previous to subsidence, forms nuclei for the accrescence of other matter, and disposes the saline compounds, such as the sulphate of magnesia, chloride of sodium, to chrystallize and precipitate much sooner than they otherwise would."

These are Dr. Ritterbandt's own words, and he is perfectly right in the opinion conveyed by them, viz, that it is heat which causes the salts, both in marine and land boilers to pass from the soluble to the insoluble state.

The Doctor admits or rather declares, that the heat is the grand cause of the disease; and, in order to alleviate it, he puts into the boiler a quantity of ammoniacal salt. Now in this treatment of the patient I do not agree. Instead of attempting to alleviate, I have succeeded in removing the cause-the heat.

The temperature to which water is raised in the act of being converted into steam in a boiler is not sufficient to cause the salts to pass from the soluble to the insoluble state. This change is produced by the over-heated state of the metal of the common kinds of boilers; an over-heating which is not at all a necessary accompaniment of the converting of water into steam by the application of fire to a boiler.

The existence of minute globules of steam on the metal of boilers is the cause of its becoming over-heated. By constructing boilers so that currents of water are formed in them, which sweep along the heating surface, and remove from it the globules of steam the instant they are formed, the metal of the boilers never becomes hotter than the water it confines; and consequently, the salts in a boiler thus constructed can nowhere receive that amount of heat which is necessary to make them pass from the soluble to the insoluble state.

I have had a seven-horse power boiler constructed according to this plan, working for the last thirteen months; a portion of that time it was worked with sea-water, and the remaining time with fresh water, which naturally contains a quantity of lime; yet no deposit has ever formed within the boiler.

The fact, that the red-lead paint with which the interior of the furnace and flues were painted thirteen months ago, is still in good condition, is a most convincing proof that the metal never exceeds in temperature the water which it confines.

The subjoined accounts of some experiments on this subject would I think, be new and interesting to some of your readers, if you can. find room for them.

Willow Park, Greenock,

July 3, 1845.

JAMES JOHNSON.

Experiment First.-Put fifty cubic inches of sea water into a vessel, and place it in a water bath over a fire; the temperature of the vessel containing the sea water can never exceed 212° Fahr., as the intervening water of the bath prevents it. After the bulk of the water has been diminished by evaporation until there is only ten cubic inches of water remaining, then allow it to cool, and it will be found that no crystals or hard deposit of any kind has been formed: a considerable quantity of a soft flocculent substance will have settled to the bottom of the vessel; but the slightest motion of the vessel

causes it to move about, and on again applying heat to the vessel the flocculent matter rises up and is dispersed throughout the water. Let the evaporation now be carried on until there be between seven and eight cubic inches of water remaining, and then allow it to cool; it will now be evident, by the existence of a few small crystals, that the water is a saturated solution.

This experiment proves that no hard or injurious deposit will be formed in a vessel in which sea water is evaporated, until the water has become a saturated solution, provided the vessel containing the water be not heated beyond 212° Fahr.

Experiment Second.-Place other fifty cubic inches of the same sea water in a Florence flask over a gas light for evaporation; after the water has become thoroughly heated it will assume a milky white appearance, and immediately after the flocculent matter will become visible throughout the water, the same as in the first experiment.Continue the boiling of the water until there is about nineteen cubic inches remaining; if the bottom of the flask be now carefully watched. small specks of a hard scaly deposit will be observed forming; count their number, observe their size, and allow the flask with the water to cool; when cold it will be evident that no increase has taken place in the quantity of the hard deposit, but the soft flocculent matter wil have settled down on the bottom of the flask. The specific gravity

of the water will now be 1070, and it will float one of Twaddell's hydrometers to the 14°. Apply the flame of the gas again to the bottom of the flask; all the flocculent matter will immediately rise up from the bottom and be dispersed throughout the water, without any increase being made to the scaly deposit previous to the water having commenced to boil; but as soon as ebullition has fairly commenced, then the scaly deposit will begin to increase and go on as it did at first.

Now this experiment shows that scaly deposit is only formed during the time the water is boiling, and where the ebullition is greatest. There is another fact connected with this experiment which I must make known, If the hard deposit was formed in consequence of the water being saturated with matter, it is reasonable to suppose that the deposit would fall to the bottom in a circular or round shape, as the bottom of the flask is spherical, but this is not the shape that it assumes. The gas flame which I used was what is called a swallow. tailed burner; the top surface of the flame next the flask is a flat narrow strip, and, what is very remarkable, this hard deposit inside the flask is formed in the shape of a narrow strip crossing the bottom of it, and exactly coinciding with the flame outside. From this it is evident that the hard scale is formed in consequence of the overheating of the part of the flask acted on by the flame.

When the salt water comes in contact with the over-heated plates, they produce or cause a premature crystallization. The process is a production of salt from water which is not saturated.

Can any other circumstance connected with the working of boilers

but the over-heating of the plates be assigned as the reason why the water in marine boilers deposits salt, although it is not a saturated solution?

After a little reflection on these observations, the following question is likely to be suggested, and I shall therefore answer it.

As the salt is produced from water that is not saturated, by the over-heating of the plates; why does the salt not re-dissolve after the producing cause has been removed, by the boiler being allowed to to cool?

Now the fact is, that salt thus produced does re-dissolve into the water from which it was taken, provided another change be not allowed to take place; for after that change it is insoluble even in fresh water.

I frequently repeated the second experiment, and in all my trials of it but one, I found that the scale of salt formed, gradually dissolved in the course of twelve hours after the boiling ceased. The trial in which the scale did not re-dissolve at first excited my curiosity, but I soon found out the cause. The person that I sent for the sea-water on this occasion brought it in a rusty iron vessel, and the water had absorbed a considerable quantity of the oxide of iron from it. When the water was boiled, the oxide of iron united with the scale, and formed an insoluble compound. In the other trials pure sea-water was used, and it could receive no oxide of iron from the vessel in which it was boiled, as it was a glass flask; therefore in those repetitions of the experiment the scale dissolved when the water cooled. Marine boilers are themselves the source whence the scale receives the metallic oxide, and becomes an insoluble compound.

In making the above experiments, I observed that the coating of deposit, after being formed, did not continue to increase in thickness so rapidly as I supposed it would, judging by the rapid manner in which the first coating was formed, and the reason why it did not do so, is owing to the first coating being attached to a firm substance (the glass;) whereas the succeeding coatings required to attach themselves to the first coating, a substance which was, comparatively speaking, soft, and from which it was easy for the bursting bubble of steam to detach portions of the newly-formed scale.

On carrying on the experiments for a length of time, I found that there was a considerable quantity of those detached portions of scale moving about in the water. Now I think it will be admitted, that the same circumstances will occur in marine boilers: viz. that when scale has been and is continuing to be formed in a marine boiler, there will always be suspended in the water minute portions of detached newly-formed scale, and as you are aware that this scale has an affinity for iron, the only conclusion that can be drawn is, that those portions of scale are carried by the ebullition of the water to the comparatively speaking quiescent side parts of the boiler, where they have undisturbed freedom to satiate themselves with their favorite, the iron composing the shell, with which they soon become united in the form of a hard crust.

From a careful inspection of the two kinds of scale, it is evident

that scale from the flues of a marine boiler is of a hard crystalline nature, whereas that from the shell is of a soft, chalky structure,-facts which corroborate my experiments and theory.

I now consider it an established fact, that the premature crystallization in marine boilers, from water which is not saturated with salt, is caused by the local action of the over-heated plates on the water which comes in contact with the plates immediately after each bubble of steam has detached itself from the plates. Mechanics' Mag.

Embankment across the Valley of the Brent.

Mr. Colthurst exhibited and described, three sections of the embankment across the valley of the Brent, at Hanwell, on the line of the Great Western Railway.

The embankment, which was formed of gravel, was 54 feet in height; it rested on vegetable soil, beneath which was a thickness of 4 feet of alluvial clay; then occurred a bed of gravel, varying from 3 feet to 10 feet in thickness, resting upon the London clay which was traversed in all directions by slimy beds or joints.

The surface of the country sloped gradually towards the Brent, which was at a level of about 20 feet below the south side of the embankment.

The subsidence of the embankment commenced during the night of the 21st of May, 1837; the next morning, the foundation was discovered to have given way, and a mass of earth, 50 feet in length by 15 in width, was forced from beneath the north or lower side of the embankment, towards the Brent. For four months this protruded mass increased in dimensions, and the subsidence of the embankment continued, until the surface assumed an undulating outline, which, on being cut through, showed that the subjacent beds corresponded accurately with the curvatures produced at the surface by the dis turbance. The state of the seams or strata beneath the surface, was ascertained by sinking trenches at right angles to the embankment.

The symptoms of failure in the embankment, at this period, were confined to a subsidence of about 15 feet, with a fissure extending all along the top of the south slope, at the side opposite to where the foundation had yielded. From the dip of that fissure, Mr. Colthurst inferred the nature and inclination of a rupture of the ground under the embankment.

Immediately on the commencement of the slip, Mr. Brunel directed a terrace to be formed, on the swollen surface, at the north foot of the embankment; the weight of the mass thus placed, succeeded effectually, in stopping the further progress of the subsidence, which up to that period, had exceeded 30 feet. The swollen ground extended over nearly 400 feet in length, by about 80 feet in width, and was elevated nearly 10 feet, with a horizontal movement of about 15 feet. The general disturbance, ranged to a distance of 220 feet from the foot of the slope, towards the river Brent, the south bank of which, was forced forward about 5 feet.

At a meeting of the Institution of Civil Engineers, London.

The rupture of the ground beneath the embankment, was indicated by the crack near the upper part of the south slope.

In a letter received recently from Mr. Bertram, one of the engineers on the Great Western Railway, it was stated, that the Brent embankment had subsided very little for several years; indeed, from the nature of the material, there was naturally less sinking, than in loosely formed clay embankments; a coating of ballast from 6 inches to 9 inches in thickness, applied once a year, was found sufficient for all purposes.

The slips which occurred in embankments formed of clay, occasioned trouble at first, by their immediate effect on the road above, and the difficulty of adding material to them. Mr. Bertram had found in many such instances, in the London clay district, that a temporary measure, of forming the softened mass which had slipped down, into large raised beds or ridges from 8 feet to 12 feet wide, by dressing with the spade, surface punning, &c., had the effect of keeping rainwater out, allowing the raised parts to dry, and retaining the mass in its place, until better weather and matured arrangements, permitted. the more permanent proceeding of forming an extended footing and working up the mass with additional material, so as to fill up the space with an increased slope.

When the Acton cutting slipped about 3 years since, Mr. Bertram was induced (from the difficulty of bringing gravel to the spot, and the quantity of surplus stuff in the cutting,) to try burnt clay for the drains, for forming an open backing to collect water, and also for mixing with the soft clay in punning up again; from what he then saw, he gave a decided preference to that material, over any kind of gravel, for mixing with clay, to retain it in its place. When gravel was used, there was generally a slight subsidence and opening at the top, but with burnt clay neither occurred. The usual system pursued, was to form with that mixed material, continuous abutments and revetments, upon the original face, and in all cases to make sure of thorough drainage from the back.

He had always been able to trace an immediate connexion between courses of the septaria and the slips at Acton. Those courses were not sufficiently open to act as natural drains: he had made many surface and deep drains leading from them, but the quantity of water drawn off, was not equal to that which was obtained by the means before described.

At Ruscombe, he had removed the gravel stratum from the top, laying bare and well draining the surface of the clay, using the gravel as a footing or buttress below, at such portions of the cutting as had been forced up by previous slips; when there was under drainage from longitudinal culverts, that plan answered very well.

At that portion of the Sonning cutting, which slipped so suddenly two years ago, the stratum of gravel was found to be broken into, by an upraised bank or dam of clay, which, after much wet weather, kept a reservoir of water penned back, until it broke out the mass of clay, down to the next stratum: the dam had been cut across at different points in the slope for the purpose of drainage, and when that