intensly heated and unsaturated steam can, by the projection of water into it, produce highly elastic vapor," and satisfied themselves that in no case was an increase of elasticity produced by injecting water into hot and unsaturated steam, but the reverse; and, in general, that the greater quantity thus introduced, the more considerable was the diminution in the elasticity of the steam.*

A fourth hypothesis has been advanced in a communication to this office, from Mr. N. Sawyer, mechanical engineer of Baltimore, which may perhaps deserve the test of experiment. It supposes the water in a steam boiler to be permanently thrown out of level by the unequal pressure on its surface, resulting from the escape of the steam through the throttle valve and at one end, and the consequent diminution of pressure at that point. This alteration of level, of course, exposes a portion of the boiler,to become unduly heated, and when the working of the engine is stopped, the restoration of level by gravity brings a quantity of water in contact with the over-heated metal, producing highly elastic steam, which, according to the author, may result in an explosion. The existence of the difference of level, here supposed, is supported by the testimony of Mr. C. Evans, who remarks, in an article on the causes of explosion, that "wherever the steam is taken from to supply the engine, there will be the greatest ebullition, and the water will be higher there than in any other part of the boiler."+ The Franklin committee, in their experiments to ascertain "whether, on relieving water, heated to or above the boiling point, from pressure, any commotion is produced in the fluid," found that, "on making an opening in the boiler, even when the pressure did not exceed two atmospheres, a local foaming commenced at the point of escape, followed soon by a general foaming throughout the boiler, more violent in proportion as the opening was increased." Though the difference of level, thus produced, could not it is believed, be sufficient to account for the production of a quantity of steam great enough to result in an explosion, yet the extent to which it would operate could be determined only by experiment.

An hypothesis which has been lately advanced and to which the attention of Congress has been asked, in a published letter from its author, addressed to the honorable John Davis, of the Senate, would not have been noticed here but for the latter circumstance. This explanation of explosion, so far as it can be gathered from the pamphlet in which it is set forth,§ attributes the phenomena to the action of liberated caloric, set free by relieving the presure under which its combination with water is stated to be alone possible. This hypothesis seems to have been suggested by the supposed impossibility of accounting for the phenomena of explo

* Journ. Frank. Inst., vol. xvii., p. 19.

† Pittsburg Morning Chronicle, March 2, 1848.

Frank. Journ. vol. xvii., p. 8.

$ Pamphlet on "The Causes and Effects of Explosions in Steam Engines" &c., by John Wilder, New York, 1847.

sion from the gradually increased elasticity of steam by heat. That a gradual increase of pressure can produce all the effects of the most violent explosions, has been conclusively proved by the experiments of the committee of the Franklin Institute; in which it was shown that the very effects which, in the pamphlet alluded to, are considered impossible to arise from such a cause, did actually follow the gradually increased tension of the steam.* The existence of one sufficient cause, fully supported by the experiments of men so distinguished for scientific ability as were the members of that committee, renders unnecessary a resort to vague hypotheses, unfortified by facts, having no foundation in physical analogy, and, as in the present case, based upon an assumption in contradiction of a well known physical law. Such hypotheses can only serve to divert the minds of practical men from the true causes of these fatal disasters, and thus still farther to embarrass the question of their proper remedy.

The question of the causes of steam boiler explosions, if ever solved, is to find its solution in the researches of men of true science, conducted by the order of the government and at its expense. The experiments necessary to its full elucidation require too heavy an outlay to be within the reach of the means of private individuals or even institutions; and the nature of the interests to be protected, have made them, by the acknowledgement of all, a proper object of public appropriation. The most valuable contributions to our knowledge of the causes of explosions have been made by the scientific labors of the committee of the Franklin Institute, undertaken at the request and prosecuted at the expense, so far as the apparatus was concerned, of the Treasury Department in 1831. The services of the committee were gratuitous, although rendered at the expense of much time and labor, and the chairman, Dr. A. D. Bache, the present superintendent of the coast survey, devoted the greater part of his leisure during four years to the investigation. Previous to commencing the experiments, the committee addressed a circular to every engineer known to them as connected with the practical application of steam, and who had any personal knowledge of the explosion of a boiler. The answers to these circulars, though containing many crude hypotheses, furnished a valuable collection of facts and suggestions which served to guide the researches of the committee.† The undersigned cannot but believe that the institution of a new series of experiments on the same subject, conducted in the same spirit and on the same liberal scale as those of the committee alluded to, would serve to confirm the knowledge already acquired and throw still further light upon this interesting and important inquiry.

The causes of explosions, as laid down by the committee, are: 1. Excessive pressure within a boiler, the a boiler, the pressure being gradually increased.

2. The presence of unduly heated metal within a boiler. 3. Defects in the construction of a boiler or its appendages. 4. The carelessness or ignorance of those entrusted with the management of the steam engine.*

ce,

The causes, so far as they are stated in the returns to this office may all be included under one or other of these classes, and they have accordingly been so classified in the summary found at the end of the table of explosions. [Appendix C.]

How far these causes are correctly assigned, the undersigned is not able to say, as the means by which they were ascertained are not known to this office. It is desirable that in case of future explosions, provision should be made for obtaining full and accurate information of the nature and causes of the accident. Entire reliance cannot be placed on the testimony of ordinary eye witnesses, however honest their intentions, and still less confidence can be reposed in that of the engineers and officers of the boat on which the accident occurs, as they are directly interested in showing that it was the result of no dereliction of duty on their part. But, improper influences aside, persons unacquainted with the nature and properties of steam and the steam engine are not competent to form an opinion of any value upon the causes of an explosion. Nor are those who have a merely practical knowledge of these matters in a better situation for this purpose. They have too often preconceived opinions or hypotheses to which they are anxious to make facts agree, and, without intention to deceive, they are apt to seize upon those facts which seem to justify their prejudices, and disregard others of a different tendency. It is to an investigation by scientific men alone, men who value hypotheses only as they are in accordance with, and serve to account for, facts, and who are ready to abandon them whenever and as often as they are found to conflict with well ascertained facts, that we are to look. for a satisfactory determination of the causes of these disasters. It might be made the duty of the inspectors of steam boilers, to keep an accurate record of the explosions which occur in their several districts; they might be authorized to procure an examination in every such case by persons having the necessary scientific qualifications, and required to report at stated periods to an appropriate bureau at Washington. The collating and comparing of the information thus furnished would doubtless lead to valuable deductions, and have an important bearing upon future legislation with reference to this subject. In a great majority of the cases embraced in the accompanying table it is known that no scientific investigation was had, and the returns, so far as the question of causes is concerned, are probably, therefore, based upon common report, or the mere opinions of engineers.

That "undue pressure within a boiler, gradually increased," is one of the most frequent causes of explosions has been proved, as well by recorded cases as by the experiments of the committee.

* General Report on Explosions, p. 6.

One of the points of inquiry to which their attention was specially directed, was to ascertain the sort of bursting produced by a gradual increase of pressure; and after several decisive experiments, they came to the conclusion that all the circumstances attending the the most violent explosions may occur without a sudden increase of pressure within a boiler.* This gradually increased pressure may result from the accidental failure of the apparatus intended to relieve it, or from this being intentionally prevented from operating by the reckless hands which sometimes have the control of it. Of the ninety-eight cases in which the causes of explosions are stated in the returns herewith submitted, sixteen (being 16 per cent. of the whole,) are assigned to this cause. The term "excessive pressure" in the table indicates these cases.

The cases referred to "the presence of unduly heated metal within the boiler" are also sixteen. The danger attending the overheating of the boiler arises from the diminished tenacity of the metal thereby produced, which renders it incapable of any longer sustaining the ordinary working pressure; and from the metal's being made itself a reservoir of heat, capable of generating an increased quantity of highly elastic steam, as soon as water shall be brought in contact with it. This overheating of the boiler may be occasioned by suffering the water to become too low, or by allowing sediment to accumulate on the bottom. The former cause operates by exposing a portion of the surface next the fire to the action of heat while uncovered by water, and the latter by interposing a medium of but low conducting power between the fire and the water, thereby suffering the metal to acquire heat faster than that heat can be conveyed to the water. In either case the metal is technically said to be "burnt," and even if no serious consequences should occur at the time when such burning takes place, still the tenacity of the metal is thereby permanently diminished, its thickness decreased, and the liability to explosion, therefore, greatly enhanced. In boilers where flues are employed their collapse is the ordinary result of this state of things.

The deficiency of water may be the consequence of the pumps being obstructed, so as to throw in less than the requisite supply; or from their being heated, so as to inject steam; or from their not being in action when the engine is stopped and steam blowing off. The sudden removal of these causes of deficient supply, while the intensity of the fire continues at the same time undiminished, is very likely to produce an explosion for reasons already explained. Hence the great number of these accidents which occur immediately upon starting the engine after a stoppage at a landing or for repairs. Mr. Evans expresses the opinion that "eight out of ten explosions take place just at the time of starting from a landing, the engine making generally but one or two revolutions."

Frank. Jour., vol. xvii., p. 225.

The diminution of tenacity from overheating was found by the sub-committee on the strength of materials to be equal to about one-third of the original strength. Article in Pittsburgh Chronicle before referred to.

The subject of deposits, in connexion with the over-heating of the metal of the boiler, is one of great importance, and one which is still comparatively open as a field of research. The Franklin committee made it one point of their inquiry to ascertain by experiment the effect of deposites in boilers. They admitted the collection and hardening of such deposits on the bottom of the boiler, and attributed the danger from them to the production of exfoliations of oxide, which gradually reduced the thickness of the metal, or to the weakening increase of temperature in the metal which they permit.*

With regard to incrustations and deposites in the interior of steam boilers, it may in general be admitted that these must differ with the character of the water used. In boilers using "hard" water, they consist chiefly of the carbonates of lime and iron mixed with oxide of iron; containing, besides, the earthy salts from the water. Boilers using ocean water are found to detect the differences existing in different parts of the ocean in regard to the composition of its water. It appears that certain "scales" which were taken from the boilers of the United States steamer "Marcy," and subjected to analysis by Professor Johnson, were found to collect in the boilers while running over the Bahama Banks; and the experience of the "Marcy" is confirmed by that of other steamers which have traversed the same tract of the Atlantic. Professor Johnson's analysis showed this salt to be di-hydrated gypsum, and led him to the application of the acetate of potassa as a solvent.†

The introduction of oil or fat into a boiler may result in the production of another class of deposites entirely different from the scales above alluded to, and which appear to result from a combination of fatty materials with earthy bases. Such an incrustation was found in the interior of a steam boiler at Burlington, N. J., and was submitted to examination by Professor Johnson, who supposes it to be a species of soap, formed between the earthy oxides and the acids of animal fat.‡

In streams which, like the Mississippi and its tributaries, flow for thousands of miles through an alluvial country, and which are subject to freshets, not unfrequently producing alterations in depth of from thirty to fifty feet, the quantity of earthy, calcareous, and other matters held in suspension is great, almost beyond conception. Mr. Cist gives the estimate of an intelligent engineer, that in a twelve days' trip on the Mississippi, the quantity of mud injected into the boilers was fifty-one thousand six hundred gallons by measurement, or over two hundred tons in weight. This calculation is based upon the supposition that the sediment in the water amounted to ten per cent., a rate said to be below the fact, at least as regards the Mississippi. This sediment collecting on the bottom of a boiler, becomes, owing to the admixture of calca

* Frank. Jour., vol. xvii., p. 233.

† Frank. Jour., vol. xliv., p. 226, (Oct., 1847.

Frank. Jour., vol. xliv., p. 398.

Cist's Com., Appendix [G] No. 1.