of comparison for the above table. We see, in table No. XI, that the greatest resistance obtained by the mixture of lime and pipe clay only, was 187 lbs. The experiments of tables Nos. XI and XII differ only in substituting in the latter, potash for soda. We see that the results obtained with potash are superior to those obtained with soda. No. 3 of the above table being cracked, I am uncertain whether the greatest resistance belonged to No. 3 or No. 4. No. 5 shows that the proportion of one-half potash, diminished the resistance. I made a sixth experiment by putting one-half of potash at 10°: the mortar resulting broke under a load of 242 lbs. It is remarkable that the mortars which gave the greatest resistance were those which hardened the most slowly. Several of the preceding tables the same result; it was quite striking with the Boulogne pebbles of table No. VIII. It will there be seen that No. 1 which hardened in twelve hours, supported a weight of only 108 lbs., while No. 12. which supported 286 lbs., required 12 days to harden. An inspection of the preceding tables will show that, in general, the mortars which gave good results, took from 8 to 15 days to indurate, whether made of natural or artificial hydraulic lime. The experiments of table No. XII were made in the month of January. I also made essays with the soda and potash of commerce, which, as is known, are composed of sub-carbonates, sulphates and muriates of these substances: the results obtained were not so good.

gave

I made several experiments with lime and solutions of potash and soda without other admixture. I operated in this way: I took several pieces of fat lime as they came from the kiln, and slaked them to dry powders thus, on the first piece I sprinkled the fourth of its volume of solution of soda standing at 5° of the pese-acid; on the second, the same quantity of solution of the sub-carbonate of soda of commerce, of the same strength; on the third, the same quantity of an equally strong solution of potash: on the fourth a like quantity of solution of the same strength of the sub-carbonate of potash of commerce; and on the fifth an equal quantity of solution of muriate of soda, also standing at 5°. I left these several powders in the air for one month; I then added the quantity of water requisite to produce a paste, and had them reburnt in a lime kiln. As they were drawn from the kiln I slaked them with a little water, and left them in the air, for some days: I afterwards made mortars of them, by mixing them with sand, and having allowed them to harden in the air for twelve hours, I placed them in water: the two first mortars hardened in two days; the third, the fourth and the fifth hardened in a day and a half. At the end of a year, I wished to break these mortars: and I was much surprised to find that, notwithstanding they had hardened so promptly, they had become entirely soft. As the boxes containing these mortars were covered with others, I had not been able to observe them at the period to their beginning to lose hardness.

It follows from experiments made up to this time, that it is clay,—that is to say, a mixture of alumine and silex, united with the lime, that gives good hydraulic lime by calcination. (An exception exists in the lime of Senonches, which is said to be very hydraulic, and which appears to owe this property to silex alone.) We do not as yet know the proportions in which silex and alumine ought to exist, to give the best results. Hydraulic limes are improved by adding to the mixture of clay and lime, a little soda: the result is still better if potash be used instead of soda. We have seen that the Boulogne pebbles gave a lime which hardened very quickly, but which would not bear mixture with any other substance, in making

* This mortar was cracked on being cut, which diminished its strength. Au.

mortar this being the only lime which gave this result. I before observed that it appeared to me these pebbles contained a little soda: but we have just seen that limes which I caused to be reburned with a little soda or potash gave good mortars when mixed with a considerable quantity of sand: it seems, therefore, as I have remarked, that these pebbles contain some substance which escaped analysis, or else that some of the substances known to be present are modified by the sea-water.

In comparing artificial hydraulic limes, I have been led to remark that better results were obtained when, in lieu of making the mixture of clay, with lime just from the kiln, I first slaked the lime with one fifth of its bulk of water, and left it, in this state, in the air for a month or two. I do not attribute the better effect solely to a more minute division of the lime: but I think that the slaked lime exposed to the air, absorbing much oxygen, as I have shown in page 31, is capable, in this state of oxidation, of combining better with the clay which is mixed with it. I proposed making experiments to ascertain what might be the advantages of this process, and also during what time it would be proper to have exposed to the air, lime that had been slaked to powder, or lime designed to be left to spontaneous slaking; but I was obliged to quit Strasburg, before having it in my power to make these experiments.

CHAPTER V.

On Hydraulic Mortars made of fat lime and trass, or fat lime and puzzalona. Trass is a substance obtained from the village of Brohl, near Andernach, on the Rhine; this village is situated at the foot of an extinct volcano. Trass is of a grayish colour, much resembling gray clay which has been calcined. I have seen several pieces of trass which were covered with lava. This last substance differs much from trass: the separation is distinct: the lava which covers the trass is of a blackish colour, and its surface is full of asperities and cavities, showing that it has sustained a very high heat, and very rapid cooling: trass seems to have been exposed to a much lower heat.

Puzzalona is likewise a burned clay-deriving its name from the village of Puzzoles, at the foot of Vesuvius; it is found at or near the surface. According to Mr. Sganzin, there are a great many varieties of puzzalona: it is found white, black, yellow, gray, brown, red, and violet.

It is not known whether trass and puzzalona result from banks of clay which have been highly heated by incumbent lava, or whether they have been thrown out of volcanos during irruptions; it is possible that both suppositions may be true; an examination on the spot can alone remove uncertainty in this respect.

Puzzalona is, according to Mr. Sganzin, also obtained near Rome: this is of a red-brown, and contains brilliant particles having a metallic lustre.

The Dutch have a great trade in Trass. They get out this substance in masses, and reduce it to a very fine powder, by means of wind mills. Much has been sent to France, to the North, and to England; but it seems that the commerce has diminished a little. Some authors call this substance the Terras of Holland.

Arches, in several of the gothic churches, in Belgium were built of pieces of the Andernach Trass: as was the case in several edifices on the Rhine. The great lock of Slikens in Belgium, was rebuilt with pieces of the arch of a demolished church; and by pulverizing the same material, a very good

trass was obtained.

On the shores of the Mediterranean, much puzzalona, furnished from the environs of Rome and Naples, is used. The good effects of this substance were known to the Romans, for Vitruvius, Book II., Chapter VI., has the following passage. "There is a kind of powder to which nature has given an admirable quality: it is found near Baie, and in the grounds around Mount Vesuvius. This powder mixed with lime and stones, not only renders common masonry very firm, but it even unites and hardens in a remarkable manner, masonry at the bottom of the sea."

I had occasion to appreciate the good effects of trass in the great works of Vesel. It was brought in powder, and as it was transported in batteaux descending the Rhine, it was not dear. Before becoming acquainted with the excellent qualities of the Obernai hydraulic lime, I caused some of it to be brought to Strasburg, and the first hydraulic works that I constructed at Strasburg were made of mortar composed of fat lime, sand, and trass. But the trass cost much more at Strasburg than at Vesel, on account of the export duties that the Prussians levied on it, and because of the difficulty of ascending the Rhine. A cubic metre (35.34 cub. feet) cost 120 francs (about $23.00) at Strasburg, which raised the cost of the hydraulic mortar enormously. I had begun some researches with a view to make artificial trass, when I became aware of the hydraulic properties of the Obernai lime. This lime was afterwards used without trass in the foundations of revetments, in the construction of bridges, and in all masonry exposed to the air; but in the foundation of locks and dams and in the construction of caps for under-ground casemates, a little trass was always mixed in the mortar, with the lime and sand. I continued my experiments on the forming of artificial trass, less for our particular advantage at Strasburg, therefore, than for those places which do not afford good natural hydraulic limes. I succeeded, in the end, as will be seen in the following Chapter, in making artificial trass equal in quality to natural trass or puzzalona. The cost of a cubic metre (35.34 cub. feet) was about 30 francs (about $6.00.) I caused the puzzalona to be brought from Naples, it was red, and much resembled brick or tile dust.

An analysis of trass and puzzalona, made many years ago, may be found in the Précis du cours de construction of Mr. Sganzin of the Polytechnic school. Mr. Berthier asked me for samples of the trass and puzzalona which I used at Strasburg, and made an analysis of them. Both analyses are given below.

In Mr. Berthier's analysis, there was in the Puzzalona, oxide of iron and titanium 0.120.-AUTHOR.

We see that the new analysis of trass and puzzalóna by Mr. Berthier differs in several points from the preceding analysis. According to that of Mr. Berthier, trass and puzzalona contain potash and soda which were not before noticed in those substances. According to the old analysis, puzzalona contains much more alumine than trass, but according to Mr. Berthier, about the same. The proportion of alumine to silex is greater, for both substances, in the old, than in the new analysis. I will add that Mr. Berthier discovered that the trass and puzzalona which I sent him attracted the magnetic needle.

We ought not to be surprised that these analyses differ sensibly, because it has been known for a long time that different puzzalonas give very different results, and it is the same with trass. We can conceive that these substances may vary considerably, because the several strata of clay in the same bank often vary sensibly. I proceed to give different experiments made with fat lime, puzzalona, and trass.

Observations on the experiments of Table No. XIII.

The lime of the first twelve numbers was measured in powder and mixed with the substances indicated in the table. It will be seen that I used various fat limes of the environs of Strasburg. I satisfied myself that these limes had no hydraulic property. The first four numbers were designed to ascertain whether puzzalona is preferable to trass. The puzzalona gave me rather the best results-the lime being the same in all. The experiments 9, 10, 11, and 12 confirm the first four. We see that I had a greater resistance by mixing sand with the lime and puzzalona, or lime and trass, than when there was no sand. The limes of Nos. 13 and 14 were measured in paste, and we see that I obtained a better result by augmenting the proportion of puzzalona.

No. 15 contains the results of many trials made with 1 part of fat lime measured in powder and 2 parts of trass. The time of hardening varied from 3 to 16 days, and the resistance from 231 to 580 lbs. But I had few mortars giving a result so feeble as 231 lbs., or so strong as 580 lbs. We might say that the mean resistance with mortars composed of fat lime, sand and trass, was between 352 and 374 pounds. The mortar which gave 580 pounds contained lime and trass only. If, with the same substances, I had composed a mortar of lime, sand, and trass, I should have had, probably, a still better result. I found fewer variations with puzzalona than with trass, probably because, having but a small quantity of the former, my experiments were few. We shall, however, see in table No. XX., that a mortar like No. 13 of the above table, broke under a weight of 286 lbs. I am uncertain whether this weak result was due to particular circumstances or to the quality of the puzzalona. It appears to me, in general, that there was no great difference between the trass and the puzzalona, that I used.

In comparing the results of table XIII., with those of the preceding tables in which mortars were made of natural or artificial hydraulic lime and sand only, we see that these last hardened much more slowly, and gave much less resistance, than mortars made of fat lime, sand, and trass or puzzalona. Mr. Gauthey was therefore in error, as I noticed page 6, when he stated that mortars made of fat lime and puzzalona did not take consistence in water and remained pulverulent.

In the observations made by Mr. Vicat on the pamphlet which I published in 1824, and of which I have already spoken, this Engineer remarks as follows: "The greatest resistance that Mr. Treussart found for trass mortars immersed for one year, is 17 lbs. for four-tenths of an inch square, and I have made mortars with lime and sand, which, after a year's exposure on a roof to all the inclemencies of the seasons gave a resistance of 31 lbs. and even of 40 lbs." The results that Mr. Vicat and I obtained are very different. These differences can only be attributed to our having used means of measuring the resistance, or rather the tenacity, of mortars, which do not admit of comparison. I will observe, however, that the greatest resistance I found was 579 lbs., for mortar made of artificial trass, and that the section being four square inches, there results 23.1 lbs. for four-tenths of an inch square instead of 17 lbs., as Mr. Vicat supposes. He does not say what were the dimensions of the mortars he submitted to proof in order to compare his results with mine; and as a year had not elapsed from the publication of my pamphlet to the appearance of his observations, I conclude that he did not use the same method in breaking the mortars of which he