cording to the quality of the iron, and some bars much more. In making such small castings as these experimented upon, you can always ensure a pretty sound casting; whereas, in making large castings, you cannot depend on its being perfectly sound throughout; neither is the iron itself so strong, the grain being generally larger and

coarser.

It has therefore always been my practice, as a general average of cast-iron, to take one ton as my datum for the weight that will break a bar one inches square and one foot long, supported and weighted as above, exclusive of the effect of the weight of the bar itself in producing the fracture.

Now, although it would have been more scientific to have taken a higher number or greater weight as a datum, and considered the weight of the beam itself as a part of the weight with which it was loaded, yet this condition would have involved two considerations; for it would always have been necessary to have found the weight of the beam itself, and subtracted that portion of its weight which contributed to its load, from the calculated weight which would produce fracture before the precise weight to be applied could be determined. I therefore chose the lower number, or less weight, as we practical men wish to arrive at our conclusions with the least trouble, and by the shortest rules.

I have, however, had such extensive practice in the use of cast-iron, and had so many opportunities of breaking cast-iron beams of all sizes, from a load of one ton to twenty, that I now place the most implicit confidence in my data, having found them, on large beams in general, true to less than the twentieth of the calculated weight; and more frequently the calculation has overrated the weight that produced fracture, than come under it.

It has also been my practice, as it is of every prudent and experienced engineer, never to expose a cast-iron beam, under any circumstances, to more than one-sixth of the weight that would break it; and, from calculations I have occasionally made, on castings made by other engineers, I find it accords very nearly with their general practice.

With this object in view, and to ensure security, I always prove every casting to double the weight that I intend there should ever come upon it; for I wish it to be distinctly understood, that I consider no casting safe till it has undergone a satisfactory proof; for a casting shall appear perfect and sound on the surface, yet air-holes and such like imperfections may be within,

and nothing but a severe proof can detect theni.

Having calculated the weight applied at the centre that will break the beam or casting to be proved, I take one-third, of that weight, and apply it to the centre of the beam, and while this weight is upon it, I have the beam forcibly struck all over its upper edge with a sledge-hammer, and afterwards all over its flat side with moderate blows by the same hammer. Previous to applying the weight, we strike a chalk line upon the beam, and after the weight has been applied, and the beam well hammered, we apply the chalk like again, and find what deflection has been produced by the weight. The weight is then suffered to remain upon the beam for a short time, and if the deflection does not increase, the weight is removed, and the chalk line applied again, and then, if the beam has recovered its deflection, we conclude it is perfect enough for the intended purpose, and it is accordingly sent off to its destination.

Should the beam, however, have not recovered its deflection, we conclude it has some imperfections; and it is again put to a more severe trial by hammering with the same weight upon it; and if the deflection does not increase, the beam is passed, but if the deflection does increase, the proof is continued, by adding more weight, and well hammering, till it gives way; and I seldom remember an instance of a beam giving way under such circumstances that did not turn out to have a material defect in it

With regard to the weight of the hammer with which the beam is struck, it has been in general discretionary; but the workmen who have made the proof for me have never been very nice about it, always taking the first hammer that came to hand, and not being very ceremonious about the use of it, concluding, according to my observations to them, that it is much better that the beam should be broken in the proof, if imperfect, than be sent off, and found defective afterwards

A good general rule may be, to have the weight of the hammer one pound for every ton, applied as a proof.

The deflection which one-third of the breaking weight occasions to a beam, varies from one six-hundredth to one five-hundredth of its length, according to the quality of the iron; but as a ready means of calculation for the proof-weight, I take the one five-hundredth of the length, which I find near enough for all practical purposes, observing that the deflection is directly as the weight. When a beam is loaded with about one-half its breaking weight, permanent deflection begins to take place, consequently

a beam should never be loaded to such an

excess.

I need hardly observe to you, gentlemen, that the strength of any beam is directly as the square of the depth and thickness, and inversely as the length, a truth you are no doubt fully acquainted with; but I thought it best to insert it, that the report on this subject might be complete.

The proving of beams of large dimensions is a very tedious and laborious operation, and was attended with great expense, till, having so much proving to do, I contrived a method by which the business was performed with certainty and despatch, and at a comparatively trifling expense One of your association (Mr. Smirke) has seen the apparatus, and witnessed a proof, and can speak as to the efficacy of it.

I have stated, that I never exposed a beam when applied to the purpose for which it is intended to a greater load than one-sixth of its breaking weight, but that I always prove it to twice that weight; take an example: suppose I find by calculation that the strength of the given beam, or the weight that would break it, was fifteen tons; I then apply five tons on its centre as a proof: now five tons on the centre, is equal to ten tons uniformly distributed over the whole beam; but the load is to be equal to or 2 tons applied at the centre, which would be equal to five tons uniformly distributed over the beam; consequently, the beam having been proved with five tons on its centre, will be equal to double the weight that should ever come upon it when the load is not more than onesixth of its breaking strength.

Having now given you a detailed account of my system of proof, and the several reasons thereof, which I considered as essentially necessary, that you might duly appreciate my proceedings, and also, that you might the better be enabled to draw your own conclusion on the following results, and further, that you might be assured I did not proceed at random, but that there was reason and method in what I did ;

Everything having been prepared as before described, I measured the beam we were about to prove, being the first from the centre beam on the east side of it, and I found the distance between the walls on which it rested, 35 feet, from which I calculated that it would require 47 tons 17 cwt. 1 qr. 17lbs. to break it; and consequently, the proof having to be one-third of this weight, there would have to be applied 15 tons 19 cwt. 15lbs. at the centre of the beam.

All parties present being now prepared, we commenced at twelve o'clock at noon to

put on the weights, taking the deflections at certain intervals (the particulars of which will be given in a table after the trial of the second beam, as one table will answer for both, and shew the effects on each at one view.) The operation was performed with as much dispatch as was possible, only just allowing time to take the various deflections; nevertheless it was a quarter past three o'clock by the time we had got on 15 tons 13 cwt. 1 qr. 16lbs. (the time employed being three hours and a quarter); when a question being raised whether there was not already more than double the weight upon it that could ever be brought upon the beam, I agreed to let the weight remain upon it all night, and decide this question in the morning; the deflection was now of an inch; the doors were locked, and every thing made secure, and the weight left to hang upon it all night.

The same parties that met yesterday having assembled again this morning, the beam being examined the deflection was found to be exactly the same as when we left it last night; the weight was removed, and the deflections noted with the same weights at those they were taken at in putting the weights on, when the whole of the weight was removed, and the scale taken of; the beam regained its original position within one-tenth of an inch. It was however found, that on applying the chalk-line to a line that had been struck on the beam, before beginning the proof yesterday, that it exactly corresponded, so that it was evident no permanent deflection had taken place; and the apparent deflection, as indicated by the diagonal scale, was accounted for, by supposing the weight to have settled the beam one-tenth of an inch closer, and more firmly down upon the wall.

The question now came to be decided, whether or not I had proved the beam to double the weight that could ever be brought upon it (for this being a very defective casting, I was desirous of proceeding to prove it up to twenty tons); and if so, whether it was advisable to proceed to such an extremity.

The principal difference amongst us was as to the weight that a dense mass of people would bring upon the floor: to decide this point, I had a square of ten feet struck out upon the floor, and by getting as many of the workmen together, upon a portion of it, as could stand close wedged together, it appeared that seventy people might stand upon a square of an hundred superficial feet; this gave 1.43, or about 1 superficial feet for each person; ten of the men taken promiscuously were weighed, and found to be 13

This I call 200lbs. per square foot for the greatest weight that can be brought upon the floor, for we must be secure against the greatest extremity; it is in such extremity, that security is most desirable; for should an accident happen in a crowded assembly, the consequences would be most deplorable.

Each beam supports a portion of the floor 35 feet long and 4 feet 9 inches wide, equal to 166 superficial feet, which, at 200 lbs. per square foot, will be 14 tons 16 cwt. 3 qrs. 14 lbs. ; the proof I had put the beam to was 15 tons 13 cwt. 1 qr. 16 lbs. suspended at the centre, which was equal to 31 tons 6 cwt. 3 qrs. 4 lbs., equally distributed over the beam; consequently it was evident the beam had been proved to more than double the weight that could ever be brought upon it; under these circumstances, I declined going to any further proof on this beam.

I proceeded to the proof of the next beam, which is nearly in the centre of the bow. The proof was made in the same manner as with the first beam; the weight, 15 tons 13 cwt. 1 qr. 16 lbs., was suffered to remain upon it all night, under a deflection of of an inch, and next morning we found the deflection had increased the part of an inch during the night; when the weight was taken off, and the chalk line applied as in the former case, we found the beam had recovered its original

98 lbs.

100 lbs.

198 ditto.

position, and we concluded that the apparent deflection had been from the beam bedding itself more firmly down upon the wall.

The following Table shows the weights upon the beams, with the deflection which those weights produced; it will be observed that the deflections on the removal of the weights did not come back to the same point as when they were put on; but this may be accounted for, first, from the beams having settled themselves a little down upon the walls; and secondly, from less time being given for the beam to recover itself, as the weights were taken off in about onehalf the time in which they were put on, for the beams ultimately recovered their original position. It may be further noticed, the deflections are nearly as the weight, and if the half of the weight of the beam itself was taken into the account, the deflections would be found to agree much nearer with the ratio of the weights.

The ends of the beam, No. 1, had no weight of any importance upon them; but the ends of the beam, No. 2, were built solidly into the wall, and the iron angular pilaster, with a weight of about 100 tons upon it, rested on the top edge; this circumstance no doubt gave a greater degree of stiffness to the beam, No. 2, as it will be found, on inspection of the following Table, to have had much less deflection with the same weight than the beam, No. 1.

PREVENTING ACCIDENTS ON RAIL

WAYS.

The repeated accidents which have occurred on the Grand Junction Railway since its opening, have produced considerable excitement amongst those connected with the railway system, both as proprietors and travellers. It has been stated that more serious accidents have happened upon this line during the short time it has been worked, than on the Liverpool and Manchester during all the years it has been at the service of

the public. Numerous plans have been from time to time suggested for the prevention of such mishaps, and not a few have been published in our pages; surely from the whole of them some efficient method could be designed. We add to the list of suggestions the following: Plan by P. Lecount, Esq. C.E., F.A.S. (From the Railway Times.)

"Sir-I shall be glad to lay before the public, through your valuable publication, the following plan for conducting railway travelling, which will, I think, render accidents to passengers next to impossible; and

I should be glad if those persons most interested in the question will point out any objection to it, that they may, if possible, be obviated.

"I propose, instead of attaching the engine close to the foremost carriage, to fix the train to it by a chain of a sufficient length to allow the train to be stopped by the brakesmen when any accident happens to to the engine. The engine first meeting all obstacles, and being subject to upsetting, getting off the rails, &c., is the means, as railways are now worked, of doing all the ultimate damage to the carriages; but by my plan it will be next to impossible, with a proper look out, for any damage to happen to the carriages at all. The chain should be fixed to a roller, so that it could be wound up when the engine backed astern to keep it clear of the wheels; and when the train approaches the station, the engine and train should be gradually approximated, in order that the train might be brought into the station; this can be done by the roller with ease, as I have found by trial, that I can draw up a train of loaded earth-waggons with one hand when being towed by an engine, so as to render the connecting chain quite slack.

"Yours, truly,

Februrary 26, 1838.

"P. LECOUNT."

Plan by R. Prosser, Esq., C. E.

(From the Birmingham Advertiser.) "No system of telegraphic signals of the ordinary kind can be serviceable on a railway, because that implies that the road should be divided into stations, and an accident occurring between stations could not be communicated without great loss of time, and the velocity which the trains acquire on a railway will not admit of this loss of time.

"Any system of light upon a railway would be useless in foggy weather, for no artificial light with which I am acquainted, can penetrate a dense fog; either of these methods are clearly inapplicable.

In tra

velling by railway, it has always appeared to me that too much was exacted from the engineer, who, in my estimation, has quite sufficient to attend to in working his machine. I would therefore propose that the engineer be placed, as in steam-boats, under the control of a captain or person whose sole business should be the direction of the train. I would furnish this person with as long a speaking-trumpet as it might be found convenient to carry.. I should place another person behind each train with a similar apparatus; if these trumpets were mounted upon light wheels, I believe that a velocity of at least 20 miles per hour could be com

municated to the apparatus, upon the principle of the velocipede.

"In the event of an accident occurring between stations, either one or both of these apparatus could be started for assistance, and the trumpet used as a means of alarm, long before his arrival.

"By applying the ear to the small end, in the event of the train to which he belonged standing still, he would be able to hear the approach of another train in time to prevent accident, and before they hove in sight, could communicate his wants. I should also recommend that each of the stations be supplied with two speakingtrumpets, which might be used in the same manner as the others, for speaking or hearing.

"The above is merely an outline of the method 1 propose to your consideration: I shall now procede to give you the proofs upon which I ground its recommendation: -Soon after the accident of the 9th of September, I had a speaking-trumpet constructed, and with an assistant, I have held distinct conversations at more than a mile distance; and, judging from the effect, this was by no means the limit, but I could not conveniently command a greater distance, or spare time from other avocations; of course, any artificial sound communicated through the tube would be heard at a distance depending upon the intensity of the original sound.

"I lay no claim to discovery, because I am aware that the speaking-trumpet is of remote origin, and that useful publication, the Mechanics' Magazine, No. 750, contains an extract from Blackwood's Magazine, under the head "Acoustic Telegraph," which I enclose, because it will shorten this communication, and it contains in a short compass much which I had quoted at greater length from Nicholson, Somerville, Lieu. tenant Foster, M. M. Biot, Arnot, Young, Perrole, Chladin, La Grange, Derham, Morland, Gough, Hassenfratz, Dr. Moyse, Monsieur Charles, Walker, Robinson, Hutton, &c. &c. The modern application of the Stethoscope is a beautiful illustration of a mode of rendering audible, sounds too feeble for the human ear to discriminate without artificial aid.

"This communication might be made much longer. I have said sufficient to put you in possession of my notions, and although I have apparently added to the expense of each trip, I am sure it will be found not really so; but I am quite confident that, by the adoption of this plan, a further sacrifice of life may be prevented, at a cost too trifling to form an objection, and furnishing a means of telegraphic despatch along the