Apple Identification Practice – Re-Examining a ‘Rosemary Russet’

There’s an apple tree growing in the walled garden at work – National Trust Quarry Bank – that has always – as far as I know, and as far as our garden records database is concerned – been labelled ‘Rosemary Russet’. This year is has produced dozens, if not hundreds of good-sized, fresh-tasting, crisp, juicy apples, which – I’m fairly certain – aren’t actually ‘Rosemary Russet’ after all.

I had my suspicions a while back, and so I’ve been keeping an eye on them, watching them develop over the past couple of months, and as they’ve got closer and closer to ripeness, the more convinced I am that they’re definitely not true-to-type.

As you can see from the photo above, and the next couple, the apples on this tree don’t really show much in the way of russeting – the brown, faintly rough-textured skin that’s distinctive of “russet” varieties – and are distinctly red-striped, particularly on the side of the fruit that’s facing the sun.

Here are a few apples that I picked from this tree earlier this week, so they’re three weeks riper than the ones pictured above. They haven’t developed much more in the way of russeting though – as you’d expect, because russeting tends to be a feature of the fruit from when it begins to develop, unlike reddening which is often a response to sunlight exposure – and they are still showing those distinct red stripes, the red hasn’t merged into a wider patch.

By way of comparison, the ‘Rosemary Russet’ reference photos in the National Fruit Collection database, show an apple that’s a bit more conical, has distinctly more russeting, isn’t as strongly striped, and has a stem / stalk that’s a noticeably longer than the mini-stalks on our apples.

Likewise, the bank of images on FruitID.com show apples that are perhaps similar in shape to ours, but they’ve nearly all ripened to a deeper yellow base colour, rather than pale green; their skins have much more russeting, with patches of red blush rather than striping; more pronounced lenticels (the grey-brown dots on the skin); and again, they’ve got much longer stalks.

I’m also growing two ‘Rosemary Russet’ cordon apples on my allotment plot. One of them was grafted a few years ago with a scion from the tree at work; unsurprisingly the one apple that it managed to produce this year is identical to the ones above. The other cordon tree was grafted with a scion from a parent tree that’s growing down in the Cotswolds¹If I remember it right, I think the scion was provided by my Twitter-buddy Steve., and the apples on this second cordon are, once again, noticeably different to the examples from the walled garden tree at Quarry Bank:

As you can see, these apples are much more heavily russeted, with red patches rather than stripes, and you can just make out the fairly long stalk on the apple at the bottom left, and the eye of the two larger apples is much shallower than those on the apples from Quarry Bank.

Of course, it’s always important to bear in mind that the same variety of apple growing in different locations – therefore on different soils containing different quantities and balances of nutrients, and exposed to different weather conditions – will often vary, usually in size and degree of ripening. And indeed, you can see some distinct variance in the the three apples that I’m seeking to identify.

Still, you’d expect the key characteristics of a variety or cultivar²A cultivar is a cultivated variety; one that has been deliberately bred by cross-pollination and subsequent selection, rather than a chance seedling that has been discovered and then retained for onward propagation. to be retained to a reasonable degree. Otherwise, if it’s simply not true-to-type, then it has to be considered to be a different variety. All-in-all, I think there’s enough visual evidence to conclude that Quarry Bank’s “Rosemary Russet” tree is not true-to-type, and therefore must actually be… something else.

The question then is: what else could it be? My first thought was that perhaps two trees had been mixed up when they were planted, and the labels transposed. But the rest of the trees in the main part of the walled garden orchard do seem to be true-to-type, particularly the ‘Lucombe’s Pine’ next door to the tree in question, which would have been the most likely candidate for an accidental switch when they were both planted, back in 2015.

Instead, I think it’s much more likely that there was a mix-up further up the supply chain, which could have happened at any stage in the process of sourcing scions, grafting, planting out, growing on, lifting the tree for sale, labelling it for delivery… there are too many possibilities to narrow it down.³It doesn’t really matter when or how the mix-up occurred. These things happen.

But that does leave us with a mystery tree to try to identify. And that’s where the fun starts.

A couple of years ago I wrote a guide to identifying apples and pear varieties in which I described a step-by-step method – based on and compiled from numerous pomological sources – for working towards a positive identification of an unknown apple. The method begins with building a profile of the fruit’s key characteristics. So let’s start there, and see where it takes us.

Step 1: Build a Profile

There are certain physical attributes or features that all apples have in common – size, general shape, colouration, skin markings, length of stalk, depth of eye or cavity, and so on – and it’s the differences in those features that marks one variety our from another. By closely examining the apple and describing those features, I can begin to build an identification profile. However, it’s important to build that profile based on a consensus observation of a few examples of a particular variety – three or more apples at a time – rather than a single specimen, because, as we’ve already seen, three apples picked from the same tree at the same time can vary quite widely from each other, as per the images below:

Drawing on the elements that are common to all three specimen apples – it really helps that I can hold the apples in my hand, as well as see them in photographs, so I can examine them from all angles – I can build the following profile:

General Size and Shape – Approx. 7cm in diameter, 5-6cm in height. Shape is flat-round, rather than conical. Varying degrees of ridging, some almost perfectly round, others more pentagonal.
Cheek – The skin of the apple. In this case it’s pale green, shading to yellow. Reddened on the side nearest the sun, but not deeply so, more an orange-red colour. Distinct, almost blotchy striping. Pale lenticels. Smooth-skinned, with only light russeting. No sign of a longitudinal hairline.
Stalk – The stem that attached the fruit to the tree. In this case very short indeed, just 2-3mm in length, embedded deep in the cavity.
Cavity – The area around the stalk. Quite deep, lightly russeted.
Base – The area around the cavity. Lightly ridged. Russeting doesn’t extend all that far outwards from the cavity.
Sepals – The remains of the apple blossom. Two examples are fully closed, one is open. The sepals aren’t particularly large.
Eye – The area around the sepals. Again, quite deep and wide, and lightly russeted.
Apex (Crown) – The area surrounding the eye. Noticeable but not particularly distinct crowning (bumps). Light russeting on one example, hardly any on the other two.

Step 2: Match the Criteria

Armed with the information above, it’s time to hit the FruitID.com identification database, plug in a few parameters, and see what sort of results are returned.

Focusing on the characteristics that are both the most distinct and the most consistent across all three apples, I set the following parameters from among the dozens that are available to specify within FruitID’s search function:

Appearance – Bultitude Group : Group 4 (striped, sometimes flushed) or Group 7 (some russeting, flushed and striped)
Fruit Diameter – medium, 50mm to 80mm
Fruit Shape – Profile : flat-round
Usable When Picked: yes (as opposed to: needs to be matured off the tree)
Skin – Ground Colour : yellow green
Skin – Hue of Over Colour : orange red
Skin – Stripes : may have strong stripes
Cavity – Stalk Length : short and stubby

As I added each criteria the number of possible matches in the database reduced, until with those nine active criteria, only five likely matches were offered.

And they were (click for a larger image):

1. ‘Ashmead’s Kernel‘
2. ‘Curl Tail‘
3. ‘Gravenstein‘
4. ‘Mitchelson’s Seedling‘
5. ‘Rival‘

Step 3: Draw Conclusions

‘Ashmead’s Kernel‘ I can immediately discount, mainly because we also grow that variety at Quarry Bank, and the apples on our tree, which are definitely true-to-type, are quite distinct from the ones I’m trying to identify here.

‘Curl Tail‘ and ‘Mitchelson’s Seedling‘ I can likewise rule out, mainly on the grounds that the FruitID reference photos show apples with much shallower cavities and eyes than ours. Both of them also have much more pronounced lenticels than our apples.

‘Rival‘ is perhaps closer, but again, the base colour of the reference photos is a much deeper yellow, the cavity and eye is shallower, and the red colouration is more flushed, less striped.

That leaves ‘Gravenstein‘ … and you know, I do believe we might be on to something here. The image below is a screenshot of the reference photos from FruitID.com:

Note the pronounced red striping and patchy orange-red flush over a light yellow-green base colour; the varying degrees of irregularity of the fruit’s general shape; the depth and light russeting of the cavity and eye; and of course those very short, stubby stalks. I think that all looks rather familiar to me.

By way of a second opinion, here’s the ‘Gravenstein’ reference photo from the National Fruit Collection Database:

And here’s a quick visual reminder of the apples that I picked from the Quarry Bank kitchen garden tree:

That’s a pretty close match, wouldn’t you say? I can’t be a hundred per cent certain, at least not without a DNA test against the examples in the National Fruit Collection, and I really would like a second opinion or two if I can get them. But for now, I think I can safely say that ‘Gravenstein’ is a much likelier match for our apples than ‘Rosemary Russet’ was.

If it is a ‘Gravenstein’, then that’s quite an interesting apple to have in any orchard; a true heritage variety that some sources suggest might date to the seventeenth century in Denmark, or Germany, or Italy, which was apparently introduced to the U.K. c. 1820⁴I suspect that means it was added to the National Fruit Collection when it was first established by the Horticultural Society of London at their Chiswick Garden… I can feel another research project coming on!. And of course, we now need to source a ‘Rosemary Russet’ to fill the gap in our collection. Luckily, I know of a true-to-type tree that might produce a suitable scion or two this coming winter…

What about you? Do you think ‘Gravenstein’ seems like a good match for these apples, or can you suggest a stronger candidate? Have you tried to identify any mystery apples of your own? If so, have you been able to get any good matches via FruitID.com? Can you recommend any other good, reliable information sources? Do please let me know, via the comments.

Footnotes

• 1
  If I remember it right, I think the scion was provided by my Twitter-buddy Steve. ↩︎
• 2
  A cultivar is a cultivated variety; one that has been deliberately bred by cross-pollination and subsequent selection, rather than a chance seedling that has been discovered and then retained for onward propagation. ↩︎
• 3
  It doesn’t really matter when or how the mix-up occurred. These things happen. ↩︎
• 4
  I suspect that means it was added to the National Fruit Collection when it was first established by the Horticultural Society of London at their Chiswick Garden… I can feel another research project coming on! ↩︎