Contact Us

Use the form on the right to contact us. 


123 Street Avenue, City Town, 99999

(123) 555-6789


You can set your address, phone number, email and site description in the settings tab.
Link to read me page with more information.

The Great Marshmallow Experiment


Recipes, thoughts and rants from Mallow & Hop.

The Great Marshmallow Experiment

Karen Marshall

I sent in my Class B permit application on June 1st, 2015 - the date I had indicated on my seller's permit and business license that I'd be starting my business. On June 11th I got a call from a very kind city employee informing me that the county's fiscal year ends in June, so if I processed my application now I'd be hit with a bill for the next year as soon as July, essentially paying $254 for a 3 week permit. He asked if I wanted to delay the process until July, I said yes.

That, of course, leaves me with three extra weeks of time to stew on my labels, freak myself out, and generally over prepare. After hours of staring at google docs calculating the cost and nutritional information of each flavor of marshmallow and hours and hours of trying to create a Wholesale Linesheet... I needed to give myself a rest. And you know what I can do in my sleep? Make marshmallows.


Two years ago I had started taking the HarvardX course Science & Cooking: From Haute Cuisine to Soft Matter Science. I lasted a week before the math scared me away - lots of logarithms! Now I'm trying again. (The math is still making my brain hurt, but I'm determined this time.) The course involves completion of a final project, an experiment involving the food science concepts learned in the course. BUT WHY WAIT? 

 I might have been Harvard material 10 years ago, but that was before 10 years of the internet and caring far too much about TV shows turned my brain to fudge.

I might have been Harvard material 10 years ago, but that was before 10 years of the internet and caring far too much about TV shows turned my brain to fudge.

I've been, until now, just experimenting based around others' recipes. I'll substitute and I'll add and I'll subtract, but the results are a crapshoot: some batches never set, some batches sweat, some just taste bad or bland. I've always wondered - why does the ratio of sugar to syrup change when you add certain ingredients? Why are recipes involving fruit cooked to 250° and not just to 240°? What result does an extra tablespoon of gelatin have to the finished product?

I'm going to find out


To determine the effect of each element of the process of cooking marshmallows on the finished product, in taste, texture, and durability. There are 6 obvious variables for the base recipe:

  • Sugar
  • Syrup 
  • Water in the base
  • Water in the bloom
  • Gelatin
  • Heat


I am not the first to embark on this mission... in fact it is actually a stock "Science Fair" experiment offered over on But I have been unable to find via Google-Fu any substantial results from said experiments, and nothing in the depth I plan to go (this is only Group A of planned recipes). In fact, in doing some basic research on the science behind marshmallows, one of the first results I found was a question I had asked on the Food52 message boards 2 years ago with a few untested but not unhelpful answers. 

A bit of marshmallow history. Marsh Mallow is actually a plant, of the genus Althaea. Doesn't the "marsh" part of it make so much more sense knowing that?

Confectioners in early 19th century France pioneered the innovation of whipping up the marshmallow sap and sweetening it, to make a confection similar to modern marshmallow. The confection was made locally, however, by the owners of small sweet shops. They would extract the sap from the mallow plant's root, and whip it themselves. The candy was very popular, but its manufacture was labour-intensive. In the late 19th century, French manufacturers thought of using egg whites or gelatin, combined with modified corn starch, to create the chewy base. This avoided the labour-intensive extraction process, but it did require industrial methods to combine the gelatin and corn starch in the right way. - Wikipedia

Gelatin, as well as having the benefit listed above of not needing to be extracted from the marsh mallow plant (a rather absurd piece of writing considering the common source of gelatin, but I digress), is a protein that binds together the sugar molecules in a pattern that traps air, resulting in the light texture. Let me let a science blog repeat that, but more science-y:

Gelatin makes marshmallows chewy by forming a tangled 3-D network of polymer chains.  Once gelatin is dissolved in warm water (dubbed the “blooming stage”), it forms a dispersion, which results in a cross-linking of its helix-shaped chains.  The linkages in the gelatin protein network, called “junction zones” trap air in the marshmallow mixture and immobilize the water molecules in the network . The result? The famously spongy structure of marshmallows! - Science and Food UCLA

The trapped air is also what turns the translucent solution you get when you first add your boiling syrup to your bloomed gelatin into the opaque white pillows we recognize as marshmallows. The air traps the light and reflects it back at you!

As for the syrup- although I use my own homemade sugar syrup in Mallow & Hop marshmallows, I will be using corn syrup for this experiment to ensure consistency. Corn syrup is ostensibly added to prevent the crystallization of sugar molecules in the cooling (and whipping) process - I've never encountered this, but I'm excited for the possibility of doing so in the "lesser corn syrup" recipe.

One way to prevent the crystallization of sucrose in candy is to make sure that there are other types of sugar—usually, fructose and glucose—to get in the way. Large crystals of sucrose have a harder time forming when molecules of fructose and glucose are around. [...] A simple way to get other types of sugar into the mix is to “invert” the sucrose (the basic white sugar you know well) by adding an acid to the recipe. Acids such as lemon juice or cream of tartar cause sucrose to break up (or invert) into its two simpler components, fructose and glucose. Another way is to add a nonsucrose sugar, such as corn syrup, which is mainly glucose. Some lollipop recipes use as much as 50% corn syrup; this is to prevent sugar crystals from ruining the texture.- The Science of Cooking

This brings up an interesting possibility - would marshmallows made with fruit puree in the base not have to have an invert syrup at all? That's a test for another day.

And finally: heat. I'll go back to ScienceBuddies on this one:

"As the [sugar] solution boils, water evaporates, and the solution becomes even more concentrated with sugar, and the boiling point gets pushed even higher. For example, an 85 percent sugar solution has a boiling point about 28°F higher than the boiling point of pure water. The final concentration of sugar in the syrup determines the structure of the candy that forms when the syrup cools. Think about the difference between the texture of caramels and lollipops—the caramel is softer and chewier, while the lollipop is hard and cracks when it is bitten. The syrup used to make caramels is cooked until it reaches 240–250°F, at which point it has an 87 percent concentration of sugar in solution and enters the firm ball stage (see the Exploratorium link in the Bibliography for more information about this). The syrup used to make lollipops, though, is cooked until it reaches a 99 percent concentration of sugar in solution (at a temperature of 300–310°F), which makes drops of the syrup turn into hard, brittle, easy-to-break threads when placed in cold water. When making marshmallows, the syrup is cooked until it reaches 235–240°F, or 85 percent concentration of sugar in solution." -

So, what will happen? Will higher heat result in firmer marshmallows? Will less corn syrup make them accidentally crunchy? What affect does the water in the recipes have?

One last thing: there are two elements to any marshmallow recipe. I'll use them liberally in the upcoming sections.

  • The Base - this is the mix of water, sugar and syrup that is heated
  • The Bloom - this is the bloomed gelatin (water + gelatin) to which the base is added 


If this were nearer to the end of the Food Science class, I might be able to come up with one. As it is... ¯\_(ツ)_/¯ 

Here are a few guesses though

  • I think altering the amount of water in the base will have a negligable affect on the end product, and will only affect the cooking time (which I will not measure).
  • Bases cooked to a higher heat will result in a firmer marshmallow, cooked lower will be lighter (or fail to set).


I start with Eileen Talanian's Vanilla Marshmallows recipe. Eileen Talanian's book Marshmallows is the most consistent and delicious marshmallow book out there, and I've purchased more than a few. Then, recipes are minorly adjusted. As there are 6 initial variables, there are 6 factorial (720) recipes to try, and that is with only an On/Off function, not the sliding scale of 1 cup, 1.5 cups, 2 cups that a recipe deals with. The possibilities are quite literally endless, at least in my lifetime. But we'll get to batches that aren't recognizable as marshmallows eventually. Starting small, I am taking each variable and adjusting it +1/3 and -1/3, hoping that is a large enough percentage to make a noticeable difference, but small enough to result in an edible product.

Potential hitches in the "scientific" "process":

  • Recipes will be done by measurements and not weight. To quoth the Harvard edX class: "Careful cooking protocols use weights of ingredients instead of volumes. This is because, as we have emphasized in this week's lectures, what one really cares about is the number of molecules of each ingredient in the recipe. This is more accurately controlled with weight measurements." Sorry buddy - calculating the number of moles of protein in a litre of buttermilk used up all the math energy I had this weekend. A more careful experiment awaits in the future.

  • Some recipes may have results that are invisible in the finished product, like "takes longer to reach target temperature." This will not be documented as I am working with a gas stove and cannot rely on my ability to apply the same amount of heat to each batch.

  • The whipping process and determining when each batch is done and ready to set will be determined by my subjective sense of when a batch is done and ready to set. I have never trusted recipes on this front. The recipes say to whip for at least 10 minutes, my batches are always done at 5.

  • Many of the criteria I'm judging by are subjective. Some will melt in your mouth more than others, some feel denser.

  • I do not have a way to measure humidity. As the day wears on, water boiled off from the many marshmallow bases will increase humidity in the room. 

Some attempts to limit the variables include: spending lots on AC to keep my apartment a constant-ish temperature, calibrating thermometers before each day of cooking, allowing each batch to set for the same amount of time, using mass produced corn syrup to provide consistency.

The cooking starts.

 An immediate casualty. Rest in peace sweet candy thermometer, you served me well.

An immediate casualty. Rest in peace sweet candy thermometer, you served me well.

 So. Many. Batches.

So. Many. Batches.


Each batch will be judged on moisture level, "melt", density and durability. My boyfriend and I ate a lot of marshmallows over the course of this weekend. Our belly-aches are your gain. Ratings are listed on a scale of 1 to 10, with Mine / Eric's. For moistness - 1 is dry, 10 is almost liquid; you may also want to think of "moistness" as "stickiness". For "melt" - 1 is doesn't, 10 is very melty. For density - 1 is light, 10 is dense. This won't be an interesting read; I would skip if I were you. I promise to TL;DR.

 All that stuff visually. Pink are my ratings, blue are Eric's. Enjoy the gender normativeness.

All that stuff visually. Pink are my ratings, blue are Eric's. Enjoy the gender normativeness.

  • Batch A - The Control.
    • Moist Level: 2 / 3
    • Melt Level: 5 / 5
    • Density: 5 / 7
      • Batch A are a tad sticky, which betrays a slight humidity in my kitchen. 
  • Batch B - More Syrup (More Overall Sugar)
    • Moist Level: 7 / 7
    • Melt Level: 7 / 7
    • Density: 6 / 7
  • Batch C - Less Syrup (Less Overall Sugar)
    • Moist Level: 1 / 2
    • Melt Level: 2 / 4
    • Density: 8 / 6
  • Batch D - More Sugar (More Overall Sugar)
    • Moist Level: 1 / 3
    • Melt Level: 5 / 4
    • Density: 5 / 6
  • Batch E - Less Sugar (Less Overall Sugar)
    • Moist Level: 9 / 7
    • Melt Level: 4 / 4
    • Density: 4 / 7
  • Batch F - Higher Sugar : Syrup Ratio
    • Moist Level: 1 / 3
    • Melt Level: 6 / 3
    • Density: 6 / 3
  • Batch G - Lower Sugar : Syrup Ratio
    • Moist Level: 8 / 8
    • Melt Level: 9 / 7
    • Density: 7 / 7
  • Batch H - More Base Water
    • Moist Level: 4 / 7
    • Melt Level: 7 / 6
    • Density:  7 / 6
  • Batch I - Less Base Water
    • Moist Level: 2 / 4
    • Melt Level: 8 / 4
    • Density: 5 / 8
      • This has notes as "Karen's favorite." Odd how different our scores are.
  • Batch J - More Gelatin
    • Moist Level: 3 / 7
    • Melt Level: 5 / 3
    • Density: 4 / 6
  • Batch K - Less Gelatin
    • Moist Level: 9 / 9
    • Melt Level: 10 / 10
    • Density: 3 / 9
      • The biggest disparity in our ratings by far. Also, my notes here say "Yum."
  • Batch L - More Bloom Water
    • Moist Level: 4 / 7
    • Melt Level: 2 / 5
    • Density: 8 / 7
  • Batch M - Less Bloom Water
    • Moist Level: 3 / 3
    • Melt Level: 6 / 6
    • Density: 10 / 6
      • This batch whipped very quickly and and was very sticky and stringy when being transferred to the pan.
  • Batch N - Higher Heat
    • Moist Level: 3 / 5
    • Melt Level: 8 / 6
    • Density:  5 / 7
  • Batch O - Lower Heat:
    • Moist Level: 9 / 7
    • Melt Level: 6 / 7
    • Density: 3 / 7
      • This batch took forever to whip up.
  • Batch P - Control W/ Mold Inhibitor
    • Moist Level: 3 / 4
    • Melt Level: 5 / 4
    • Density: 6 / 8
      • These results should be the same as group A. They are not for two reasons - one, batch P was likely cooked at a slightly higher humidity than A, being at the end of the day. Two, human error. After 16 marshmallows you forget what you're rating.

So what does this mean?

Let's compare the recipes that have variants in the amounts of sugar.

 In descending order of Sugar:Syrup ratio.

In descending order of Sugar:Syrup ratio.

The higher the ratio of Sugar:Syrup, the more likely the resulting marshmallow will be drier (and that lower moisture content should result in a longer lasting mallow - to be updated when they start molding in a few months). A higher corn syrup also seems to result in a "meltier" marshmallow, which is a desirable trait. If you further adjust for the higher Sugar:Syrup ratios while separating the recipes by total Sugar content, you get this:

Screen Shot 2015-06-23 at 1.42.00 PM.png

So, high sugar : syrup ratio = drier, less melty marshmallows. Low sugar : syrup ratio = wet, sticky, but perfectly melt in your mouth marshmallows. It looks like from here, the Control is the happy medium.

I'm a little shocked by the differing numbers here. There are so few reasonable conclusions to be drawn here! More base water results in a moister marshmallow, which makes sense initially, until you refer back to the Science buddies, which says that the base should always reach 240 when there is 85-87% sugar concentration. As for the melt and density? Eric and I disagreed on both.

Screen Shot 2015-06-23 at 1.39.13 PM.png

The disparity in density in K is, I believe, caused by the fact that K is not really recognizable as a marshmallow - it's light for a marshmallow, dense for fluff, and dat melt score... More gelatin results in a lighter marshmallow. I would hypothesize that the results of bloom water recipes are only related to how effectively that amount of water can "bloom" the gelatin. The standard recommended amount is 1/4 cup of water per 1tbsp of gelatin. 

Screen Shot 2015-06-23 at 1.37.59 PM.png

Higher heat results in a slightly denser marshmallow, but that doesn't seem to affect its "melt" factor - in fact the high heat batch tied for 3rd for meltiness. I'd like to try to repeat this result in a future experiment.


Totally non conclusive findings, ie, my TL;DR

  • High sugar : syrup ratio = drier, less melty marshmallows. Low sugar : syrup ratio = wet, sticky, but perfectly melt in your mouth marshmallows. There is definitely a recipe or two of mine that I plan to adjust slightly based on this knowledge.
  • More "active" gelatin results in lighter marshmallows, less "active" gelatin results in denser marshmallows. I'm interested in adding more gelatin to some recipes to see where the effect caps out. Definitely make sure there is enough water in your base to bloom the gelatin, or the results are undesirable.
  • Cooking the base to under 240° may not result in marshmallows. Effect of additional heat inconclusive.
  • I personally enjoy marshmallows with less water in the base, it seems?
  • Marshmallows are very sugary and eating in excess of 8 at a time may give you a headache and cavities.
  • Humans are bad at comparisons.
  • (Which recipes result in longer lasting mallows - TBD!)

Here are a few experiments I'd like to take on in the future:

  • Repeating the experiment measuring by weight.
  • Playing with recipes where the activated gelatin is varied - so a bloom with 1/2 cup water & 2 tbsp gelatin, 3/4 cup & 3 tbsp, 1 cup & 4 tbsp, etc.
  • How does sugar in the bloom affect the finished product?
  • How does higher heat affect a recipe with more total sugar in the base?
  • How does additional gelatin affect a recipe with more total sugar in the base?
  • Does alcohol affect any of these factors?
  • Can fruit purees (fructose) replace a majority of the corn syrup while retaining "melt"?
  • Constructing the perfect marshmallow.

Now, if anyone has an idea for what to do with all these marshmallows...