Nothing has as yet been said about the arrangement of the wires, and the charging of the barrels and tanks. Two long wires, one insulated, pass through the whole system of rubber tubes and their intermediate connections, and are eventually led up to the crib. An Able cartridge is now placed about the middle of the gas-pipe tube, which is in each barrel, and from the cartridge lead two wires, each about 22 inches long, one of which is fastened to the long insulated wire, and the other to the other long wire, as shown in Fig. 3 of Plate IX.

That portion of the gas-pipe tube which was within the barrel was perforated with holes one-eighth of an inch in diameter, and about 4 inches apart. This tube was filled with rifle powder. The barrels were filled through the bung with the nitrate of soda powder. By this arrangement it was supposed that when a cartridge was ignited, the quickly acting rifle powder would ignite the more slowly acting soda powder in many places, thus insuring the burning of all of it. When filled, the barrels were securely plugged, and an iron hoop was driven around each of them over the bung to prevent it from being driven into the cask from the pressure of the water tamping when the chamber and shaft were filled with water.

The wooden cask torpedoes were placed against the edges of the excavation, and about 8 feet apart. The iron tank torpedoes were distributed in the interior near the rock columns. When all were in position it appeared as though each torpedo was connected with the adjoining ones by an India-rubber hose. Each torpedo was braced and fastened in position by means of timber, so that when the chamber became filled with water, they would not float about.

It was feared that when the water was introduced into the chamber its pressure would crush some of the barrels, unless some counteracting force prevented. An airpump was, therefore, brought into requisition. In order to see that there was no leakage, air was forced through the hose to the various packages of powder. The test was satisfactory.

All the arrangements having been completed, it was publicly announced that at noon of April 23 the explosion would take place, and immense crowds collected on Telegraph Hill, and other prominent places in the city, to witness the novel sight. On the morning of that day a hole, about 5 inches in diameter, was cut in the iron case about 2 feet above low-water mark. As the tide rose the water flowed through the hole, and at noon of that day the chamber was two-thirds full of water. The indicator of the airpump at that time indicated a pressure of 5 pounds to the square inch. At 2 p. m., everything being ready, an insulated wire was connected with the one brought up though the hose, (the end of the other wire being in the water,) and a boat containing the Beardslee's magneto-electric battery and the coil of insulated wire pushed off from the crib, paying out the wire as we proceeded. When about 800 feet from the crib the wire was cut, and the end connected with one of the poles of the battery, the circuit being completed by a wire connected with the other pole, dragging in the water. One turn of the battery crank and the explosion instantly followed. A column of water, variously estimated at from 200 to 300 feet high, rose majestically in the air. The diameter of the body of water thus thrown up was probably 200 feet. Around the base of this column was another simultaneous outburst of water, probably 70 feet high, whose flood seemed to roll outward. High above the mass of water could be seen rocks and pieces of timber. The highest jet of water was that which came through the shaft, and appeared as black as ink.

An examination, made shortly after the blast, showed that the shoalest part of the water over the rock was but 14 feet deep at mean low tide. As this shoal place was but of small area it was thought to be the old dump-pile, as a pole could be thrust for several feet into the mass of broken rock of which it was composed. It was also thought that the tides would carry off a considerable portion of this broken mass; but after waiting several days and again examining the locality, it was found to be in nearly the same condition as immediately after the blast. It was therefore necessary that the contractor should take some active steps toward removing the débris. For that purpose he constructed a rake, (see Plate X.) It was made of heavy wrought iron, weighing 24 tons, and its general appearance was that of an ordinary garden rake, except that all its parts were very much heavier. It was 8 feet wide. Each tooth was about 28 inches long, and slightly curved, with a width of 3 inches and an average thickness of about 5 inches. The space between two teeth was 2 inches. Its handle was of heavy iron, and about 8 feet in length. When ready for use it was lowered through the well, which was in the middle of the scow, by means of three ropes, two of which passed over a windlass at the rear part of the well, and were fastened to the two sides of the rake, while the third rope, coming from a point on the front part of the top of the well, was attached to the handle of the rake. While in position the handle of the rake was horizontal and its teeth nearly vertical. To prevent its sagging to the rear when the scow was moved forward, a chain was attached to each side of it, and the two led forward to the bow of the scow, where they were fastened.

When a steam-tug attached to the scow was moved back and forth over the rock the rake generally scraped the débris into deep water, and the depth of water over

the rock was increased a few feet. Occasionally pieces of timber would float to the surface, and as they all proved to have been part of the crib-work, packages of powder were placed on the surface of the rock and exploded, with the hope of breaking up the remainder of the crib. The result was, that numerous other pieces, varying from a few inches to 12 feet in length, were disengaged from the mass of the débris, and came up to the surface. This led to the conclusion that at the time of the explosion the lower part of the crib-work, which consisted of a box filled with 200 tons of rock, must have broken in two, and fallen down bodily into the crater which the blast had produced. In raking over the rock, the small loose pieces of stone would be dragged along until the rake came in contact with the interlaced timbers of the crib-work. This would temporarily check the progress of the tug, and it was supposed that it 'caused the rake to jump over the obstacle, and form a new dump. A diver was several times sent down on the rock. He stated that it was broken to pieces, the largest of which would not measure more than half a cubic yard, and that their average size was about that of a hen's egg. He moreover reported that at his last descent he could find no timber. Charges of powder, varying from 25 to 75 pounds, were occasionally lowered to the spot and exploded, with results similar to those of previous blasts. At last, on the 25th of May, the contractor informed us that he thought he had obtained the requisite depth of water.

An experiment was made a few days afterward, by passing the rake over the surface of the rock, noting the depth of water at the time, and then ascertaining the height of the tide at the time of the experiment from the self-registering tide-gauge at Fort Point. It was found that the surface of the rock was uneven, and that many of the lumps did not have 24 feet of water over them at mean low tide. The contractor therefore resumed work as before.

It was found very inconvenient to go to Fort Point, about four miles distant, to examine the readings of the tide-gauge there, and consequently the contractor placed a floating gauge at Cozzens's Wharf in the city, a place directly opposite the rock. The zero of this gauge was determined from the observations taken during a few days, and this zero at that time was supposed to be mean low tide. From time to time during a few months after the blast the contractor requested that an experiment be made, which was done, resulting in determining that the necessary depth of water had not been obtained. In the mean time the method of testing the depth of water by means of the rake was not satisfactory to all the parties interested, and Lieutenant Heuer devised what he called a boom. (See Plate X.)

Two pieces of timber, each about 25 feet long and 12 inches square, were nailed to two cross-pieces of timber in such way that the space between the two long pieces was about 2 feet. These cross-pieces were secured at points about 5 feet from the end of the logs, and the whole constituted the floating or buoyant part of the apparatus. On the cross-pieces, and midway between the logs, holes were bored corresponding with those in the two cast-iron cylinders, standing on flat rectangular bases, which were screwed to the cross-pieces. The cylinders were 7 inches high with an interior diameter of 2 inches, the thickness of the iron being inch. Through the cylinders and the holes through the cross-pieces, gas-pipes, 2 inches in diameter and 30 feet long, passed. The gas-pipes hung vertically, and could be fastened at any desired depth by set screws attached to the cylinders. The lower end of each pipe terminated in a T, through which a bar of iron, 30 feet long by 2 inches in diameter, passed and was fastened. The gas-pipes were graduated, and when both were lowered the same distance, the iron bar was horizontal. Experiments were made with this apparatus, referring the soundings to the contractor's gauge.

As the rock was gradually removed and the required 24 feet of water at mean low tide was nearly attained, it became a question whether the contractor's gauge, the zero of which had been determined from observations taken during a few days only, was correct or not; and also whether the water-level at low tide at the gauges at Fort Point Wharf and Cozzens's Wharf, four miles apart, were practically on the same horizontal plane. Lieutenant Heuer, therefore, ran a line of levels between the two points, thereby, in fact, carrying the mean low-tide mark at Fort Point to Cozzens's Wharf. The contractor run a similar line of levels to test the accuracy of Lieutenant Heuer's work, and the results by the two differed by less than an inch. With these data, a United States gauge was established at Cozzens's Wharf, the zero of which corresponded, as nearly as possible, with the zero of the gauge at Fort Point. On comparing the two gauges at Cozzens's Wharf, it was found they differed over a foot, the United States gauge giving the highest reading. The latter gauge was not a floating one, but was merely a graduated rod fastened to a pile, and it was difficult to read it with accuracy in ordinary conditions of the bay, and almost impossible to do so when the bay was rough; still it was impossible to conceal the fact that this discrepancy existed between the two gauges, and although the experiment made on the 20th of August showed that there was 24 feet of water over the rock at mean low tide by the contractor's gauge, it was necessary to inform him that he had not yet fulfilled his

contract.

Other official duties required us to visit Oregon. We sailed on the 2d of September and returned on the 14th of October, and during this interval tidal observations were taken at all low tides by an observer employed by the United States. On our return it was ascertained that the work had been suspended by the contractor, who insisted that he had complied with the terms of his contract. A careful examination was made of the observations taken in the interval, which were compared with corresponding ones at Fort Point, and the result was a full confirmation of the fact that the contractor was mistaken. Upon being convinced of this, he acknowledged his mistake, and resumed the work of removing the rock.

The discussion of the subject was quite interesting, bringing out several points of importance not thought of in the beginning of the work, and therefore it is thought best to insert here the following extract of the official letters written to the contractor, and on the strength of which he resumed work on the rock:

"SAN FRANCISCO, October 20, 1870. "SIR: The first matter of investigation was the error in the position of the tin pointer of your gauge, which you supposed to read zero when at mean low water, (defined by the Coast Survey to be the level of average lowest low water.) From the discussion of the observations taken with your gauge at Cozzens's Wharf, from September 5 to October 14, inclusive, I find the mean lowwater line to be -0.357; that is, the tin pointer must be moved down 0.357 of a foot in order to correspond with the mean low-water line deduced from the above-mentioned observations. A mean of the 25 observations from September 5 to September 30, inclusive, gives the number -0.377 instead of -0.357.

"The mean low-water level deduced from this short series of observations is different from that deduced from a series of many years' duration. I therefore have compared the observations taken at the Fort Point gauge with the mean low-water line of the Coast Survey, determined from a mean of fourteen years' observation, and I find that during the twenty-five days commencing on the 5th of September, (the Coast Survey observations for October cannot be obtained until the end of the month,) the mean of the lowest low water was + 0.694; that is, the average height of the lowest low waters during that period was 0.694 feet above the Coast Survey mean low-water line. Therefore the difference between your tin pointer and the mean low-water line of the Coast Survey is 0.694-(-0.377)=1.071.

"In the following table will be found a copy of the low-tide observations taken in September at Fort Point, and at your gauge on Cozzens's Wharf:

"The number 1.071 represents the distance your tin pointer must be moved down in order that your gauge reading corresponds to that of the Coast Survey at mean low tide at Fort Point. The scale reading of the United States gauge at Cozzens's Wharf and the Fort Point gauge would be the same if the water between the two places at low tide were on the same horizontal plane, and they would be at mean low water 3.411, that being the reading of the Fort Point scale at mean low water. The mean of the low tide readings at Fort Point from September 5th to September 30th, inclusive, is 3.411 +0.694-4.105. The corresponding reading of the United States gauge is 4.435. The difference is 0.330. If the lower ends of the two gauges are on the same horizontal plane, as is believed to be the case, and if the observations were correctly taken, then this difference can only be accounted for by a slope of the waters between the two places. I must state, however, that the observations at the United States gauge were quite difficult in rough weather, and could not be relied upon within a tenth of a foot.

"Supposing that this 0.33 of a foot in your favor were allowed by the Chief of Engineers, you have still to remove 0.741 of a foot before the required 24 feet of water over the rock is obtained. The number 0.741-1.071-0.330.

"I state for your information that I have learned officially from the Coast Survey Office in Washington that the mean rise and fall of the tide at Fort Point is 3.76 feet, and at Meiggs's Wharf, near Cozzens's Wharf, it is 3.86 feet.

"A copy of this letter will be forwarded to the Chief of Engineers. Very respectfully,

"Mr. A. W. VON SCHMIDT, C. E.,

"Contractor for Removing Blossom Rock."

66

"R. S. WILLIAMSON, Major United States Engineers.

It is difficult to conceive that in an open bay like that of San Francisco there should be a perceptible difference of level between Cozzens's Wharf and the wharf at Fort Point; but being confident that the lines of level run by the contractor, and Lieutenant Heuer on the part of the United States, gave essentially the same results, and being unable to conceive of any other cause for the difference of 0.33 of a foot, except an actual difference of level of the waters at low tide between the two points, this 0.33 of a foot has been given in favor of the contractor, as, in fact, have all doubtful points been given to him.

Being convinced of the truth of the facts contained in the above letter, the contractor resumed his work with the view of removing three-fourths of a foot of the rock at such parts of it as the removal was necessary. On December 6, 1870, I received from him the following letter:

"SAN FRANCISCO, December 6, 1870.

"SIR: I have the honor to state that I have removed from the surface of Blossom Rock, since your last survey, 700 cubic yards of stone.

“So far as I am able to ascertain, I have 24 feet of water over the rock at mean low water.

"I most respectfully ask you to order a survey to be made for your own information, and hope that you will be pleased to accept the work.

"I would propose that at 10 o'clock to inorrow would be a good time for the survey, on account of the tide.

66

'Very respectfully,

"A. W. VON SCHMIDT, "Civil Engineer, &c., Contractor for Removing Blossom Rock. "Major R. S. WILLIAMSON, "United States Engineer."

In consequence of this letter we were at the wharf at the time appointed, where we found the tug, the scow, and the boom ready for us. The boom was set to the proper depth of water; on it stood Lieutenant Heuer with a lead line in hand, and the apparatus was very slowly and frequently towed over the rock, until every portion of it must have been passed over by the boom. The boom did not strike the rock, and every sounding indicated 24 or more feet of water over it at mean low tide. On the following day the work was accepted, and the contract money paid to the contractor.

The officer of the Coast Survey in charge on this coast, Assistant George Davidson, had intimated that, as soon as our operations were over he would make a survey and chart of the portion of the bay where the rock formerly stood. The survey was made, but before the chart was finished the subordinate officer of the Coast Survey who made the hydrographic survey became sick, and the work was delayed until he should recover. An examination of the soundings taken by the United States Coast Survey showed that there were four lumps on the rock which had not quite 24 feet of water over them. I therefore directed Lieutenant Heuer to make the survey of the rock.

METHOD OF MAKING THE FINAL SURVEY OF BLOSSOM ROCK.

A scow, 46 feet by 28 feet, was moored on the rock by means of four anchors and lines. Two observers on shore, and distant about one mile from the scow and from each other, took angles simultaneously with theodolites, on a flag staff erected on the SCOW. The direction of its axis at the same instant was determined by compass on board. As soon as the theodolite men had observed its position, two men with lead lines (one on each side and end) would sound all around it. When these sets of soundings were completed, its position would be slightly changed, the theodolite men would again be signaled, would observe its new position, and soundings would be made as before. This operation was repeated many times, until the whole surface of the rock was covered with soundings. We obtained over 1,800 soundings, about 1,300 of which were plotted.

After the soundings were reduced to mean low-water level and plotted, it was found that there were seven places where there was less than 24 feet of water. At these places there appeared to be small masses of loose rock, about the size of a cubic foot or less, rising above its general level. They were so small that it was often difficult to find them a second time. In previous examinations which had been made by us by means of the sounding float or "boom" we had found no place where there was too little water, and therefore the money had been paid to the contractor. As soon as the detailed survey above described was completed, the contractor was informed that these seven lumps still existed on the rock; he therefore, at his own expense, and with as little delay as possible, sent down a diver in armor, who found the lumps. He reported that the surface of the rock was uneven, like furrows, and was covered with stones, usually quite small, and seldom as large as a cubic foot. All the lumps on these ridges were scraped off into deeper water. He went down twice, remaining nearly an hour each time. He could only go down about slack water.

We made a resurvey of that portion of the rock which had contained the lumps, and found over each of them 24 feet or more of water. Some of these lumps may have existed when that survey was made which resulted in the money being paid to the contractor, because they were so small, and being detached masses were not detected. Some of them, however, were made during the recent survey, when, several times, rocks, used as sinkers to buoys, had been lowered on the rock, and left there by the parting of the lines to which the buoys had been attached. The survey could only be made when the bay was smooth, and only fourteen days suitable for the purpose were found during three months. The soundings over the rock at the final survey are shown on Plate XI.

CONCLUDING REMARKS.

In concluding this report a few words with regard to the manner of conducting the work, the defects in it, &c., may be of value, though they are the natural deductions from the operations already detailed. The main defect was in firing the charges of powder before a sufficient amount of excavation had been made. A glance at the original plan, as partly shown on Plate VI, (the stone pillars were to be replaced by wooden ones,) and the plan actually followed out as shown on Plate VII, shows at once that in the former case the amount of excavated rock was so great that when the shell of the rock was broken and lifted up by the force of the explosion, the greater portion of the débris would fall into the excavation, and there would be at least the. required 24 feet of water over the rock or débris; and in the latter case, the amount excavated was not half as much as the amount of rock remaining above the line of 24 feet at mean low water. Consequently, just after the blast there was only 14 feet of water over the débris. In order that the mass of rock over the 24 feet line should equal the mass excavated, about ten feet more of excavation should have been made. The contractor and many others supposed that the tide would sweep away the loose débris, but this idea proved fallacious.

It was supposed that a crust of only 6 feet thick, supported by the pillars of rock or wood, would be admissible; but, in fact, a thickness of from 14 to 18 feet was required to prevent the water from entering the chamber in quantities greater than could be easily controlled; but there was no practical difficulty in excavating more in the central part of the rock, though toward the edges suspicious leaks had developed themselves. The removal of the débris probably cost more than the amount expended up to and including the time of the blast. Had he made this 10 feet of additional excavation, it is probable he would have been perfectly successful. Again, the chamber containing the powder being only about two-thirds full of water at the time of the blast, made the resistance very much less than it would have been had the chamber as well as the shaft been filled, so that the water would have stood on the same level in the shaft as on the outside. To that fact may be attributed some of the contractor's want of perfect success.

Mr. Von Schmidt deserves a great deal of credit for the work he has achieved. His daring character is shown by his accepting a contract in which he was to receive no