The collection

1The collection of drawings is catalogued at the Bibliothèque Centrale, Muséum National d’Histoire Naturelle, as Manuscript n^o 5007. Most of the drawings are watercolors, some are blackand-white pencil or ink, and some have both colored and black-and-white portions. The vast majority of them depict mature, actually overripe or even senescing fruits. A portion of the drawings show more than one fruit. Almost all of the fruits are numbered in series from 1 through 98. Most of the other drawings are designated in a series with most of the Latin letters from A through T (R is absent), and depict flowers and foliage. Although catalogued as Manuscript N^o 5007, the collection of drawings was unaccompanied by any written material until recently, when it was supplemented with six documents (Table 1; Paris 2002a), all of which were prepared long after all but one of the drawings were made.

2The drawings of Manuscript n^o 5007 are divided into two voluminous portfolios. Portfolio Ms5007(1) has four parts: I, drawing series 1 to 33, consisting of 89 drawings on 46 sheets; II, drawing series 34 to 47, consisting of 52 drawings on 34 sheets; III, drawing series 48 to 63 consisting of 40 drawings on 39 sheets; IV, drawing series 64 to part of 76 consisting of 38 drawings on 32 sheets. Portfolio Ms5007(2) has three parts: V, rest of drawing series 76 to part of 90 consisting of 76 drawings on 47 sheets; VI, rest of drawing series 90 to 98 plus two more recent drawings, consisting of 48 drawings on 39 sheets; VII, drawings A to T (without R) and two which bear neither a number nor a letter, consisting of 21 drawings on 21 sheets. Total: 364 drawings of which 361 were made by Duchesne himself, two (N^o 96* and an unnumbered fruit of turban gourd) were by the artist, Jean Louis Prévost, and one (in black-and-white, unnumbered fruits apparently openpollinated progeny of a turban gourd) by his daughter, Lucette Duchesne. They are mounted on 258 sheets which are quite large, 64 x 48 cm (reproduced at sixty percent of the original size in the present volume). Altogether, 615 fruits are illustrated. The drawings of large fruits as well as groups of small fruits were made directly on the sheets whilst the drawings of the small, individual fruits were most often made separately and then mounted on the sheets, often more than one to a sheet.

The purpose of the drawings

3The main purpose of these drawings was documentation. Prior to the invention of photography, it was often necessary to draw highly accurate and detailed depictions of experimental materials and/or results for the same reason that photography has been employed in more recent times, that is, for documentation and for serving as a basis for further experimentation. Duchesne had been taught scientific drawing by his father. He used this artistic skill to document his work with strawberries (Staudt 2003). His drawings with Cucurbita are all the more magnificent because, unlike the strawberries, most of the illustrations of Cucurbita are aquarelles, that is, in watercolor.

Dates and sites of the drawings

4Most of the depictions of fruits bear dates. Of these, the vast majority were drawn during the months of December 1769 through 1773 and January, February, and March of 1770 through 1774. These, then, were drawn from fruits that grew on plants that had been raised during the previous summers, 1769 through 1773. There is one magnificent watercolor of part of a plant bearing an unripe fruit from September, 1775. There are also thirteen black-and-white sketches of plants or plant parts that bear dates, most of which are from August and September of 1773, and some from 1779 and 1780. The two pictures of turban gourds were added in 1785 and 1786. The black-andwhite drawing by Lucette Duchesne was added much later, in 1796. The sub-title on the second page of Document 3 with the drawings, the Voucher, indicates where many or all of the drawings were made:… dessinées et coloriées d’après nature, tant au Château de Meudon qu’à celui de Trianon à Versailles… (… drawn and colored from nature, both at the Château of Meudon and at that of the Trianon of Versailles...).

Literary material accompanying the drawings

5As already described, six documents related to the drawings were found and added to the collection of drawings at the Muséum national d’Histoire naturelle in 2000 and two other documents related to the acquisition of the drawings by the Muséum were found in the Bibliothèque municipale de Versailles in 2005 (Table 1). There are also three publications by Duchesne in which quite a few of the drawings are specifically mentioned. Two of these publications are his contributions to Lamarck’s Encyclopédie Méthodique, Botanique (Duchesne 1786c) and Tessier and Thouin’s Encyclopédie Méthodique, Agriculture (Duchesne 1793). These two differ by alluding to different drawings, the former referring to the series of numbers from 1 through 79 and the latter being restricted to the series of numbers beginning with 80.

6Of the three publications, the most important includes all of the allusions to the drawings made in both of the encyclopedia contributions. This is of course the Essai sur l’Histoire Naturelle des Courges, described earlier.

7The Essai and the contribution to Lamarck’s encyclopedia begin with an overview of the gourd family, the Cucurbitaceae, as described earlier. This overview of the cucurbits is followed by a section comprised of three paragraphs headed Caractère Générique, on the distinguishing characteristics of the genus, almost the entire scope of which focuses on the floral parts. As a number of the drawings are of the floral parts, I will append this section of the Essai to the presentation and brief description of those drawings.

8After the section of Caractère Générique is inserted a passage written by the editor, Lamarck, who added some background information on the drawings. His introductory remarks appear in the encyclopedia contribution and are offset in the Essai with a heading, as follows:

The Editor announces at this point the use that he made of the manuscript that had been communicated to him, and does it in these terms:
“Mr. Duchesne, who has cultivated over several years the plants of this genus in the view of establishing the effects of cross-fertilizations on their different races, and has described and drawn all of the fruits that he has seen them produce, and as having agreed to communicate this interesting work that he has done on this matter, we are going to present it in almost its entirety in this article. It will be observed that the numbers cited in the descriptions indicate those of the colored drawings that Mr. Duchesne has made according to nature, which are numerous drawings, well-done, and which, presented in 1779 before the Royal Academy of Sciences in support of a highly detailed treatise from which this article is extracted, and having precedingly been placed at Trianon before the eyes of Louis XV, are found today, according to the intentions of His Majesty, deposited in the Cabinet of Stamps of the Royal Library, where they can be consulted.”

The numbering of the drawings

9Lamarck, in his introductory remarks to Duchesne’s contribution to the Encyclopédie Méthodique, Botanique, wrote: On observera que les numéros cités dans les descriptions, indiquent ceux des dessins coloriés que M. Duchesne a faits d’après nature… [It will be observed that the numbers cited in the descriptions indicate those of the colored drawings that Mr. Duchesne made according to nature]. By comparing the numbers used in the descriptions in the text of that article as well as in the text of the Essai sur l’histoire naturelle des courges with those of the paintings of fruits possessing easily discernable characteristics, such as striping, bicolor pattern, and warts, it can be verified with certainty that the numbers do indeed correspond exactly, except for several errors in typesetting. The numbers can be sorted into three categories.

10One category of numbers is the number alone (e.g. 1, 14, 28, 43, 68). The collection includes drawings labelled with all numbers alone from 1 through 98, except for n^o 48 and n^o 89. Numbers alone designate fruits of the different cultivars, as is evident from the text … les autres numéros de 15 à 28, donnent de véritables variétés qui doivent toutes se rapporter à la race des Cougourdettes … […the other numbers 15 to 28 given of true varieties which all have to belong to the race of Egg/Pear Gourds…]. Here it is necessary to point out that the term variétés, as used at the time, could mean varieties, that is, cultivars, or could mean variants, a term that was coined later on. Duchesne’s use of véritables variétés [true varieties] could only mean the former.

11The second category of numbers are those having suffixed letters (e.g. 1ba, 43c, 68a). These refer to offspring of cultivars, that is, variants produced by cross-pollination, as for n^o 43: Le n^o 43, qui se trouvoit ovale, chargé de peu de bosselures, mais vert & marqué seulement vers la queue d’un très-grand panache jaune, m’a produit des variétés nombreuses & assez différentes: un seul 43a avoit sa forme, sa grosseur & ses bosselures, mais sans panache…; 43b étoit petit & rond, 43f, long & triple de grosseur; 43c, avec les mêmes couleurs & la même forme ronde, n’avoit point de bosselures… [n^o 43, which is seen to be oval and bearing a few warts, but green and marked only near the peduncle with a very large yellow patch, produced for me numerous and quite different variants: only 43a had its shape, its size, and its warts, but without a patch…; 43b was small and round, 43f long and triple the size; 43c with the same colors and the same round shape but with no warts at all…]. Each letter suffixed represents one generation of crosspollination. Those described above for n^o 43 represent the first-generation progeny of cross-pollination. Numbers 43aa, 43ab, and 43ba resulted from two generations of cross-pollination but the former two are from a different family than the last: 43aa and 43ab were cross-pollinated offspring of 43a whilst 43ba was a cross-pollinated offspring of 43b. Another example from the text of the Essai: En seconde génération, des variations biens plus grandes... 82ba étoit devenu une Cougourdette, en Coloquinelle; 82ca au contraire étoit un Giraumon blanc… [In the second generation, the variations were a great deal larger… 82ba had become a Egg/Pear Gourd, in the form of a Round Gourd; 82ca on the contrary was a white Long Squash…]. Indeed, Mendelian genetics would dictate that the second generation of forced crosspollination would have involved highly heterozygous, first-generation parents and would therefore be expected to produce tremendous variation in the progeny.

12Many of the drawings bear dates. It can be noticed from the dates that, for a given family, there rarely occur cases in which fruits having more letters suffixed are dated with a year earlier than those having fewer. The fruits that Duchesne obtained from his research plots in the late summer or autumn of each year could be kept for months, perhaps a year for some small ornamental gourds but it would be highly unlikely that they could be kept for longer than that. In a few of the pictures, more than one fruit is depicted and each fruit is designated with a letter. This letter is to be understood as the last letter in a suffix. The fruits in these pictures are from the same generation, and their being depicted together was to save space.

13The third category of numbers, which appears only in association with no 14, has a superscript number (e.g. 14¹, 14⁴). What Duchesne, in his use of superscripts, had in mind was slight variation within the same cultivar: Sous le n^o 14 se trouvent de légères differences de forme, de grosseur & de panaches, qui ne peuvent porter le nom de variétés... [Within n^o 14 are found slight differences in shape, size, bicolor patches, which cannot carry the name of varieties…]. N^o 15-28, which bear some similarity to n^o 14, were given separate numbers by Duchesne because he recognized them as “véritable variétés”, that is, separate cultivars.

14There are some cases of numbers being supplemented with an asterisk, for example, there is a n^o 83 and a n^o 83*, which are similar. The asterisk seems to have been used to indicate a fruit from a second plant of the same cultivar, or more accurately, a second plant grown from a seed originating from the same seed packet, of which there are instances of very different fruits, such as between 56 and 56*. But there does not seem to be a consistent usage for the asterisk. In some cases, it seems to replace a question mark (?), indicating uncertain ancestry or identification. The supplemented letter x also seems to designate an unknown or questionable ancestry or identification.

15Duchesne labelled a number of the drawings, usually ascribing an illustrated fruit to a literary source, most often the Historiae Plantarum Universalis of Jean Bauhin (1651). In some cases, these phrases also appeared in subheadings of the Essai and of the Encyclopédie Méthodique, Botanique. N^o 84, an orange-yellow lemon-shaped fruit, is labelled Cucurbita clypeata affinis ad citrum nonnihil accedens and is attributed in the text to J. Bauhin. Another, n^o 93n, of a long, curved, striped cocozelle fruit and part of the plant bearing it, is labelled Cucurbita aspero folio flore luteo semiarcus in morem intorta J. B. Some comments, though, are not attributions to J. Bauhin or anyone else. The most significant of these, borne by n^o 96*, is Cucurbita moschata.

Numbers of individual fruits

16In a number of the drawings containing more than one fruit, four-digit numbers appear beside the fruit, in addition to the letter suffix of the series number. For example, the progeny of 82c (82cb, cc, cd, ce, etc.) have beside them numbers, respectively 3320, 3314, 3158, 3313, etc. The four-digit numbers are of the original numbering system used by Duchesne, before he transcribed them to the 1-98 series.

The initial numbering system and dates

17From Document 6 accompanying the drawings, the catalogue listing entitled Travail des Pépons, can be deciphered Duchesne’s initial numbering system. Of the many entries, that of n^o 82cb, which initially had been designated as 3320, can serve as simple example: cb. & c. de 1035. de 127.9 et 11 fevr. 1773. The inscription de 1035. de 127 (from 1035, from 127) designates the maternal ancestry of n^o 82cb, with n^o 82c initially having been designated as 1035 and n^o 82 initially having been designated as 127. The depiction that includes n^o 82cb had been drawn on 9 and 11 February 1773. From comparison of the numbers and dates, it is apparent that numbers of less than 1000 were drawn during the winter of 1769-1770 or the months just before or after, from fruits grown in 1769; the 1000-series numbers were drawn during or just prior or after the winter of 1770-1771 (from fruits grown in 1770); the 2000-series numbers were drawn during or around winter 1771-1772, the 3000-series in winter 1772-1773 and the 4000-series in winter 1773-1774. For drawings on which the date is not inscribed, I have included in my commentary dates that can be deduced from the initial numbers listed in Document 6.

Conclusions concerning the numbering

18The number series of 1-98 is arranged in accordance with Duchesne’s taxonomic treatment of Cucurbita. N^os 1-93 are of his Cucurbita polymorpha (= C. pepo), in five races: n^os 1-13 of C. p. Colocyntha, n^os 14-31 of C. p. pyridaris, n^os 32-44, 46, 47 of C. p. verrucosa, n^os 45, 48-79 of C. p. oblonga, and n^os 8093 of C. p. Melopepo. N^os 94-96 are of C. moschata and n^o 97 and 98 are of C. maxima. The illustrations of the turban gourds of C. maxima, late additions to the collection, do not bear numbers.

19Although it is clear from the drawings bearing dates when they were made, it is not clear exactly when the numbers of the drawings were transcribed from the initial numbering system to the 1-98 series. Duchesne could not have known ahead of time precisely how to classify each of the individual cultivars (varieties), and therefore the numbers would likely have been transcribed some time after the completion of his experimentation with cross-pollination. The larger fruits were drawn directly on large folio sheets whilst most of the smaller ones were drawn on thinner paper that was later cut and mounted onto large folio sheets. The transcription to the number series of 1-98 seems to have been done before mounting, as one of the drawings, n^o 12a, has the “1” cut off. My guess is that the transcription to the number series of 1-98 was followed immediately or shortly afterwards by the mounting of the drawings on the sheets, and this was in anticipation of and in order to facilitate the presentation of his research before the Royal Academy of Sciences in 1779.

20There seem to have been a few errors that were made in transforming the initial numbers to the number series. For example, there is duplication of n^o 14a and 14b. The two illustrations of no 14a were drawn on the same date, 25 April 1770. The fruits are so different, however, that they could not possibly have come from the same plant. Document 6, the Travail des Pépons, is of no help in this particular case, listing no n^o 14a but one 14xa. The same situation occurred concerning the n^o 14b. From Documents 5 and 6, it can be learned that there was also duplication of n^o 93. Given the prodigious number of illustrated fruits, it is certainly possible that errors could have occurred in various ways.

21Paintings of n^o 48 and n^o 89 are missing even though their fruits are described in the text. Their depictions were missing when Document 5 and Document 6 were prepared (ca. 1796). Although Duchesne took notes on these, it is possible that the drawings were simply misplaced or perhaps the fruits rotted before Duchesne found time to draw them; in several drawings, fruits are depicted as beginning to rot.

22There are over a dozen cases of first-generation fruits missing. For example, illustrations exist for n^o 1 and for n^o 1aa, 1ab, 1ac, 1ad, 1ae, and 1af; so there must have been a fruit no 1a but there is no illustration of it in the collection. Here again, it is possible that fruit n^o 1a decayed or became discolored to the extent that it was not worthy of being illustrated before its seeds were extracted, or that a drawing of it was simply lost from the collection. Another example is the existence of n^o 46 and n^os 46aa and 46ab, but n^o 46a is absent. There is one case of a second-generation fruit missing: n^os 92a and 92aaa are in the collection but n^o 92aa is missing. None of these three missing fruits are listed in Documents 5 and 6. Document 5 gives short descriptions of n^os 14d, 14e, 14f, and 14f*, but these are not included in the collection of drawings. Altogether, there are 615 fruits depicted in the collection. Duchesne (1786a, p. 6) mentioned that he raised between 1000 and 1200 pepos but it is unclear if he meant he raised that number of plants or drew that number of fruits.

23Duchesne began drawing fruits in the months following the summer of 1769 and there are a few instances of numbers bearing a letter suffix that were drawn during the early months of 1770, for example 6d, 11b, 14b, 40a, 42b, 47a, 76a, and 85e. This indicates that Duchesne must have grown these fruits in 1769. It follows that his experimentation with cross-pollination must have begun no later than in 1768 by his planting seeds of at least some of the cultivars that he collected, specifically n^os 6, 11, 14, 40, 42, 47, 76, and 85. Likewise, he must have planted remnant seeds of these same cultivars again, in 1769 or later, in order to illustrate their mature fruits.

24There are also several printing errors in the text of the Essai and the contribution to Lamarck’s encyclopedia. I have included the text of the Essai, below, to escort the drawings and where I have recognized printing errors within, I have given the correct numbers and placed the wrong numbers in brackets {}.

Other Explanatory Remarks

25All of the drawings are to scale. Most of them are at their actual size but a few are scaled down and were specifically mentioned as such by Duchesne on the drawings themselves. The dimensions given in my descriptions of the individual fruits I made directly from the drawings.

26Moreover, in the following descriptions of the drawings, each drawing that bears a date is specifically mentioned as “Dated”. For each which does not bear a date but for which the date is listed in or can otherwise be deduced from Document 6 (Table 1), the date is given but the word “Dated” is not included in the description. Each drawing for which the date cannot be ascertained is described as “Undated”.

27The numerical order of the drawings does not always follow the numerical order of the plates in the portfolios. In the descriptions that follow, the numerical sequence of the drawings is maintained in order to make the results of Duchesne’s experimentation more easily comprehensible. Thus, the order of the plates has been compromised on a number of occasions.