There are no set-and-forget dials and buttons on this state-of-the-art oven advertised in a 1916 Columbus, North Dakota, community cookbook.

I had an interesting sense of déjà vu watching Glen Powell of Glen and Friends use a 1914/1915 cake recipe recently. In my post about quiet ovens I wrote

The ability to trust a steady temperature in an oven isn’t just an improvement. It’s a paradigm shift. Their oven preparation process, and the terminology they used to describe it, was built around a process that simply doesn’t exist in modern kitchens.

Glen had trouble with a pre-self-regulating oven cake recipe in part because he treated his oven as a set-and-forget appliance—which, being a modern oven, it was—rather than a tool in a process, as the recipe assumed.

We have chained time in our kitchens. We have halted the decay of highly perishable foods and we have lashed baking to a schedule. We don’t have to worry about hot ovens and slack ovens and quick ovens and quiet ovens. We can literally set a temperature and a time and go do something else. And we increasingly forget that time once ruled us instead of us ruling time.

We can barely imagine the technology that inspired those old oven terms. So we make little charts corresponding oven terms to temperatures and shove our cakes into the darkness hoping for a correspondence.

Used as a rule of thumb, these charts make it possible to experiment with recipes from before modern kitchens. But when we forget that they are a rule of thumb, they instead make it more difficult to make recipes from books before the modern home refrigerator and oven. Glen’s burnt cake was a matter of running into this problem from multiple directions.

Glen recognizes that all those oven descriptor to temperature charts are inaccurate. In his 160-year-old Victorian pudding video, he said:

If you go to certain websites, if you go to Wikipedia, there’ll be a chart that tells you what a quick oven is, and that’s kind of the 2026 interpretation. In reality… this book doesn’t give a chart. A lot of them do, and when you look at the charts in the books from the time period, in that transitionary time period between ovens that had no temperature control to ovens that did have temperature control, the charts are often at odds with each other of what is a quick oven, what’s a fast oven, a hot oven, a slow oven, a medium oven.

There are, however, times he doesn’t seem to recognize why they’re inaccurate, that they reflect a process that is entirely alien to us today. Because of that, in this vintage cake-making attempt he assumed that setting the temperature to a “low end” from the charts¹ and then coming in a little early to check would be all he needed to do.

The thing is, the reason those charts are inaccurate because they are an attempt to describe a thing that is not a temperature as if it were a temperature. This isn’t necessarily obvious in the case of terms like a “hot” oven or a “cold” oven or even a “moderate” oven. But what temperature does a “quick” oven describe? What temperature does a “quiet” or a “slack” oven describe? These are not descriptions of a steady state. They are descriptions of movement toward an end.

Modern ovens use a steady supply of power combined with a thermostat to maintain a steady temperature. Old ovens did not have a steady supply of power. Nor did they have a thermostat. They were steady neither in fuel nor in fire. This terminology never described a temperature. It described a process. At best, it was a measurable starting point in a process, but it was still a process, not a steady state. Our modern ovens are not just incredibly advanced compared to the ovens used when those terms were coined, they are incredibly different.

In many ways our ovens are more understandable to someone from the nineteenth century than their ovens are to us. In my take on what is a quiet oven, I quoted the 1876 Centennial Buckeye Cook Book:

Many test their ovens in this way: if the hand can be held in from twenty to thirty-five seconds (or while counting twenty or thirty-five) it is a quick oven, from thirty-five to forty-five seconds is “moderate,” and from forty-five to sixty seconds is “slow;” thirty-five seconds is a good oven for large fruit cakes. All systematic housekeepers will hail the day when some enterprising Yankee or Buckeye girl shall invent a stove or range with a thermometer attached to the oven, so that the heat may be regulated accurately and intelligently.

That invention wouldn’t, apparently, come into general production until the twentieth century. Information is difficult to find, but it appears that, for example, the American Stove Company marketed their first self-regulating oven in 1915. This appears to be about the time that other companies introduced self-regulating ovens as well. It would take a good decade for them to become common in home kitchens.

“Milk and cream delivered daily” from the farm, in the same cookbook as the stove ad; there are no refrigerator ads in this 1916 cookbook.

The earliest self-regulating oven manuals represented on Michigan State’s Little Cookbooks site are from 1925 and 1926. In my own cookbooks, it’s at 1925 that I start to see parenthetical temperatures along with the terms (though that particular cookbook has no “hot oven”, only “moderately hot oven”).

The F.W. McNess’ Cook Book of 1935 has a very interesting outlier: one of the two “hot oven” mentions puts it at 500°. That’s for a pumpkin pie and it is at that temperature only for ten minutes. After that, the pie should be finished in a slow oven of unspecified temperature.² Similarly, the Imperial Sugar Company’s Romantic Recipes has a pie to be baked in a hot oven at 450°, but again the heat is reduced to 400° after ten minutes.

This highlights one of the big differences between modern ovens and the older ovens these charts refer to. Until the self-regulating oven, it was up to the baker to regulate the oven. The baker had to adjust the flow of fuel, perhaps even turning it off. In fact, reading that quote from the Centennial Buckeye Cook Book again for this post, I noticed something I didn’t notice before: the author wasn’t even asking for a thermostat. All they were asking for was a thermometer so that the home baker, not the machine, could “accurately and intelligently” regulate their oven. Baking would still require constant attention or a slow cooling.

The Battersea Household Cookery Recipes cookbook was first published in 1914, in London; Glen’s copy is the second printing from 1915. If my research is right³, that would have been just before, or contemporaneous with, the first self-regulating ovens entering the home kitchen market. And well before self-regulating ovens became common in home kitchens.

The recipe Glen was trying to follow called for a hot oven and a baking time of one and a half to two hours. He set the oven to 425° and left the room. When he returned to check on the cake after only 45 minutes, it was burned black top and bottom. This was not the minor variation in oven temperatures we might expect today. The cake was a charred brick despite only being in the oven for half the minimum time specified in the recipe.

Later in the video, Glen makes the supposition that the recipe was made for someone who stayed in the kitchen, or at least near the kitchen. They would then, he suggests, have been able to smell that the cake was done in thirty minutes instead of an hour and thirty minutes. This is almost certainly true. It does not explain the estimated time being so far off.

The wider, deeper truth is that the recipe was made for someone who stayed in the kitchen because they had to. Their technology was not something that could be left alone. Even ignoring the danger of fire spreading from the oven to the house, their ovens needed constant monitoring because using their ovens was a process. Their ovens did not feature the set-and-forget buttons and dials of our modern ovens.

If you think of your oven as a process you won’t walk away. It would be like a potter walking away from the potter’s wheel. Bakers had to constantly adjust the oven’s fuel, unless they were allowing the oven to cool. And even when letting an oven cool, the baker remained to add different things to the oven at different times in the cooling process.

It’s entirely possible that the hour-and-a-half baking time is because the baker was expected to bring the oven to hot and then let it cool. As noted above, there’s a remnant of this in our modern recipes for pies; we often put a pie in at a temperature above 400°, and then quickly turn the temperature down to a temperature below 400°.⁴

Could that be what was going on here? The oven starts hot, and then over the next hour or two cools off? I don’t know. We really have lost a lot of understanding about how people baked before modern appliances. They didn’t have appliances, things designed to handle a task automatically. They had tools, things designed to help a human perform that task.

A refrigerator before 1926, if you even had one, was basically an icebox with an air-conditioner attached.

There’s a very similar recipe to Glen’s 1914/1915 Battersea recipe in May Byron’s 1914/1915 Pot-luck; or The British home cookery book:

740. AUNT-MARY CAKE (Ireland)

One and a half pounds of flour, half a pound of butter, one and a half pounds of sultanas, one and a half pounds of currants, one dessertspoonful of bread-soda (bi-carbonate ?), one pint of sour milk. Rub butter thoroughly into flour and mix all the dry ingredients, putting the bread-soda in last and mixing it well in. Last of all, put the milk. Mix all well, and bake in a greased tin for two and a half to three hours, in a hot oven to start with. When the cake has risen, the heat should be less.

I found this recipe, incidentally, by going over all of the old cookbooks I used when writing Quiet Ovens. I searched for the phrase “hot oven”. This was neither difficult nor time consuming because “hot oven” does not appear to have been a common term. Some of the hits I found weren’t even a term. The only mention of “hot oven” in Mrs. Black’s Household Cookery, for example, is that “solid cakes require a rather hot oven”. Even for the standards of the time this is less a specific description than just a recommendation that the oven, you know, should be hot.

The Aunt-Mary Cake is a different recipe than the Battersea recipe. It lacks eggs, for example, which in my very limited experience can make a big difference in cooking time. It uses nearly twice as much dried fruit. But it also shares several features with the Battersea “Housekeeper’s Cake”. It mixes the butter with the flour instead of with the sugar, what Glen calls “reverse creaming”. It calls for baking soda with no cream of tartar. And it calls for a hot oven with a multi-hour baking time. It specifies even more time than Glen’s recipe, but then, Glen’s Battersea Housekeeper’s Cake is about a third the size of the Aunt-Mary Cake: a half pound of flour instead of a pound and a half of flour, and a half pound of dried fruit instead of three pounds of dried fruit.

The really interesting parts are where the older instructions are expanded. The Aunt-Mary Cake hot oven is “a hot oven to start with”. The temperature is explicitly reduced later—and in a weirdly passive way, merely saying that “the heat should be less”.

There’s also a difference in how the milk is described, which leads to the other difference between 1915 and now.

There’s a point in Glen’s video where he ponders what the baking soda is for. After all, there’s no acid in the recipe. There is, however, milk in the recipe. Someone pointed this out in the YouTube comments, and his response was:

While yes there are acids in milk—milk will not activate/react with baking soda. Go into your kitchen, put some soda in a glass and pour in some milk. Nothing happens.

Emphasis mine. He’s leaving something very much unsaid in that description, which is that when you go into your kitchen to get a glass of milk to test this theory, you will first open your refrigerator and take pasteurized milk out. That’s why there’s practically no useful acid in it. Modern refrigerated milk can be several days old and still fresh.⁵

Not only is our refrigeration better, but our milk started its commercial life pasteurized. Pasteurization doesn’t appear to have become universal, in the United States at least, until 1947 on a statewide basis, and 1973 on a national basis.

Glen’s recipe is a “Housekeeper’s Cake” in a 1914 cookbook. The first modern home refrigerator was the General Electric Monitor Top in 1927. In 1915, there would have been more acid in the milk. In my opinion they would very likely have used yesterday’s milk, or the day before’s depending on whether they had an icebox and what kind if they did. But regardless of how fresh the milk is, it was not taken from what we would consider to be a home kitchen refrigerator. It started its life unpasteurized, already on the inevitable road to sour.

What would happen if we combined the sparse instructions from the Battersea Polytechnic cookbook with the more detailed instructions from the May Byron cookbook?

My reasoning is this: cakes like this don’t normally take that long to cook. If you want a slow-baked cake, you need a slow oven. And though this may sound like madness, there is method in it: this is a Housekeeper’s Cake. It might very well be designed so that the housekeeper of a boarding house or other communal living environment could put the cake into the hot oven after the final dinner item was removed, and turn off all fuel sources. The cake will be ready an hour and a half to two hours later after the table has been set, the diners have sat down and eaten, people have talked, and it’s time for a simple dessert.

In other words, this cake might specifically be designed so that the baker doesn’t have to remain in the kitchen and keep an eye on the oven.

Raisins are often soaked ahead of time in the era, so often that I suspect the technique often goes without saying. Before refrigeration, drying fruit was one means of preserving it, and soaking dried fruit was how it was reconstituted. Eating raisins raw was, I suspect, a little like eating sugar from the jar. I like whiskey-soaked raisins. So I’m soaking the raisins in whiskey. A “housekeeper” might very well soak them in hot water or even cold water.

And since it’s a one-third recipe, I’m going with two-thirds of a cup of milk (one third of a pint), somewhat sour, minus about one egg’s worth of liquid. The amount of milk in the Battersea recipe isn’t specified, only “a little bit” and a texture of “rather stiff”. Glen has a great joke about this around 06:30 in the video.

How much is a little bit of milk? This is pre-metric. English cookbook. So obviously an imperial little bit.

I’m translating “rather stiff” as “not quite stiff”. My version turned out basically cookie dough consistency. A stiff dough would just be described as stiff, not rather stiff.⁶ So I’m going with “still very pliable”. Your mileage may vary.

For all of these interpretations, your mileage may vary.

The other changes are “what do I have in my kitchen” changes. The substitution of candied pineapple for candied orange peel, or raisins and dried cranberries for raisins and sultana raisins shouldn’t make much of a difference, other than making a cake I like better.

Aunt Battersea’s Leave It Be Cake

Servings: 16
Preparation Time: 2 hours, 30 minutes
Baking A Vintage Cake: Fire, Failure & Redemption (Glen & Friends Cooking)
Pot-Luck or The British Home Cookery Book (ebook, Internet Archive)

Ingredients

• 4 oz raisins
• 4 oz dried cranberries
• ½ cup bourbon
• 8 oz flour
• 1 tsp baking soda
• ½ tsp cinnamon
• 3 oz butter
• 4 oz sugar
• 1 oz candied pineapple
• 1 egg
• milk (to ⅔ cup)
• sugar for sprinkling

Steps

1. Soak the raisins and cranberries in bourbon for at least an hour and up to overnight.
2. Preheat oven to 425°.
3. Line an 8x8 cake pan with buttered parchment paper.
4. Sift the flour, baking soda, and cinnamon together.
5. Mix in the butter “until like breadcrumbs”.
6. Mix in the sugar.
7. Mix in the raisins, cranberries, and candied pineapple.
8. Beat egg in a measuring cup.
9. Add enough milk to make ⅔ cup liquid.
10. Mix egg and milk into flour.
11. Pat mixture into prepared cake pan.
12. Sprinkle sugar over top.
13. Bake in preheated oven until cake just starts to rise, about three to five minutes.
14. Turn oven off. Bake another hour and a half or so, removing from oven when needed.
15. Serve warm.

I put an oven thermometer at the front of the oven; it dropped to about 300° very quickly. An hour into the baking time it was still about 250°, and by the end it was 150°. It could very well have been left in the oven at that point until needed. If your oven doesn’t lose heat as quickly—say, if you’re using a 1915 Monarch Range made of heavy iron—you may want to turn it off immediately on putting the cake in rather than waiting a few minutes. I also suspect that the cake is at least a little forgiving regarding the initial temperature.

Now, if this is the correct interpretation—and it certainly made a wonderful spice cake—should it have been written differently in the cookbook? Yes. From our perspective it should have, and I find it hard to believe from those instructions that the Battersea Polytechnic meant to start with a hot oven and then turn it off while baking. But ovens then were so different from ovens now that I can’t be certain, especially in an instructional book that might assume there’s an instructor explaining things.⁷

A vintage cookbook about vintage cooking!

I should have reserved a few teaspoons for cookies!

Not bad looking after over an hour and a half in the oven…

Really, a beautiful and tasty spice cake.

Despite it’s apparent popularity I was unable to find an online scan of the Battersea Polytechnic cookbook.

Circa 1926, this Calumet Model Kitchen contains what we would call a very primitive oven. I don’t think it contains a refrigerator at all; there may be an icebox to the far right, but it’s probably just a cabinet. Not that there was much difference before GE introduced their Monitor Top.

Baking in 1915 was a process. Much of our modern advances have been about arresting these processes, about holding back the inevitable entropy of life. Just as oven terminology wasn’t a steady state but a process, food itself wasn’t just an ingredient but a process. Milk especially was a process, something any housekeeper of the time would have recognized. We arrest the process that is milk through pasteurization and refrigeration. In 1915 they didn’t have our time-stopping appliances, and pasteurization was decades into the future.

Modern bakers have completely forgotten what life was like before modern appliances, before ubiquitous and reliable power. Even if they were using an icebox, the milk had started its life with all of its bacteria intact. And whether using an early form of pasteurization or not, it hasn’t been kept cold enough to stop it from souring anyway. When I made the suggestion that the milk you “go into your kitchen” to get should not have been in a refrigerator, Glen replied that:

If the recipe writer wanted ‘sour milk’, then the recipe would/should have asked for sour milk. It didn’t.

In modern times, this is true.⁷ We have stopped time in the kitchen, and our day-old milk is still sweet. But there was no refrigerator in the target kitchen of Household Cookery Recipes. In that era, in the era before home refrigeration and pasteurization, milk was always in the process of souring. This is why recipes from the era often call for “sweet milk” in recipes where we would just call for “milk”.⁸

Absent modern refrigeration, the default is that milk is noticeably along the process of souring. Moderately soured milk doesn’t need to be specified, because before refrigeration, milk is always in the process of souring. It’s just a matter of how far along in that process it is.

While it’s possible that a “housekeeper” might have access to refrigeration in 1915, it would have been nothing like modern refrigerators. Even as late as 1926 a “refrigerator” was at best literally an icebox with a central compressor blowing cold air into it.⁹

Further, because milk is always in the process of souring, milk that we would consider “sour”, they would have considered merely “not sweet”. There would be no need to call it sour milk because it’s just milk. It will still have more acid than milk that’s spent the same amount of time under modern refrigeration.

I have a strong suspicion that people back then often made cakes and pies specifically to preserve milk that was on the edge of undrinkable.

One of the amazing things to me about early home refrigerators is how the manufacturers initially treated them as cooking appliances. You were expected to use them for cold cooking, setting the dial and leaving it just as you would an oven, but for cooking things like ice creams instead of things like cakes.

It is as if the concept of setting a dial and then leaving it was still an amazing new thing in the thirties, a marketing point that could be expected to attract customers.

We live a Jetson’s life compared to the world before modern refrigerators and self-regulating ovens. Everything for us is a pushbutton or a dial or a magic cabinet. We prepare the ingredients, put them into a magic box, push a few buttons, and let them be. We halt the natural processes of entropy and decay so that we no longer need fear the once-inevitable processes of life.¹⁰

We’ve had that ability for so long that we’ve forgotten what it means. We’ve forgotten that these processes were once so inevitable there was no need to mention them in a recipe. This is all because our ancestors did the work to create and maintain ubiquitous, reliable power. Without reliable power, we return to a world where milk was always souring, and babysitting ovens went without saying.