Week Ten: Sherbet
(Note: I amended the final recipe instructions on 31/03/2024 to correct the amount of zest; you want 1.6%, or multiply the total weight by 0.016)
Sherbet is essentially a creamy sorbet; it has a legal definition in the US, although I don’t know if I’m reading it correctly because it looks as though the margin for butterfat is between 1% and 2% which is a) narrow and b) low (Dana Cree, writing in 2017, gave the US definition as between 2% and 4% which is better.) Not so in the UK, the closest thing we come to defining is a milk ice. Not one of the Gunter’s confectioners had a recipe for anything like a sherbet. Sherbet as a frozen dessert isn’t much known in the UK - we have another definition for sherbet which is a fizzy sweet powder, supposedly to be mixed as a drink but more often just eaten by itself - so I went to the other side of the pond to find historic sherbet recipes via Google Books advanced searches.
The difficult part of trying to find early sherbet recipes is the term has been used in many different ways. The word comes from sharbat which is a drink from Persia (modern day Iran) and I found a few of these recipes; I also found several recipes for an alcoholic punch drink. Then comes the use of sherbet to mean a sorbet. The earliest recipe I found that could be recognised as a sherbet as in a creamier sorbet is in Directions for Cookery published in 1840. It doesn’t use milk or cream though; it uses boiled veal bones, or essentially gelatine. It’s not the only one, there are a lot of 19th century sherbet recipes that use gelatine, although most use powdered gelatine that was patented in 1845 (as ‘portable gelatine’) rather than boiling a knuckle. From 1875 sherbet recipes using egg whites start appearing, some add tartaric acid too. I also found one recipe using arrowroot from 1855, but only the one so I discounted it. The earliest milk sherbet I found was from 1893, although I found some earlier ones that used milk/cream and egg whites, mostly from this 1879 book, so obviously there was some overlap, and gelatine and egg white recipes continued to be published after the milk ones started.
Milk sherbet is what we recognise as a sherbet now, but there were enough gelatine and egg white recipes that they were obviously common practice once. Bear in mind neither gelatine nor egg white will contribute much fat at all, their main role is adding protein which binds water and leads to a less icy and therefore more smooth and ‘creamy’ (although there’s no butterfat) texture.
I used Mary Johnson Lincoln's gelatine recipe, she was a well known educator and author, plus I liked that her recipe uses measures I can get; what I call a pint and what an American calls a pint are very different and, whilst it was easy for me to get US measuring cups, it’s really difficult to get a US measuring jug in the UK without shipping it from America. A standard egg recipe was much harder to find, they tended to appear in books that were compilations of housewives’ recipes. There was a wide variety in the recipes, as you’d expect; I ended up using a recipe from Housekeeping in Old Virginia mainly based on it using exact measures for every ingredient rather than ‘sugar to taste’, and in measures I could access (four cups to a quart, and the US and UK share a pound) although I halved it because half a gallon of water is rather a lot. I also decided to use four oranges based on a different recipe from the same person using twelve oranges and three lemons for a gallon of water - lemons are roughly replaced with one and a half as many oranges, so that’s about sixteen oranges for a gallon of water, I’m using quarter of a gallon. I used the 1893 milk sherbet recipe, increased to six oranges rather than four lemons (although again, I halved the entire recipe.) Cree has a blood (or blush) orange sherbet recipe in Hello My Name Is Ice Cream, so I didn’t need to do much research here; I did most of my early ice cream making - even before going to the old recipes, right back when I was first learning - using her recipes.
A few things to point out:
1. Gelatine is an animal product and, whilst I don’t personally have to boil animal bones to use it, that is how it’s obtained. So, not suitable for vegetarians or vegans.
2. The egg whites aren’t cooked at any point. Eggs are pasteurised in the UK so I don’t worry about eating raw egg, but if you don’t have access to pasteurised eggs or are more vulnerable then avoid that recipe.
3. A bit like how I use medium eggs because eggs were smaller back then, I used medium sized oranges rather than large ones. These are also about the same size as blood oranges, which meant I could use them as a like for like replacement for Cree’s recipe whilst I was testing out (normal oranges cost less than blush oranges, if I’m making sherbet after sherbet I won’t use the expensive oeanges.)
The egg white recipe was a complete absolute disaster. I won’t wholly blame the recipe because it’s a poor workman who blames the tools, but I think it can carry some of the blame here. There’s no instruction on how much to beat the egg whites, in this or any recipe that I found, so I was winging it - I didn’t want to end up trying to pour foam into an ice cream machine, but maybe I should have beaten for longer? It also at no point tells you to juice the fruit but it does refer to ‘lemonade’ which at least heavily implies they get juiced and mixed with water at some point. I probably misinterpreted at least parts of the recipe if not the entire thing, it’s one of those recipes that I think relies on ‘pre-known knowledge’ which goes along with my suspicion that the egg white recipes are grass roots, but certainly the way I did it was not good. It froze as clumps of ice, and it froze to my paddle to the point it wouldn’t turn anymore. I couldn’t lift the paddle out of the bowl, I was actually worried I broke my machine. I still put it in the freezer, it became the texture of slushy icy snow. It tasted ok as a mixture, but once frozen it had no taste. An all around failure.
In contrast, the gelatine recipe worked surprisingly well. The obvious catch is it forms a jelly. There isn’t enough gelatine in it to fully set the amount of water used, but it goes quite thick and ‘semi jelly’ and I think if I were making it again I wouldn’t chill it for as long as I normally do. There comes a point where a mixture is chilled as much as it will; I usually leave it beyond that because a) that point is often in the middle of the night, b) it can help flavours if it infuses for longer and c) most ice cream mixtures aren’t harmed by chilling for too long. But, with gelatine, I’d want to stop it at the just chilled point rather than let it set. As I poured it into the machine I wasn’t hopeful, and for the first twenty five minutes I thought I was right … and then, like magic, it all seemed to happen in about five minutes and I ended up with a nice end product. I was expecting it to be gummy but it wasn’t at all. It didn’t have the rich orange taste of HMNI but it did taste of orange.
But neither of those is sherbet, really. And so it comes down to 1893 milk sherbet and Hello My Name Is blood orange sherbet. And the amazing thing is, how close they are; 1893 is 2% fattier (7% vs 5%) and 2% less sugary (total 25.5% vs 27.5%; added sugar, ie not sugar naturally present, kept the same difference) than HMNI. Water is virtually identical. 1893 tasted a lot less orangey, in fact I could only really pick it out because I knew it was there (but it did have a taste, unlike the egg white sherbet.) It used getting on for half the amount of orange juice than HMNI so this isn’t surprising, but I think some of this will come down to how much steeping and infusing HMNI had compared to 1893. The ingredients I don’t think need to change at all; it doesn’t quite meet the ‘maximum allowed’ fat or water for a sherbet as given by Cree, but only by 1%. I’m happy with that.
Apart from having more orange juice in the recipe, HMNI does two things that 1893 doesn’t to increase the flavour - the zest of the oranges is steeped in the hot dairy mixture, and the juicy bits in the orange juice remain in the mixture as it’s chilling (Cree never actually says to strain it out before churning, but I like smooth textures so I did.) A lot of the taste of an orange is in the oils in the zest, using it would elevate the flavour. The final question was, to heat or not to heat? The 1893 recipe does include scalding (heating to just below boiling point) the cream, but then it’s chilled again before being put back into the mixture so I didn’t heat it on the first try - it didn’t seem to have any point. Heating dairy opens up the proteins in it and they can ‘grab on’ to flavour, so heating the cream with the zest in it could produce a more intense flavour. But it has downsides too, cooked dairy is a different taste (although the aim is to not reach the point of cooking it) and evaporation will result in a loss of water and therefore a change in final texture. I did it twice more, once heating the cream and once not, and you’re about to see the conclusion I came to.
Thank you for following me this far in my not-so-internal sherbet tinkering thoughts, and I hope you enjoy 1893 blush orange sherbet.
Ingredients
I made 1000g, as always use the percentages to make what you like.
505.9g whole (3.7% butterfat) milk (50.59%)
127.7g whipping (39.5% butterfat) cream (12.27%)
213.6g caster sugar (21.36%)
152.8g blush orange juice (15.28%)
(Also 1.6% of the total weight in grated zest, so in this case 16g; this gets strained out so doesn’t feature in the mixture calculations)
Method
1. Put the cream and grated zest in a pan, and quickly heat to about 80C or when bubbles just start.
2. Take the pan off the heat and leave to stand, covered, for 30 minutes.
3. Whilst it’s standing, combine the milk and orange juice in a mixing bowl.
4. Once the cream/zest has finished standing, add it to the milk/orange juice.
5. Slowly add the sugar whilst whisking (I use a stand mixer and balloon whisk.)
6. Leave the mixture to chill for at least four hours, and preferably overnight. (You can check if you need an ice bath to bring it to 20C before putting it in the fridge, but I doubt it will be needed.)
7. Rewhisk just prior to churning.
8. Put the mixture through a sieve to strain out the zest and bits of orange.
9. Churn according to your ice cream machine’s instructions, stopping at soft serve texture.
10. Put into a container and freeze for several hours, preferably overnight.
This was an adventure for me - I’ve never had sherbet before, nor have I made it before. I’ve made some good and some bad sherbets in the last week as I experimented, and ultimately I think this is a successful recipe. I have limited ability to quantify that because the only sherbets I’ve eaten are ones I’ve made and maybe I make really bad sherbet? But if the classification is, did I enjoy it, then it’s a successful sherbet. I’m looking forward to experimenting with different fruits and adding 1893 sherbet to my ice cream repertoire.
Comments
Post a Comment