Ancient Recipe: Savillum (Cheesecake) (Roman, 1st century BCE)

“Make a savillum thus:  Mix half a libra* of flour and two and a half librae of cheese, as is done for libum [another kind of cheesecake].  Add 1/4 libra of honey and 1 egg.  Grease an earthenware bowl with oil.  When you have mixed the ingredients well, pour into the bowl and cover the bowl with an earthenware testo [lid].  See that you cook it well in the middle, where it is highest.  When it is cooked, remove the bowl, spread with honey, sprinkle with poppy, put it back beneath the testo for a moment, and then remove.  Serve it thus with a plate and spoon.” ~ From Cato’s De Agri Cultura (“Concerning Agriculture”), 160 BCE
img_3041.jpg

Savillum is a Roman recipe found in De Agri Cultura, the earliest-known work of Roman prose. It was written by the Roman politician Cato the Elder, a man noted for his devotion to simplicity and love of country life. Fitting its author’s lifestyle, De Agri Cultura is a straightforward instructional manual on farming. Those recipes which appear are just as simple and rustic as savillum.

This is one of several Roman dishes that could be called “cheesecake”, although it lacks a crust on the bottom. I frequently choose to make it for ancient food-themed events and parties because it’s an easy Roman dish to love, as it doesn’t deviate too far from a modern Western palate. It’s amazingly simple, with a batter made from just four ingredients: honey, fresh cheese (ricotta or a farmer’s cheese), flour and egg. After baking, the savillum is topped with a spice that was as well-known to the Romans as it is to us: poppy seeds (papaver). One time I ran out of poppy seeds and used black sesame seeds, and it was just as delicious.

Savillum would have been served at the end of a Roman meal, in keeping with the Roman dining customs that we follow to this day: appetizer, main course and dessert, called gustatio (tasting), prima mensa (first plate) and secunda mensa (second plate). Like many Roman desserts, this recipe makes extensive use of honey (mel), the favorite Roman sweetener. In fact, other than fruits like dates and figs, honey was the only Roman sweetener. Sugar was first refined from sugarcane in ancient India around 350 CE, centuries after this recipe was recorded. Even then, sugar did not penetrate far into the Roman world. Its faraway origin made it too expensive for daily use, and the Roman historian Pliny the Elder writes in the 1st century that sugar is to be used “only for medicinal purposes”, as it was said to soothe stomach pain and other ailments. Presumably if sugar had been more widely available to the Romans, they would have experimented enough to learn how to cook with it.

Honey, on the other hand, was widely available because it could be produced in so many places. The islands of Malta and Sicily were main centers of Roman beekeeping, with one Maltese apiary examined by archaeologists harboring over 100 hives. Romans were well-aware of regional differences creating unique flavors and qualities of honey. The Greek city of Cecropia and the island of Corsica were infamous for their inferior honey, while the Greek cities of Hybla and Hymettus were said to produce the best. In his Epigrams (86-103 CE), the poet Martial uses the reputations of these various honeys to make a metaphor about writing: don’t expect good poetry from lousy material, just as you wouldn’t expect Hymettian honey out of a Cecropian bee.

Romans preserved food in honey, used it in sauces for meat and delicate desserts like savillum, and mixed it with water and spices to make a refreshing non-alcoholic beverage called hydromel (honey-water), although they drew the line at fermenting honey into mead, regarded as the practice of foreign enemies. Like rock sugar, honey was believed to have medicinal properties, and the physician Galen wrote that it “warms and clears wounds and ulcers in any part of the body.” Despite this seeming honey obsession, the Roman diet was still low in sugar by modern standards, and Roman burial remains show strong, healthy teeth.

There are many modern recreations of this recipe, but I use Cathy Kaufman’s reconstruction in her book Cooking in Ancient Civilizations (2006).

THE RECIPE

The savillum will puff up into a golden brown mound as it cooks, which looks pretty cool but is unfortunately ruined by poking holes for extra honey to soak in. It’s mushy, so serve with a spoon, hot or room temperature.

3 1/2 cups ricotta or farmer’s cheese, drained and densely packed

3/8 of a cup honey, plus another 3/4 of a cup

1 1/4 cup flour, whole-wheat (more authentic) or white

1 beaten egg

poppyseeds

Preheat the oven to 350 degrees. In a bowl, mix all the ingredients except the 3/4 cup honey and the poppyseeds. Pour into a deep pie dish or springform cake pan and cook for 1 hour and 40 minutes. When the cake is firm, poke some holes to allow the additional honey to seep in. Top with 3/4 cup of honey and the poppyseeds and bake for 10 minutes more.

THE VERDICT

This is one of my favorite Roman recipes I’ve tried. You could easily serve it at a modern dinner-party and none would be the wiser. X out of X.

*Sometimes called the “Roman pound”, one libra was actually only .72 of a pound, or 329 grams.

 

 

Ancient Recipe: Kamakh Rijal (Countertop Cheese) (Arab & Persian, ca. 10th century CE)

There are several varieties of this, but all follow the same recipe, only differing in ingredients. First, take a large, dry pumpkin-shell…put in 5 ratls* of sour milk, 10 ratls of fresh milk, and 1 1/2 ratls of fine-brayed salt, and stir. Cover, and leave for some days in the hot sun. This is first made in June, at the beginning of the mid-summer. Each morning add 3 ratls of fresh milk, and stir morning and evening. Add milk as the liquid lessens, until the beginning of August…cover until the beginning of October: then remove from the sun until set, and serve. ~ from Kitab-al Tabikh (The Book of Dishes), 10th century

IMG_3370Before I leave for a vacation that will likely delay my next blog post, I’m excited to share a long-term project I’ve been working on: kamakh rijal, a kind of spreadable cheese from the Medieval Islamic world that’s incredibly easy to make. All you need is milk, yogurt, salt, and the courage to eat a dairy product that’s been sitting at room temperature for more than a month. My stomach feels fine, thanks.

Originally a Persian recipe, kamakh rijal is found in many Medieval Arab cookbooks, including the earliest known Arabic-language cookbook, the 10th-century Book of Dishes. Despite its former popularity, the unique method of cheese production has been largely forgotten in modern times. The modern-day expert on kamakh rijal and rediscoverer of the technique is food historian Charles Perry. He has written about it several times, including for the premiere issue of the fermentation magazine Cured and the book Medieval Cuisine of the Islamic World by Lilia Zaouali (2009). This is the same book that gave me a recipe for a Roman-influenced chicken dish which I recently wrote about for Sarah Lohman’s blog, Four Pounds Flour.

In typical cheesemaking, milk is curdled to separate solid curds from liquid whey. Any remaining moisture is pressed out of the curds to protect the good bacteria that transforms them into cheese. In kamakh rijal, it is the salt and the pre-existing good bacteria in the yogurt that ensure fermentation and keep off bad microbes. There is no curdling or pressing involved, so the end product is a semisolid. We would call it spreadable cheese; 10th-century Persians called it “something eaten by licking“, the literal translation of kamakh rijal. I used cow’s milk, but I would be curious to try with sheep or goat’s milk, both of which would have been readily available in the Medieval Islamic world.

Below is my kamakh rijal journal, written over the course of six weeks. Following the instructions of Perry and The Book of Dishes, the cheese is edible after six weeks, but it’s actually meant to age for 14-16 weeks total. Mine isn’t going anywhere just yet, so stay tuned for a part 2.

Friday, April 28th, 2017

I bought a quart of whole milk and a container of the most ultra-probiotic, microorganism-rich yogurt I could find (Russian kefir). Sharing a Brooklyn studio apartment with another person as I do, it’s not always easy to find sufficient space for these kind of projects. Perry’s recipe in Cured says to mix 1 cup yogurt, 5 cups milk and 3 tablespoons kosher salt. My only available glass container wasn’t quite big enough, so I had to do some eye-watering math to shrink the recipe by one-fifth.

After my calculations I ended up with a milk/yogurt/salt mixture that completely filled up the container, and I was just barely able to fit the container like a puzzle piece into the back of my extremely crowded counter-top. Taking this as a good sign, I covered the container with cheesecloth. For the next week I will stir and taste the mixture daily, and top it off with more milk (Perry suggests canned evaporated milk) as it dries out. At this stage, it tastes like salty milk.

Friday, May 5th (1 week in)IMG_3141

My boyfriend has started to complain about the smell in the apartment, which he notices as soon as he walks in and I don’t notice unless I’m right next to the bowl (what does that say about me?) But it’s definitely a cheese smell and not the smell of milk gone bad; the salt in the mixture sees to that. By the end of a full week, the kamakh rijal is starting to look a little yellowish and is definitely thicker, with an even texture. When I stir with a chopstick, it comes out coated. The flavor is increasingly tangy, more like yogurt than cheese but definitely approaching cheesiness. Exciting!

Friday, May 13th (2 weeks)IMG_3227

This week I topped off the mixture with a can of evaporated milk per the recipe instructions. Evaporated milk is more yellowish than fresh milk, and the kamakh rijal is now definitely a pale whitish yellow, the color that Benjamin Moore would call “eggshell.” Although it was liquid last week, it’s now so thick a chopstick can stand upright in it (see image). It’s also started to taste like yeast. On the one hand, my boyfriend has stopped complaining about the smell. On the other hand, now I’m starting to notice it.

Friday, May 19th (3 weeks)IMG_3245

Up close, the kamakh rijal smells kind of like feet. But hey, it’s cheese, that’s a good thing, right? A thin layer spread on a cracker with a slice of tomato is quite good. It is even thicker than last week; I added a new can of evaporated milk to top it off but I noticed that the milk does not mix evenly into the kamakh rijal as readily as it did before, just sits on top. Kind of looks like curds floating in the milk now.

Monday, May 29th (4 weeks + 3 days)IMG_3314

Couldn’t check on the kamakh rijal until Monday. It’s gone solid again, and while the texture is still a bit lumpy, it spreads nicely on a piece of bread, like soft butter. The taste appears to have neutralized. It’s still tangy but not yeasty anymore, and for the first time I’m reminded of cheddar cheese. My boyfriend surprised me by asking to taste it (direct quote: “you haven’t gotten the shits yet, so it’s probably fine, right?”) Then he surprised me further by saying he actually liked it. Like the taste, the smell appears to have peaked a week or two ago and is now mellowing out.

Added some more evaporated milk. The outer edges of the glass container have developed a pretty gnarly yellow crust. Blegh.

Tuesday, June 6th (5 weeks + 4 days)IMG_3336

I’ve been a little lax in adding fresh milk, and the kamakh rijal has undergone yet more changes. It’s so thick now that if you stir it or stick a knife into it, it will not regain its original shape. The outer surface is drying out and starting to look like a crusty lunar landscape, if the moon really was made out of cheese. I try to be careful about the dry parts when tasting, but the occasional little crumb gets mixed into the smooth majority of the cheese. Gross. If I were serving it to another person, I would scrape off the entire top quarter-inch where it’s dried out to avoid the gritty texture.

Friday, June 9th (6 weeks)IMG_3367

By this point I couldn’t stand looking at the yellow crust around the edge of the container anymore, so I broke it off. That plus the addition of more milk revived the mixture. It’s smoother and softer again now, and overall looks much more…..edible. I was expecting the taste to get stronger and stronger over time, but I’m fascinated by how it peaked and then started to get weaker. I wonder if that represents one type of microbe taking over another (bacteria vs. fungus?)

The kamakh rijal having reached its minimum age for cheesiness, I have a few different options. The recipe quoted above suggests adding mint, garlic and the anise-like spice nigella before allowing it to age further, which sounds like a delicious combination. Another recipe calls for dried rose petals, which intrigues me; I don’t like rose-flavored desserts, but I’ve never tried a savory preparation. I plan to let it age another six weeks before dividing it up into parts and trying both the rose and nigella/garlic/mint versions.

IMG_3340When I was visiting Paris, I ate a meal (okay, several meals) that consisted entirely of wine, bread and cheese. With this in mind, I spread some of my six-week-old kamakh rijal on bread and poured myself a glass of red wine. Not only did it go together well, it felt like a scene from a medieval Persian poem. “A loaf of bread, a jug of wine, and Thou“; I always thought Omar Khayyam was talking to a lover, but maybe he was just addressing the spreadable cheese. VIII out of X.

* One 10th-century ratl equals 406.25 grams, or .90 of a pound.