Culinary History Mystery #6: Tomato Eggs

Chinese Eggs with Tomatoes
Chinese Tomato Eggs

Tomato Eggs is a home-cooked Chinese dish that reminds students, travelers, and those living abroad of home.  Just a whiff of this cooking and folks will tell tales of sitting in or near the kitchen as a kid as a parent made this dish – and how good it tasted!  it is simple, elegant, and savory, and less than 10 – 15 minutes from wok to table.  Chopped green onions are almost always used. Sometimes garlic or onion is added, and often there is a blast of shaoxing, rice vinegar, or even oyster sauce to add flavor.  Some recipes also add sugar to counter the acidity of the tomatoes, but the memorable taste of the dish usually just comes from the combined flavors of the fresh ingredients.

The form of the dish can be dry, like in the picture above, or is can be moist with a thin tomato sauce, or even soupy. It is often served over or with rice or fresh noodles. My travels tend to make me think that presentation varies mostly by individual preference and not by geography, because I have had both dry and wet forms in a number of different places.

I’ve enjoyed this dish all over China, from Beijing and Xian to places much further west and south.  Although the Chinese regard this as a quintessential Chinese dish, my favorite thing about tomato eggs (蕃茄炒蛋/西红柿炒蛋) is that it is probably Arab in origin.

Saudi Tomato Eggs: Shakshouka
Saudi Tomato Eggs: Shakshuka

The Arab dish that Tomato Eggs most resembles is Shakshuka. This dish is eaten all over the Saudi Peninsula, North Africa, and the Levant. Turkey even has its own version called Menemen. Although the form varies a great deal, from the dry, Saudi version pictured here, to poached eggs over a spiced tomato sauce as in Egypt and Israel, to a complex ragout of vegetables (with lots of tomatoes) and sometimes bits of meat or sausage bound together by eggs.   It is almost always served with pita bread or naan.  Onions are almost always used and sometimes garlic is as well.  Spicing can be just salt and pepper with a little bit of chopped parsley or cilantro as in Oman to a dish flavored with cumin, or dishes with oregano and other herbs.  Chili peppers or ground chilies are often added, but I have never had a Shakshuka that I could call hot.  These days, cheese is sometimes added, but that is a modern addition and not found in traditional recipes for the dish in any of the cultures that now enjoy it.

A comparison of the Saudi and Chinese recipes show that the recipes are nearly identical, although the Chinese use a two-step cooking process:

2 tablespoons olive oil 3 tablespoons cooking oil
2 medium tomatoes, diced 4 large eggs
2 garlic cloves, minced 1/2 teaspoon salt
1 small white onion, minced 1 teaspoon shaoxing wine (optional)
½ teaspoon salt 3 dashes white pepper powder
½ teaspoon black pepper 8 oz. fresh tomato (cut into thin wedges)
1 teaspoon cumin 1 teaspoon sugar
4 eggs 2 tablespoons water
Some chopped green onions
1/4 teaspoon sesame oil (garnish, optional)
Break eggs over mixture and cook for another 3-5 minutes or until done. Stir with a spatula to mix or slide onto a plate and serve. Heat up a wok with 2 tablespoons cooking oil. Add the egg mixture into the wok, and use your spatula to spread the eggs. Keep stirring until the eggs form lumps. Gently break the lumps into smaller pieces. As soon as the eggs are cooked, dish out and set aside. Clean the wok and heat it up again with 1 tablespoon cooking oil. Drop the tomato wedges into the wok and do a few quick stirs. Add sugar and water into the tomatoes. Cover it with the lid and let it cook for about 30 seconds. Transfer the eggs and chopped scallions into the tomatoes, stir-fry for 30 seconds or so, dish out and serve immediately.

The main reasons why this is probably another west-to-east spread of a recipe is the commonality and variations of the dish in the Muslim Mediterranean, Suez and Persian Gulf,and the unusual nature of the dish in China’s litany of egg recipes. Another reason why it is likely a dish with an “Arab” origin is that Muslim people took to the tomato very early on in its introduction in the Old World. While the Europeans were generally skittish about eating this member of the nightshade family, and raised them as curiosities or ornamental garden plants, the Muslims dove right in and cultivated them as food early on in their arrival in the Eastern Hemisphere. The Spanish were the only Europeans who generally took to eating the tomato in the 16th Century. This is probably because they saw them being cultivated and eaten in the New World and knew that they were not harmful.

Apples of Love (Tomatoes), Gerarde, 1597
Apples of Love (Tomatoes), Gerarde, 1597

The Spanish were said to particularly enjoy them with cooked with oil, salt, and pepper as a sort of stew, and also to make a sauce out of them with vinegar added to the ingredients above and to use that sauce on their meats (Gerarde, History of Plants, 1597). Gerarde also notes that tomatoes or “Love Apples” grow well in warm climates like Spain and italy. Of interest, perhaps, is that Gerarde describes both red and yellow tomatoes.

Although there is mention that the Italians also ate tomatoes in 17th Century botanicals, this is repetition of incorrect information. The original citation says that the Italians ate, “Eggplants”. This, even in historical documents became misreported as, “tomatoes”, and the error continues to proliferate today.

Evidence for the early Arab love of eating tomatoes can be found in John Parkinson’s 1629 Earthly Paradise, in which he reports that tomato plants grow well in hot climates like those in, “Barbary and Ethiopia”. Parkinson’s 1640 Theatrum Botanicum expands this range of growth in the Old World to, “easterly countries such as Egypt, Syria and Arabia.” His 1629 work notes that tomatoes are much eaten in the hot countries where they grow well.

Lancelot Addison’s 1671 work, An Account of West Barbary, notes that tomatoes are eaten raw with oil along with other, “salads.” In 1710, Dr. William Salmon’s Herbal notes that the Spanish ate tomatoes boiled in vinegar with pepper and salt, and served up with oil and lemon juice (possibly a poached tomato); and that they also eat tomatoes raw with oil, vinegar, and pepper.

Red Tomato
Red Tomato

By comparison, the earliest European mention of tomatoes growing in Asia (Malaysia) can be found in Georg Rumphius’s 1747 work Herbarium Amboinense. Rumphius notes that the natives cultivate two varieties and that both are used in cooking. In 1790, a brief mention of tomatoes growing well in the fields and gardens of Cochin, China is found in Louriero’s Flora Cochinchinensis.

So, from all this information, we can infer that the Arabs were eating tomatoes in the 16th century – at least the Morisco’s in Spain were – and possibly so were people across a broad swath of the Muslim World from North Africa to the Arabian Peninsula and into the Levant. More evidence for Tomato Eggs having Arab roots. Tomatoes may have been eaten in Ming China, but the earliest evidence I can find in a language I can read comes from well into the period of Qing rule. That said, however, we know that the Chinese were trading with the pre-Islamic Arabs and that trade between the peoples only flourished after the adoption of Islam, with the influence of foreign Muslim peoples in China reaching its peak probably in the Yuan Dynasty.

What I love most about Culinary History Mysteries like that is that hundreds of years later, the history of the interaction between the Chinese and the “Arabs” lives on in the foods people eat. Another enduring testament to the power of the Silk Road in the lives of the people.

(Words and recipe analysis by Laura Kelley. Photo of Chinese Eggs with Tomatoes by; Photo of Saudi Tomato Eggs: Shakshuka by Noor AlQahtani. Recipe for Saudi Shakshuka from Noor’s site, Ya Salam; Recipe for Chinese Tomato Eggs from Rasa Malaysia)

The Origins of Curry Powder

Where did curry powder come from? There is no real equivalent in authentic subcontinental cuisines for a ready-made powder. The closest thing to a curry powder is a masala, and that is almost always more of a paste than a powder because of the addition of wet and dry ingredients to the mix. On the subcontinent, seeds and roots, etc. are roasted, ground and mixed in varying proportions according to the needs of the recipe. Although the origins of curry powder are unclear, the advertisement below gives us a firm data point of the mid 1780s for a commercial curry powder for sale in London.

First British Curry Advertisement
First British Curry Advertisement

The advertisement, which ran in the Morning Post (now incorporated into the Daily Telegraph) says that this curry powder was brought back from the East Indies by Solander. Now, Solander was the great Swedish naturalist who was botanist on Captain Cook’s Endeavour expedition to the Pacific. Despite the claim, this is probably just a marketing ploy – like Mrs. Pepperidge – because the closest the Endeavour ever got to India was actually Indonesia (Batavia/Jakarta) and it returned to Britain in 1771, some 13 years before the advertisement. Solander, on the other hand, did meet an untimely death in 1784, and was something of a celebrated figure at the time. So, it was good business sense by the maker of the curry powder to use Solander’s name to conjure images of exploration and the exotic cuisines of the east.

It isn’t completely clear which company manufactured this powder, but I have one data point that indicates that it was Crosse and Blackwell – S&B – still makers of chutneys, relishes, and sauces. The problem with this is that they weren’t incorporated until 1830 when the men behind the initials S&B bought the business from its proprietors West and Wyatt. West and Wyatt, on the other opened its doors for business in 1706, so it indeed could have been their curry powder for sale at Sorlie’s Perfumery Warehouse in 1784.

The advertisement claims that the curry powder will help you make sumptuous sauces for East-Indian dishes. It also says that the curry powder promotes good digestion, good circulation, a vigorous mind and . . . wait for it . . . a strong libido. Who doesn’t want more of all that? How could anyone resist?

Dean Mahomet
Sake Dean Mahomet

However, because our early data point shows commercial curry powder for sale in 18th Century England, it doesn’t necessarily mean that it is an English invention. People from the subcontinent were already immigrating in the 17th Century, with the earliest baptism of an Indian-born Asian man in 1616, and by the 18th Century, Indian sailors were commonplace on East India Company ships, hired to replace men who had died on the voyage east. The passage for the Indian sailors was often one-way, from east to west, with the sailors attempting to start a life in a less-than-welcoming England. Usually, however, they wound up in transient, low-wage jobs or living by the good will of others. The cooks on the ships who fed these sailors, sometimes fared better than the sailors themselves, and wound up as tavern and pub cooks, slinging British food as well as the occasional curry to the hungry English populous.

In 1773 the Norris Street Coffee House in Haymarket started selling a prepared curry, and by 1810, Sake Dean Mahomet opened the first Indian-owned and operated Indian restaurant in Britain with the Hindustan Coffee House at 34 George Street, Portman Square. In Mahomet’s restaurant, British patrons could enjoy hookahs with ‘real Chilim tobacco’ as well as a wide selection of curries.

Fortunately, perhaps, or not, this article has no firm conclusions to offer about the definitive origins of curry powder, but it does place some good data on the table. Despite my wanting to keep the door open to the contributions of Anglo-Asians in the formulation of curry powders, my instinct tells me that commercial, prepared curry powder is probably not their contribution to world cuisine. If it were an Asian or an Anglo-Asian invention, I would think that the taste of curries made with curry powder would be a lot more authentically “Indian”. Still, I’ll keep digging to see there are further strands to pull, so stay tuned. (Words by Laura Kelley, Newspaper clipping of first British Curry Powder Ad from the British Library and Portrait of Sake Dean Mohamet from the Wellcome Archive.)

Curry Through Foreign Eyes #4: Dr. Kitchiner

Today’s exploration of Indian Curry through Foreign Eyes takes us back to early 19th Century England to The Cook’s Oracle by Dr. Kitchiner, which was first published in London by Samuel Bagster in 1817. The original title of the book is Apicius Redivivus, or Apicius Reborn, so it is clear that the publisher thought that this book was a masterpiece of gourmet dining. Either that, or he simply wanted to cash in on the image of Apicius’s legendary dining habits in the sales of Dr. Kitchiner’s book.

The Cook's Oracle, 1831 (American)
The Cook’s Oracle, 1830 (American)

The Kitchiner recipe for curry powder is an important one, and is cited as the basis of many recipes since then, including Mrs. Beeton’s and the curry powder used when the British introduced “Indian” curry to the Japanese in the late 19th Century.

To begin, the 1817 edition of The Cook’s Oracle has two recipes for curry powder (Nos. 454 and 455). These change and combine a great deal across editions of the book, with recipe No. 455 (with some variation) becoming the recipe that endures in later editions, including the American editions. In the 1817 edition of the book, Kitchiner observes that these recipes were given to him by a friend and he cannot vouch for their flavor or authenticity (imagine writing THAT in a cookbook today)! However in later editions of the book, he swears to the authenticity of recipe No. 455 for “Cheap Curry Powder”. So I chose to work with this recipe both for its terrific name as well as for its lasting quality.

In working with the Kitchiner recipes (No. 455 from both the 1817 and 1830 editions), I also think I have figured out why so many early curries and so many modern commercial curry powders have much more turmeric than any modern or historical Indian curry out there. The answer is simple: The confusion of grated, fresh turmeric root with dried and ground turmeric powder.

I have never seen an authentic Indian curry with more than a fraction of turmeric relative to the amounts of coriander and cumin. For example, if the recipe calls for 2-3 teaspoons of ground cumin and/or coriander, it will usually only call for about ¼-to- ½ -teaspoon of turmeric. Most Indian recipes use turmeric judiciously, almost in the way a bit of saffron is used to take the sharp edges off of the flavor of the other spices. On the other hand, try to find a mainstream, commercial curry powder that isn’t bright yellow or orange from the amount to turmeric in the mix. I have long wondered about this, and now think that adhearance to “traditional” historical recipes may be the reason for this.

Turmeric, Two Forms
Turmeric, Two Forms

To try to prove this hypothesis, I cooked the Kitchiner curries with three ounces of fresh, grated turmeric root and found them to taste much more like and Indian curry than curries cooked with ground turmeric. This is not simply the difference between fresh and dried spice – a difference we all are aware of – but also of the relative proportion of the wet, grated root to the baked and dried powder in the recipe as a whole. An ounce of fresh root is much less turmeric than an ounce of ground turmeric, and the resulting flavor of the curry is radically different. It’s fascinating to me how a likely mistake in the 18th and 19th Centuries can still resonate today. Try it sometime with a favorite historical recipe and see if you agree about the turmeric issue. On to the recipes.

The 1817 Recipe for “Cheap Curry Powder” calls for four ounces of coriander seed, three ounces of turmeric, one ounce each of black pepper, ginger, and lesser cardamoms, and one-quarter ounce of cinnamon and cayenne. This recipe becomes a little gentler as time goes on, with later editions calling for three ounces of coriander seed and turmeric, one ounce of black pepper, mustard (an addition) and ginger, and half an ounce of lesser cardamoms, and a quarter ounce of cumin seed. Later American editions call for the addition of a half-ounce of allspice as well. Dr. Kitchiner observes in the later editions that the omission of the cayenne pepper from the recipe is to allow for cooks to add more curry powder according to taste without making the dish too hot. Written in modern form the recipes looks like this:

Dr. Kitchiner’s Curry Powder No. 455 (1817)
8 tablespoons coriander seed
6 tablespoons turmeric
2 tablespoons black peppercorns
2 tablespoons ginger
2 tablespoons green cardamom seeds
1.5 teaspoons ground cinnamon
1.5 teaspoons cayenne pepper


Dr. Kitchiner’s Curry Powder No. 455 (1830)
6 tablespoons coriander seed
6 tablespoons turmeric
2 tablespoons black peppercorns
2 tablespoons mustard (an addition)
2 tablespoons ginger
1 tablespoon green cardamom
1 tablespoon allspice
1.5 teaspoons cumin seed.

The direction is to place all ingredients in a cool oven overnight, then to grind in a granite mortar and pass through a silk sieve. The sieving makes this a fine powder as opposed to a coarser, rustic grind.

Another reason for working with recipe No. 455 is that there is no specific recipe for a curry in the 1817 version of Dr. Kitchiner. Rather he suggests making curry sauces by adding curry powder a bit at a time to gravy or butter until a sauce pleasing to taste unfolds. There are recipes for deviled eggs, a bare-bones mulligitawny and a couple of curry-flavored forcemeats as well a a calf’s-head broth, but no meat stewed in liquid as the British had come to interpret as curry. I had to turn to a later edition if I wanted the Kitchiner curry recipe, and used the recipe from the 1830 edition instead.

Here is the original recipe for curries in the 1830 edition of Dr. Kitchiner’s The Cooks Oracle:

Curries (No. 497)
Cut fowls or rabbits into joints, and wash them clean: put two ounces of butter into a stew-pan; when it is melted, put in the meat, and two middling-sized onions sliced, let them be over a smart fire till they are of a light brown, then put in half a pint of broth; let it simmer twenty minutes.

Put in a basin one or two table-spoonfuls of curry powder (No. 455), a tea-spoonful of flour, and a tea-spoonful of salt; mix it smooth with a little cold water, put it into the stew-pan, and shake it well about till it boils: let it simmer twenty minutes longer; then take out the meat, and rub the sauce through a tamis or sieve: add to it two table spoonfuls of cream or milk; give it a boil up; then pour it into a dish, lay the meat over it: send up the rice in a separate dish.

Written in a more modern form, the ingredients looks like this:

Dr. Kitchiner’s Curries (1830)
1 – 1.5 pounds boneless fowl or rabbit (more if using meat on the bone)
4 tablespoons unsalted butter
2 large yellow onions
1 cup chicken broth
2 tablespoons curry powder (No. 455)
1 teaspoon flour
1 teaspoon salt
water to make a thin paste of the above three ingredients
2 tablespoons of whole milk or cream

The method from the original recipe is fairly straightforward. I made a couple of changes, searing the meat and removing it from the pan before adding the onions to the remaining butter, I added a bit more curry powder than called for, didn’t really boil the curry after adding the dairy, and I didn’t sieve the sauce before serving.

Dr. Kitchiner's Chicken Curry, 1830
Dr. Kitchiner’s Chicken Curry, 1830

Note that the “cowboy roux” or “white wash” used at the end is a mix of flour, water, curry powder and salt and is used to thicken the sauce before finishing it with a bit of whole milk or cream. Because the Kitchiner recipe is so influential in the development of other western recipes for curry, I suspect that this recipe is probably where East Asian curries adopted their “curry roux” from, because the British introduced their version of Indian curry to Japan in the late 19th Century. More about that in future posts.

So what do these curries taste like? To me, the Kitchiner curry using the 1830 curry powder tastes like a more robust version of the Hannah Glasse curry (1774) which used only turmeric, ginger and black pepper (with a little lemon juice) for spice. It’s good, but it’s very turmeric heavy and almost completely lacks any cumin flavor, which is understandable given the proportionally miniscule amount in the curry powder. It also has none of the nutmeg and mace that Mary Randolph wrote about in 1824. The 1817 version of the powder that has the extra 2 tablespoons of coriander seeds, the two tablespoons of green cardamom seeds, and 1.5 teaspoons each of cinnamon and cayenne has a nice kick to it that is lacking in the 1830 curry powder. The overwhelming flavor of turmeric is less overwhelming in the earlier version. Its a pity that this earlier version of the curry powder didn’t endure.

Both recipes also taste more authentically “Indian” with the use of three ounces fresh turmeric instead of three ounces of dried powder. (Words and historical recipe development by Laura Kelley; Photo of The Cook’s Oracle from Gunsight Antiques; Photo of Turmeric, Two Forms from Wikipedia and merged by Laura Kelley; Photo of Dr. Kitchiner’s Chicken Curry by Joseph

Culinary History Mystery # 5: A Loaf of Leavened Mesopotamian Bread

Something wonderful and unexpected happened yesterday. After a long day of tromping around historical archaeology sites in St. Mary’s City with the family, I arrived home to find a long-expected, but immediately unanticipated e-mail from a fellow food lover in England. Cid is a purveyor of fine foods and an expert breadmaker. Some time ago, I asked her to help me solve a historical food puzzle that has been vexing me for some time. Namely, did the Mesopotamians enjoy leavened as well as unleavened bread?

Because we are lacking explicit evidence for the use of yeast to leaven bread in Mesopotamia and many other ancient cultures, modern cooks reconstructing the cuisines of these cultures have assumed that all of the bread in these cultures was flat and dense like hardtack. An unfortunate assumption.

What Cid has beautifully demonstrated is that spelt, which is not too dissimilar to the emmer wheat used by Mesopotamians, makes a great big loaf of sourdough bread using a starter based only on wild yeast from the environment.

For those of you new to the “starter” concept, it is simply that a grain providing a carbohydrate source is mixed with water and allowed to attract microorganisms from the environment.  As the microorganisms consume the food given them by the flour (carbohydrate and sugar) they reproduce and must be “fed” with the addition of new flour (and sometimes water).  Sometimes an additional sugar source is added by soaking macerated fruit in the water to be mixed with the flour to start the starter culture more rapidly.  This process continues until a stable community of yeast and bacteria is established in the liquid or semi-solid “starter”.  This starter is added to flour, water and other ingredients, kneaded, folded, proofed and baked and sourdough bread results.

The importance of Cid’s demonstration is that even if the ancients didin’t explicitly know and write about yeast as an ingredient, they might have known how to use the wild type organism.  Cid’s talents as a baker show that wild-yeast leavened breads were possible in the ancient world.  She writes:

“For the past few weeks I’ve been feeding a spelt ferment starter with organic spelt flour every day. Next to my other ‘white’ ferments the spelt had a different smell and didn’t bubble up as quickly.

Wild Yeast Spelt Starter

It’s not clear whether the ancient bakers used a ‘poolish’ method for their bread, which is basically an amount of starter ferment mixed with water and flour left to further ferment over night before adding more flour the next day. It seemed like a reasonable idea so that’s what I did. It bubbled well enough to make me believe it would produce a good loaf. So at breakfast the following day I mixed in enough spelt flour and two teaspoons of salt and about a tablespoon of olive oil.

The feel of the dough was very different to the normal sourdoughs I make on a weekly basis. Despite kneading and resting, the dough never felt elastic and rose only slightly. You see, the gluten in modern wheat flour produces stretchy dough that rises well and if ‘folded’ at regular intervals rather than kneaded, will give the crumb its familiar large holes and crackling crust.

Unbaked Spelt Sourdough

As I got the dough properly formed up into a loaf, I fired up my oven to 240°C with baking stone in place.

As you can see the baked loaf is rather flat and dark looking. It’s consistency is much more like a scone than modern day brown bread and it has a sour tang. The texture is too heavy for my taste and the ferment too sour. A portion of this type of bread would be very filling and full of natural fibre.

A lighter texture could possibly have been achieved by mixing the spelt with other grains known to Mesopotamians, such as rye, oats or the numerous wild grasses they incorporated into their diets. Sieving the milled grain would have given a whiter, presumably lighter weight end result as well.”

Wild Yeast Spelt Loaf

Wonderful work, Cid! Your demonstration is not only proof of concept that these ancients could have enjoyed leavened wild-yeast sourdough breads, but it is also significant for understanding ancient beermaking and winemaking as well. Many of the people and companies who have tried to reconstruct these recipes have been left wondering how to ferment the grain and malt mixtures that have been described on the ancient tablets. Your wild yeast spelt starter gives them an excellent way to introduce yeast into the alcohol mixture. The ancients may not have known what yeast was, but I’m betting they knew how to cultivate and use it for bread and alcohol production. (To beer makers out there – lets talk about that aromatic “wort” a bit – I may have some ideas on that score as well.)

Another thing that is important to the flavor of bread, beer and wine is that wild yeast starters are complex cultures of local yeasts not the uniform commercial cultures of Sacccharomyces cerviseae one can buy in the market. Additionally, these starters all have complex communities of local bacteria in them. The difference between species and community diversity in commercial versus wild starters affects proofing time, texture and flavor of the products made.

If any of you start experimenting with spelt (or farro which is emmer wheat used by the Mesopotamians), I suggest you consider flavoring your loaves with spices to balance the strong flavor of the spelt or farro. Some spices that are authentic to Mesopotamian bread can be found in the Ninda-gal recipe (JCS Vol. 29, No. 3) are onion seeds, sumac and saffron. You could also troll some of the Mesopotamian recipes on Silk Road Gourmet for some other spice combinations as well. (Words by Laura Kelley and Cid, and Photographs of Wild Yeast Spelt Starter, Unbaked Sourdough and Leavened Spelt Loaf by Cid).