Milk and Honey are the only articles of diet whose sole purpose is food.
-University of California Food Sciences Department
Think about that.
A long time--
In all the world, these are the only two things uniquely intended to continue the life-cycle without interrupting any other life-cycle.
Now think about mixing them together.
Cooking with milk is prehistoric. Exodus (34:26) exhorts the Israelites not to "boil a kid in its mother's milk" hinting that milk as an ingredient was used well before the scholars codified the law. (Truly, Talmudic scholars do not fully understand this prohibition, whose discussion continues today.) In spite of the Talmudic prohibition, raw milk is a fickle ingredient for cooking. It comes fresh from the cow with all of the micro-organisms to make cheese, or yoghurt, or butter, or kefir, or be fermented into kumiss. Cooking it 'cauldron-of-hell' style on an open fire led invariably to boil-overs and a less than fresh tasting result. All of this changed when sugar was available, either as hard-to-collect honey or as cheap cane sugar from the Indes.
When this happened, the delicacy joined a new word that was just coming into being, first in Catalonia, then throughout the Spanish world:
manjar. Catalonia's richest city, Barcelona, was port city to the western Mediterranean. Food chosen for a voyage into Barcelona had to be profitable for the merchant. It would be rare and expensive. The merchants labeled new foods with their place of origin "
menjar dels Árabes" or "
menjar dels Muslmanes", using the Latin root menducar, meaning to chew (it later became mangiare=EAT!). Thus, the signs only said: "This is what the Arabs eat". In an early bow for consumer marketing, the purchasers soon came to associate
menjar with 'delicacy'. However, around this time, manjar became associated with another delicacy:
Manjar de los ángeles, the food of the angels. Its ingredients: boiled milk and honey.
The only Latin Americans who feel they are on equal (or maybe a bit superior) to angels are the Argentines, and notably they are one of the few countries that truly follow the most natural practice of making manjar de los ángeles, or dulce de leche. The recipe allows fresh, unpasteurized milk. There is no corn flour (Cuba), dark caramel (Colombia), or mixing of goat and cow milk (Mexico). The recipe is a
national standard that is available for your personal download and printing. This clever use by the Argentine government of the Internet to virtually publish useful information saves immense amounts of time in the government's printing office, thereby allowing it to physically print virtually worthless bonds.
Using fresh milk available from our favorite natural grower,
Siembra Tres Vidas, we used this modified Argentine recipe
- 100 units fresh milk
- 24 units cane sugar (modified upwards from 20 parts. See discussion.)
- 1 units invert sugar syrup (modified downwards from 4 parts.)
- 0.05 sodium bicarbonate (speed Maillard reaction)
- 0.06 vanilla
Even though I had sufficient invert syrup to meet the original recipe, it was a dark syrup. As this was the first calibration run, I wanted to observe the natural color changes as the milk evaporated. Later runs should use more invert sugar to prevent crystallization.
Comverting the ratios as follows:
- 1 liter milk (1000 gm)
- 240 gm white sugar (Snow White--still packed in Puerto Rico and the original sugar made by Serallés...)
- 0.5 gram sodium bicarbonate (Baking Soda)
- 1/2 a Dominican vanilla bean (see more of the story below)
Set the milk in a 3-liter double boiler and allow to reach its maximum temperature. This will normally not exceed 95C. When the milk reaches 95C, add the 240 grams of sugar and weigh the total pan. At this point you have 1000 gm of milk and 240 gm of sugar for a total weight of 1240 grams. Dulce de Leche is a condensation process that needs to proceed until the remaining dulce is at least 70% solids (70 on the Brix scale). If you have a honey refractometer, you do not need to weigh the pan and compute the evaporation.
Weigh the milk and heat to the steam temperature (~92C) then add sugarWhen you start, you will have
- 240 grams of sugar
- About 11.5% solids in the whole milk (sugars, fats and proteins) = 115 gm
This means about 355 grams of solids that should be about 70% of the total weight. This means a final weight of dulce de leche when done of 500 gm, or half the original milk weight.
After one hour of cooking, about 30% of the water will have evaporated. At this point, add the bicarbonate of soda. A good amount of fizzing will occur, indicating acids are being reduced. The reduction of acids allows a much more rapid browning of milk proteins, without having to raise the temperature to the 'cauldron of hell' stage.
Add baking soda (NaHCO3) and watch the fizz
Now it becomes a game of patience. Keep gently stirring the milk and sugar every 10-15 minutes. Using steam as the cooking medium eliminates the possibility of milk proteins of coagulating and creating lumps in the finished dulce.
After about 40% of the liquid has been evaporated (45 Brix), add essence of vanilla, or in this case, one-half of a natural vanilla bean from the island of Dominica. (You buy these well away from the tourist track in Saint Thomas from a nice lady who can also turn your son into a goat.) Fresh picked vanilla beans have such an aroma that the entire kitchen turns into a candy reverie.
The nice lady in Saint Thomas who sells natural vanilla beans from abandoned plantations in DominicaKeep stirring, and finally add the invert syrup when solids are reaching 65-70%. Invert syrup is composed of glucose and fructose that will disrupt crystallization of the pure cane sugar.
See here.
For this batch, I computed a target weight of pan and dulce of approximately 1240 grams. This point represented where the solids would be 75-80%. Again, there is a good deal of estimation in this. It is best to use a refractometer.
Stir frequently and keep the steam jacket filled
Look closely at the color change in the last 30 cc of evaporation. This happens in 15-20 minutes in a steam pan. Be careful! The progression of the dulce is shown below. Overall time is about 4 hours to make about 500 grams in a test batch.
| Weight in pan (gm) | Estimated % solids | Temperature C | Notes | Time |
| 1000 | 11.50% | 20 | Fresh Milk |
|
| 1240 | 28.63% | 90 | Add sugar | 11:34:00 AM |
| 901 | 40.62% | 92 | Add bicarbonate-fizzing | 12:27:00 PM |
| 793 | 46.15% | 92 |
| 12:52:00 PM |
| 687 | 53.28% | 92 | Dropped Thermometer in tank | 01:26:00 PM |
| 613 | 61.34% |
| Added invert syrup | 01:48:00 PM |
| 560 | 67.14% |
|
| 02:04:00 PM |
| 525 | 71.62% |
|
| 02:21:00 PM |
| 506 | 74.31% |
|
| 02:31:00 PM |
| 486 | 77.37% |
|
| 02:58:00 PM |
| 476 | 78.99% |
|
| 03:10:00 PM |
| 468 | 80.34% | 91
| Repaired thermometer
| 03:24:00 PM |
The finished dulce de leche while at 60C. At this temperature it is still fluid; as it cools it became nearly solid.
Excellent in an Italian espresso for an adult velvet sugar bomb. Or pour on vanilla ice cream. Or spread on bread, which reminds me...This would be a biga based bread in two sub-batches. The intent was to split the original biga, making it the 'mother' of two 2.4 kg batches. Splitting the batches permitted pacing the dough to the makeup bench and the single oven. Having only one oven means that bread near the end of a batch is over-risen and usually good for pizza.
The biga was a 55% AR dough:
- 450 gm of WW flour
- 550 gm of Amapola Harina por Pan
- 8 gm of Fleischman's IDY yeast dissolved in
- 600 gm of water at 40C with
- 10 gm of invert syrup
This was allowed to ferment overnight for 13 hours. Then the resulting ~1.55 kg biga was divided in two.
Mixing and rising The final dough was made to a 60% AR recipe in a 2.4 kg batch. Per batch:
- Final total flour: 1.5 kg (the biga already has 500 gm)
- Final total water: 900 gm (the biga already has 275 gm)
- Salt per batch: 30 gm (2% of flour weight)
To make the total weight, add the biga to an additional 625 gm of water and 30 gm of salt and mix on first speed until the biga (and its yeast) are liquefied. Then add the additional 1 kg of flour and mix for 7 minutes on first speed.
Take the finished dough, round it on the bench, and place in a 6-8 quart sealed tub for first rising.
The biga after 13 hours. Very firm yet yeasted.
Dough after first mixing, then after first rising (1-hr @ 35C--sunny)
The second batch was prepared in the same way, but one hour behind the first. This allows one to rise while the second is still in preparation.
Allow to rise 1+ hr, then knock down and allow to rise another 30 minutes before forming
Forming and BakingThe batches divided into 100 gm rolls and 400 gm small loaves.
- Divide dough and form boules. Allow to rest 20 minutes at about 27C.
- Bench boules into Kaiser knots and pistolet rolls, let rest 10 minutes.
- Bench loaves into mid-length loaves, not baguette. Let rise 30 minutes on flat.
The rolls were baked at 225C for 14 minutes. The loaves were baked for 20 minutes at 225C, then allowed to cool on the hot baking stones to improve the crust.
100gm divisions, ready for rounding
12 pistolet, 12 Kaiser knots, 6 loaves
to channel 26 (526 MHz), near the geometric mean of the channels available in San Juan
