188 



THE CIVIL ENGINEER AND ARCHITECTS JOURNAL. 



[JvNE, 



be any reason why the gj'psums should differ from the carbonate of 

 lime in this respect. Indeed, we find that the law holds good with 

 the English pypsums, for the Derby stone makes a stronger plaster 

 than that of Newark, just as it is harder than the latter. Dumas 

 a^ees witli Guy Lussac, in supposing that no other chemical action 

 takes place with the gypsum, than the evolution of its water of 

 crystallisation. 



In Paris, the mode of using plaster is to employ it pure and free 

 from mixture. The very low price at which it is sold, and the 

 comparatively high price of sand, dispense with the motives of 

 economy which render mixtures almost indispensable in our case. 

 The town of Paris pays for its municipal works, at the rate of 

 12s. 9gd. per ton of plaster, whereas it cannot yet be had in London 

 for less than about iOs. per ton. Whilst the practice in France is 

 to use plaster pure, I am disposed to think that the mixture of 

 sand, so far from being prejudicial, is even desirable, if confined 

 within reasonable limits. We find that in reassuming the state of 

 liydrated sulphate of lime, the plaster goes through an imperfect 

 crystallisation; and this action is accompanied by a singular re- 

 arrangement of the molecules. This causes the plaster to swell 

 when used alone, and to such an extent, that it is impossible even 

 to finish a ceiling close up to a wall at once. Now the introduc- 

 tion of a body so full of inequalities as the coarse, sharp sands, 

 must afl'ord room for the free action of this expansion; and, at the 

 same time, the facettes of the sand must offer, as is were, nuclei, 

 which cannot but be favourable to the crystallisation. It is, doubt- 

 lessly, on these principles that we can explain the superiority of 

 the plaster containing the wood brees, which does become harder 

 than the purer plasters, if used alone. Too large a proportion of 

 sand should be avoided; but very fair work can be executed even 

 with a mixture in the proportions of two of sand and one of plaster. 

 Under any circumstances, the finishing coat should be pure. Sub- 

 sequent experience will decide, whether the use of two materials 

 of this kind does not expose the work to unequal contractions, 

 likely to cause fissures, or cracks. 



The plaster made near Paris sets with a rapidity very much 

 greater than any material we are accustomed to for plastering 

 purposes; and, for very large uniform surfaces, perhaps this is a 

 difficulty. The workmen have not the time to work the floating 

 coats with the mathematical correctness we usually exact in our 

 country. But, to a certain extent, this objection may be obviated, 

 by slight differences in the mode of preparing the plaster, or by 

 altering the quantity of water in proportion to the positions in 

 which the material is required to be used. Thus, if all the 

 strength of the plaster is needed, the smallest quantity of water 

 is introduced; about as much in bulk as the plaster itself occupied. 

 This is called by the workmen, "gacher serre" (stiff guaged). 

 AV^hen it is necessary to work and re-work the face, as in setting 

 coats, more water is added, or the plaster is said to be "gache 

 clair" (guaged thin). Habit alone can fix the precise proportions, 

 for it is impossible to arrive constantly at the same results in the 

 burning. For the very finest works, the workmen make what they 

 call a "coulis;" this is run in, in a semi-fluid state. Plaster which 

 has been thus treated, with an excess of water, does not acquire 

 the t3nacity, nor the hardness of that treated in such a way as only 

 to present to it the water of crystallisation. 



The extraordinary forces of adherence, &e., of the Paris plaster, 

 enables the work on ceilings or partitions to be executed with far 

 less expense of lathing than similar works executed with our 

 lime and hair. Rondelet made experiments to ascertain the limits 

 of these forces, and he obtained the following resiilts: — A parallelo- 

 pipedon of plaster, ivith a base measuring 1 in each way, supported 

 a weight of VUlb., acting so as to tear it asunder; this he called the 

 force of adhesion. Similar figures resisted a crushing weight of 

 722lb.; so that the ratio of the resistance of plaster to an effort of 

 traction, compared to one of extension, is as 1-9^. Rondelet found 

 that there was a sensible difference in the manner in which plaster 

 adhered to lirick or stone, from the action of mortar under similar 

 circumstances. For, when cubes, joined by the respective materials, 

 were subjected to forces tending to tear them asunder, the mortar 

 liroke through the centre of the joint, leaving particles attached to 

 the upper and under surfaces; the plaster, on the contrary, left the 

 surfaces perfectly clean. In new works, the plaster adheres to 

 other materials, with about half the force necessary to tear it 

 asunder; mortar, for several years at least, only attains one-third 

 of the same force. This ratio does not continue; for, after ten or 

 twelve years, the plaster loses its strength, whilst, at the same 

 epoch, we find the adhesion of the mortar toother substances to be 

 equal to the force of adhesion of the cubes themselves. The 

 »ubsequent ratios are in inverse progression; mortar always hardens 



by time— plaster loses strength. As these remarks only apply to 

 its use as a mortar externally, it should never be employed per- 

 manently for such positions; internally the loss of strength is 

 not so rapid, for it depends upon the absorption of moisture from 

 the atmosphere. For temporary works; for internal works, requiring 

 great rapidity of execution, however, the use of Paris plaste is 

 invaluable. 



The usual practice in Paris (as I had the honour of observing in 

 a paper I read last year), is to execute the work intended to be 

 plastered with rubble stone, set in plaster mortar. If possible, the 

 principal elevations are executed in ashlar; externally, plaster is 

 never used if it can be avoided, for its use requires care and 

 numerous precautions. Firstly, the plaster coat must be entirely 

 out of the ground; it must be removed from all weatherings, 

 where the capillary action would allow the absorption of water; 

 the upper surfaces must be covered with zinc, or other metal; and, 

 if it be expected to stand for many years, the whole must be 

 painted. When, however, plaster is to be applied on walls, exter- 

 nally or internally, the course followed is to clear out the joints of 

 the masonry, and to wet the surface. Plaster, gauged stiff, is laid 

 on with a broom, or in any similar expeditious manner, and it is 

 brought to a tolerably uniform face by use of the trowel. This is 

 called 'faire le crepi,' a term equivalent to our "rendering." The 

 floating coat, or 4 enduit,' is applied by the trowel, and dressed 

 off with a rule, in somewhat a similar manner to the system fol- 

 lowed by our own workmen; but it is in the execution of this work 

 that the greatest ditficulty arises, from the rapidity with which the 

 plaster sets. The stuff' is gauged thin, but not sufficiently so to 

 allow much manipulation. \Vhen the face is floated, as described, 

 the plasterer passes over the surface with a sort of toothed trowel, 

 called 'la truelle bretelee;' using, firstly, the toothed side, to 

 remove any asperities, and finishing with the knife edge on the 

 other. A thin setting coat is lastly added, to stop up all the pores 

 or inequalities. The time required to complete such plastering 

 on wall is very short compared with what we are accustomed to. 

 The floating coat may be applied within four days of the rendering, 

 under favourable conditions; and the whole work easily completed 

 in a week. 



Partitions are usually executed in a manner essentially different 

 from our own. A sort of wood frame-work is made, without much 

 complication of carpentry, by the way, for the French, very wisely, 

 prefer a wall where we too often place large trussed partitions. 

 The French partitions rarely consist of more than upright posts, 

 with stouter ones for doorways, and a few discharging braces, or 

 horizontal ties. The upright posts, 'les poteaux,'are spaced about 

 1 ft. 4 in. apart ; the door-posts are usually planed so as to form 

 the architraves of the doors; they are called 'les poteaux d'huisserie.' 

 Upon the common quarters laths are nailed (mostly of poplar, or 

 fir), which are from 3 to t inches wide, and spaced about i^ inches 

 apart. The interior is filled in with old plaster rubble, or light 

 stone, and the outer surfaces rendered, as for walling. Such par- 

 titions answer admirably for the purposes of keeping out sound, 

 and are tolerably light. From the immense quantities of plaster 

 rubble to be met with in Paris, they are also, comparatively 

 speaking, economical. Close lathing is very rarely executed; nor, 

 in fact, do the oak laths used in France allow such work to be well 

 done. Some masons in Paris use a sort of tile, cast beforehand 

 purposely for this use, and made of plaster. This system is not 

 so solid as the usual one of only employing rubble, for the plaster 

 does not adhere so well to the smooth faces of the tiles; but it 

 avoids a very considerable amount of humidity. 



( 1 ) Ilourde Plein. 



Ceilings are executed in several manners. — (1.) The space 

 between the joists is filled in solid, with plaster, or stone, rubble 

 carried on rather wide laths underneath; the lower surface is then 

 rendered like a wall wonld be, and a bed is formed on the top to 

 receive the tiles, or sleeper joists and flooring are added. This is 

 said to be 'hourde pleiu.' 



