58 ANNUAL OF SCIENTIFIC DISCOVERY. 



the chances of accident were remote; but that any repeated deflexion, 

 either at intervals, or continued so long as to induce a permanent 

 depression, must be productive of danger, which could only be adverted 

 by altering, or replacing the parts deficient in strength, and maintaining 

 a rigid supervision, whether of beams when loaded, or of parts of 

 machinery, or of railway stock after working. By such means, accidents 

 woiild be prevented, and a greater degree of confidence be established in 

 structures in which metal was employed. Loud. Mechanics' Magazine. 



DILATATION OF CAST-IRON BY SUCCESSIVE HEATINGS. 



The remarkable phenomenon that cast iron presents after being 

 heated, of not returning, on cooling, to its original dimensions, but of 

 presenting constantly an increase of this volume, and by consecutive 

 heatings and coolings, of acquiring a permanent volume, larger and larger, 

 was first observed by Prinsep, in 1829. This chemist found that a retort 

 of cast-iron, of which the capacity had been measured with care by the 

 weight of mercury it contained, gave the following results. Before ever 

 being heated, the retort contained 9.13 cubic inches of mercury; after the 

 first heating and cooling, the contents were increased to 9.64 cubic inches ; 

 and after three successive heatings to the fusing point of silver, the con- 

 tents were 10.16 cubic inches. The cubic dilatation produced then was 

 11.28 per cent., or a lineal dilatation of nearly 3.73 per 100. Since this, 

 there has been occasion to observe more frequently, and to investigate this 

 property of cast-iron. It has been remarked, in effect, that all grate-bars 

 which sustained a high heat became curved, little by little, that they elon- 

 gated more and more, until finally they would push out the bars that sus- 

 tained them. 



M. Brii, in a work he has recently published, entitled Researches on the 

 Calorific Power of the Principal Combustibles found in Prussia) has made 

 known some experiments on this subject. By the aid of numerous 

 measurements, he has found that its permanent length augments after a 

 heating, but that this augmentation was so much the less as the bar had 

 been heated more often, and finally ceased. Thus, a grate-bar of 3.5 

 feet in length, after three days of a moderate fire, had taken a permanent 

 elongation of three-sixteenths of an inch, (equal to 0.446 per cent. ;) at 

 the end of seventeen days, this elongation was seven-sixteenths of an 

 inch, (1.042 per cent,) and at the end of thirty days had reached thirteen- 

 sixteenths of an inch, (nearly two per cent.,) and did not yet appear to 

 have attained its maximum. Another bar, of the same kind, after a long 

 service, had preserved a permanent elongation of 1.2o inches, or nearly 

 three per cent. The bars, while in the fire, experience another elonga- 

 tion, whicli is temporary, and contract as the heat is diminished ; and it may 

 hence be concluded with M. Brix, that it is proper to give to each new bar 

 a play, longitudinally, of about one twenty-fifth of an inch,' or four per 

 cent., to allow for this permanent and temporary elongation. In all cases, 



