Introduction

Snake melons, Cucumis melo L. subsp. melo Flexuosus Group (Cucurbitaceae), are widely grown in the Old World tropics and subtropics, from northern Africa to the Near East to the Indian subcontinent (Chakravarty 1966; Walters and Thieret 1993; Pandey et al. 2010). While most melon fruits have a nearly 1:1 ratio of length to width, cultivars of the Flexuosus Group are distinguished by their having fruits of a 4:1 or higher ratio (Pitrat et al. 2000; Burger et al. 2010). This great elongation of the fruits is characteristic of cucurbits that have been intensively cultivated and selected for the consumption of their immature fruits (Paris 1989).

Cultivars of snake melons differ from one another in the exterior color of the immature fruits, which can be dark green, light green, or striped dark and light green (Chakravarty 1966; Pandey et al. 2010). They can also differ for longitudinal furrowing of the fruits from very shallow to very deep. Moreover, the 4:1 length-to-width ratio is often greatly exceeded as, for example, fruits over 50 cm long and only 5 cm wide are a common occurrence. Unlike the familiar dessert melons, C. melo subsp. melo Reticulatus, Cantalupensis, and Inodorous Groups, melons of the Flexuosus Group do not become sweet as they ripen, but instead become insipid or sour (Burger et al. 2006).

Like cucumbers, Cucumis sativus L., snake melons are harvested when they are immature and green, and typically are used when fresh or pickled (Walters and Thieret 1993; Pandey et al. 2010). Like other melons, snake melons flourish in warm, sunny climates but, unlike cucumbers, they are ill-adapted to cooler climates. For these reasons and more, scholars from temperate regions identified cucumbers as the elongated fruits appearing in depictions from ancient Egypt and from various Mediterranean civilizations of Roman times. Likewise, the literary allusions of antiquity, to the Greek sikyos, the Latin cucumis, and the Hebrew qishu’im, were thought to be to cucumbers. Unlike cucumbers, however, fruits of C. melo subsp. melo, including snake melons, are distinctly hirsute, softly hairy, when immature (Fig. 1; Kirkbride 1993; Jeffrey 2001). As pointed out by Feliks (1967), the descriptions by the Roman writers Columella and Pliny of the hairy downiness of the cucumis fruits are certainly not consistent with the fruits of C. sativus, which are glabrous, but instead with C. melo. Hebrew writings of Jewish Oral Law known as the Mishna and Tosefta, dating to the the second and third centuries CE, respectively, repeatedly mention that the qishu’im (also qishu’in, sing. qishut) underwent piqqus prior to being eaten. Saul Lieberman (1955), scholar of ancient Greek and Jewish Oral Law, observed that the ancient Greek pekos was the source of the Hebrew piqqus, and meant the removal of soft hairs. This Hebrew epithet, in functioning as a verb, becomes lefaqqes (inf.), mefaqqes (first pers. sing.), etc., and thus appears to be the origin for the Arabic epithet faqqous (Paris and Janick 2008a), which is used to the present day to denote the snake melon across the extensive Arabic-speaking geographical region south and east of the Mediterranean Sea (Issa Bey 1930; Hassib 1938).

Fig. 1
figure 1

Part of harvested, young, snake melon fruits, faqqous, of Cucumis melo subsp. melo ‘Armenian Striped’ (left) and ‘Tortarello’ (right) showing the hairs, piqqus, on the fruit surface

Full size image

The primary objective of this study was to assess the historical importance and value of snake melons by tracing the written records of the epithets piqqus and faqqous in the late classical and medieval periods. Another, secondary objective was to clarify how exactly piqqus, removal of the hairs of the fruit, was accomplished, based on the results of a modest experiment and supported by literary evidence. First, though, I will review the literature pertaining to the Mediterranean history of the snake melon from ancient Egypt through the Roman period.

Overview of the snake melon in ancient and classical times

The geographical origin of Cucumis melo is controversial, as plants of this species have been reported as growing wild on several continents. New evidence has emerged that strongly indicates the Indian subcontinent as the central place of origin (Sebastian et al. 2010). Wild or primitive melons have also been described from elsewhere, including Sudan; the indigenous, cultivated melons growing there were described as small, striped, and varying from oblate to oval, and eaten fresh, pickled, and used as cooking ingredients (Mohamed and Yousif 2004).

The observations from Sudan are significant, given the close proximity of Sudan to Egypt and the Mediterranean Sea. Melons are the cucurbit most frequently encountered in Egyptian paintings and Mediterranean artwork and literature of antiquity (Zohary and Hopf 1993; Janick et al. 2007). Most of the melons illustrated on Egyptian wall paintings over 3,000 years old have a length-to-broadest width ratio of 2:1 or 3:1 (Keimer 1924). These, therefore, were proportionately longer and narrower than the indigenous melons of Sudan and typical of C. melo subsp. melo Chate Group (syn. Adzhur Group). However, a few of the melons found in these paintings were longer and narrower, having a length-to-broadest width ratio between 4:1 and 6:1 and therefore typical of the Flexuosus Group, and one painting shows these long melons as distinctly striped and another depicts them as furrowed (Manniche 1989). This is evidence that people had begun selecting long-fruitedness in C. melo over 3,000 years ago for culinary use of young fruits (Paris 1989). These melons are the qishu’im longed for by the Children of Israel during their exodus from Egypt (Numbers 11:5), and subsequently widely grown in Israel itself as a special word for a field of this crop, miqsha, appears in prophetic literature (Isaiah 1:8) (Feliks 1967; Paris 2009).

In Roman times, snake melons were a widely grown, much-appreciated cucurbit crop in lands around the Mediterranean Sea, as is plain from the collection of images presented by Janick et al. (2007). Another image of snake melons, not included by those authors, is found in a partially preserved wall painting from the ruins of Ercolano (Herculanum), a city destroyed by the eruption of Mount Vesuvius in 79 CE. In this painting, serpent-like striped melon fruits, closely resembling those of the extant ‘Striped Armenian’ (Goldman 2002) are depicted in and adjacent to a large glass jar. The representation of a glass jar suggests usage of snake melons for pickling. Obviously, snake melons, at least when not cut into pieces, would be ill-adapted for this purpose as quite a large jar would be required to hold them. Apparently, the snake melons were most often eaten fresh, though for out-of-season use they would have to have been preserved as pickles.

The esteem of snake melons is reflected in the Roman literature of the first century CE. Columella wrote of the hairy, snake-like cucumis (Forster and Heffner 1955) and Pliny wrote that the cucumis is covered with white down and can be made to grow into all sorts of long shapes (Rackham 1950). According to Pliny, the Emperor Tiberius of Rome commanded his subjects to make the hairy, snake-like cucumis available to him all year round (Paris and Janick 2008b).

The appreciation of snake melons is also apparent in the Jewish literature of the classical and early medieval period. The qishu’im (or qishu’in, sing. qishut) are discussed more than any other cucurbit crop in the codices of Jewish Oral Law known as the Mishna and the Tosefta, compiled in Israel circa 200 CE and 300 CE, respectively. One of the tractates in these codices (Mishna, ‘Oqazin 2:1) even contains a pun, keshut shel qishut, meaning the dense, soft downiness of the snake melon. One aspect concerning the qishut in these codices was the need for it to undergo piqqus, that is, the removal of its downy hairs, prior to eating. This discussion among rabbinical sages was repeated and elaborated in the Jerusalem Talmud, completed in Israel circa 400 CE, and the Babylonian Talmud, completed in Iraq circa 600 CE. In the latter work (Tractate Berakhot 57b), it is written that the great sage and compiler of the Mishna, Yehuda the President, had qishu’in on his table all-year-round to honor his many visitors.

Piqqus: rubbing off the hairs of young cucurbit fruits

The analysis, over the ages, of how piqqus was accomplished has been reviewed and evaluated exhaustively by the great Talmudic scholar Saul Lieberman (1955). I will not repeat his lengthy treatment. Overall, it is agreed that piqqus is accomplished by rubbing the fruits, but one question seems to have remained. In Jewish Oral Law, piqqus is mandated for two cucurbits harvested when still hairy, and therefore immature, the qishu’im and the delu’im, the latter being edible, long-fruited bottle gourds, Lagenaria siceraria (Molina) Standl. Similarly, sheliqa, generally agreed to be a dipping in boiling water, is mandated for two other cucurbits, used when ripe, avattiah, watermelon, Citrullus lanatus (Thunb.) Matsum. et Nakai, and melafefon, round melon, most likely C. melo subsp. melo Adana Group. In his voluminous, critical treatment of the Tosefta, Lieberman (1955, p. 671) suggested that sheliqa might have been used to remove the hairs on a wide variety of vegetables. Feliks (2005), who identified correctly numerous plants mentioned in Scripture and Jewish Oral Law, took this comment one step further, and wrote that maybe piqqus was synonymous with sheliqa, that is, dipping of the fruits in boiling water would remove the hairs and might also improve their taste.

Given the doubt suggested by Feliks (2005), I conducted two modest experiments, in separate years, on young melon fruits. In the first trial, I dipped in boiling water a fruit of ‘Carosello Barese’ (Cucumis melo subsp. melo Chate Group), which is very hairy, and part of a fruit of ‘Striped Armenian’ (C. melo subsp. melo Flexuosus Group), which is moderately hairy. I observed that the dipping did not accomplish piqqus but rather quite the opposite, it appressed the hairs to the epidermis, making them more difficult to remove. I did this a second time, but this time I kept the fruits in boiling water for 5 or 10 min. No piqqus occurred; the hairs remained appressed to the fruit surface and were difficult to remove. I did not sense any improvement in flavor of the boiled fruits over the raw ones. In the second experiment, I used ‘Striped Armenian’ and ‘Tortarello’ (Flexuosus Group) (Fig. 1), doing the same as I did in the first experiment, and with the same results. Hence, piqqus (removal of the hairs) could not be accomplished by dipping in or prolonged treatment with boiling water. Piqqus and sheliqa, therefore, are not synonymous. Feliks (2005) suggested that watermelons and mature, round melons were mandated to undergo sheliqa to allow easy peeling of the rind. I did not test this, but there may be other explanations. Brief exposure of mature melons to hot water lengthens their shelf life (Fallik et al. 2000). Also, small insects and grit can get stuck in the hairs of immature cucurbit fruits and could only be removed efficiently together with the hairs; as I observed, the hairs can only be removed by rubbing, piqqus. However, merely dipping in boiling water, sheliqa, could be sufficient to remove insects and grit from “bald”, ripe fruits.

Corroborating evidence exists in a medieval source. The Mishne Tora, the great work of Moshe Ben-Maymon (Maimonides, 1138—1204), rabbi, philosopher, doctor, scientist, and lexicographer, was completed in what is now Egypt in 1178 (Kraemer 2008). The Mishne Tora is comprised of Maimonides’ analysis of the entire Oral Jewish Law, that is, inclusive of the Mishna, Tosefta, and Jerusalem and Babylonian Talmuds, and written in well-structured, clearly understood Mishnaic Hebrew. The Mishne Tora gathers together information from disparate parts of those four works and lucidly rearranges them into 14 books. In the Mishne Tora, many sentences are copied almost exactly as they appear in the four earlier works, with only slight changes to improve the grammar, sentence structure, or clarify the meaning.

Although the vocabulary used in the Mishne Tora is strikingly similar to that used in the Mishna and Tosefta, as Maimonides wanted his work to be understandable to as many readers of Hebrew as possible, there are some differences (Kraemer 2008). Of particular reference to piqqus are its conjugational forms in passages from the tractate Ma‘asrot of the Mishna, Tosefta, and Jerusalem Talmud, and tractates Beza and Bava Mezi‘a from the Babylonian Talmud. These passages are arranged together in the Mishne Tora, in Book Zera‘im, Chapter Ma‘asrot. In all cases, the conjugational forms of piqqus are replaced with the corresponding conjugational forms of shifshuf, rubbing. Here is a brief example:

We’im eno mefaqqes, misheya‘amid ‘arema

And if he does not conduct piqqus, then from [the time] when a pile is gathered

Jerusalem Talmud, Ma‘asrot, 1: 5

We’im eno meshafshef, misheya‘amid ‘arema

And if he does not conduct shifshuf, then from [the time] when a pile is gathered

Mishne Tora, Book Zera‘im, Chapter Ma‘asrot, 3: 9

Maimonides did not use the word lefaqqes, found in various conjugational forms in the Talmud, in his Mishne Tora. In its place, he used the word leshafshef, which means to rub. Clearly, Maimonides, the rabbi and linguist, knew what piqqus was and substituted it with a word that he thought would be more widely understood by his contemporary Hebrew-reading audience. Piqqus does indeed mean rubbing off of the hairs.

Usage of the epithet faqqous in the medieval period

From the above, Maimonides replaced the word piqqus and its conjugational forms with shifshuf for easier comprehension by the intended Jewish audience. I am unaware if the word piqqus survived in Hebrew documents beyond the time of the writing of the Babylonian Talmud, compiled in Iraq and finished circa 600 CE. After what appears to be a hiatus of some three centuries, its Arabic derivative, faqqous, made an appearance.

It seems strange that Abu Hanifa ad-Dinawari (d. 895), who studied in Iraq and compiled a lexicography of plants (Lewin 1960), did not use this epithet even though he briefly mentioned the qitha (Hamidullah 1973, 1993), the Arabic equivalent of the Hebrew qishu’im, which in both languages can be a generic term applying to both chate and snake melons, or specifically to snake melons. Faqqous, however, refers exclusively to snake melons (Issa Bey 1930; Hassib 1938).

Ar-Razi, ca. 920 CE

Muhammad ibn Zakariyya Ar-Razi (865–925), of Rayy, northern Persia, travelled to Baghdad to study medicine. In his Manafi‘ al-Aghdhiya wa-Daf‘ Madarriha [Treatise on Foods and Correctives], he wrote that the faqqous were a bad, indigestible food when large, but that the young, tender faqqous were much better (Ar-Razi 1982). This literary use of the epithet faqqous, in the first quarter of the tenth century, is the earliest known to me.

Al-Warraq, ca. 950 CE

Ibn Sayyar al-Warraq wrote a comprehensive cookbook, Kitab al-Tabikh, describing the cuisine of the royalty and dignitaries of mid-10th-century Baghdad (Nasrallah 2007). In this book, the faqqous is mentioned as a garnish for a summer dish.

Al-Ghafiqi, ca. 1160 CE

Ahmad Al-Ghafiqi, a prominent Andalusian physician, wrote his Kitab al-Adwiya al-Mufrada (Book of Simple Drugs) around 1160. Only an abridged version of this work has survived. The original version very likely alluded to faqqous, as Meyerhof and Sobhy (1932) wrote that Ibn al-Baytar’s book of pharmacology, written almost a century later (see below), was derived almost entirely Al-Ghafiqi’s now lost book.

Maimonides, 1168 CE

Maimonides completed a major work on Jewish Law, his Commentary on the Mishna, in Egypt a decade prior to his Mishne Tora (Kraemer 2008). Although the Mishna itself was written in Hebrew, Maimonides used Judeo-Arabic (Arabic written in Hebrew characters) for his Commentary, as Arabic was the lingua franca of the Middle East and northern Africa. For Kil’ayim (Crossing) 2:1, Maimonides explained that the qishut is called (in Arabic) faqqous. For Ma‘asrot (Tithing) 1:5, he explained that the down which grows on the qishu’in, of little fine hairs, is called fiqqis. For ‘Oqazin (Fruit stems) 2:1, he explained that the down (keshut) of the qishut refers to the little fine hairs (Ar. zughb) on the faqqous (Qafah 1963, 1968).

Maimonides, ca. 1200 CE

Maimonides also wrote, in Arabic, a book on pharmacy, Sharh Asma al-‘Uqqar (Glossary of Drug Names), in Egypt at the end of the twelfth century (Meyerhof 1940; Rosner 1995; Kraemer 2008). This is a lexicon and includes only those items that have synonymous names. The manuscript known today is a copy probably written several decades later by Ibn al-Baytar, who may have also edited the original work. There is an entry for qitha, as follows: “It is as-sawaf in the language of the Arabs. Its round species is called ____ (omitted by the copyist); al-faqqus is the long species.” Evidently, the long-fruited qitha, the faqqous, was much more common as only it is given as synonymous with the qishut in the Commentary on the Mishna.

‘Abd al-Latif, ca. 1200 CE

‘Abd al-Latif al-Baghdadi (1162–1231), born in Baghdad, is thought to have conversed with Maimonides around the year 1200 (De Sacy 1810). Much of the terminology used by ‘Abd al-Latif in his Account of Egypt is the same as that of Maimonides. He described the faqqous as a small qitha, not longer than the length of a finger. This description is in contrast to that of Maimonides and modern writers (Issa Bey 1930; Hassib 1938), who defined the faqqous as the common long type of qitha.

Ibn al-Baytar, ca. 1240 CE

Ibn al-Baytar of Andalusia is best known for his large pharmacological compendium (Leclerc 1883). This work, however, is not considered to be original but almost entirely a copy of that of Al-Ghafiqi (Meyerhof 1940). The citations are exhaustive and the short text on faqqous is one of many passages credited to Ar-Razi.

Eshtori HaFarhi, 1322 CE

Eshtori HaFarhi was a scholar of Jewish Law who spent his early years in what are now France and Spain (Luncz 1897). He emigrated to Israel in 1313, stopping briefly in Egypt on the way, and settled in the northern town of Bet She’an. In his book Kaftor waFerah (Button and Flower), written in Hebrew, he identified the cucurbits appearing in tractate Kil’ayim (Crossing) with their contemporary names. Significantly, he wrote that “the qishut is al-faqqous”. HaFarhi, writing in Hebrew about Jewish Law, obviously intended his book for a Jewish audience. Nonetheless, he used the contemporary Arabic word faqqous to clarify the meaning of the Mishnaic Hebrew qishut. It does not seem possible to determine whether he was familiar with the epithet faqqous from Israel, Spain, or Egypt (the last is unlikely, as his stay in Egypt was brief), but as Arabic was the lingua franca at the time in all of these places, it appears likely that the faqqous was a familiar word over a wide geographical area. In any case, Eshtori HaFarhi assumed that his Jewish audience was familiar with this word.

Anonymous, ca. 1350 CE

An anonymous cookbook in Arabic called Kanz al-Fawa’id fi Tanwi’ al-Mawa‘id was written circa 1350, probably in Egypt (Marin and Waines 1993). The index has listings for various cucurbits. There are two allusions to faqqous and five each for qitha and ‘ajjur. From descriptions of modern writers (Issa Bey 1930; Hassib 1938), it would seem that ‘ajjur or ‘aggur most often was used in reference to short-fruited qitha, that is, adzhur (chate) melons.

Besaychi, ca. 1490 CE

Eliyyahu Besaychi, a scholar from a leading Qara’ite Jewish family initially from Adrianopoli and then Constantinople, wrote a book in Hebrew on Jewish Law that after his death became known as Adderet Eliyyahu (Anqori 1966). For tractate Kil’ayim, Besaychi discussed a number of crop plants in Hebrew alphabetical order. Under the letter Qof he wrote “Qishut is called in Arabic al-faqqun.” Although this Hebrew sentence contains a mis-spelling, it is clear that Besaychi, writing in Turkey, had made the same identification of the Mishnaic Hebrew qishut with the contemporary Arabic faqqous that Eshtori HaFarhi had made in Israel approximately 170 years before him.

Conclusions

The Hebrew word piqqus was in use from the classical period to the beginning of the medieval period, as is evident from the codices of Jewish Law that were compiled from the second to the sixth centuries in Israel and Iraq (Mekhon Mamre 2010). Subsequently, this word fell into disuse, being replaced by the word shifshuf, as indicated in a late twelfth century work of Maimonides, the Mishne Tora. The Hebrew piqqus was preserved in the Arabic faqqous, not for the process of rubbing the hairs off fruit called qishu’im (He.) or qitha (Ar.), but rather for the fruit itself.

Only a few extant Arabic works antedate the tenth century. This situation is, in large part, responsible for the apparent disconnection in time, 300 years of the Dark Ages, between the last recorded Hebrew use of piqqus and the first known recorded Arabic use of faqqous. From the early tenth century to the close of the medieval period, however, the faqqous were mentioned in documents from across the Islamic Empire, from Iraq to Andalusia. These documents were not concerned with horticulture, but instead with lexicography, religious law, pharmacology, medicine, diet, and cookery. Specific references to faqqous and other melons used when young do not appear in the late medieval Andalusian horticultural guides of Ibn Bassal (Millas and Aziman 1955) and Ibn al-‘Awwam (Clément-Mullet 1866). These guides do employ the term qitha which, although more generic for immature melons, short or long (Issa Bey 1930; Hassib 1938), from Maimonides’ Commentary it is apparent that the long-fruited faqqous was the more common.

Cucumis sativus is better adapted to temperate climates than C. melo. Cucumbers were introduced to Europe after the Roman period (Janick et al. 2007). Given their similarity to chate (adzhur) and snake melons in appearance, flavor, and texture but better adaptation to temperate Europe, cucumbers replaced these cucumber-like melons, relegating them in Europe to the status of a relict crop (Hammer et al. 1986; Laghetti et al. 2008). To the present day, though, the faqqous, the snake melons, are widely grown and enjoyed across much of the Arabic-speaking part of the world and warm regions beyond.