whose ends are flat and well bedded, the formula becomes 



"03.5 .73.5 



^1.63 



instead of 



J)3.55 fp.55 



to= 46.65 



Li.T 



as given in the first paper. 



It is a matter of observation long recorded, both by Mr. Hodgkinson and 

 other experimentalists, that the metal in large castings is not uniform in density, 

 the density diminishing from the outside of the casting to the centre. Hence it 

 was justly inferred that the crushing, tensile, and transverse strength of large 

 castings would vary, being the greatest towards the outside and less towards the 

 centre. In cast-iron pillars of 2^ inches diameter, the crushing force varied from 

 39 tons per square inch outside to 33^ tons per square inch centre. Mr. Hodg- 

 kinson discovered that the difference in the strength between the outside and 

 centre of large castings is much less than in small ones. Colonel Sir Henry 

 James found that the central part of bars of iron planed was much weaker to 

 bear transverse strain than liars of the same size. By jilaning out three-fourth 

 inch bars from the centre of two-inch square and three-inch square bars, the 

 central portion was little more than half the strength of that from an inch bar. 



The fall of the railway bridge over the river Dee at Chester, when several 

 lives were lost, led Mr. Hodgkinson to investigate the position of the tension- 

 rods, which were intended as auxiliary supports to the structure. The particu- 



