Retombé Tendu and Retombé Plié

The following is a tentative draft of a short section from “The Movement of the Legs,” chapter 4 of the work in progress The Technique of Eighteenth-Century Ballet (© Edmund Fairfax 2018). Some aspects are discussed in detail in other sections or chapters, and the reader would simply consult those sections in a completed work, but these are not available here at the present time. To sidestep the high fees demanded by institutions and publishers for the use of copyrighted images, I use careful tracings of original illustrations instead, with irrelevant detail from background, costume, etc. omitted when desirable, and these tracings are presented below.


4.2.7 Landing From Jumps

The retombé (‘fall back down’) in eighteenth-century ballet was the land upon the floor from a jump into the air, or as Gallini (1772?: 163) puts its, “the falling is the return of the feet to the ground, by the natural gravitation of the body.” In the sources, this action is normally expressed by some form of the verb ‘to fall,’ although other expressions are sometimes used as well: in German fallen (‘to fall’), niederfallen or herabfallen (‘to fall down’), herabsinken (‘to sink down’), or niederspringen (literally ‘to spring down’); in Italian cadere (‘to fall’) or discendere (‘to descend’); in French retomber (‘to fall back down’); and in English to fall, to (a)light, or to come down. The Retombé Tendu and the Retombé Plié

The dance of the eighteenth century regularly employed two kinds of lands. The first was the retombé tendu (‘stretched land’), which was a land onto the toe(s) with the knee(s) held straight, and with the heel(s) either kept off the floor or straightway set down on the floor, as reconstructed in figs. 1 and 2. (The term retombé tendu is not extant but is used here to fill a nomenclatural lacuna and ease discussion.) The reconstruction shows time-lapsed phases, moving from the left to the right. It is to be understood that all the movements occur on the same spot rather than spread out as shown; the separation of the stages is merely for clarity’s sake. To be more explicit, the leftmost leg shows the dancer in the air; the second from the left, the touch-down of the toes; the middle leg, the rolling down onto the ball; the second from the right, the lowering of the heel towards the floor without bending the knee; the rightmost, the raising of the heel again.

Fig. 1. A reconstruction of the retombé tendu, with the heel raised.

Fig. 2. A reconstruction of the retombé tendu, with the heel set down.

The second was the retombé plié (‘bent land’), which was a land onto the toe(s) with the knee(s) well bending, and with the heel(s) either kept off the floor or straightway set down on the floor, as reconstructed in figs. 3 and 4.

Fig. 3. A reconstruction of the retombé plié, with the heel set down.

Fig. 4. A reconstruction of the retombé plié, with the heel raised.

These two kinds of lands are only fleetingly mentioned in the eighteenth-century sources, here and there in connection with individual steps. More often than not, the manner of landing from a jump is simply not given in the descriptions of individual steps or even in broader discussions of technique.

Let us look at some examples. Rameau (1725a: 160) alludes to the retombé tendu in his description of the composite version of the pas de rigaudon, writing that “great care must be taken in doing this step, that your legs are well stretched when you raise them and that when you spring [in the concluding pas sauté], you land on the toes of both feet with the legs stretched, which makes you appear lighter.” In the contretemps de côté, Ferriol (1745: 1/111) instructs the reader to land in a retombé tendu, to wit, to “bend on the former [i.e., the left foot], springing with the left foot, and keep the right off the floor and keep the knees straight.” Magri (1779: 1/58) alludes to these two ways of landing in each of the two constituent jumps of the échappé: a retombé plié in the first and, as implied, a retombé tendu in the second, for the whole step “is made up of two little springs and between these a land in a bend,” and not two lands in a bend, one after each jump.

These two different manners of landing are perhaps most clearly delineated and contrasted in descriptions of the chassé, assemblé, and sissonne. In his discussion of the chassé double, or two chassés performed one after the other, Rameau (1725a: 177) states that “you land from the first spring bent and in the same moment spring a second time.” Ferriol (1745: 1/113) expands upon this description, noting explicitly that a retombé tendu is performed in the second chassé: “since two are regularly done in succession, you land in a bend in the first spring, and in the same moment you spring a second time, landing with the knees straight.”

In his description of the assemblé found in the pas de gaillarde, Rameau (1725a: 145 alludes again to a retombé tendu: here the initial assemblé is done by “landing from this spring on both feet with the knees well stretched.” Tomlinson provides two examples of the assemblé – or “close” as he calls – in two different contexts, each with a different land. He notes that a dancer is to land with straight legs in the assemblé when followed by a pause, while in the assemblé figuring in the bourrée en l’air liée the executant is to land in a bend. He writes (1735: 88) that “instead of the Close’s ending either in the first or third Position with the Knees straight, as in the former [i.e., the assemblé on its own], it here comes down behind with the knees bent.”

The two manners of landing are also alluded to in connection with the composite step the pas de sissonne, specifically in two different versions. In his discussion of the initial assemblé figuring in one version, Rameau (1725b: 75) explicitly indicates that “you land in a bend on both feet crossed,” while in the variant he calls the sissonne coupée (1725a: 158) “you land on both your feet without bending the knees; you bend them afterwards, however, in order to do the second spring.” Tomlinson (1735: 47) gives the timing for this “coupé” execution of the sissonne and makes clear that the dancer is to land with the knees straight, and then only on the following count is he to bend the knees in order to rise: “The first Spring is made upon the Time or first Note; the Sink for the second is in the second Note, which second Spring is performed to the third Note; and the fourth is in the Sink preparing for the succeeding Step.”

These different ways of landing in the assemblé of the sissonne are clearly distinguished – or, rather, can be clearly distinguished – in Beauchamps notation. In the sissonne coupée (example on the right in fig. 5), the bend-signs representing the second preparatory bend are placed on the following characters, while in the regular sissonne (example on the left in fig. 5), with a retombé plié in the initial assemblé, the bend-signs marking the preparatory bend for the second jump are placed on the characters representing the land from the opening assemblé.

Fig. 5. Left: pas de sissonne (Feuillet 1700a: 81); right: pas de sissonne coupée (Feuillet 1704: 3), after originals.

Normally, the characters marking the second constituent jump of the sissonne show no bend of the knees in the final land from the jump. As such, this absence of any plié-signs is in agreement with extant verbal descriptions, which explicitly state that the executant lands from the last jump with the knees stretched. Sol (1725: 72), for example, notes in his description of the sissonne that the dancer lands in the final jump with “the knees straight,” Dufort (1728: 73) similarly indicating that the step ends “with the knees well stretched.” “Lands on the Toes” Mentioned in Reviews

Reviews of theatrical dance performances from the period at times remark upon the skill with which dancers could land neatly on the toes, and some of these are almost certainly references to the retombé tendu. Zacharias von Uffenbach (1935: 31), for example, notes of a dancing Scaramouche seen at London’s Drury Lane Theatre on the thirteenth of June 1710 that

the most amazing of all was that he danced a ‘Chique’ [i.e., jig] with great agility on the tips of his toes with his feet turned entirely inwards, so that one cannot conceive how he was able to bend his feet thus backwards, stand on tiptoes, and spring about without straining his feet or breaking them at the ankle-joints. He jumped so high in the air and with such frequency, alighting each time on his toes.

A further example is afforded by Sonnenfels (1784: 6/337-338 = 17 Dec. 1768), who almost certainly alludes to the retombé tendu in his review of an eleven-year-old dancer by the name of Delphini. He remarks in particular upon the girl’s “strength in the knees, which shows itself especially in lands on the toe.” Such a comment can hardly describe a land in a demi-plié.

This girl is a wonder of nature and art. In her eleventh year, she not merely bids fair to be a great dancer but is one already and brings together all the talents which Terpsichore otherwise bestows singly upon her favorites. In the heroic as well as the comic, in expression and gestures, as well as in that which is properly called dance and which consists in the lightness, quickness, and justness of steps, in the height and brilliance of entrechats, and in the strength when landing on the toe, even in the play of the countenance and the eyes with which she paints forth joy, sadness, displeasure, waggishness, and all the subtle blends thereof, strong in all alike, she amazes all newcomers and exasperates all fellow artistes. One could not have received from nature a greater talent for dance: a good build, strength in the knees, which shows itself especially in lands on the toe, in pirouettes, and rises; her tendons, like little elastic springs, raise her quickly and keep her on the tip of her foot for a unbelievable length of time. Rolling Through the Foot

Fig. 6. Detail from the frontispiece of Bonin (1712), after an original.

The exact mechanics of how these two different kinds of lands were to be done are only hinted at in the sources. It is clear that no matter what kind of land the dancer was to do, the toes of the feet were to be pointed straight down to the floor in descending from the jump. Indeed, the foot, except in only a few cases, was normally to be well stretched and pointed when raised off the floor (§3.11). Noverre (1760: 328) indicates that failure to do so naturally resulted in mishaps, for “by faulty positions of the feet, which not pointing directly toward the floor when the body comes down, [the feet] twist, bend, and give way under the weight that they receive.” An illustration of a springing dancing master from the frontispiece to Bonin’s 1712 handbook shows the feet directed downward in agreement with Noverre’s remarks (fig. 6).

It follows from the foregoing that the tips of the toes were the first part of the foot to hit the floor in landing. The dancer then was evidently to roll smoothly down onto the ball of the landing foot. Such articulation of the joints of the foot in landing from a jump is certainly in agreement with the normal manner of simply transferring weight in non-springing steps extant for the period, with the dancer holding his foot stretched, and then “first setting it down stiff on the toe, lowering it thence onto the ball” (Taubert 1717: 512). Magri (1779: 1/129) notes explicitly in connection with capers that the foot was to be well articulated in cushioning the body from the shock of hitting the floor, that “in coming down, [you] land in second position [in the pirouette en l’air], coming down first onto the toes of the feet and then the rest, as you must take care in all the lands from capers.” The late-eighteenth-century dancer Roller (1843: 109-110) also notes broadly that “in landing from a hop or spring, one must come down onto the very tips of the toes. Moreover, not only does a land onto the flat foot look heavy and is unpleasant to the ear but the body receives a jolt and loses its lissom fluidity for the following movement.”

Once the dancer had rolled down onto one ball, or both balls, the rest of the impact from the land was dissipated in two ways, depending on whether the retombé was executed tendu or plié. In the case of the retombé plié, the dancer clearly softened the land both by bending the knees as well as by working through the insteps, that is, by going straightway into a plié. Whether the heels were set down on the floor or not hinged upon whether the plié was regular or forced. As already discussed above (§4.2.6), normally the heels were lowered to the floor in a plié, with some exceptions, however, the most notable one being a retombé plié preparatory to a grand jump, in which case, the heels do not appear to have touched the floor in order to ensure that the springing movement did not happen at the end of the muscles’ range of motion, at odds with current ballet practice.

In the case of the retombé tendu, the heels were evidently lowered towards the floor without touching it, only to be quickly raised back up again while keeping the knees straight, if the dancer was immediately to continue dancing on the toes. Thus in the retombé tendu, the impact of the land was cushioned by the elastic and resilient articulation of the toes and instep(s) alone. Such a practice is certainly in agreement with early-nineteenth-century usage. Helmke (1829: 144) makes clear in his description of the assemblé figuring in the combination of a jeté and assemblé (which he calls a saut-croix) that the heels were not to touch the floor in landing on the toes in the final assemblé: “lower the body, bending the knees, spring up again, beat the left foot behind the right and the right in front of the left in fifth position (fig. 12 [not reproduced here]); in doing so, the heels always remain raised and must not touch the floor.”

Fig. 7. Springing onto the toe (g), heel (h), and flat of the foot (i) (Feuillet 1700a: 20), after an original.

It was possible, however, to land on the heel(s) or flat(s) of the feet, although these options appear to have been seldom met with in more lofty dancing and were likely confined mainly to the comic and grotesque styles. Feuillet (1700a: 20) both mentions and gives notational characters (fig. 7) for the three options of landing on the toe (g), heel (h), or flat of the foot (i), in the latter case, with the heel and ball coming down at the same time. Landing on the heels is in fact prescribed in one handbook for one version of the échappé (ch. 25). On the whole, however, the notational characters for these three distinctive ways of landing are almost never found in extant dances recorded in Beauchamps notation. In practice chorégraphie typically dispenses with showing the part of the foot onto which the dancer is to land, presumably because it was needless and uneconomical to mark in such predictable information. Lowering the Heel to the Floor

There is some suggestion that the execution of the retombé tendu in ballroom dancing could differ from that normal on stage: to wit, in rolling down off the toes onto the ball(s), the ballroom dancer could lower the heel(s) to the floor straightway as well, as he not uncommonly did after a relevé in non-springing steps, in order to eschew having to balance too long (§4.2.4). (It should also be borne in mind that recreational dancers did not always dance in shoes fit for dancing (click here for further discussion).) Indeed, all of Rameau’s illustrations of lands from jumps (1725a) show the dancer on the flats of the feet rather than on the toes, like those found in Tomlinson (1735). This practice would appear then to be an instance of the “Artful Carelessness” that Weaver (1712: 65) indicates was to be cultivated by the amateur dancer and would appear to be an extension of the “present négligent and thoroughly absurd fashion” of dancing in the ballroom “with the feet flat,” alluded to by Taubert (1717: 506). In fact, it appears to have been normal to lower both heels to the floor at once in the assemblé of the ballroom (ch. 19), a practice prescribed by Sol (1725: 78), Tomlinson (1735: 37-38), and – in connection with the assemblé found in the pas de gaillarde – Magri (1779: 1/72).

Lowering the heel(s) to the floor in a land from a jump when performed on the ballroom floor was evidently of some age already by the beginning of the eighteenth century, one alluded to by Esquivel (1642: 11r) in connection with the jump the salto al lado, wherein “you must spring onto the toe, lowering the whole foot forthwith.” Early-nineteenth-century sources at times mention this option as well, such as Lambert (1820?: 16):

The Assemblé is a spring from one foot upon both, place the right foot to the second position, balancing the body upon the left, and in springing from the ground, bring the right foot to the fifth or third position, falling upon both toes and flattening the feet immediately afterwards; to make the Assemblé backwards, the movement is made from the second position, throwing the right foot behind the left, to the fifth position instead of forward.

Lambert (1820?: 15) also has the heel(s) set down in certain places in non-springing steps, even dismissing a position on the toes as too unsuitably “theatrical” for some movements in the ballroom.  “In making the Coupé, observe to keep the foot flat – the style is theatrical and affected when it is made with the heel raised.”

In some cases, a prescription to lower the heel(s) at the end of a step may be part of a long tradition of beginning and ending a step flat-footed when the step was described in isolation (still the norm in descriptions of contemporary ballet steps). But the illustrations in Rameau (1725a) and Tomlinson (1735), which show flat-footed dancers, and Taubert’s remark (1717: 506) given above about the “thoroughly absurd fashion” suggest that setting the heel(s) down on the floor in order to avoid balancing was not uncommon in the eighteenth-century ballroom.

The prescribed norm, however, was to perform most terre-à-terre leg movements of formal dance on the toes, or better said, on the balls of the feet with the heels lifted off the floor. Hänsel (1755: 71), for example, writes that broadly “a man or woman dancer must have the heels little or not at all on the floor.” Pasch (1707: 435) likewise indicates that

as dancing is also moving in the lightest of possible ways, he who has knowledge here knows that everything which is least fixed to the earth can be moved most easily, and for this reason, and not out of pure unbridled whim, a dancer is mostly on the toes of the feet.

The same remark can be found almost verbatim in Behr (1713: 22). Thanks to the prescribed practice of pulling up the body, that is, stretching it upwards from the feet to the head (§3.2), and dancing on the toes (when the feet were not in the air), eighteenth-century dance was “largely nothing more than almost a light hovering” (Behr 1713: 24), a view also found in Pasch (1707: 436) and Taubert (1717: 518). Or as Vieth puts it (1794: 2/399), “stretching the foot, which allows one to rise onto the toes, so needed in fine dancing, must not be practiced any less; without being good on the toes, there can be no lightness in the steps.” Use of the Retombé Tendu and the Retombé Plié

The question then arises of how these two different manners of landing were to be used. The little information extant suggests that normally dancers were to come down onto the toes with the legs held straight. In other words, the retombé tendu was the norm in springing, i.e., with jumps performed in isolation. Indeed, Beauchamps notation generally symbolizes jumping steps simply with a character for an initial bend followed by one for a rise into the air, without any further symbol attached to the composite graph showing a cushioning bend for the land. The retombé plié, on the other hand, was a special case, figuring either as a distinctive feature of a given step, or as the result of performing a combination wherein a retombé tendu was immediately followed by a plié. That is to say, in the latter case, the land from a jump would flow freely and swiftly into the preparatory plié of the immediately following step, such that the retombé tendu became in effect a retombé plié by virtue of the quick unbroken combination of movements. But if a pause – or, indeed, some other movement – intervened between the land from one jump and the preparatory plié of some subsequent movement, then the dancer would land on straight legs (unless, of course, a retombé plié was called for as a distinctive part of the step).

In practice, however, the number of retombés pliés in a dance likely outnumbered that of retombés tendus in the airborne styles of the half-serious, comic, and grotesque, wherein commonly many jumping steps were strung together (Fairfax 2003: 103-166), such as “twenty entrechats in succession without stopping, beaten à huit,” which Charles de Brosses records having seen in 1739-40 (cited in Sand 1862: 1/172). And so from a statistical point of view, one might say that the retombé plié was in fact the norm, at least in the aforementioned styles. Indeed, concluding a string of high jumps with a pose – the latter, according to Magri (1779: 1/110-111), usually assumed by first bending and rising – would have allowed a dancer to avoid the more difficult retombé tendu even in the last jump.

An example of a step wherein a land in a bend was a defining feature is afforded by the “regular” pas de sissonne. Here, the dancer was to “land in a bend on both feet” (Rameau 1725b: 75) in the initial constituent element, the assemblé, even if the dancer paused before doing the next constituent jump. When performed in the ballroom minuet, for example, the dancer, at least according to Taubert, lands in a plié from the opening assemblé on the downbeat of the first measure and stays down for the second count only to rise and land from the ensuing jump on the third count. As Taubert writes (1717: 711),

one springs with the right crossed behind the left, with the knees bent and well turned out, and without moving, stays down in this position for the second quarter note and springs up gently again on [the right foot] on the third quarter note and at the same time thrusts the left leg away to the side.

As noted above, however, it was also possible, in the variant called the sissonne coupée, to land from the assemblé on straight legs and then bend again after a short pause, as described in both Rameau (1725a: 158) and Tomlinson (1735: 47). The choice of land then becomes a defining feature for these versions of the sissonne.

A further example is afforded by a variant of the bourrée, one beginning with a demi-contretemps and called a soubresaut by Magri (1779: 1/107) – not to be confused with the modern step of the same name. The dancer lands in a bend from the initial hop, thereby going at once into the following demi-coupé of the ensuing bourrée proper, for he notes that “it is not like the other combined steps that are made up of the movements of the steps that go together with it because the spring in the demi-contretemps and the beginning of the bourrée are done in the same instant [here].”

Elsewhere, the retombé plié was simply the by-product of the choreography, wherein the land from one jump was immediately followed by the preparatory bend belonging to another movement. This is evident from Ferriol’s remark (1745: 1/113) that when a number of chassés were performed in succession, for example, they were all to be done with a retombé plié except the last one, for the bends of the retombés pliés in this string were, strictly speaking, the beginnings of the following steps: “note that if many are done one after the other, as in l’allemande, you land in all of them with the knees bent except the last one.”

This postulated practice would explain why Tomlinson (1735: 88, 39) has the dancer in the assemblé figuring in the bourrée en l’air liée land in a bend (because the immediately following step, a changement de pieds, begins at once with a plié), but has the dancer in the assemblé on its own when followed by a pause land with the legs straight (for in the latter case, Tomlinson has “This Step generally [take] up a Measure, that is to say, with the Time you rest and stand still”). Likewise, in his description of a sequence of two pas sautés, Tomlinson’s remarks on timing (1735: 41) also point to a retombé plié as preparation for the immediately following jump: “The Coming down of the first Spring, as I said before, marks the Time or First Note; the Sinking or Bending of the Knees, in order for the second Rise or Spring, answers the same Note; and the third is the Coming down of the Weight in the Sink &c.”

In a similar vein, Taubert (1717: 700) notes in his description of the contretemps double that “the first time, you fall gently with the knees a little bent and thus take tempo [i.e., do the preparatory bend] at the same time for the second spring.” And the composite pas de rigaudon was not uncommonly followed by a pause or a single marché (ch. 19), at least in the ballroom, that is, was not followed by any plié, and thus the dancer in the concluding sauté needed to “land on the toes of both feet with the legs stretched” (Rameau 1725a: 160). Problems with Landing

The retombé tendu was a difficult way of landing from a jump. Keeping the knees straight and cushioning the impact through the articulation of the feet alone places greater strain on the knees, calves, and Achilles tendons. Not surprisingly, among the injuries common among dancers in Noverre’s day (1760: 327) figured the “tearing of the Achilles tendon.”

Furthermore, the attempt to keep the legs straight in hitting the floor could easily cause the dancer to bounce up a little, be thrown off balance, and fall to the floor, particularly in lands from grand jumps. As Magri warns (1779: 1/129), “if the limbs remain full of vigor and rigidity, the body will spring up again upon landing therefrom, resulting in a lose of upright balance and occasioning a fall to the floor.”

At the same time, however, the dancer needed to guard against landing flaccidly and needed to keep the weight of the body pulled up to lessen some of the strain on the lower parts. As Magri writes (1779: 1/138), “a ballerino out of breath and utterly flaccid can never land lightly or rise neatly but needs strength to rise powerfully and to hold himself in coming down from on high onto the toe of the foot onto which he lands.” Historical Context

The practice outlined above is in marked contrast to that of contemporary classical ballet, wherein the dancer is always to land in a demi-plié, in order to reduce the potential for injury occasioned by the impact of the land and, in lowering the heels to the floor, to provide the dancer with a more secure base upon which to balance. So entrenched is this practice of landing in a bend today that the prospect of landing on straight legs is scarcely imaginable to the modern dancer. Indeed, readers schooled in contemporary ballet technique may well find themselves tempted to dismiss these fairly clear references to the retombé tendu as somehow incomplete descriptions and to assume that a land in a plié is somehow always implied. It should be noted, however, that this manner of landing was not unique to eighteenth-century dance but in fact was a practice already established in the seventeenth century and one that continued right up to the end of the nineteenth century. In fact, the convention of landing on straight legs is even more clearly stated in the sources both predating and postdating the eighteenth century.

The practice is quite clearly alluded to in a few dance handbooks extant from the seventeenth century. De Lauze (1623: 47), for example, notes generally in his discussion of the cabriole that “one must never bend the knees except in taking one’s temps in order to rise,” i.e., except when performing a preparatory bend in order to rise or jump. In a similar vein, Esquivel (1642: 19r) notes broadly in his discussion of the retirada that the dancer was normally not to land in a bend in any dance steps other than a select few:

the retiradas are done in two ways, with the carrerilla or without it. Either the one or the other must be done by taking the right foot to the fore as if giving a kick and drawing it back behind the same breadth that it was taken to the fore, or a little less, and setting it down in a bend at the same time, and then doing the carrerilla if required (only here and in the rompidos that are done with carrerillas do bends of the knees look good, and they must be done with such skill and care that one can clearly make out that this movement stems not from feebleness but from nimbleness, for he who is by nature slack in the knees will not dance well.

The restriction of lands with bent knees to a select number of jumps is evident in even earlier sources. Both Caroso (1581) and Negri (1602) prescribe a slight bend of the knees only when landing from some high jumps described in their handbooks, and such bending is presented as a mere embellishment, as a way of giving “grace” to the jump in question, and not as a biomechanical necessity. In ending the salto riverso, for example, Caroso (1581: 12r) writes that one is to “do a salto tondo, coming down onto the toes tightly together, with the knees slightly open [i.e., bent] in order to give them a little grace.” (A prescription to set the heels down in landing is also conspicuously absent). Santucci (1614), however, prescribes a small bend of the knees at the end of springing steps more freely, but again, this appears to be done as an embellishment, not out of any biomechanical need.

The nineteenth-century sources are even more explicit about the practice of generally landing from jumps on the toes with the knees held straight. Théleur (1831: 58), for example, writes that

with few exceptions, steps should terminate with the knees straight, if both feet are on the ground, or if the step terminates on one foot, then the knee of the leg supporting the body should be straight, thus making each step perfect in itself before the commencement of another.

Descriptions of individual jumps in the handbooks from this period also make fleeting reference to straight-kneed lands. In his description of the assemblé, Gourdoux-Daux (1817: 32), for example, instructs the reader, in landing from the jump, “to alight on the toes and let the heels come to the floor gradually, holding the knees straight.”

In a number of the steps from his handbook, Roller explicitly instructs the reader to land on the toes without bending the knees. In his description of the entrechat, for example, he writes (1843: 102) that “the land, however, must be onto the tips of the toes with the knees wholly stretched, which must not slacken or bend, for that would look dull; there must be no lack of strength visible with the entrechat.” Likewise in his saut tiré, he (1843: 212-213) has the dancer “jump up strongly in a vertical line, bend the knees and draw the feet under the body as much as possible, and in landing, you must be well stretched in the knees and high on the toes.” Again, in his spaccata (or spajato as he misspells it), he (1843: 200) writes that “in landing on the very tips of the toes, one must be in the closed position again without, however, bending in the least.” Similarly with the salto cavallo (1843: 218), “in landing, you must be again stretched on the toes of the feet in first position.”

Fig. 8. Notation of an entrechat à quatre (Saint-Lèon 1852: 24), after an original.

Examples of isolated jumps recorded in the notational systems developed by Saint-Léon (1852) and Zorn (1887) regularly show the dancer in such jumps landing on the toes with the legs held straight. Indeed, Saint-Léon’s verbal description of the entrechat accompanying his notational characters (1852: 24) makes explicit reference to a land onto straight legs (fig. 8): “in fifth position bent, with the left leg in front, movement into the air beaten behind and in front, land in fifth stretched, with the left foot in front, which makes up an entrechat à quatre.” See also fig. 11, which shows a changement de pieds, again landing on straight legs. Perhaps the most compelling evidence for this kind of land is found in Emmanuel (1895), who shows, in sequential line-drawings or in strings of time-lapsed photographs, a (rather flaccid) dancer executing a number of basic jumps in isolation, which end not in a plié, as in contemporary classical ballet, but on the toes tendu (fig. 9).

Fig. 9. A jeté dessus (Emmanuel 1895: 141), after an original.

As with the eighteenth-century sources, those from the nineteenth century also speak of lands onto bent knees in certain circumstances. For example, Roller includes a retombé plié only as a distinctive part of a given step, in the échappé, for example, or in the preparatory jump he calls the pas tombé (not to be confused with the eighteenth-century tombé) preceding the sissonne and other steps (such as the sundry versions of the splits, the salto tondo, the saut tortue, the saut genou, the salto cavallo, the saut de pendu, and the saut raid et rompu, for example). According to Roller (1843: 111), this preparatory jump landing in a bend was performed as follows:

the Pas Tombé [i.e., Retombé]. A fallen step – or one could also say a deeply sprung step – is done as follows in all the positions. It is assumed that the pupil is standing in first position. He springs lightly, enough that both feet leave the floor, but at that very moment [of landing], he bends, the body coming quickly down and is in the position as if he had merely bent, as in 14 [showing a grand plié, with the heels off the floor, not reproduced here] but not as deeply as is shown in the illustration. He stretches forcefully and quickly at once but without springing, the body returning to the position on the toes as is shown in 13 [illustrating a relevé on the toes, not reproduced here]. This exercise must be repeated in all positions until it is executed with extraordinary lightness, as the tombé is used in the enchaînement of many steps. In order to do this step well and surely, the body must never lose its perpendicular line.

Fig. 10. Position of dancer after landing from the first jump in the sissonne (1834: pl. 3, fig. 3), after an original.

Likewise, Costa (1831: 154), in his description of the sissonne, instructs the dancer to land from the opening assemblé “bent on the toes of the feet” and refers the reader to fig. 10. This is in agreement with the execution of the same part in the eighteenth-century counterpart: cf. Rameau’s comment (1725b: 75) “land in a bend on both feet,” as well as Siris’s remark (1706: 32) “Spring either forwards or backwards to the Third Position, upon the two Points of the Toes, and at the same instant you must Bend your two Knees to rebound,” i.e., on the toes with the knees bent.

When jumps were performed in unbroken succession, the nineteenth-century sources typically have all the jumps of the sequence – or better said, the jumping steps that terminate with a land – end in a retombé plié except the last one, which ends in a retombé tendu, if the last one is not immediately followed by a bend. This is in full agreement with the practice fleetingly referred to in the eighteenth-century handbooks. Gourdoux-Daux (1817: 44), for example, notes that when two grands coupés – his term for beaten hops – were performed in succession, the first was done with a retombé plié: “In order to perform successively one or more grand coupes, sink upon the left foot at the moment you alight on it, then rise, unfolding the right leg and foot again as above and bring it behind the left leg for the second movement.”

Saint-Léon (1852: 43, 48) gives two different notations and descriptions of the changement de pieds. In one combination (fig. 11), wherein a single changement is followed by a simple rise onto pointe, the dancer lands tendu from the jump (represented by the first symbol in bar 5 of the figure):

during the two last eighth notes of the fourth measure, there are two movements: the first with the legs bent in the same position as the foregoing [i.e., fifth with the dancer’s right in front], the second with the legs stretched in first in the air [shown by the superscript symbol above the 1]. . . . And during the first quarter note of the fifth measure, the legs stretched on the floor, with the left leg in front.

Fig. 11. Notation of a changement de pieds with retombé tendu (Saint-Léon 1852: pl. 3), after an original.

Later he shows three changements in a row followed at once by an échappé plié (preparation for an ensuing pirouette). Here the dancer lands in a plié in each of the jumps (fig. 12), as a preparatory bend is needed for each ensuing step of the combination. He writes (1852: 48),

the eighteenth example shows: preparatory position in fifth, legs stretched, right in front, bras bas, body épaulé to the left. During the quarter-note rest: legs bent in fifth, right in front, body back to de face, shown by the natural. During the first eighth note: legs stretched in first in the air. During the first eighth note of the first measure: legs bent in fifth, left leg in front. During the two sixteenth notes: legs stretched in first in the air. During the other two quarters of the measure, the same movements are repeated. During the first eighth note of the second measure: legs bent in fifth, left in front. During the second eighth note of the same measure: legs stretched in the air and in second, bras ouverts. During the two following counts of the second measure: legs bent in second, demi-bras with the left arm, right arm extended, body épaulé to the right. During the third measure: cou-de-pied position, left leg in the air, sign “turn to the left,” four turns, bras bas. This sign continues to the third quarter of the fourth measure. During the last count of the fourth measure, the enchaînement starts again with the other leg [and the remainder of the notation is not shown here]. It ends with the legs stretched in fourth, left leg in front, arms in what is called au public. This sequence of movements constitutes a combination of three changements de pieds, échappé, pirouette sur le cou-de-pied, posé devant, and the same with the other leg.

Fig. 12. Notation of changements de pieds with retombés pliés (Saint-Léon 1852: pl. VI), detail after the original.

As discussed in detail in the foregoing section (§4.2.6), the early-nineteenth-century dancer could come down into a forced bend in his retombé plié, that is, with the knees bent and the heels off the floor (fig. 10). Isolated references to this practice in Roller (1843) suggest, moreover, that a land in a forced bend was also seen as the usual preparation for a grand jump, although in such a case, the heels would need to be close to the floor out of biomechanical necessity (see the rightmost foot in fig. 4).

The current practice of invariably landing from a jump in a true demi-plié (with knees bent and the soles of the feet set down on the floor) appears to have come into existence only in the twentieth century, although even in the early part of that century elements of the old practice were, it seems, still discernible in some schools of ballet. It is apparent from Vaganova’s handbook of 1934 and the Craske-Beaumont Cecchetti handbook of 1930 that a land in a demi-plié was the norm in those schools by the 1930s. The latter source (1972: 9) writes that

in alighting after a pas d’élévation, the tips of the toes should be the first to reach the ground, the sole of the foot following immediately, taking care that the heel is always completely lowered. Failure to lower the heel and to take a strong demi-plié results in a loss of ballon, that is, the smooth falling and rising of the feet in the passage from step to step.

Bruhn (1963: 50-51, 9), however, writes that landing from a jump in a soft demi-plié was only a relatively recent development in the Bournonville school, which formerly cultivated a more “brittle” manner of landing that could strain the knees:

not many years ago, it was usually a rather brittle, bouncy jump, which sometimes proved hard on the knees. . . . in recent years, through better teaching [introduced in the 1950s by the Russian teacher Vera Volkova], the Danish dancers have rediscovered [sic] the proper way to land, through the foot and with a soft deep smooth demi-plié, without lessening the height or brilliance of the jump.

In Bournonville choreography, the work-around for this old “brittleness” was an avoidance of holding a position after a jump, or arranging steps, as Bruhn adds, “so that the landing from a jump was disguised by the quick take-off for another jump or some other rapid movement; a jump was seldom followed by a sustained position which would expose the abrupt manner of landing.” Theory Versus Practice

Despite his statements cited above that the knees were not to bend at all in landing from such jumps as the entrechat, spaccata, salto cavallo, Roller (1843: 109-110) elsewhere notes broadly that a little slackening in the knees was possible in a retombé, presumably the retombé tendu, although his wording suggests that such yielding was not to look like a true demi-plié. That is, the knees relaxed only “a little” to help absorb some of the impact from the land (and were presumably quickly stretched again after the impact had been deadened):

in landing from a hop or spring, one must come down onto the very tips of the toes. Moreover, not only does a land onto the flat foot look heavy and is unpleasant to the ear but the body receives a jolt and loses its lissom fluidity for the following movement. In an elastic land onto the toes, the ankle joint[s], and right after the knees, give way a little, and the land is deadened, as it were.

Elsewhere Roller (1843: 39) again notes broadly that the dancer was to land on the toes and could soften the knees a little as well:

to spring off and land upon the whole sole is clumsy and strenuous, but springing lightly off the toes and coming back down onto the toes is easier. Landing on the toes softens the fall while the joints of the feet, of the knees, and those at the hips together yield. Coming down onto the heels, however, produces a hard jolt. These methods should also be strictly observed by gymnasts in jumping and swinging (or vaulting).

The same inconsistency can be found in Bartholomay (1838).

The question then arises if this slight softening of the knees in the nineteenth-century retombé tendu was also permitted in the eighteenth-century counterpart, if theory and practice in our period of study could in like manner be somewhat at odds, at least concerning the lands from grand jumps. It seems rather likely, however, that a “compromised” retombé tendu was also known at least in the last few decades of the eighteenth century. Roller (1775-1844) had in fact received his dance training in the late 1780s and had been a professional dancer in the serious style until 1799. In a few places in his handbook, Roller draws the reader’s attention to changes in dance conventions. Significantly, he does not indicate that this softer manner of landing was a novelty, although absence of such a remark is no proof that such lands were not new in his day. Certainly such a compromising practice is not at odds with Magri’s warning (1779: 1/129) that

you must see to it that at the end of the spring the body is freed of strength and thus will be aplomb where it lands, because if the limbs remain full of vigor and rigidity, the body will spring up again upon landing therefrom, resulting in a lose of upright balance and occasioning a fall to the floor.

Magri’s words are reminiscent of Noverre’s statement (1760: 330) that

the sudden change from relaxation to a strong tension and from flexion to a violent extension is thus the cause of a multitude of mishaps which would doubtless be less frequent if the dancer were to yield, as it were, to the fall and if the weak parts did not try to resist a weight that they cannot support or overcome.

If Noverre is referring here to a compromised retombé tendu, then this practice may have been new in his day – his Lettres were written apparently in the 1750s – as he seems to suggest that many dancers did not “yield” to the fall (i.e., slacken all the leg-joints somewhat) and that the “weak parts” (the feet alone) were responsible for “resisting” the weight of body.

Given that a compromised retombé tendu may have been used with grand jumps at least in the latter part of the eighteenth century, I have included an illustrated reconstruction of it in fig. 13. The degree of slackening in the leg-joints would presumably have been dependent upon the height of the jump and the dancer’s strength.

Fig. 13. A reconstruction of a compromised retombé tendu.